British Journal of Dermatology

December 2007 - Vol. 157 Issue 6 Page I-1335

Contents pages I–XI

Snippets

Research Snippets pages xxi–xxi

Topical review

Paraneoplastic hypertrichosis lanuginosa acquisita: uncommon or overlooked?

P.H.T.J. Slee, R.I.F. van der Waal, J.H. Schagen van Leeuwen, R.A. Tupker, R. Timmer, C.A. Seldenrijk and M.A.M. van Steensel pages 1087–1092

Review articles

The relationships between exposure dose and response in induction and elicitation of contact hypersensitivity in humans

P.S. Friedmann pages 1093–1102

Psoriasis: evolution of pathogenic concepts and new therapies through phases of translational research

E. Guttman-Yassky and J.G. Krueger pages 1103–1115

Guidelines

Guidelines for evaluation and management of urticaria in adults and children

C.E.H. Grattan and F. Humphreys on behalf of the British Association of Dermatologists Therapy Guidelines and Audit Subcommittee pages 1116–1123 Original articles

Cutaneous biology

Cathelicidin LL-37 induces the generation of reactive oxygen species and release of human α- defensins from neutrophils

Y. Zheng, F. Niyonsaba, H. Ushio, I. Nagaoka, S. Ikeda, K. Okumura and H. Ogawa pages 1124–1131

Microarray analysis of aberrant gene expression in actinic keratosis: effect of the Toll-like receptor-7 agonist imiquimod

A. Torres, L. Storey, M. Anders, R.L. Miller, B.J. Bulbulian, J. Jin, S. Raghavan, J. Lee, H.B. Slade and W. Birmachu pages 1132–1147

Biphasic expression of stromal cell-derived factor-1 during human wound healing

A. Toksoy, V. Müller, R. Gillitzer and M. Goebeler pages 1148–1154

Original articles

Clinical and laboratory investigations

CTACK /CCL27 expression in psoriatic skin and its modification after administration of etanercept

A. Campanati, G. Goteri, O. Simonetti, G. Ganzetti, K. Giuliodori, D. Stramazzotti, D. Morichetti, M.L. Bernardini, B. Mannello, G. Fabris and A. Offidani pages 1155–1160

Prevalence of Staphylococcus aureus toxins and nasal carriage in furuncles and impetigo

F. Durupt, L. Mayor, M. Bes, M.-E. Reverdy, F. Vandenesch, L. Thomas and J. Etienne pages 1161–1167

Increased nuclear β-catenin in suprabasal involved psoriatic epidermis

P.J. Hampton, O.K. Ross and N.J. Reynolds pages 1168–1177

Cutaneous Malassezia flora in atopic dermatitis differs between adults and children

Y. Takahata, T. Sugita, H. Kato, A. Nishikawa, M. Hiruma and M. Muto pages 1178–1182

Reduction of immunosuppression for transplant-associated skin cancer: thresholds and risks

C.C. Otley, M.D. Griffin, M.R. Charlton, B.S. Edwards, M. Neuburg and T. Stasko for the Reduction of Immunosuppression Task Force of the International Transplant Skin Cancer Collaborative pages 1183–1188

Morphoea: a manifestation of infection with Borrelia species?

K. Eisendle, T. Grabner and B. Zelger pages 1189–1198

Original articles

Contact dermatitis and allergy

Filaggrin null alleles are not associated with hand eczema or contact allergy

A. Lerbaek, H. Bisgaard, T. Agner, K. Ohm Kyvik, C.N.A. Palmer and T. Menné pages 1199–1204

Original articles

Dermatopathology

The applicability and prognostic value of the new TNM classification system for primary cutaneous lymphomas other than mycosis fungoides and Sézary syndrome: results on a large cohort of primary cutaneous B-cell lymphomas and comparison with the system used by the Dutch Cutaneous Lymphoma Group

N.J. Senff and R. Willemze pages 1205–1211

Involvement of E-cadherin, β-catenin, Cdc42 and CXCR4 in the progression and prognosis of cutaneous melanoma

M.G. Tucci, G. Lucarini, D. Brancorsini, A. Zizzi, A. Pugnaloni, A. Giacchetti, G. Ricotti and G. Biagini pages 1212–1216

Original articles

Epidemiology and health services research

Environmental factors, parental atopy and atopic eczema in primary-school children: a cross- sectional study in Taiwan

Y-L. Lee, C-W. Li, F-C. Sung, H-S. Yu, H-M. Sheu and Y.L. Guo pages 1217–1224

Original articles

Paediatric dermatology

Novel EBP gene mutations in Conradi–Hünermann–Happle syndrome

P.M. Steijlen, M. van Geel, M. Vreeburg, D. Marcus-Soekarman, L.J.M. Spaapen, F.C.M. Castelijns, M. Willemsen and M.A.M. van Steensel pages 1225–1229

Original articles

Photobiology

Melanocortin 1 receptor (MC1R) genotype influences erythemal sensitivity to psoralen– ultraviolet A photochemotherapy

G. Smith, M.J.V. Wilkie, Y.Y. Deeni, P.M. Farr, J. Ferguson, C.R. Wolf and S.H. Ibbotson pages 1230–1234

Original articles

Therapeutics

Treatment of pyoderma gangrenosum with intravenous immunoglobulin

D.L. Cummins, G.J. Anhalt, T. Monahan and J.H. Meyerle pages 1235–1239

Isotretinoin therapy and the incidence of acne relapse: a nested case–control study

L. Azoulay, D. Oraichi and A. Bérard pages 1240–1248

Original articles

Concise communication

Psoriasis patients show signs of insulin resistance

S. Boehncke, D. Thaci, H. Beschmann, R.J. Ludwig, H. Ackermann, K. Badenhoop and W-H. Boehncke pages 1249–1251

Case reports

Unusual molecular findings in Kindler syndrome

K. Arita, V. Wessagowit, A.C. Inamadar, A. Palit, H. Fassihi, J.E. Lai-Cheong, C. Pourreyron, A.P. South and J.A. McGrath pages 1252–1256 A sporadic case of early-onset sarcoidosis resembling Blau syndrome due to the recurrent R334W missense mutation on the NOD2 gene

P. Coto-Segura, S. Mallo-Garcia, M. Costa-Romero, J.I. Arostegui, J. Yague, E. Ramos-Polo and J. Santos-Juanes pages 1257–1259

Gene corner

COL7A1 mutational analysis in Korean patients with dystrophic epidermolysis bullosa

S-W. Oh, J.S. Lee, M.Y. Kim and S-C. Kim pages 1260–1264

A novel homozygous mutation of the EVER1/TMC6 gene in a Japanese patient with epidermodysplasia verruciformis

S. Aochi, G. Nakanishi, N. Suzuki, N. Setsu, D. Suzuki, K. Aya and K. Iwatsuki pages 1265–1266

Correspondence

A case of primary anetoderma in an infant

H-J. Yu, H. Shin, M-S. Kang and J-S. Kim pages 1267–1269

Successful treatment of severe psoriatic natal cleft fissuring with tissue adhesive

V.H. Smith and S.K. Jones pages 1269–1270

Recovery from Sézary syndrome following Mycobacterium avium spondylitis

A. Yamada, O. Yamasaki, K. Asagoe, K. Tsuji, T. Hamada, Y. Ota and K. Iwatsuki pages 1270–1271

Late-onset neutropenia following rituximab treatment in patients with autoimmune diseases

R. Rios-Fernández, M.T. Gutierrez-Salmerón, J.-L. Callejas-Rubio, M. Fernández- Pugnaire and N. Ortego-Centeno pages 1271–1273

Hypertriglyceridaemia during treatment with adalimumab in psoriatic arthritis

G. Stinco, F. Piccirillo and P. Patrone pages 1273–1274

Adalimumab for treatment of pyoderma gangrenosum

R.G. Pomerantz, M.E. Husni, E. Mody and A.A. Qureshi pages 1274–1275 Effects of etanercept therapy on fatigue and symptoms of depression in subjects treated for moderate to severe plaque psoriasis for up to 96 weeks

R. Krishnan, D. Cella, C. Leonardi, K. Papp, A.B. Gottlieb, M. Dunn, C.F. Chiou, V. Patel and A. Jahreis pages 1275–1277

Cutaneous Langerhans cell histiocytosis in an elderly man successfully treated with narrowband ultraviolet B

S. Imafuku, S. Shibata, A. Tashiro and M. Furue pages 1277–1279

Pyoderma gangrenosum and interleukin 8

M. Oka pages 1279–1281

Recurrent KIND1 (C20orf42) gene mutation, c.676insC, in a Brazilian pedigree with Kindler syndrome

B.C.F. Martignago, J.E. Lai-Cheong, L. Liu, J.A. Mc Grath and T.F. Cestari pages 1281–1284

Transient macular erythema with eosinophilia in a patient carrying the FIP1L1-PDGFRA fusion gene

H. Yahara, T. Satoh, T. Hashimoto and H. Yokozeki pages 1284–1287

A case of milia en plaque successfully treated with oral etretinate

N. Ishiura, M. Komine, T. Kadono, K. Kikuchi and K. Tamaki pages 1287–1289

Limited cutaneous systemic sclerosis associated with discoid lupus erythematosus in two Japanese patients with anticentromere antibodies

T. Kawakami, K. Kawasaki and Y. Soma pages 1289–1291

Coexistence of primary cutaneous anaplastic large cell lymphoma and mycosis fungoides in a patient with B-cell chronic lymphocytic leukaemia

M. Marschalkó, J. Csomor, N. Erős, Á. Szigeti, J. Hársing, J. Szakonyi, M. Désaknai, A. Matolcsy, J. Demeter and S. Kárpáti pages 1291–1293

Usefulness of the QuantiFERON® test in the confirmation of latent tuberculosis in association with erythema induratum

J. Angus, C. Roberts, K. Kulkarni, I. Leach and R. Murphy pages 1293–1294 Respiratory involvement in toxic epidermal necrolysis portends a poor prognosis that may not be reflected in SCORTEN

J.S. Hague, J.M.R. Goulding, T.M.W. Long and B.C. Gee pages 1294–1296

Hydroxyzine-induced acute generalized exanthematous pustulosis

Y-S. Tsai, M-E. Tu, Y-H. Wu and Y-C. Lin pages 1296–1297

Malignant melanoma in a woman with LEOPARD syndrome: identification of a germline PTPN11 mutation and a somatic BRAF mutation

M. Seishima, Y. Mizutani, Y. Shibuya, C. Arakawa, R. Yoshida and T. Ogata pages 1297–1299

Anonychia, hyponychia and spontaneous amputation of the distal phalanges as a consequence of ischaemic necrosis of the extremities after umbilical catheterization

I. Neri, F. Savoia, F. Giacomini and A. Patrizi pages 1299–1301

Basal cell carcinomas of the inner canthus: incidence of incomplete excision according to topographical localization of tumours

F. Boriani, F. Marconi pages 1301–1302

News and Notices

News and Notices pages 1303–1303

Corrigenda

Corrigenda pages 1304–1304

Corrigenda pages 1304–1304

Erratum

Erratum pages 1304–1304 Original article printed in: British Journal of Dermatology 157:4 p. 788 doi: 10.1111/j.1365-2133.2007.08094.x Author Index

Author index pages 1305–1323

Subject Index

Subject index pages 1324–1335

92 Lipomas after blunt soft tissue trauma: are they real? Contents Analysis of 31 cases M.C.AUST, M.SPIES, S.KALL, A.GOHRITZ, P.BOORBOOR, P.KOLOKYTHAS AND Volume 157, Number 1, July P.M.VOGT Epidemiology and health services research Snippets 100 Impetigo in epidemic and nonepidemic phases: an Research Snippets incidence study over 4½ years in a general population Clinical Snippets S.RØRTVEIT AND G.RORTVEIT 106 Hairdressing is associated with scalp disease in African Editorial schoolchildren 1 The alcohol hand rub: a good soap substitute? N.P.KHUMALO, S.JESSOP, F.GUMEDZE AND R.EHRLICH G.A.JOHNSTON AND J.S.C.ENGLISH Photobiology Review articles 111 Variable pulsed light is less painful than light- 4 The evolution of the psoriatic lesion emitting diodes for topical photodynamic therapy of P.C.M.VAN DE KERKHOF actinic keratosis: a prospective randomized controlled 16 Skin manifestations of intravascular lymphoma mimic trial inflammatory diseases of the skin P.BABILAS, R.KNOBLER, S.HUMMEL, C.GOTTSCHALLER, T.MAISCH, M.KOLLER, ¨ ¨ J.ROGLIN AND A.BOER M.LANDTHALER AND R-M.SZEIMIES Original articles Therapeutics Cutaneous biology 118 Use of oral glycopyrronium bromide in hyperhidrosis 26 DNA repair capacities of cutaneous fibroblasts: effect of V.BAJAJ AND J.A.A.LANGTRY sun exposure, age and smoking on response to an 122 Oral R115866 in the treatment of moderate acute oxidative stress to severe facial acne vulgaris: an exploratory S.SAUVAIGO, M.BONNET-DUQUENNOY, F.ODIN, F.HAZANE-PUCH, study N.LACHMANN, F.BONTE´, R.KURFU¨ RST AND A.FAVIER C.J.VERFAILLE, M.COEL, I.H.BOERSMA, J.MERTENS, M.BORGERS AND 33 A cytotoxic analysis of antiseptic medication on skin D.ROSEEUW substitutes and autograft 127 Mycophenolate mofetil for severe childhood atopic Q.LE DUC, M.BREETVELD, E.MIDDELKOOP, R.J.SCHEPER, M.M.W.ULRICH AND dermatitis: experience in 14 patients S.GIBBS M.HELLER, H.T.SHIN, S.J.ORLOW AND J.V.SCHAFFER Clinical and laboratory investigations 133 Vehicle-controlled, randomized, double-blind study 41 Differences in survivin location and Bcl-2 expression in to assess safety and efficacy of imiquimod 5% cream CD30+ lymphoproliferative disorders of the skin applied once daily 3 days per week in one or two compared with systemic anaplastic large cell courses of treatment of actinic keratoses on the lymphomas: an immunohistochemical study head G.GOTERI, O.SIMONETTI, S.RUPOLI, G.PICCININI, C.RUBINI, D.STRAMAZZOTTI, A.ALOMAR, J.BICHEL AND S.MCRAE F.FAZIOLI, C.CAPOMAGI, P.LEONI, A.M.OFFIDANI AND L.LOMUZIO 142 Corticosteroid-induced clinical adverse events: 49 Association of functional gene variants in the frequency, risk factors and patient’s opinion regulatory regions of COX-2 gene (PTGS2) with L.FARDET, A.FLAHAULT, A.KETTANEH, K.P.TIEV, T.GE´NE´REAU, C.TOLE´DANO, nonmelanoma skin cancer after organ transplantation C.LEBBE´ AND J.CABANE M.GOMEZ LIRA, S.MAZZOLA, G.TESSARI, G.MALERBA, M.ORTOMBINA, L.NALDI, 149 A multicentre, randomized, controlled study of the G.REMUZZI, L.BOSCHIERO, A.FORNI, C.RUGIU, S.PIASERICO, G.GIROLOMONI efficacy, safety and cost-effectiveness of a combination AND A.TURCO therapy with amorolfine nail lacquer and oral 58 Sentinel lymph node biopsy in melanoma: a terbinafine compared with oral terbinafine alone for micromorphometric study relating to prognosis and the treatment of onychomycosis with matrix completion lymph node dissection involvement ´ S.DEBARBIEUX, G.DURU, S.DALLE, O.BEATRIX, B.BALME AND L.THOMAS R.BARAN, B.SIGURGEIRSSON, D.DE BERKER, R.KAUFMANN, M.LECHA, 68 Prevalence of metabolic syndrome in patients with J.FAERGEMANN, N.KERROUCHE AND F.SIDOU psoriasis: a hospital-based case–control study P.GISONDI, G.TESSARI, A.CONTI, S.PIASERICO, S.SCHIANCHI, A.PESERICO, Concise communications A.GIANNETTI AND G.GIROLOMONI 158 Endothelial cells in infantile haemangiomas originate from the child and not from the mother Contact dermatitis and allergy (a fluorescence in situ hybridization-based 74 How irritant is alcohol? study) H.LO¨ FFLER, G.KAMPF, D.SCHMERMUND AND H.I.MAIBACH ´ 82 Artificial reduction in transepidermal water loss S.REGNIER, N.DUPIN, C.LE DANFF, M.WASSEF, O.ENJOLRAS AND S.ARACTINGI 161 Transforming growth factor-b receptor II is improves skin barrier function I.BURACZEWSKA, U.BROSTRO¨ M AND M.LODE´N preferentially expressed in the companion layer of the human anagen hair follicle Dermatological surgery and lasers H.M.SOWDEN, R.O.S.KAROO AND D.J.TOBIN 87 Randomized, double-blind, prospective study to 165 Associations of promoter region polymorphisms in the compare topical 5-aminolaevulinic acid methylester tumour necrosis factor-a gene and early-onset psoriasis with topical 5-aminolaevulinic acid photodynamic vulgaris in a northern Polish population therapy for extensive scalp actinic keratosis B.NEDOSZYTKO, A.SZCZERKOWSKA-DOBOSZ, M.ZABŁOTNA, J.GLEN´ , K.RBAŁA F.J.MOLONEY AND P.COLLINS AND J.ROSZKIEWICZ II Contents

Case reports 207 A novel deletion mutation in the EDAR gene in a 168 Acquired palmoplantar keratoderma and immuno- Pakistani family with autosomal recessive hypohidrotic bullous disease associated with antibodies to desmocollin 3 ectodermal dysplasia M.C.BOLLING, J.R.MEKKES, W.F.M.GOLDSCHMIDT, C.J.M.VAN NOESEL, M.TARIQ, N.WASIF AND W.AHMAD M.F.JONKMAN AND H.H.PAS 209 Mondor’s phlebitis after using tadalafil 174 Bartonella-related pseudomembranous angiomatous C.GUARNERI AND F.GUARNERI papillomatosis of the oral cavity associated with 210 Cutaneous angiokeratoma and venous malformations in allogeneic bone marrow transplantation and oral a Hispanic-American patient with cerebral cavernous graft-versus-host disease malformations C.VASSALLO, M.ARDIGO` , V.BRAZZELLI, M.ZECCA, F.LOCATELLI, B.J.ZLOTOFF, R.H.BANG, R.S.PADILLA AND L.MORRISON P.E.ALESSANDRINO, M.LAZZARINO, S.CORONA, P.LANZERINI, M.BENAZZO, 213 Folate with methotrexate: big benefit, questionable M.FABBI AND G.BORRONI cost Gene corner I.BROWNELL AND B.E.STROBER 179 Novel COL7A1 mutations in a Japanese family with 213 News and Notices transient bullous dermolysis of the newborn associated with pseudosyndactyly H.NAKANO, Y.TOYOMAKI, S.OHASHI, A.NAKANO, H.JIN, T.MUNAKATA, N.AKITA, K.TAMAI AND Y.MITSUHASHI Correspondence Volume 157, Number 2, August 183 Cutaneous lesions in neurofibromatosis 1: confused terminology Snippets S.BARBAROT, C.NICOL, C.VOLTEAU, D.LE FORESTIER, J.M.N’GUYEN, E.MANSAT, P.WOLKENSTEIN AND J. F.STALDER Research Snippets 184 Lentigo maligna involving the tumour nests and Clinical Snippets stroma of a nodular basal cell carcinoma Editorial S.M.TAIBJEE, B.C.GEE, D.S.A.SANDERS, A.SMITH AND R.A.CARR 215 Sunbeds, beauty and melanoma 188 Eruptive vellus hair cysts presenting as bluish-grey B.DIFFEY facial discoloration masquering as naevus of Ota K.H.N.CHAN, W.Y.M.TANG, W.Y.LAM AND K.K.LO Review articles 189 Angiomyofibroblastoma of the vulva with a penile 217 Proposal of a new classification system for melanocytic appearance naevi T.WATANABE, Y.YOSHIDA AND O.YAMAMOTO G.ARGENZIANO, I.ZALAUDEK, G.FERRARA, R.HOFMANN-WELLENHOF AND 191 Successful treatment of severe psoriatic arthritis with H.P.SOYER infliximab in an 11-year-old child suffering from 228 The oncogenic potential of human papillomaviruses: a linear psoriasis along lines of Blaschko review on the role of host genetics and environmental S.ROTT, R.M.KU¨ STER AND U.MROWIETZ cofactors 192 Two siblings with neonatal pemphigus vulgaris V.K.MADKAN, R.H.COOK-NORRIS, M.C.STEADMAN, A.ARORA AND S.K.TYRING associated with mild maternal disease Original articles T.UGAJIN, H.YAHARA, Y.MORIYAMA, T.SATO, K.NISHIOKA AND H.YOKOZEKI Cutaneous biology 194 An infantile case of pityriasis lichenoides et 242 Mechanisms of melanogenesis inhibition by varioliformis acuta 2,5-dimethyl-4-hydroxy-3(2H)-furanone D.HOSHINA, M.AKIYAMA, K.HAMASAKA AND H.SHIMIZU J.LEE, E.JUNG, J.LEE, S.HUH, Y.C.BOO, C.G.HYUN, Y-S.KIM AND D.PARK 196 Gomez–Lopez–Hernandez syndrome: another 249 Antitumour necrosis factor-a chimeric antibody consideration in focal congenital alopecia (infliximab) inhibits activation of skin-homing CD4+ D.J.PURVIS, A.RAMIREZ, N.ROBERTS AND J.I.HARPER and CD8+ T lymphocytes and impairs dendritic cell 198 A case of phosphaturic mesenchymal tumour (mixed function connective tissue variant) that developed in the C.BEDINI, F.NASORRI, G.GIROLOMONI, O.DE PITA` AND A.CAVANI subcutaneous tissue of a patient with oncogenic 259 Adhesion of peripheral blood mononuclear cells and osteomalacia and produced fibroblast growth factor 23 CD4+ T cells from patients with psoriasis to cultured M.OKA, T.KAMO, E.SASAKI, H.KAJI, H.NISHIZAWA, Y.IMANISHI AND C.NISHIGORI endothelial cells via the interaction between 200 Unilateral periorbital oedema due to sarcoid infiltration lymphocyte function-associated antigen type 1 and of the eyelid: an unusual presentation of sarcoidosis intercellular adhesion molecule 1 with facial nerve palsy and parotid gland enlargement D.WATABE, H.KANNO, A.YOSHIDA, A.KUROSE, T.AKASAKA AND T.SAWAI M.YAOSAKA, R.ABE, H.UJIIE, Y.ABE AND H.SHIMIZU 202 Pityriasis rubra pilaris in a mother and two daughters Clinical and laboratory investigations M.A.THOMSON AND C.MOSS 266 Dermoscopic pattern of intermediate stage in 204 Cutaneous Mycobacterium neoaurum infection causing seborrhoeic keratosis regressing to lichenoid keratosis: scarring alopecia in an immunocompetent host report of 24 cases L.K.MARTIN, R.LAWRENCE, S.KOSSARD AND D.F.MURRELL P.ZABALLOS, S.BLAZQUEZ, S.PUIG, E.SALSENCH, J.RODERO, J.M.VIVES AND 206 The use of intravenous immunoglobulin in cutaneous J.MALVEHY and recurrent perforating intestinal Degos disease 273 Long-term culture of multibacillary leprosy (malignant atrophic papulosis) macrophages isolated from skin lesions: a new model K.J.ZHU, Q.ZHOU, A.H.LIN, Z.M.LU AND H.CHENG to study Mycobacterium leprae–human cell interaction Contents III

D.F.MOURA, R.M.B.TELES, M.M.RIBEIRO-CARVALHO, R.B.TELES, I.M.C.F.SANTOS, 369 Topical becocalcidiol for the treatment of psoriasis H.FERREIRA, T.O.FULCO, J.A.C.NERY, E.P.SAMPAIO AND E.N.SARNO vulgaris: a randomized, placebo-controlled, double- 284 Transformed mycosis fungoides: clinicopathological blind, multicentre study features and outcome Y.R.HELFRICH, S.KANG, T.A.HAMILTON AND J.J.VOORHEES E.BARBERIO, L.THOMAS, F.SKOWRON, B.BALME AND S.DALLE 290 Men with Kennedy disease have a reduced risk of Concise communications androgenetic alopecia 375 No association between MDM2 SNP309 promoter R.SINCLAIR, K.J.GREENLAND, S.VAN EGMOND, C.HOEDEMAKER, A.CHAPMAN polymorphism and basal cell carcinoma of the skin AND J.D.ZAJAC S.WILKENING, K.HEMMINKI, P.RUDNAI, E.GURZAU, K.KOPPOVA, A.FO¨ RSTI AND R.KUMAR Contact dermatitis and allergy 378 Addition of topical pimecrolimus to once-daily mid- 295 Fragrance ingredient labelling in products on sale in potent steroid confers no short-term therapeutic the U.K. benefit in the treatment of severe atopic dermatitis; a D.A.BUCKLEY randomized controlled trial 301 Use of complementary and alternative treatment for J.M.SPERGEL, M.BOGUNIEWICZ, A.S.PALLER, A.A.HEBERT, P.R.GALLAGHER, allergic contact dermatitis C.MCCORMICK, A.PARNEIX-SPAKE AND T.HULTSCH E.NOIESEN, M.D.MUNK, K.LARSEN, M.HØYEN AND T.AGNER Case report Dermatological surgery and lasers 382 Multiple eruptive myxoid dermatofibromas: report of 306 Alteration of extracellular matrix modulators after first case and review of literature nonablative laser therapy in skin rejuvenation A.ANTAL, B.ZELGER, J.REIFENBERGER, T.NIEHUES, O.FEYEN, M.MEGAHED, J.OH, N.KIM, S.SEO AND I-H.KIM T.RUZICKA AND B.HOMEY Dermatopathology Gene corner 311 Acral lentiginous melanoma: histopathological 386 A novel mutation of the SLC39A4 gene causing prognostic features of 121 cases acrodermatitis enteropathica ´ A.PHAN, S.TOUZET, S.DALLE, S.RONGER-SAVLE, B.BALME AND L.THOMAS S.S.KILIC, M.GIRAUD, S.SCHMITT, S.BE´ZIEAU AND S.KU¨ RY 319 Stromelysin-3 expression in the differential diagnosis of dermatofibroma and dermatofibrosarcoma Correspondence protuberans: comparison with factor XIIIa and CD34 388 Severe persistent pemphigoid gestationis: long-term H.J.KIM, J.Y.LEE, S.H.KIM, Y.J.SEO, J.H.LEE, J.K.PARK, M.H.KIM, Y.W.CINN, remission with rituximab ` K.H.CHO AND T.Y.YOON G.CIANCHINI, C.MASINI, F.LUPI, R.CORONA, O.DE PITA AND P.PUDDU 325 A study of the secretion mechanism of the sebaceous 389 Livedo racemosa and digital necrosis in a patient with gland using three-dimensional reconstruction to primary seronegative antiphospholipid syndrome and examine the morphological relationship between the fibromuscular dysplasia of peripheral arteries sebaceous gland and the arrector pili muscle in the K.EISENDLE, W.JASCHKE AND N.SEPP follicular unit 392 Hypocomplementaemic urticarial vasculitis associated W-C.SONG, K-S.HU, H-J.KIM AND K-S.KOH with non-Hodgkin lymphoma and treatment with intravenous immunoglobulin Epidemiology and health services research D.SHAH, A.W.ROWBOTTOM, C.L.THOMAS, P.CUMBER AND 331 Fabry disease and the skin: data from FOS, the Fabry M.M.U.CHOWDHURY outcome survey 394 Consecutive use of different biological therapies in the C.H.ORTEU, T.JANSEN, O.LIDOVE, R.JAUSSAUD, D.A.HUGHES, G.PINTOS- treatment of psoriasis MORELL, U.RAMASWAMI, R.PARINI, G.SUNDER-PLASSMAN, M.BECK AND A.B. A.COSTANZO, M.PAPOUTSAKI, A.MAZZOTTA AND S.CHIMENTI MEHTA ON BEHALF OF THE FOS INVESTIGATORS 394 Cutaneous mercury deposits after henna dye 338 Trends in melanoma epidemiology suggest three application in the arm different types of melanoma G.MOUZOPOULOS, V.TSOUPAROPOULOS, M.STAMATAKOS, I.MIHELARAKIS, D.LIPSKER, F.ENGEL, B.CRIBIER, M.VELTEN AND G.HEDELIN D.PASPARAKIS AND E.AGAPITOS Photobiology 396 Development of Crohn disease in a patient on 344 The challenge of follow-up in narrowband ultraviolet etanercept for psoriasis B phototherapy K.AHMAD AND S.ROGERS B.L.DIFFEY AND P.M.FARR 397 Leptomeningeal melanoma and multiple cutaneous 350 Quantitative risk assessment of sunbeds: impact of new melanocytic naevi high power lamps J.PONNAMPALAM, J.NICHOLSON AND D.ATHERTON H.OLIVER, J.FERGUSON AND H.MOSELEY 398 Fibronectin and laminin expression in sentinel lymph nodes of patients with malignant melanoma Therapeutics A.GRADILONE, P.GAZZANIGA, E.CIGNA, F.VASATURO, B.VINCENZI, O.GANDINI, 357 Efficacy and tolerability of a Chinese herbal medicine I.SILVESTRI, D.RIBUFFO, S.SCARPA, N.SCUDERI AND A.M.AGLIANO` concoction for treatment of atopic dermatitis: a 401 A randomized study of minimal curettage followed by randomized, double-blind, placebo-controlled study topical photodynamic therapy compared with surgical K.L.E.HON, T.F.LEUNG, P.C.NG, M.C.A.LAM, W.Y.C.KAM, K.Y.WONG, K.C.K.LEE, excision for low-risk nodular basal cell carcinoma Y.T.SUNG, K.F.CHENG, T.F.FOK, K.P.FUNG AND P.C.LEUNG L.BERROETA, C.CLARK, R.S.DAWE, S.H.IBBOTSON AND C.J.FLEMING 364 Combination of surgical avulsion and topical therapy 403 Lymphocytic infiltration (Jessner–Kanof): lupus for single nail onychomycosis: a randomized erythematosus tumidus or a manifestation of controlled trial borreliosis? C.GROVER, S.BANSAL, S.NANDA, B.S.N.REDDY AND V.KUMAR M.KAATZ, B.ZELGER, J.NORGAUER AND M.ZIEMER IV Contents

405 Prognostic value of sentinel lymph node biopsy 441 On the role of the epidermal differentiation H.M.SHAW AND J.F.THOMPSON complex in ichthyosis vulgaris, atopic dermatitis and 405 Granulomatous slack skin with clonal T-cell receptor-c psoriasis gene rearrangement in skin and lymph node S.HOFFJAN AND S.STEMMLER I.E.BELOUSOVA, S.M.NIKONOVA, R.SIMA AND D.V.KAZAKOV 407 Fatal disseminated angioinvasive Fusarium falciforme Original articles infection in a patient with acute myeloid leukaemia Cutaneous biology S.J.YUN, M-G.SHIN, C.CHOI, H-J.KIM, J-B.LEE, S-J.KIM, S-C.LEE AND Y.H.WON 450 Selective biodegradation in hair shafts derived from 409 Maximizing the quality of review articles archaeological, forensic and experimental contexts A.S.WILSON, H.I.DODSON, R.C.JANAWAY, A.M.POLLARD AND M.J.SLADDEN AND C.S.SLADDEN 410 Erythrokeratoderma variabilis without GJB3 or GJB4 D.J.TOBIN mutation: a review of Japanese patients 458 Impaired cutaneous wound healing in granulocyte ⁄ M.NAKAMURA AND O.ISHIKAWA macrophage colony-stimulating factor knockout 411 Lichen sclerosus et atrophicus-like lesions in mycosis mice fungoides Y.FANG, S-J.GONG, Y-H.XU, B.D.HAMBLY AND S.BAO E.PARERA, A.TOLL, F.GALLARDO, B.BELLOSILLO, R.M.PUJOL AND R.MARTI´ Clinical and laboratory investigations 413 Does spontaneous autoimmunity improve survival in 466 Serum chemokine profiles in patients with alopecia visceral metastatic melanoma? areata S.VERCAMBRE-DARRAS, S.DUBUCQUOI, I.FAJARDY, F.PIETTE AND L.MORTIER Y.KUWANO, M.FUJIMOTO, R.WATANABE, N.ISHIURA, H.NAKASHIMA, 415 Filaggrin expression and the pathogenesis of epidermal Y.OHNO, S.YANO, N.YAZAWA, H.OKOCHI AND K.TAMAKI cysts 474 Mite-related bacterial antigens stimulate inflammatory I.KUROKAWA, K.UMEDA, K.NISHIMURA, K-I.YAMANAKA, A.HAKAMADA, K- cells in rosacea I.ISODA, A.TSUBURA AND H.MIZUTANI N.LACEY, S.DELANEY, K.KAVANAGH AND F.C.POWELL 417 Treatment-resistant classical epidermolysis bullosa 482 Reverse transcriptase activity in human normal and acquisita responding to rituximab psoriatic skin samples E.SADLER, B.SCHAFLEITNER, C.LANSCHUETZER, M.LAIMER, G.POHLA-GUBO, J.-P.MOLE`S, A.TESNIERE AND J.-J.GUILHOU R.HAMETNER, H.HINTNER AND J.W.BAUER 487 The vulval vestibular mucosa—morphological effects of 419 Urachal sinus presenting as periumbilical dermatitis oral contraceptives and menstrual cycle G.A.COX, I.CHAN, J.LLOYD, R.O.WITHEROW AND J.N.LEONARD U.JOHANNESSON, B.BLOMGREN, M.HILLIGES, E.RYLANDER AND 420 Ciclosporin improves quality of life in Kimura’s disease N.BOHM-STARKE M.VOURC’H, S.BARBAROT AND J-F.STALDER 494 Topical application of acidified nitrite to the nail 421 Treatment of lentigo maligna with topical 1% cidofovir renders it antifungal and causes nitrosation of cysteine D.CALISTA groups in the nail plate 423 Serum tissue polypeptide antigen correlating with M.J.FINNEN, A.HENNESSY, S.MCLEAN, Y.BISSET, R.MITCHELL, I.L.MEGSON AND clinical course in a patient with mycosis fungoides: a R.WELLER potential disease marker? 501 Molecular heterogeneity of familial porphyria cutanea S.C-S.HU, G-S.CHEN, C-S.WU AND C-C.E.LAN tarda in Spain: characterization of 10 novel mutations 425 On confusing prima facie validity with true validity in the UROD gene V.W.BERGER AND E.GEE M.ME´NDEZ, P.POBLETE-GUTIE´RREZ, M.GARCI´A-BRAVO, T.WIEDERHOLT, 426 Lymphomatoid granulomatosis complicating other M.J.MORA´N-JIME´NEZ, H.F.MERK, M.C.GARRIDO-ASTRAY, J.FRANK, haematological malignancies A.FONTANELLAS AND R.ENRI´QUEZ DE SALAMANCA F.M.MULLER, S.LEWIS-JONES, S.MORLEY, N.KERNOHAN, D.MEIKLEJOHN, 508 The impact of changes in clinical severity on J.R.GOODLAD AND A.EVANS psychiatric morbidity in patients with psoriasis: a Book review follow-up study 429 Photodermatology. Henry W. Lim, Herbert Ho¨nigsmann & F.SAMPOGNA, S.TABOLLI, D. ABENI AND THE IDI MULTIPURPOSE PSORIASIS John L.M. Hawk (editors). New York: Informa RESEARCH ON VITAL EXPERIENCES (IMPROVE) INVESTIGATORS Healthcare, 2007 Contact dermatitis and allergy A.V.ANSTEY 514 Delayed-type hypersensitivity to low molecular weight 430 News and Notices heparins and heparinoids: cross-reactivity does not depend on molecular weight 431 Corrigenda R.H.GRIMS, W.WEGER, H.REITER, E.ARBAB, B.KRA¨ NKE AND W.ABERER 518 Occupational dermatitis related to chromium and Volume 157, Number 3, September cobalt: experience of dermatologists (EPIDERM) and occupational physicians (OPRA) in the U.K. over an 11-year period (1993–2004) Snippets P.ATHAVALE, K.W.SHUM, Y.CHEN, R.AGIUS, N.CHERRY AND Research Snippets D.J.GAWKRODGER ON BEHALF OF EPIDERM Clinical Snippets Dermatopathology ink4a Review articles 523 p16 expression decreases during imiquimod 433 The optimal use of bexarotene in cutaneous T-cell treatment of anal intraepithelial neoplasia in human lymphoma immunodeficiency virus-infected men and correlates R.GNIADECKI, C.ASSAF, M.BAGOT, R.DUMMER, M.DUVIC, R.KNOBLER, with the decline of lesional high-risk human A.RANKI, P.SCHWANDT AND S.WHITTAKER papillomavirus DNA load Contents V

A.KREUTER, U.WIELAND, T.GAMBICHLER, P.ALTMEYER, H.PFISTER, K.TENNER- Gene corner RACZ, P.RACZ, A.POTTHOFF AND N.H. BROCKMEYER, FOR THE GERMAN 602 New KRT10 gene mutation underlying the annular NETWORK OF COMPETENCE HIV ⁄ AIDS variant of bullous congenital ichthyosiform 531 Evaluation of tumour-infiltrating CD4+CD25+FOXP3+ erythroderma with clinical worsening during regulatory T cells in human cutaneous benign and pregnancy atypical naevi, melanomas and melanoma metastases N.SHETH, D.GREENBLATT AND J.A.MCGRATH V.MOURMOURAS, M.FIMIANI, P.RUBEGNI, M.C.EPISTOLATO, V.MALAGNINO, 605 Recurrent p.N767S mutation in the ATP2A2 gene in a C.CARDONE, E.COSCI, M.C.DE NISI AND C.MIRACCO Japanese family with haemorrhagic Darier disease Epidemiology and health services research clinically mimicking epidermolysis bullosa simplex 540 Association or lack of association between tetracycline with mottled pigmentation class antibiotics used for acne vulgaris and lupus T.HAMADA, S.YASUMOTO, T.KARASHIMA, N.ISHII, H.SHIMADA, Y.KAWANO, erythematosus S.IMAYAMA, J.A.MCGRATH AND T.HASHIMOTO D.J.MARGOLIS, O.HOFFSTAD AND W.BILKER Correspondence 547 Community-based study of acne vulgaris in adolescents 609 Neonatal follicular mucinosis in Singapore S.DALLE, K.MARROU, B.BALME AND L.THOMAS H-H.TAN, A.W.H.TAN, T.BARKHAM, X-Y.YAN AND M.ZHU 610 Recessive dystrophic epidermolysis bullosa associated 552 Incidence of hand eczema in a population-based twin with dilated cardiomyopathy cohort: genetic and environmental risk factors S-W.OH, J.S.LEE, M.Y.KIM, J.Y.CHOI AND S-C.KIM A.LERBAEK, K.O.KYVIK, H.RAVN, T.MENNE´ AND T.AGNER 612 A case of milium-like syringoma with focal Paediatric dermatology calcification in Down syndrome 558 Extensive venous/lymphatic malformations causing S-H.SEO, C-K.OH, K-S.KWON AND M-B.KIM life-threatening haematological complications 614 Coccygeal polypoid eccrine naevus J.MAZEREEUW-HAUTIER, S.SYED, R.I.LEISNER AND J.I.HARPER S-W.OH, T-W.KANG, Y.C.KIM AND W.LEW 615 First case series on the use of calcipotriol– Therapeutics betamethasone dipropionate for morphoea 563 Intramuscular immunoglobulin for recalcitrant M.T.DYTOC, I.KOSSINTSEVA AND P.T.TING suppurative diseases of the skin: a retrospective review 618 Systemic sclerosis in association with peristomal of 63 cases pyoderma gangrenosum B.GOO, H.J.CHUNG, W.G.CHUNG AND K.Y.CHUNG M.FUJIKURA, T.OHTSUKA AND Y.OYAMADA 569 Double-blind, randomized, placebo-controlled study of 619 Actinic prurigo deterioration due to degradation of a lotion containing triethyl citrate and ethyl linoleate DermaGard window film in the treatment of acne vulgaris A.C.KERR AND J.FERGUSON A.CHARAKIDA, M.CHARAKIDA AND A.C.CHU 620 Successful treatment of primary cutaneous Concise communications follicle centre lymphoma with topical 5% 575 Inhibition of tumour necrosis factor-a secretion from imiquimod EpiDermTM tissues by a novel small molecule, UTL-5d A.STAVRAKOGLOU, V.L.BROWN AND I.COUTTS J.SHAW, C.LIU, R.MARTIN, B.CHEN, R.HOLTZ, W-H.HUANG AND A-R.LEE 622 Hydroxycarbamide: a treatment for lichen sclerosus? 580 Frequency of allergic contact dermatitis to isoeugenol N.TOMSON AND J.C.STERLING 622 Metastatic calcinosis cutis presenting as bilateral vulval is increasing: a review of 3636 patients tested from cysts 2001 to 2005 A.BISWAS, J.COOPER AND B.LATIFAJ J.M.L.WHITE, I.R.WHITE, A.GLENDINNING, J.FLEMING, D.JEFFERIES, 624 Calcified subcutaneous nodules: a long-term D.A.BASKETTER, J.P.MCFADDEN AND D.A.BUCKLEY 583 Increased lipopolysaccharide-induced tumour necrosis complication of interferon beta-1a therapy factor-a, interferon-c and interleukin-10 production in A.E.MACBETH, B.R.KENDALL, A.SMITH, G.SALDANHA AND K.E.HARMAN atopic dermatitis 625 Remission of photosensitivity following treatment of D.SIMON, L.R.BRAATHEN AND H-U.SIMON psoriasis vulgaris with tumour necrosis factor 587 Chronic hepatitis B reactivation: a word of caution inhibitors regarding the use of systemic glucocorticosteroid M.VIGUIER, M.JEANMOUGIN, E.BEGON, O.VEROLA, L.DUBERTRET AND therapy H.BACHELEZ C-H.YANG, T-S.WU AND C-T.CHIU 627 Recalcitrant lithium-induced psoriasis in a suicidal Case reports patient alleviated by tumour necrosis factor-a 591 Primary cutaneous marginal zone B-cell lymphoma in inhibition a patient with chronic lymphocytic leukaemia T.WACHTER, W.M.MURACH, E-B.BRO¨ CKER AND M.P.SCHO¨ N E.ROBAK, D.JESIONEK-KUPNICKA, T.ROBAK, A.HOLUB, E.WAWRZYNIAK, 629 Coexistence of two discordant B-cell lymphomas in the J.BARTKOWIAK, A.BEDNAREK, M.CONSTANTINU AND H.URBANSKA-RYS skin and lymph node: report of a case with primary 596 Oral steroid improves bullous pemphigoid-like clinical cutaneous follicle-center lymphoma and nodal mantle- manifestations in non-Herlitz junctional epidermolysis cell lymphoma bullosa with COL17A1 mutation G.GOTERI, S.RUPOLI, D.STRAMAZZOTTI, G.DISCEPOLI, A.R.SCORTECHINI, E.MABUCHI, N.UMEGAKI, H.MUROTA, T.NAKAMURA, K.TAMAI AND A.GIACCHETTI, D.MORICHETTI, A.TASSETTI, S.PULINI, S.MULATTIERI, I.KATAYAMA A.STRONATI AND P.LEONI 599 Giant bathing trunk naevus with lymphadenopathy and 631 Cutaneous vasculitis and T-large granular lymphocyte unusual pathology leukaemia with parallel evolution E.A.WEST, J.L.MCPARTLAND, H.RIGBY AND R.A.G.PARSLEW N.MEYER, J.DUFOUR, T.LAMY AND J.CHEVRANT-BRETON VI Contents

633 News and Notices Contact dermatitis and allergy 713 The incidence of occupational skin disease as reported 634 Errata to The Health and Occupation Reporting (THOR) network between 2002 and 2005 S.TURNER, M.CARDER, M.VAN TONGEREN, R.MCNAMEE, S.LINES, L.HUSSEY, Volume 157, Number 4, October A.BOLTON, M.H.BECK, M.WILKINSON AND R.AGIUS 723 Nickel allergy: relationship between patch test and Snippets repeated open application test thresholds ´ Research Snippets L.A.FISCHER, J.D.JOHANSEN AND T.MENNE 730 High frequency of contact allergy to gold in patients Clinical Snippets with endovascular coronary stents Review articles S.EKQVIST, C.SVEDMAN, H.MO¨ LLER, M.KEHLER, C.M.PRIPP, J.BJO¨ RK, 637 A systematic review of juvenile-onset clinically B.GRUVBERGER, E.HOLMSTRO¨ M, C.G.GUSTAVSSON AND M.BRUZE amyopathic dermatomyositis Dermatological surgery and lasers P.GERAMI, H.W.WALLING, J.LEWIS, L.DOUGHTY AND R.D.SONTHEIMER 739 Repeat liposuction-curettage treatment of axillary 645 Practical issues on interpretation of scoring atopic hyperhidrosis is safe and effective dermatitis: the SCORAD index, objective SCORAD and F.G.BECHARA, M.SAND, N.S.TOMI, P.ALTMEYER AND K.HOFFMANN the three-item severity score 744 The punch and graft technique: a novel method of A.P.ORANJE, E.J.GLAZENBURG, A.WOLKERSTORFER AND F.B.DE WAARD- surgical treatment for chondrodermatitis nodularis VAN DER SPEK helicis Meeting report N.RAJAN AND J.A.A.LANGTRY 649 Obesity in psoriasis: the metabolic, clinical and Dermatopathology therapeutic implications. Report of an interdisciplinary 748 Pseudoxanthoma elasticum: biopsy of clinically normal conference and review skin in the investigation of patients with angioid W.STERRY, B.E.STROBER AND A.MENTER ON BEHALF OF THE INTERNATIONAL streaks PSORIASIS COUNCIL S.J.BROWN, S.J.TALKS, S.J.NEEDHAM AND A.E.M.TAYLOR Original articles 752 The expression pattern of interferon-inducible proteins Cutaneous biology reflects the characteristic histological distribution of 656 Skin wound healing in the SKH-1 female mouse infiltrating immune cells in different cutaneous lupus following inducible nitric oxide synthase inhibition erythematosus subsets R.R.BELL, R.W.DUNSTAN AND N.K.KHAN J.WENZEL, S.ZAHN, S.MIKUS, A.WIECHERT, T.BIEBER AND T.TU¨ TING 662 Histone deacetylase inhibitors preferentially augment 758 A retrospective study addressed to understanding what transient transgene expression in human dermal predicts severe histological dysplasia ⁄early melanoma fibroblasts in excised atypical melanocytic lesions K.YASUKAWA, D.SAWAMURA, M.GOTO, H.NAKAMURA AND H.SHIMIZU R.M.STRAUSS, F.ELLIOTT, P.AFFLECK, A.P.BOON AND J.A.NEWTON-BISHOP Clinical and laboratory investigations Epidemiology and health services research 670 Quantitative analysis of Malassezia in the scale of patients 765 A scoring system for mucosal disease severity with with psoriasis using a real-time polymerase chain special reference to oral lichen planus reaction assay M.ESCUDIER, N.AHMED, P.SHIRLAW, J.SETTERFIELD, A.TAPPUNI, M.M.BLACK Y.TAKAHATA, T.SUGITA, M.HIRUMA AND M.MUTO AND S.J.CHALLACOMBE 674 Assessment of atopic eczema: clinical scoring and Concise communications noninvasive measurements 771 Functional redundancy of extracellular matrix protein 1 E.A.HOLM, H.C.WULF, L.THOMASSEN AND G.B.E.JEMEC in epidermal differentiation 681 Diagnosis of common dermatophyte infections by a S.SERCU, Y.POUMAY, F.HERPHELIN, J.LIEKENS, L.BEEK, A.ZWIJSEN, novel multiplex real-time polymerase chain reaction V.WESSAGOWIT, D.HUYLEBROECK, J.A.MCGRATH AND J.MERREGAERT detection ⁄identification scheme 776 Paradoxical effects of b-estradiol on epidermal M.ARABATZIS, L.E.S.BRUIJNESTEIJN VAN COPPENRAET, E.J.KUIJPER, G.S.DE permeability barrier homeostasis HOOG, A.P.M.LAVRIJSEN, K.TEMPLETON, E.M.H.VAN DER RAAIJ-HELMER, M.TSUTSUMI AND M.DENDA A.VELEGRAKI, Y.GRA¨SER AND R.C.SUMMERBELL 780 C16 laminin peptide increases angiotropic extravascular 690 Assessment of abnormal blood flow and efficacy of migration of human melanoma cells in a shell-less treatment in patients with systemic sclerosis using a chick chorioallantoic membrane assay newly developed microwireless laser Doppler C.LUGASSY, H.K.KLEINMAN, S.E.VERNON, D.R.WELCH AND R.L.BARNHILL flowmeter and arm-raising test M.KIDO, S.TAKEUCHI, S.HAYASHIDA, K.URABE, R.SAWADA AND M.FURUE Case reports 698 Fast and sensitive detection of Trichophyton rubrum DNA 783 Nephrogenic systemic fibrosis: a case series suggesting from the nail samples of patients with onychomycosis gadolinium as a possible aetiological factor by a double-round polymerase chain reaction-based J.A.MORENO-ROMERO, S.SEGURA, J.M.MASCARO´ JR, S.E.COWPER, M.JULIA`, assay E.POCH, A.BOTEY AND C.HERRERO A.K.GUPTA, M.ZAMAN AND J.SINGH 788 Human papillomavirus type 26 infection causing 704 Ceramide analogue 14S24 selectively recovers multiple invasive squamous cell carcinomas of the perturbed human skin barrier fingernails in an AIDS patient under highly active K.VA´VROVA´, A.HRABA´LEK, S.MAC-MARY, P.HUMBERT AND P.MURET antiretroviral therapy Contents VII

A.HANDISURYA, A.RIEGER, A.BANKIER, A.KOLLER, A.SALAT, G.STINGL AND 826 Rapid onset of multifocal human papillomavirus R.KIRNBAUER 72-associated oral intraepithelial neoplasia in a human 795 Flexural allergic contact dermatitis to benzalkonium immunodeficiency virus-infected patient chloride in antiseptic bath oil A.KREUTER, N.H.BROCKMEYER, P.ALTMEYER, H.PFISTER AND U.WIELAND FOR S.HANN, T.M.HUGHES AND N.M.STONE THE GERMAN COMPETENCE NETWORK HIV/AIDS Correspondence 828 Infliximab, as sole or combined therapy, induces rapid 799 Bortezomib-associated cutaneous vasculitis clearing of erythrodermic psoriasis X.GARCIA-NAVARRO, L.PUIG, M.T.FERNA´NDEZ-FIGUERAS, J.DALMAU, E.ROE M.D.F.TAKAHASHI, L.G.M.CASTRO AND R.ROMITI AND A.ALOMAR 831 Novel homozygous nonsense TMC8 mutation detected 801 The Wnt signalling ligand RSPO4, causing inherited in patients with epidermodysplasia verruciformis from anonychia, is not mutated in a patient with congenital a Brazilian family nail hypoplasia ⁄aplasia with underlying skeletal defects P.L.RADY, W.R.P.DE OLIVEIRA, Q.HE, C.FESTA, E.A.RIVITTI, S.B.TUCKER AND C.S.SEITZ, M.VAN STEENSEL, J.FRANK, J.SENDEREK, K.ZERRES, H.HAMM AND S.K.TYRING C.BERGMANN 833 Giant genital variant of folliculosebaceous cystic 803 Increasing the dose of cetirizine may lead to better hamartoma: successful management by CO2 laser and control of chronic idiopathic urticaria: an open study acitretin therapy of 21 patients J-J.BRU¨ CHER, I.FRANKE, J.ULRICH, H.GOLLNICK AND M.LEVERKUS Y.KAMEYOSHI, T.TANAKA, S.MIHARA, S.TAKAHAGI, N.NIIMI AND M.HIDE 835 Cold urticaria: tolerance induction with cold baths 804 Circumscribed palmar hypokeratosis: partial remission Y.A.VON MACKENSEN AND M.STICHERLING by photodynamic therapy 836 Aplasia cutis associated with coarctation of the aorta: S.BENOIT, C.S.SEITZ, H.HAMM, C.S.VETTER-KAUCZOK AND E-B.BRO¨ CKER could this be an incomplete form of Adams–Oliver 806 Primary essential cutis verticis gyrata with hyaluronic syndrome? acid deposition C.HERAS MULERO, R.BARTRALOT SOLER, L.RODRI´GUEZ-CANO, J.MOLLET ´ ˜ M.NISHIKAWA, M.TANIOKA, E.ARAKI, T.YAMAMOTO, T.SAKURAI, Y.MIYACHI SANCHEZ, L.PALACIO ALLER, G.APARICIO ESPANOL, D.BODET CASTILLO, ´ AND A.UTANI P.BASSAS FREIXAS AND V.GARCIA-PATOS 808 Transglutaminase 1 deficiency and corneocyte collapse: 838 Plexiform schwannoma mimicking haemangioma: an indication for targeted molecular screening in pitfalls in clinical diagnosis and histological autosomal recessive congenital ichthyosis interpretation G.ESPOSITO, G.TADINI, F.PAPARO, A.VIOLA, L.IENO, W.PENNACCHIA, S.LO, P.HOW AND A.L.M.MOSS F.MESSINA, L.GIORDANO, A.PICCIRILLO AND L.AURICCHIO 839 Idiopathic eruptive macular pigmentation in a 810 Photodynamic therapy of acne using methyl 7-year-old girl: case report and discussion of aminolaevulinate diluted to 4% together with low differences from erythema dyschromicum doses of red light perstans ¨ L.MAVILIA, G.MALARA, G.MORETTI, M.LO RE AND A.P.GUERRA A.VOLZ, D.METZE, M.BOHM, L.BRUCKNER-TUDERMAN AND D.NASHAN 811 Postkala-azar dermal leishmaniasis coexisting with 841 Acute cutaneous T-cell lymphoma transformation borderline tuberculoid leprosy during treatment with alemtuzumab S.BANSAL, A.GOEL, K.SARDANA, V.KUMAR AND N.KHURANA S.FAGUER, F.LAUNAY, L.YSEBAERT, C.MAILHOL, O.ESTINES-CHARTIER, 813 Lip verrucous carcinoma in a pregnant woman L.LAMANT AND C.PAUL successfully treated with carbon dioxide laser surgery 842 Recurrence of classical juvenile pityriasis rubra pilaris C.K.HSU, J.Y.Y.LEE, C.H.YU, M.M.L.HSU AND T.W.WONG in adulthood: report of a case 815 Central nervous system involvement in stage 1b J-B.HONG, H-C.CHIU, S-H.WANG AND T-F.TSAI mycosis fungoides 844 News and Notices A.LALLY, K.HOLLOWOOD, S.WHITTAKER AND R.TURNER 816 ‘Turkey ear’: a diagnosis or a physical sign? C.WILLIAMS, A.MITRA AND S.WALTON Volume 157, Number 5, November 818 A novel mutation of CYLD in a Chinese family with multiple familial trichoepithelioma and no CYLD protein expression in the tumour tissue Snippets Y-G.ZUO, Y.XU, B.WANG, Y-H.LIU, T.QU, K.FANG AND M.G.HO Research Snippets 821 Impact of a public service advertisement about Clinical Snippets onychomycosis on the health behaviour of the Greek Editorials population with nail disorders 847 The role of self-tests in the diagnosis of hair dye S.GREGORIOU, D.KALOGEROMITROS, G.LARIOS, M.MAKRIS AND allergy D.RIGOPOULOS J.M.L.WHITE AND I.R.WHITE 822 Photodistributed eruptive telangiectasia: an uncommon 849 ‘Benchside-to-bedside’ adverse drug reaction to venlafaxine H.TSAO AND J.ENGLISH M.VACCARO, F.BORGIA, O.BARBUZZA AND B.GUARNERI 824 An outbreak of creeping eruption in southern France Review articles P.DEL GIUDICE, E.CAUMES, C.BOISSY, F.LEDUFF, P.DELAUNAY, V.BLANC- 850 Psoriasis and psoriatic arthritis: separate or one and the AMRANE, S.L.GOFF-LEVAN, P.MARTY, Y.LE FICHOUX AND E.COUNILLON same? 825 A salutary case of Fumaderm potentially masking the D.H.CIOCON AND A.B.KIMBALL symptoms of bowel cancer and partial bowel 861 The safety of tacrolimus ointment for the treatment of obstruction atopic dermatitis: a review S.Y.NG AND J.WILKINSON M.H.A.RUSTIN VIII Contents

Original articles Dermatopathology From bench to bedside 970 Matrix metalloproteinases as mediators of tissue injury 874 Photoageing: mechanism, prevention and therapy in different forms of cutaneous lupus erythematosus M.YAAR AND B.A.GILCHREST T.M.JA¨RVINEN, P.KANNINEN, L.JESKANEN, S.KOSKENMIES, J.PANELIUS, T.HASAN, A.RANKI AND U.SAARIALHO-KERE Cutaneous biology 888 Potassium channel openers accelerate epidermal barrier Epidemiology and health services research recovery 981 Hairdressing and the prevalence of scalp disease in M.DENDA, M.TSUTSUMI, K.INOUE, D.CRUMRINE, K.R.FEINGOLD AND African adults P.M.ELIAS N.P.KHUMALO, S.JESSOP, F.GUMEDZE AND R.EHRLICH 989 Risk factors for acute generalized exanthematous Clinical and laboratory investigations pustulosis (AGEP)—results of a multinational 894 Nail thickness measurements using optical coherence case–control study (EuroSCAR) tomography and 20-MHz ultrasonography A.SIDOROFF, A.DUNANT, C.VIBOUD, S.HALEVY, J.N.BOUWES BAVINCK, M.MOGENSEN, J.B.THOMSEN, L.T.SKOVGAARD AND G.B.E.JEMEC L.NALDI, M.MOCKENHAUPT, J-P.FAGOT AND J-C.ROUJEAU 901 Clinical features and natural course of Behc¸et’s disease 997 The clinical effect of topical phenytoin on wound in 661 cases: a multicentre study healing: a systematic review E.ALPSOY, L.DONMEZ, M.ONDER, S.GUNASTI, A.USTA, Y.KARINCAOGLU, J.SHAW, C.M.HUGHES, K.M.LAGAN AND P.M.BELL B.KANDI, S.BUYUKKARA, O.KESEROGLU, S.UZUN, U.TURSEN, M.SEYHAN AND A.AKMAN Therapeutics 907 The significance of multiple blue-grey dots 1005 Comparison of cutaneous tolerance and efficacy of ) (granularity) for the dermoscopic diagnosis of calcitriol 3 lgg 1 ointment and tacrolimus ) melanoma 0Æ3mgg 1 ointment in chronic plaque psoriasis R.P.BRAUN, O.GAIDE, M.OLIVIERO, A.W.KOPF, L.E.FRENCH, J-H.SAURAT AND involving facial or genitofemoral areas: a double-blind, H.S.RABINOVITZ randomized controlled trial 914 Abnormal activator protein 1 transcription factor Y.H.LIAO, H.C.CHIU, Y.S.TSENG AND T.F.TSAI expression in CD30-positive cutaneous large-cell lymphomas Concise communications X.MAO, G.ORCHARD, R.RUSSELL-JONES AND S.WHITTAKER 1013 The ‘follicular trochanter’: an epithelial compartment 922 Pathophysiology of nocturnal scratching in childhood of the human hair follicle bulge region in need of atopic dermatitis: the role of brain-derived further characterization neurotrophic factor and substance P S.TIEDE, J.E.KLOEPPER, D.A.WHITING AND R.PAUS K-L.E.HON, M-C.A.LAM, K-Y.WONG, T-F.LEUNG AND P-C.NG 1017 A clinical assessment of a patch test kit marketed to 926 Can automated dermoscopy image analysis instruments U.K. hairdressers for detecting hair dye allergy provide added benefit for the dermatologist? A study D.I.ORTON comparing the results of three systems 1021 Missense mutation in exon 7 of TRPS1 gene A.PERRINAUD, O.GAIDE, L.E.FRENCH, J-H.SAURAT, A.A.MARGHOOB AND in an Italian family with a mild form of R.P.BRAUN trichorhinophalangeal syndrome type I 934 Association of human herpesvirus 6 reactivation A.ROSSI, V.DEVIRGILIIS, V.PANASITI, R.G.BORRONI, M.CARLESIMO, M.GENTILE, with the flaring and severity of drug-induced F.CARIOLA AND S.CALVIERI hypersensitivity syndrome Case reports M.TOHYAMA, K.HASHIMOTO, M.YASUKAWA, H.KIMURA, T.HORIKAWA, 1025 Chromoblastomycosis caused by Chaetomium funicola: K.NAKAJIMA, Y.URANO, K.MATSUMOTO, M.IIJIMA AND N.H.SHEAR a case report from Western Panama 941 Pityriasis lichenoides: the differences between children M.PIEPENBRING, O.A.CA´CERES MENDEZ, A.A.ESPINO ESPINOZA, R.KIRSCHNER and adults AND H.SCHO¨ FER S.WAHIE, E.HISCUTT, S.NATARAJAN AND A.TAYLOR 1030 Late presentation of erythropoietic protoporphyria: 946 Differentiation of tumour-stage mycosis fungoides, case report and genetic analysis of family members psoriasis vulgaris and normal controls in a pilot study L.BERROETA, I.MAN, D.R.GOUDIE, S.D.WHATLEY, G.H.ELDER AND using serum proteomic analysis S.H.IBBOTSON E.W.COWEN, C-W.LIU, S.M.STEINBERG, S.KANG, E.C.VONDERHEID, H.S.KWAK, 1032 Post-kala-azar dermal leishmaniasis as an immune S.BOOHER, E.F.PETRICOIN, L.A.LIOTTA, G.WHITELEY AND S.T.HWANG reconstitution inflammatory syndrome in a patient 954 A randomized controlled trial of pimecrolimus cream with acquired immune deficiency syndrome 1% in adolescents and adults with head and neck S.ANTINORI, E.LONGHI, G.BESTETTI, R.PIOLINI, V.ACQUAVIVA, A.FOSCHI, atopic dermatitis and intolerant of, or dependent on, S.TROVATI, C.PARRAVICINI, M.CORBELLINO AND L.MERONI topical corticosteroids Gene corner D.F.MURRELL, S.CALVIERI, J.P.ORTONNE, V.C.HO, S.WEISE-RICCARDI, N.BARBIER AND C.F.PAUL 1037 A novel point mutation in the gene encoding capillary 960 T-cell receptor repertoire in pyoderma gangrenosum: morphogenesis protein 2 in a Japanese patient with evidence for clonal expansions and trafficking juvenile hyaline fibromatosis T.N.BROOKLYN, A.M.WILLIAMS, M.G.S.DUNNILL AND C.S.PROBERT A.HATAMOCHI, T.SASAKI, T.KAWAGUCHI, H.SUZUKI AND S.YAMAZAKI Contact dermatitis and allergy Correspondence 967 Steroid allergy in patients with inflammatory bowel 1040 A pilot open trial evaluating the efficacy of low-dose disease aminopterin in the canine homologue of human atopic M.MALIK, A.-M.TOBIN, F.SHANAHAN, C.O’MORAIN, B.KIRBY AND J.BOURKE dermatitis Contents IX

T.OLIVRY, J.S.PAPS, P.BIZIKOVA, K.M.MURPHY, H.A.JACKSON AND 1071 Topical imiquimod in the treatment of a long-standing J.ZEBALA capillary malformation 1042 Cutaneous pseudolymphoma in association with D.J.KOUBA, D.YIP, E.F.FINCHER AND R.L.MOY Leishmania donovani 1073 Acrodermatitis continua of Hallopeau due to oral M.J.FLAIG AND R.A.RUPEC terbinafine 1044 Erythema annulare centrifugum associated with chronic F.NISHIWAKI, Y.MATSUMURA, N.MORITA, S.KORE-EDA, Y.MIYACHI AND lymphocytic leukaemia M.OMOTO I.HELBLING, R.WALEWSKA, M.J.S.DYER, M.BAMFORD AND K.E.HARMAN 1074 Extensive skin pigmentation caused by deposits of 1045 Erythema annulare centrifugum revealing chronic metallic particles following total elbow arthroplasty: lymphocytic leukaemia metallosis or not? J.STOKKERMANS-DUBOIS, M.BEYLOT-BARRY, B.VERGIER, K.BOUABDALLAH A.ASAHINA, H.FUJITA, S.FUKUDA, H.KAI, M.YAMAMOTO, N.HATTORI AND AND M.S.DOUTRE T.MORI 1047 Admissions to a U.K. teaching hospital with 1076 The rapid onset of multiple squamous cell carcinomas nonnecrotizing lower limb cellulitis show a marked during etanercept treatment for psoriasis seasonal variation L.LY AND D.CZARNECKI S.F.HAYDOCK, S.BORNSHIN, E.C.WALL AND R.M.CONNICK 1078 A case of do-it-yourself Mohs’ surgery using bloodroot 1048 Designing a validated patient information website: a obtained from the internet quality-controlled information portal illustrated by skin A.G.AFFLECK AND S.VARMA cancer 1079 Male genital lichen sclerosus and tacrolimus M.S.LLOYD, A.CLARK, A.PARIKH AND P.BUTLER C.B.BUNKER 1050 Multidisciplinary evaluation of patients with psoriasis 1080 The ‘handprint’ approximates to 1% of the total body presenting with musculoskeletal pain: a dermatology: surface area whereas the ‘palm minus the fingers’ does rheumatology clinic experience not E.MODY, M.E.HUSNI, P.SCHUR AND A.A.QURESHI C.L.THOMAS AND A.Y.FINLAY 1051 Therapeutic effect of lipoprostaglandin E1 on livedoid 1081 Resistance to hydroxychloroquine due to smoking in a vasculitis associated with essential cryoglobulinaemia patient with lupus erythematosus tumidus T.KAWAKAMI, K.KAWASAKI, M.MIZOGUCHI AND Y.SOMA R.HU¨ GEL, T.SCHWARZ AND R.GLA¨SER 1053 Epidermal cell necrosis with direct epidermal infiltration of Epstein–Barr virus (EBV)-encoded small 1083 News and Notices nuclear RNA-positive T lymphocytes in a patient with EBV-associated haemophagocytic syndrome Y.KAWACHI, M.ITOH, Y.FUJISAWA, J.FURUTA, Y.NAKAMURA, T.BANNO, Volume 157, Number 6, December T.TAKAHASHI AND F.OTSUKA 1056 Severe linear form of granuloma annulare along Blaschko’s lines preceding the onset of a classical form Snippets of granuloma annulare in a child Research Snippets F.MORICE-PICARD, F.BORALEVI, S.LEPREUX, C.LABRE`ZE, D.LACOMBE AND Clinical Snippets A.TAI¨EB Topical review 1058 Arndt–Gottron scleromyxoedema: successful therapy 1087 Paraneoplastic hypertrichosis lanuginosa acquisita: with intravenous immunoglobulins uncommon or overlooked? P.GHOLAM, M.HARTMANN AND A.ENK P.H.T.J.SLEE, R.I.F.VAN DER WAAL, J.H.SCHAGEN VAN LEEUWEN, R.A.TUPKER, 1060 Concordant lymphoma of cutaneous anaplastic large R.TIMMER, C.A.SELDENRIJK AND M.A.M.VAN STEENSEL cell lymphoma and systemic B-cell leukaemia H.SATO, Y.NAKAMURA, T.TAKAHASHI AND F.OTSUKA Review articles 1061 Radiation recall dermatitis in a patient affected with 1093 The relationships between exposure dose and response pheochromocytoma after treatment with lanreotide in induction and elicitation of contact hypersensitivity A.BAUZA´, L.J.DEL POZO, J.ESCALAS AND F.MESTRE in humans 1063 Sock-line bands in infancy P.S.FRIEDMANN D.R.BERK AND S.J.BAYLISS 1103 Psoriasis: evolution of pathogenic concepts and new 1064 CD8+ poikilodermatous mycosis fungoides with a therapies through phases of translational research nonaggressive clinical behaviour and a good response E.GUTTMAN-YASSKY AND J.G.KRUEGER to psoralen plus ultraviolet A treatment Guidelines ¨ ¨ S.ADA AND A.TULIN GULEC¸ 1116 Guidelines for evaluation and management of urticaria 1067 Disseminated cryptococcal infection in a patient with in adults and children severe psoriasis treated with efalizumab, methotrexate C.E.H.GRATTAN AND F.HUMPHREYS ON BEHALF OF THE BRITISH and ciclosporin ASSOCIATION OF DERMATOLOGISTS THERAPY GUIDELINES AND A.J.TUXEN, M.K.YONG, A.C.STREET AND C.DOLIANITIS AUDITSUBCOMMITTEE 1068 PTPN22 R620W polymorphism is not associated with pemphigus Orginal articles K.MEJRI, M.KALLEL- SELLAMI, E.PETIT-TEIXEIRA, O.ABIDA, H.MBAREK, Cutaneous biology M.ZITOUNI, M.BEN AYED, V.H.TEIXEIRA, M.MOKNI, B.FAZZA, H.TURKI, 1124 Cathelicidin LL-37 induces the generation of reactive F.TRON, D.GILBERT, H.MASMOUDI, F.CORNELIS AND S.MAKNI oxygen species and release of human a-defensins from 1070 ‘Smoker’s acne’: a new clinical entity? neutrophils B.CAPITANIO, J.L.SINAGRA, M.OTTAVIANI, V.BORDIGNON, A.AMANTEA AND Y.ZHENG, F.NIYONSABA, H.USHIO, I.NAGAOKA, S.IKEDA, K.OKUMURA AND M.PICARDO H.OGAWA X Contents

1132 Microarray analysis of aberrant gene expression in Photobiology actinic keratosis: effect of the Toll-like receptor-7 1230 Melanocortin 1 receptor (MC1R) genotype influences agonist imiquimod erythemal sensitivity to psoralen–ultraviolet A A.TORRES, L.STOREY, M.ANDERS, R.L.MILLER, B.J.BULBULIAN, J.JIN, photochemotherapy S.RAGHAVAN, J.LEE, H.B.SLADE AND W.BIRMACHU G.SMITH, M.J.V.WILKIE, Y.Y.DEENI, P.M.FARR, J.FERGUSON, C.R.WOLF AND 1148 Biphasic expression of stromal cell-derived factor-1 S.H.IBBOTSON during human wound healing Therapeutics A.TOKSOY, V.MU¨ LLER, R.GILLITZER AND M.GOEBELER 1235 Treatment of pyoderma gangrenosum with intravenous Clinical and laboratory investigations immunoglobulin 1155 CTACK ⁄CCL27 expression in psoriatic skin and its D.L.CUMMINS, G.J.ANHALT, T.MONAHAN AND J.H.MEYERLE modification after administration of etanercept 1240 Isotretinoin therapy and the incidence of acne relapse: A.CAMPANATI, G.GOTERI, O.SIMONETTI, G.GANZETTI, K.GIULIODORI, a nested case–control study D.STRAMAZZOTTI, D.MORICHETTI, M.L.BERNARDINI, B.MANNELLO, G.FABRIS L.AZOULAY, D.ORAICHI AND A.BE´RARD AND A.OFFIDANI 1161 Prevalence of Staphylococcus aureus toxins and nasal Concise communication carriage in furuncles and impetigo 1249 Psoriasis patients show signs of insulin resistance S.BOEHNCKE, D.THACI, H.BESCHMANN, R.J.LUDWIG, H.ACKERMANN, F.DURUPT, L.MAYOR, M.BES, M-E.REVERDY, F.VANDENESCH, L.THOMAS AND K.BADENHOOP AND W-H.BOEHNCKE J.ETIENNE 1168 Increased nuclear b-catenin in suprabasal involved Case reports psoriatic epidermis 1252 Unusual molecular findings in Kindler syndrome P.J.HAMPTON, O.K.ROSS AND N.J.REYNOLDS K.ARITA, V.WESSAGOWIT, A.C.INAMADAR, A.PALIT, H.FASSIHI, 1178 Cutaneous Malassezia flora in atopic dermatitis differs J.E.LAI-CHEONG, C.POURREYRON, A.P.SOUTH AND J.A.MCGRATH between adults and children 1257 A sporadic case of early-onset sarcoidosis resembling Y.TAKAHATA, T.SUGITA, H.KATO, A.NISHIKAWA, M.HIRUMA AND Blau syndrome due to the recurrent R334W missense M.MUTO mutation on the NOD2 gene 1183 Reduction of immunosuppression for transplant- P.COTO-SEGURA, S.MALLO-GARCIA, M.COSTA-ROMERO, J.I.AROSTEGUI, associated skin cancer: thresholds and risks J.YAGUE, E.RAMOS-POLO AND J.SANTOS-JUANES C.C.OTLEY, M.D.GRIFFIN, M.R.CHARLTON, B.S.EDWARDS, M.NEUBURG AND T.STASKO FOR THE REDUCTION OF IMMUNOSUPPRESSION TASK FORCE OF Gene corner THE INTERNATIONAL TRANSPLANT SKIN CANCER COLLABORATIVE 1260 COL7A1 mutational analysis in Korean patients with 1189 Morphoea: a manifestation of infection with Borrelia dystrophic epidermolysis bullosa species? S-W.OH, J.S.LEE, M.Y.KIM AND S-C.KIM K.EISENDLE, T.GRABNER AND B.ZELGER 1265 A novel homozygous mutation of the EVER1 ⁄TMC6 gene in a Japanese patient with epidermodysplasia Contact dermatitis and allergy verruciformis 1199 Filaggrin null alleles are not associated with hand S.AOCHI, G.NAKANISHI, N.SUZUKI, N.SETSU, D.SUZUKI, K.AYA AND eczema or contact allergy K.IWATSUKI A.LERBAEK, H.BISGAARD, T.AGNER, K.OHM KYVIK, C.N.A.PALMER AND T.MENNE´ Correspondence 1267 A case of primary anetoderma in an infant Dermatopathology H-J.YU, H.SHIN, M-S.KANG AND J-S.KIM 1205 The applicability and prognostic value of the new 1269 Successful treatment of severe psoriatic natal cleft TNM classification system for primary cutaneous fissuring with tissue adhesive lymphomas other than mycosis fungoides and Se´zary V.H.SMITH AND S.K.JONES syndrome: results on a large cohort of primary 1270 Recovery from Se´zary syndrome following Mycobacterium cutaneous B-cell lymphomas and comparison with the avium spondylitis system used by the Dutch Cutaneous Lymphoma A.YAMADA, O.YAMASAKI, K.ASAGOE, K.TSUJI, T.HAMADA, Y.OTA AND Group K.IWATSUKI N.J.SENFF AND R.WILLEMZE 1271 Late-onset neutropenia following rituximab treatment 1212 Involvement of E-cadherin, b-catenin, Cdc42 and in patients with autoimmune diseases CXCR4 in the progression and prognosis of cutaneous R.RIOS-FERNA´NDEZ, M.T.GUTIERREZ-SALMERO´ N, J-L.CALLEJAS-RUBIO, melanoma M.FERNA´NDEZ-PUGNAIRE AND N.ORTEGO-CENTENO M.G.TUCCI, G.LUCARINI, D.BRANCORSINI, A.ZIZZI, A.PUGNALONI, 1273 Hypertriglyceridaemia during treatment with A.GIACCHETTI, G.RICOTTI AND G.BIAGINI adalimumab in psoriatic arthritis Epidemiology and health services research G.STINCO, F.PICCIRILLO AND P.PATRONE 1217 Environmental factors, parental atopy and atopic 1274 Adalimumab for treatment of pyoderma gangrenosum R.G.POMERANTZ, M.E.HUSNI, E.MODY AND A.A.QURESHI eczema in primary-school children: a cross-sectional 1275 Effects of etanercept therapy on fatigue and symptoms study in Taiwan of depression in subjects treated for moderate to severe Y-L.LEE, C-W.LI, F-C.SUNG, H-S.YU, H-M.SHEU AND Y.L.GUO plaque psoriasis for up to 96 weeks Paediatric dermatology R.KRISHNAN, D.CELLA, C.LEONARDI, K.PAPP, A.B.GOTTLIEB, M.DUNN, 1225 Novel EBP gene mutations in Conradi–Hu¨nermann– C.F.CHIOU, V.PATEL AND A.JAHREIS Happle syndrome 1277 Cutaneous Langerhans cell histiocytosis in an elderly P.M.STEIJLEN, M.VAN GEEL, M.VREEBURG, D.MARCUS-SOEKARMAN, man successfully treated with narrowband ultraviolet B L.J.M.SPAAPEN, F.C.M.CASTELIJNS, M.WILLEMSEN AND M.A.M.VAN STEENSEL S.IMAFUKU, S.SHIBATA, A.TASHIRO AND M.FURUE Contents XI

1279 Pyoderma gangrenosum and interleukin 8 1296 Hydroxyzine-induced acute generalized exanthematous M.OKA pustulosis 1281 Recurrent KIND1 (C20orf42) gene mutation, c.676insC, Y-S.TSAI, M-E.TU, Y-H.WU AND Y-C.LIN in a Brazilian pedigree with Kindler syndrome 1297 Malignant melanoma in a woman with LEOPARD B.C.F.MARTIGNAGO, J. E.LAI-CHEONG, L.LIU, J.A.MCGRATH AND T.F.CESTARI syndrome: identification of a germline PTPN11 1284 Transient macular erythema with eosinophilia in a mutation and a somatic BRAF mutation patient carrying the FIP1L1-PDGFRA fusion gene M.SEISHIMA, Y.MIZUTANI, Y.SHIBUYA, C.ARAKAWA, R.YOSHIDA AND H.YAHARA, T.SATOH, T.HASHIMOTO AND H.YOKOZEKI T.OGATA 1287 A case of milia en plaque successfully treated with oral 1299 Anonychia, hyponychia and spontaneous amputation etretinate of the distal phalanges as a consequence of ischaemic N.ISHIURA, M.KOMINE, T.KADONO, K.KIKUCHI AND K.TAMAKI necrosis of the extremities after umbilical 1289 Limited cutaneous systemic sclerosis associated with catheterization discoid lupus erythematosus in two Japanese patients I.NERI, F.SAVOIA, F.GIACOMINI AND A.PATRIZI with anticentromere antibodies 1301 Basal cell carcinomas of the inner canthus: incidence T.KAWAKAMI, K.KAWASAKI AND Y.SOMA of incomplete excision according to topographical 1291 Coexistence of primary cutaneous anaplastic large cell localization of tumours lymphoma and mycosis fungoides in a patient with F.BORIANI AND F.MARCONI B-cell chronic lymphocytic leukaemia 1303 News and Notices M.MARSCHALKO´ , J.CSOMOR, N.ERO}S, A´.SZIGETI, J.HA´RSING, J.SZAKONYI, M.DE´SAKNAI, A.MATOLCSY, J.DEMETER AND S.KA´RPA´TI 1304 Corrigenda 1293 Usefulness of the QuantiFERON test in the confirmation of latent tuberculosis in association with 1304 Erratum erythema induratum 1305 Author Index J.ANGUS, C.ROBERTS, K.KULKARNI, I.LEACH AND R.MURPHY 1294 Respiratory involvement in toxic epidermal necrolysis 1324 Subject Index portends a poor prognosis that may not be reflected in SCORTEN J.S.HAGUE, J.M.R.GOULDING, T.M.W.LONG AND B.C.GEE RESEARCH SNIPPETS DOI 10.1111/j.1365-2133.2007.08347.x

Biphasic expression of stromal cell-derived factor-1 during human wound healing Chemokines such as SDF-1/CXCL12 not only regulate the distribution of invading leucocytes but also control tissue remodelling. Employing in situ hybridization, Toksoy ne et al. studied SDF-1/CXCL12 expression after artificial wounding of human skin. While it was evenly expressed by endothelial cells and fibroblasts in undisturbed skin, a ns remarkable spatial pattern became obvious after injury: SDF-1/CXCL12 was upregulated at the wound margins whereas it was completely suppressed in the fibrous neostroma that replaced the provisional matrix covering the initial wound defect. Altogether, the data suggest that SDF-1/CXCL12 exerts multiple functions during wound repair that include regulation of neoangiogenesis, fibroblast proliferation and epithelialization. Toksoy A, Mu¨ller V, Gillitzer R, Goebeler M. Biphasic expression of stromal cell-derived factor-1 during human wound healing. Br J Dermatol 2007; 157:1148–54. The role of filaggrin null alleles in hand eczema and contact allergy Carriers of one of the variant filaggrin alleles strongly associated with atopic dermatitis, R501X and 2282del4, have varying degrees of impaired skin barrier. Genetic risk factors influence the risk of hand eczema, and an impaired skin barrier facilitates the penetration of contact allergens. Thus, a possible role of the variant alleles in the pathogenesis of hand eczema and contact allergy could be hypothesized. However, in this genetic-epidemiological study no association between the variant filaggrin alleles and hand eczema or contact allergy could be demonstrated. Lerbaek A, Bisgaard H, Agner T et al. Filaggrin null alleles are not associated with hand eczema or contact allergy. Br J Dermatol 2007; 157:1199–204. The applicability and prognostic value of the new TNM classification system for primary cutaneous lymphomas other than mycosis fungoides and Se´zary syndrome Research in primary cutaneous lymphomas (PCL) has been hampered by inconsistent classifications and lack of a consistent system to define disease extent. Recently, a TNM classification for PCL other than mycosis fungoides and Se´zary syndrome was proposed for consistent reporting of disease extent. Senff et al. tested the applicability and prognostic significance of this TNM classification on 300 patients with an indolent or aggressive type of cutaneous B-cell lymphoma, as recognized in the WHO–EORTC classification. The TNM system was found to be useful in documenting disease extent and showed prognostic significance in the aggressive, but not in the indolent, subgroups. Senff NJ, Willemze R. The applicability and prognostic value of the new TNM classification system for primary cutaneous lymphomas other than mycosis fungoides and Se´zary syndrome: results on a large cohort of primary cutaneous B-cell lymphomas and comparison with the system used by the Dutch Cutaneous Lymphoma Group. Br J Dermatol 2007; 157:1205–11. MC1R genotype influences erythemal sensitivity to PUVA Psoralen–ultraviolet A (PUVA) photochemotherapy is widely used to treat psoriasis and other common skin diseases. There are unpredictable interindividual differences in the acute erythemal effects of PUVA and its longer-term risk of skin cancer. Polymorphisms in the melanocortin 1 receptor (MC1R) have been associated with increased sensitivity to UV radiation and skin cancer risk. Smith et al. studied MC1R genotype in patients commencing PUVA and found that inheritance of the Val60Leu, Arg163Gln and multiple MC1R SNPs was associated with increased erythemal sensitivity. These data suggest that MC1R genotype influences PUVA sensitivity and increase our understanding of genetic factors that predict photochemotherapy responses. Smith G, Wilkie MJV, Deeni YY et al. Melanocortin 1 receptor (MC1R) genotype influences erythemal sensitivity to psoralen– ultraviolet A photochemotherapy. Br J Dermatol 2007; 157:1230–4.

Isotretinoin therapy and the incidence of acne relapse 1 Isotretinoin is an effective treatment for severe nodular acne. However, data regarding 0·8 its long-term benefits are lacking. Using a population-based cohort of 17 351 first- time isotretinoin users followed for up to 20 years, Azoulay et al. found that 41% of 0·6 subjects experienced an acne relapse necessitating further treatment with an antiacne 0·4 agent (isotretinoin or other). The authors also identified several predictors of 0·2 experiencing an acne relapse. Thus, in view of isotretinoin’s relatively high relapse rate

Proportion of patients in the cohort 0 and important side-effects profile, these data could be of prognostic value to clinicians 0 50 100 150 200 250 who treat patients with acne. Time (months) Azoulay L, Oraichi D, Be´rard A. Isotretinoin therapy and the incidence of acne relapse: a nested case–control study. Br J Dermatol 2007; 157:1240–8. TOPICAL REVIEW DOI 10.1111/j.1365-2133.2007.08253.x Paraneoplastic hypertrichosis lanuginosa acquisita: uncommon or overlooked? P.H.T.J. Slee, R.I.F. van der Waal,* J.H. Schagen van Leeuwen, R.A. Tupker,* R. Timmer, C.A. Seldenrijk§ and M.A.M. van Steensel– Departments of Internal Medicine, *Dermatology, Obstetrics and Gynaecology, Gastroenterology and §Pathology, St Antonius Hospital, PO Box 2500, 3430 EM Nieuwegein, the Netherlands –Department of Dermatology, University Medical Hospital Maastricht, the Netherlands

Summary

Correspondence Acquired hypertrichosis lanugo-type or hypertrichosis lanuginosa acquisita (HLA) P.H.T.J. Slee. is often associated with metabolic and endocrine disorders and use of certain E-mail: [email protected] drugs. The occurrence of HLA with malignancy was first noted in 1865, and it has since been described in 56 patients as a paraneoplastic syndrome both in Accepted for publication 13 August 2007 women and in men. Sometimes HLA occurs concurrent with acanthosis nigri- cans, papillary hypertrophy of the tongue, and glossitis. The predominance of Key words female cases is striking. Malignancy-associated HLA seems to occur especially in acquired hypertrichosis lanugo-type, hypertrichosis the age group 40–70 years. In women with HLA the most frequent malignancy lanuginosa acquisita, malignancy, paraneoplastic is colorectal cancer, followed in order by lung cancer and breast cancer; in men Conflicts of interest lung cancer is the malignancy most frequently associated with HLA, followed by None declared. colorectal cancer. In 3 years we saw 10 patients with HLA, in whom the malig- nancy was usually metastasized. Only one patient had local disease; after removal of the primary tumour it took 2 years before the lanugo hair recurred. The aeti- ology of the syndrome is not clear: no specific hormonal or biochemical abnor- malities have been identified as yet. The difference between hirsutism and lanugo-type hypertrichosis is discussed. It is stressed that the appearance of lanugo-type hypertrichosis in body areas previously perceived by patients as ‘hairless’ is highly indicative of internal malignancy.

Hypertrichosis, defined as excessive hair growth, is a problem Lanugo hairs physiologically grow in utero, contain no that may present itself in various clinical patterns. Recognition medulla, are not pigmented, are long and are shed during the of the type of hypertrichosis is very important for the diagno- last months of pregnancy up to the first months after birth.1 sis and therapy of the underlying disease. In contrast to hirsutism, lanugo-type hypertrichosis does not show a gender-specific distribution pattern. Patients with Hypertrichosis acquired hypertrichosis lanugo-type or hypertrichosis lanugin- osa acquisita (HLA) grow lanugo-type hair near their eye- Hypertrichosis may involve lanugo hair, vellus hair or terminal brows and on their forehead, ears and nose. Some patients hair.1 Lanugo hairs are long, thin and unpigmented, like have extensive involvement that includes the extremities, axil- wool. Vellus or intermediate hairs are short, unpigmented lae and trunk, but the palmoplantar, suprapubic and genital hairs; they are variably medullated. Terminal hairs contain a areas are rarely involved.2 medulla, and are longer, thick, and pigmented. These hairs Hypertrichosis lanuginosa may occur as congenital or are involved in hirsutism, which is defined as an adult male- acquired forms. Only the lanugo-type of hypertrichosis will pattern hair overgrowth in children or women. Most authors be discussed here. For the other types excellent reviews are consider hirsutism as a subclass of hypertrichosis. This type of available.1 hypertrichosis appears predominantly on body areas with androgen-sensitive follicles, e.g. the chest, beard and mous- Hypertrichosis lanuginosa congenita tache regions. Hirsutism occurs as a result of androgen activ- ity: either as a consequence of elevated systemic androgen Hypertrichosis lanuginosa congenita or congenital lanugo levels, or by means of local elevated sensitivity for androgens, hypertrichosis (CLH) often represents a normal variation in or a combination of both mechanisms. In hirsutism endocrine premature neonates that have not gone through the in utero evaluation is warranted. shedding phase because of premature birth, but is rarely

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 1087 1088 Paraneoplastic hypertrichosis lanuginosa acquisita, P.H.T.J. Slee et al. encountered in full-term children. In CLH the hair is spread epithelial tumours, e.g. proteins from the Wingless family and over the entire body, sparing only the palms, soles and b-catenin. Especially the latter is of interest, for it is capable of mucous membranes. It can be inherited as an autosomal dom- initiating de novo hair growth in adult mice.60 inant feature, but sporadic cases also occur.1,3 The recent observation that treatment with epidermal growth factor (EGF) receptor (EGFR) antagonists can result in 61 Hypertrichosis lanuginosa acquisita hypertrichosis is also quite intriguing in this respect. EGF signalling is known to play an important role in the regulation HLA may occur in metabolic and endocrine diseases, e.g. por- of hair follicle growth and development. Mutations in the phyria and hypo- or hyperthyroidism, or may be induced by gene coding for the EGFR in the mouse cause disturbances of drugs such as ciclosporin, penicillamine, psoralens, glucocorti- hair growth.62 Considering the involvement of EGF signalling costeroids, diazoxide, interferon, minoxidil, phenytoin or in many solid malignancies such as lung cancer, it is tempting cetuximab.1,3 Furthermore, HLA can be associated with neo- to speculate that the production of EGFR ligands by the plastic disease, as outlined below (Tables 1 and 2). tumour could contribute to HLA. The skin symptoms that are associated with it, such as acanthosis nigricans, conceivably Hypertrichosis lanuginosa acquisita as reflect skin overgrowth caused by activation of EGF signalling. paraneoplastic manifestation The occurrence of lanugo-type hypertrichosis on body regions that a patient previously has perceived as ‘hairless’ should be In 1865, Turner reported the association of HLA with malig- interpreted as an important marker of an underlying malig- nancy in a woman with breast cancer.4 Several case reports nancy, after ruling out the above-mentioned other factors have appeared since, leading to a total of only 56 reported (hormones, drugs) that may be involved in HLA.19 Although patients to date.4–56 A comprehensive search of scientific pub- HLA may arise as a paraneoplastic manifestation in various lications via PubMed (1950–2005) and Google Scholar was cancers, it has predominantly been reported in lung and colo- carried out. The search terms used were [(‘hypertrichosis’) rectal cancer (Table 1). The predominance of these cancers AND (‘lanuginosa’ OR ‘lanugo’ OR ‘lanugo-type’)] AND may partially be explained by their higher prevalence rates. [‘acquisita’ OR ‘acquired’] AND [‘malignancy’ OR ‘Neo- Therefore, in patients with HLA in the absence of malignancy, plasms’ (Mesh) OR ‘paraneoplastic’]. All articles were checked the following work-up is advised, apart from history taking for related links and citations. Articles were checked to find and physical examination: blood examination, a chest X-ray out whether another aetiology was (more) reasonable. Essen- and colonoscopy.3 tial data of these patients are outlined in Table 1. HLA is strik- ingly predominant in women (female to male ratio 2Æ3) and, Own experience according to reported cases, it seems to occur mainly between ages 40 and 70 years (mean 54).4–56 In women, colorectal In the past 3 years 10 patients with paraneoplastic HLA were carcinoma is the most observed associated malignancy, fol- diagnosed in our hospital (Table 2). A combined pattern of lowed by lung cancer and breast cancer. In men, lung cancer lanugo-type hypertrichosis and hirsutism was encountered in is the most frequently encountered malignancy, followed three of them. Two of them had a sex cord stromal tumour of by colorectal carcinoma.4–56 Hair growth may occur from the ovary as underlying neoplasia, and a third was diagnosed 2Æ5 years before the tumour is identified up to 5 years after with an adrenocortical tumour. Production of androgens by diagnosis. Patients with HLA usually have metastatic disease at these three malignancies probably explains the combined pat- the time of diagnosis and, therefore, a poor prognosis. How- tern of hypertrichosis. Only one of the seven patients with ever, successful antitumour therapy is associated with regres- lanugo-type hypertrichosis without hirsutism had localized sion of the hair growth. disease. After removal of the primary tumour, an adenocarci- For reasons unknown, paraneoplastic HLA usually pro- noma of the endometrium, it took 2 years before the lanugo gresses in a craniocaudal direction and may be accompanied hair regrew. Data of all patients are presented in Table 2.63 by acanthosis nigricans, hypertrophy of tongue papillae, and A characteristic picture of patient 2 is given in Figure 1. glossitis.1 The pathogenesis of paraneoplastic HLA has not yet Hypertrichosis was also recognized in patients who were been elucidated. Hypothetically, tumour-produced cytokines previously treated with chemotherapy and who experienced may not only stimulate tumour growth, but may also promote alopecia. During regrowth of hair hypertrichosis was noted, as proliferation of other cells, including those of the hair folli- is reported in the literature. Such patients were not included cles. The concurrence of HLA and acanthosis nigricans might in Table 2. Currently, this is the largest series of patients with point towards the involvement of an insulin-like growth fac- paraneoplastic HLA in a single institution, bringing the total tor, but this has not been identified to date.57 Furthermore, number of reported patients to 63. In Table 3 an overview of various fibroblast growth factors (FGFs) are known to be all reported patients with paraneoplastic HLA is provided. In involved in the regulation and differentiation of hair view of the fact that this many patients were detected in just growth.58,59 FGF production has been demonstrated in lung 3 years, paraneoplastic HLA may be substantially under- tumours. Several other growth factors that play a role in the reported. This is probably mainly due to unfamiliarity of clini- initiation of hair follicle growth are produced by malignant cians with this particular paraneoplastic syndrome. Also,

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 Paraneoplastic hypertrichosis lanuginosa acquisita, P.H.T.J. Slee et al. 1089

Table 1 Overview of all reported cases in the literature

Age Advanced First author (year) Sex (years) or localized Localization Malignancy Turner (1865)4 F 42 Adv Breast Carcinoma Lyell (1951)5 F 35 Adv Urinary bladder Transitional cell carcinoma Dingley (1957)6 F 56 Adv Ovary Adenocarcinoma Fretzin (1967)7 M 69 Adv Lung Anaplastic carcinoma Herzberg (1969)8 M 65 Adv Gallbladder Adenocarcinoma Hensley (1969)9 M 41 Adv Lung Small cell carcinoma Chadfield (1970)10 F 78 Adv Rectum Adenocarcinoma Djajadiningrat (1970)11 F 43 Adv Rectum Adenocarcinoma Hegedus (1972)12 F 45 Adv Colon Adenocarcinoma F 56 Adv Colon Adenocarcinoma van der Lugt (1973)13 F 73 Adv Colon Adenocarcinoma Anderson (1973)14 M 63 Adv Lung Squamous cell carcinoma Reinhold (1974)15 F 60 Adv Colon Adenocarcinoma Samson (1975)16 F 66 Adv Endometrium Adenoacanthoma F 35 Adv Mediastinum Histiocytic lymphoma Rzempoluch (1976)17 F 44 Adv Lung Anaplastic carcinoma Kaiser (1976)18 F 46 Adv Endometrium Adenocarcinoma Wadskow (1976)19 F 54 Loc Breast Carcinoma McLean (1977)20 F 19 Loc Pancreas Islet cell carcinoma Ikeya (1978)21 M 69 Loc Lung Undifferentiated carcinoma Davies (1978)22 F 59 Loc Colon Carcinoid Ricken (1979)23 F 24 Adv Leukaemia Chronic lymphocytic Goodfellow (1980)24 M 61 Loc Lung Polygonal cell carcinoma Gonza´lez (1980)25 M 71 Adv Urinary bladder Transitional cell carcinoma Shee (1981)26 M 57 Adv Lung Adenocarcinoma Knowling (1982)27 M 62 Adv Lung Adenocarcinoma F 51 Adv Lung Adenocarcinoma Sindhuphak (1982)28 F 32 Adv Unknown primary Adenocarcinoma Ulrich (1983)29 F 34 Adv Lung Carcinoma George (1983)30 F 46 Adv Breast Duct carcinoma Price (1985)31 F 63 Adv Colon Adenocarcinoma Kassis (1985)32 F 54 Loc Endometrium and lunga Anaplastic adenocarcinoma Jemec (1986)33 F 48 Adv Lymphoma Follicular Skaf (1986)34 F 62 Adv Colon Adenocarcinoma Hovenden (1987)35 F 76 Adv Lung Clear cell carcinoma Dyall-Smith (1987)36 M 48 Adv Colon Adenocarcinoma Carratala´ (1987)37 F 69 Adv Lung Adenocarcinoma Mengori (1989)38 M 69 Adv Colon Adenocarcinoma Rodriguez (1990)39 F 30 Adv Lung Carcinoma Salazar (1990)40 M 50 Adv Lung Undifferentiated carcinoma De Clercq (1990)41 F 58 Adv Breast Duct carcinoma McKenna (1992)42 F 65 Adv Endometrium Adenocarcinoma Begany (1992)43 F 32 Adv Skin Melanoma Brinkmann (1992)44 M 30 Loc Colon Adenocarcinoma Duncan (1994)45 M 69 Adv Kidney Renal cell carcinoma Toyoki (1998)46 M 75 Adv Rectum Adenocarcinoma Farin˜a (1998)47 F 66 Adv Breast Duct carcinoma Maier (1999)48 F 27 Adv Parotid gland Planocellular carcinoma Pe´rez-Losada (2001)49 F 62 Loc Sarcoma Extraskeletal Ewing sarcoma Bauer (2001)50 M 70 Adv Leukaemia Acute myeloid Sanchez-Estella (2005)51 F 50 Loc Cervix Squamous cell carcinoma Lorette (2006)52 F 68 Loc Rectum Adenocarcinoma Pruijm (2007)53 F 58 Adv Unknown primary Adenocarcinoma Saad (2007)54 F 51 Loc Stomach Adenocarcinoma Wyatt (2007)55 M 93 Adv Prostate Adenocarcinoma Vulink (2007)56 F 57 Adv Breast Adenocarcinoma

Loc, localized disease; Adv, advanced disease. aFour months later lung cancer was diagnosed.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 1090 Paraneoplastic hypertrichosis lanuginosa acquisita, P.H.T.J. Slee et al.

Table 2 Series of patients with paraneo- Age Tumour plastic hypertrichosis lanuginosa acquisita at Patient Sex (years) localization Tumour type St Antonius Hospital 1 F 53 Lung Advanced nonsmall cell carcinoma 2 F 62 Colon Advanced adenocarcinoma 3 F 27 Colon Advanced adenocarcinoma 4 F 61 Colon Advanced adenocarcinoma 5 F 60 Endometrium Localized adenocarcinoma 6 F 56 Breast Advanced adenocarcinoma 7 F 38 Breast Advanced neuroendocrine carcinoma 8 F 80 Ovary Localized sex cord stromal tumour 9 F 53 Ovary Advanced sex cord stromal tumour 10 F 63 Adrenal gland Advanced adrenocortical tumour

Table 3 Overview of all reported patients with paraneoplastic hypertrichosis laniginosa acquisita, including our series (in brackets)

Cumulative Tumour localization Men Women reported cases Lung 8 7a (+1) 16 Colorectum 4 10 (+3) 17 Breast 0 6 (+2) 8 Uterus 0 4a (+1) 5 Urinary bladder 1 1 2 Lymphoma 0 2 2 Chronic lymphocytic 01 1 leukaemia Acute myeloid 10 1 leukaemia Ovary 0 1 1 Unknown primary 0 2 2 Pancreas 0 1 1 Gallbladder 1 0 1 Melanoma 0 1 1 Kidney 1 0 1 Parotid gland 0 1 1 Ewing sarcoma 0 1 1 Cervix 0 1 1 Stomach 0 1 1 Prostate 1 0 1 Total 17 47 64

aThe case of endometrial and lung carcinoma32 has been coun- ted twice.

taking, physical examination and further diagnostics when HLA is diagnosed. If hypertrichosis is seen in an adult without Fig 1. Hypertrichosis in patient 2 (with permission of J Clin Oncol63). associated metabolic disorders or a history of taking a drug known to cause this phenomenon, it is imperative to search for a malignancy. hypertrichosis in men is less obvious against a more hairy background and is cosmetically more accepted than in women. This may be the reason why most cases reported so References far have been in women. Furthermore, in women hypertri- 1 Wendelin DS, Pope DN, Mallory SB. Hypertrichosis. J Am Acad chosis may be missed because of the ready availability of Dermatol 2003; 48:161–79. various hair removal techniques. 2 Poole S, Fenske NA. Cutaneous markers of internal malignancy. In this context Goethe is cited, who wrote: ‘What one II. Paraneoplastic dermatoses and environmental carcinogens. JAm knows, one sees.’ We reiterate the importance of history Acad Dermatol 1993; 28:147–64.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 Paraneoplastic hypertrichosis lanuginosa acquisita, P.H.T.J. Slee et al. 1091

3 Cohen PR, Kurzrock R. Mucocutaneous paraneoplastic syndromes. 28 Sindhuphak W, Vibhagool A. Acquired hypertrichosis lanuginosa. Semin Oncol 1997; 24:334–59. Int J Dermatol 1982; 21:599–601. 4 Turner M. Case of a woman whose face and body in two or three 29 Ulrich LG, Munk-Jensen N. Acquired hypertrichosis lanuginosa. weeks’ time became covered with a thick crop of short and white Review of literature and description of a case. Ugeskr Laeger 1983; downy hair. Med Time Gaz 1865; 2:507. 145:2586–7. 5 Lyell A, Whittle CH. Hypertrichosis lanuginosa, acquired type. 30 George SL, Whitton A, Plowman PN. Hair regrowth after cancer Proc R Soc Med 1951; 44:576–7. therapy. Hum Toxicol 1983; 2:465–72. 6 Dingley ER, Marten RH. Adenocarcinoma of the ovary presenting 31 Price ML, Hall-Smith SP. Hypertrichosis lanuginosa acquisita. as acanthosis nigricans. J Obstet Gynaecol Br Emp 1957; 64:898– Clin Exp Dermatol 1985; 10:255–7. 900. 32 Kassis V, Kassis E, Keiding L, Thomsen HK. Hypertrichosis lanugin- 7 Fretzin DF. Malignant down. Arch Dermatol 1967; 95:294–7. osa acquisita associated with multiple malignancies. J Am Acad 8 Herzberg JJ, Potjan K, Gebauer D. Acquired hypertrichosis lanugin- Dermatol 1985; 12:1106–7. osa: a new cutaneous paraneoplastic syndrome. Ann Dermatol Syphiligr 33 Jemec GB. Hypertrichosis lanuginosa acquisita. Report of a case (Paris) 1969; 96:129–34. and review of the literature. Arch Dermatol 1986; 122:805–8. 9 Hensley GT, Glynn KP. Hypertrichosis lanuginosa as a sign of 34 Skaf RA, Anthony MM. Acquired hypertrichosis lanuginosa: a case internal malignancy. Cancer 1969; 24:1051–6. report. J Reprod Med 1986; 31:629–32. 10 Chadfield HW, Khan AV. Acquired hypertrichosis lanuginosa. Trans 35 Hovenden AL. Acquired hypertrichosis lanuginosa associated with St Johns Hosp Dermatol Soc 1970; 56:30–4. malignancy. Arch Intern Med 1987; 147:2013–18. 11 Djajadiningrat AP, van der Lugt L, van Gilse HA, Siddre WJ. 36 Dyall-Smith D, Varigos G, Thomas R. Hypertrichosis lanuginosa Acquired hypertrichosis lanuginosa. Ned Tijdschr Geneeskd 1970; acquisita and adenocarcinoma of the colon. Australas J Dermatol 1987; 114:639–41. 28:1–6. 12 Hegedus SI, Schorr WF. Acquired hypertrichosis lanuginosa and 37 Carratala´ J, Ribera M, Martin A, Niubo R. Hipertrichosis lanuginosa malignancy. A clinical review and histopathologic evaluation with adquirida y cancer diseminado. Med Clin (Barc) 1987; 88:564. special attention to the ‘mantle’ hair of Pinkus. Arch Dermatol 1972; 38 Mengori P, Rosales O. Hypertrichosis lanuginosa in a man with 106:84–8. colon adenocarcinoma. Arch Intern Med 1989; 149:471. 13 van der Lugt L, Dudok de Wit C. Hypertrichosis lanuginosa acqui- 39 Rodriguez LV, Velasco JT, Vasconcellos R. Hypertrichose lanugine- sita. Dermatologica 1973; 146:46–54. use acquise parane´oplastique associe´ea` une scle´rodermie. Ann 14 Anderson G. Paramalignant Syndromes in Lung Cancer. London: Heine- Dermatol Venereol 1990; 117:605–10. mann Books, 1973; 79–86. 40 Salazar FM, Perez ES, Albroch JRC et al. Carcinoma indiferenciado 15 Reinhold HM, Schlu¨ter E, Schlaak M. Hypertrichosis lanuginosa bronchopulmonar asociado a esclerodermia e hipertricosis lanugin- acquisita in colon carcinoma. Inn Med 1974; 4:281–4. osa. An Med Intern (Madrid) 1990; 7:258–60. 16 Samson MK, Buroker TR, Henderson MD et al. Acquired hypertri- 41 De Clercq D, Iriarte Ortabe JI, Reychler M. Manifestations bucco- chosis lanuginosa. Report of two new cases and a review of the faciales de l’hypertrichosis lanuginosa. Rev Stomatol Chir Maxillofac literature. Cancer 1975; 36:1519–21. 1990; 91:465–8. 17 Rzempoluch E, Szczurek Z, Osiecki Z. Hypertrichosis lanuginosa 42 McKenna KE, Hayes D, McMillan JC. Subacute cutaneous lupus ery- acquisita in the course of bronchial cancer. Przegl Dermatol 1976; thematosus-like gyrate erythema and hypertrichosis lanuginosa 63:199–203. acquisita associated with uterine adenocarcinoma. Br J Dermatol 18 Kaiser IH, Perry G, Yoonessi M. Acquired hypertrichosis lanuginosa 1992; 127:443–4. associated with endometrial malignancy. Obstet Gynecol 1976; 43 Begany A, Nagy-Vezekenyi K. Hypertrichosis lanuginosa acquisita. 47:479–82. Acta Derm Venereol (Stockh) 1992; 72:18–19. 19 Wadskow S, Bro-Jørgensen A, Søndergaard J. Acquired hypertri- 44 Brinkmann J, Breier B, Goos M. [Hypertrichosis lanuginosa chosis lanuginosa. A skin marker of internal malignancy. Arch acquisita in ulcerative colitis with colon cancer]. Hautarzt 1992; Dermatol 1976; 112:1442–4. 43:714–16. 20 McLean DI, Macaulay JC. Hypertrichosis lanuginosa acquisita asso- 45 Duncan LE, Hemming JD. Renal cell carcinoma of the kidney and ciated with pancreatic carcinoma. Br J Dermatol 1977; 96:313–16. hypertrichosis lanuginosa acquisita. Br J Urol 1994; 74:678–9. 21 Ikeya T, Izumi A, Suzuki M. Acquired hypertrichosis lanuginosa. 46 Toyoki Y, Satoh S, Morioka G et al. Rectal cancer associated with Dermatologica 1978; 156:274–82. acquired hypertrichosis lanuginosa as a possible cutaneous marker 22 Davies RA, Newman DM, Phillips MJ et al. Acquired hypertrichosis of internal malignancy. J Gastroenterol 1998; 33:575–7. lanuginosa as a sign of internal malignant disease. Can Med Assoc J 47 Farin˜a MC, Tarı´n N, Grilli R et al. Acquired hypertrichosis lanugin- 1978; 118:1090, 1095–6. osa: case report and review of the literature. Surg Oncol 1998; 23 Ricken KH. Hypertrichosis lanuginosa et terminalis acquisita and 68:199–203. pseudo-ichthyosis acquisita as paraneoplastic syndrome in chronic 48 Maier S, Arlt W, Wiebecke S et al. Paraneoplastic hypertrichosis lymphatic leukemia (CLL). Z Hautkr 1979; 54:869–72. lanuginosa et terminalis in a 27-year old woman with parotid 24 Goodfellow A, Calvert H, Bohn G. Hypertrichosis lanuginosa acqui- carcinoma. Internist (Berl) 1999; 40:1323–6. sita. Br J Dermatol 1980; 103:431–3. 49 Pe´rez-Losada E, Pujol RM, Domingo P et al. Hypertrichosis lanugin- 25 Gonza´lez JJ, Ungaro PC, Hooper JW, Werk EE Jr. Acquired hyper- osa acquisita preceding extraskeletal Ewing’s sarcoma. Clin Exp trichosis lanuginosa: a rare manifestation of urinary bladder carci- Dermatol 2001; 26:182–3. noma. Arch Intern Med 1980; 140:969–70. 50 Bauer HI, Kaatz M, Elsner P. Circumscribed hypertrichosis lanugin- 26 Shee CD, Graham VA. Acquired hypertrichosis lanuginosa and car- osa in acute myeloid leukaemia. Dtsch Med Wochenschr 2001; cinoma of the bronchus. Thorax 1981; 36:153–4. 126:845–6. 27 Knowling MA, Meakin JW, Hradsky NS, Pringle JF. Hypertrichosis 51 Sanchez-Estella J, Yuste M, Santos JC, Alonso MT. [Acquired para- lanuginosa acquisita associated with adenocarcinoma of the lung. neoplastic hypertrichosis lanuginosa]. Actas Dermosifiliogr 2005; Can Med Assoc J 1982; 126:1308–10. 96:459–61.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 1092 Paraneoplastic hypertrichosis lanuginosa acquisita, P.H.T.J. Slee et al.

52 Lorette G, Maruani A. Images in clinical medicine. Acquired hyper- 58 Fuchs E. Beauty is skin deep: the fascinating biology of the epider- trichosis lanuginosa. N Engl J Med 2006; 354:2696. mis and its appendages. Harvey Lect 1998–99; 94:47–77. 53 Pruijm MC, van Houtum WH. An unusual cause of hypertrichosis. 59 Soutter AD, Ngyen M, Watanabe H, Folkman J. Basic fibroblast Neth J Med 2007; 65:42, 45. growth factor secreted by an animal tumor is detectable in urine. 54 Saad N, Hot A, Ninet J et al. [Acquired hypertrichosis lanuginosa Cancer Res 1993; 53:5297–9. and gastric adenocarcinoma]. Ann Dermatol Venereol 2007; 134:55– 60 Gat U, DasGupta R, Degenstein L, Fuchs E. De novo hair follicle 8. morphogenesis and hair tumors in mice expressing a truncated 55 Wyatt JP, Anderson HF, Greer KE, Cordoro KM. Acquired hypertri- beta-catenin in skin. Cell 1998; 95:605–14. chosis lanuginosa as a presenting sign of metastatic prostate cancer 61 Kerob D, Dupuy A, Reygagne P et al. Facial hypertrichosis induced with rapid resolution after treatment. J Am Acad Dermatol 2007; by cetuximab, an anti-EGFR monoclonal antibody. Arch Dermatol 56:S45–7. 2006; 142:1656–7. 56 Vulink AJ, ten Bokkel Huinink D. Acquired hypertrichosis lanugin- 62 Du X, Tabeta K, Hoebe K et al. Velvet, a dominant EGFR mutation osa: a rare cutaneous paraneoplastic syndrome. J Clin Oncol 2007; that causes wavy hair and defective eyelid development in mice. 25:1625–6. Genetics 2004; 166:331–40. 57 Cruz PDJ, Hud JAJ. Excess insulin binding to insulin-like growth 63 Slee PHTJ, Verzijlbergen FJ, Schagen van Leeuwen JH, van der factor receptors: proposed mechanism for acanthosis nigricans. Waal RIF. Acquired hypertrichosis: a rare paraneoplastic syndrome J Invest Dermatol 1992; 98:S82–91. in various cancers. J Clin Oncol 2006; 20:523–4.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1087–1092 REVIEW ARTICLE DOI 10.1111/j.1365-2133.2007.08162.x The relationships between exposure dose and response in induction and elicitation of contact hypersensitivity in humans P.S. Friedmann Dermatopharmacology Unit, Southampton General Hospital, Southampton SO16 6YD, U.K.

Summary

Correspondence Like all physiological systems, the human immune system exhibits dose–response Peter S. Friedmann. relationships in its reactions. The strength of sensitization is related to the E-mail: [email protected] potency of the immunogen and the dose that reaches the immune system. In ) skin, as sensitizing dose per unit area (lgcm 2) is increased on a log scale, there Accepted for publication 15 June 2007 is a sigmoid dose–response curve for subsequent reactivity. Similarly, the response to elicitation shows a classical sigmoid response to increasing challenge Key words dose, with the dose per unit area again being the determinant. There is a clear contact sensitization, dose per unit area, inverse correlation between the strength of sensitization and the subsequent dose dose–response, patch test, repeated open of antigen to which an individual will respond. This is reflected in the different application test, sensitizing dose challenge systems used to diagnose the existence of allergic contact sensitization Conflicts of interest to a given allergen. The occluded patch test aims to use the highest concentration None declared. possible to detect the weakest degrees of allergy, whereas the repeated open application test uses much lower concentrations similar to those encountered in real life, applied repeatedly but without occlusion, to assess clinical relevance. Many authors have attempted to use the lowest concentrations to which rare, highly sensitized individuals can react to define the concentrations which might be free of risk in terms of inducing allergic sensitization. However, it is clear that the dose–response relationships for induction of sensitivity by repeated low-dose exposures must be carefully defined in future studies. This article reviews the dose–response relationships of human contact sensitization.

In the field of contact allergy one of the central questions is: hydrates and nucleic acid components. These microbial why does the human immune system develop unwanted constituents are recognized by a diverse set of so-called pat- immune hypersensitivity to environmental chemicals (xenobi- tern recognition receptors which include toll-like receptors otics) that results in apparently valueless skin inflammation? and scavenger receptors. Secondary questions relate to quantitative aspects of the Adaptive responses are designed to establish highly specific immune system and its responses: for example, how low an immunological memory of the relevant antigen ⁄immunogen exposure to an immunogen can the immune system detect but they are slow to be initiated as they involve the prolifera- and respond to? The human immune system functions mainly tion of clones of T lymphocytes specifically able to recognize to protect us from harm by pathogens. In order to fulfil this the antigen. In order for the immune system to ‘invest’ in the function it has to detect and respond to foreign substances complex process of waking up naı¨ve T cells and causing them appropriately – that is, it must generate an effective response to proliferate, activate B cells to make antibody and establish when the foreign substance is perceived as dangerous, but the immunological memory, the immune system must sense that system is programmed not to respond to nondangerous sub- it is important and necessary so to do. In fact, the alerting sig- stances. The immune system mounts two major types of nals that say ‘this is important’ are the danger signals that acti- response – the ‘innate’ response and the ‘adaptive’ or acquired vate the innate responses. The cross-talk between the cellular response. components of the innate and adaptive immune systems is Innate responses are designed primarily to generate rapid crucial in determining whether T cells are activated and the defences to danger signals without identifying the activating nature of the effector mechanisms that are generated. stimulus in great detail. The agents that are perceived as dan- The induction of a new immune response involves the col- gerous are mainly microbes and their proteins, lipids, carbo- lection ⁄uptake by professional antigen-presenting dendritic

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 1093 1094 Dose–response relationships in contact sensitization, P.S. Friedmann cells (DCs) of foreign material which they process and associ- chemical ⁄xenobiotic. These factors have been reviewed in ate with the major histocompatibility molecules, HLA class I detail elsewhere4 and space does not permit further review and II. The DCs ‘present’ the major histocompatibility com- here. Local factors which render subjects susceptible to sensiti- plex (MHC)-associated immunogen to T lymphocytes so that zation include the stratum corneum barrier, xenobiotic metab- the cell(s) bearing the highly specific cognate receptor for the olism in skin, and concomitant stimuli (e.g. exposure to immunogen can recognize it. In order to be activated to pro- ultraviolet radiation); see below. liferate, the T cell must receive additional signals from the DC via costimulatory molecules and cytokines. Sensitizing potency of chemicals The cell types that respond most energetically to ‘danger’ are macrophages and DCs but most cell types can also partici- Having considered the general processes underlying recogni- pate in innate defensive responses. Cells respond by upregulat- tion of foreign, microbe-derived antigens, the question arises ing all the molecules that might be required for defensive as to what is the basis for the immune system recognizing action. However, in the case of the DCs, because they have and responding to small xenobiotic molecules such as contact been alerted to the stimulus being dangerous and hence sensitizers. The T-cell receptor is designed to recognize the important, there is also a maturation and augmented expres- structure of a peptide of nine to 15 amino acids lying in the sion of key immunostimulatory cytokines including inter- groove of the MHC class I or II molecule. In order to be leukin (IL)-1b, IL-12 and type 1 interferons as well as ‘seen’ by the T cell, small molecules must bind to proteins, costimulatory molecules including CD40, CD80 and CD86. hence acting as haptens. Hapten-conjugated peptides can then These highly activated, ‘alerted’ DCs are able to present com- be associated with the MHC groove, making them recogniz- ponents of the perturbing agent ⁄organism to naı¨ve T lympho- able by T cells with the appropriate cognate receptor. Some cytes with a set of very powerful activation signals that initiate chemicals are intrinsically reactive, while others may be proliferation and clonal expansion – resulting in an active rendered reactive through the actions of drug-metabolizing immune response. If the DCs present molecules to T cells and detoxifying enzymes. It is thought that the more without having been alerted to potential danger, various out- protein-reactive a chemical the more potent it is as an comes are possible: the T cells will not proliferate and might immunogen ⁄sensitizer. So some chemicals including dinitro- even undergo apoptosis – resulting in absence of immune chlorobenzene (DNCB), diphenylcyclopropenone, oxazolone reactivity to that ‘antigen’. Alternatively, a different type of and squaric acid dibutyl ester are very potent immunogens, activation may occur in which T cells can act as regulatory T inducing contact sensitization in 100% of normal humans. cells, preventing effector T cells from being generated or from Much work has gone into devising assays that can predict responding to the ‘antigen’, thus establishing a form of active the sensitizing potency of different chemicals. Thus, tests in immunological tolerance. In mice and humans contact hyper- animals include the guinea pig maximization test, the Bueh- sensitivity responses involve activation of T cells of both ler test and the mouse ear swelling test.5 Some of these CD4+ Th1 and CD8+ Tc1 types.1–3 experimental systems involve administering the chemical Once clones of memory ⁄effector T cells have been gener- with an adjuvant which gives the ‘danger’ signal and hence ated they are available to react much more rapidly should maximizes the chance of the immune system responding. their specific target antigen be re-encountered. Importantly, Implicit in this is that the chemical on its own may not be memory cells and cytotoxic T cells can be activated by many able to induce an immune response but requires the ‘danger cell types capable of showing ⁄presenting the antigen in associ- signals’ from the adjuvant. Kligman developed a modified ation with MHC class II or I, respectively. Thus, macrophages human ‘maximization’ test which involved pretreating a skin and B cells which express MHC class II but which do not area with 5% sodium lauryl sulphate (SLS), a detergent with express all the costimulatory molecules or even the key cyto- irritant properties. The agent under test was then applied to kines IL-1b or IL-12 are very good at presenting antigens to the area under occlusion for 48 h.6 The local lymph node memory T cells. assay (LLNA), which involves quantification of proliferation of lymphoid cells in lymph nodes draining sites of topical Factors contributing to sensitizing efficiency application of putative sensitizers, has become widely adopted as a standard mouse-based screening test.7,8 The ultimate determinants of whether a chemical induces sen- Many chemicals found in everyday life are potential sensi- sitization are the resultant of a range of host-related factors tizers; thus Kligman showed 100% sensitization of human that influence susceptibility, which may be systemic or local, volunteers by para-phenylenediamine (PPD) using high the intrinsic sensitizing potency of the chemical and the dose concentrations of this substance (1 g of 10% PPD in petro- ) that reaches the cells of the immune system. leum jelly to give 711 lgcm 2 compared with up to 6% in hair dyes).9 Indeed, in the LLNA it is more potent than DNCB.10 However, the doses in which most chemicals are Susceptibility to sensitization encountered in the normal environment are either too low to Age, sex and genetic factors such as HLA haplotype can all activate a ‘positive’ immune response manifesting as allergic influence whether and to what degree an individual reacts to a hypersensitivity, or they may even induce immunological

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 Dose–response relationships in contact sensitization, P.S. Friedmann 1095 tolerance, probably mediated by regulatory T cells. As indi- healthy human volunteers with different doses of DNCB on an cated above, whether an active hypersensitivity or a state of area of constant size and, in other groups, comparing nonreactivity is induced may be a reflection of how the chem- responses evoked by varying the area of application of the ical interacts with the innate immune system to produce dan- sensitizing dose. Thus, initially, five groups received 62Æ5, ger signals. Most potent sensitizers are also irritants and there 125, 250, 500 or 1000 lg of DNCB on a circle of forearm is evidence that it is the irritant properties that exert the ‘adju- skin 3 cm in diameter.13 As DNCB remains bound in the skin vant’ effect of stimulating the innate immune response, hence for some weeks, it is available for recognition as soon as a potentiating the activation of the adaptive ⁄acquired immune clone of memory T cells specific for DNCB enters the systemic response. For example, Allenby and Basketter showed that pre- circulation from the lymph node. The site of initial application treatment with an irritant, SLS, potentiated allergic responses of DNCB subsequently becomes inflamed with a contact to nickel.11 However, SLS is also a surfactant and the possibil- dermatitis-like reaction – the so-called delayed flare. This is an ity that it simply compromised the epidermal barrier and facil- indicator not only that the individual is now clinically sensi- itated absorption of nickel is a possible alternative explanation. tized, but the time at which it occurs is a reflection of the This ‘danger signal’ effect resulting from the irritant properties time taken for the clonal expansion of specific T cells that of sensitizers is a major area of interest but to date the rele- occurs during the generation of clinical sensitization. The pro- vant mechanisms are obscure. portion of individuals sensitized by increasing doses of DNCB showed a sigmoid log-dose response curve, with 100% being sensitized by 500 lg DNCB and above. There was an inverse Factors determining the dose that the immune system relationship between the sensitizing dose and the time taken encounters for the delayed flare to appear – thus it took a mean of Many factors affect the dose (number of molecules) of a 15Æ5 days to appear for those who received 62Æ5 or 125 lg chemical that reaches the viable layers of the epidermis in but decreased for each group as the sensitizing dose increased, which the DCs are found. These include the concentration so that those who received 1000 lg showed the delayed flare (quantity per unit area) applied to the surface, the duration after only 8 days.14 for which it is applied, the molecular weight of the com- Four weeks after sensitization the degree of reactivity was pound (those larger than 500 Da do not penetrate into normal assessed by quantifying responses to challenge with four skin), the intactness of the stratum corneum barrier, occlu- simultaneous eliciting doses: 3Æ125, 6Æ25, 12Æ5 and 25 lgof sion,12 the solubility characteristics (lipid-soluble molecules DCNB. These challenges were applied to the other forearm penetrate much better than water-soluble ones), the vehicle on 1-cm paper discs (Al test patches; Epitest, Hyryla, and the partition coefficient of the molecule in relation to the Finland). The patches were left in place for 6 h but the constituents of the vehicle and the stratum corneum. Addi- responses were assessed at 48 h both clinically (grade 2 was tional factors may include metabolic processing of the com- an erythematous and palpably indurated response) and by pound which may either reduce its effective concentration or, thickness which was quantified by use of Harpenden skinfold alternatively, might generate an intermediate compound which callipers. The proportion of individuals in each sensitization is the actual sensitizer. These factors will all contribute at group that gave grade 2 or stronger (i.e. definitely sensitized) every encounter with a chemical – both during the initial sen- responses followed the same sigmoid curve with log-sensitiz- sitizing exposure and also at subsequent elicitation ⁄challenge ing dose as was seen for the delayed flare. Thus the propor- exposure(s). Indeed, the diagnostic occluded patch test chal- tions sensitized rose from 8% by 62Æ5 lg through 62% by lenge deliberately maximizes the exposure dose by using high 125 lg, 83% by 250 lg to 100% by 500 lg and above. This concentrations applied for 48 h under occlusion. Additional reveals that there is a normal distribution to the human factors may come into play when there is repeated exposure immune response – some people being low ⁄weak respond- to a compound either through daily use or in a repeated chal- ers, others being high ⁄strong responders. Within each sensiti- lenge. These will be discussed below. zation group the normal distribution was evident – with some people being weakly sensitized or apparently unsensi- Dose–response relationships for induction tized, others being moderately reactive and yet others being of sensitization strongly reactive. The responses (increase in skinfold thick- ness) plotted against log-challenge dose formed the lower or Use of potent contact sensitizers such as DNCB has taught us central portions of sigmoid dose–response curves. When the much about the quantitative dose–response relationships of linear segments of the family of challenge dose–response the human immune system. Apart from the chemical property curves were analysed by a generalized linear interactive of actual or potential reactivity with proteins, it is now clear model, the slopes of the curves were parallel13 (Fig. 1). This that the major determinant of how the immune system revealed that whether individuals were strongly or weakly responds to a sensitizer is the intensity of the sensitization sensitized, increasing the challenge dose by a given increment stimulus, that is the dose of sensitizer in relation to the area (doubling) was associated with a constant factor of augmen- of exposure – the dose per unit area. The experimental work tation of the response – the slope of the dose–response that underpins this conclusion involved sensitizing groups of curve. By extrapolation of the regression lines it could be

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 1096 Dose–response relationships in contact sensitization, P.S. Friedmann

Dose-responses to elicitation challenge in Increase in responsiveness to DNCB 5 groups of normal volunteers as sensitising dose increases 3·0 2·5

2·0 2·5 1·5

1·0 2·0 0·5 Skinfold thickness (mm)

0 1·5 0·812 1·625 3·125 6·25 12·5 Response DNCB challenge dose (µg) 1·0 Fig 1. Five groups of healthy volunteers received initial sensitizing (increase in thickness, mm) doses of 1000 (circles), 500 (diamonds), 250 (inverted triangles), 125 (triangles) or 62Æ5 lg (squares) of dinitrochlorobenzene (DNCB). Four weeks later they were challenged with four doses of DNCB 0·5 (from 3Æ125 to 25 lg) and responses measured as thickness with skinfold callipers. The regression lines of the linear portion of the dose–response curve are plotted as solid lines. The calculated 0·0 extrapolation to y = 0 (theoretical threshold elicitation dose) is plotted as dotted lines. 62·5 125 250 500 1000 Sensitising does of DNCB (µg) calculated ⁄predicted that the lowest dose (threshold) of Fig 2. Increase in responsiveness to dinitrochlorobenzene (DNCB) as DNCB to which each of the groups might respond would be sensitizing dose increases. Reactivity for each of the five groups 3Æ9 lg for the group that received the lowest sensitizing dose shown as response at challenge with 12Æ5 mg (measured as and 3Æ0, 2Æ0, 1Æ4 and 0Æ8 lg, respectively, for the other skinfold thickness) plotted against initial sensitizing dose. Results are groups in increasing order of sensitizing dose. These absolute shown as mean ± SEM. Reproduced from Friedmann14 with doses applied on 1-cm diameter paper discs convert to 5, permission. ) 3Æ8, 2Æ5, 1Æ8 and 1 lgcm 2, respectively. As mentioned above, one of the important variables deter- mining elicitation responses is the duration of exposure – Dose per unit area these DNCB challenge doses were applied only for 6 h, but in pilot work it was established that longer application periods The above set of dose–response relationships was the result of were associated with much stronger responses. So it would be applying varying doses of sensitizer to a constant area of skin – wrong to think that this represents an absolute lowest (thresh- 3 cm diameter = 7Æ1cm2. Therefore the sensitizing doses can ) old) dose which can detect the presence of sensitization. also be expressed as the dose per unit area – lgcm 2 Finally, it was possible to represent the relationship between (Table 1). We next examined the effects of varying the area increasing sensitizing dose and degree of reactivity (strength of the sensitizing application while maintaining a constant of sensitization) by plotting on a log scale the sensitizing dose dose per unit area. Over a wide range of areas – from a circle against the elicitation response for each group, represented as 4Æ25 cm in diameter down to 1 cm – if the dose per unit area the skinfold thickness at the 12Æ5 lg challenge dose. This were kept constant there was no effect of area, even though showed a log-linear increase in sensitization with increasing the total doses varied by up to 10 fold (Table 1). It was only sensitizing dose (Fig. 2). The same dose–response relation- when the area was reduced to 3 mm (by application of a ships were examined in individuals who spontaneously devel- 3-mm diameter paper disc impregnated with DNCB at the oped three or more unrelated contact allergies (detected in the desired concentration) that area had an effect and only 26% patch test clinic). These individuals were shown to be ‘high were sensitized compared with 93% by the same dose per unit responders’ in that, as a group, they were sensitized to a area on a 1-cm paper disc. Overall, these results can be inter- much greater degree by a given dose of DNCB.15 This was preted in relation to Langerhans cell (LC) numbers. As the ) shown to be a quantitative rather than a qualitative difference, mean density of LCs in forearm skin is about 750 mm 2,an namely, the high-responder end of the normal distribution, as area 1-cm diameter (0Æ78 cm2) contains about 59 000 LCs.16 individuals with a single contact sensitivity were intermediate After application of DNCB up to 20% of LCs migrate down in their sensitizability. into the dermis17 – say 6000–12 000 LCs. This number of

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 Dose–response relationships in contact sensitization, P.S. Friedmann 1097

Table 1 Relationship between proportions sensitized, area of application, and sensitizing Application site Sensitizing dose dose of dinitrochlorobenzene (DNCB) Diameter Area Total Concentration Number of Percentage ) Row (cm) (cm2) (lg) (lgcm 2) subjects sensitized 13 7Æ1 1000 142 24 100 23 7Æ1 500 71 40 100 33 7Æ1 250 35Æ43083 43 7Æ1 125 17Æ73063 53 7Æ162Æ58Æ8248 61Æ51Æ862Æ535Æ47 86 72Æ13Æ55816Æ42255 83 7Æ1 116 16Æ43450 94Æ25 14Æ2 232 16Æ41566 10 1 cm paper 0Æ830382893 11 3 mm paper 0Æ08 3 38 15 26

Data are from several studies with DNCB.13,21,23,51 The first five rows are the normal sub- jects from Figures 1 and 2. Row 6 gives the same dose per unit area as row 3. In rows 7–9 the areas were half and double the standard 7Æ1cm2 but the dose per unit area was constant. Reproduced from Friedmann52 with permission.

LCs is sufficient to sensitize a human fully. Also, for a given Factors affecting the responses to elicitation number of molecules of DNCB, the potency of sensitization is challenge greater if few LCs present many molecules per cell rather than having the same number of molecules presented by many Factors affecting the response to elicitation challenges include more LCs at fewer molecules per cell. Similar observations the strength of sensitization, whether the immune response to were made by Macatonia et al. in mice and the relationship this antigen has stabilized or is still maturing, the effective between sensitizing dose and LCs reaching the lymph node dose received by the immune system, local tissue factors such defined in quantitative terms.18 as body site and other chemical perturbing influences includ- There have not been many other studies attempting to ing irritancy and whether the challenge is encountered as a quantify induction of contact sensitization in humans. Cardin single or repeated contact. et al. examined the sensitizing potency of Kathon [methylchlo- roisothiazolinone ⁄methylisothiazolinone (MCI ⁄MI)], applying Strength of sensitization the compound to normal volunteers in a range of concentra- tions from 5 to 20 p.p.m. (= 0Æ0005–0Æ002%).19 However, The principle factor that determines responses to elicitation the Kathon was applied on occluded patches at actual doses challenges with contact sensitizers is the strength of sensitiza- ) per unit area of 1Æ14 or 2Æ9 lgcm 2 depending on which tion – whether the individual is sensitized to a weak or a patch system was used. Moreover, repeated patch applications strong degree. This has been dealt with above. were made thrice weekly for nine applications. Although the authors reported the results as ‘20 p.p.m. sensitized 2 ⁄45 vol- ) Boosting effect of repeated challenge unteers’ it is clear that a cumulative dose of 20 lgcm 2 was applied under occlusion over the 3 weeks. In a large American During the early weeks and possibly months following initia- study 495 individuals were patch tested with a mix of tion of sensitivity, repeated challenges with the sensitizer can MCI ⁄MI at 250 p.p.m., which is equivalent to 0Æ025%. The boost the strength of sensitization and hence the reactivity. material was applied on Finn chamber patches (presumably We showed that individuals who received the 50% effective 7 mm) which were occluded for 48 h. While 13 individuals sensitizing dose of DNCB (116 lg applied to a 3-cm diameter were already allergic, three of 495 were sensitized de novo by circle), who failed to give positive responses to the first elici- this concentration.20 Although no indication was given of the tation challenge, gave very strong positive responses when quantity of MCI ⁄MI applied to each patch it is possible to esti- exposed to a second challenge some time later.21 This indi- mate the exposure. The normal quantity of allergen applied to cated that the first challenge had actually boosted the degree a 7-mm Finn chamber is about 5–10 mg of white soft paraf- of sensitization from a subclinical level to a strong degree of fin (WSP) (see below), or 10–15 lL of a solution. Ten micro- clinical allergic sensitivity. However, when people have been litres of a 0Æ025% solution contains 2Æ5 lg of MCI ⁄MI which, sensitized for a long time, as in the case of nickel sensiti- on a 7-mm Finn chamber, gives a dose per unit area of vity, repeated challenges are reproducibly the same and there ) 6Æ4 lgcm 2. is no sign that repeated or continued exposures augment

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 1098 Dose–response relationships in contact sensitization, P.S. Friedmann reactivity.22 The effect of repeated exposures to low concen- the intensity scores of reactions for chambers loaded with trations of potential sensitizers such as those in personal 5 mg or more, although the area of the response for cham- products ⁄cosmetics has yet to be clearly defined, but there is bers loaded with 15 mg was sometimes greater. The probable clearly the possibility that repeated exposures will induce and explanations for these observations are as follows. Firstly, boost sensitization. 2Æ5 mg is not sufficient to cover the chamber so the reaction is smaller. Five and 10 milligrams are contained by the chamber as a ‘column’ of ointment of different thicknesses The dose of antigen that penetrates the epidermis or heights. Fifteen milligrams filled the chambers to the brim This is determined by the same variables as operated for the and even spilled over, producing slightly larger areas of sensitizing dose – the stratum corneum permeability barrier, application than the size of the chambers, but the thick- the vehicle, and the solubility and partition coefficient of the ness ⁄height of the column of ointment may also have been compound. In the same way as for the induction of sensitiza- greater than for the other loading quantities. However, the tion, the dose per unit area is a crucial determinant of crucial consideration is: what is happening at the skin sur- response to elicitation challenge. Thus, when a series of DNCB face? A number of molecules of the antigen will diffuse into doses, as used by the author for the normal dose–response re- the stratum corneum, the number being determined by the lationship studies above, was applied to 1-cm diameter paper concentration in the vehicle and the partition coefficients and discs (Al test) of area 0Æ78 cm2, absolute doses of 3Æ125, relative solubilities of the substance in the vehicle and the 6Æ25, 12Æ5 and 25 lg were actually 3Æ98, 7Æ96, 15Æ91 and stratum corneum. The loss of substance from the vehicle into ) 31Æ83 lgcm 2. This yielded dose-related responses, with all the epidermis will lower the concentration of the substance four challenges eliciting positive responses in strongly sensi- in the vehicle in a narrow zone where it is in contact with tized individuals, and the responses, measured as increase in the skin, but a steady state will be reached leaving much of thickness, showed a linear relationship with log of challenge the substance still in solution in the vehicle. So although 5, dose13,15,23 (Fig. 1). In those studies, the DNCB challenges 10 or 15 mg of a 5% concentration was applied in columns were applied for only 6 h for two reasons: DNCB is a strong of increasing heights, the same number of molecules pene- irritant and if patch tests on DNCB-naı¨ve individuals are left trate into the epidermis from each column – hence the reac- ) on for longer, then doses of 15Æ9 lgcm 2 and above induce tions of equal intensity. Whereas, if all the nickel in 5, 10 or irritant responses which are erythematous but lack substantial 15 mg had been deposited on to the skin surface in a volatile oedema (thickness). Secondly, longer-duration applications solvent that left all the solute on the surface, then the effect evoked significantly stronger responses which, at those con- would have been very different and would have given a centrations, may blister and from which it is impossible to dose–response curve. The above observation is of critical obtain meaningful calliper readings. There are data showing importance in the patch test clinic as it shows that as long as that as duration of application of elicitation challenges an adequate quantity of the ointment containing the antigen increases, there is a reciprocal drop in the concentrations is applied, reproducible results will be obtained. The corol- required to elicit positive responses. This was clearly shown lary is that workers attempting to demonstrate dose–response for PPD.24,25 Similarly, it has been shown for nickel that 48-h relationships should not apply the compound under investi- application of elicitation challenges evokes a higher frequency gation in a nonvolatile vehicle as, although dose–response of positive responses than 24-h application.26 This observation curves can be obtained, the actual amounts delivered to the reflects the fact that longer periods of application result skin surface are uncertain. in greater effective doses penetrating through the stratum corneum. Body site The crucial point about the above is that the DNCB was applied in an acetone solution which, when applied directly There are clear variations in reactivity at different body sites. on to skin, evaporates and deposits all the DNCB on to the Thus forearm skin is significantly less responsive to inflamma- skin, or, when applied to a paper disc, the relative lipid solu- tory challenges including contact hypersensitivity elicitation bility of DNCB ensures that it partitions out of the paper and with nickel (Fig. 3)22 or ultraviolet-induced erythema.27,28 into the stratum corneum. However, when DNCB or another Also, measurement of irritant responses induced by dithranol lipid-soluble molecule is applied in a semisolid base such as showed significant site variations between the medial and lat- WSP, the position is quite different. This was examined by eral sides of the forearm.29 When body sites were compared applying different total quantities of nickel sulphate in WSP in repeated open application tests (ROATs), the lower arm on 7-mm Finn chambers in nickel-allergic individuals.22 was less responsive than the upper arm, while back skin was Thus, replicate chambers loaded with 2Æ5, 5, 10 and 15 mg most reactive.30 Importantly, recent exposure to ultraviolet of ointment containing 5% nickel sulphate were applied to radiation or the use of topical immunosuppressive agents on the back in duplicate and the positive responses assessed at any body site can reduce responses at that site. These points 48 h. Responses at chambers loaded with 2Æ5 mg were smal- all mean that performance of challenges on the forearm are ler and usually one grade weaker than those for chambers less likely to yield positive results compared with other, more loaded with 5 mg. However, there were no differences in reactive sites.

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 Dose–response relationships in contact sensitization, P.S. Friedmann 1099

potentiation by irritancy. One hapten was applied at subelicit- Differences in contact sensitivity ing doses in the presence of irritant concentrations of a differ- 0·125 responses on forearm and back ent hapten to which the mice had not been sensitized,34 thus allowing separation of irritant effects from possible detergent effects. 0·100 Repeated applications

Normal patch tests are designed to detect low degrees of sensi- 0·075 tization – i.e. not to miss the presence of weak allergy. They are a highly artificial exposure situation using the maximal nonirritant concentration of a given allergen applied under occlusion, usually for 48 h. In personal products the concen- 0·050 trations of constituents such as preservatives and fragrances are Erythema index often well below the concentrations used in standard patch tests. Therefore, in efforts to determine whether the actual real-life exposure concentrations are of clinical significance in 0·025 elicitation of product-related allergic contact dermatitis, researchers have developed challenges designed to mimic real- life use of personal products by repeated application of the product to unoccluded skin – the so-called ROAT. Unfortu- 0·000 nately this area of study is confused by two different agendas. The first deals with efforts to determine how best to establish 0·07 0·31 1·25 5 the clinical relevance of allergic sensitivity detected by patch Nickel concentration (%) tests. The second and inter-related issue is the question of how strongly individuals are sensitized and what are the lowest con- Fig 3. Differences in contact sensitivity responses on forearm and centrations of sensitizers to which they can react. The second back. Erythema responses as measured with an erythema meter in agenda attempts to use the lowest levels that highly sensitized response to challenge with increasing concentrations of nickel individuals can react to – the threshold concentrations for elici- sulphate applied 48 h before. Squares = responses on back; tation – to define the safe concentrations for common expos- triangles = responses on forearm. Results are shown as mean ± SEM. Reproduced from Memon and Friedmann22 with permission. ure, presuming that this extrapolates to sensitizing potency. The ROAT has mainly been used to examine reactivity to fra- grances,35–39 preservatives40–44 and metals,11,45,46 although there are reports using ROAT to examine reactivity to other Augmentation of tissue responses either through irritant allergens.47–49 Because quite varying techniques and approaches effects or by the presence of other sensitizers have been taken by different investigators, the present review In proportion to their potency as irritants or sensitizers, xeno- will attempt to distil the main points that have emerged. Thus, biotics applied to the skin activate innate immune responses individuals are usually selected to receive ROATs on the basis with induction of a wide range of cytokines and adhesion that they have already given positive responses to standard molecules and activation of DCs, all important for maximizing patch test challenge. The agent to be tested is made up in vari- the efficiency of immune surveillance mechanisms.31 This tis- ous vehicles including ethanol, water or cream, or the actual sue response to chemical perturbation primes the tissues so product itself is applied. Applications are normally made twice they can respond to lower concentrations of antigen. Thus, daily to areas of skin between 5 · 5 and 10 · 10 cm, usually Shuster et al. observed that when a concentration of a contact on the volar forearm or antecubital fossa, for 7–14 days. As sensitizer too low to elicit a clinical response was applied to mentioned above, the volar forearm – the most common site of sensitized recipients in combination with a subirritant the ROAT application – is the least sensitive site! Application is concentration of an irritant such as SLS, positive patch test usually discontinued once positive reactions develop. In some responses could be evoked;32 as mentioned above, SLS is not studies, if a given concentration fails to elicit a positive only an irritant but also a detergent and so will be likely to response after 7–14 days the concentration is increased and so alter the absorption of the sensitizer. Similarly, if two allergens on. The overall agreements are that strongly sensitized individ- were mixed at concentrations below their elicitation thresh- uals react more quickly to lower concentrations of the relevant olds and applied to individuals allergic to both allergens, the allergen – all of which is completely predictable from the mixture elicited positive responses.33 This has particular rele- experimental observations on the dose–response relationships vance to challenges performed with patients’ own products, of the allergic contact hypersensitivity response.50 either in patch tests or in ROATs (see below). Grabbe et al. When expressed as the concentration or percentage that performed similar experiments in mice, and confirmed the elicits a positive ROAT or patch test, the literature suggests

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 1100 Dose–response relationships in contact sensitization, P.S. Friedmann

Table 2 The quantities of contact sensitizer used in repeated open application tests (ROATs) expressed as concentrations and doses per unit area

Dose per unit ROAT Amount area per Total dose conc Conc applied Area application No. of per unit area % positive ) ) Reference Allergen (%) (p.p.m.) (lL) (cm2) (lgcm 2) applications (lgcm 2) responders Johansen Lyral 3 30 000 50 9 166Æ7 28 4667 89 et al.38 0Æ5 5000 50 9 27Æ8 28 778 61 Farm48 Colophony 20 200 000 5 3Æ142 318 9 2863 77 1 10 000 5 3Æ142 15Æ9 9 143 31 0Æ1 1000 5 3Æ142 1Æ69 14Æ38 Schnuch MDBGN 0Æ025 250 500 25 5 28 140 62 et al.43 0Æ01 100 500 25 2 28 56 54 0Æ005 50 500 25 1 28 28 33 Zachariae MCI ⁄MI 0Æ00075 7Æ5509 0Æ047 56 2Æ63 56 et al.44 0Æ0002 2 50 9 0Æ0125 56 0Æ728 Johansen Chloroatranol 0Æ0025 25 50 9 0Æ14 28 3Æ9 100 et al.37 0Æ0005 5 50 9 0Æ03 28 0Æ84 92

Data from five studies in which sufficient quantitative data were given to allow these calculations. Many authors report ROAT concentrations as p.p.m. In this table the concentrations, total doses applied and doses per unit area have been calculated to allow comparisons. MDBGN, methyldibromoglutaronitrile; MCI ⁄MI, methylchloroisothiazolinone ⁄methylisothiazolinone.

that very low concentrations of allergens can elicit positive is an accumulation of the compound in the stratum corneum responses. Thus, for example, when a dilution series of colo- reservoir so the actual ⁄effective concentration ends up signifi- phony from 20% down to 0Æ1% was made in acetone ⁄arachis cantly higher than that applied to the surface. It is often hard oil and used for occluded patch tests and ROATs, four of 13 to extract sufficient data from the published reports, but a colophony-allergic individuals reacted to both challenges recent paper by Zachariae et al. expressed both diagnostic patch down to 1% (Table 2).48 The antibacterial agent methyldi- test and ROATs as dose per unit area and included all the rele- bromoglutaronitrile elicited positive ROATs in 13 of 39 aller- vant data. First, they showed that standard patch tests with ) gic individuals at a concentration of 0Æ005% (50 p.p.m.).43 MCI ⁄MI at 0Æ01% (100 p.p.m.) applied as 3 lgcm 2 elicited Chloroatranol, an allergen found in the natural fragrance positive responses in all 28 allergic patients tested.44 They material ‘oak moss absolute’, was able to elicit positive ROATs then performed ROATs at two different concentrations, 2 and in 12 of 13 oak moss-allergic patients at a concentration of 5 7Æ5 p.p.m. applied for 4 weeks. Seven of 25 reacted to the p.p.m. (0Æ0005%).37 Lyral [3-cyclohexene-1-carboxaldehyde, first test and all these and an additional seven reacted to the 4-(4-hydroxy-4-methyl pentyl)] evoked positive responses in higher dose. Although the daily doses were 0Æ025 and ) two of 17 allergic patients at 29 p.p.m. and in nine of 18 at 0Æ094 lgcm 2 the total doses applied over 4 weeks amounted ) 662 p.p.m.38 However, when the details of the test procedure to 0Æ7 and 2Æ63 lgcm 2 (Table 2). If the total dose applied are used to express these concentrations as doses per unit area accumulates in the stratum corneum ⁄epidermal reservoir, then an interesting picture emerges (Table 2). Johansen et al. have these doses are much nearer the dose applied as a diagnostic calculated doses in relation to area of application. They patch test. Clearly not all of it will be available because of loss showed that applications of 15 lL of allergen at 29 or 662 by shedding of squames. Second, the tissue priming effects p.p.m. to 0Æ8-cm Finn chamber discs gave doses per unit area of chemical perturbation (activation of DCs, upregulation ) of 0Æ9 and 20 lgcm 2, respectively.38 For the fragrance of chemokines and adhesion molecules) resulting in locally ingredient chloroatranol, the same group calculated that the enhanced immune surveillance may facilitate immune most strongly sensitized individuals reacted in ROAT to responses to lower concentrations of an antigen. These tissue ) applications of 0Æ03 lgcm 2.37 The individuals most strongly perturbing effects may be the result of the contact allergen sensitized by DNCB were calculated above to be able to react itself or other co-administered substances such as solvents or ) in occluded patch tests to 1Æ02 lgcm 2 although this was excipients. Also, the chronic stimulus of repeated applications only applied for 6 h. Thus, there is a significant caveat relating in the ROAT may well generate stronger innate signals. to the interpretation of ROATs and patch tests which emerges from the failure of many authors to consider the absolute Assumptions about elicitation thresholds doses applied or the doses per unit area; some examples have and exposure levels been compiled into Table 2 to illustrate the point. When a contact allergen is applied repeatedly to the skin One recurrent theme in the literature is the relationship there may be several factors which might allow lower concen- between the concentrations of contact allergens to which sen- trations to elicit positive responses. First, it may be that there sitized individuals can react (elicitation) and the risks of naı¨ve

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 Dose–response relationships in contact sensitization, P.S. Friedmann 1101 individuals becoming sensitized by exposure to such concen- with either method. (iii) Within any group, the responses trations. It is hard to be precise about the subject because exhibit a normal distribution with individuals who may react there may be significant differences in the dose ⁄concentration much more weakly or more strongly than the majority. of sensitizer required to induce sensitization by a single expo- (iv) The fact that there are individuals who respond to vari- sure compared with that required to sensitize by multiple ous contact sensitizers, mainly fragrance ingredients and pre- exposures. Thus, it is generally clear from experimental sensi- servatives, at the level of 5–10 p.p.m. indicates that they are tization as above, that a potent sensitizer such as DNCB given sensitized to a high degree. There are two possible explana- as a single exposure will sensitize more than 50% of people at tions for this: firstly, that these are exceptional ‘high respond- ) concentrations as low as 16Æ4 lgcm 2. Although a few indi- ers’ who are sensitized by chemicals to which most people viduals sensitized by this dose could respond to challenge do not react; secondly, as the relevant sensitizers can induce ) doses as low as 3Æ125 lg(4lgcm 2), individuals sensi- such strong sensitization when they are normally encountered ) tized by 70 or 124 lgcm 2 could react to challenge with at low levels in personal products, this indicates that they are doses as low as 0Æ8 lg (applied on a 7-mm Finn cham- actually potent sensitizers. Furthermore, because the exposure ) ber = 1Æ02 lgcm 2). These individuals were clinically very concentration is low, it implies that repeated exposures to strongly allergic and these eliciting doses are not very different such low concentrations can still induce high degrees of sen- ) from the 2Æ63 lgcm 2 to which some of the MCI ⁄MI- or fra- sitivity. Therefore future work should concentrate on two grance-allergic individuals appear able to react. The hypothesis main areas: one is defining the dose–response relationships that emerges is that repeated doses of low concentrations may for sensitization by repeated applications of very low doses of sensitize to a very high degree so that sensitized individuals antigen. The other is investigation of the effects of different can show elicitation responses at very low antigen doses in times of application of elicitation challenge under both open the region of a few p.p.m. Apart from the simple consider- and occluded conditions. ation of antigen potency, two other factors mentioned above could play an important role in potentiating the sensitizing References effect of low concentrations of any agent: (i) repeated expo- sures will result in accumulation of the chemical in the stra- 1 Moulon C, Wild D, Dormoy A et al. MHC-dependent and -indepen- tum corneum ⁄epidermal reservoir which would have the same dent activation of human nickel-specific CD8+ cytotoxic T cells effect as a single application at a much higher dose; and from allergic donors. J Invest Dermatol 1998; 111:360–6. 2 Sieben S, Hertl M, Al Masaoudi T et al. Characterization of T cell (ii) the tissue perturbation that follows repeated applications responses to fragrances. Toxicol Appl Pharmacol 2001; 172:172–8. of low levels of chemicals would result in activation of innate 3 Pickard C, Smith AM, Cooper H et al. Investigation of mechanisms immune responses including increased expression of cytokines underlying the T-cell response to the hapten 2,4-dinitrochloroben- and chemokines together with augmented recruitment of cells zene. J Invest Dermatol 2007; 127:630–7. of the immune system. This would very likely have significant 4 Menne´ T, Wilkinson JD. Individual predisposition to contact der- adjuvant effects on the potency of antigen-presenting DCs to matitis. In: Textbook of Contact Dermatitis (Rycroft RJG, Menne´ T, present very low concentrations of the relevant sensitizer. Frosch P, eds), 2nd edn. Berlin: Springer-Verlag, 1995; 123–30. 5 Patrick E, Maibach HI. Predicitive assays: animal and man, and These factors will be inextricably interlinked so distinguishing in vitro and in vivo. In: Textbook of Contact Dermatitis (Rycroft RJG, their contributions would require careful experiments in Menne´ T, Frosch P, eds), 2nd edn. Berlin: Springer-Verlag, 1995; which the reciprocity of particular doses of sensitizer were 705–47. compared with regard to activation of innate immune 6 Kligman AM. The identification of contact allergens by human responses. Thus, for example, one application of 100 lg could assay. 3. The maximization test: a procedure for screening and rat- be compared with 10 applications of 10 lg. These will be fac- ing contact sensitizers. J Invest Dermatol 1966; 47:393–409. tors that operate in real life and not just in artificial or experi- 7 Loveless SE, Ladics GS, Gerberick GF et al. Further evaluation of the local lymph node assay in the final phase of an international col- mental situations. laborative trial. Toxicology 1996; 108:141–52. 8 Gerberick GF, Ryan CA, Kimber I et al. Local lymph node assay: Concluding observations validation assessment for regulatory purposes. Am J Contact Dermat 2000; 11:3–18. The contact hypersensitivity response demonstrates the fol- 9 Kligman AM. The identification of contact allergens by human lowing characteristics of the dose–response relationships of assay. II. Factors influencing the induction and measurement of the human immune system. (i) The magnitude of the allergic contact dermatitis. J Invest Dermatol 1966; 47:375–92. 10 Warbrick EV, Dearman RJ, Lea LJ et al. Local lymph node assay immune response (degree of sensitization) is directly propor- responses to paraphenylenediamine: intra- and inter-laboratory tional to the log of the sensitizing dose, represented by the evaluations. J Appl Toxicol 1999; 19:255–60. dose per unit area. (ii) The corollary is that the stronger 11 Allenby CF, Basketter DA. An arm immersion model of compro- the degree of sensitization, the lower the challenge dose of mised skin (II). Influence on minimal eliciting patch test concen- the sensitizer to which a sensitized individual will react. This trations of nickel. Contact Dermatitis 1993; 28:129–33. is seen both with standard occluded patch test challenges and 12 Menne´ T. Quantitative aspects of nickel dermatitis. Sensitization with ROATs. It is not surprising that there is a general corre- and eliciting threshold concentrations. Sci Total Environ 1994; 148:275–81. lation between the threshold concentrations for elicitation

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 1102 Dose–response relationships in contact sensitization, P.S. Friedmann

13 Friedmann PS, Moss C, Shuster S et al. Quantitative relationships matory effects of haptens determine the concentration-dependent between sensitising dose of DNCB and reactivity in normal elicitation of allergic contact dermatitis. J Clin Invest 1996; subjects. Clin Exp Immunol 1983; 53:709–11. 98:1158–64. 14 Friedmann PS. Graded continuity, or all or none – studies of the 35 Johansen JD, Andersen KE, Menne´ T. Quantitative aspects of isoeuge- human immune response. Clin Exp Dermatol 1991; 16:79–84. nol contact allergy assessed by use and patch tests. Contact Dermatitis 15 Moss C, Friedmann PS, Shuster S et al. Susceptibility and amplifica- 1996; 34:414–18. tion of sensitivity in contact dermatitis. Clin Exp Immunol 1985; 36 Andersen KE, Johansen JD, Bruze M et al. The time–dose–response 61:232–41. relationship for elicitation of contact dermatitis in isoeugenol aller- 16 Ford GP, Friedmann PS, White SI et al. Possible inhibitory mecha- gic individuals. Toxicol Appl Pharmacol 2001; 170:166–71. nisms for contact sensitization by DNCB following UV-B induced 37 Johansen JD, Andersen KE, Svedman C et al. Chloroatranol, an damage to Langerhans cells. Br J Dermatol 1984; 111:701–2. extremely potent allergen hidden in perfumes: a dose–response 17 Cumberbatch M, Clelland K, Dearman RJ et al. Impact of cutaneous elicitation study. Contact Dermatitis 2003; 49:180–4. IL-10 on resident epidermal Langerhans’ cells and the development 38 Johansen JD, Frosch PJ, Svedman C et al. Hydroxyisohexyl 3-cyclo- of polarized immune responses. J Immunol 2005; 175:43–50. hexene carboxaldehyde – known as Lyral: quantitative aspects and 18 Macatonia SE, Edwards AJ, Knight SC. Dendritic cells and the initia- risk assessment of an important fragrance allergen. Contact Dermatitis tion of contact sensitivity to fluorescein isothiocyanate. Immunology 2003; 48:310–16. 1986; 59:509–14. 39 Bruze M, Johansen JD, Andersen KE et al. Deodorants: an experi- 19 Cardin CW, Weaver JE, Bailey PT. Dose–response assessments of mental provocation study with isoeugenol. Contact Dermatitis 2005; Kathon biocide. (II). Threshold prophetic patch testing. Contact 52:260–7. Dermatitis 1986; 15:10–16. 40 Flyvholm MA, Hall BM, Agner T et al. Threshold for occluded 20 Rietschel RL, Nethercott JR, Emmett EA et al. Methylchloroisothia- formaldehyde patch test in formaldehyde-sensitive patients. Rela- zolinone-methylisothiazolinone reactions in patients screened for tionship to repeated open application test with a product contain- vehicle and preservative hypersensitivity. J Am Acad Dermatol 1990; ing formaldehyde releaser. Contact Dermatitis 1997; 36:26–33. 22:734–8. 41 Gruvberger B, Andersen KE, Brandao FM et al. Repeated open appli- 21 Friedmann PS, Rees J, White SI et al. Low-dose exposure to antigen cation test with methyldibromo glutaronitrile, a multicentre study induces sub-clinical sensitization. Clin Exp Immunol 1990; 81:507–9. within the EECDRG. Contact Dermatitis 2005; 52:19–23. 22 Memon AA, Friedmann PS. Studies on the reproducibility of aller- 42 Jensen CD, Johansen JD, Menne´ T, Andersen KE. Methyldibromo gic contact dermatitis. Br J Dermatol 1996; 134:208–14. glutaronitrile contact allergy: effect of single versus repeated daily 23 White SI, Friedmann PS, Moss C et al. The effect of altering area of exposure. Contact Dermatitis 2005; 52:88–92. application and dose per unit area on sensitization by DNCB. Br J 43 Schnuch A, Kelterer D, Bauer A et al. Quantitative patch and Dermatol 1986; 115:663–8. repeated open application testing in methyldibromo glutaronitrile- 24 Hextall JM, Alagaratnam NJ, Glendinning AK et al. Dose–time rela- sensitive patients. Contact Dermatitis 2005; 52:197–206. tionships for elicitation of contact allergy to para-phenylenediamine. 44 Zachariae C, Lerbaek A, McNamee PM et al. An evaluation of Contact Dermatitis 2002; 47:96–9. dose ⁄unit area and time as key factors influencing the elicita- 25 McFadden JP, Wakelin SH, Holloway DB et al. The effect of patch tion capacity of methylchloroisothiazolinone ⁄methylisothiazolinone duration on the elicitation of para-phenylenediamine contact (MCI ⁄MI) in MCI ⁄MI-allergic patients. Contact Dermatitis 2006; allergy. Contact Dermatitis 1998; 39:79–81. 55:160–6. 26 Kalimo K, Lammintausta K. 24 and 48 h allergen exposure in patch 45 Nielsen NH, Menne´ T, Kristiansen J et al. Effects of repeated skin testing. Comparative study with 11 common contact allergens and exposure to low nickel concentrations: a model for allergic contact

NiCl2. Contact Dermatitis 1984; 10:25–9. dermatitis to nickel on the hands. Br J Dermatol 1999; 141:676– 27 Rhodes LE, Friedmann PS. A comparison of the ultraviolet 82. B-induced erythemal response of back and buttock skin. Photoderma- 46 Wahlberg JE, Liden C. Cross-reactivity patterns of cobalt and nickel tol Photoimmunol Photomed 1992; 9:48–51. studied with repeated open applications (ROATS) to the skin of 28 Waterston K, Naysmith L, Rees JL. Physiological variation in the guinea pigs. Am J Contact Dermat 2000; 11:42–8. erythemal response to ultraviolet radiation and photoadaptation. 47 Chang YC, Clarke GF, Maibach HI. The provocative use test (PUT) J Invest Dermatol 2004; 123:958–64. [repeated open application test (ROAT)] in topical corticosteroid 29 Lawrence CM, Howel D, Shuster S. Site variation in anthralin allergic contact dermatitis. Contact Dermatitis 1997; 37:309–11. inflammation on forearm skin. Br J Dermatol 1986; 114:609–13. 48 Farm G. Repeated open application tests (ROAT) in patients allergic 30 Hannuksela M. Sensitivity of various skin sites in the repeated open to colophony – evaluated visually and with bioengineering tech- application test. Am J Contact Dermat 1991; 2:102–4. niques. Acta Derm Venereol (Stockh) 1998; 78:130–5. 31 Friedmann PS, Strickland I, Memon AA et al. Early time course of 49 Krasteva M, Cristaudo A, Hall B et al. Contact sensitivity to hair recruitment of immune surveillance in human skin after chemical dyes can be detected by the consumer open test. Eur J Dermatol provocation. Clin Exp Immunol 1993; 91:351–6. 2002; 12:322–6. 32 McLelland J, Shuster S, Matthews JN. ‘Irritants’ increase the 50 Villarama CD, Maibach HI. Correlations of patch test reactivity and response to an allergen in allergic contact dermatitis. Arch Dermatol the repeated open application test (ROAT) ⁄provocative use test 1991; 127:1016–19. (PUT). Food Chem Toxicol 2004; 42:1719–25. 33 McLelland J, Shuster S. Contact dermatitis with negative patch tests: 51 Rees JL, Friedmann PS, Matthews JN. The influence of area of the additive effect of allergens in combination. Br J Dermatol 1990; application on sensitization by dinitrochlorobenzene. Br J Dermatol 122:623–30. 1990; 122:29–31. 34 Grabbe S, Steinert M, Mahnke K et al. Dissection of antigenic and 52 Friedmann PS. Clinical aspects of allergic contact dermatitis. In: irritative effects of epicutaneously applied haptens in mice. Evi- Immunopharmacology and Immunotoxicology (Dean JH, Luster MI, Munson dence that not the antigenic component but nonspecific proinflam- AE et al., eds), 2nd edn. New York: Raven Press, 1994; 589–616.

2007 The Author Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1093–1102 REVIEW ARTICLE DOI 10.1111/j.1365-2133.2007.08135.x Psoriasis: evolution of pathogenic concepts and new therapies through phases of translational research E. Guttman-Yassky and J.G. Krueger Laboratory for Investigative Dermatology, The Rockefeller University, Box 178, 1230 York Avenue, New York, NY 10021, U.S.A.

Summary

Correspondence Psoriasis is perhaps unique for a disease studied through translational science in Emma Guttman-Yassky. that there is not an accepted animal model, yet many rounds of bidirectional E-mail: [email protected] translation have taken place that have helped to define disease pathogenesis and to advance therapy. In this review, we illustrate the evolution of new pathogenic Accepted for publication 28 May 2007 concepts and the testing of new therapeutic agents through translational research in humans. We present a current view of disease pathogenesis that stems from Key words research in patients and animal models, but with the perspectives (i) that disease dendritic cell, keratinocyte, psoriasis, T cell, models can advance or hinder the overall translational enterprise and (ii) that the translational research research process must be firmly grounded in the pathophysiology of the actual Conflicts of interest human condition. None declared.

Translational research does not have a unified definition, but disease pathogenesis that stems from research in patients and it is a process which uses scientific investigation to advance animal models, but with the perspectives (i) that disease mod- the understanding of human physiology or disease and which els can advance or hinder the overall translational enterprise then seeks to improve health by translating observations of and (ii) that the research process must be firmly grounded in experimental science to new therapies. There is not a fixed the pathophysiology of the actual human condition. process to translational discoveries and it can begin with an interesting clinical observation that leads to scientific investiga- Before translational research: serendipity tion or with a basic laboratory finding that is subsequently and therapeutic development found to be related to human physiology or disease. A current conceptualization of translational research is that it encom- Psoriasis is a disease of many cell types, and recognition of passes bidirectional flow of information from the bedside or abnormal growth ⁄differentiation ⁄structure formed by these clinic to the laboratory and vice versa. cell types led to many pathogenic hypotheses. Hypotheses of A conventional vision of translational research is that a dis- abnormal function of keratinocytes, dermal fibroblasts, vascu- ease is studied for cellular and molecular mechanisms in lar growth ⁄structure and immune cells were all proposed.1–3 model systems and that the testing of new therapeutic However, the first highly effective therapeutics for psoriasis approaches occurs in these models before proceeding to clini- such as ultraviolet (UV) B and psoralen plus UVA (PUVA) cal trials in humans. This approach is often termed bench- were developed though an empirical approach that was not to-bedside clinical research. Psoriasis vulgaris is a disease that based on specific hypotheses of disease pathogenesis or on breaks this stereotypical view, as key studies that led to new mechanistic properties of the therapeutics.4 Subsequently, therapeutics were developed through human interface attempts were made to study how cellular ⁄histological disease research, a process that might be called reverse translation, or features were affected by potent therapeutics, but this bedside-to-bench (and back) clinical research. In this review, approach did little to sort out pathogenic roles for different we will illustrate the evolution of new pathogenic concepts cell types, as no mechanistic specificity or cellular specificity and the testing of new therapeutic agents through translational was known for these agents. research in humans. Psoriasis is perhaps unique for a disease Phototherapy of psoriasis with artificial light sources has a studied through translational science in that there is not an history of more than 75 years, after being introduced in 1925 accepted animal model, yet many rounds of bidirectional by Goeckerman, in combination with topical crude coal tar.4 translation have taken place that have helped to define disease UVB was developed without any understanding of its mecha- pathogenesis and to advance therapy. In this review, we will nism, while much later studies revealed immunosuppressive discuss the evolution of translational science in psoriasis over properties5 and selective induction of apoptosis in T lympho- several stages of evolution. We will present a current view of cytes infiltrating psoriasis plaques.6

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1103 1104 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger

PUVA was introduced in 1974, after it was shown that the retrospect, it was not appreciated that the phenotypes of psori- combination of orally administered 8-methoxypsoralen and atic skin can vary tremendously based on different mouse subsequent exposure to a new high-intensity UVA radiation backgrounds – in fact, later work transplanting psoriatic skin source was a highly effective treatment for psoriasis7 without to other mouse strains totally changed this view of disease any rationale for specific cellular effects and also without a pathogenesis. specificity for psoriasis, as it proved an efficacious therapy for 8 a variety of different skin disorders. Translation phase II: new cell subsets and molecular activation pathways are defined Early translational research: a focus as monoclonal antibody technology emerges on the epidermis and keratinocyte growth ⁄differentiation The production of monoclonal antibodies was first described in 197522 which led to the development of antibodies that Development of a therapy for psoriasis based on a prior could be used to mark specific populations of leucocytes in understanding of cellular pathogenesis could be considered to humans. Bos et al. were the first to apply monoclonal anti- be the start of translational research. In 1951, Gubner bodies for the immunohistochemical detection of leucocyte described a series of patients with psoriasis who were treated subsets in psoriasis lesions. They identified marked infiltration with aminopterin, based on the principle that it targeted of psoriasis lesions by T cells23 and were the first to propose a hypermetabolic epithelial tissues and with clear knowledge potential role of the cellular immune system in disease patho- that epidermal hyperplasia was one of the most obvious histo- genesis. The appearance of psoriatic lesions was shown to cor- pathological features of psoriasis vulgaris.9 Accordingly, metho- relate with the epidermal influx and activation of T cells.24 trexate, a potent inhibitor of the enzyme dihydrofolate The involvement of many other immune-related molecules reductase and related to aminopterin, was developed as an was discovered in this period, including upregulation of HLA- antagonist of keratinocyte hyperplasia and its clinical dosing DR,25,26 and induction of intercellular adhesion molecule-1 schedule was developed with an understanding of altered pro- and E-selectin, that facilitate T-cell trafficking etc.27–30 liferation kinetics of keratinocytes in psoriasis lesions.10–13 The Immunophenotyping of T cells in psoriasis showed that they assumption that methotrexate specifically targeted keratinocyte are mainly activated memory T cells: CD2+, CD3+, CD5+, hyperplasia led to development of other epithelial-targeted CLA+, CD28lo, CD45RO, with a majority expressing activa- therapeutics, e.g. synthetic retinoids and vitamin D derivatives, tion markers HLA-DR, CD25 [interleukin (IL)-2 receptor] and which (somewhat paradoxically) inhibit or stimulate differen- CD27.27 T-cell differentiation was found to be strongly polar- tiation of epidermal keratinocytes.14–17 ized towards the type 1 pathway, with production of the The therapeutic success with methotrexate was the likely cytokines interferon (IFN)-c and tumour necrosis factor trigger for many clinical and laboratory investigations of (TNF)-a.28 All of these observations led to a progressive view altered keratinocyte growth and differentiation in psoriasis that an activated cellular immune system was a common lesions. In this period, the main techniques to study cellular feature of psoriatic lesions and that inappropriate immune abnormalities in psoriasis were conventional pathology activation might be pathogenic (Fig. 1). However, pathogenic approaches with some use of electron microscopy and radio- hypotheses that came from this phase of bedside-to-bench active tracers for DNA synthesis. By detailed examination of translation were limited by the inability to test those concepts cell proliferation kinetics, rapid cell cycling and rapid matura- directly in patients, as the needed test agents were not avail- tion of epidermal keratinocytes were established for epidermal able for human use. Fortunately, the refinement of monoclo- psoriatic cells, in comparison with normal epidermal cells.18 nal antibody technologies and associated molecular biology The concept that epidermal hyperplasia cannot occur without led to development of key reagents for critical testing of vascular proliferation was contributed by Braverman and hypotheses soon thereafter. Sibley.19 The altered pattern of epidermal differentiation in psoriasis, a process similar to regenerative maturation in skin Translation phase III: pathogenic concepts wounds, was clearly shown using immunohistochemistry to begin to be tested and refined by the testing define induced molecular markers associated with alternate 3,20 of targeted therapeutics in patients. Rationally epidermal differentiation. While the regenerative pathway designed therapeutics enter the clinic in skin wounds was clearly reversible upon completion of repair, the regenerative pathway appeared to be fixed in an Bidirectional translational research on psoriasis began in ear- ‘on’ state in psoriasis lesions and many held the view that nest when experiments on disease pathogenesis had pro- psoriatic keratinocytes were intrinsically growth dysregulated. gressed far enough to propose specific molecular or cellular In fact, this view was supported by experiments in which alterations as disease causes and when experimental therapeu- involved and uninvolved psoriatic skin was transplanted to tics to those pathways were available for testing in the clinic. nude mice.21 While transplants of lesional skin became less By the early 1980s there were two main proposals of disease acanthotic, both nonlesional and lesional psoriatic skin became pathogenesis, one linked to aberrant activation of keratinoctyes hyperplastic in weeks following the initial transplantation. In (as a primary defect) and another to the pathogenic role of

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1105

Fig 1. An early model of immune-regulating molecules expressed in psoriasis lesions and the keratinocyte response. DC, dendritic cell; TCR, T-cell receptor; LFA-1, lymphocyte function-associated antigen-1; IFN, interferon; TNF, tumour necrosis factor; KC, keratinocyte; ICAM-1, intercellular adhesion molecule-1; EGF-R, epidermal growth factor receptor; KGF-R, keratinocyte growth factor receptor; IGF-1R, insulin-like growth factor-1 receptor; IL, interleukin. cellular immunity. Kragballe and Voorhees made the observa- pathology was suppressed, but not reversed, by ciclosporin tion that arachidonic acid derivatives played a role in epi- treatment. Essentially two pathogenic models were consistent dermal growth regulation and were highly upregulated in with the summed therapeutic, histological and other scientific psoriasis lesions, and thought them to be a property of psori- studies of psoriasis at this time (Fig. 2). In one model, a pri- atic keratinocytes.31 This led to several studies that tested topi- mary growth defect in keratinocytes led to regenerative activa- cally applied ⁄oral 5-lipoxigenase inhibitors,32 which modestly tion and, as activated keratinocytes were shown to make many reduced scaling and erythema in lesions. While overall activity immune-activating cytokines,41 the cellular immune infiltra- was not sufficient to proceed with clinical development,33 tion ⁄activation in psoriasis lesions could stem from the kerati- these studies are probably the first large-scale exercise of nocyte defect. In a second model, activated leucocytes could bidirectional translational research in psoriasis. come to infiltrate focal skin regions, and induce reactive, In a similar vein, studies began to test the role of activated regenerative changes in epidermal keratinocytes. However, cellular immunity in psoriasis using therapeutic agents that direct links between immune-derived cytokines or cellular were being developed for preventing immune rejection of products and keratinocyte growth activation were only begin- transplanted organs in humans. In an early study from 1989, ning to be elucidated, with IL-142–44 and IL-645,46 identified a patient with long-standing psoriasis was shown to have sev- as potential mitogens for keratinocytes, but with IFNs47 and eral disease-free months after a short treatment with a CD3 TNF48 identified as growth-suppressive molecules. monoclonal antibody termed clone OKT3 and subsequently Clearly work in this phase of translation is beginning to given the name muromonab-CD3.34 In the early 1990s, a few deepen the understanding of psoriasis by coupling targeted case reports suggested the therapeutic benefit of chimeric CD4 therapeutics with histological and biochemical studies of pso- monoclonal antibodies35,36 on psoriasis, with histology show- riasis tissue taken from patients undergoing experimental ther- ing loss of T cells from the inflammatory infiltrate. However, apeutics. Hence there is bidirectionality to the research being full disease reversal and the overall extent of immune involve- conducted. Two key findings from emerging therapeutic stud- ment were inconclusive. ies came to support the view that immune cells are the main The strongest arguments for psoriasis as a T cell-mediated inducers of the epidermal reaction in psoriasis. Firstly, it was autoimmune disease came from experimental treatment with shown that regenerative epidermal activation in psoriasis calcineurin antagonists such as ciclosporin37,38 and tacrolimus plaques, as defined by specific molecular markers, could be fully (or FK506),39 where tremendous improvements in the clinical reversed, and homeostatic growth restored, when immune appearance of psoriasis were evident. While these data were infiltrates were ablated by effective therapies such as UVB and interpreted as proving a role for activated cellular immunity in PUVA.49 Thus, like the observations of Mansbridge and Knapp psoriasis, two serious caveats to this view were discovered: that cutaneous wounds ceased regenerative differentiation (i) ciclosporin at physiological concentrations attained in the upon healing,20 psoriasis could now be viewed as a reversible skin had direct antiproliferative effects on keratinocytes40 and condition of altered epidermal activation, where possibly infil- thus it was not a pure immune antagonist, and (ii) while trating immune cells contributed to the phenotype in a signifi- histological features of psoriasis improved and immune infil- cant way. Because the therapies that established this principle trates were decreased, clinically resolved lesions continued to were not specifically targeted to one cell type or molecular show regenerative (keratin 16+) keratinocytes. Hence, one pathway, one could only establish correlations, but not causal- could still make the argument that fundamental disease ity, between disease activity and infiltrating leucocytes and

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1106 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger

excess keratinocyte proliferation and regenerative epidermal (I) growth.50 Hence, this was the first clear proof that regenera- tive activation of keratinocytes was triggered by activated T lymphocytes. The T-cell pathogenesis hypothesis was extended by two types of translational experiments. In one approach, which is discussed more extensively in a section on skin transplantation models below, nonlesional psoriatic skin was grafted to SCID mice and shown to maintain a normal appearance (unlike in earlier studies on nude mice). Injection of activated mono- nuclear leucocytes into the grafts caused conversion to active psoriasis plaques. Secondly, testing of targeted immune antago- nists in patients with psoriasis continued with antibodies and fusion proteins that blocked key cell surface molecules involved in T-cell activation. It is this series of clinical experiments, which coupled clinical testing of targeted therapeutics with detailed studies of cellular and molecular pathways affected in psoriasis lesions, that begins iterative cycles of bidirectional translational research which progressively expand and refine our understanding of pathogenic immunity in psoriasis. T-cell (II) costimulation is hindered by monoclonal antibodies to CD80 and CD86, as well as by CTLA4-Ig.51,52 CTLA4-Ig is a fusion protein composed of a combination of the extracellular domain of cytotoxic T-lymphocyte-associated antigen-4 (CTLA4) with the IgG heavy-chain sequence creating a soluble protein that binds with high affinity to both CD80 and CD86. Therapy with CTLA4-Ig fusion protein resulted in clinical and pathological disease reversal in patients with psoriasis.53 Therapy with CTLA4-Ig also resulted in a reduction of activated T cells and dendritic cells (DCs) infiltrating the psoriatic plaques.54,55 These observations were highly important, as they not only establish that T-cell stimulation is mandatory for ongoing dis- ease activity, but also imply an important role for proteins expressed on activated DCs as potential future targets in psoria- sis.28 Several studies reported relatively good results in psori- atic patients with the use of monoclonal antibodies against the a chain of the IL-2 receptor (CD25) (basiliximab ⁄ daclizumab).56,57 Daclizumab was reported consistently to block CD25 in peripheral blood and tissue, that correlated with disease improvement. Interestingly, the absolute T-cell counts Fig 2. Two alternative models of psoriasis pathogenesis: a passive were not affected by this therapy.57 immunity model and an active immunity model. DC, dendritic cell; The logical extension of immune targeting of psoriasis led PMN, polymorphonuclear leucocyte. to the development of new biologic therapies. These therapies have only recently been approved (during the last 3 years). associated molecules. Hence, the second key clinical experi- The uniqueness of these therapies is the fact that they target ment was to demonstrate that a highly selective antagonist of specific molecules involved in defined cell activation pathways. activated T cells (one without direct keratinocyte-suppressive Examples of the newly designed biologic therapies are listed effects) could fully reverse the regenerative epidermal pheno- in Table 1. Such biologic therapies include alefacept (LFA- type that defines disease pathology in psoriasis. This result was 3-TIP, Amevive, Biogen), efalizumab (anti-CD11a, Raptiva, reported in 1995 using the fusion toxin DAB389IL-2, a protein Genentec, Xoma, Serono), etanercept (Enbrel , Amgen, which acts only on cells expressing functional IL-2 receptors, Wyeth), infliximab (Remicade, Centocor) and adalimumab 58–60 and causes those cells to undergo apoptosis. DAB389IL-2 selec- (Humira , Abbott). The perception of psoriasis as a tively inhibited the growth of activated T lymphocytes, but T cell-mediated disease led to a revisiting of the thera- not keratinocytes, and it produced a depletion of T cells peutic mechanism of older conventional therapies, includ- in skin lesions of psoriasis, leading to reversal of all defin- ing PUVA, UVB (including narrow-band UVB) and able pathological features of psoriasis lesions, including methotrexate.6,61,62

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1107

Table 1 Examples of newly designed biologic therapies for psoriasis vulgaris

Generic (trade) name Target Status Early agents used for establishing the immune basis of psoriasis Denileukin diftitox ⁄DAB389IL-2 (Ontak ) CD25 Not currently used for psoriasis, approved for cutaneous T-cell lymphoma Abatacept ⁄CTLA4-Ig (Orencia) CD80 and CD86 Investigational and not currently used treatment for psoriasis, approved for rheumatoid arthritis Tacrolimus ⁄FK506 (Prograf) Calcineurin Not currently used for psoriasis, approved for organ transplantation (systemic), atopic dermatitis (topically) Daclizumab (Zenepax) CD25 (antagonist) Investigational and not currently used treatment for psoriasis, approved for organ transplantation Basixilimab (Simulect) CD25 (antagonist) Investigational and not currently used treatment for psoriasis, approved for organ transplantation Approved biologic agents for psoriasis Alefacept (Amevive)a CD2 FDA approved Efalizumab (Raptiva)a,b CD11a (LFA-1) FDA and EMEA approved Infliximab (Remicade)a,b TNF FDA and EMEA approved Etanercept (Enbrel)a,b TNF, lymphotoxin FDA and EMEA approved Drugs under investigation (in human ⁄murine clinical trials) Adalimumab (Humira)a,b TNF FDA approved only for psoriatic arthritis Pimecrolimus Calcineurin In clinical trials Cent-1275a,b IL-12 ⁄23p40 In clinical trials ABT-874a,b IL-12 ⁄23p40 In clinical trials 146B7a,b IL-15 In clinical trials

FDA, Food and Drug Administration; EMEA, European Agency for the Evaluation of Medicinal Products; TNF, tumour necrosis factor; IL, interleukin. aCytokine inhibitors; bmonoclonal antibodies.

Translation phase IV: the interface of science, human genome and by technology-intensive methods that technology and medicine. Fundamental give us the power to detect and discover new molecules that discoveries in human biology and are expressed selectively in psoriasis. Flow cytometry and cell- disease-specific therapeutics sorting methods, as well as genomic methods for study of mRNAs and DNA, have contributed significantly to an Translational research seeks to use the discoveries of basic sci- advanced understanding of pathogenic pathways in psoriasis, ence to advance the treatment of human diseases. Hence, a and these methods now set up new hypotheses which are major component of translational research is integrating tested directly in clinical studies or with human skin in xeno- knowledge of human disease with basic information that is graft models, which are described later. If the eventual goal of gathered in model systems. The course of discovery that led to translational science is to develop personalized therapeutics, introduction of targeted biologics for psoriasis depended heav- we are now progressing towards that goal by attempting to ily on fundamental discoveries in basic science laboratories of develop disease-specific treatments that are based on restricted immunological pathways. Subsequently, antibodies or fusion expression of regulatory immune molecules in psoriasis. We proteins to specific immune molecules proved to be effective will now discuss some immune pathways which were discov- in models of inflammation or organ transplantation and ered in psoriasis by technology-intensive approaches, where ‘humanized’ molecules were created, but few were intended resulting hypotheses were translated backwards into clinical for psoriasis as the main indication. Fortunately, the relevant trials – some successful and some not, but all advancing our pathways were expressed in psoriasis lesions and many, but understanding of pathological immunity. not all, of the intended therapeutics had positive activity in patients with psoriasis. Our currently approved biologic thera- Immune deviation and new T-cell subsets peutics mostly followed this pathway of discovery and imple- mentation, while at the same time giving us some new The idea that immune responses are polarized into Th1 vs. insights into disease biology. Th2 T-cell responses has framed immunological thought for Although we have completed several rounds of bidirectional more than a decade. For example, psoriasis has been classified scientific discovery in psoriasis, we have now entered a new as a Th1-dominant disease, while atopic dermatitis (AD) is stage of translational science where discovery in human sys- more of a Th2-dominant disease.63,64 In turn, such classifica- tems is on par with those in model systems and where the tions are based on the ability to detect cytokine synthesis at unique features of human disease can be discovered. In large the level of individual T cells and on gene expression mea- part this has been made possible by the sequencing of the sures for defining cytokines of Th1 and Th2 T cells.65–69

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1108 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger

Experiments showing strong Th1 polarization of T cells in provided important information about molecular pathways of psoriasis led directly to clinical trials in which IL-4, IL-10 and pathogenic inflammation86 and real-time RT-PCR has provided IL-11 were given to patients with psoriasis in an attempt to a highly sensitive means to quantify inflammatory genes with deviate the immune balance more towards Th2, which is con- therapeutic studies in human skin.74 Many DC- and T cell- sidered to be a relative inhibitor of the Th1 response. Indeed, related genes, including cytokines and chemokines that several studies showed disease improvements associated with regulate these cell types, have been elucidated and we are use of the deviating cytokines, but long-lasting alterations beginning to understand that disease-related genes are regu- were not produced by these approaches.70–72 In the context of lated by ‘master’ transcription factors such as STAT1, STAT3 these studies, upregulation of IL-12 receptors and one com- and nuclear factor jB that control large sets of inflammatory ponent of the cytokine IL-12 was discovered in psoriasis genes. In turn, IFNs, TNF and IL-20 family cytokines have lesions. IL-12 is a cytokine that promotes Th1 T-cell differen- been identified as likely upstream activators of the implicated tiation and production of IFN-c, so logically inhibitors were transcription factors.69,77,87,88 Future therapeutics based on developed and tested to the p40 subunit of IL-12 with consid- inhibitory microRNAs (small interfering RNAs) or decoy erable success.73,74 oligonucleotides could be used directly to target these trans- Emerging data from humans must be integrated in a rapid cription factors or downstream gene products. fashion with new information emerging from basic science studies for translational science to succeed. As one example, a Identification of inflammatory dendritic cells new group of T cells termed Th17 and characterized by the production of IL-17 and IL-22 is central to the pathogenesis DCs are the bridge between innate and adaptive immunity, of autoimmune inflammatory diseases in animal models. In and DCs have a critical role in presenting antigens to T cells turn, IL-23 is a newly discovered cytokine which is related to and thus converting naı¨ve T cells to skin-homing mem- IL-12 through sharing of a common p40 subunit, and which ory ⁄effector T cells. Recently, new functions of DCs have been stimulates development of Th17 T cells.73,74 It now appears discovered and cells mediating direct effector (innate) that IL-23 is the dominant ‘p40’ cytokine in psoriasis and evi- immune functions have been termed ‘inflammatory’ DCs dence in model systems points to probable pathogenic activity (IDCs).89 IDCs, which incude myeloid (CD11c+) and plasma- of Th17-derived cytokines.75,76 Hence, the attempt to target cytoid (BDCA-2+) lineages, probably serve to contain some IL-12 in psoriasis may have inadvertently targeted IL-23 and types of infection.89 For example, a new CD11c+ DC termed Th17 T cells as the dominant pathway.77–79 Thus far, elucida- TIP-DC [TNF- and inducible nitric oxide synthase (iNOS)- tion of IL-23 and Th17 cytokines in psoriasis lesions has producing DC] was discovered in Listeria-infected mice and depended critically on genomic methods, which are described appears to be necessary to clear this organism from infected in the next section. spleens.90 Plasmacytoid DCs produce high levels of IFN-a dur- In addition, a distinct subset of regulatory T cells, termed ing viral infection and thus may directly restrain viral replica- ‘Tregs’, has been identified in several animal models of allergy tion and boost antiviral immunity.91 or autoimmunity.80,81 These cells are CD4+, CD25+ and also The identification of iNOS as a highly upregulated gene in express the transcription factor Foxp3, the costimulation psoriasis (initially detected on gene arrays) led to identifica- receptor CTLA4, and neuropilin-1.82,83 CD4+ CD25+ Tregs tion of the protein product within DCs of psoriasis lesions also control autoreactive effector cells and prevent autoimmunity. expressing high levels of TNF. This iNOS+ and TNF+ DC, These cells can suppress T-lymphocyte responses in a non- which appears to be the human equivalent of the TIP-DC, is antigen-specific manner. Sugiyama et al. recently demonstrated greatly increased in psoriasis lesions (in comparison with nor- that in psoriasis CD4+ CD25+ Tregs have reduced suppressor mal skin) and in some cases exceeds the number of infiltrating activity.84 However, there is need to refine definitions of Tregs T cells in psoriasis plaques.92 TNF associated with TIP-DCs in psoriasis further, as CD25+ T cells in humans can be a may be an autocrine or paracrine cytokine for activating iNOS mixed population of true Foxp3+ Tregs and recently activated mRNA transcription in this cell type, as TNF blockade with Foxp3– effectors.85 etanercept rapidly shuts off iNOS production.93 In addition, CD11c+ TIP-DCs synthesize IL-20 and IL-23 in psoriasis Genomics lesions, which gives this cell type the ability to activate inflam- matory gene products in keratinocytes (via IL-20)87 and ⁄or The human genome is now a means for fundamental discov- activate Th17 T cells via IL-23. In model systems, IL-23 ery. Sequence mining has led to discovery of new cytokines, induces profound epidermal hyperplasia94 and effects appear e.g. IL-20, and psoriasis was the first disease in which this to be mediated by IL-22, which is a product of activated Th17 cytokine was implicated as a pathogenic contributor. Gene T cells.94,95 Hence, cytokines produced by TIP-DCs may bridge arrays and real-time reverse transcription–polymerase chain innate and adaptive immune activation in psoriasis lesions. reaction (RT-PCR), both based on genomic information, have Another type of IDC, the plasmacytoid DC, has been proposed enabled large-scale analyses of gene expression in the skin and as a key early trigger of inflammation in developing psoriasis other organs. Global maps of genes expressed in psoriatic lesions through production of IFN-a,96 a cytokine that also lesions, including inflammatory cytokines and chemokines, bridges innate and adaptive immune functions.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1109

Fig 3. Tumour necrosis factor (TNF)-a and inducible nitric oxide synthase (iNOS)- positive dendritic cells (TIP-DCs), plasmacytoid dendritic cells (PDCs) and T cells together with their induced cytokines and chemokines are orchestrating a Th1- and Th17-driven immune response. Interleukin (IL)-20, IL-23, IL-12, IL-17 and IL-22 and innate defence products are major components in the psoriatic plaque. KC, keratinocyte; MMP, matrix metalloproteinase; IFN, interferon.

Another facet of IDCs, which has emerged recently from and then critically tested in human systems. In this way, cur- the study of AD,97 is that these cells may have alternative dif- rent research in psoriasis is a model for molecular medicine ferentiation pathways that stimulate differing T-cell subsets and for the application of scientific principles to the develop- that typify Th1 vs. Th2 (and also Th17) skin diseases. The ment of new therapeutics. Still, there is the strong need for TIP-DC is the major type of myeloid DC in psoriasis, whereas integrating information from model inflammatory diseases this cell type is mostly absent in AD. Instead, myeloid DCs in and from genetic studies of psoriasis to testable clinical AD appear to be polarized to produce a different set of hypotheses, as discussed in the two following sections. chemokines that can direct T-cell recruitiment ⁄activation into the Th2 pathway. Both psoriasis and AD also contain marked Translational models: the attempt to reverse populations of ‘mature’ DCs, that are highly active antigen- engineer psoriasis in model systems presenting cells, so probably conventional T-cell activation via DCs is also ongoing in both diseases. Potentially, myeloid DCs Several transgenic models of cytokine expression in the skin can have both inflammatory and antigen-presenting functions have established that keratinocyte hyperplasia can be induced with differing degrees of differentiation ⁄maturation influenc- by overexpression of single growth factors, e.g. transforming ing the balance of these two functions.97 growth factor-a,98 amphiregulin99 or IL-20,100 while vascular Figures 2 and 3 illustrate some of the new cellular and endothelial growth factor produced by keratinocytes can stim- molecular pathways which have been implicated in psoriasis ulate not only epidermal hyperplasia, but also vascular growth by technology-intensive translational approaches. However, as and leucocyte infiltration in skin. Likewise, several types of a continuing part of the translational process, we want to immune cytokines (IL-12 ⁄IL-23) expressed as transgenes101 or emphasize that this scheme must be considered only as a cur- injected directly into murine skin induce enhanced cutaneous rent hypothesis of pathogenesis and that some parts still need immunity and keratinocyte hyperplasia.74 Experiments like to be established in psoriasis lesions in human skin. Further- these help us to understand the potential functions of individ- more, the functions of individual cells and molecules in ual cytokines that are overexpressed in psoriasis lesions, but at human skin must be established by selective antagonism with the same time there are inherent limitations in the ability of new test agents in clinical trials or xenograft systems. Poten- fixed transgenes to model the process of reversible hyperplasia tially, psoriasis could be effectively treated by selective antago- in which levels of inflammatory cytokines are modulated, nism of TIP-DCs or Th17 T cells (or selective products according to active or suppressed disease. In addition, psoriasis thereof), but this must be tested directly in future studies. is a case where numerous products are simultaneously over- Fortunately, the technology-intensive approaches which help expressed and where interactions with 30 or more upregulat- to establish the new cells and molecular products in psoriasis ed cytokines and chemokines may set up interactive circuits lesions can be used in a powerful way to help dissect effects that are reproduced in single transgene models. Unfortunately, of specific antagonists during proof-of-concept clinical trials we do not know how any model of epidermal hyperplasia and thus make experimentation at the clinical interface an induced by cytokine overexpression reproduces the genomic extension of solid science in which hypotheses are generated signature of psoriasis in which expression of > 1300 genes is

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1110 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger altered. However, potentially the most serious limitation We want to comment on two other recent models of epi- imposed in transgene or knockout murine models is imposed dermal hyperplasia produced by alterations in transcription by the dissimilar structure of murine vs. human skin.77 Virtu- factors. In one model, overexpression of constitutively active ally all models with epidermal hyperplasia in murine skin have STAT3 induces epidermal hyperplasia and immune infiltration been called ‘psoriasis-like’, even though cellular and histo- in the skin, with a dependence of the disease phenotype on logical features do not match changes in psoriasis. Follicular both keratinocytes and T cells.109,110 STAT3 is a transcription hyperplasia, often associated with alopecia, is a nearly univer- factor that is generally overexpressed in psoriasis plaques and, sal feature of these mouse models and often the immune infil- in other recent work, STAT3 activation in keratinocytes is trates are less intense and with a differing mixture of leucocytes clearly linked to IL-20 family cytokines that are upregulated in compared with psoriasis plaques. An excellent and thoughtful psoriasis lesions. Hence, even though keratinocyte hyperplasia review of mouse models has been recently published,102 so we in follicular epithelium is dominant in this mouse model, it will not aim to describe all murine models, but want to has helped to direct research towards disease-related cytokines comment on strengths and weaknesses of modelling using that could regulate STAT3 activation in human skin. several described models as examples. Subsequently we will In another model, epidermal hyperplasia has been induced discuss the study of psoriasis in transplanted skin models. by targeted deletion of the related transcription factors junB With striking changes in skin structure produced in many and c-jun in epidermal keratinocytes.111 Human JunB is a mouse models, some have asked whether the ‘fine’ differences component of the activator protein-1 transcription factor that with clinical or cellular features of psoriasis lesions really mat- is localized in the PSORS6 locus. The mouse model is based ter. The answer depends to some extent on the use of infor- on an observed reduction in JunB in a small number of mation that derives from the models. If one is seeking to patients reported by the model’s authors. This finding, and identify molecular pathways that alter cellular behaviours the model in general, are of questionable relevance to psoria- which are common to many different skin diseases, then the sis, as three other studies with much larger numbers of psori- fine details of resulting skin changes matter less. However, if asis patients show increased expression of JunB in psoriasis one is seeking to understand alterations that are specific to and virtually no change in c-jun levels.112–114 Hence a model one disease process, then the details matter more. Consider based on molecular changes opposite to those occurring in the that psoriasis and AD are believed to be very different clinical human disease could seriously mislead efforts to understand diseases, but recent work has shown profound similarities in key pathophysiology and to develop targeted therapies to rele- pathological epidermal hyperplasia and cellular immune infil- vant molecules. trates in both diseases, even to the extent that histopathology 103 cannot distinguish individual cases with certainty. Patholog- Xenografts as models of psoriasis ical overlaps create the case that some targeted therapies, e.g. calcineurin antagonists and the antibody efalizumab, are effec- These models are unique, in the sense that uninvolved tive in both diseases.104–106 Alternatively, there is also some (or nonlesional) psoriatic skin is transplanted on to severely therapeutic specificity to both diseases and the future direction immunodeficient mice, trying to mimic the genetic, immune of translational ⁄therapeutic research is to seek more specific and pathological characteristics of psoriasis. Initial reports causes of each disease, for example causes of different T-cell, focused on athymic nude mice, that when engrafted with pso- DC or gene expression polarities. Models of epidermal hyper- riatic skin showed a regression of the histological and immu- plasia in mice are often as similar to AD as to psoriasis (with nological features of psoriasis, correlating with the elimination histological features found in neither) and the underlying of the immune infiltrate of human T cells within the psoriatic polarities of cells and inflammatory gene products that really skin.21,115 Boehncke et al.116 and Wrone-Smith and Nickol- distinguish these two human diseases are unknown in the off117 demonstrated that psoriasis could be induced in human models. Hence, from the standpoint of developing disease- skin grafts on SCID mice by injection of superantigen-activated selective treatments, the models are of questionable value, at leucocytes. These SCID models also provided strong evidence least until pathology is known at a molecular level, and with for the importance of T cells in psoriasis. Of note is the fact the additional problem that presently some irreconcilable dif- that SCID mice lack both humoral and cellular immunity, and ferences in immune cell markers within DC lineages exist differ from nude mice in that infiltrating lymphocytes are less between humans and mice. Hence, two recent models which readily eliminated from human grafts, potentially explaining suggest that inflammatory macrophages ‘drive’ psoriasis107,108 the partial maintenance of psoriatic features in transplanted have unknown similarities to the human disease because of lesional skin for weeks and even months. SCID mice were the following: the available markers in the mouse do not have widely used for psoriasis research for the evaluation of the human cognates, the markers have cellular overlaps between efficacy of novel biological agents in preclinical studies.118 macrophages and IDCs, and considerable work is needed to A novel xenotransplantation model was recently des- understand molecular overlaps between macrophages, IDCs cribed.119 This model uses AGR129 mice, that is a triple and conventional DCs in human skin. Still, the models are knockout mouse (deficient in type I and II IFN receptors) and provocative and suggest the need for deeper research in has recombination-activating gene 2 knocked out (rag2).119 human tissues. These mice lack T cells and B cells, and their natural killer

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1111

(NK) cells are severely impaired. When uninvolved skin is potentially have a role in tolerance or inflammation and their engrafted on to AGR129 mice, plaques spontaneously develop, manipulation may increase the risk of autoimmune pheno- without exogenous delivery of CD4+ T cells. The resident mena.126,128 Another psoriasis-susceptibility locus identified T cells (especially CD8+ T cells) undergo extensive prolife- within PSORS2 encodes RAPTOR (regulatory-associated ration and TNF-a production upon engraftment. The T cells protein of MTOR).129,130 The significance of this association is are probably crucial to the development of the psoriatic not yet understood.126 phenotype (blocking T cells or TNF-a inhibited the conver- sion to lesional skin). The final link in translational research: To conclude, there are many parallels between mice and defining limits and doing required ‘homework’ humans, but despite this there exist significant differences in immune system development, activation and response to chal- We are clearly at a stage where scientific discoveries in transla- lenge, in both the innate and the adaptive arms. However, as tional studies have advanced the understanding and therapy of there are many things in common, there has been a tendency psoriasis. The lessons learned over the last decade can now to ignore the differences and even to make the assumption speed therapeutic development in other inflammatory diseases that what is true for mice is also necessarily applicable for of the skin, or even other organs. humans. By making such an assumption, we might overlook While many of our new biologic therapies seem to have aspects of human immunology that cannot be modelled in favourable safety profiles and may offer better long-term dis- mouse models. This also pertains to differences that may pre- ease management than older conventional agents, we still have clude a successful clinical trial in mice becoming a successful important information to gather about the functional conse- clinical trial in humans.120 quences of blocking specific immune molecules and pathways in humans. Future work must establish not only the long-term Genetic susceptibility: understanding the safety of selective immune antagonism, but we must also interplay of genes and the environment strive to learn more about the specific effects of new therapeu- tic agents on the human immune system. Hence, as a final Several putative loci for genetic susceptibility to the disease topic, we want to discuss some real limitations of therapeutic have been reported on the basis of genome-wide linkage stud- targeting with respect to demonstrating or testing pathogenic ies, not always with reproducible results.121 However, one concepts. The species specificity of most protein-based thera- locus in the major histocompatibility complex region on peutics often precludes testing in most nonhuman species chromosome 6 was replicated in several populations, being favoured for laboratory study. Mechanisms of these therapeu- reported first in a Finnish population in 1980.122–124 This tics are assumed from model systems, but in most cases very locus, termed psoriasis susceptibility 1 (PSORS1), is currently limited information in humans is available. considered the most important susceptibility locus.60 Two To link a clinical effect with an assumed mechanism genes within this interval were studied in respect to confer- requires actual demonstration of underlying assumptions in ring a susceptibility to psoriasis. Those genes are helix coiled- humans. In some instances, the underlying assumptions of coil rod homolog (HCR) and corneodesmosin (CDSN ⁄S). pharmacological mechanisms may have been misleading or CDSN is expressed in terminally differentiated keratinocytes wrongly interpreted. We will cite three examples, intended and in the inner root sheath of hair follicles and, apart from only to illustrate the scientific hazard of assuming too much, the skin, is detected in significant amounts only in the pla- and to make the appeal for doing sufficient ‘homework’ in centa and thymus.125,126 Current mapping of PSORS1 showed humans to make the translational process completely valid. 127 strong allelic association with HLA-Cw6. In the original work with DAB389IL-2 in psoriasis, it was Nevertheless, the exact location of PSORS1 gene remains assumed that the IL-2 receptor-bearing T lymphocytes, acting controversial because of linkage disequilibrium across this as pathogenic effectors, were the cellular target of this toxin.50 region.126,127 The penetrance of PSORS1 locus is estimated to Today we know that T cells with the highest levels of IL-2 be less than 15%, implying that other genetic ⁄environmental receptors are likely to be Tregs, so elimination of this cell factors may also contribute to the liability for the disease.126 population might have contributed to variable clinical out- Other potential loci are PSORS2, RUNX1 and RUNX3, comes seen with this agent in early trials. In addition, further RAPTOR, PSORS3, PSORS4 and PSORS9.126 Further potential work in the laboratory (our unpublished results) showed that psoriasis-susceptibility loci were recently discovered within activated DCs bear functional IL-2 receptors and can be killed the PSORS2 locus on chromosome 17q25. Those loci, runt- by DAB389IL-2. Hence, clinical improvements induced by related transcription factor 1 and 3 (RUNX1 and RUNX3), DAB389IL-2 perhaps can be interpreted only to the level that encode for a gene involved in blood cell (including immune IL-2 receptor-expressing leucocytes are pathogenic effectors in cells) development. A model of RUNX3 knockout mice lacks psoriasis. Langerhans cells in the epidermis, but not epidermal DCs. In As a second example, efalizumab was introduced for treat- addition, RUNX3-deficient cytotoxic T cells have defective ment of psoriasis on the basis that it would prevent trafficking response to antigens. RUNX3 also has a role in the develop- of T cells from the peripheral circulation into skin lesions by ment of cd T cells in the skin. Therefore, RUNX3 pathways blocking LFA-1-mediated firm adhesion to endothelial cells.131

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1112 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger

Work from the psoriasis skin transplantation models now sug- 5 Parrish JA. New concepts in therapeutic photomedicine: photo- gests that T-cell trafficking into skin may not be required to chemistry, optical targeting and the therapeutic window. J Invest initiate or sustain psoriasis lesions and other work from our Dermatol 1981; 77:45–50. 6 Ozawa M, Ferenczi K, Kikuchi T et al. 312-nanometer ultraviolet B laboratory shows unanticipated effects of efalizumab on light (narrow-band UVB) induces apoptosis of T cells within pso- expression of surface proteins on T cells that are unrelated to riatic lesions. J Exp Med 1999; 189:711–18. 97 LFA-1. In addition, there is variability to the therapeutic 7 Parrish JA, Fitzpatrick TB, Tanenbaum L et al. Photochemotherapy benefit seen in individual patients (although the primary target of psoriasis with oral methoxsalen and longwave ultraviolet light. is blocked in all patients) which does not yet have a clear N Engl J Med 1974; 291:1207–11. explanation. 8 Honig B, Morison WL, Karp D. Photochemotherapy beyond psori- Finally, and perhaps the largest mystery, is the variable ther- asis. J Am Acad Dermatol 1994; 31:775–90. 9 Gubner R. Effect of ‘aminopterin’ on epithelial tissues. AMA Arch apeutic response seen to alefacept in individual patients. Pro- Derm Syphilol 1951; 64:688–99. posed mechanisms of this agent (i) as a costimulation 10 Stewart WD, Wallace SM, Runikis JO. Absorption and local action antagonist and (ii) as a cytotoxic agent for T cells via activat- of methotrexate in human and mouse skin. Arch Dermatol 1972; ing NK lymphocytes do not explain the clinical variability in 106:357–61. the response and in intact T-cell responses to neoantigens in 11 Syed TA, Hadi SM, Qureshi ZA et al. Management of psoriasis patients undergoing active treatment. More recent genomic vulgaris with methotrexate 0Æ25% in a hydrophilic gel: a placebo- measures of T-cell responses in patients undergoing treatment controlled, double-blind study. J Cutan Med Surg 2001; 5:299–302. 12 Bjerring P, Beck HI, Zachariae H et al. Topical treatment of psoriat- with alefacept show unexpected agonist activities that seem to 132 ic skin with methotrexate cream: a clinical, pharmacokinetic, and differentiate responding from nonresponding patients. histological study. Acta Derm Venereol (Stockh) 1986; 66:515–19. Hence, a new mechanism has been proposed for alefacept that 13 Mueller W, Herrmann B. Cyclosporin A for psoriasis. N Engl J Med could only have been discovered by direct study of target leu- 1979; 301:555. cocytes in patients undergoing treatment with relevant doses 14 Saurat JH. Retinoids and psoriasis: novel issues in retinoid phar- of the agent. macology and implications for psoriasis treatment. J Am Acad Psoriasis is probably the best example of a common skin Dermatol 1999; 41:S2–6. 15 Smith EL, Walworth NC, Holick MF. Effect of 1 alpha,25- disease where rapid advances in pathogenic understanding and dihydroxyvitamin D3 on the morphologic and biochemical differ- development of new therapeutics have occurred through the entiation of cultured human epidermal keratinocytes grown in translational science approach. To reach the ultimate goals of serum-free conditions. J Invest Dermatol 1986; 86:709–14. disease-specific therapeutics and then personalized therapy 16 Kragballe K. Treatment of psoriasis by the topical application with the most appropriate agent(s), more work is needed to of the novel cholecalciferol analogue calcipotriol (MC 903). Arch refine the pathogenic hypotheses which have emerged from Dermatol 1989; 125:1647–52. recent work and to elucidate the basis of individual response 17 Kragballe K, Wildfang IL. Calcipotriol (MC 903), a novel vitamin D3 analogue stimulates terminal differentiation and inhibits pro- variability. The work on psoriasis points the way to the liferation of cultured human keratinocytes. Arch Dermatol Res 1990; potential for translational science to improve the under- 282:164–7. standing and treatment of other common skin diseases such as 18 Weinstein GD, McCullough JL, Ross P. Cell proliferation in nor- AD or alopecia areata, which also have complex genetics, mul- mal epidermis. J Invest Dermatol 1984; 82:623–8. tiple cell types, and environmental triggers acting to create an 19 Braverman IM, Sibley J. Role of the microcirculation in the treat- end phenotype in the skin which we recognize as a discrete ment and pathogenesis of psoriasis. J Invest Dermatol 1982; 78:12– disease. In a larger sense, we are at a critical juncture in the 17. 20 Mansbridge JN, Knapp AM. Changes in keratinocyte maturation classification and treatment of all skin diseases where the during wound healing. J Invest Dermatol 1987; 89:253–63. power of translational science and associated technologies 21 Krueger GG, Chambers DA, Shelby J. Involved and uninvolved stands to create a new taxonomy of skin diseases based on skin from psoriatic subjects: are they equally diseased? Assessment a real understanding of gene functions in the skin. Indeed, by skin transplanted to congenitally athymic (nude) mice. J Clin we are now in an age of molecular medicine that stands to Invest 1981; 68:1548–57. revolutionize dermatology. 22 Kohler G, Milstein C. Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 1975; 256:495–7. 23 Bos JD, Hulsebosch HJ, Krieg SR et al. Immunocompetent cells in References psoriasis. In situ immunophenotyping by monoclonal antibodies. Arch Dermatol Res 1983; 275:181–9. 1 de Mare S, de Jong E, van Erp PE et al. Markers for proliferation 24 Baker BS, Swain AF, Valdimarsson H et al. T-cell subpopulations in and keratinization in the margin of the active psoriatic lesion. Br J the blood and skin of patients with psoriasis. Br J Dermatol 1984; Dermatol 1990; 122:469–75. 110:37–44. 2 Braverman IM, Fonferko E. Studies in cutaneous aging: II. The 25 Baker BS, Swain AF, Fry L et al. Epidermal T lymphocytes and microvasculature. J Invest Dermatol 1982; 78:444–8. HLA-DR expression in psoriasis. Br J Dermatol 1984; 110:555–64. 3 Mansbridge JN, Knapp AM. The binding of Helix pomatia and Ulex 26 Gottlieb AB, Lifshitz B, Fu SM et al. Expression of HLA-DR mole- europeus agglutinins to normal and psoriatic skin. J Invest Dermatol cules by keratinocytes, and presence of Langerhans cells in the 1984; 82:170–5. dermal infiltrate of active psoriatic plaques. J Exp Med 1986; 4 Honigsmann H. Phototherapy for psoriasis. Clin Exp Dermatol 2001; 164:1013–28. 26:343–50.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1113

27 Bos JD, De Rie MA. The pathogenesis of psoriasis: immunological 47 Hattori N, Komine M, Yano S et al. Interferon-gamma, a strong facts and speculations. Immunol Today 1999; 20:40–6. suppressor of cell proliferation, induces upregulation of keratin 28 Krueger JG. The immunologic basis for the treatment of psoriasis K6, one of the inflammatory- and proliferation-associated keratins. with new biologic agents. J Am Acad Dermatol 2002; 46:1–23. J Invest Dermatol 2002; 119:403–10. 29 Reichrath J, Muller SM, Kerber A et al. Biologic effects of topical 48 Felli N, Pedini F, Zeuner A et al. Multiple members of the TNF calcipotriol (MC 903) treatment in psoriatic skin. J Am Acad Derma- superfamily contribute to IFN-gamma-mediated inhibition of tol 1997; 36:19–28. erythropoiesis. J Immunol 2005; 175:1464–72. 30 Vandermeeren M, Janssens S, Borgers M et al. Dimethylfumarate is 49 Krueger JG, Wolfe JT, Nabeya RT et al. Successful ultraviolet an inhibitor of cytokine-induced E-selectin, VCAM-1, and ICAM-1 B treatment of psoriasis is accompanied by a reversal of keratino- expression in human endothelial cells. Biochem Biophys Res Commun cyte pathology and by selective depletion of intraepidermal 1997; 234:19–23. T cells. J Exp Med 1995; 182:2057–68. 31 Kragballe K, Voorhees JJ. Modulation of epidermal cell division 50 Gottlieb SL, Gilleaudeau P, Johnson R et al. Response of psoriasis and growth by oxygenation products of arachidonic acid. J Allergy to a lymphocyte-selective toxin (DAB389IL-2) suggests a primary Clin Immunol 1984; 74:426–9. immune, but not keratinocyte, pathogenic basis. Nat Med 1995; 32 Henderson WR Jr. The role of leukotrienes in inflammation. 1:442–7. Ann Intern Med 1994; 121:684–97. 51 Tepper MA, Linsley PS, Tritschler D et al. Tolerance induction by 33 Degreef H, Dockx P, De Doncker P et al. A double-blind vehicle- soluble CTLA4 in a mouse skin transplant model. Transplant Proc controlled study of R 68 151 in psoriasis: a topical 5-lipoxygenase 1994; 26:3151–4. inhibitor. J Am Acad Dermatol 1990; 22:751–5. 52 Blazar BR, Taylor PA, Linsley PS et al. In vivo blockade of 34 Weinshenker BG, Bass BH, Ebers GC et al. Remission of psoriatic CD28 ⁄CTLA4: B7 ⁄BB1 interaction with CTLA4-Ig reduces lethal lesions with muromonab-CD3 (orthoclone OKT3) treatment. JAm murine graft-versus-host disease across the major histocompatibil- Acad Dermatol 1989; 20:1132–3. ity complex barrier in mice. Blood 1994; 83:3815–25. 35 Prinz J, Braun-Falco O, Meurer M et al. Chimaeric CD4 mono- 53 Abrams JR, Lebwohl MG, Guzzo CA et al. CTLA4Ig-mediated clonal antibody in treatment of generalised pustular psoriasis. blockade of T-cell costimulation in patients with psoriasis vulgaris. Lancet 1991; 338:320–1. J Clin Invest 1999; 103:1243–52. 36 Nicolas JF, Chamchick N, Thivolet J et al. CD4 antibody treatment 54 June CH, Ledbetter JA, Gillespie MM et al. T-cell proliferation of severe psoriasis. Lancet 1991; 338:321. involving the CD28 pathway is associated with cyclosporine-resis- 37 Ellis CN, Gorsulowsky DC, Hamilton TA et al. Cyclosporine impr- tant interleukin 2 gene expression. Mol Cell Biol 1987; 7:4472–81. oves psoriasis in a double-blind study. JAMA 1986; 256:3110–16. 55 Abrams JR, Kelley SL, Hayes E et al. Blockade of T lymphocyte 38 Baker BS, Griffiths CE, Lambert S et al. The effects of cyclosporin A costimulation with cytotoxic T lymphocyte-associated antigen on T lymphocyte and dendritic cell sub-populations in psoriasis. 4-immunoglobulin (CTLA4Ig) reverses the cellular pathology of Br J Dermatol 1987; 116:503–10. psoriatic plaques, including the activation of keratinocytes, den- 39 Jegasothy BV, Ackerman CD, Todo S et al. Tacrolimus (FK 506) – dritic cells, and endothelial cells. J Exp Med 2000; 192:681–94. a new therapeutic agent for severe recalcitrant psoriasis. Arch 56 Mrowietz U, Zhu K, Christophers E. Treatment of severe psoriasis Dermatol 1992; 128:781–5. with anti-CD25 monoclonal antibodies. Arch Dermatol 2000; 40 Khandke L, Krane JF, Ashinoff R et al. Cyclosporine in psoriasis 136:675–6. treatment. Inhibition of keratinocyte cell-cycle progression in G1 57 Krueger JG, Walters IB, Miyazawa M et al. Successful in vivo block- independent of effects on transforming growth factor alpha ⁄ ade of CD25 (high-affinity interleukin 2 receptor) on T cells by epidermal growth factor receptor pathways. Arch Dermatol 1991; administration of humanized anti-Tac antibody to patients with 127:1172–9. psoriasis. J Am Acad Dermatol 2000; 43:448–58. 41 Nickoloff BJ, Mitra RS, Riser BL et al. Modulation of keratinocyte 58 Gottlieb AB. Psoriasis: emerging therapeutic strategies. Nat Rev Drug motility. Correlation with production of extracellular matrix mole- Discov 2005; 4:19–34. cules in response to growth promoting and antiproliferative 59 Kupper TS. Immunologic targets in psoriasis. N Engl J Med 2003; factors. Am J Pathol 1988; 132:543–51. 349:1987–90. 42 Ristow HJ. A major factor contributing to epidermal proliferation 60 Schon MP, Boehncke WH. Psoriasis. N Engl J Med 2005; in inflammatory skin diseases appears to be interleukin 1 or a 352:1899–912. related protein. Proc Natl Acad Sci USA 1987; 84:1940–4. 61 Coven TR, Walters IB, Cardinale I et al. PUVA-induced lymphocyte 43 Sauder DN, Stanulis-Praeger BM, Gilchrest BA. Autocrine growth apoptosis: mechanism of action in psoriasis. Photodermatol Photo- stimulation of human keratinocytes by epidermal cell-derived thy- immunol Photomed 1999; 15:22–7. mocyte-activating factor: implications for skin aging. Arch Dermatol 62 Jeffes EW III, McCullough JL, Pittelkow MR et al. Methotrexate Res 1988; 280:71–6. therapy of psoriasis: differential sensitivity of proliferating lym- 44 Groves RW, Rauschmayr T, Nakamura K et al. Inflammatory and phoid and epithelial cells to the cytotoxic and growth-inhibitory hyperproliferative skin disease in mice that express elevated levels effects of methotrexate. J Invest Dermatol 1995; 104:183–8. of the IL-1 receptor (type I) on epidermal keratinocytes. Evidence 63 Farrell AM, Antrobus P, Simpson D et al. A rapid flow cytometric that IL-1-inducible secondary cytokines produced by keratinocytes assay to detect CD4+ and CD8+ T-helper (Th) 0, Th1 and Th2 in vivo can cause skin disease. J Clin Invest 1996; 98:336–44. cells in whole blood and its application to study cytokine levels in 45 Grossman RM, Krueger J, Yourish D et al. Interleukin 6 is atopic dermatitis before and after cyclosporin therapy. Br J Dermatol expressed in high levels in psoriatic skin and stimulates prolifera- 2001; 144:24–33. tion of cultured human keratinocytes. Proc Natl Acad Sci USA 1989; 64 Nickoloff BJ, Nestle FO. Recent insights into the immunopatho- 86:6367–71. genesis of psoriasis provide new therapeutic opportunities. J Clin 46 Elder JT, Sartor CI, Boman DK et al. Interleukin-6 in psoriasis: Invest 2004; 113:1664–75. expression and mitogenicity studies. Arch Dermatol Res 1992; 65 Szabo SK, Hammerberg C, Yoshida Y et al. Identification and 284:324–32. quantitation of interferon-gamma producing T cells in psoriatic

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 1114 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger

lesions: localization to both CD4+ and CD8+ subsets. J Invest Der- mechanism underlying unrestrained pathogenic effector T cell matol 1998; 111:1072–8. proliferation. J Immunol 2005; 174:164–73. 66 Austin LM, Ozawa M, Kikuchi T et al. The majority of epidermal 85 Xu Q, Lee J, Jankowska-Gan E et al. Human CD4+ CD25low adap- T cells in psoriasis vulgaris lesions can produce type 1 cytokines, tive T regulatory cells suppress delayed-type hypersensitivity dur- interferon-gamma, interleukin-2, and tumor necrosis factor-alpha, ing transplant tolerance. J Immunol 2007; 178:3983–95. defining TC1 (cytotoxic T lymphocyte) and TH1 effector popula- 86 Zhou X, Krueger JG, Kao MC et al. Novel mechanisms of T-cell tions: a type 1 differentiation bias is also measured in circulating and dendritic cell activation revealed by profiling of psoriasis on blood T cells in psoriatic patients. J Invest Dermatol 1999; 113:752– the 63,100-element oligonucleotide array. Physiol Genomics 2003; 9. 13:69–78. 67 Szegedi A, Aleksza M, Gonda A et al. Elevated rate of Thelper1 87 Wang F, Lee E, Lowes MA et al. Prominent production of IL-20 (T(H)1) lymphocytes and serum IFN-gamma levels in psoriatic by CD68+ ⁄CD11c+ myeloid-derived cells in psoriasis: gene regu- patients. Immunol Lett 2003; 86:277–80. lation and cellular effects. J Invest Dermatol 2006; 126:1590–9. 68 Asadullah K, Sterry W, Trefzer U. Cytokines: interleukin and 88 Wolk K, Kunz S, Witte E et al. IL-22 increases the innate immu- interferon therapy in dermatology. Clin Exp Dermatol 2002; nity of tissues. Immunity 2004; 21:241–54. 27:578–84. 89 Shortman K, Naik SH. Steady-state and inflammatory dendritic-cell 69 Lew W, Bowcock AM, Krueger JG. Psoriasis vulgaris: cutaneous development. Nat Rev Immunol 2007; 7:19–30. lymphoid tissue supports T-cell activation and ‘type 1’ inflamma- 90 Serbina NV, Salazar-Mather TP, Biron CA et al. TNF ⁄iNOS-produc- tory gene expression. Trends Immunol 2004; 25:295–305. ing dendritic cells mediate innate immune defense against bacte- 70 Ghoreschi K, Thomas P, Breit S et al. Interleukin-4 therapy of pso- rial infection. Immunity 2003; 19:59–70. riasis induces Th2 responses and improves human autoimmune 91 Asselin-Paturel C, Trinchieri G. Production of type I interferons: disease. Nat Med 2003; 9:40–6. plasmacytoid dendritic cells and beyond. J Exp Med 2005; 71 Asadullah K, Sabat R, Friedrich M et al. Interleukin-10: an impor- 202:461–5. tant immunoregulatory cytokine with major impact on psoriasis. 92 Lowes MA, Chamian F, Abello MV et al. Increase in TNF-alpha Curr Drug Targets Inflamm Allergy 2004; 3:185–92. and inducible nitric oxide synthase-expressing dendritic cells in 72 Trepicchio WL, Ozawa M, Walters IB et al. Interleukin-11 therapy psoriasis and reduction with efalizumab (anti-CD11a). Proc Natl selectively downregulates type I cytokine proinflammatory path- Acad Sci USA 2005; 102:19057–62. ways in psoriasis lesions. J Clin Invest 1999; 104:1527–37. 93 Gottlieb AB, Chamian F, Masud S et al. TNF rapidly down-regu- 73 Kauffman CL, Aria N, Toichi E et al. A phase I study evaluating lates multiple proinflammatory pathways in psoriasis plaques. the safety, pharmacokinetics, and clinical response of a human J Immunol 2005; 175:2721–9. IL-12 p40 antibody in subjects with plaque psoriasis. J Invest 94 Zheng YL, Li BS, Kanungo J et al. Cdk5 modulation of mitogen- Dermatol 2004; 123:1037–44. activated protein kinase signaling regulates neuronal survival. 74 Chamian F, Lowes MA, Lin SL et al. Alefacept reduces infiltrating Mol Biol Cell 2007; 18:404–13. T cells, activated dendritic cells, and inflammatory genes in psori- 95 Cho ML, Kang JW, Moon YM et al. STAT3 and NF-kappaB signal asis vulgaris. Proc Natl Acad Sci USA 2005; 102:2075–80. pathway is required for IL-23-mediated IL-17 production in spon- 75 McKenzie BS, Kastelein RA, Cua DJ. Understanding the IL-23– taneous arthritis animal model IL-1 receptor antagonist-deficient IL-17 immune pathway. Trends Immunol 2006; 27:17–23. mice. J Immunol 2006; 176:5652–61. 76 Teunissen MB, Koomen CW, de Waal Malefyt R et al. Interleukin- 96 Nestle FO, Conrad C, Tun-Kyi A et al. Plasmacytoid predendritic 17 and interferon-gamma synergize in the enhancement of pro- cells initiate psoriasis through interferon-alpha production. J Exp inflammatory cytokine production by human keratinocytes. J Invest Med 2005; 202:135–43. Dermatol 1998; 111:645–9. 97 Guttman-Yassky E, Lowes MA, Fuentes-Duculan J et al. Major dif- 77 Lowes MA, Chamian F, Abello MV et al. Eruptive papules during ferences in inflammatory dendritic cells and their products distin- efalizumab (anti-CD11a) therapy of psoriasis vulgaris: a case guish atopic dermatitis from psoriasis. J Allergy Clin Immunol 2007; series. BMC Dermatol 2007; 7:2. 119:1210–17. 78 Aggarwal S, Ghilardi N, Xie MH et al. Interleukin-23 promotes a 98 Elder JT, Fisher GJ, Lindquist PB et al. Overexpression of trans- distinct CD4 T cell activation state characterized by the production forming growth factor alpha in psoriatic epidermis. Science 1989; of interleukin-17. J Biol Chem 2003; 278:1910–14. 243:811–14. 79 Anderson EJ, McGrath MA, Thalhamer T et al. Interleukin-12 to 99 Cook PW, Piepkorn M, Clegg CH et al. Transgenic expression of interleukin ‘infinity’: the rationale for future therapeutic cytokine the human amphiregulin gene induces a psoriasis-like phenotype. targeting. Springer Semin Immunopathol 2006; 27:425–42. J Clin Invest 1997; 100:2286–94. 80 Umetsu DT, Akbari O, Dekruyff RH. Regulatory T cells control 100 Blumberg H, Conklin D, Xu WF et al. Interleukin 20: discovery, the development of allergic disease and asthma. J Allergy Clin Immu- receptor identification, and role in epidermal function. Cell 2001; nol 2003; 112:480–7. 104:9–19. 81 Wraith DC, Nicolson KS, Whitley NT. Regulatory CD4+ T cells 101 Kopp T, Lenz P, Bello-Fernandez C et al. IL-23 production by and the control of autoimmune disease. Curr Opin Immunol 2004; cosecretion of endogenous p19 and transgenic p40 in keratin 16:695–701. 14 ⁄p40 transgenic mice: evidence for enhanced cutaneous 82 Fontenot JD, Gavin MA, Rudensky AY. Foxp3 programs the immunity. J Immunol 2003; 170:5438–44. development and function of CD4+ CD25+ regulatory T cells. Nat 102 Gudjonsson JE, Johnston A, Dyson M et al. Mouse models of Immunol 2003; 4:330–6. psoriasis. J Invest Dermatol 2007; 127:1292–308. 83 Bruder D, Probst-Kepper M, Westendorf AM et al. Neuropilin-1: a 103 de Jongh GJ, Zeeuwen PL, Kucharekova M et al. High expression surface marker of regulatory T cells. Eur J Immunol 2004; 34:623– levels of keratinocyte antimicrobial proteins in psoriasis compared 30. with atopic dermatitis. J Invest Dermatol 2005; 125:1163–73. 84 Sugiyama H, Gyulai R, Toichi E et al. Dysfunctional blood and 104 Boguniewicz M, Schmid-Grendelmeier P, Leung DY. Atopic target tissue CD4+ CD25high regulatory T cells in psoriasis: dermatitis. J Allergy Clin Immunol 2006; 118:40–3.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 Psoriasis: translational research, E. Guttman-Yassky and J.G. Krueger 1115

105 Farshidi A, Sadeghi P. Successful treatment of severe refractory 119 Boyman O, Hefti HP, Conrad C et al. Spontaneous development of atopic dermatitis with efalizumab. J Drugs Dermatol 2006; 5:994–8. psoriasis in a new animal model shows an essential role for resi- 106 Hassan AS, Kaelin U, Braathen LR et al. Clinical and immunopatho- dent T cells and tumor necrosis factor-alpha. J Exp Med 2004; logic findings during treatment of recalcitrant atopic eczema with 199:731–6. efalizumab. J Am Acad Dermatol 2007; 56:217–21. 120 Mestas J, Hughes CC. Of mice and not men: differences between 107 Wang H, Peters T, Kess D et al. Activated macrophages are essen- mouse and human immunology. J Immunol 2004; 172:2731–8. tial in a murine model for T cell-mediated chronic psoriasiform 121 Davidson A, Diamond B. Autoimmune diseases. N Engl J Med 2001; skin inflammation. J Clin Invest 2006; 116:2105–14. 345:340–50. 108 Stratis A, Pasparakis M, Rupec RA et al. Pathogenic role for skin 122 Tiilikainen A, Lassus A, Karvonen J et al. Psoriasis and HLA-Cw6. macrophages in a mouse model of keratinocyte-induced psoriasis- Br J Dermatol 1980; 102:179–84. like skin inflammation. J Clin Invest 2006; 116:2094–104. 123 Trembath RC, Clough RL, Rosbotham JL et al. Identification of a 109 Sano S, Chan KS, Carbajal S et al. Stat3 links activated keratinocytes major susceptibility locus on chromosome 6p and evidence for and immunocytes required for development of psoriasis in a further disease loci revealed by a two stage genome-wide search novel transgenic mouse model. Nat Med 2005; 11:43–9. in psoriasis. Hum Mol Genet 1997; 6:813–20. 110 Sano S, Chan KS, Kira M et al. Signal transducer and activator of 124 Burden AD, Javed S, Bailey M et al. Genetics of psoriasis: paternal transcription 3 is a key regulator of keratinocyte survival and pro- inheritance and a locus on chromosome 6p. J Invest Dermatol 1998; liferation following UV irradiation. Cancer Res 2005; 65:5720–9. 110:958–60. 111 Zenz R, Eferl R, Kenner L et al. Psoriasis-like skin disease and 125 Tazi Ahnini R, Camp NJ, Cork MJ et al. Novel genetic association arthritis caused by inducible epidermal deletion of Jun proteins. between the corneodesmosin (MHC S) gene and susceptibility to Nature 2005; 437:369–75. psoriasis. Hum Mol Genet 1999; 8:1135–40. 112 Haider AS, Duculan J, Whynot JA et al. Increased JunB mRNA and 126 Liu Y, Krueger JG, Bowcock AM. Psoriasis: genetic associations protein expression in psoriasis vulgaris lesions. J Invest Dermatol and immune system changes. Genes Immun 2007; 8:1–12. 2006; 126:912–14. 127 Nair RP, Stuart PE, Nistor I et al. Sequence and haplotype analysis 113 Johansen C, Kragballe K, Rasmussen M et al. Activator protein 1 supports HLA-C as the psoriasis susceptibility 1 gene. Am J Hum DNA binding activity is decreased in lesional psoriatic skin com- Genet 2006; 78:827–51. pared with nonlesional psoriatic skin. Br J Dermatol 2004; 128 Brenner O, Levanon D, Negreanu V et al. Loss of Runx3 function 151:600–7. in leukocytes is associated with spontaneously developed colitis 114 Kulski JK, Kenworthy W, Bellgard M et al. Gene expression profil- and gastric mucosal hyperplasia. Proc Natl Acad Sci USA 2004; ing of Japanese psoriatic skin reveals an increased activity in 101:16016–21. molecular stress and immune response signals. J Mol Med 2005; 129 Hara K, Maruki Y, Long X et al. Raptor, a binding partner of target 83:964–75. of rapamycin (TOR), mediates TOR action. Cell 2002; 110:177– 115 Gilhar A, Etzioni A, Assy B et al. Response of grafts from patients 89. with alopecia areata transplanted onto nude mice, to administra- 130 Capon F, Helms C, Veal CD et al. Genetic analysis of PSORS2 tion of interferon-gamma. Clin Immunol Immunopathol 1993; 66:120– markers in a UK dataset supports the association between RAPTOR 6. SNPs and familial psoriasis. J Med Genet 2004; 41:459–60. 116 Boehncke WH, Dressel D, Zollner TM, Kaufmann R. Pulling the 131 Jullien D, Prinz JC, Langley RG et al. T-cell modulation for the trigger on psoriasis. Nature 1996; 379:777. treatment of chronic plaque psoriasis with efalizumab (Raptiva): 117 Wrone-Smith T, Nickoloff BJ. Dermal injection of immunocytes mechanisms of action. Dermatology 2004; 208:297–306. induces psoriasis. J Clin Invest 1996; 98:1878–87. 132 Haider AS, Lowes MA, Gardner H et al. Novel insight into the 118 Villadsen LS, Schuurman J, Beurskens F et al. Resolution of psoria- agonistic mechanism of alefacept in vivo: differentially expressed sis upon blockade of IL-15 biological activity in a xenograft genes may serve as biomarkers of response in psoriasis patients. mouse model. J Clin Invest 2003; 112:1571–80. J Immunol 2007; 178:7442–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1103–1115 GUIDELINES DOI 10.1111/j.1365-2133.2007.08283.x Guidelines for evaluation and management of urticaria in adults and children C.E.H. Grattan and F. Humphreys* on behalf of the British Association of Dermatologists Therapy Guidelines and Audit Subcommittee Department of Dermatology, Norfolk and Norwich University Hospital, Norwich NR4 7UY, and St John’s Institute of Dermatology, St Thomas’ Hospital, London SE1 7EH, U.K. *Department of Dermatology, Warwick Hospital, Warwick CV34 5BW, U.K.

Summary

Correspondence Appropriate management of urticaria depends on the correct evaluation of clinical Clive Grattan. patterns and causes where these can be identified. Guidance for treatment is pre- E-mail: [email protected] sented, based on the strength of evidence available at the time of preparation. As many of the recommendations relate to the off-licence use of drugs, it is particu- Accepted for publication 4 July 2007 larly important that clinicians should be familiar with dosing and side-effects of treatment in the context of managing urticaria. Key words guidelines, treatment, urticaria

Conflicts of interest None declared.

Members of the British Association of Dermatologists Therapy Guidelines and Audit Subcommittee are: A.D. Ormerod (Chair), H. Bell, F. Humphreys, D. Mitchell, R. Bull, M. Tidman, D.J. Eedy, S. Joseph and S. Wagle.

Disclaimer Clinical classification

These guidelines have been prepared for dermatologists on For clinical purposes it is often more helpful to classify urti- behalf of the British Association of Dermatologists and reflect caria by presentation than by aetiology, which is often diffi- the best data available at the time the report was prepared. cult to establish. A classification based on clinical features may Caution should be exercised in interpreting the data; the be used to guide appropriate investigation and management. It results of future studies may require alterations of the conclu- is usually possible to distinguish clearly recognizable patterns sions or recommendations in this report. It may be necessary of urticaria on the clinical presentation, supported, where or even desirable to depart from the guidelines in the interests appropriate, by challenge tests and skin biopsy (Table 1). The of specific patients and special circumstances. Just as adherence presentation of urticaria in childhood is similar to that in to the guidelines may not constitute defence against a claim of adults. Clinical and aetiological classifications should be com- negligence, so deviation from them should not necessarily be plementary rather than exclusive: for example, chronic ordin- deemed negligent. ary urticaria (COU) is most appropriate when the aetiology remains uncertain. Where there is evidence of histamine- Definition releasing autoantibodies the patient has autoimmune COU (syn. chronic autoimmune urticaria) but where there is no The term urticaria is widely used to describe an eruption of evidence of functional autoantibodies the patient has idio- weals. It is now also increasingly being used to define a dis- pathic COU (syn. chronic idiopathic urticaria). ease characterized by short-lived itchy weals, angio-oedema or Ordinary urticaria is the commonest pattern, presenting with both together. Most patients with urticaria do not have sys- spontaneous weals anywhere on the body with or without temic reactions, but allergic and some physical urticarias may angio-oedema. Although the underlying tendency to urticaria occasionally progress to anaphylaxis. Conversely, urticaria is is spontaneous it is often possible to identify aggravating often a feature of anaphylactic and anaphylactoid reactions. factors, such as heat or pressure from clothing, that appear

2007 The Authors 1116 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. 1117

Table 1 Clinical classification of the urticarias latter may occasionally progress to anaphylaxis in a highly sensitized individual (e.g. latex allergy). Ordinary urticaria Urticarial vasculitis presents with urticaria clinically but small Acute (up to 6 weeks of continuous activity) vessel vasculitis histologically. Other features of this systemic Chronic (6 weeks or more of continuous activity) disease may include joint and renal involvement. Episodic (acute intermittent or recurrent activity) Autoinflammatory syndromes presenting with urticaria typically Physical urticarias (reproducibly induced by the same develop spontaneous weals, pyrexia and malaise, with other physical stimulus) Mechanical features that define the disease phenotype (such as renal amy- Delayed pressure urticaria loidosis and sensorineural deafness in Muckle–Wells syn- Symptomatic dermographism drome). The inherited patterns usually present in early Vibratory angio-oedema childhood. Thermal The duration of individual weals can be very helpful in dis- Cholinergic urticaria tinguishing between these clinical patterns: weals typically last Cold contact urticaria from 2 to 24 h in ordinary urticaria and up to 2 h in contact Localized heat urticaria Other urticaria. The weals of physical urticaria are gone within Aquagenic urticaria an hour except those in delayed pressure urticaria, which take Solar urticaria 2–6 h to develop and up to 48 h to fade. The weals of urti- Exercise-induced anaphylaxis carial vasculitis usually persist for days. Angio-oedema may Angio-oedema without weals last up to 3 days without treatment. Idiopathic Drug-induced C1 esterase inhibitor deficiency Aetiology Contact urticaria (contact with allergens or chemicals) Urticarial vasculitis (defined by vasculitis on skin biopsy) Despite thorough evaluation many cases remain unexplained Autoinflammatory syndromes (‘idiopathic’) but it may be possible to assign a specific aetiol- Hereditary ogy to individual cases of urticaria (Table 2). Cryopyrin-associated periodic syndromes (CIAS1 mutations) Acquired Schnitzler syndrome Immunological urticaria

At least 30% of patients with COU have histamine-releasing autoantibodies. These degranulate mast cells and basophils in vitro by activating high-affinity IgE receptors directly or IgE to encourage urticarial lesions. It may follow an acute, epi- bound to them.1 Patients with evidence of functional autoanti- sodic (syn. intermittent) or chronic course. Weals occur bodies are increasingly being regarded as having an auto- continuously every day or almost daily while the disease is immune subset of urticaria. Cross-linking of specific IgE on active. cutaneous mast cells by allergens can cause contact urticaria, Physical urticarias are triggered reproducibly by one or more anaphylaxis and some cases of acute or episodic ordinary urti- physical stimuli. Swellings are induced rather than spontane- caria, but experience shows that allergy is not the cause ous. Defining the stimulus provides an opportunity to mini- of chronic continuous disease in adults. Urticarial vasculitis mize or prevent urticaria through lifestyle changes. The and acute urticarial reactions to drugs or blood products most readily identifiable triggers are mechanical or thermal. (serum sickness) are thought to result from the lodging of Some authorities distinguish cholinergic urticaria from the immune complexes in small blood vessels. The angio-oedema physical urticarias because it is primarily induced by the of C1 inh deficiency is mediated by kinins resulting from stimulus for sweating rather than overheating per se (even though the usual reason for sweating is a raised core Table 2 Aetiologies of urticaria temperature). Angio-oedema without weals should be distinguished from angio- Idiopathic oedema occurring with weals as it may be caused by angio- Immunological tensin-converting enzyme (ACE) inhibitors or be a presenta- Autoimmune (autoantibodies against FcRI or IgE) tion of C1 esterase inhibitor (C1 inh) deficiency. Patients with Allergic (IgE-mediated type I hypersensitivity reactions) Immune complex (urticarial vasculitis) C1 inh deficiency may present with abdominal pain without Complement-dependent (C1 esterase inhibitor deficiency) obvious angio-oedema. Angio-oedema without weals may also Nonimmunological be idiopathic. Direct mast cell-releasing agents (e.g. opiates) Contact urticaria occurs only when the eliciting substance is Aspirin, nonsteroidal anti-inflammatories and dietary absorbed percutaneously or through mucous membranes. It is pseudoallergens never spontaneous. Percutaneous or mucosal absorption of an Angiotensin-converting enzyme inhibitors allergen may result in a localized or a systemic reaction. The

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 1118 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. complement activation and bradykinin formation rather than performed in all patients. Relevant clinical and laboratory tests histamine. for the different clinical patterns of urticaria are summarized in Table 3. Nonimmunological urticaria Acute or episodic ordinary urticaria Degranulation of mast cells and basophils can occur indepen- dently of IgE receptor activation after exposure to certain No investigations are required except where suggested by drugs (e.g. codeine) and other agents (e.g. radiocontrast the history. IgE-mediated reactions to environmental allergens media). The mechanism by which aspirin, nonsteroidal anti- (such as latex, nuts or fish) as a cause of acute allergic or inflammatory drugs (NSAIDs) and dietary pseudoallergens contact urticaria can be confirmed by skin-prick testing (such as salicylates, azo dyes and food preservatives) cause or (where there are facilities) and CAP fluoroimmunoassay (pre- aggravate urticaria remains uncertain but probably involves viously radioallergosorbent tests, RAST) on blood. Results of leukotriene formation as well as histamine release. Angio- both have to be interpreted in the clinical context. Single- oedema due to ACE inhibitors is believed to result from inhi- blind oral challenge with food additives or aspirin may be bition of kinin breakdown by ACE. appropriate in the evaluation of episodic urticaria in the appropriate clinical setting in centres where challenge cap- sules are available. Associations

Thyroid autoimmunity in COU (14%) is more prevalent than Chronic ordinary urticaria in population controls (6%)2 (Quality of evidence II-ii; see Appen- dix 1). A significantly higher prevalence of coeliac disease in No investigations are required for the majority of patients children and adolescents with severe chronic urticaria than with mild disease responding to H1 antihistamines. A useful in case-matched controls has been reported.3 Associations screening profile for nonresponders with more severe disease between chronic urticaria and occult infection (e.g. dental could include a full blood count and white cell differential abscess4 and gastrointestinal candidiasis5) have been proposed (for instance, to detect the eosinophilia of bowel helminth but there is little evidence to support them (Quality of evidence infections or the leucopenia of systemic lupus erythematosus), III). A meta-analysis of therapeutic trials for Helicobacter pylori and erythrocyte sedimentation rate (usually normal in COU found that resolution of chronic urticaria was more likely but may be raised in urticarial vasculitis and always raised in when antibiotic therapy was successful than when it was not autoinflammatory syndromes). Thyroid autoantibodies and (Quality of evidence I, Strength of recommendation B).6 There is no stat- thyroid function tests should be performed, especially if an istical association between malignancy and urticaria7 (Quality autoimmune aetiology of urticaria is likely. There is currently of evidence II-ii) although individual case reports have been no routine laboratory test for histamine-releasing autoantibod- published. ies but intradermal injection of autologous serum (the auto- logous serum skin test, ASST) offers a reasonably sensitive and 9 Appropriate investigations specific screening test in centres with experience of doing it. The basophil histamine release assay remains the gold standard The diagnosis of urticaria is primarily clinical.8 Any investi- investigation for functional autoantibodies in centres where it gations should be guided by the history and should not be is available.

Table 3 Relevant investigations Skin Physical FBC ESR TA ⁄TFT IgE C4 biopsy challenge Ordinary urticaria Acute ⁄episodic ))) (+) )) ) Chronic (+) (+) (+) ))) ) Physical urticaria ))) ))) + Angio-oedema without weals ))) )+ )) Contact urticaria ))) (+) )) ) Urticarial vasculitis + + ))++ ) Autoinflammatory syndrome + + )))))

FBC, full blood count; ESR, erythrocyte sedimentation rate; TA, thyroid autoantibodies; TFT, thyroid function tests; IgE, specific IgE (CAP) or skin prick tests; C4, component of complement as a marker for C1 esterase inhibitor deficiency and in hypocomplementaemic urticarial vasculitis; (+), discretionary investigations.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. 1119

Physical urticarias Pharmacological agents

Physical urticarias may occur alone or coexist with ordinary Antihistamines urticaria. International standards for the diagnosis of physical urticarias and definitions of challenge testing have been The efficacy and safety of antihistamines in urticaria is undis- proposed.10 puted although not all patients respond and some, very occa- sionally, become worse. The outcome of randomized controlled studies of nonsedating H1 antihistamines has been Angio-oedema without weals summarized.12 Seven nonsedating H1 antihistamines are cur- Serum C4 should be used as an initial screening test for rently licensed for urticaria in the U.K. Cetirizine, deslorata- hereditary and acquired C1 inh deficiency. A low C4 level dine, fexofenadine, levocetirizine, loratadine and mizolastine between and during attacks (< 30% mean normal) has a are taken once daily. Acrivastine is taken three times a day in very high sensitivity but low specificity for C1 inh defi- view of its short half-life (T½). It is now available in the U.K. ciency.11 If low, C1 inh deficiency can be confirmed by only in a nonprescription presentation. Cetirizine (the active quantitative and functional C1 inh assays. Immunochemical metabolite of hydroxyzine) may be sedating, especially at and functional C1 inh are both low in type I hereditary higher doses. Mizolastine is contraindicated in clinically angio-oedema (HAE) whereas only functional activity is low significant cardiac disease and when there is prolongation of in type II HAE. C1q is also reduced in acquired C1 inh the Q-T interval. It should not be taken concurrently with drugs deficiency. that inhibit hepatic metabolism via cytochrome P450 (includ- ing macrolide antibiotics and imidazole antifungals) and with drugs that have potential arrythmic properties (including tricy- Urticarial vasculitis clic antidepressants, such as doxepin). Cetirizine has the short- Lesional skin biopsy is essential to confirm the presence of est time to attain maximum concentration, which may be an small-vessel vasculitis histologically (leucocytoclasia, endothe- advantage where rapid availability is clinically important. Des- lial cell damage, perivascular fibrin deposition and red cell loratadine has the longest elimination T½ at 27 h and should extravasation are key changes although there is no single therefore be discontinued 6 days before skin prick testing. defining feature). Patients with urticarial vasculitis need a All patients should be offered the choice of at least two full vasculitis screen, including serum complement assays nonsedating H1 antihistamines because responses and toler- for C3 and C4 to distinguish normocomplementaemic ance vary between individuals (Strength of recommendation A). It from hypocomplementaemic disease, which carries a worse has become common practice to increase the dose above the prognosis. manufacturer’s licensed recommendation for patients who do not respond when the potential benefits are considered to out- Interventions weigh any risks (Quality of evidence III, Strength of recommendation C). ‘Antiallergic’ effects on mast-cell mediator release of possible clinical importance have been shown with cetirizine13 and General measures loratadine,14 especially at higher doses. Adjustments to the Nonspecific aggravating factors, such as overheating, stress, timing of medication can be helpful to ensure that the highest alcohol and drugs with the potential to worsen urticaria drug levels are obtained when urticaria is anticipated. The use (e.g. aspirin and codeine) should be minimized. The risk of of sedating antihistamines as monotherapy is now less com- cross-reactions between aspirin and NSAIDs is difficult to mon because of concerns about reduced concentration and quantify but may relate to potency of cyclooxygenase inhi- performance but they can be effective and well tolerated by bition and dose. NSAIDs should be avoided in aspirin-sensi- some individuals. Doxepin has useful antihistaminic properties tive patients with urticaria. ACE inhibitors should be but has sedating and anticholinergic side-effects. Addition of a avoided in patients with angio-oedema without weals and sedating antihistamine at night [e.g. chlorphenamine (chlor- used with caution in urticaria if angio-oedema is also pres- pheniramine) 4–12 mg, hydroxyzine 10–50 mg] to a non- ent. Oestrogens should be avoided in HAE. Cooling antipru- sedating antihistamine by day may help patients sleep better ritic lotions, such as calamine or 1% menthol in aqueous although it probably has little additional clinical effect on urti- cream, can be soothing (Quality of evidence III, Strength of recom- caria if the H1 receptor is already saturated. The off-licence mendation A). Clearly written information sheets, such as the addition of an H2 antihistamine, on the other hand, may British Association of Dermatologists’ publication on urti- sometimes give better control of urticaria than an H1 anti- caria and angio-oedema, can be very helpful to patients. It histamine taken alone (Quality of evidence II, Strength of recommenda- is important to explain to the patient that a cause of the tion C)15,16 although, in practice, it may be more helpful for condition is unlikely to be found but the prognosis for dyspepsia that may accompany severe urticaria. eventual recovery from ordinary, physical and vasculitic urticarias is excellent. Some physical urticarias may be Renal impairment. Acrivastine should be avoided in moderate ) especially persistent. renal impairment (creatinine clearance 10–20 mL min 1) and

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 1120 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. the dose of cetirizine, levocetirizine and hydroxyzine should Epinephrine (syn. adrenaline) be halved. Cetirizine, levocetirizine and alimemazine (trime- prazine) should be avoided in severe renal impairment (creati- Intramuscular epinephrine can be life saving in anaphylaxis ) nine clearance < 10 mL min 1). Loratadine and desloratadine and in severe laryngeal angio-oedema but should be used with should be used with caution in severe renal impairment. caution in hypertension and ischaemic heart disease. Dosing is weight dependent. The British National Formulary recommends Hepatic impairment. Mizolastine is contraindicated by significant 0Æ5 mL of 1 : 1000 (500 lg) epinephrine by intramuscular hepatic impairment. Alimemazine should be avoided in hepa- injection for adults and adolescents older than 12 years. Fixed- tic impairment because it is hepatotoxic and may precipitate dose epinephrine pens delivering 300 lg for adults or 150 lg coma in severe liver disease. Chlorphenamine and hydroxyzine in children between 15 and 30 kg may be carried by patients should also be avoided in severe liver disease because their for emergency self-administration if the history indicates that sedating effect is inappropriate. the individual is at risk of further life-threatening attacks. If after the first dose of epinephrine there is no significant relief Antihistamines in pregnancy. It is best to avoid all antihistamines of symptoms, a further dose should be given. Epinephrine is in pregnancy, especially during the first trimester, although not considered helpful for angio-oedema caused by C1 inh none has been shown to be teratogenic in humans. Hydroxy- deficiency (Quality of evidence III). There is currently no licensed zine is the only antihistamine to be specifically contraindi- epinephrine aerosol inhaler available in the U.K., although Pri- cated during the early stages of pregnancy in its current U.K. matene Mist (Wyeth, Madison, NJ, U.S.A.) is available as a manufacturer’s Summary of Product Characteristics. Avoidance named patient import from the U.S.A. where it is licensed for or caution is recommended for the others, particularly in the asthma. It should be sprayed directly on to the affected area of first trimester and during lactation. Chlorphenamine is often the mouth rather than inhaled or used sublingually with the chosen by clinicians in the U.K. when antihistamine therapy intention of achieving systemic absorption. is necessary because of its long safety record. Loratadine and cetirizine are classified as U.S. Food and Drug Administration Immunomodulating therapies Pregnancy Category B drugs, implying there is no evidence of harm to the fetus during pregnancy, although well-con- Ciclosporin has been the best studied immunsuppressive drug trolled studies in humans are not available to exclude harmful for COU to date. It was effective in about two thirds of patients effects. with severe autoimmune urticaria unresponsive to antihist- ) amines at 4 mg kg 1 daily19 for up to 2 months (Quality of Antihistamines in childhood. None of the currently licensed antihis- evidence I, Strength of recommendation A) but only 25% of the tamines is contraindicated in children 12 years and older. As responders remained clear or much improved 4–5 months dosing and age restrictions for individual products vary in later. In a recent large multicentre study, there were fewer younger children, it is recommended that the relevant Data therapeutic failures when ciclosporin was taken for 16 weeks Sheets are consulted before prescribing. than 8 weeks.20 Optimal patient selection, dose and duration of treatment still need to be defined. Some patients with chronic urticaria without evidence of functional autoantibodies Antileukotrienes (with a negative ASST) also respond, although this is not well Antileukotrienes may be taken in addition to an H1 antihist- documented in the literature and a beneficial outcome from amine for poorly controlled urticaria but there is little evidence immunosuppressive treatment is less predictable. Similar over- that they are useful as monotherapy. They appear more likely all responses have been seen in open studies of tacrolimus21 to benefit aspirin-sensitive and ASST-positive COU than other and mycophenolate mofetil.22 Plasmapheresis23 and intra- patterns of urticaria17 but a good response is unpredictable. venous immunoglobulins24 may also be effective in severe Montelukast is usually chosen. autoimmune chronic urticaria (Quality of evidence II-ii) but are expensive and not widely available. There have been anecdotal reports of success with methotrexate25 and cyclophospha- Corticosteroids mide.26,27 Resolution of cold urticaria has been noted in a Oral corticosteroids may shorten the duration of acute urti- patient treated with omalizumab for asthma,28 improvement of caria (e.g. prednisolone 50 mg daily for 3 days in adults18) delayed pressure urticaria occurred during treatment of psoria- although lower doses are often effective. Parenteral hydrocorti- sis with etanercept29 but no improvement was seen in a patient sone is often given as an adjunct for severe laryngeal oedema with severe corticosteroid-dependent COU given rituximab.30 and anaphylaxis although its action is delayed. Short tapering courses of oral steroids over 3–4 weeks may be necessary for Other interventions urticarial vasculitis and severe delayed pressure urticaria (Qual- ity of evidence III) but long-term oral corticosteroids should not Although some food additives and natural salicylates may be used in chronic urticaria (Strength of recommendation A) except aggravate aspirin-sensitive chronic urticaria31 the value of in very selected cases under regular specialist supervision. avoidance is controversial. In one prospective open study of

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. 1121 inpatients with chronic urticaria, 73% of 64 improved within for patients with symptomatic recurring angio-oedema or 2 weeks of a strict pseudoallergen diet but confirmed exacer- related abdominal pain. Anabolic steroids are the treatment of bations on provocation testing with individual pseudoallergens choice for most adults (Quality of evidence III, Strength of recommen- were demonstrated in only 19% of them32 (Quality of evidence dation B) but should be avoided in children if possible. Viriliz- III, Strength of recommendation B). Oral sodium cromoglycate is not ing side-effects may occur even at the low doses needed absorbed from the gastrointestinal tract and is not effective for for long-term maintenance. Regular monitoring for hepatic urticaria. Nifedipine has been shown to reduce pruritus and inflammation and hepatocellular adenomas is essential. Tran- wealing in idiopathic COU33 (Quality of evidence II-i, Strength of examic acid may be used for maintenance but is contraindi- recommendation C), but the benefit in clinical practice is usually cated in patients with a history of thrombosis. Regular eye disappointing. Thyroxine treatment of euthyroid patients with examinations and liver function tests are recommended by idiopathic COU and with evidence of thyroid autoimmunity the manufacturer in the long-term treatment of HAE. Prophy- may occasionally result in improvement of urticaria34 (Quality laxis before planned surgery or dental procedures includes of evidence III, Strength of recommendation C). Although the published taking tranexamic acid 2 days before and afterwards or evidence for using sulfasalazine or dapsone in delayed pressure increasing the dose of established maintenance therapies with urticaria is anecdotal, they may be successful in otherwise cor- tranexamic acid or anabolic steroids. C1 inh concentrate ticosteroid-dependent cases. Sulfasalazine has also been should be given for emergency treatment of serious angio- reported to benefit idiopathic COU in a retrospective review35 oedema attacks or as prophylaxis before surgery, especially (Quality of evidence III, Strength of recommendation C) but there is a when intubation or dental extractions are necessary. Fresh risk of aggravating urticaria in aspirin-sensitive patients. Some frozen plasma may be used as a substitute in an emergency if patients with idiopathic COU have responded to warfarin36 C1 inh is not available. (Quality of evidence III, Strength of recommendation C). Idiopathic angio-oedema without weals may respond to tranexamic Prognosis acid37 (Quality of evidence II-ii, Strength of recommendation B). A dou- ble-blind randomized placebo-controlled study appeared to A comprehensive survey published in 1969 before the advent show a benefit from stanozolol with cetirizine over placebo of nonsedating antihistamines showed that 50% of patients with cetirizine38 (Quality of evidence II-i, Strength of recommendation with chronic urticaria attending a hospital clinic with weals C). Hydoxychloroquine improved the quality of life scores but alone were clear by 6 months. By contrast, over 50% of did not reduce the requirement for other medication in patients with weals and angio-oedema still had active disease patients with idiopathic COU.39 Psoralen photochemothera- after 5 years46 and therefore had a poorer outlook. A retro- py,40 ultraviolet B phototherapy41 and relaxation therapies42 spective survey in 1998 did not address prognosis directly but for chronic urticaria have yielded inconsistent results (Quality of found that 44% of hospitalized patients with urticaria reported evidence VI, Strength of recommendation D) although narrow-band a good response to antihistamines.47 It is possible that the ultraviolet B phototherapy may be more promising.43 Using a more potent antihistamines now available result in better dis- very potent topical steroid in a foam vehicle on the most ease control although the prognosis for complete recovery has affected area has been reported for delayed pressure urti- probably not changed over 40 years. caria,44 and some immediate benefit was noted at the site of application of a potent topical steroid under occlusion for Key points 2 weeks in patients with idiopathic COU,45 but the routine use of topical steroids is not recommended. 1 Urticaria can usually be classified on the clinical presenta- tion without extensive investigation. The weals of physical urticaria usually last less than 1 h (except delayed pressure Treatment of C1 esterase inhibitor deficiency urticaria) whereas those of ordinary urticaria typically last The management of C1 inh has been comprehensively from 2 to 24 h. Urticarial vasculitis should be sought by skin reviewed11 (Table 4). Maintenance therapy is only necessary biopsy if weals last longer.

Table 4 Summary of treatments for C1 esterase inhibitor deficiency

Drug Maintenance Short-term prophylaxis Emergency Stanozolola 2 mg alternate days to 10 mg daily 10 mg, 48 h before and after procedure – Danazolb 200 mg Mon–Fri to 400 mg daily 600 mg, 48 h before and after procedure – Tranexamic acid 0Æ5–3Æ0 g daily £ 4Æ5 g, 48 h before and after procedure – C1 esterase inhibitor concentrate – 1000 U, 1 h before procedure 500–1500 U Fresh frozen plasma – – 3 units

aNo longer available in the U.K. but obtainable through IDIS World Medicines (Weybridge, U.K.). bNot licensed for hereditary angio- oedema in the U.K. Dose ranges given are for adults only.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 1122 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al.

2 Urticaria often remains idiopathic after allergic, infectious, 9 Sabroe RA, Grattan CEH, Francis DM et al. The autologous serum physical and drug-related causes have been excluded as far as skin test: a screening test for autoantibodies in chronic idiopathic possible. At least 30% of patients with the ordinary presenta- urticaria. Br J Dermatol 1999; 140:446–53. 10 Kobza Black A, Lawlor F, Greaves MW. Consensus meeting on the tion of chronic urticaria appear to have an autoimmune aetiol- definition of physical urticarias and urticarial vasculitis. Clin Exp ogy. The ASST is a reasonably sensitive and specific marker for Dermatol 1996; 21:424–6. histamine-releasing autoantibodies in this group. 11 Gompels MM, Lock RJ, Abinun M et al. C1 inhibitor deficiency: 3 Advice on general measures and information can be helpful consensus document. Clin Exp Immunol 2005; 139:379–94. for most patients with urticaria, especially if an avoidable 12 Wedi B, Kapp A. Chronic urticaria: assessment of current treat- physical or dietary trigger can be identified. Over 40% of hos- ment. Exp Rev Clin Immunol 2005; 1:459–73. pitalized patients with urticaria show a good response to anti- 13 Spencer CM, Faulds D, Peters DH. Cetirizine. A reappraisal of its pharmacological properties and therapeutic use in selected allergic histamines, which are the mainstay of therapy. disorders. Drug 1993; 46:1055–80. 4 It has become common practice to increase the dose of sec- 14 Bousquet J, Czarlewski W, Danzig MR. Antiallergic properties of ond-generation H1 antihistamines above the manufacturer’s loratadine: a review. Adv Ther 1995; 12:283–98. licensed recommendation for patients when the potential ben- 15 Bleehen SS, Thomas SE, Greaves MW et al. Cimetidine and chlor- efits are considered to outweigh any risks. pheniramine in the treatment of chronic idiopathic urticaria: a 5 Combinations of nonsedating H1 antihistamines with other multi-centre randomized double-blind study. Br J Dermatol 1987; agents, such as H2 antihistamines, sedating antihistamines at 117:81–8. 16 Paul E, Bo¨deker RH. Treatment of chronic urticaria with terfena- night or the addition of antileukotrienes, can be useful for dine and ranitidine: a randomized double-blind study in 45 resistant cases. patients. Eur J Clin Pharmacol 1986; 31:277–80. 6 Oral corticosteroids should be restricted to short courses for 17 Di Lorenzo G, Pacor ML, Mansuetto P et al. Is there a role for anti- severe acute urticaria or angio-oedema affecting the mouth, leukotrienes in the management of urticaria? Clin Exp Dermatol 2006; although more prolonged treatment may be necessary for 31:327–34. delayed pressure urticaria or urticarial vasculitis. 18 Zuberbier T, Iffla¨nder J, Semler C, Henz BM. Acute urticaria: clini- 7 Immunomodulating therapies for chronic autoimmune urti- cal aspects and therapeutic responsiveness. Acta Derm Venereol (Stockh) 1996; 76:295–8. caria should be restricted to patients with disabling disease 19 Grattan CEH, O’Donnell BF, Francis DM et al. Randomized double- who have not responded to optimal conventional treatments. blind study of cyclosporin in chronic ‘idiopathic’ urticaria. Br J Dermatol 2000; 143:365–72. Audit points 20 Vena GA, Cassano N, Colombo D et al. Cyclosporine in chronic idio- pathic urticaria: a double-blind, randomised, placebo controlled 1 The use of investigations above the minimum standard for trial. J Am Acad Dermatol 2006; 55:705–9. the different clinical presentations of urticaria. 21 Kessel A, Bamberger E, Toubi E. Tacrolimus in the treatment of severe chronic idiopathic urticaria: an open-label prospective study. 2 Use of antihistamines above the manufacturers’ recom- J Am Acad Dermatol 2005; 52:145–8. mended dose. 22 Shahar E, Bergman R, Guttman-Yassky E, Pollack S. Treatment of severe chronic idiopathic urticaria with oral mycophenolate mofetil References in patients not responding to antihistamines and ⁄or corticosteroids. Int J Dermatol 2006; 45:1224–7. 1 Niimi N, Francis DM, Kermani F et al. Dermal mast cell activation 23 Grattan CEH, Francis DM, Slater NGP et al. Plasmapheresis for by autoantibodies against the high affinity IgE receptor in chronic severe, unremitting, chronic urticaria. Lancet 1992; 339:1078–80. urticaria. J Invest Dermatol 1996; 106:1001–6. 24 O’Donnell BF, Barr RM, Kobza Black A et al. Intravenous immuno- 2 Leznoff A, Sussman GL. Syndrome of idiopathic chronic urticaria globulin in autoimmune chronic urticaria. Br J Dermatol 1998; and angioedema with thyroid autoimmunity: a study in 90 138:101–6. patients. J Allergy Clin Immunol 1989; 84:66–71. 25 Gach JE, Sabroe RA, Greaves MW, Kobza Black A. Methotrexate- 3 Caminiti L, Passalacqua G, Magazzu G et al. Chronic urticaria and responsive chronic idiopathic urticaria: a report of two cases. associated coeliac disease in children: a case–control study. Pediatr Br J Dermatol 2001; 145:340–3. Allergy Immunol 2005; 16:428–32. 26 Bernstein JA, Garramone SM, Lower EG. Successful treatment of 4 Resch CA, Evans RR. Chronic urticaria and dental infection. Cleve autoimmune chronic idiopathic urticaria with intravenous cyclo- Clin Q 1958; 25:147–50. phosphamide. Ann Allergy Asthma Immunol 2002; 89:212–14. 5 James J, Warin RP. An assessment of the role of Candida albicans and 27 Asero R. Oral cyclophosphamide in a case of cyclosporin and food yeasts in chronic urticaria. Br J Dermatol 1971; 84:227–37. steroid-resistant chronic urticaria showing autoreactivity on auto- 6 Federman DG, Kirsner RS, Moriarty JP, Concato J. The effect of logous serum skin testing. Clin Exp Dermatol 2005; 30:578–602. antibiotic therapy for patients infected with Helicobacter pylori who 28 Boyce JA. Successful treatment of cold-induced urticaria ⁄anaphyl- have chronic urticaria. J Am Acad Dermatol 2003; 49:861–4. axis with anti-IgE. J Allergy Clin Immunol 2006; 117:1415–18. 7 Lindelo¨f B, Sigurgeirsson B, Wahlgren CF, Eklund G. Chronic urti- 29 Magerl M, Philipp S, Manasterski M et al. Successful treatment of caria and cancer: an epidemiological study study of 1155 patients. delayed pressure urticaria with anti-TNF-a. J Allergy Clin Immunol Br J Dermatol 1990; 123:453–6. 2007; 119:752–4. 8 Kozel MM, Mekkes JR, Bossuyt PMM, Bos JD. The effectiveness 30 Mallipeddi R, Grattan CEH. Lack of response of severe steroid- of a history-based diagnostic approach in chronic urticaria and dependent chronic urticaria to rituximab. Clin Exp Dermatol 2007; angioedema. Arch Dermatol 1998; 134:1575–80. 32:333–4.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 Guidelines for evaluation and management of urticaria, C.E.H. Grattan et al. 1123

31 Doeglas HMG. Reactions to aspirin and food additives in patients 45 Ellingsen AR, Thestrup-Pedersen K. Treatment of chronic idiopathic with chronic urticaria, including the physical urticarias. Br J Dermatol urticaria with topical steroids. Acta Derm Venereol (Stockh) 1996; 1975; 93:135–43. 76:43–4. 32 Zuberbier T, Chantraine-Hess S, Harmann K, Czarnetski BM. 46 Champion RH, Roberts SOB, Carpenter RG, Roger JH. Urticaria Pseudoallergen-free diet in the treatment of chronic urticaria. A and angio-oedema: a review of 554 patients. Br J Dermatol 1969; prospective study. Acta Derm Venereol (Stockh) 1995; 75:484–7. 81:588–97. 33 Bressler RB, Sowell K, Huston DP. Therapy of chronic idiopathic 47 Humphreys F, Hunter JAA. The characteristics of urticaria in 390 urticaria with nifedipine: demonstration of beneficial effect in a patients. Br J Dermatol 1998; 138:635–8. double-blinded, placebo controlled, crossover trial. J Allergy Clin Immunol 1989; 83:756–63. 34 Rumbyrt JS, Katz JL, Schocket AL. Resolution of chronic urticaria Appendix 1 in patients with thyroid autoimmunity. J Allergy Clin Immunol 1995; Strength of recommendations and quality of evidence 96:901–5. 35 McGirt LY, Vasagar K, Gober LM et al. Successful treatment of recal- citrant chronic idiopathic urticaria with sulfasalazine. Arch Dermatol Strength of recommendations 2006; 142:1337–42. A There is good evidence to support the use of the procedure 36 Parslew R, Pryce D, Ashworth J, Friedmann PS. Warfarin treatment B There is fair evidence to support the use of the procedure of chronic idiopathic urticaria and angio-oedema. Clin Exp Allergy C There is poor evidence to support the use of the procedure 2000; 30:1161–5. D There is fair evidence to support the rejection of the use 37 Munch EP, Weeke B. Non-hereditary angioedema treated with tran- of the procedure examic acid. Allergy 1985; 40:92–7. E There is good evidence to support the rejection of the use 38 Parsad D, Pandhi R, Juneja A. Stanozolol in chronic urticaria: a of the procedure double-blind, placebo-controlled trial. J Dermatol 2001; 28:299– Quality of evidence 302. I Evidence obtained from at least one properly designed, 39 Reeves GEM, Boyle MJ, Bonfield J et al. Impact of hydoxychloro- randomized controlled trial quine on chronic urticaria: chronic autoimmune urticaria study II-i Evidence obtained from well-designed controlled trials and evaluation. Int J Med 2004; 34:182–6. without randomization 40 Olafsson JH, Larko¨ O, Roupe G et al. Treatment of chronic urticaria II-ii Evidence obtained from well-designed cohort or case– with angio-oedema with PUVA or UVA plus placebo: a double- control analytical studies, preferably from more than blind study. Arch Dermatol Res 1986; 278:228–31. one centre or research group 41 Hannuksela M, Kokkonen E-L. Ultraviolet light therapy in chronic II-iii Evidence obtained from multiple time series with urticaria. Acta Derm Venereol (Stockh) 1985; 65:449–50. or without the intervention. Dramatic results in 42 Shertzer CL, Lookingbill DP. Effects of relaxation therapy and uncontrolled experiments (such as the results of the hypnotizability in chronic urticaria. Arch Dermatol 1987; 123:913– introduction of penicillin treatment in the 1940s) 16. could also be regarded as this type of evidence 43 Berroeta L, Clark C, Ibbotson SH et al. Narrow-band (TL-01) ultra- III Opinions of respected authorities based on clinical violet B phototherapy for chronic urticaria. Clin Exp Dermatol 2004; experience, descriptive studies or reports of expert 29:91–9. committees 44 Vena GA, Cassano N, D’Argento V, Milani M. Clobetasol propio- IV Evidence inadequate owing to problems of nate 0.05% in a novel foam formulation is safe and effective in the methodology (e.g. sample size, or length or short-term treatment of patients with delayed pressure urticaria: a comprehensiveness of follow-up or conflicts in evidence) randomized double-blinded placebo-controlled trial. Br J Dermatol 2006; 154:353–6.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1116–1123 CUTANEOUS BIOLOGY DOI 10.1111/j.1365-2133.2007.08196.x Cathelicidin LL-37 induces the generation of reactive oxygen species and release of human a-defensins from neutrophils Y. Zheng,* F. Niyonsaba,* H. Ushio,* I. Nagaoka, S. Ikeda, K. Okumura*§ and H. Ogawa* *Atopy (Allergy) Research Center and Departments of Dermatology, Host Defense and Biochemical Research and §Immunology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan

Summary

Correspondence Background Psoriasis is characterized by epidermal infiltration of neutrophils that Franc¸ois Niyonsaba. destroy invading microorganisms via a potent antimicrobial arsenal of oxidants E-mail: [email protected] and antimicrobial agents. In contrast to atopic dermatitis, psoriasis exhibits low levels of skin infections due to the presence of antimicrobial agents, including Accepted for publication 29 June 2007 cathelicidin LL-37. LL-37 kills a broad spectrum of microbes, and activates neu- trophil chemotaxis. Key words Objective To determine whether or not LL-37 could regulate additional neutrophil cathelicidin LL-37, a-defensin, mitogen-activated functions such as production of cytokines ⁄chemokines, reactive oxygen species protein kinase, neutrophil, reactive oxygen species and release of neutrophil antimicrobial peptides. Conflicts of interest Methods Human peripheral blood neutrophils were used in this study. The produc- None declared. tion of interleukin (IL)-8 and release of a-defensins were analysed by enzyme- linked immunosorbent assay, and real-time polymerase chain reaction (PCR) was used to quantify a-defensin gene expression. Phosphorylation of mitogen- activated protein kinase (MAPK) was determined by Western blotting. The gener- ation of reactive oxygen species was examined using flow cytometry, and intracellular Ca2+ mobilization was measured using a calcium assay kit. Results LL-37 enhanced the production of IL-8 under the control of MAPK p38 and extracellular signal regulated kinase (ERK), as evidenced by the inhibitory effects of p38 and ERK1 ⁄2 inhibitors on LL-37-mediated IL-8 production. Fur- thermore, LL-37 induced phosphorylation of p38 and ERK. We also revealed that LL-37 stimulated the generation of reactive oxygen species dose- and time- dependently, most probably via NADPH oxidase activation and intracellular Ca2+ mobilization. Finally, LL-37 induced both mRNA expression and protein release of a-defensins, known as human neutrophil peptide 1–3. Conclusion Taken together, we suggest that in addition to its microbicidal proper- ties, LL-37 may contribute to innate immunity by enhancing neutrophil host defence functions at inflammation and ⁄or infection sites.

The skin participates in the innate immune system not only been found in various mammals such as the pig, sheep and through its physical barrier against invading microorganisms, cows, humans have only one cathelicidin gene.3 The unique but also via the production of a vast arsenal of antimicrobial human cathelicidin known as human cationic antibacterial peptides and proteins. Among these microbicidal agents, it protein of 18 kDa (hCAP18) has been identified thus far, is increasingly evident that defensins and cathelicidins are and its mature antibacterial peptide is termed LL-37, because key players in skin-mediated host defence.1 Cathelicidins are it begins with two leucine residues, and is 37 amino acid characterized by an N-terminal signal peptide, a highly con- residues long. LL-37 was first identified in neutrophils and served cathelin domain and a structurally variable cationic later shown to be expressed in various epithelia, lympho- antimicrobial peptide at the C-terminus that becomes active cytes, monocytes, keratinocytes and various glands.3 Process- after cleavage.2 Although multiple cathelicidin genes have ing of LL-37 from the cathelicidin precursor is essential for

2007 The Authors 1124 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al. 1125 activation of its biological activity, and is accomplished by Materials and methods proteinase 3.4 LL-37 has a broad-spectrum antimicrobial activity against a variety of Gram-positive and Gram-negative Reagents bacterial, fungal and viral pathogens.1,3 Recently, it is has been proven that LL-37 not only acts as Synthetic LL-37 (LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPR- an endogenous antibiotic, but also displays additional roles TES) and its scrambled form, sLL-37 (RSLEGTDRFPFVRLK- such as regulation of inflammatory and immune responses, NSRKLEFKDIKGIKREQFVKIL),17 were synthesized by the chemoattraction of immune or inflammatory cells to wound solid-phase method on a peptide synthesizer (model PSSM-8; or infection ⁄inflammation sites, acceleration of angiogenesis, Shimadzu, Kyoto, Japan). Rabbit polyclonal antiphosphorylated promotion of wound healing and re-epithelization, and p38 and ERK1 ⁄2 antibodies, and p38 and ERK1 ⁄2 antibodies binding and neutralizing of lipopolysaccharides.1 In fact, we were obtained from Cell Signaling Technology (Beverly, MA, have previously shown that LL-37 chemoattracts mast cells,5 U.S.A.). The inhibitors SB203580 (Sigma-Aldrich, St Louis, activates these cells to release proinflammatory mediators,6 MO, U.S.A.) and PD98059 (Cell Signaling Technology) were and enhances vascular permeability in vivo through activation used to study the MAPK pathway involved in the activation of of mast cells.7 In addition, LL-37 has been reported to stim- neutrophils. 5-(and 6-) Chloromethyl-2’,7’-dichlorohydrofluo- ulate monocytes, T cells, dendritic cells and keratinocytes, rescein diacetate acetyl ester (CM-H2DCFDA) was obtained suggesting its crucial role in both innate and adaptive from Molecular Probes (Eugene, OR, U.S.A.), and diphenylene immunity.8–10 The in vivo importance of LL-37 in prevention iodonium chloride (DPI) and 1,2-bis (o-amino phenoxyl) eth- of infections is supported by its low expression in skin ane-N,N,N’,N’-tetraacetic acid acetoxymethyl ester (BAPTA- lesions caused by atopic dermatitis, correlating with AM) were obtained from Sigma-Aldrich. enhanced susceptibility to skin infections in this disease.11,12 Moreover, the deficiency of LL-37 in patients with a congen- Purification and stimulation of peripheral blood ital neutropenia known as morbus Kostmann, is thought to neutrophils be an origin of severe infection in this disorder.13 Further direct evidence of the importance of cathelicidins was pro- According to Juntendo University ethical committee approval, ven by the finding that mice lacking CRAMP (cathelicidin- informed consent was obtained from healthy volunteers, and related antimicrobial peptide), a LL-37 homologue, were blood was drawn from the cubital vein. Neutrophils were iso- more susceptible to skin infections caused by group A strep- lated from heparinized blood by polymorphprep (Axis-Shield, tococci than wild-type mice.14 Consistently, wounds in pigs Oslo, Norway) following the manufacturer’s instructions, and in which the processing of porcine cathelicidins to their the purity was >98%. Cells at a final concentration of 2 · 106 ) mature peptides was inhibited also showed increased suscep- cells mL 1 were treated with various doses of LL-37 at 37 C tibility to infection.15 for 2–12 h in RPMI 1640 medium (Nissui Pharmaceutical, Neutrophils are the effector cells in innate immunity, and Tokyo, Japan) supplemented with 10% fetal calf serum (FCS)

LL-37 has been reported to induce the migration of these in 5% CO2 atmosphere. Following incubation, cells were cen- cells.8 Recruited neutrophils may also release additional host trifuged, and the supernatants were used for enzyme-linked defence modulators such as antimicrobial peptides, a-defen- immunosorbent assay (ELISA), while the cell pellet was used sins and cathelicidins, to amplify innate immune responses for total RNA extraction. against invading pathogens. Furthermore, it was recently shown that LL-37 suppresses apoptosis of neutrophils, result- Enzyme-linked immunosorbent assay ing in the prolongation of their life span and strengthening their host defence against microbial invasion.16 IL-8 and HNP1–3 released in the cell-free supernatants from Based on these observations, and the fact that LL-37 is nonstimulated or stimulated cultures with various doses of highly overexpressed in active psoriatic lesions where high LL-37 for the indicated time periods were measured with amounts of infiltrating neutrophils have been observed, we appropriate ELISA kits from R&D Systems (Minneapolis, MN, speculated that LL-37 might have additional modulating U.S.A.) for IL-8, and from HyCult Biotechnology (Uden, the activities on human neutrophil functions. Supporting our Netherlands) for HNP1–3. Supernatants were stored at )20 C hypothesis, we demonstrated that LL-37 stimulates neutro- until used for ELISA according to the manufacturer’s instruc- phil interleukin (IL)-8 production via activation of mitogen- tions. In some experiments, cells were pretreated with activated protein kinase (MAPK) p38 and extracellular signal PD98059 or SB203580 for 2 h prior to stimulation with regulated kinase (ERK), and mediates the generation of LL-37, and ELISA was performed as above. reactive oxygen species (ROS) most probably via nicotin- amide adenine dinucleotide phosphate (NADPH) oxidase Western blot analysis and intracellular Ca2+ mobilization. In addition, LL-37 ) induces mRNA expression as well as extracellular release Neutrophils were incubated with 20 lgmL 1 LL-37 for the of a-defensins known as human neutrophil peptide (HNP) indicated time periods. After stimulation, the lysates were ) 1–3. obtained by lysing cells in lysis buffer [50 mmol L 1 Tris–HCl

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 1126 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al.

)1 (pH 8), 150 mmol L NaCl, 0Æ02% NaN3,0Æ1% sodium according to the manufacturer’s instructions. First-strand ) dodecyl sulphate (SDS), 1% NP-40, containing 1 lmol L 1 cDNA was synthesized from 3 lg of total RNA with oligo )1 - phenylmethylsulphonyl fluoride (PMSF), 10 lgmL leu- (dT)12–18 primers using Superscript II RNase H reverse trans- ) ) peptin, 10 lgmL 1 pepstatin-A, 50 lgmL 1 aprotinin and criptase (Life Technologies), as described before.9,10 Real-time ) 2 mmol L 1 sodium orthovanadate]. Equal amounts of total PCR was performed using the TaqMan Universal PCR Master protein were subjected to 12Æ5% SDS-polyacrylamide gel elec- Mix (Applied Biosystems, Branchburg, NJ, U.S.A.). Amplifica- trophoresis. Then, nonspecific binding sites were blocked, and tion and detection of HNP1–3 mRNA were analysed by 7500 the blots were incubated with polyclonal antibodies against Real-Time PCR System (Applied Biosystems), according to the phosphorylated p38 and ERK1 ⁄2 or unphosphorylated p38 manufacturer’s specifications. HNP1–3 primer ⁄probe set was and ERK1 ⁄2 overnight, according to the manufacturer’s obtained from Applied Biosystems Assays-on-DemandTM.To instructions. The membrane was developed with an enhanced standardize mRNA concentrations, transcript levels of the chemiluminescence detection kit (Amersham Pharmacia Bio- housekeeping gene GAPDH were determined in parallel for tech, Piscataway, NJ, U.S.A.). each sample, and relative transcript levels were corrected by normalization based on the GAPDH transcript levels. Changes in gene expression were reported as fold increases relative to Measurement of intracellular reactive oxygen species untreated controls. production

) Neutrophils at a density of 1 · 106 cells mL 1 were sus- Statistical analysis pended in 150 lL of RPMI 1640 containing 10% FCS, and in- cubated at 37 C for the time indicated, with or without The statistical analysis between groups (stimulated vs. nonsti- 50 lL of titrated concentrations of LL-37. The cells were mulated neutrophils) from different donors was performed ) washed and resuspended in Opti-MEM containing 1 lmol L 1 using a two-sided Student’s t-test, and results were considered DCFDA, a fluorescent dye used for detection of ROS oxida- to be statistically significant for P <0Æ05. The results are tion.18,19 Following incubation for 30 min at 37 C, cells shown as mean ± SD. were washed in ice-cold PBS, and then analysed by a FACScan (Becton-Dickinson, Bedford, U.K.). In inhibition experiments, Results the agents tested were added to cells 30 min before stimula- tion with LL-37. LL-37 induces interleukin 8 production by neutrophils

2 To determine the stimulatory activities of LL-37 on human Measurements of intracellular Ca + mobilization neutrophils, we first evaluated its ability to induce the produc- Intracellular Ca2+ mobilization from neutrophils was measured tion of proinflammatory cytokines and chemokines from neu- by a no-washing method using a FLIPR Calcium Assay Kit trophils, as determined by ELISA. As shown in Figure 1, the ) (Molecular Devices, Sunnyvale, CA, U.S.A.). Neutrophils activation of neutrophils with 10–40 lgmL 1 LL-37 resulted (100 lL) were seeded at a density of 2 · 105 cells per well in a dose-dependent production of IL-8. Furthermore, the into poly-D-lysine-coated 96-well black-walled clear bottom microtitre plate (Becton-Dickinson). Cells were then loaded for 1 h at 37 C in an equivalent volume of Hank’s balan- 100 ** )1 ced salt solution (HBSS) containing 20 mmol L HEPES, ) 80 ** ) –1 2Æ5 mmol L 1 probenecid (Sigma-Aldrich), and Calcium 3 60 Reagent (Molecular Devices, Menlo Park, CA, U.S.A.) at pH 7Æ4, prepared according to the manufacturer’s instructions. To 40 * * * * form a uniform monolayer of cells on the bottom of the wells, IL-8 (pg mL 20 the microplate was gently centrifuged for 5 min with low 0 2 4 812 acceleration and without break. The cell-containing plate was placed into FlexStation II (Molecular Devices), and a volume of Time (h) 50 lL per well of LL-37 diluted in assay buffer was added to achieve the final indicated concentration. Maximum change in Fig 1. LL-37 induces the generation of interleukin (IL)-8 from neutrophils. Human neutrophils were stimulated with LL-37 at fluorescence over baseline was used to determine agonist ) ) concentrations of 10 lgmL 1 (horizontal stripes), 20 lgmL 1 response, as quantified using SoftMax Pro (version 5) software. ) ) (diagonal stripes) or 40 lgmL 1 (zigzag lines), or with 40 lgmL 1 of a scrambled form of LL-37, sLL-37 (dots) for 2–12 h. The amounts Total RNA extraction and real-time quantitative of IL-8 released into the culture supernatants were determined by polymerase chain reaction enzyme-linked immunosorbent assay. Values are compared between stimulated and nonstimulated cells (incubated with medium alone, Total RNA was extracted from neutrophils using TRIzol open bars). *P <0Æ05, **P <0Æ01. Each bar represents the reagent (BRL, Life Technologies, Rockville, MD, U.S.A.), mean ± SD of five independent experiments from five donors.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al. 1127 stimulatory effect of LL-37 seemed to be time-dependent, (a) 0 5 10 20 30 (min) reaching a maximum at 12 h of incubation. Longer incubation (18 and 24 h) of neutrophils with LL-37 also resulted in more p-p38 enhanced production of IL-8; however, this production was p38 accompanied by a dramatic decrease of cell viability, as evalu- ated by trypan blue exclusion (data not shown). A scrambled p-ERK1/2 peptide with the same amino acid content but not the same sequence as LL-37, sLL-37, was used as a control to exclude a ERK1/2 nonspecific effect of LL-37. sLL-37 had no effect on IL-8 pro- duction by neutrophils. (b) 80

LL-37 induces phosphorylation of mitogen-activated ) –1 60 protein kinase p38 and extracellular signal regulated ** kinase required for interleukin 8 production 40

We next explored the mechanism by which LL-37 induces 20 ** IL-8 (pg mL IL-8 production by investigating its effect on phosphorylation 0 of MAPK in neutrophils. Cells were stimulated with Med Ctrl SB PD ) 20 lgmL 1 LL-37 for 5–30 min, and both the phosphory- lated and unphosphorylated p38 and ERK1 ⁄2 levels were Fig 2. LL-37 induces phosphorylation of mitogen-activated protein determined using Western blot analysis. LL-37 markedly kinase (MAPK) p38 and extracellular signal regulated kinase (ERK), induced phosphorylation of p38 at 10 min, reaching a peak at which are further required for neutrophil activation. (a) Neutrophils 6 )1 )1 20 min, as compared with nonstimulated cells. Similar to (2 · 10 cells mL ) were stimulated with 20 lgmL LL-37 for the p38, the phosphorylation of ERK1 ⁄2 was also observed at 10 indicated time periods, and phosphorylated p38 (p-p38) or ERK1 ⁄2 and 20 min, before drastically decreasing at 30 min (Fig. 2a). (p-ERK1 ⁄2) and unphosphorylated p38 (p38) or ERK1 ⁄2 (ERK1 ⁄2) levels in cellular lysates were determined by Western blot analysis. To determine whether activation of p38 and ERK1 ⁄2 was Shown is one representative of six separate experiments with required for LL-37-mediated production of IL-8, cells were similar results. (b) To determine the role of p38 and ERK in LL-37- incubated with specific inhibitors for p38 or ERK1 ⁄2 for 2 h mediated neutrophil activation, neutrophils were preincubated with ) ) before stimulation with LL-37. The presence of SB203580 20 lmol L 1 of p38-specific inhibitor SB203580 (SB), 20 lmol L 1 (p38 inhibitor) and PD98059 (ERK1 ⁄2 inhibitor) almost of ERK1 ⁄2-specific inhibitor PD98059 (PD), or medium alone (Ctrl, completely suppressed the production of IL-8 (Fig. 2b). We control) for 2 h, after which the cells were stimulated for 12 h with ) confirmed that the treatment of neutrophils with MAPK 40 lgmL 1 LL-37. Interleukin (IL)-8 released into supernatants was inhibitors did not affect cell viability (data not shown). determined by enzyme-linked immunosorbent assay. Values are the mean ± SD of four separate experiments from different donors. **P <0Æ01 as compared between the presence and absence of each LL-37 stimulates the generation of reactive oxygen MAPK inhibitor. species by neutrophils

The kinetics of LL-37-stimulated ROS generation in neutroph- ils was measured using the cell permeant, oxidation-sensitive selective species of oxidants. DPI, a broad-spectrum inhibitor dye DCFDA. In dose–response experiments, we found that of flavoprotein-containing oxidoreductases, has been reported LL-37 significantly induced DCFDA oxidation in a dose- to reduce the release of ROS by blocking NADPH oxidase ) dependent fashion from 5 to 20 lgmL 1 (Fig. 3a). As activity.21 Consistently, various doses of DPI markedly inhibit- ) expected, the control peptide sLL-37 (20 lgmL 1) had no ed LL-37-mediated ROS production in a dose-dependent man- ) effect on generation of ROS by neutrophils. Furthermore, ner (Fig. 4a). DPI at concentrations as low as 2Æ5 lmol L 1 LL-37 induced significant increases in DCFDA oxidation in a was effective, and the inhibitory effect was maximal at ) time-dependent manner from 2Æ5 to 20 min, with a maximal 10 lmol L 1. Because the intracellular Ca2+ chelator BAPTA- oxidation at 20 min (Fig. 3b). Using dihydroethidium (DHE) AM has been shown to inhibit ROS generation potently in as a selective probe for superoxide anion,20 LL-37 could only neutrophils,22 its possible involvement was also examined. As slightly induce DHE oxidation at various doses and time seen in Figure 4b, BAPTA-AM also markedly reduced LL-37- periods tested (data not shown). induced ROS generation. To confirm the role of intracellular Ca2+in LL-37-induced ROS generation, the effect of LL-37 on intracellular Ca2+ Effects of antioxidants on LL-37-mediated reactive mobilization was tested. Supporting our observation, LL-37 oxygen species generation significantly increased intracellular Ca2+ concentrations in neu- To further characterize the ROS generated upon LL-37 stimu- trophils in a dose–response fashion (Fig. 5). Together, these lation in neutrophils, antioxidants were used as scavengers of results suggest that the LL-37-mediated ROS generation is

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 1128 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al.

(a) 100 (a) 30 5 µg mL–1 20 µg mL–1 80 20 sLL-37 * * * * 60 10

40 Med Fluoresence mean 0 10 µg mL–1 0102030 40 20 DPI (µmol L–1) (b) 0 30 100 101 102 103 104 DCFDA fluoresence intensity 20

(b) 100 * 10 * * * * 80 Fluoresence mean 0 0 010203040 60 20 2·5 BAPTA-AM (µmol L–1) 10

Counts40 Counts 5 Fig 4. Effects of diphenylene iodonium chloride (DPI) and 1,2-bis (o-amino phenoxyl) ethane-N,N,N’,N’-tetraacetic acid acetoxymethyl 20 ester (BAPTA-AM) on LL-37-mediated reactive oxygen species generation. Dichlorofluorescein diacetate (DCFDA) oxidation was ) 0 determined after 20-min stimulation with 20 lgmL 1 LL-37 in 0 1 2 3 4 10 10 10 10 10 the presence or absence of titrated concentrations of (a) DPI or DCFDA fluoresence intensity (b) BAPTA-AM. Data are corrected from spontaneous DCFDA oxidation, and are expressed as fluorescence mean. The data represent Fig 3. LL-37 induces reactive oxygen species (ROS) generation in the average of six experiments. *P <0Æ01 as compared between the neutrophils dose- and time-dependently. (a) Representative FACS presence and absence of inhibitor. profile of four independent experiments for various doses of LL-37- induced dichlorofluorescein diacetate (DCFDA) oxidation at 20 min with nonstimulated cells (black lines, Med: medium alone); ) 5 lgmL 1 LL-37-induced ROS generation is represented by red lines, 50 000 40 µg mL–1 )1 )1 10 lgmL (blue lines), 20 lgmL (orange lines), whereas 20 40 000 )1 lgmL sLL-37 used as a control for LL-37 is represented by green –1 30 000 20 µg mL lines. (b) Representative data of five separate time-course experi- ) ments of 20 lgmL 1 LL-37-induced ROS oxidation. Nonstimulated 20 000 10 µg mL–1 (0, 0 min) are represented by black lines, 2Æ5-min incubation 10 000 (2Æ5, green lines), 5-min incubation (5, red lines), 10-min incubation

(10, blue lines) and 20-min incubation (20, orange lines). Fluorescence change 0 0 60 120 180

Time (s) mediated by a flavoenzyme, most probably an NADPH oxi- 2+ dase, and via increase of intracellular Ca2+ concentrations. Fig 5. LL-37 induces intracellular Ca mobilization in neutrophils. Neutrophils (2 · 105 cells per well) were incubated for 1 h at 37 C in Hank’s balanced salt solution containing HEPES, probenecid and ) LL-37 increases the gene expression and extracellular Calcium 3 Reagent, and then stimulated with 10–40 lgmL 1 LL-37 release of human a-defensins as described in Materials and methods. Data are representative of four separate experiments from different donors, and shown as It has been shown previously that stimulation of neutrophils fluorescence change corrected from spontaneous fluorescence by from guinea pigs resulted in extracellular release of antimicro- nonstimulated cells. Arrow indicates the addition of LL-37. bial agents such as a-defensins and cathelicidin CAP11.23 Thus, we envisaged that LL-37 might also activate human neu- trophils to produce and release a-defensins. To this end, we with predominance of HNP1, 2 and 3.24–26 Incubation of ) first evaluated the effect of LL-37 on mRNA expression of neutrophils with 10–40 lgmL 1 of LL-37 resulted in dose- HNPs, a type of a-defensins located in neutrophil granules, dependent increases of gene expression of HNP1–3. This effect

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al. 1129

such as human b-defensins and LL-37 in psoriatic lesions. For example, the concentrations of LL-37 have been estimated at ) ) ~304 lmol L 1 (1Æ38 mg mL 1) in psoriatic skin lesions,11 ) ) and 1 lmol L 1 (4Æ5 lgmL 1) in normal skin.27 However, human b-defensins and LL-37 are downregulated in atopic dermatitis, which demonstrates increased susceptibility to bac- terial and viral infections.11,12 Since psoriasis is histopathologi- cally characterized by a preferential epidermal infiltration of neutrophils,28 we anticipated that LL-37 might have stimula- tory effects on neutrophils, in addition to the chemotaxis and suppression of neutrophil apoptosis reported previously.8,16 Supporting our hypothesis, the current study revealed that LL-37 mediates the production of IL-8 from neutrophils through activation of MAPK p38 and ERK, and induces ROS generation via activation of an NADPH oxidase and intra- cellular Ca2+ mobilization. Furthermore, LL-37 increases mRNA expression and extracellular release of neutrophil- derived antimicrobial peptides HNP1–3. It has been reported that after the onset of an inflammatory reaction, large numbers of neutrophils are still present at the inflamed site long after the cessation of neutrophil influx.29 Fig 6. LL-37 increases the gene expression and protein release These infiltrating cells have been shown to be activated by a of human neutrophil peptide (HNP) 1–3. (a) Neutrophils were wide variety of stimuli to release inflammatory mediators such )1 incubated with medium alone (open bars), 40 lgmL sLL-37 (dots) as cytokines and chemokines. In the current study, we showed )1 )1 and LL-37 at 10 lgmL (horizontal stripes), 20 lgmL (diagonal that LL-37 activates highly purified neutrophils to produce l )1 stripes) or 40 gmL (zigzag lines) for 2–12 h. Following the IL-8 selectively, but not other cytokines or chemokines such incubation, total RNA was extracted, converted into cDNA, and real- as IL-6, tumour necrosis factor (TNF)-a, macrophage inflam- time polymerase chain reaction was performed to analyse the changes matory protein (MIP)-1a, MIP-1b and MIP-3a (data not in gene expression. Each bar shows the mean ± SD from five different donors. Values represent fold increases in gene expression above cells shown) that have been demonstrated to contribute to the 30,31 incubated with medium alone. (b) Supernatants from cells stimulated pathogenesis of psoriasis. This suggests that LL-37- ) with 40 lgmL 1 sLL-37 (control peptide, dots) or LL-37 at mediated IL-8 production is selective. Because IL-8 has been ) ) 10 lgmL 1 (horizontal stripes), 20 lgmL 1 (diagonal stripes) or identified as a major chemoattractant present in psoriatic ) 40 lgmL 1 (zigzag lines) for indicated time periods were used for lesions,32 LL-37 may not only directly, but also indirectly detection of released concentrations of HNP1–3 proteins by enzyme- enhance the chemotaxis of neutrophils by inducing the linked immunosorbent assay. Values are compared between stimulated production of IL-8 from this cell population. and nonstimulated cells (open bars) from various donors. *P <0Æ05, We attempted to characterize the cellular mechanisms by Æ **P <001. Each bar represents the mean ± SD of five independent which LL-37 activates neutrophils by focusing on the path- experiments. ways of the MAPK cascades p38 and ERK, known to partici- pate in a large variety of cellular activities such as cell was also time-dependent, reaching a peak at 8 h before survival and proliferation, and the expression of proinflam- decreasing after a 12-h incubation (Fig. 6a). matory cytokines and chemokines.33,34 Although the activa- Because LL-37 increases the gene expression of HNP1–3, we tion of MAPK by LL-37 has been reported in several postulated that it could also stimulate neutrophils to produce inflammatory and immune cells including mast cells,7 kerati- and release extracellularly their respective proteins. After stimu- nocytes9 and monocytes,35 the present study is the first lation of cells with various doses of LL-37 for 2–12 h, the report to provide evidence that LL-37 also induces phos- release of HNPs in cell-free supernatants was determined by a phorylation of p38 and ERK in human neutrophils. Both specific ELISA kit. As shown in Figure 6b, LL-37 dose-depen- p38 and ERK were required for LL-37-mediated neutrophil dently induced the extracellular release of HNP1–3 after 8- and activation, as evidenced by the inhibitory effects of 12-h incubations. Thus, LL-37 specifically enhances both the p38- and ERK-specific inhibitors on IL-8 production. We gene expression and release of HNPs; its control sLL-37 had no confirmed that the specific inhibitor for c-Jun N-terminal stimulatory effect on neutrophils (Fig. 6a,b). kinase could not suppress the neutrophil activation by LL-37 (data not shown), ruling out its involvement in the stimula- Discussion tory effect of LL-37 on neutrophils. Being the archetypical phagocytic cells that actively destroy In contrast to atopic dermatitis, psoriatic skin exhibits low pathogenic microorganisms, neutrophils deploy a potent levels of infection, due to the presence of antimicrobial agents antimicrobial arsenal that includes oxidants and microbicidal

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 1130 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al. peptides.36 Neutrophils possess a multicomponent enzyme References complex termed the NADPH oxidase which, when activated, 1 Niyonsaba F, Nagaoka I, Ogawa H. Human defensins and cathelici- generates large quantities of ROS that contribute to host dins in the skin: beyond direct antimicrobial properties. Crit Rev 37,38 defence. The evidence that LL-37 is involved in neutrophil Immunol 2006; 26:545–76. host defence functions was first demonstrated by finding that 2 Zanetti M, Gennaro R, Romeo D. Cathelicidins: a novel protein this peptide stimulated neutrophils to generate significant family with a common proregion and a variable C-terminal antimi- amounts of ROS. LL-37-mediated ROS generation was most crobial domain. FEBS Lett 1995; 374:1–5. probably via NADPH oxidase activation as evidenced by the 3 Niyonsaba F, Hirata M, Ogawa H, Nagaoka I. Epithelial cell-derived b inhibitory effect of DPI, which is known to suppress NADPH antibacterial peptides human -defensins and cathelicidin: multi- functional activities on mast cells. Curr Drug Targets Inflamm Allergy oxidase. Moreover, the observation that an intracellular Ca2+ 2003; 2:224–31. chelator BAPTA-AM markedly reduced LL-37-induced ROS 4 Sørensen OE, Follin P, Johnsen AH et al. Human cathelicidin, generation suggested that this generation was also dependent hCAP-18, is processed to the antimicrobial peptide LL-37 on intracellular Ca2+ mobilization. Since ROS generation can by extracellular cleavage with proteinase 3. Blood 2001; 97: also be mediated by mitochondrial respiration39 or xanthine 3951–9. oxidase,40 neutrophils were pretreated with rotenone (a po- 5 Niyonsaba F, Iwabuchi K, Someya A et al. A cathelicidin family of tent inhibitor of the mitochondrial electron transport chain) human antibacterial peptide LL-37 induces mast cell chemotaxis. Immunology 2002; 106:20–6. or allopurinol (inhibitor for xanthine oxidase). Neither rote- 6 Niyonsaba F, Someya A, Hirata M et al. Evaluation of the effects none nor allopurinol affected LL-37-mediated ROS generation of peptide antibiotics human b-defensins-1 ⁄-2 and LL-37 on

(data not shown), which excluded the involvement of mito- histamine release and prostaglandin D2 production from mast cells. chondrial respiration or xanthine oxidase in LL-37-induced Eur J Immunol 2001; 31:1066–75. ROS generation. 7 Chen X, Niyonsaba F, Ushio H et al. Human cathelicidin LL-37 The contribution of LL-37 to the neutrophil-mediated host increases vascular permeability in the skin via mast cell activation, defence was further confirmed by its ability to induce the and phosphorylates MAP kinases p38 and ERK in mast cells. J Dermatol Sci 2006; 43:63–6. release of neutrophil antimicrobial granule components. 8 Yang D, Chen Q, Schmidt AP et al. LL-37, the neutrophil granule- Among neutrophil granule components, four different human and epithelial cell-derived cathelicidin, utilizes formyl peptide a-defensin molecules (HNP1, HNP2, HNP3 and HNP4) have receptor-like 1 (FPRL1) as a receptor to chemoattract human been described so far, and shown to play a crucial role in peripheral blood neutrophils, monocytes, and T cells. J Exp Med both innate and adaptive immunity.24–26,41 HNP1, HNP2 and 2000; 192:1069–74. HNP3, which differ only in the first amino acid, account for 9 Niyonsaba F, Ushio H, Nagaoka I et al. The human b 5–7% of total neutrophil proteins.25 In contrast, HNP4, with -defensins (hBD-1, -2, -3, -4) and cathelicidin LL-37 induce interleukin-18 secretion through p38 and ERK MAPK activation an amino acid sequence distinct from other HNP sequences, in primary human keratinocytes. J Immunol 2005; 175:1776–84. 41 comprises less than 2% of total defensins in neutrophils. The 10 Niyonsaba F, Ushio H, Nakano N et al. Antimicrobial peptides current study demonstrated that LL-37 significantly enhanced human b-defensins stimulate epidermal keratinocyte migration, both HNP1–3 gene expression and their respective protein proliferation and production of proinflammatory cytokines and release. Thus, LL-37 activates both the oxidative (ROS) and chemokines. J Invest Dermatol 2007; 127:594–604. nonoxidative (HNPs) killing mechanisms of neutrophils to 11 Ong PY, Ohtake T, Btandt C et al. Endogenous antimicrobial pep- support host defence. tides and skin infections in atopic dermatitis. N Engl J Med 2002; 347:1151–60. Taken together, our finding that LL-37 activates human 12 Howell MD, Novak N, Bieber T et al. Interleukin-10 downregulates neutrophils to generate cytokines and ROS, and to extracellu- anti-microbial peptide expression in atopic dermatitis. J Invest Derma- larly release and antimicrobial peptides provides insight into tol 2005; 125:738–45. the novel mechanism by which LL-37 contributes to the mod- 13 Putsep K, Carlsson G, Boman HG, Andersson M. Deficiency of anti- ulation of host defence, particularly at inflammation and infec- bacterial peptides in patients with morbus Kostmann: an observa- tion sites. tion study. Lancet 2002; 360:1144–9. 14 Nizet V, Ohtake T, Lauth X et al. Innate antimicrobial peptide pro- tects the skin from invasive bacterial infection. Nature 2001; Acknowledgments 414:454–7. 15 Cole AM, Shi J, Ceccarelli A et al. Inhibition of neutrophil elastase We would like to thank the members of Atopy (Allergy) prevents cathelicidin activation and impairs clearance of bacteria Research Center and Department of Immunology of Juntendo from wounds. Blood 2001; 97:297–304. University School of Medicine for their excellent discussion, 16 Nagaoka I, Tamura H, Hirata M. An antimicrobial cathelicidin pep- comments and encouragement. We are also very grateful to all tide, human CAP18 ⁄LL-37, suppresses neutrophil apoptosis via the volunteer blood donors, and to Michiyo Matsumoto for secre- activation of formyl-peptide receptor-like 1 and P2X7. J Immunol 2006; 176:3044–52. tarial assistance. This work was supported in part by Grant-in- 17 Kandler K, Shaykhiev R, Kleemann P et al. The antimicrobial pep- Aid for Scientific Research from the Ministry of Education, tide LL-37 inhibits the activation of dendritic cells by TLR ligands. Culture, Sports, Science and Technology, Japan to F.N., and Int Immunol 2006; 18:1729–36. Atopy (Allergy) Research Center, Juntendo University, Tokyo, 18 LeBel CP, Ischiropoulos H, Bondy SC. Evaluation of the probe Japan. 2’,7’-dichlorofluorescin as an indicator of reactive oxygen

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 Effects of cathelicidin LL-37 on neutrophils, Y. Zheng et al. 1131

species formation and oxidative stress. Chem Res Toxicol 1992; cytes ⁄macrophage colony-stimulating factor. J Exp Med 1999; 5:227–31. 190:923–34. 19 Crow JP. Dichlorodihydrofluorescein and dihydrorhodamine 123 30 Arican O, Aral M, Sasmaz S, Ciragil P. Serum levels of TNF-a, are sensitive indicators of peroxynitrite in vitro: implications for IFN-c, IL-6, IL-8, IL-12, IL-17, and IL-18 in patients with active intracellular measurement of reactive nitrogen and oxygen species. psoriasis and correlation with disease severity. Mediators Inflamm Nitric Oxide 1997; 1:145–57. 2005; 2005:273–9. 20 Rothe G, Valet G. Flow cytometric analysis of respiratory burst 31 Pastore S, Mascia F, Mariotti F et al. Chemokine networks in inflam- activity in phagocytes with hydroethidine and 2’,7’-dichlorofluore- matory skin diseases. Eur J Dermatol 2004; 14:203–8. scin. J Leukoc Biol 1990; 47:440–8. 32 Duan H, Koga T, Kohda F et al. Interleukin-8-positive neutrophils 21 Ellis JA, Mayer SJ, Jones OT. The effect of the NADPH oxidase in psoriasis. J Dermatol Sci 2001; 26:119–24. inhibitor diphenyleneiodonium on aerobic and anaerobic microbi- 33 Ballif BA, Blenis J. Molecular mechanisms mediating mammalian cidal activities of human neutrophils. Biochem J 1988; 251:887–97. mitogen-activated protein kinase (MAPK) kinase (MEK)-MAPK cell 22 Devadas S, Zaritskaya L, Rhee SG et al. Discrete generation of super- survival signals. Cell Growth Differ 2001; 12:397–408. oxide and hydrogen peroxide by T cell receptor stimulation: selec- 34 Chang L, Karin M. Mammalian MAP kinase signaling cascades. tive regulation of mitogen-activated protein kinase activation and Nature 2001; 410:37–40. Fas ligand expression. J Exp Med 2002; 195:59–70. 35 Bowdish DM, Davidson DJ, Speert DP, Hancock RE. The human 23 Yomogida S, Nagaoka I, Yamashita T. Comparative studies on the cationic peptide LL-37 induces activation of the extracellular sig- extracellular release and biological activity of guinea pig neutrophil nal-regulated kinase and p38 kinase pathways in primary human cationic antibacterial polypeptide of 11 kDa (CAP11) and defen- monocytes. J Immunol 2004; 172:3758–65. sins. Comp Biochem Physiol B Biochem Mol Biol 1997; 116:99–107. 36 Lehrer RI, Ganz T, Selsted ME et al. Neutrophils and host defense. 24 Ganz T, Selsted ME, Szklarek D et al. Defensins. Natural peptide Ann Intern Med 1988; 109:127–42. antibiotics of human neutrophils. J Clin Invest 1985; 76:1427–35. 37 Lambeth JD. NOX enzymes and the biology of reactive oxygen. 25 Selsted ME, Harwig SS, Ganz T et al. Primary structures of three Nat Rev Immunol 2004; 4:181–9. human neutrophil defensins. J Clin Invest 1985; 76:1436–9. 38 Fialkow L, Wang Y, Downey GP. Reactive oxygen and nitrogen 26 Selsted ME, Harwig SS. Determination of the disulfide array in the species as signaling molecules regulating neutrophil function. Free human defensin HNP-2: a covalently cyclized peptide. J Biol Chem Radic Biol Med 2007; 42:153–64. 1989; 264:4003–7. 39 Li Y, Trush MA. Diphenyleneiodonium, an NAD(P)H oxidase 27 Murakami M, Ohtake T, Dorschner RA et al. Cathelicidin anti- inhibitor, also potently inhibits mitochondrial reactive oxygen spe- microbial peptide expression in sweat, an innate defense system cies production. Biochem Biophys Res Commun 1998; 253:295–9. for the skin. J Invest Dermatol 2002; 119:1090–5. 40 Grum CM, Gross TJ, Mody CH, Sitrin RG. Expression of xanthine oxi- 28 Wahba A, Cohen H, Bar-Eli M, Callily R. Neutrophil chemotaxis in dase activity by murine leukocytes. J Lab Clin Med 1990; 116:211–18. psoriasis. Acta Derm Venereol 1979; 59:441–5. 41 Wilde CG, Griffith JE, Marra NN. Purification and characterization 29 Coxon A, Tang T, Mayadas N. Cytokine-activated endothelial cells of human neutrophil peptide 4, a novel member of the defensin delay neutrophil apoptosis in vitro and in vivo: a role for granulo- family. J Biol Chem 1989; 264:11200–3.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1124–1131 CUTANEOUS BIOLOGY DOI 10.1111/j.1365-2133.2007.08218.x Microarray analysis of aberrant gene expression in actinic keratosis: effect of the Toll-like receptor-7 agonist imiquimod A. Torres, L. Storey, M. Anders, R.L. Miller,* B.J. Bulbulian,* J. Jin,* S. Raghavan,* J. Lee, H.B. Slade and W. Birmachu* Dermatology Office, Loma Linda University Medical Center, Loma Linda, CA, U.S.A. *Pharmacology and Medical & Scientific Affairs, 3M Pharmaceuticals, 270-2S-06 St Paul, MN 55144, U.S.A.

Summary

Correspondence Background The molecular events leading to actinic keratosis (AK) are not well Woubalem Birmachu. understood. E-mail: [email protected] Objective To identify and compare gene expression changes in AK lesions and in sun-exposed nonlesional skin and to determine the effect of imiquimod 5% Accepted for publication 14 May 2007 cream on these changes. Method A double-blind, vehicle-controlled, randomized study was conducted to Key words evaluate the molecular changes in AK treated with imiquimod. Seventeen male actinic keratosis, confocal microscopy, gene subjects with ‡ 5 AK lesions on the scalp applied vehicle or imiquimod three expression, imiquimod, microarray times a week for 4 weeks. Gene expression analysis using Affymetrix oligonu- Conflicts of interest cleotide arrays was performed on shave biopsies of lesions taken before and after R.L.M., B.J.B., J.J., S.R., J.L., H.B.S. and W.B. treatment. Confocal microscopy was performed on the study area as an adjunc- were employed by 3M Pharmaceuticals, the tive diagnostic procedure. manufacturers of Aldara, imiquimod cream. Results We identified gene expression changes which occur in sun-exposed, non- lesional skin as well as in AK lesions. These changes include, but are not limited to, the overexpression of oncogenic and proliferative genes and diminished expression of tumour suppressor genes. The gene expression changes observed in AK lesions and in sun-exposed, nonlesional skin were consistent with the confo- cal microscopy observations, which showed abnormalities in the sun-exposed, nonlesional skin, similar in nature but less pronounced than abnormalities seen in AK. Imiquimod partially or totally reversed the aberrant expression of some of the genes observed in AK, consistent with clearing of lesions and normalization of confocal cellular images. Conclusions The data show that profound gene expression changes occur in sun- exposed, nonlesional skin which progress further in AK lesions. The data also suggest that imiquimod may play a role in normalizing gene expression and cellular morphology in sun-damaged skin.

Actinic keratoses (AK) are common, cutaneous, precancerous squamous cell carcinoma (SCC).4 Although a single pathway lesions appearing as rough, dry, scaly patches on the sun- to neoplastic transformation has not been defined for AK, sim- exposed skin of middle-aged and elderly people.1,2 The patho- ilarities between AK and SCC extend to histological and genesis of AK is directly related to the extent of exposure to genetic derangements, including alterations in the expression ultraviolet (UV) radiation, with chronic exposure causing of tumor suppressors and proto-oncogenes.8–10 AK appear cutaneous and systemic immunosuppression.3,4 Studies on capable of full malignant transformation to SCC, with a risk UV-exposed keratinocytes have shown changes in the expres- for transformation ranging from 0Æ25% on an annual basis to sion of genes with growth regulation, apoptosis, DNA repair, 20% over a 10-year period.2,11,12 cell adhesion and cytokine activity.5–7 It is believed that once Standard treatment of AK includes various types of surgical apoptosis is avoided, an actinically damaged cell may accumu- and chemical treatments,10,13 which are often associated with late further derangements, which, in the absence of adequate scarring and infection, and which may not address subclinical local immune surveillance, eventually lead to the phenotype of lesions that become clinically evident later. Imiquimod is

2007 3M Pharmaceuticals 1132 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1133 approved for the treatment of AK,14 superficial basal cell carci- confocal microscopy was used to characterize the specific AK noma,15,16 and external genital and perianal warts caused by lesions biopsied for gene expression analysis. human papillomavirus.17 In contrast to current therapy, the UV radiation is known to be carcinogenic. Cumulative DNA efficacy of imiquimod in AK is related to its immune modula- damage caused by exposure to UV can lead to mutations in tion. Imiquimod, a member of a class of drugs termed promoter regions of genes, which can lead to alterations immune response modifiers, has been shown to be a Toll-like in the gene expression profiles in the skin. These genetic receptor-7 (TLR7) agonist,18 whose antiviral and antitumour derangements can induce changes in important pathways activity is believed to be primarily due to the activation of the involved in cellular proliferation, cell survival, terminal differ- innate immune response. Imiquimod has been shown to acti- entiation and apoptosis.3,6 In order to identify these pathways vate antigen-presenting cells such as monocytes, macrophages, in AK, we evaluated changes in gene expression in sun- and plasmacytoid and myeloid dendritic cells. The activation exposed, nonlesional skin and in AK lesions before treatment leads to the production of interferon a and other cytokines with imiquimod. and chemokines, as well as T-cell costimulatory molecules important for the induction of an adaptive immune Materials and methods response.19,20 The results reported here were part of a study designed pri- Institutional review board and informed consent marily to investigate the mechanism of the antilesional activity of imiquimod using global gene expression analysis.21 The This single center study was conducted at Loma Linda Uni- study dealt with genes that were expressed at normal levels in versity Medical Center, Loma Linda, CA, U.S.A. The study pretreatment samples from AK and were altered in expression protocol, informed consent documents for subjects, and after treatment with imiquimod. Gene expression was analysed information documents for subjects were approved by the in AK lesions before, during and after treatment with imiqui- study centre’s Institution Review Board. This study was con- mod 5% cream. The topical application of imiquimod to AK ducted according to the Code of Federal Regulations of the lesions was shown to increase the expression of chemokine United States Food and Drug Administration (21 CFR Part 56, and cytokine genes and a large number of interferon-inducible Institutional Review Boards, and Part 50, Protection of Human genes with proapoptotic and growth inhibitory activity. The Subjects) and the International Conference on Harmonization observed gene expression profile was indicative of immune- Edition 6, Guideline for Good Clinical Practice. cell mediated destruction of AK lesions induced by imiqui- mod. In addition, this study showed the development of an Study conduct adaptive immune response, consistent with the observation of CD4+, CD8+, and CD11c+ cellular infiltrates reported with This was a phase II, double-blind, vehicle-controlled, random- topical application of imiquimod.14–16,22 ized parallel group study. Patients were recruited randomly The present analysis extends these findings by comparing from a southern California dermatology university clinic. All gene expression profiles between sun-exposed, nonlesional patients were men, with a history of male pattern baldness, skin and AK lesions to determine the molecular changes in Fitzpatrick skin type I–III and had a history of chronic sun sundamaged skin. In addition, gene expression changes in AK exposure. Some subjects had a history of skin cancer but not lesions were compared with the expressions after imiquimod in the area of treatment. Subjects were free of any significant treatment to assess the effect of imiquimod on the aberrant inflammatory skin conditions, such as psoriasis and eczema, at gene expression inherent in AK lesions. the application site, did not use tanning beds or sun parlours Since most AK lesions are diagnosed and followed-up clini- and did not use moisturizers and over-the-counter retinol cally (i.e. by visual and tactile examination), and confirmed products containing alpha- or beta-hydroxyacids during the by histology, a noninvasive method to diagnose AK lesions study period. Enrolled subjects were also free of immunomod- and to determine lesion clearance could decrease the need for ulators or immunosuppressive medication such as inhaled or biopsy. The successful use of reflectance confocal microscopy oral corticosteroids, interferon or interferon inducers and oral for the diagnosis of skin dysplasia has now been docu- retinoids (for 1 month before treatment), or systemic cancer mented23 and specific applications for AK,24 basal cell carci- chemotherapy or radiation (for 6 months before treatment). noma25 and melanoma26,27 have been described. The specific Subjects were required to have at least five clinically visible confocal microscopy parameters characteristic of AK have been AK lesions within a 25 cm2 area on the balding scalp. Confo- described previously and include irregular hyperkeratosis of cal as well as clinical and histopathological assessment was the stratum corneum, and enlarged pleomorphic nuclei with performed on the same lesion at the prestudy screening visit haphazard orientation in the stratum spinosum and stratum to establish a correlation of the confocal images and their basale.24 In this study, due to the limitation of biopsy mat- respective nonsun-exposed nonlesional skin, sun-exposed non- erial, histological analysis could not be performed on each lesional skin, AK and SCC. All subjects with lesions histologic- biopsy taken for gene expression analysis. Therefore, after ally identified as having a degree of dysplasia suggestive of initial comparison of AK lesions by histology and confocal SCC were disqualified from the study. All sites identified as AK microscopy and thus setting up a benchmark comparison, lesions were marked and a plastic template of the location

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1134 Microarray and confocal microscopy analysis, A. Torres et al. made for exact identification at a later time. Enrolled subjects those identified as AK within the target treatment area as were randomized to receive treatment with imiquimod or characterized by histological and confocal microscopy vehicle cream. Subjects applied the study cream to the treat- parameters as defined in the confocal microscopy section ment area 3 times per week for 4 weeks. The study cream below. By the criteria used, AK biopsies consisted of 75– was applied before normal sleeping hours and remained on 100% dysplastic cells. An additional biopsy of a different the skin for approximately 8 h before it was removed. AK lesion was taken at each subsequent study visit (treat- At the treatment initiation visit before performing biop- ment weeks 1, 2, and 4) and at 4 weeks after treatment. sies, confocal microscopy was performed on untreated AK Thus, each biopsy was of a different AK lesion. In cases lesions and untreated nonlesional sun-exposed skin near the where there was no remaining lesion, a biopsy was taken treatment area. Confocal microscopy was also performed on from a site originally marked as an AK lesion. The non- untreated, nonsun-exposed, nonlesional skin located on the lesional, nonsun-exposed site biopsy was used to establish a upper inner arm to establish a baseline for comparison and baseline for comparison of gene expression changes in AK the assessment of sun-induced skin changes compared with before and after treatment with imiquimod, as well as AK normal skin, as well as for comparison and assessment of compared with actinic-induced skin changes with no clinical changes in AK lesions after treatment with imiquimod or evidence of AK. A total of seven shave biopsies of these vehicle cream. Confocal microscopy was also performed at various skin conditions were taken per subject. Shave biop- each study visit at treatment weeks 1, 2 and 4, and sies were immediately placed in RNALater (Ambion, Austin, 4 weeks after the last treatment before biopsy. Only sites to TX, U.S.A.), equilibrated at room temperature for 1 h, kept be biopsied were imaged. Clinical assessment was also made at 4 C for 24 h, and then stored at )20 C before RNA 4 weeks after treatment. extraction. At the treatment initiation visit, a shave biopsy was taken Since lesions were removed for biopsy during the course of of an untreated AK lesion, a nonlesional, sun-exposed site, the study, the percentage clearance was not calculated for each and a nonlesional, nonsun-exposed site from the upper subject. Instead the percentage of subjects who had one or inner arm area for gene expression analysis. In an effort to more lesions cleared after treatment was calculated and standardize the amount of tissue that was removed at each reported in Table 1. biopsy, the same size punch was used to score the skin Safety evaluations were made at all treatment and post-treat- surrounding all the lesions to be biopsied for gene expres- ment visits, and included monitoring of adverse events and sion studies with an attempt to shave-biopsy the lesion at local skin reactions, as well as photographing the treatment the papillary dermis level. The lesions to be biopsied were area and reviewing any concomitant medications.

Table 1 Clinical and confocal microscopy assessment of actinic keratosis lesions

Confocal assessment Clinical assessment

Cells visualized ⁄ atypia ⁄diagnosis not clear Undeterminablea Normal epidermis Normal epidermis Actinic Treatment keratosis No. of Subjects No. of Subjects No. of Subjects No. of Subjects Study visit group (%) subjects (%) No.b subjects (%) No.b subjects (%) No.b subjects (%) No.b Prestudy Imiquimod 13 (100) 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0)c,d — Vehicle 4 (100) 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0)d — Initiation Imiquimod 13 (100) 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0)d — Vehicle 4 (100) 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0)d — Week 1 Imiquimod 12 (92Æ3) 1 (7Æ7) 002 0Æ0 (0) — 0Æ0 (0) — NA — Vehicle 4 (100) 0Æ0 (0) — 0Æ0 (0) — 0Æ0 (0) — NA — Week 2 Imiquimod 11 (84Æ6) 1 (7Æ7) 001 1 (7Æ7) 006 0Æ0 (0) — NA — Vehicle 3 (75) 1 (25) 007 0Æ0 (0) — 0Æ0 (0) — NA — Week 4 Imiquimod 9 (69Æ2) 2 (15Æ4) 014, 017 1 (7Æ7) 015 1 (7Æ7) 002 NA — Vehicle 4 (100) 0Æ0 (0) — 0Æ0 (0%) — 0Æ0 (0) — NA — Week 4 Imiquimod 7 (53Æ8) 2 (15Æ4) 011, 017 1 (7Æ7) 009 3 (23) 001, 002 4 (31)d 001, 002, post-treatment 006 004, 006 Vehicle 3 (75) 1 (25) 007 0Æ0 (0) — 0Æ0 (0) — 1 (25)d 007

a‘Undeterminable’ is a composite of the ‘Cells visualized but diagnosis undeterminable’ and ‘Unable to visualize cells clearly’ in the confocal microscopy assessment results. NA designates no clinical assessment was done. bSubject number with designated assessment. cHistology assessment. dClinical assessment.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1135

and the second cycle using Enzo BioArray High Yield RNA Confocal microscopy Transcript Labeling Kit (Affymetrix). The biotinylated cRNA Actinic keratosis lesions and nonlesional skin were imaged with was hybridized to Affymetrix U133A and U133B GeneChip the commercial confocal microscopy system VivaScope 1000 arrays containing 22 253 probe sets each. The hybridized (Lucid Inc., Henrietta, NY, U.S.A.) as described.25 The images probe arrays were stained with streptavidin-phycoerythrin were obtained with a near-infrared 830 nm diode laser of conjugate and images acquired according to Affymetrix power less than 40 mW at the tissue level. The lens used was a instructions. Images were analysed using MicroArray Suite ·30 water immersion lens, 0Æ9 NA #96008 (LOMO, Vermont Version 5 (MAS5). Chips were normalized to a global average Optechs, Charlotte, VT, U.S.A.) with water or water-based gel intensity of 150 to allow chip-to-chip comparison. The quality immersion media that had a refractive index of 1Æ33 for the ·30 of the images was ascertained by monitoring the noise, back- lenses. For the purposes of this study, only the horizontal ‘sec- ground, per cent transcript present, and the 3¢ :5¢ ratio for tions’ were used, although confocal microscopy can produce the housekeeping gene GAPDH. The average background and both vertical and horizontal virtual sections. Histology was noise for the arrays, the per cent transcript present, and the performed as described.24 Initial benchmark comparison of 3¢ :5¢ratios for GAPDH were within the accepted values.30,31 histology and confocal microscopy parameters included the following criteria: (i) polymorphism (AK) vs. monomorphism Analysis of Affymetrix GeneChip data (normal), (ii) cellular disorder (AK) vs. orderly orientation of cells (normal), (iii) parakeratosis of the stratum corneum (AK) Signals from the Affymetrix GeneChip images of biopsy sam- vs. none (normal) and (iv) partial thickness polymorphism ples for untreated normal skin obtained from nonsun-exposed (AK) vs. full thickness polymorphism (SCC). sites were used as control for the calculation of changes in expression for the other biopsy samples: pretreatment AK lesions; sun-exposed, untreated, non-AK skin; and AK lesions RNA extraction and purification during and after treatment (treatment weeks 1, 2, and 4, and Total RNA from the biopsy samples was extracted and purified 4 weeks after treatment). The statistical algorithm in MAS5 using Qiagen RNeasy Mini Kit Protocol for the Isolation of evaluates the image for the expression signal, the absent ⁄pres- Total RNA from Heart, Muscle, and Skin Tissue (Qiagen, ent call, and the P-value associated with the signal. It also eval- Valencia, CA, U.S.A.) according to the manufacturer’s instruc- uates the fold change of the sample relative to the designated tions. control sample expressed as the signal log2 ratio, the P-value associated with the fold change, and the direction of change (increased, I; decreased, D; or no call, NC). Using the Affyme- TaqManTM real-time reverse transcriptase polymerase trix data mining tool software, the expression data from MAS5 chain reaction were filtered on the basis of specific criteria to identify differ- TaqManTM (Applied Biosystems, Foster City, CA, U.S.A.) real- entially expressed genes. A given gene was selected if one time reverse transcriptase polymerase chain reaction (RT-PCR) sample from the series passed the following criteria: signal was performed for a number of the genes to confirm the detection P £ 0Æ01, signal log2 ratio £ )2or‡ 2, and a microarray results. cDNA was reverse transcribed from total change call designation of ‘increased’ (I) or ‘decreased’ (D). RNA using Invitrogen Superscript First-Strand Synthesis System for RT-PCR (Invitrogen, Carlsbad, CA, U.S.A.) according to Statistical analysis the manufacturer’s instructions. Real-time RT-PCR was per- formed using the Applied Biosystems 7900HTTM sequence The natural log of the fold change values from the Affymetrix detection instrument (Applied Biosystems), essentially as gene expression was used in an analysis of variance (ANOVA) described.28 TaqMan gene-specific primers and probes for to determine statistically significant changes in differential selected genes were purchased from Applied Biosystems. The expression between sample groups. Expression fold changes housekeeping gene glyceraldehyde-3-phosphate dehydrogenase were calculated with respect to expression in nonsun-exposed, (GAPDH) was used to normalize each sample. nonlesional skin samples. A 2-way ANOVA using a blocking fac- tor to account for repeat observations on the same subject was used to compare: (i) the fold change for sun-exposed, non- Microarray analysis lesional samples and pretreatment AK samples, (ii) the fold Samples for microarray analysis were prepared by two rounds change for pretreatment AK samples for subjects who were of linear target amplification according to the Affymetrix treated with imiquimod and samples from the same subjects instructions for eukaryotic small sample preparation.29,30 during treatment with imiquimod, and (iii) the fold change Briefly, double-stranded cDNA was synthesized from 100 ng for pretreatment AK samples for subjects who were treated of total RNA with oligo(dT)24 T7 primer (Affymetrix, Santa with imiquimod and imiquimod-treated samples taken Clara, CA, U.S.A.), followed by two cycles of in vitro transcrip- 4 weeks after the last imiquimod treatment. Differences tion of cRNA. The first cycle of in vitro transcription was per- between sample groups were considered significant if formed using a T7 polymerase (MEGAscript T7 Kit; Ambion) P<0Æ05 for the ANOVA.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1136 Microarray and confocal microscopy analysis, A. Torres et al.

Cluster analysis was performed using Spotfire DecisionSite- nuclei. This AK skin sample also exhibited a haphazard orienta- 8.1 for functional genomics (Spotfire Inc, Somerville, MA, tion and a honeycomb pattern of cells that was not uniform U.S.A.), using the unweighted pair-group method with arith- and was more consistent with the disordered differentiation metic mean (UPGMA) and the Euclidean similarity measure. typical of AK lesions. The sun-exposed, nonlesional skin Functional categorization of genes was based on analysis using (Fig. 1c) also exhibited pleomorphic cells and an irregular the Ontology Browser in Spotfire and gene descriptions at the honeycomb pattern, and, although it was to a lesser degree NCBI website (http://www.ncbi.nih.gov). Gene ontology files than sun-exposed, lesional (AK) skin, it was still consistent were obtained from the gene ontology website (http:// with an abnormal pattern of differentiation of epidermal cells. www.geneontology.org). The Ontology browser calculates a Thus, the difference in confocal pattern between AK and non- Fisher’s exact test probability, which reflects the chance that lesional sun-exposed skin was largely a matter of degree of the gene ontology category is represented by random involvement, since the pattern was similar for both. Figure 1d chance.32 P <0Æ05 was considered significant. shows the confocal image for the treatment site where the lesions had cleared at the end of the study. The image exhibited Results smaller cell size with a more uniform honeycomb pattern con- sistent with a differentiation pattern similar to nonsun-exposed skin, indicating that treatment with imiquimod results in Demographics reversion of abnormal epidermis. Seventeen white men were randomized to receive treatment A summary of the confocal and clinical assessments for all (13 to imiquimod 5% cream and four to vehicle cream). The subjects from initiation of the study to the end of the study is median number of AK lesions at baseline was 10 per subject shown in Table 1. At initiation of the study, all 17 subjects had (range 6–13 lesions). The mean age was 75 years (range 62– AK as diagnosed by histology, clinical assessment and confocal 89 years). assessment. None of the subjects cleared their lesions during week 1 and week 2 treatments. Because the number of AK lesions was progressively reduced because of biopsy of lesions, Confocal and clinical assessment accurate assessment of the percentage clearance could not be The intent of evaluating AK lesions with confocal microscopy made for each subject. Therefore, the proportion of subjects was to monitor noninvasively and document any changes in exhibiting normal epidermis is reported in Table 1. At the end- the epidermis that resulted from treatment. Although most of-treatment (week 4 treatment), Subject 002 had a confocal AK are diagnosed and followed clinically, suspicious AK are assessment of the post-treatment site which was consistent with biopsied for confirmation of diagnosis. Therefore, a noninva- a diagnosis of normal epidermis, defined as confocal patterns sive method to evaluate a suspicious lesion could potentially similar to those of untreated, nonlesional, nonsun-exposed skin. reduce the need to perform biopsies, and thus reduce potential At 4 weeks after treatment, three subjects (Subjects 001, 002 complications from procedures. At the prestudy screening visit and 006) had confocal assessments of the post-treatment site all AK lesions in the treatment area were diagnosed clinically that were consistent with a diagnosis of normal epidermis. None and by confocal microscopy and a representative lesion biop- of the subjects in the vehicle treatment arm had a confocal sied and confirmed histologically for each subject. The criteria assessment consistent with a diagnosis of normal epidermis. At used for diagnosis of AK were as described above. This 4 weeks after treatment, five subjects (Subjects 001, 002, 004, provided benchmark confocal images of clinically and histo- 006 and 007) showed complete clinical clearance of AK, with logically confirmed AK lesions. The confocal assessment of AK the skin exhibiting a more pleasing cosmetic appearance, as lesions exhibited similar and reproducible patterns in all of the defined by a more uniform pigmentation and smoothness at the subjects as did the confocal images of nonsun-exposed non- treatment site similar to nonsun-exposed skin (normalization). lesional skin. Thus, correlations between confocal images and There was good agreement between the confocal and clinical their respective nonsun-exposed nonlesional skin, and AK assessments, with the exception of Subject 007 (placebo group) lesions were established. Thereafter, at initiation of the study and Subject 004 (imiquimod group), who were diagnosed as and at 1, 2 and 4 weeks treatment time, lesions were diag- clearing lesions at 4 weeks after treatment by clinical assess- nosed by confocal microscopy. In addition, 4 weeks after treat- ment, but not by confocal assessment. This may be related to ment, lesions were diagnosed clinically as well as by confocal the greater sensitivity of confocal microscopy in detecting microscopy (Table 1). Figure 1 shows the confocal images for abnormality of the epidermis compared with clinical assessment Subject 001, whose lesions had cleared at the end of the study performed by visual and tactile examination. (week 4 after treatment). Nonsun-exposed, nonlesional skin (Fig. 1a) was characterized by an image exhibiting a uniform Gene ontology classification of aberrantly expressed honeycomb pattern of cells consistent with normal differentia- genes in sun-exposed, nonlesional skin and in actinic tion from the basal layer to the surface. Sun-exposed, lesional keratosis lesions (AK) skin (Fig. 1b) exhibited a stratum corneum showing irregular hyperkeratosis and parakeratosis, with a stratum Three hundred and fifty-one transcripts identified from MAS5 spinosum and basal layer showing enlarged pleomorphic were found to be altered in expression in pretreatment AK

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1137

(a) Uniform honeycomb pattern is smaller, consistent with differentiation from basal layer to surface

Uniform small cell stratum spinosum honeycomb pattern (b) Stratum spinosum and basale shows enlarged, pleomorphic nuclei with haphazard orientation

(c) Irregular honeycomb pattern of cells & cells with pleomorphic nuclei persist

(d) Smaller cell size with more uniform honeycomb pattern; more like sun- unexposed skin Smaller cells with greater depth, as with sun-unexposed skin

Fig 1. Confocal microscopy images of various skin biopsy sites for imiquimod-treated Subject 001. (a) Nonsun-exposed, nonlesional skin, (b) untreated, sun-exposed, lesional skin prior to treatment (pretreatment AK), (c) untreated, sun-exposed, nonlesional skin, and (d) sun-exposed, nonlesional skin in treatment area at the site of a previous lesion (cleared lesion) 4 weeks post-treatment. lesions and sun-exposed nonlesional skin in at least 60% of the Evaluation for genes that were differentially expressed in subjects. This list was used in a two-way ANOVA to determine sun-exposed, nonlesional skin and (or) AK lesions pretreat- differential expression between (i) pretreatment AK lesions and ment in at least 60% of subjects resulted in 265 genes that sun-exposed, nonlesional skin, (ii) pretreatment AK lesions and were (i) increased or decreased in expression compared with lesions during treatment with imiquimod and (iii) pretreatment normal nonsun-exposed skin, and (ii) had a median expres- AK lesions and 4 weeks after treatment with imiquimod. sion fold change (n = 17) with respect to nonsun-exposed

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1138 Microarray and confocal microscopy analysis, A. Torres et al. normal skin of > 2 for increased expression or <)2 for of two sun-exposed, nonlesional skin samples, Sun-016 and decreased expression. These data are summarized in Table S1 Sun-017, which have different expression profiles from the (Supplemental Table 1). ANOVA analysis of the log transformed rest of the samples. Thus, the two-way hierarchical cluster- expression fold change for the AK lesion group with respect ing of the 141 genes differentiates most of the sun-exposed, to the sun-exposed, nonlesional skin group showed 141 genes nonlesional skin samples from the AK lesions. that were significantly different in their expression between Extensive confirmation of genes altered in expression in this the two groups (P <0Æ05). Forty-nine genes which were sup- study by real-time RT-PCR was not possible because of limita- pressed in expression or were normal in expression in sun- tions in the amount of biopsy material available. Therefore, exposed, nonlesional skin were significantly more suppressed only a few genes were selected for confirmation by RT-PCR. in AK lesions. Ninety-one genes that were increased or normal Figure 3 shows a comparison of the fold change in expression in expression in sun-exposed, nonlesional skin were further obtained by using quantitative RT-PCR and Affymetrix Gene- increased in expression in AK lesions. The expression of 125 Chip analysis for the genes DTR (now HBEGE), IGFBP5, WNT5A unique genes (12 suppressed and 113 overexpressed) was not and CRISP3. The magnitude and direction of changes in statistically different between the two groups. expression determined by the two methods are comparable. Figure 2 shows a dendogram from a two-way hierarchical In order to determine the biological pathways and processes cluster analysis of the 141 genes that were significantly differ- represented by the aberrantly expressed genes, we evaluated ent in their expression between sun-exposed, nonlesional skin the gene ontology classification of the genes differentially and pretreatment AK lesions. The figure shows two main expressed in sun-exposed, nonlesional skin and in pretreat- sample clusters which are composed of (i) predominantly ment AK lesions. The functional groups with P <0Æ05 for pretreatment AK lesion samples (designated as AK) and statistically significant representation in the specified ontology (ii) sun-exposed, nonlesional skin samples (designated as category are summarized in Table 2. The highest representa- Sun). Figure 2 also shows a separate smaller cluster consisting tion of functions under biological processes was development,

161766 141

18·6 0 AK11 AK 09 AK 17 AK 16 AK 02 AK 14 AK 10 AK 15 AK 01 AK 04 AK 03 AK 07 AK 12 AK 13 AK 06 AK_08 AK_05 Sun 17 Sun 06 Sun 15 Sun 07 Sun 12 Sun 10 Sun 14 Sun 11 Sun 03 Sun 02 Sun 04 Sun 08 Sun 05 Sun 09 Sun 01 Sun 13 Sun_16 –6·6 6·6

Dendogram colour scale

Fig 2. Cluster analysis of genes altered in expression in sun-exposed, nonlesional skin and in pretreatment AK lesions. Hierarchical clustering was performed as described in the Methods section. ‘AK’ designates pretreatment AK samples; ‘Sun’ designates sun-exposed, nonlesional samples. Numbers designate subjects. Hierarchical clustering was performed using the WPGMA method and the Euclidean similarity measure. Red, white and green indicate upregulated, unchanged, and downregulated genes, respectively. The colour scale on the dendogram is log2 of fold change and ranges from –16Æ6 (bright green) to 16Æ6 (bright red).

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1139

of S100A12, which is overexpressed in AK lesions only. The * * * 10 expression of both classes of genes was significantly higher in 8 * pretreatment AK lesions than in sun-exposed, nonlesional skin. 6 * * Tissue proteases and their inhibitors play important roles in * * 4 * maintaining the structural integrity of the epidermis and the 35 2 dermo-epidermal junction. Several tissue proteases (Table 2)

0 are altered in expression in AK lesions, and in sun-exposed,

nonlesional skin. These include five matrix metalloproteinases

–2

* * * * * * * * (MMPs), MMP1, MMP3, MMP9, MMP10, and MMP12, and two

–4 *

Fold change in expression * kallikreins (KLKs), KLK6 and KLK13. In addition to the over- –6 DTR DTR IGFBP5 IGFBP5 WNT5A WNT5A CRISP3 CRISP3 RTPCR Affy RTPCR Affy RTPCR Affy RTPCR Affy expression of the proteases, several inhibitors of these genes –8 were suppressed in expression including TIMP3 (an inhibitor of MMP3). SERPINA3 (Clade A), SERPINB3 (SCCA1, SCC antigen Fig 3. Confirmation of expression changes observed by Affymetrix 1) and SERPINB4 (SCCA2, SCC antigen 2) were overexpressed GeneChip analysis using real-time RT-PCR for DTR, IGFBP5, WNT5A in sun-exposed, nonlesional skin and in AK lesions, and SERP- and CRISP3. The y axis depicts the median fold change in expression INA3, SERPINB3, SERPINB4, SERPINB13 (Hurpin) and SERPINE1 determined for 13 subjects. Fold change in expression were determined relative to nonsun-exposed nonlesional skin. Grey bars, (PAI1) were overexpressed in pretreatment AK lesions only. sun-exposed nonlesional skin; black bars, pretreatment AK lesions and white bars, samples take 4-weeks post-treatment with imiquimod. Deregulation of genes important in development: growth *Indicates expression difference that had P <0Æ05 between sample factors and genes implicated in cell survival pathways groups as determined by ANOVA. Alterations in the expression of growth factors and their with 46 and 56 genes altered in expression in sun-exposed, receptors and various oncogenes are associated with the growth nonlesional skin and in pretreatment AK lesions, respectively. and development of human tumors. In this study, we found that The aberrantly expressed genes include those involved in mor- several genes with key roles in proliferative signalling pathways phogenesis, histogenesis and epidermis development. The next were increased in expression in AK lesions, including genes in most highly represented branch of biological process was the insulin growth factor (IGF1) pathway, epidermal growth immune response, with 42 genes aberrantly expressed in AK factor (EGF) signalling pathways, WNT (wingless-type) signal- lesions, including inflammatory response genes, regulation of ling pathway and the FOS oncogene signalling pathway lymphocyte activation, and complement activation. (Table S1 and Table 3). Changes in the IGF1 pathway included increased expression of IGF1, which was increased in sun-exposed skin samples and Deregulation of expression of genes important to in pretreatment AK samples; and a decrease in expression of epidermal differentiation and homeostasis a negative regulator of this pathway, IGFBP5.36,37 IGFPB5 is Deregulation of epidermal homeostasis in AK lesions is indi- decreased in sun-exposed, nonlesional skin, with a further cated by the aberrant expression of several genes involved in decrease in expression in AK lesions. Alteration in the expres- epidermal differentiation and proliferation. Twenty-three sion of IGF1 has been shown to be associated with the growth genes in the gene ontology category of structure molecule and development of human tumors.38 Changes in the EGF activity (Table 2 and Table S1) were altered in expression in receptor signalling pathway include increased expression of pretreatment AK lesions, with a subset also being altered in two EGF receptor ligands, DTR and AREG39 and ADAM12,a expression in sun-exposed, nonlesional samples. These include positive regulator of the EGFR signalling system and a member extracellular matrix structural constituents [COL1A1, COL1A2, of the disintegrin and metalloprotease (ADAM) family.40 COL3A1, COL11A1, COMP, CHI3L1, FBLN1, and MAGP2 (now Members of the activator protein 1 (AP-1) transcription named MFAP5)] and structural constituents of cytoskeleton factors, FOSB and FOSL1 (Fra-1) which are involved in cell (KRT6A, KRT6B, KRT9, KRT13, KRT16, KRT17, and NEFL). All of development and differentiation were overexpressed in pre- the keratin genes were higher in expression in pretreatment treatment AK samples but not in sun-exposed, nonlesional AK lesions than in sun-exposed, nonlesional skin. The epider- skin samples. mal differentiation complex (EDC) at chromosome 1q21 in The expression of the wingless-type MMTV integration site humans contains several classes of genes important to epider- family, member 5A (WNT5A), was increased in pretreatment mal differentiation and homeostasis, including the SPRR (small AK lesions and sun-exposed, nonlesional samples. Concomi- proline-rich protein) genes and the calcium binding 9 genes, tant with the increased expression of WNT5A, two negative which are components of the keratinocyte cornified enve- regulators of the WNT pathway, WIF1 (WNT inhibitory fac- lope.33,34 Several members of the SPRR genes (SPRR1A, SPRR1B, tor) and CTNNBIP1 (catenin, beta interacting protein 1) were SPRR2B, and SPRR2C) and the S100 genes (S100A7, S100A8, decreased in expression in pretreatment AK samples. CTNNBIP1 S100A9, and S100A12) were increased in expression in sun- has been shown to be decreased in expression in some malig- exposed, nonlesional skin and in AK lesions, with the exception nant melanomas.41

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1140 Microarray and confocal microscopy analysis, A. Torres et al.

Table 2 Gene ontology of genes altered in sun-exposed nonlesional skin and in pretreatment actinic keratosis (AK) lesionsa

Number of genes altered in expression Examples of genes altered in expression

Sun-exposed nonlesional Pretreatment Gene ontology skin AK lesions Increased in expression Suppressed in expression Developmentb 46 56 ADAM12, ALOX15B, CCL2, CD3G, COL11A1, ALDH3A2, CTNNBIP1, DSCR1L1, EMX2, COL1A1, COL1A2, COL3A1, COMP, FGFR3, FHL1, GATA3, HOXC6, IGFBP5, CRISP3, CRISP2, DHRS9, DTR, EHF, FADS1, KIAA1233, LEPR, MYH11, NRLN1, FOSB, FOXE1, HYAL4, IFRD1, IGF1, PDGFC, PMP22, SCEL, TGFBR3, WIF1, INHBA, KLK6, KRT13, KRT16, KRT17, ZNF145 KRT6A, KRT6B, KRT9, NDRG4, NEFH, PITX1, PITX2, PPARG, PTN, RUNX1, S100A7, SPP1, SPRR1A, SPRR1B, SPRR2B, STMN2, WNT5A, ZIC1 Histogenesis 15 18 KRT6A, KRT6B, KRT9, KRT13, KRT16, ALDH3A2, SCEL, ZNF145 KRT17, COL1A1, COL11A1, SPRR1A, SPRR1B, S100A7, FOXE1, INHIBA, SPP1 Epidermis development 710KRT6A, KRT6B, KRT9, KRT13, KRT16, ALDH3A2, SCEL KRT17, COL1A1, SPRR1A, SPRR1B, S100A7 Immune responseb 27 42 BRDG1, CCL19, CCL2, CCL20, CD24, CD3G, AQP9, DAF, DF, CCL27, HF1, IL1F7 CLECSF12, CRISP3, CXCL13, CXCL2,CXCL9, DEFB4, HPSE, IFI27, IGHG3, IGHM, IGL@, INHBA, LTP, LTF, MS4A1, NR4A2, OAS1, RGS1, S100A12, S100A8, S100A9, SAA2, SERPINA3, SERPINB4, SPP1, UBD Lymphocyte activation 46CD3G, CLECSF12, INHBA, MS4A1, RGS1, — SPP1 Humoral immune response 25CCL2, CD24, LTF, BRDG1 — Complement activation 23— DAF, DF, HF1 Inflammatory response 812CCL19, CCL2, CCL20, CXCL13, CXCL2, — CXCL9, HPSE, S100A12, S100A8, S100A9, SERPINA3, SPP1 Cytokine activity 12 14 CCL2, CCL20, CCL19, CXCL2, CXCL9, CCL27, IL1F7 CXCL13, IL1F9, PTN, SPP1, LTB, INHIBA, AREG Chemokine activity 57CCL2, CCL20, CCL19, CXCL2, CXCL9, CXCL13 Growth factor activity 710IGF1, PTN, SPP1, DTR, INHIBA, IL1F9, AREG BTC, IL1F7, PDGFC Peptidase 10 16 ADAM12, ADAMEC1, GZMB, KLK6, KLK13, DF, KIAA1233 KYNU, LTF, MMP1, MMP3, MMP9, MMP13, SERPINE1 Protease inhibitor 48PI3, SERPINA3, SERPINB3, SERPINB4, TIMP3, TFPI SERPINE1, SERPINB13 Oxidoreductase activity 16 25 ACADM, AKR1B10, ALDH1A3, ALOX15B, ADH1B, ALDH3A2, CYP4B1, CYP39A1 CH25H, CYP4F8, DIO2, DHRS9, FADS1, FADS2, FMO2, GCLM, GPD1, GLDC, GPX2, HSD3B1, HSD17B2, RRM2, SC5DL, SCD, SOD2 Structure molecule activity 16 23 CHI3L1, COL1A1, COL1A2, COL3A1, MAGP2, FBLN1, LAMB4 COL11A1, COMP, KRT6A, KRT6B, KRT9, KRT13, KRT16, KRT17, NEFH, NEFL, SPRR1A, SPRR1B, SPRR2B, MYO6

aGene ontology classification was performed using the gene ontology browser in Spotfire using ontology files from http://www.geneontology. org. bSubcategories of this main category are indented below.

In addition to the tumour suppressor genes in the IGF1 and lesions. These include PDGFRL (platelet-derived growth factor WNT signalling pathways discussed above, the expression of receptor-like), TGFBR3 (transforming growth factor beta recep- other tumour suppressor genes was also downregulated in AK tor 3) and FGFR3 (fibroblast growth factor receptor 3). These

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1141

Table 3 Genes whose expression was altered in pretreatment actinic keratosis (AK) and partially or totally reversed by treatment with imiquimod (IMIQ)

Fold Fold P-value Fold P-value Gene change change sun vs. Fold change P-Value change AK vs. Probe set ID symbol suna AKb AKc IMIQd AK vs. IMIQe poste postc Gene title 205357_s_at AGTR1 )2Æ1 )3Æ3 0Æ000 )3Æ10Æ804 )2Æ2 0Æ018 angiotensin II receptor, type 1 205239_at AREG 2Æ64Æ4 0Æ000 4Æ10Æ979 3Æ1 0Æ007 amphiregulin (schwannoma- derived growth factor) 219099_at C12orf5 2Æ32Æ5 0Æ020 2Æ80Æ274 2Æ1 0Æ026 chromosome 12 open reading frame 5 216379_x_at CD24 1Æ92Æ7 0Æ000 2Æ60Æ092 1Æ8 0Æ004 CD24 antigen (small cell lung carcinoma cluster 4 antigen) 206166_s_at CLCA2 1Æ62Æ3 0Æ000 1Æ90Æ335 1Æ8 0Æ038 chloride channel, calcium activated, family member 2 207802_at CRISP3 6Æ410Æ3 0Æ024 2Æ3 0Æ007 6Æ6 0Æ046 cysteine-rich secretory protein 3 205382_s_at DF )2Æ5 )3Æ2 0Æ000 )4Æ00Æ102 )2Æ6 0Æ021 D component of complement (adipsin) 202994_s_at FBLN1 )1Æ9 )3Æ1 0Æ000 )4Æ40Æ061 )2Æ1 0Æ001 fibulin 1 201540_at FHL1 )2Æ1 )3Æ6 0Æ000 )2Æ70Æ834 )2Æ5 0Æ004 four and a half LIM domains 1 202768_at FOSB )1Æ14Æ0 0Æ016 1Æ60Æ262 )1Æ5 0Æ001 FBJ murine osteosarcoma viral oncogene homolog B 204420_at FOSL1 1Æ37Æ4 0Æ000 5Æ00Æ262 2Æ9 0Æ013 FOS-like antigen 1 202147_s_at IFRD1 1Æ82Æ4 0Æ000 1Æ30Æ216 1Æ5 0Æ003 interferon-related developmental regulator 1 211959_at IGFBP5 )2Æ2 )2Æ9 0Æ002 )1Æ3 0Æ002 )1Æ5 0Æ001 insulin-like growth factor binding protein 5 211762_s_at KPNA2 1Æ62Æ2 0Æ000 2Æ5 0Æ043 1Æ6 0Æ006 karyopherin alpha 2 (RAG cohort 1, importin alpha 1) 209125_at KRT6A 6Æ17Æ4 0Æ002 5Æ80Æ062 4Æ5 0Æ045 keratin 6A ⁄⁄⁄keratin 6C ⁄⁄⁄ keratin 6E 219181_at LIPG 1Æ92Æ9 0Æ002 2Æ60Æ433 2Æ3 0Æ040 lipase, endothelial 218888_s_at NETO2 1Æ82Æ4 0Æ000 2Æ30Æ602 1Æ9 0Æ029 neuropilin (NRP) and tolloid (TLL)-like 2 204621_s_at NR4A2 1Æ03Æ7 0Æ007 1Æ20Æ505 1Æ4 0Æ011 nuclear receptor subfamily 4, group A, member 2 209763_at NRLN1, CHRDL1 )2Æ0 )2Æ6 0Æ001 )2Æ00Æ975 )1Æ9 0Æ021 chordin-like 1 213568_at OSR2 )2Æ1 )3Æ1 0Æ006 )3Æ00Æ079 )1Æ9 0Æ036 odd-skipped related 2 (Drosophila) 205226_at PDGFRL )1Æ3 )3Æ4 0Æ000 )1Æ90Æ339 )2Æ1 0Æ014 platelet-derived growth factor receptor-like 210138_at RGS20 2Æ02Æ2 0Æ008 1Æ4 0Æ001 1Æ8 0Æ021 regulator of G-protein signaling 20 205064_at SPRR1B 3Æ44Æ0 0Æ016 3Æ30Æ183 3Æ2 0Æ045 small proline-rich protein 1B (cornifin) 204731_at TGFBR3 )1Æ8 )2Æ5 0Æ001 )2Æ50Æ419 )2Æ1 0Æ011 transforming growth factor, beta receptor III 209387_s_at TM4SF1 1Æ62Æ1 0Æ001 1Æ50Æ406 1Æ7 0Æ021 transmembrane 4 L six family member 1 203798_s_at VSNL1 2Æ83Æ6 0Æ001 4Æ50Æ568 2Æ4 0Æ005 visinin-like 1 205990_s_at WNT5A 4Æ05Æ1 0Æ008 4Æ5 0Æ047 4Æ4 0Æ002 wingless-type MMTV integration site family, member 5A 206683_at ZNF165 1Æ82Æ70Æ060 2Æ80Æ464 1Æ5 0Æ008 zinc finger protein 165

aMedian fold change for sun-exposed, nonlesional skin. bMedian fold change for pretreatment AK. cP-values are for 2-way subject controlled ANOVA analysis. Statistically significant P-values are in bold font; n = 17 for comparison of Sun vs. AK and n = 13 for comparison of AK vs. IMIQ and AK vs. Post. Fold change was calculated relative to expression in sun-unexposed, nonlesional skin samples. dMedian fold change for the maximum fold change value of week 1, week 2 and week 4 treatments with imiquimod. eMedian fold change for samples at week 4 post-treatment with imiquimod.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1142 Microarray and confocal microscopy analysis, A. Torres et al. genes have been shown to be frequently decreased in expres- chemokines, two genes of the interleukin (IL)-1 cytokine sion in various cancers and have been postulated to act as family, IL1F7 and IL1F9, were aberrantly expressed in sun- tumour suppressors.42–44 exposed, nonlesional skin and to a greater extent in pretreat- ment AK samples. The expression of IL1F7 decreased, whereas that of IL1F9 was increased. Aberrant expression of genes involved in immune response Reversal of aberrantly expressed genes after treatment In this study, we observed deregulation of the expression of with imiquimod several immune response genes in sun-exposed, nonlesional skin (27 genes) and in pretreatment AK lesions (42 genes) Treatment of AK lesions with the TLR7 agonist imiquimod has (Table 2). The chemokines CCL2, CCL19, CCL20, CXCL2, CXCL9, been shown to result in changes in gene expression profile CXCL13 were increased in expression and CCL27 was decreased indicative of activation of the innate immune system and in expression in AK lesions. CCL20, a chemokine chemotactic immune cell-mediated destruction of AK lesions.21 Here we for immature dendritic cells and lymphocytes, was increased ask the question if treatment with imiquimod was able to in AK lesions and sun-exposed, nonlesional skin. CCL2 and reverse or attenuate the aberrant gene expression inherent in CCL19 and CXCL9, chemokines chemotactic for monocytes and AK lesions. ANOVA analysis comparing the fold change of dendritic cells, and for T and B cells, were overexpressed in expression of genes in pretreatment AK samples with their AK lesions only. CCL27 was decreased in expression in sun- expression 4 weeks after treatment with imiquimod resulted exposed, nonlesional skin and further decreased in pretreat- in 28 genes which were significantly different in expression ment AK lesions. CCL27 has been shown to be chemotactic for in the two groups (P <0Æ05). The data are summarized in skin-associated memory T lymphocytes and may play a role in Table 3. In most cases, the median fold change for the post- mediating homing of lymphocytes to cutaneous sites.45 CCL27 treatment values is near the value for sun-exposed, nonlesional has also been shown to possess antitumour activity46 and its skin, indicating a progression towards normalization of gene expression shown to be suppressed in SCC.7 In addition to expression in skin treated with imiquimod. Figure 4 shows a

P-value P-value

1123 3 8 AK vs. FC IMIQ- AK vs Probe ID Gene symbol FC Sun FC AK Sun post post 15·1 0 201540_at FHL1 –2·1 –3·6 0·000 –2·5 0·004 202994_s_at FBLN1 –1·9 –3·1 0·000 –2·1 0·001 211959_at IGFBP5 –2·2 –2·9 0·002 –1·5 0·001 213247_at SVEP1 –1·5 –2·3 0·078 –1·5 0·007 202768_at FOSB –1·1 4·0 0·016 –1·5 0·001 206683_at ZNF165 1·8 2·7 0·0601·5 0·008 211762_s_at KPNA2 1·6 2·2 0·000 1·6 0·006 202147_s_at IFRD1 1·8 2·4 0·000 1·5 0·003 216379_x_at CD24 1·9 2·7 0·000 1·8 0·004 203798_s_at VSNL1 2·8 3·6 0·001 2·4 0·005 205239_at AREG 2·6 4·4 0·000 3·1 0·007 205990_s_at WNT5A 4·0 5·1 0·008 4·4 0·002 AK 04 AK 11 AK 01 AK 16 AK 12 AK 05 AK 02 AK_06 AK 15 AK_17 AK_08 AK_09 AK_14 IMIQ 4WK Post_16 IMIQ 4WK Post 15 IMIQ 4WK Post_06 IMIQ 4WK Post 01 IMIQ 4WK Post 08 IMIQ 4WK Post_02 IMIQ 4WK Post 14 IMIQ 4WK Post_12 IMIQ 4WK Post 04 IMIQ 4WK Post_11 IMIQ 4WK Post 05 IMIQ 4WK Post 17 IMIQ 4WK Post_09

–3 4·5

Dendogram colour scale

Fig 4. Cluster analysis of genes which were aberrantly expressed in pretreatment AK and attenuated or reversed in expression in samples at 4 weeks post-treatment with imiquimod. Gene expressions were significantly different between the 2 groups, as determined by the ANOVA analysis (P <0Æ01). Hierarchical clustering was performed as described in the Methods section. ‘AK’ designates pretreatment AK samples; ‘Post’ designates samples at 4 weeks post-treatment with imiquimod. Numbers designate subjects who were treated with imiquimod (n = 13). FC designates fold change. Hierarchical clustering was performed using the WPGMA method and Euclidean similarity measure. Red, white and green indicate ) upregulated, unchanged, and downregulated genes, respectively. The colour scale is log2 of fold change and ranges from 1Æ3 (bright green) to 4Æ5 (bright red).

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1143

2-way hierarchical clustering of the genes which were differ- (a) entially expressed between pretreatment AK samples and sam- 5·0 ples obtained 4 weeks post-treatment and had P <0Æ01. Two main sample clusters are apparent. The smaller cluster, which 2·5 is dominated by high expression of FOSB, includes pretreat- ment AK samples (designated as AK) for Subjects 005, 009, 0·0 FC FOSB 008, 012, 014, and 017 and 4 weeks post-treatment samples 2

(designated as Post) for Subjects 009 and 017. Subjects 009 Log –2·5 and 017 showed very little gene expression response during P = 0·001 treatment with imiquimod, as assessed by global gene expres- –5·0 Sun AK IMIQ Post 21 sion. The second and larger major cluster is characterized by Treatment near normal or depressed levels of FOSB and consists of post- treatment samples for Subjects 001, 002, 004, 005, 006, 008, (b) 1 011, 012, 014, 015, and 016 and pretreatment AK samples for Subjects 001, 002, 004, 006, 011, and 015. This cluster 0 consists of post-treatment samples from subjects who had cleared lesions at 4 weeks post-treatment as judged by confo- –1 FC IGFBP5 cal assessment and clinical assessment (Subjects 001, 002 and 2 006) and by clinical assessment only (Subject 004) (Table 1). –2 Log The subjects in this cluster showed moderate to strong gene –3 P = 0·001 expression changes during treatment with imiquimod as com- pared with Subjects 009 and 017.21 It is interesting that the Sun AK IMIQ Post expression levels of the oncogene FOSB in the pretreatment AK Treatment samples of subjects who were diagnosed as having cleared lesions at 4 weeks post-treatment were lower than those sub- Fig 5. Box plot of selected genes (a, FOSB,b,IGFBP5) whose jects who had not cleared their lesions. The significance of this expression was attenuated by treatment with imiquimod. ‘AK’, ‘Sun’ needs further study. and ‘Post’ designate pretreatment AK samples; sun-exposed, nonlesional skin samples; and samples at week 4 post-treatment with Examples of changes in expression after treatment with imiquimod, respectively. ‘IMIQ’ designates the maximum fold change imiquimod are shown in Figure 5a (FOSB) and Figure 5b value from treatment weeks 1, 2, and 4 with imiquimod. Fold change (IGFBP5). The median expression level of FOSB decreased in was calculated with respect to nonsun-exposed normal skin. A fold the post-treatment samples and was close to the value for sun- change of 1 (log2 FC = 0) indicates the same expression as in exposed, nonlesional skin samples. The median expression nonsun-exposed normal skin. Boxes indicate the median 95% level of IGFBP5 increased from a value of )2Æ9 in AK lesions confidence intervals, asterisks designate outliers, and the lines connect to a value of )1Æ5 after imiquimod treatment (P =0Æ001). median values. Probabilities are given for 2-way ANOVA comparing Many of the genes that were highly expressed at baseline and pretreatment AK to week 4 post-treatment for imiquimod-treated attenuated after imiquimod treatment included genes that have subjects (n = 13). been reported to be involved in proliferative pathways such as the WNT pathway (WNT5A), IGF pathway (IGFBP5), epidermal growth factor receptor pathways (AREG), and FOSB. Attenua- 2 tion of the downregulation of tumour suppressor genes such as PDGFRL and TGFBR3 would also be predicted to decrease 0 S–001 S–006 S–002 S–005 S–012 S–017 proliferation. It is not clear whether these pathways are –2 reversed in the cells or whether the attenuation of the expres- sion patterns is a reflection of a decrease in the number of –4 dysplastic keratinocytes. In either case, the decrease indicates –6 that imiquimod treatment results in the normalization in expression of oncogenic genes and tumour suppressor genes –8

in sun-damaged skin. Fold change in expression –10 Because of the short duration of the post-treatment period (4 weeks), resulting in a low percentage of subjects whose Fig 6. Comparison of the expression of IGFBP5 in cleared AK lesions lesions cleared (four of 13 treated), attempts to correlate gene biopsied 4 weeks post-treatment with imiquimod (Subjects S001, expression with lesion resolution was not made. In general, S002 and S006) and AK lesions that did not clear (Subjects S005, subjects whose lesions cleared by 4 weeks post-treatment S012 and S017). Fold change in expression was determined relative to tended to have a higher degree of attenuation of aberrant gene nonsun-exposed skin samples taken from each subject. Grey bars, expression. Figure 6 shows an example for IGFBP5. Subjects sun-exposed nonlesional skin; black bars, pretreatment AK lesions and who had cleared lesions 4 weeks post-treatment (Subjects white bars, 4 weeks after treatment with imiquimod.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1144 Microarray and confocal microscopy analysis, A. Torres et al.

001, 002 and 006) showed larger differences in expression kines in tumours is complex. While some chemokines may between pretreatment AK and 4 weeks post-treatment com- enhance innate or specific host immunity against tumours (e.g. pared with subjects who had AK lesions which did not clear CCL27, IL1F7); others (e.g. CCL2, CCL20) may favour tumour (Subjects 005, 012 and 017). growth and metastasis by promoting tumour cell proliferation, migration, or vascularization.61,62 In this study, the increased Discussion expression of chemokines associated with tumour proliferation (e.g. CCL2, CCL20) and the decrease in expression of those that Chronic exposure to UV radiation has been associated with enhance innate or specific host immunity (e.g. CCL27, IL1F7) progressive genetic and tissue damage. Many of these genetic indicate that AK lesions possess an immune environment abnormalities can lead to the loss of important cellular func- conducive to tumour growth and immune evasion. tions by altering the expression of genes that regulate cellular Although the exact mechanism of pathogenesis of AK devel- proliferation and tumour suppressor activity. In this study, we opment is unknown, part of the pathogenesis may involve have shown that sun-exposed, nonlesional skin and AK lesions suppression of the immune response against dysplastic cells.63 exhibit aberrant expression of genes important in epidermal It is believed that prolonged UV exposure changes the homeostasis, integrity and differentiation and in cellular pro- immune surveillance mechanism of the skin, contributing to liferation. In addition, the expression of several immune func- the tolerance of tumour cells.4 tion genes with proliferative as well as invasive potential are This study found aberrant expression of genes in three main increased in expression whereas genes which are important to proliferative pathways. Both positive and negative regulators cellular immunity are downregulated. This is consistent with of the IGF1 pathway are aberrantly expressed in AK (increased earlier findings suggesting chronic UV exposure leads to sup- expression of IGF1 and decreased expression of IGFBP5), indi- pression of the cutaneous immune response, perhaps partly by cating the importance of this pathway in the pathogenesis of the induction of tolerance to dysplastic cells. AK development. The WNT signalling pathway controls the The increased expression of structural constituents of the transcription of genes that regulate cellular proliferation64 and epidermis and genes of the epidermal differentiation complex is frequently deregulated in cancer.41,65,66 As in the IGF1 indicate increased proliferation of keratinocytes in AK lesions pathway, overexpression of a positive driver of the WNT path- and deregulation of normal epidermal differentiation. KRT6, way (WNT5A) with a concomitant decrease in the expression KRT14, KRT16, KRT17, and KRT13 have been shown to be UV- of negative regulators of this pathway (WIF1 and CTNNBIP1) inducible.5,47 KRT13, KRT16, and KRT17 are associated with confirms the importance of the proliferative WNT pathway in hyper-proliferation of keratinocytes in psoriatic lesions.48,49 AK lesions. KRT17 is increased in breast carcinoma50 and in cervical carci- The activator protein 1 (AP-1) transcription factor consists noma,51 and KRT16 and KRT17 have also been shown to be of homo- or heterodimers of members of the Fos and Jun increased in head and neck SCC.48,52 The expression of SPRR proteins.67 AP-1 plays critical roles in many biological pro- and S100 genes has been shown to be increased upon expos- cesses including cell proliferation, oncogenic transformation, ure to UV,7 in psoriasis,48,49 in atopic dermatitis53 and in and apoptosis,68 as well as keratinocyte differentiation.69 SCC.7,48 Taken together, these data show that AK lesions exhi- Deregulation of the expression of AP1 genes has been bit an abnormal epidermal phenotype similar to proliferative observed in some cancers, including overexpression of Fra-1 diseases such as psoriasis and SCC. (now FOSL1) in SCC70 and FOS, FOSB, JUN and JUNB in inflam- In addition to their normal functions, tissue proteases are matory breast cancer.71 The overexpression of AP1 suggests implicated in various cancers. KLK6 and KLK13 are overexpres- this signalling pathway may play an important role in the sed in various cancers, including SCC.50,54,55 MMPs are also pathogenesis of AK development. implicated in tumour metastasis and invasiveness,56,57 and Many of the genes altered in expression in sun-exposed MMP1 and MMP10 have been shown to have increased expres- nonlesional skin and in AK lesions have also been shown to sion when fibroblasts and keratinocytes are exposed to UV be altered in human SCC, especially those involved in epider- radiation.5,58 In addition to their role in the maintenance of mal homeostasis, integrity and differentiation. The data in this the extracellular matrix (e.g. SERPINA1, SERPINE1), several SER- study suggest that many of the changes manifest in SCC have PINs, including SERPINA3, SERPINB3 and SERPINB4, are known already been initiated in sun-damaged skin and are apparent to be overexpressed in various proliferative skin diseases, in AK lesions. We also find differences in the gene expression including psoriasis, SCC and AK.50,59 Taken together, the pattern between AK and SCC. The expression of the classical altered expression of the KLKs, MMPs, and SERPINs show that tumour suppressor genes p53 (TP53) and CDKN2A was not AK lesions share many of the same derangements in gene observed to be altered in expression in this study. This is con- expression as other proliferative diseases, including psoriasis sistent with the functional mutations rather than change in and SCC. expression observed for these genes for various cancers includ- Of the aberrantly expressed IL-1 cytokine family, the genes ing SCC.48 The oncogene K-Ras (KRAS), was not observed to IL1F7 and IL1F9 possess IL-1 receptor antagonist activity; IL1F7 change in expression in AK lesions in this study, whereas also binds IL-18 receptor. IL1F7 was shown to have anti- Haider et al. report a small increase in expression in SCC tumour activity in mice.60 The role of chemokines and cyto- (fold change SCC ⁄normal = 1Æ2).48 Thus, there are inherent

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1145 differences as well as similarities in the gene expression 9 Park HR, Min SK, Cho HD et al. Expression profiles of p63, p53, profiles of AK and SCC. The similarities in gene expression survivin, and hTERT in skin tumors. J Cutan Pathol 2004; 31:544–9. observed for AK and SCC underscore the importance of treat- 10 Fu W, Cockerell CJ. The actinic (solar) keratosis: a 21st-century perspective. Arch Dermatol 2003; 139:66–70. ing AK to prevent progression to SCC. 11 Glogau RG. The risk of progression to invasive disease. J Am Acad The confocal microscopy results showed that both pretreat- Dermatol 2000; 42:23–4. ment AK lesions confirmed histologically, and sun-exposed, 12 Mittelbronn MA, Mullins DL, Ramos-Caro FA, Flowers FP. nonlesional skin had abnormal confocal image patterns com- Frequency of pre-existing actinic keratosis in cutaneous squamous pared with nonsun-exposed nonlesional skin. This observation cell carcinoma. Int J Dermatol 1998; 37:677–81. is consistent with the large number of aberrantly expressed 13 Silapunt S, Goldberg LH, Alam M. Topical and light-based treat- genes in sun-exposed, nonlesional skin that progress further in ments for actinic keratoses. Semin Cutan Med Surg 2003; 22:162–70. 14 Szeimies RM, Gerritsen MJ, Gupta G et al. Imiquimod 5% cream AK lesions. These findings indicate that there are cellular and for the treatment of actinic keratosis: results from a phase III, molecular changes in the epidermis of sun-exposed, nonlesion- randomized, double-blind, vehicle-controlled, clinical trial with al skin even before they are clinically evident. Treatment of AK histology. J Am Acad Dermatol 2004; 51:547–55. lesions with the TLR7 agonist imiquimod resulted in reversal or 15 Urosevic M, Maier T, Benninghoff B et al. Mechanisms underlying attenuation of aberrant gene expression related to regulation of imiquimod-induced regression of basal cell carcinoma in vivo. Arch cell proliferation. This is consistent with its efficacy in the treat- Dermatol 2003; 139:1325–32. ment of proliferative skin diseases such as AK and SCC. 16 Barnetson RS, Stachell A, Zhuang L et al. Imiquimod-induced regression of clinically diagnosed superficial basal cell carcinoma is In summary, the results of the confocal microscopy indicate associated with early infiltration by CD4 T cells and dendritic cells. that imiquimod treatment was associated with reversion of the Clin Exp Dermatol 2004; 29:639–43. AK lesions to more normal appearing epidermis for one of the 17 Beutner KR, Tyring SK, Trofatter KF Jr et al. Imiquimod, a patient- end-of-treatment samples and three of the 4-week post-treat- applied immune-response modifier for treatment of external genital ment samples. Confocal assessments of untreated, nonlesional, warts. Antimicrob Agents Chemother 1998; 42:789–94. sun-damaged skin also showed abnormalities similar to, but 18 Hemmi H, Kaisho T, Takeuchi O et al. Small anti-viral compounds of a lesser degree than, those seen with AK. Thus, the rever- activate immune cells via the TLR7 MyD88-dependent signaling pathway. Nat Immunol 2002; 3:196–200. sion of treated AK to normal epidermis could be indicative of 19 Miller RL, Gerster JF, Owens ML et al. Imiquimod applied topically: a reversal of actinically induced skin damage, or destruction of a novel immune response modifier and a new class of drug. Int J subclinical AK. Immunopharmacol 1999; 21:1–14. 20 Gibson SJ, Lindh JM, Riter TR et al. Plasmacytoid dendritic cells produce cytokines and mature in response to the TLR7 agonists, Acknowledgments imiquimod and resiquimod. Cell Immunol 2002; 218:74–86. We thank Rhonda Dick for help in manuscript preparation. 21 Torres A, Storey L, Anders M et al. Immune-mediated changes in actinic keratosis following topical treatment with imiquimod 5% cream. J Transl Med 2007; 5:7. References 22 Ooi T, Barnetson RS, Zhuang L et al. Imiquimod-induced regression of actinic keratosis is associated with infiltration by T lymphocytes 1 Callen JP, Bickers DR, Moy RL. Actinic keratosis. J Am Acad Dermatol and dendritic cells: a randomized controlled trial. Br J Dermatol 1997; 36:650–3. 2006; 154:72–8. 2 Marks R, Rennie G, Selwood T. The relationship of basal cell 23 Gonzalez S, Swindells K, Rajadhyaksha M, Torres A. Changing carcinomas and squamous cell carcinomas to solar keratoses. Arch paradigms in dermatology: confocal microscopy in clinical and Dermatol 1988; 124:1039–42. surgical dermatology. Clin Dermatol 2003; 21:359–69. 3 Edwards L, Levine N, Weidner M, Piepkorn M. Effect of intra- 24 Aghassi D, Anderson RR, Gonzalez S. Confocal laser microscopic lesional alpha2 interferon on actinic keratosis. Arch Dermatol 1986; imaging of actinic keratoses in vivo: a preliminary report. J Am Acad 122:779–82. Dermatol 2000; 43:42–8. 4 Yantsos VA, Conrad N, Zabawski E, Cockerell CJ. Incipient intra- 25 Torres A, Niemeyer A, Berkes B et al. 5% imiquimod cream and epidermal cutaneous squamous cell carcinoma: a proposal for reflectance-mode confocal microscopy as adjunct modalities to reclassifying and grading solar (actinic) keratoses. Semin Cutan Med Mohs’ micrographic surgery for treatment of basal cell carcinoma. Surg 1999; 18:3–14. Dermatol Surg 2004; 30:1462–9. 5 Murakami T, Fujimoto M, Ohtsuki M, Nakagawa H. Expression 26 Curiel-Lewandrowski C, Williams CM, Swindells KJ et al. Use of profiling of cancer-related genes in human keratinocytes follow- in vivo confocal microscopy in malignant melanoma: an aid in diag- ing non-lethal ultraviolet B irradiation. J Dermatol Sci 2001; nosis and assessment of surgical and nonsurgical therapeutic 27:121–9. approaches. Arch Dermatol 2004; 140:1127–32. 6 Wertz K, Hunziker PB, Seifert N et al. b-carotene interferes with 27 Langley RG, Burton E, Walsh N et al. In vivo confocal scanning laser ultraviolet light A-induced gene expression by multiple pathways. microscopy of benign lentigines: comparison to conventional J Invest Dermatol 2005; 124:428–34. histology and in vivo characteristics of lentigo maligna. J Am Acad 7 Rundhaug JE, Hawkins KA, Pavone A et al. SAGE profiling of Dermatol 2006; 55:88–97. UV-induced mouse skin squamous cell carcinomas, comparison 28 Winer J, Kwang C, Jung S et al. Development and validation of with acute UV irradiation effects. Mol Carcinog 2005; 42:40–52. real-time quantitative reverse transcriptase-polymerase chain reac- 8 Cockerell CJ. Histopathology of incipient intraepidermal squamous tion for monitoring gene expression in cardiac myocytes in vitro. cell carcinoma (‘actinic keratosis’). J Am Acad Dermatol 2000; 42:11– Anal Biochem 1999; 270:41–9. 17.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 1146 Microarray and confocal microscopy analysis, A. Torres et al.

29 Affymetrix Inc. GeneChip Expression Analysis Technical Manual. 2003. carcinoma and distinguishes malignant hyperproliferation from http://gg.bu.edu/microarray/expression_manual.pdf (accessed 7 benign hyperplasia. J Invest Dermatol 2006; 126:869–81. August 2007). 49 Bowcock AM, Shannon W, Du F et al. Insights into psoriasis 30 Schindler H, Wiese A, Auer J, Burtscher H. cRNA target prepara- and other inflammatory diseases from large-scale gene expression tion for microarrays: comparison of gene expression profiles gen- studies. Hum Mol Genetics 2001; 10:1793–805. erated with different amplification procedures. Anal Biochem 2005; 50 van de Rijn M, Perou CM, Tibshirani R et al. Expression of cytoker- 344:92–101. atins 17 and 5 identifies a group of breast carcinomas with poor 31 Klur S, Toy K, Williams MP, Certa U. Evaluation of procedures for clinical outcome. Am J Pathol 2002; 161:1991–6. amplification of small-size samples for hybridization on micro- 51 Carrilho C, Alberto M, Buane L, David L. Keratins 8, 10, 13 and arrays. Genomics 2004; 83:508–17. 17 are useful markers in the diagnosis of human cervix carcin- 32 Zeeberg BR, Feng W, Wang G et al. GoMiner: a resource for bio- omas. Hum Pathol 2004; 35:546–51. logical interpretation of genomic and proteomic data. Genome Biol 52 Choi P, Chen C. Genetic expression profiles and biologic pathway 2003; 4:R28. alterations in head and neck squamous cell carcinoma. Cancer 2005; 33 Mischke D, Korge BP, Marenholz I et al. Genes encoding structural 104:1113–28. proteins of epidermal cornification and S100 calcium-binding pro- 53 Sugiura H, Ebise H, Tazawa T et al. Large-scale DNA microarray analy- teins form a gene complex (‘epidermal differentiation complex’) sis of atopic skin lesions shows expression of an epidermal differentia- on human chromosome 1q21. J Invest Dermatol 1996; 106:989–93. tion gene cluster in the alternative pathway and lack of protective gene 34 Marshall D, Hardman MJ, Nield KM, Byrne C. Differentially expression in the cornified envelope. Br J Dermatol 2005; 152:146–9. expressed late constituents of the epidermal cornified envelope. 54 Komatsu N, Saijoh K, Toyama T et al. Multiple tissue kallikrein Proc Natl Acad Sci 2001; 98:13031–6. mRNA and protein expression in normal skin and skin diseases. 35 Amano S, Akutsu N, Matsunaga Y et al. Importance of balance Br J Dermatol 2005; 153:274–81. between extracellular matrix synthesis and degradation in basement 55 Borgono CA, Michael IP, Diamandis EP. Human tissue kallikreins: membrane formation. Exp Cell Res 2001; 271:249–62. physiologic roles and applications in cancer. Mol Cancer Res 2004; 36 Copland JA, Luxon BA, Ajani L et al. Genomic profiling identifies 2:257–80. alterations in TGFb signaling through loss of TGFb receptor 56 Ruokolainen H, Paakko P, Turpeenniemi-Hujanen T. Expression of expression in human renal cell carcinogenesis and progression. matrix metalloproteinase-9 in head and neck squamous cell carci- Oncogene 2003; 22:8053–62. noma: a potential marker for prognosis. Clin Cancer Res 2004; 37 Bandyopadhyay A, Zhu Y, Malik SN et al. Extracellular domain of 10:3110–16. TGFbeta type III receptor inhibits angiogenesis and tumor growth 57 Tanimura S, Asato K, Fujishiro SH, Kohno M. Specific blockade of in human cancer cells. Oncogene 2002; 21:3541–51. the ERK pathway inhibits the invasiveness of tumor cells: down- 38 Jang JH. Reciprocal relationship in gene expression between FGFR1 regulation of matrix metalloproteinase-3 ⁄-9 ⁄-14 and CD44. Biochem and FGFR3: implication for tumorigenesis. Oncogene 2005; 24:945–8. Biophys Res Commun 2003; 304:801–6. 39 Homey B, Alenius H, Muller A et al. CCL27-CCR10 interactions reg- 58 Ramos MC, Steinbrenner H, Stuhlmann D et al. Induction of MMP- ulate T cell-mediated skin inflammation. Nat Med 2002; 8:157–65. 10 and MMP-1 in a squamous cell carcinoma cell line by ultra- 40 Gao W, Kumar S, Lotze MT, Hanning C. Innate immunity medi- violet irradiation. Biol Chem 2004; 385:75–86. ated by the cytokine IL-1 homologue 4 (IL-1H4 ⁄IL-1F7) induces 59 Takeda A, Kajiya A, Iwasawa A et al. Aberrant expression of serpin IL-12-dependent adaptive and profound antitumor immunity. squamous cell carcinoma antigen 2 in human tumor tissues and J Immunol 2003; 170:107–13. cell lines: evidence of protection from tumor necrosis factor-medi- 41 Reifenberger J, Knobbe CB, Wolter M et al. Molecular genetic anal- ated apoptosis. Biol Chem 2002; 383:1231–6. ysis of malignant melanomas for aberrations of the WNT signaling 60 Gao JQ, Tsuda Y, Katayama K et al. Antitumor effect by interleu- pathway genes CTNNB1, APC, ICAT and BTRC. Int J Cancer 2002; kin-11 receptor alpha-locus chemokine ⁄CCL27, introduced into 100:549–56. tumor cells through a recombinant adenovirus vector. Cancer Res 42 Andress DL. Insulin-like growth factor-binding protein-5 (IGFBP- 2003; 63:4420–5. 5) stimulates phosphorylation of the IGFBP-5 receptor. Am J Physiol 61 Strieter RM, Belperio JA, Phillips RJ, Keane MP. CXC chemokines Endocrinol Metab 1988; 274:E744–50. in angiogenesis of cancer. Semin Cancer Biol 2004; 14:195–200. 43 Butt AJ, Dickson KA, McDougall F, Baxter RC. Insulin-like growth 62 Vicari AP, Caux C. Chemokines in cancer. Cytokine Growth Factor Rev factor-binding protein-5 inhibits the growth of human breast can- 2002; 13:143–54. cer cells in vitro and in vivo. J Biol Chem 2003; 278:29676–85. 63 Nghiem DX, Kazimi N, Mitchell DL et al. Mechanisms underlying 44 Oh JS, Kucab JE, Bushel PR et al. Insulin-like growth factor-1 the suppression of established immune responses by ultraviolet inscribes a gene expression profile for angiogenic factors and can- radiation. J Invest Dermatol 2002; 119:600–8. cer progression in breast epithelial cells. Neoplasia 2002; 4:204–17. 64 McMahon AP. The WNT family of developmental regulators. Trends 45 Ongusaha PP, Kwak JC, Zwible AJ et al. HB-EGF is a potent inducer Genet 1992; 8:236–42. of tumor growth and angiogenesis. Cancer Res 2004; 64:5283–90. 65 Weeraratna AT, Jiang Y, Hostetter G et al. WNT5a signaling directly 46 Mori S, Tanaka M, Nanba D et al. PACSIN3 binds ADAM12 ⁄meltrin affects cell motility and invasion of metastatic melanoma. Cancer Cell alpha and up-regulates ectodomain shedding of heparin-binding 2002; 1:279–88. epidermal growth factor-like growth factor. J Biol Chem 2003; 66 Taki M, Kamata N, Yokoyama K et al. Down-regulation of WNT-4 278:46029–34. and up-regulation of WNT-5a expression by epithelial-mesenchy- 47 Del Bino S, Vioux C, Rossio-Pasquier P et al. Ultraviolet B induces mal transition in human squamous carcinoma cells. Cancer Sci 2003; hyperproliferation and modification of epidermal differentiation in 94:593–7. normal human skin grafted on to nude mice. Br J Dermatol 2004; 67 Shaulian E, Karin M. AP-1 in cell proliferation and survival. Oncogene 150:658–67. 2001; 20:2390–400. 48 Haider AS, Peters SB, Kaporis H et al. Genomic analysis defines a 68 Jochum W, Passegue E, Wagner EF. AP-1 in mouse development cancer-specific gene expression signature for human squamous cell and tumorigenesis. Oncogene 2001; 20:2401–12.

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 Microarray and confocal microscopy analysis, A. Torres et al. 1147

69 Mehic D, Bakiri L, Ghannadan M et al. Fos and jun proteins are This material is available as part of the online article from: specifically expressed during differentiation of human keratino- http://www.blackwell-synergy.com/doi/abs/10.1111/j.1365- cytes. J Invest Dermatol 2005; 124:212–20. 2133.2007.08218.x 70 Mangone FR, Brentani MM, Nonogaki S et al. Overexpression of Fos-related antigen-1 in head and neck squamous cell carcinoma. Int J Exp Pathol 2005; 86:205–12. (This link will take you to the article abstract). 71 Bieche I, Lerebours F, Tozlu S et al. Molecular profiling of inflam- matory breast cancer: identification of a poor-prognosis gene Please note: Blackwell Publishing is not responsible for the expression signature. Clin Cancer Res 2004; 10:6789–95. content or functionality of any supplementary materials sup- plied by the authors. Any queries (other than missing mate- Supplementary material rial) should be directed to the corresponding author for the article. The following supplementary material is available for this arti- cle online:

Table S1 Genes whose expression was altered in sun-exposed nonlesional skin samples and pretreatment actinic keratosis samples

2007 3M Pharmaceuticals Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1132–1147 CUTANEOUS BIOLOGY DOI 10.1111/j.1365-2133.2007.08240.x Biphasic expression of stromal cell-derived factor-1 during human wound healing A. Toksoy,* V. Mu¨ller,* R. Gillitzer* and M. Goebeler* *Department of Dermatology, University of Wu¨rzburg, Wu¨rzburg, Germany Department of Dermatology, University Hospital Mannheim, University of Heidelberg, 68135 Mannheim, Germany

Summary

Correspondence Background Chemokines tightly regulate the spatial and temporal infiltration of Matthias Goebeler. invading leucocyte subsets during wound healing. Stromal cell-derived factor-1 E-mail: [email protected] (SDF-1 ⁄CXCL12) is a homeostatic chemokine with multiple functions; its role heidelberg.de during cutaneous wound healing, however, needs to be explored. Accepted for publication Objectives To elucidate expression of the multifunctional CXC chemokine SDF-1 ⁄ 11 July 2007 CXCL12 during human wound healing. Methods Skin biopsies were obtained from 14 volunteers between 1 and 21 days Key words after incisional wounding and processed for in situ hybridization and immuno- chemokines, CXCL12, CXCR4, endothelial cells, histochemistry. stromal cell-derived factor-1, wound healing Results We analysed the spatial and temporal distribution of SDF-1 ⁄CXCL12 after Conflicts of interest artificial wounding and detected a complete downregulation at both the mRNA None declared. and the protein level within the fibrous stroma that replaces the initial wound defect. However, increased levels of SDF-1 ⁄CXCL12 were observed at the wound margins. Focusing on mediators regulating SDF-1 ⁄CXCL12 expression in vitro we realized that both tumour necrosis factor-a and interferon-c downregulated its expression in human dermal microvascular endothelial cells and fibroblasts. Conclusions Our data suggest that SDF-1 ⁄CXCL12 is tightly regulated during wound repair. Increased expression at the wound margin may contribute to the accumu- lation of endothelial progenitor cells, thus accelerating neovascularization.

Wound healing is a highly dynamic and complex process in the immune, circulatory and central nervous system.4–7 comprising three phases: inflammation, tissue formation and SDF-1 ⁄CXCL12 is constitutively expressed in primary and sec- tissue remodelling.1 These events involve the interaction of ondary lymphatic organs where it is important for physiologi- resident cells such as keratinocytes, fibroblasts, endothelial cal lymphocyte trafficking.8 Under pathological conditions, cells and mast cells as well as of infiltrating leucocyte subsets. SDF-1 ⁄CXCL12 may contribute to the recruitment of endothe- Important factors orchestrating communication between these lial progenitor cells to sites of myocardial infarction9,10 and to cells are chemokines, a family of cytokines with the predomi- the microenvironment of tumours.11 Furthermore, the SDF-1 ⁄ nant characteristic of leucocyte subtype-specific chemoattrac- CXCL12 ⁄CXCR4 axis plays a role in controlling the metastatic tion. In the context of wound healing, chemokines not only destination of tumour cells.11,12 direct migration of infiltrating and resident cells but also con- SDF-1 ⁄CXCL12 is constitutively expressed in normal human tribute to the regulation of epithelialization, tissue remodelling skin13 and may contribute to tissue homeostasis in this organ. and angiogenesis.2 The present study was initiated to elucidate the role of SDF-1 ⁄ In contrast to the majority of chemokines that are exclu- CXCL12 in the process of tissue renewal and inflammation sively expressed under pathological conditions such as inflam- during physiological skin wound healing in healthy humans. mation, the CXC chemokine stromal cell-derived factor-1 3 (SDF-1 ⁄CXCL12) is constitutively abundant in many tissues. Materials and methods SDF-1 ⁄CXCL12 binds, unlike other chemokines, nonpromiscu- ously to its receptor CXCR4 and plays crucial roles in numer- Artificial incisional wounds and biopsy specimens ous biological processes. SDF-1 ⁄CXCL12 adjusts trafficking of haematopoietic stem cells, and mice lacking SDF-1 ⁄CXCL12 or After obtaining informed consent, incisions of 5-mm dia- its receptor CXCR4 present with severe developmental defects meter were made at the ulnar forearms of 14 healthy adult

2007 The Authors 1148 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al. 1149 volunteers as described earlier.14 Seven individual series of 6- nen, Germany) and cultured in endothelial basal medium mm punch biopsies were obtained under local anaesthesia (Clonetics) supplemented with 5% fetal bovine serum, ) ) before or at defined time intervals after wounding (i.e. at days 10 ng mL 1 human epidermal growth factor, 1 lgmL 1 ) ) 1, 2, 4, 7, 10, 14 and 21) and processed for immunohisto- hydrocortisone, 50 lgmL 1 gentamicin, 50 ng mL 1 am- ) chemistry and in situ hybridization. From some volunteers add- photericin B and 12 lgmL 1 bovine brain extract (Clo- itional biopsies were obtained between days 35 and 49 after netics) and used at passage 2. Cells were characterized as wounding. Experiments were approved by the Ethics Commit- endothelial by their expression of factor VIII-related antigen tee at the University of Wu¨rzburg. (von Willebrand factor). The fibroblast cell line F-135- 60-86, kindly provided by Dr A. Trautmann (Wu¨rzburg, Germany), was cultured in Dulbecco’s modified Eagle’s med- Immunohistochemistry ium supplemented with 10% fetal calf serum. Cells were Serial cryosections were immunostained with mouse mono- stimulated with tumour necrosis factor (TNF)-a (R&D Sys- clonal antibodies (mAbs) against SDF-1 ⁄CXCL12 (clone tems) and ⁄or interferon (IFN)-c (Peprotech, London, U.K.) 79018.111; R&D Systems, Wiesbaden, Germany), CXCR4 at the concentrations and for the time periods indicated. Sub- (clones 44716.111 and 12G5; R&D Systems), EN4 (Harlan sequently, supernatants were collected and frozen until use Sera-Lab, Crawley Down, U.K.) or vimentin (Dako, Hamburg, for enzyme-linked immunosorbent assay (ELISA). Cells were Germany) using a three-step streptavidin-biotin-peroxidase harvested, coated on glass slides and prepared for in situ procedure (streptABC-peroxidase; Dako). Biotin-conjugated hybridization. sheep antimouse antibody (1 : 200; Amersham, Braun- schweig, Germany) was used as secondary antibody. After Enzyme-linked immunosorbent assay blocking nonspecific binding sites with 20% sheep serum (Dianova, Hamburg, Germany), sections were successively Concentrations of SDF-1 ⁄CXCL12 in cell culture supernatants incubated with primary antibody, biotinylated secondary anti- were determined using a commercial sandwich ELISA (R&D body and streptABC-peroxidase. Labelling was visualized by Systems) according to the manufacturer’s instructions. incubation with 3-amino-9-ethylcarbazole and slides were then counterstained with haematoxylin. For control purposes, Flow cytometry primary antibodies were omitted and replaced by isotype- matched controls which consistently yielded negative results. Cell surface expression of CXCR4 was determined by flow cytometry using mouse mAb clones 44716.111 or 12G5. After washing and labelling with fluorescein isothiocyanate-conju- In situ hybridization gated rabbit-antimouse secondary antibody (Dako), cells were A cDNA probe for SDF-1 ⁄CXCL12 was kindly provided by T. analysed by flow cytometry using a FACScan (BD, Heidelberg, Honjo (Kyoto University, Faculty of Medicine, Kyoto, Japan).3 Germany). Sense and antisense probes were prepared as previously 14 described. Paraformaldehyde-fixed cryostat sections were Results treated with proteinase K, acetylated with acetic anhydride, dehydrated in graded concentrations of alcohol, and air-dried. During human wound healing, stromal cell-derived Thereafter, sections were overlaid with the hybridization solu- factor-1 mRNA is expressed at the wound margins but tion containing the radioactively labelled heat-denaturated not in the fibrous stroma that replaces the wound radioactive sense or antisense probes. Antisense and sense (neg- ative control) probes were hybridized with at least two sections Punch biopsies from adult human skin were obtained before of the same biopsy. Nonhybridized probes were removed by and 1, 2, 4, 7, 10, 14 and 21 days after incisional wound- several high-stringency washing procedures and digestion with ing and processed for in situ hybridization using 35S-uridine RNase A and RNase T1 (Roche Molecular Diagnostics, Mann- triphosphate-labelled SDF-1 ⁄CXCL12 antisense probes. During heim, Germany). To visualize the hybridization reaction, slides the early time points, i.e. between days 1 and 4 after were dipped in NTB-2 autoradiographic emulsion (Kodak, wounding, weak to moderate SDF-1 ⁄CXCL12 mRNA signal Rochester, NY, U.S.A.), exposed for 1–4 weeks at 4 C, and intensity was detected at the wound margins (Table 1), developed. Slides were then counterstained with Papanicolaou’s comparable with the distribution pattern of the chemokine solution and microscopically evaluated. SDF-1 ⁄CXCL12 mRNA in unaffected human skin. During that time the dermal levels were evaluated semiquantitatively as: ) (0, negative), wound defect was not yet replaced by fibrous stroma, but + (1, weak), ++ (2, moderate) or +++ (3, intense). after day 4 a hypertrophic neoepidermis started to develop at the site of the previously denuded wound surface. At day 10, the wound defect was completely replaced by fibrous Cell culture stroma in all cases. Both the intensity of SDF-1 ⁄CXCL12 Human dermal microvascular endothelial cells (HDMEC) expression as well as the overall number of SDF-1 ⁄CXCL12- were obtained from Clonetics (via Cell Systems, St Kathari- positive cells were found to be increased at the wound

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 1150 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al.

Table 1 Expression of stromal cell-derived factor-1 (SDF-1 ⁄CXCL12) (a) mRNA at the wound margin

Biopsy Healthy Day Day Day Day Day Day Day ne series skin 1 2 4 7 10 14 21 1 ++ + + ++ +++ +++ +++ +++ 2++a ++ ++ ++ ++ ++ ns 3 +++ + ++ +++ +++ +++ +++ +++ 4 + + + ++ +++ +++ ++ +++ 5 + + ++ ++ +++ +++ +++ +++ 6 ++ + + ++ +++ +++ +++ +++ 7 + + ++ ++ ++ +++ +++ +++

Mean score 1Æ51Æ01Æ52Æ12Æ72Æ92Æ72Æ9 (b) aCould not be evaluated due to technical reasons. Levels of SDF-1 ⁄CXCL12 mRNA expressed before and 1–21 days after incisional wounding of adult human skin. SDF-1 ⁄CXCL12 mRNA was visualized by in situ hybridization using 35S-uridine ne triphosphate-labelled antisense probes. Expression levels were evaluated semiquantitatively as: ) (0, negative), + (1, weak), ++ (2, moderate) or +++ (3, intense). Note that in all samples ns no SDF-1 ⁄CXCL12 message could be detected in the neostroma replacing the initial provisional matrix at the site of the wound defect. Seven individual series of 6-mm punch biopsies were obtained under local anaesthesia before or at the indicated time intervals after wounding and processed for immunohistochemis- try and in situ hybridization.

(c) margins as compared with the early phases after wounding (Table 1). Importantly, however, there was no expression of the chemokine in the fibrous neostroma while endothelial ne cells, visualized by immunolabelling with mAb EN4, were abundant (Fig. 1). To elucidate whether SDF-1 ⁄CXCL12 mes- sage would reappear at later time points during the wound healing process, additional biopsies were obtained from a ns subgroup of volunteers between 5 and 7 weeks after injury. Even at such late time points after wounding SDF-1 ⁄CXCL12 mRNA was completely absent in the wound area (data not shown).

Stromal cell-derived factor-1 mRNA expression can be Fig 1. Stromal cell-derived factor-1 (SDF-1 ⁄CXCL12) mRNA is attributed to endothelial cells and fibroblasts strongly expressed at the wound margin, but is absent in the neostroma (ns). Ten days after wounding, a prominent expression We next tried to identify the cellular source of SDF-1 ⁄ of SDF-1 ⁄CXCL12 mRNA was observed at the wound margin as CXCL12. Skin biopsies of healing wound tissue were stained visualized by in situ hybridization using 35S-uridine triphosphate- by immunohistochemistry for SDF-1 ⁄CXC12 protein or analy- labelled antisense probes (b, c). In contrast, SDF-1 ⁄CXCL12 mRNA is sed by in situ hybridization for SDF-1 ⁄CXCL12 mRNA expres- virtually absent in the ns that replaced the provisional matrix at the sion (Fig. 2). Serial sections were immunolabelled for initial wound defect while endothelial cells, immunolabelled by endothelial cells using mAb EN4 (Fig. 2a, d, j) or for vimen- monoclonal antibody EN4, are abundant in that area (a). The border between ns and adjacent skin is indicated by dotted lines. ne, tin which is expressed by fibroblasts (Fig. 2l). SDF-1 ⁄CXCL12 neoepidermis. Arrows indicate EN4+ capillaries. (a, b) bright field, was found to colocalize with EN4 staining, clearly indicating (c) dark field illumination. Bar = 100 lm. that dermal microvascular endothelium is a major site of SDF-1 ⁄CXCL12 synthesis. Furthermore, spindle-shaped vimen- tin-positive cells in the dermis, most probably fibroblasts, Expression of CXCR4, the receptor for stromal expressed SDF-1 ⁄CXCL12. Importantly, immunolabelling for cell-derived factor-1 SDF-1 ⁄CXCL12 using a mAb closely resembled mRNA expres- sion patterns. A similar pattern of SDF-1 ⁄CXCL12 expression In the next step we analysed skin expression of CXCR4, the was observed in normal skin. nonpromiscuous receptor for SDF-1 ⁄CXCL12. Using a mouse

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al. 1151

(a) (b) (c) (d)

(e) (f) (g) (h)

(i) (j) (k) (l)

Fig 2. Stromal cell-derived factor-1 (SDF-1 ⁄CXCL12) is expressed mainly by endothelial cells. Serial sections of normal human skin (a–c) or of skin samples obtained from the wound margin at 2 days (d–f) or 7 days (g–l) after incisional wounding were immunostained using the endothelial cell-specific monoclonal antibody (mAb) EN4 (a, d j), a mAb against vimentin to identify fibroblasts (l), a mAb against SDF-1 ⁄ CXCL12 (g, k), or processed for in situ hybridization using an SDF-1 ⁄CXCL12-specific antisense probe (b, c, e, f, h, i). SDF-1 ⁄CXCL12 mRNA and protein partially colocalize with EN4 staining, thus indicating endothelial cells (arrows). Furthermore, fibroblasts, identified by vimentin staining (l), express SDF-1 ⁄CXCL12 (k). Arrowheads indicate fibroblasts (e, f, k, l). (a, b, d, e, g, h, j–l) bright field, (c, f, i) dark field illumination. Bar = 100 lm.

mAb specific for CXCR4 we found that this receptor, like Tumour necrosis factor-a and interferon-c downregulate SDF-1 ⁄CXCL12, is expressed mainly by endothelial cells and the expression of stromal cell-derived factor-1 in human fibroblasts as identified by EN4 immunostaining or vimentin dermal endothelial cells and fibroblasts expression, respectively (Fig. 3a–h). In addition, CXCR4 is expressed by keratinocytes in the epidermal layer (Fig. 3b, f). While significant levels of SDF-1 ⁄CXCL12 are expressed in To confirm CXCR4 expression by an independent method, normal skin and considerably higher expression is observed at isolated HDMEC and fibroblasts were studied by flow cyto- the wound margins, the chemokine is absent in the fibrous metry. Both cell types carried substantial amounts of CXCR4 neostroma. To elucidate the mechanisms of regulation, we on their surface (Fig. 3i, j). studied expression of SDF-1 ⁄CXCL12 in its major dermal

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 1152 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al.

(a) (b) (c) (d)

(e) (f) (g) (h)

(i) (j) Events Events Events 0 64 0 64 100 101 102 103 104 100 101 102 103 104 FL1-H FL1-H

Fig 3. Colocalization of CXCR4 and its ligand stromal cell-derived factor-1 (SDF-1 ⁄CXCL12). Endothelial cells [labelled by monoclonal antibody (mAb) EN4 (a, e) and indicated by arrows] as well as fibroblasts [identified by vimentin expression (d) and indicated by arrowheads] expressed both SDF-1 ⁄CXCL12 (c, g, h) and its receptor CXCR4 (b, f). (a–f) Immunohistochemistry, (g, h) in situ hybridization; (a–g) bright field, (h) dark field illumination of tissue samples obtained at day 10 (a–d) or day 14 after wounding (e–h); bar = 100 lm. (i, j) Flow cytometric analysis of CXCR4 surface expression by human dermal microvascular endothelial cells (HDMEC) (i) and fibroblasts (cell line F-135-60-86) (j). CXCR4 surface expression on HDMEC and fibroblasts was determined by flow cytometry using mAb 12G5 and is presented as black profiles. Open profiles show cells stained with isotype control antibody. producers, HDMEC and fibroblasts, after exposure to TNF-a or by incisional wounding, SDF-1 ⁄CXCL12 shows a remarkable IFN-c in vitro. Using in situ hybridization and ELISA, both cell spatial pattern of regulation. As demonstrated in this study, types were found to express considerable amounts of SDF-1 ⁄ SDF-1 ⁄CXCL12 is upregulated at the wound margins. How- CXCL12 under basal conditions. Stimulation with TNF-a or ever, starting from day 10 after wounding, SDF-1 ⁄CXCL12 is IFN-c strongly downregulated expression of SDF-1 ⁄CXCL12 at almost completely suppressed in the fibrous neostroma that both the mRNA and the protein level (Fig. 4). Coincident replaces the provisional matrix covering the initial wound exposure to both cytokines did not further suppress SDF-1 ⁄ defect whereas endothelial cells, major producers of SDF-1 ⁄ CXCL12 expression. CXCL12, are abundant at this site. Such a spatially biphasic pattern of expression of a housekeeping chemokine during the Discussion phases of wound healing is unique and has not been shown for other constitutively expressed chemokines. Studying In normal human skin, SDF-1 ⁄CXCL12 is particularly wound healing in mouse models, Fedyk et al.15 observed expressed by resident skin cells such as endothelial cells and downregulation of SDF-1 ⁄CXCL12 while Florin et al. did not fibroblasts, which points to an important role of this chemo- detect modulation of SDF-1 ⁄CXCL12 mRNA expression during kine in preserving tissue homeostasis under physiological the first 5 days after incisional wounding.16 These apparent conditions.13 When homeostasis is disturbed by injury, e.g. differences from our data that have been obtained in humans

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al. 1153

HDMEC Fibroblasts (a)35 (b) 70 30 60 25 50 20 40 15 30 10 20 % SDF-1/CXCL12 % SDF-1/CXCL12 10

5 cells mRNA-positive mRNA-positive cells mRNA-positive 0 0 Control TNF-α IFN-γ TNF-α + IFN-γ Control TNF-α IFN-γ TNF-α + IFN-γ (c) (d) 250 1000

200 800 ) ) –1 150 –1 600

100 400 (pg mL (pg mL 50 200 SDF-1/CXCL12 protein 0 SDF-1/CXCL12 protein 0 Control TNF-α IFN-γ TNF-α + IFN-γ Control TNF-α IFN-γ TNF-α + IFN-γ

Fig 4. Expression of stromal cell-derived factor-1 (SDF-1 ⁄CXCL12) by microvascular endothelial cells and fibroblasts is suppressed by tumour necrosis factor (TNF)-a and interferon (IFN)-c. Human dermal microvascular endothelial cells (HDMEC) or fibroblasts (cell line F-135-60-86) ) ) were exposed to medium as control, 2 ng mL 1 TNF-a,25ngmL 1 IFN-c, or a combination of both for 24 h. Thereafter, cells were harvested and single-cell in situ hybridization performed using a 35S-uridine triphosphate-labelled antisense probe for SDF-1 ⁄CXCL12 (a, b). The fraction of SDF-1 ⁄CXCL12 mRNA-expressing cells was determined microscopically. Data are presented as percentage of positive cells (mean ± SEM); at least five high-power fields were evaluated per sample (a, b). Furthermore, supernatants from HDMEC (c) and fibroblasts (d) were collected and analysed for SDF-1 ⁄CXCL12 secretion using a sandwich enzyme-linked immunosorbent assay. Data are shown as mean ± SEM (c, d).

may be due to the fact that both groups studied SDF-1 ⁄ fibroblasts are the major producers of SDF-1 ⁄CXCL12 during CXCL12 expression by semiquantitative polymerase chain reac- skin wound healing. Isolated HDMEC and fibroblasts, which tion, which reflects global net expression levels, but did not both produce substantial amounts of this chemokine under consider a heterogeneous spatial distribution in distinct micro- basal conditions, downregulate its expression at the mRNA compartments. A more recent study analysing incisional level when exposed to IFN-c or TNF-a. The latter observation wounds in diabetic db ⁄db mice observed an increase of is in accordance with data by Fedyk et al.15 who observed SDF-1 ⁄CXCL12 at the wound margin but did not focus on downregulation of SDF-1 ⁄CXCL12 in fibroblasts after exposure expression levels at the site of the developing neostroma.17 to TNF-a or interleukin-1a. The mechanism of regulation is Accordingly, during the initial phase of skin recovery follow- not yet clear but appears to occur independently of NF-jB ing burns an increase of SDF-1 ⁄CXCL12 protein was observed signalling as the SDF-1 ⁄CXCL12 promoter does not contain in rats.18 binding sites for this transcription factor.3 Accordingly, During the early phase of wound healing neoangiogenesis microarray analysis of endothelial cells expressing dominant- is an important part of tissue repair. Until day 4–7 after negative IKK2, which completely blocks NF-jB-dependent injury, the number of vessels in the wounded area almost signalling, did not reveal an influence of this signalling path- doubles.14 SDF-1 ⁄CXCL12 has earlier been shown to promote way in TNF-a-stimulated cells.21 Fibroblasts lacking the AP-1 proliferation and migration of endothelial cells in vitro while transcription factor subunit JUN do not express any SDF-1 ⁄ subcutaneous injection into mice induced formation of blood CXCL12, indicating that this signalling component is a pre- vessels in vivo.19 More recent data indicated that SDF-1 ⁄ requisite for basal expression.16 Similar to endothelial cells, CXCL12 presumably plays an important role in recruiting fibroblasts proliferate during wound repair. Dynamics of pro- a proangiogenic subpopulation of haematopoietic cells to liferation correlate with changes in SDF-1 ⁄CXCL12 expression. peripheral tissues: while expression of established angiogenic Upon suppression of SDF-1 ⁄CXCL12 during the late course of factors such as vascular endothelial growth factor is involved wound healing, fibroblast proliferation ceases and matrix in mobilizing and directing such bone marrow-derived circu- synthesis is downregulated,22 resulting in transformation of lating cells to peripheral organs, SDF-1 ⁄CXCL12 fulfils the task fibroblast-rich granulation tissue into a relatively acellular scar. to position and retain these cells at such peripheral sites.20 At SDF-1 ⁄CXCL12 released by dermal cells such as fibroblasts later time points of the wound healing process neoangiogene- may further enhance keratinocyte proliferation via CXCR4 and sis ceases and the number of existing vessels subsequently potentially contribute to re-epithelialization. Coculture experi- remains constant. This event clearly coincides with suppres- ments with fibroblasts and keratinocytes demonstrated that sion of SDF-1 ⁄CXCL12 expression in the fibrous stroma of proliferation of the latter occurred in a SDF-1 ⁄CXCL12-depen- the wound. As shown in our study, endothelial cells and dent manner.16

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 1154 Expression of SDF-1 ⁄CXCL12 during wound healing, A. Toksoy et al.

Beyond its classical effects on leucocyte attraction, SDF-1 ⁄ 8 Kim CH, Broxmeyer HE. Chemokines: signal lamps for trafficking CXCL12 may thus interfere with central events occurring of T and B cells for development and effector function. J Leukoc Biol during wound repair that include regulation of neoangiogene- 1999; 65:6–15. 9 Askari AT, Unzek S, Popovic ZB et al. Effect of stromal-cell-derived sis, fibroblast proliferation and epithelialization. However, factor 1 on stem-cell homing and tissue regeneration in ischaemic there are currently conflicting data whether interference with cardiomyopathy. Lancet 2003; 362:697–703. the SDF-1 ⁄CXCL12 ⁄CXCR4 axis promotes or disturbs wound 10 Abbott JD, Huang Y, Liu D et al. Stromal cell-derived factor-1alpha healing. While the SDF-1 ⁄CXCL12 inducer dibutyryl cAMP, plays a critical role in stem cell recruitment to the heart after myo- which is used for the topical treatment of skin ulcers in Japan, cardial infarction but is not sufficient to induce homing in the promotes wound healing,17 another study provided data sug- absence of injury. Circulation 2004; 110:3300–5. gesting that inhibition of SDF-1 ⁄CXCL12 ⁄CXCR4 interaction 11 Burger JA, Kipps TJ. CXCR4: a key receptor in the crosstalk between tumor cells and their microenvironment. Blood 2006; may improve it.18 Our study that addressed the spatial and 107:1761–7. temporal dynamics of SDF-1 ⁄CXCL12 expression during 12 Muller A, Homey B, Soto H et al. Involvement of chemokine recep- human skin wound healing suggests that both site and time tors in breast cancer metastasis. Nature 2001; 410:50–6. point of interference with this ‘master’ chemokine of wound 13 Pablos JL, Amara A, Bouloc A et al. Stromal-cell derived factor is repair might affect the outcome of intervention. We hope that expressed by dendritic cells and endothelium in human skin. Am J our study will encourage further work in this sector. Pathol 1999; 155:1577–86. 14 Engelhardt E, Toksoy A, Goebeler M et al. Chemokines IL-8, Groa MCP-1, IP-10, and Mig are sequentially and differentially expressed Acknowledgments during phase-specific infiltration of leukocyte subsets in human wound healing. Am J Pathol 1998; 153:1849–60. This study is part of the M.D. thesis of A.T. and was sup- 15 Fedyk ER, Jones D, Critchley HO et al. Expression of stromal- ported by the Wilhelm-Sander-Stiftung (grant 95.064.3). The derived factor-1 is decreased by IL-1 and TNF and in dermal authors thank Sybille Schmid for performing the flow cyto- wound healing. J Immunol 2001; 166:5749–54. metry studies. 16 Florin L, Maas-Szabowski N, Werner S et al. Increased keratinocyte proliferation by JUN-dependent expression of PTN and SDF-1 in fibroblasts. J Cell Sci 2005; 118:1981–9. References 17 Asai J, Takenaka H, Katoh N et al. Dibutyryl cAMP influences endo- thelial progenitor cell recruitment during wound neovasculariza- 1 Singer AJ, Clark RAF. Cutaneous wound healing. N Engl J Med 1999; tion. J Invest Dermatol 2006; 126:1159–67. 341:738–46. 18 Avniel S, Arik Z, Maly A et al. Involvement of the CXCL12 ⁄CXCR4 2 Gillitzer R, Goebeler M. Chemokines in cutaneous wound healing. pathway in the recovery of skin following burns. J Invest Dermatol J Leukoc Biol 2001; 69:513–21. 2006; 126:468–76. 3 Shirozu M, Nakano T, Inazawa J et al. Structure and chromosomal 19 Salcedo R, Wasserman K, Young HA et al. Vascular endothelial localization of the human stromal cell-derived factor 1 (SDF1) growth factor and basic fibroblast growth factor induce expression gene. Genomics 1995; 28:495–500. of CXCR4 on human endothelial cells: in vivo neovascularization 4 Nagasawa T, Hirota S, Tachibana K et al. Defects of B-cell lympho- induced by stromal-derived factor-1alpha. Am J Pathol 1999; poiesis and bone-marrow myelopoiesis in mice lacking the CXC 154:1125–35. chemokine PBSF ⁄SDF-1. Nature 1996; 382:635–8. 20 Grunewald M, Avraham I, Dor Y et al. VEGF-induced adult neovas- 5 Tachibana K, Hirota S, Iizasa H et al. The chemokine receptor cularization: recruitment, retention, and role of accessory cells. Cell CXCR4 is essential for vascularization of the gastrointestinal tract. 2006; 124:175–89. Nature 1998; 393:591–4. 21 Viemann D, Goebeler M, Schmid S et al. Transcriptional profiling 6 Ma Q, Jones D, Borghesani PR et al. Impaired B-lymphopoiesis, of IKK2 ⁄NF-kappa B- and p38 MAP kinase-dependent gene expres- myelopoiesis, and derailed cerebellar neuron migration in sion in TNF-alpha-stimulated primary human endothelial cells. CXCR4- and SDF-1-deficient mice. Proc Natl Acad Sci USA 1998; Blood 2004; 103:3365–73. 95:9448–53. 22 Clark RAF. Wound repair: overview and general considerations. In: 7 Zou YR, Kottmann AH, Kuroda M et al. Function of the chemokine The Molecular and Cellular Biology of Wound Repair (Clark RAF, ed.), 2nd receptor CXCR4 in haematopoiesis and in cerebellar development. edn. New York: Plenum Press, 1996; 1–50. Nature 1998; 393:595–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1148–1154 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08200.x CTACK ⁄CCL27 expression in psoriatic skin and its modification after administration of etanercept A. Campanati, G. Goteri,* O. Simonetti, G. Ganzetti, K. Giuliodori, D. Stramazzotti,* D. Morichetti, M.L. Bernardini, B. Mannello, G. Fabris* and A. Offidani Dermatological Clinic, Department of Medical Sciences, and *Anatomic Pathology, Department of Neurosciences, Ancona Hospital, Polytechnic University of Marche Region, Via Conca 71, Ancona 60020, Italy

Summary

Correspondence Background Tumour necrosis factor-a upregulates the expression of a cutaneous Anna Campanati. T cell-attracting chemokine (CTACK ⁄CCL27), that promotes migration of cutane- E-mail: [email protected]; ous lymphocyte-associated antigen-positive lymphocytes into the skin. The role [email protected] of CTACK ⁄CCL27 in pathogenesis of psoriasis has recently been documented but Accepted for publication no data are available at the present time on its modification in psoriatic cutane- 23 June 2007 ous tissue after administration of etanercept. Objectives To evaluate modifications of CTACK ⁄CCL27 expression in skin of Key words patients with psoriasis after administration of etanercept and their relation with CCL27, CTACK, etanercept, psoriasis disease activity. Conflicts of interest Methods Twenty-two patients with moderate to severe psoriasis underwent clinical, None declared. histological and immunohistochemical evaluations of disease activity at baseline and at 12 and 24 weeks after starting treatment with etanercept. Results All selected patients experienced an improvement of Psoriasis Area and Severity Index (PASI) score (P <0Æ001) and Dermatology Life Quality Index score (P <0Æ001) during the treatment. Skin histological abnormalities showed statistically significant modifications during treatment (P <0Æ001). Immunohisto- chemical expression of CTACK ⁄CCL27 decreased significantly (P <0Æ001) and its relation with final PASI score was statistically significant (P <0Æ05); the pattern of distribution of CTACK ⁄CCL27 immunoreactivity significantly moved from dif- fuse and predominantly suprabasal to basal (P <0Æ001) and the restoration of basal distribution of CTACK ⁄CCL27 was also significantly related to clinical improvement of cutaneous disease (P <0Æ001). Conclusions Etanercept induces a clinical and histological improvement of psoriatic disease, promoting a reduction in CTACK ⁄CCL27 cutaneous immunostaining and favouring the restoration of physiological CTACK ⁄CCL27 epidermal expression. Moreover, CTACK ⁄CCL27 reduction in cutaneous expression during administra- tion of etanercept could be considered a favourable prognostic marker.

In psoriasis vulgaris, a type 1, autoimmune disease model,1 family, a ligand for CC chemokine receptor 10 (CCR10).5–7 histological modifications in involved skin and subsequent dis- This chemokine selectively attracts cutaneous lymphocyte- ease expression are sustained and promoted by multiple pro- associated antigen-positive (CLA+), CCR10-positive memory T inflammatory pathways, where cytokines and chemokines cells into the inflammatory sites,5,8 and its key role in patho- produced by T cells and epidermal keratinocytes play a major genesis of psoriasis has recently been highlighted by several role.2,3 studies in the literature.2–4,9,10 Among the proinflammatory cytokines with abnormal Biological response modifiers are a group of different pro- expression in psoriatic cutaneous lesions, some are directly teins able to modify the physiopathological pathways of psori- regulated by local tumour necrosis factor (TNF)-a,4 which asis, which have been successfully introduced in the treatment favours or inhibits their synthesis. In particular, TNF-a of psoriasis. In particular, etanercept is a human TNF-a recep- upregulates the expression of a cutaneous T cell-attracting tor made from the fusion of two naturally occurring TNF-a chemokine4 (CTACK ⁄CCL27) belonging to the CC chemokine receptors that binds to TNF-a with greater affinity than do

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 1155 1156 CTACK ⁄CCL27, psoriasis and etanercept, A. Campanati et al. natural receptors.11 This binding makes the bound TNF-a bio- to the basal layer or involving the suprabasal malpighian layer. logically inactive, resulting in a reduction of inflammation. Slides were evaluated three times and separately by two expert The aim of the present study was to evaluate whether the observers blind to the clinical correlates. Discordant results were administration of etanercept in patients with psoriasis is discussed until a complete agreement was reached for each related to any changes in CTACK ⁄CCL27 expression of keratin- sample. ocytes and if these modifications are related to clinical disease A nonparametric analysis of variance for repeated measure- expression. ment (Wilcoxon rank test) and a nonparametric test for bivar- iate tabular analysis (v2) were used to evaluate relations Materials and methods between PASI score, DLQI score, histological score of disease activity, immunostaining of CTACK ⁄CCL27 and CTACK ⁄CCL27 Twenty-two patients were enrolled in the study: 12 patients immunohistochemical pattern of distribution in the epidermis, received etanercept 25 mg twice weekly and 10 patients across the time points T0, T12 and T24. received 50 mg twice weekly for 24 weeks of treatment. Ethi- cal committee approval was obtained for the study and all the patients enrolled gave informed consent. 9 All patients were carefully followed up for clinical response, F(2, 63) = 55·7 8 P < 0·001 and Psoriasis Area and Severity Index (PASI) score and Derma- Mean ± tology Life Quality Index (DLQI) score were calculated at Mean SE 7 Mean ± 1·96SE baseline (T0), 12 weeks (T12) and 24 weeks (T24) of treat- ment with etanercept. All patients underwent a lesional skin 6 biopsy at T0, T12 and T24 in order to evaluate the histologic- al profile of skin lesions and all the modifications that PASI 5 occurred during treatment. All the biopsies were taken in the centre of one selected lesion located on the extensor surface 4 of the left arm. To compute histological changes of skin lesions we applied 3 the following scoring system: 0, normal skin with flattened 2 epidermis and normal papillary dermis without prominent 0 12 24 vessels and inflammatory cells; 1, skin with normal epidermis Time of biopsy and dermis, containing sparse lymphocytes around vessels; 2, skin with slight to moderate squamous hyperplasia and Fig 1. Graph showing the clinical improvement of lesions, expressed parakeratosis and active inflammation; and 3, skin with by a significant decrease of mean Psoriasis Area and Severity Index marked squamous hyperplasia and parakeratosis and active (PASI) score during administration of etanercept (Wilcoxon rank test; inflammation. P <0Æ001). We also performed immunohistochemical staining for CTACK ⁄CCL27 in the 66 skin biopsy specimens collected, fixed in formalin and embedded in paraffin. Briefly, 4–5 lm thick 18 tissue sections were placed in antigen retrieval citrate buffer pH 16 6Æ0, microwaved for 10 min at 95–98 C and allowed to cool F(2, 63) = 29·42 P < 0·001 for 20 min. After blocking endogenous peroxidase activity, 14 Mean ± sections were incubated overnight with mouse antihuman Mean SE 12 Mean ± 1·96SE CTACK ⁄CCL27 (R&D Systems, Minneapolis, MN, U.S.A.; dilution 1 : 200) at 4 C, then for 30 min at room tempera- 10 ture with antimouse immunoglobulin ⁄horseradish peroxidase DLQI (EnvisionTM ⁄HRP; Dakocytomation, Milan, Italy). Subsequently, 8 slides were washed with Tris-buffered saline and treated with 6 3,3¢-diaminobendizidine until brown staining was visible, then counterstained with Mayer’s haematoxylin, dehydrated, and 4 mounted in Permount (Biomeda, Foster City, CA, U.S.A.). 2 Immunostaining intensity was described as mild, moderate and 0 12 24 strong in each sample, and expressed semiquantitatively using Time of biopsy the following scoring system: 0, staining in up to 10% keratinocytes; 1, staining in < 25% keratinocytes (+); 2, stain- Fig 2. Graph showing the clinical improvement of lesions, expressed ing in ‡ 25% and < 50% keratinocytes (++); and 3, strong by a significant decrease in mean Dermatology Life Quality Index staining in ‡ 50% keratinocytes (+++). The pattern of distri- (DLQI) score during administration of etanercept (Wilcoxon rank test; bution of immunostaining was also described, whether limited P <0Æ001).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 CTACK ⁄CCL27, psoriasis and etanercept, A. Campanati et al. 1157

Results Histological response

Patients showed a marked improvement of the lesions also at Clinical response histological level. Before starting treatment all patients had All the patients experienced a clinical improvement of cutane- active psoriatic lesions, 21 with marked squamous hyperplasia ous disease with a significant decrease in PASI score (Fig. 1) and parakeratosis and active inflammation (score 3, Fig. 3a), and DLQI score (Fig. 2), through the 24 weeks of treatment. and one with moderately active lesions (score 2).

(a) (b)

(c) (d)

(e) (f)

Fig 3. Photographs showing the modification of the histological features (a, c, e) and CTACK ⁄CCL27 pattern of expression (b, d, f) during treatment with etanercept. (a, b) Before treatment, a markedly active lesion shows prominent epidermal hyperplasia and parakeratosis (a) and a diffuse and strong expression of CTACK ⁄CCL27 in the lesional skin with a prevalent staining of suprabasal malpighian layers (b). (c, d) After 12 weeks of treatment, a moderately active lesion shows epidermal hyperplasia and parakeratosis (c) and a diffuse and strong expression of CTACK ⁄CCL27 in the epidermis with a prevalent staining of suprabasal malpighian layers (d). (e, f) After 24 weeks of treatment, normal skin histology has been restored (e), along with basal expression of CTACK ⁄CCL27 (f).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 1158 CTACK ⁄CCL27, psoriasis and etanercept, A. Campanati et al.

3·5 3·2 3·0 3·0 F(2, 63) = 97·01 F(2, 63) = 31·28 2·8 P < 0·001 P < 0·001 Mean 2·5 Mean ± Mean ± SE 2·6 Mean SE ± Mean ± 1·96SE 2·0 Mean 1·96SE 2·4 2·2 1·5 Histology 2·0 1·0 Epidermal CCL27 1·8 0·5 1·6 1·4 0·0 0 12 24 1·2 Time of biopsy 0 12 24 Time of biopsy Fig 4. Graph showing the improvement of lesions at the histological level, expressed by a significant decrease of mean histological score Fig 5. Graph showing the reduction of CTACK ⁄CCL27 mean during administration of etanercept (Wilcoxon rank test; P <0Æ001). expression scores during administration of etanercept (Wilcoxon rank test; P <0Æ001).

At T12, the skin became histologically completely normal Table 1 Modification of the pattern of distribution of CTACK ⁄CCL27 in nine cases (score 0): the epidermis appeared flattened, the through the epidermis during the administration of etanercept papillary dermis was reduced in depth and also vessels were not prominent. Seven cases exhibited a normal architecture of Baseline 12 weeks 24 weeks Total the epidermis and dermis, but a few lymphocytes were still Basal immunostaining 0 10 20 30 present around the vessels (score 1); the other six patients Basal and suprabasal 22 12 2 36 showed squamous hyperplasia with parakeratosis and a mild immunostaining active inflammation consistent with score 2 histological disease Total 22 22 22 66 (Fig. 3c). Degrees of freedom = 2; v2 =36Æ67; P <0Æ01. At T24 the skin was histologically normal in 15 cases (score 0, Fig. 3e), and showed only sparse perivascular lymphocytes in six cases (score 1) and a slightly to moderately active dis- ease in one (score 2). (Fig. 6); moreover, the reduction in CTACK ⁄CCL27 epidermal The skin histology showed statistically significant modifica- expression directly correlated with the final PASI score tions across the time of treatment (Fig. 4). (Fig. 7).

Immunostaining for CTACK ⁄CCL27 Discussion

Epidermis was immunoreactive for CTACK ⁄CCL27 in all 66 The effect of etanercept on clinical expression of psoriatic dis- skin samples examined. Keratinocytes of the malpighian layer ease has been widely demonstrated, as well as its ability to were moderately to strongly stained in the lesional skin before improve the quality of life in treated patients.11–16 Our starting treatment, with a diffuse pattern of distribution and a data confirm that etanercept is able to induce a clinical and prevalent staining in the suprabasal malpighian layers histological improvement of psoriatic disease, with a gradual (Fig. 3b). The expression of CTACK ⁄CCL27 decreased during disappearance of epidermal hyperplasia, granulocytic and treatment with etanercept: the mean values of immunostaining lymphocytic infiltration, and dermal angiogenesis, in most scores reduced progressively at T12 and T24, and the differ- treated patients. ences were statistically significant (Fig. 5). The pattern of dis- To the best of our knowledge, no studies have been per- tribution of CTACK ⁄CCL27 immunoreactivity also tended to formed concerning the immunohistochemical modifications of change from diffuse and predominantly suprabasal (Fig. 3d) CTACK ⁄CCL27 expression in psoriasis during exposure to to basal (Fig. 3f): this pattern modification was observed in etanercept. 10 patients at T12 and in 20 patients at T24. These changes CTACK ⁄CCL27 is expressed only in the skin and not in were statistically significant (Table 1). Restoration of basal dis- other organs and is expressed constitutively especially in epi- tribution of CTACK ⁄CCL27 expression was significantly related dermal keratinocytes.5 In addition, previous in vitro data to clinical improvement of cutaneous disease manifestations revealed that production and expression of CTACK ⁄CCL27 in

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 CTACK ⁄CCL27, psoriasis and etanercept, A. Campanati et al. 1159

psoriasis and normal skin,17 Kakinuma et al.3 reported that 8·0 CTACK ⁄CCL27 is strongly expressed all over the keratinocyte 7·5 layers in patients with psoriasis, in contrast to healthy control F(1, 64) = 3·159 subjects who exhibit only moderate expression in the basal 7·0 P < 0·001 layers of epidermis. During the normal process of differentia- Mean Mean ± SE tion, the suprabasal keratinocytes lose the capability to pro- 6·5 Mean ± 1·96SE duce this protein, whereas in psoriatic skin the suprabasal cells 6·0 maintain a strong CTACK ⁄CCL27 expression due to an acceler- ated proliferation and a reduced differentiation.9

PASI 5·5 In agreement with Kakinuma et al.3 we also found an abundant production of CTACK ⁄CCL27 in psoriasis lesional 5·0 keratinocytes, particularly in the upper malpighian layers 4·5 which are expanded in the active lesions. Moreover, as CTACK ⁄CCL27 expression is upregulated by TNF-a, the 4·0 administration of the TNF-a inhibitor etanercept resulted in a significant modification of CTACK ⁄CCL27 expression in 3·5 12 psoriatic epidermis. Our data showed that etanercept induces a reduction of CTACK ⁄CCL27 immunohistochemical staining intensity and also favours the restoration of a gradient typi- Fig 6. Graph showing Psoriasis Area and Severity Index (PASI) score cal of nonpsoriatic skin with major expression in the lower mean values according to the pattern of distribution of CTACK ⁄CCL27 during administration of etanercept (Wilcoxon rank test; P <0Æ001). layers. Moreover, CTACK ⁄CCL27 gradient restoration is 1, basal and suprabasal CTACK ⁄CCL27 expression; 2, basal CTACK ⁄ related to clinical improvement of cutaneous disease, and CCL27 expression. reduction in its epidermal expression correlates with the final PASI score, so that it could be considered a favourable prognostic marker. 3·2 It appears that CTACK ⁄CCL27 provides a reason, at least in 3·0 part, for the therapeutic mechanism of action of etanercept, 2·8 and we think that larger immunohistochemical studies, includ- 2·6 ing the analysis of other cytokines simultaneously implicated 2·4 in different psoriatic pathogenetic pathways, are required, bet- 2·2 2·0 ter to elucidate the role of etanercept in disease control. 1·8 1·6 References

Epidermal CCL27 1·4 1·2 1 Lowes MA, Chamian F, Abello MV et al. Increase in TNF-alpha and 1·0 inducible nitric oxide synthase-expressing dendritic cells in psoria- 0·8 –1 0 12345 sis and reduction with efalizumab (anti-CD11a). Proc Natl Acad Sci PASI USA 2005; 102:19057–62. r = –0·4379; P = 0·04379 2 Giustizieri ML, Mascia F, Frezzolini A et al. Keratinocytes from patients with atopic dermatitis and psoriasis show a distinct chemokine production profile in response to T cell-derived cyto- Fig 7. Graph showing the correlation between final Psoriasis Area and kines. J Allergy Clin Immunol 2001; 107:871–7. Severity Index (PASI) mean score and CTACK ⁄CCL27 immunostaining 3 Kakinuma T, Saeki H, Tsunemi Y et al. Increased serum cutaneous scores (Wilcoxon rank test; r = )0Æ4379; P =0Æ04379). The red T cell-attracting chemokine (CCL27) levels in patients with atopic dotted lines indicate 95% confidence intervals and the blue circles dermatitis and psoriasis vulgaris. J Allergy Clin Immunol 2003; represent the PASI scores at 24 weeks for the corresponding CTACK ⁄ 111:592–7. CCL27 scores. There are fewer than 22 blue circles because some are 4 Banno T, Gazel A, Blumenberg M. Effects of tumor necrosis factor- in identical positions on the graph. alpha (TNF alpha) in epidermal keratinocytes revealed using global transcriptional profiling. J Biol Chem 2004; 279:32633–42. keratinocytes is upregulated by stimulation with both interleu- 5 Vestergaard C, Johansen C, Otkjaer K et al. Tumor necrosis factor- kin-12 and TNF-a9,17 and that CTACK ⁄CCL27 and its receptor alpha induced CTACK ⁄CCL27 (cutaneous T-cell-attracting chemo- CCR10 provide a specific signal for the recruitment of CLA+ kine) production in keratinocytes is controlled by nuclear factor memory T cells into the skin.9 The role of this protein in the jB. Cytokine 2005; 29:49–55. 6 Morales J, Homey B, Vicari AP et al. CTACK, a skin-associated pathogenesis of psoriasis has been widely demonstrated in the 3,4,9 chemokine that preferentially attracts skinhoming memory T cells. literature. Proc Natl Acad Sci USA 1999; 96:14470–5. Although quantitative polymerase chain reaction analysis 7 Ishikawa-Mochizuki I, Kitaura M, Baba M et al. Molecular cloning showed no significant differences in global content of of a novel CC chemokine, interleukin-11 receptor a-locus chemo- CTACK ⁄CCL27 between the lesional skin of patients with kine (ILC), which is located on chromosome 9p13 and a potential

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 1160 CTACK ⁄CCL27, psoriasis and etanercept, A. Campanati et al.

homologue of a CC chemokine encoded by molluscum contagio- 13 Papp KA, Tyring S, Lahfa M et al. Etanercept Psoriasis Study Group. sum virus. FEBS Lett 1999; 460:544–8. A global phase III randomized controlled trial of etanercept in pso- 8 Kunkel EJ, Butcher EC. Chemokines and the tissue-specific migra- riasis: safety, efficacy, and effect of dose reduction. Br J Dermatol tion of lymphocytes. Immunity 2002; 16:1–4. 2005; 152:1304–12. 9 Reiss Y, Proudfoot AE, Power CA et al. CC chemokine receptor 14 Feldman SR, Kimball AB, Krueger GG et al. Etanercept improves the (CCR) 4 and the CCR10 ligand cutaneous T cell attracting chemo- health-related quality of life of patients with psoriasis: results of a kine (CTACK) in lymphocyte trafficking to inflamed skin. J Exp Med phase III randomized clinical trial. J Am Acad Dermatol 2005; 53:887–9. 2001; 194:1541–7. 15 Gottlieb AB, Leonardi CL, Goffe BS et al. Etanercept monotherapy 10 Homey B, Alenius H, Muller A et al. CCL27–CCR10 interactions in patients with psoriasis: a summary of safety, based on an inte- regulate T cell-mediated skin inflammation. Nat Med 2002; 8:157– grated multistudy database. J Am Acad Dermatol 2006; 54:S92–100. 65. 16 Gordon K, Korman N, Frankel E et al. Efficacy of etanercept in an 11 Weinberg JM. An overview of infliximab, etanercept, efalizumab, integrated multistudy database of patients with psoriasis. J Am Acad and alefacept as biologic therapy for psoriasis. Clin Ther 2003; Dermatol 2006; 54:S101–11. 25:2487–505. 17 Homey B, Wang W, Soto H et al. The orphan chemokine receptor 12 Krueger GG, Langley RG, Finlay AY et al. Patient-reported outcomes G protein-coupled receptor-2 (GPR-2, CCR10) binds the skin asso- of psoriasis improvement with etanercept therapy: results of a ran- ciated chemokine CCL27 (CTACK ⁄ALP ⁄ILC). J Immunol 2000; domized phase III trial. Br J Dermatol 2005; 153:1192–9. 164:3465–70.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1155–1160 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08197.x Prevalence of Staphylococcus aureus toxins and nasal carriage in furuncles and impetigo F. Durupt,* L. Mayor,* M. Bes,* M.-E. Reverdy,* F. Vandenesch,* L. Thomas and J. Etienne* *INSERM, U851, 69008 Lyon, France; Universite´ Lyon 1, Centre National de Re´fe´rence des Staphylocoques, Faculte´ Laennec, 69008 Lyon, France Service de Dermatologie, Hoˆtel Dieu & Universite´ Lyon 1, 69288 Lyon cedex 02, France

Summary

Correspondence Background The precise role of Staphylococcus aureus toxins and nasal carriage in com- Franc¸ois Durupt. mon skin infections remains unclear. E-mail: [email protected] Objectives To seek correlations between toxin expression, S. aureus nasal carriage and clinical manifestations in patients with community-acquired furuncles and Accepted for publication 27 June 2007 impetigo. Methods From November 2004 to August 2005, we studied clinical data and bac- Key words teriological samples prospectively collected from 121 patients presenting with exfoliative toxins, furunculosis, impetigo, furuncles or impetigo. methicillin resistance, Panton–Valentine leukocidin, Results Sixty-four patients (31 with furuncles and 33 with impetigo) had S. aureus- Staphylococcus aureus positive skin culture. Panton–Valentine leukocidin (PVL) genes were present in Conflicts of interest 13 of 31 (42%) isolates from furuncles and were associated with epidemic None declared. furunculosis. Exfoliative toxin genes were present in 10 of 10 (100%) and 12 of 21 (57%) bullous and nonbullous impetigo isolates, respectively. Nasal carriage of S. aureus was found in 58% of patients overall. It was strongly associated with chronic furunculosis but not with simple furuncles (88% vs. 29%, P <0Æ007). Skin and nose isolates from a given patient always had identical characteristics. Methicillin-resistant S. aureus accounted for four of 64 (6%) positive skin cultures. Conclusions PVL is not involved in all types of furuncles but is associated with epi- demic furunculosis. Both bullous and nonbullous forms of impetigo are associ- ated with exfoliative toxins. Staphylococcus aureus nasal carriage is associated with the chronicity of furuncles.

Staphylococcus aureus is a frequent cause of skin and soft-tissue chronic staphylococcal infections.7,8 Nasal decontamination infections. It produces numerous virulence factors, including with mupirocin can prevent relapses of chronic furunculosis.9 adhesins and toxins. Some toxins tend to be related with In this prospective study we analysed S. aureus strains isolated particular types of skin infection. Panton–Valentine leukocidin from patients with skin infections and nasal carriage, seen by (PVL) genes tend to be expressed by S. aureus isolates from a network of dermatologists working in private practices or furuncles and skin abscesses, while exfoliative toxins (ETs) are hospital outpatient clinics. The aim was to identify links associated with bullous impetigo and the scalded skin syn- between clinical manifestations and toxin gene expression. We drome. Rare reports mention the detection of ETs in S. aureus also compared the S. aureus isolates recovered from the skin strains from nonbullous impetigo.1 Most S. aureus infections lesion(s) and nose of each patient. Finally, we determined the occurring in the community are due to methicillin-sensitive prevalence of MRSA. strains, but PVL-positive community-acquired methicillin- resistant S. aureus (CA-MRSA) strains are now emerging world- Patients and methods wide, especially in the U.S.A.2–4 ET-expressing CA-MRSA isolates have seldom been detected (mainly in Japan and Patients Switzerland).5,6 Staphylococcus aureus skin and soft-tissue infections are very fre- Between 26 November 2004 and 26 August 2005 we pro- quent in the community, and most respond rapidly to simple spectively collected standard data on patients presenting with local treatment. Some become extensive or chronic. It has furuncles or impetigo to a network of 100 dermatologists been shown that S. aureus nasal carriage is more frequent in working in private practices or hospital outpatient clinics in

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 1161 1162 S. aureus skin infections, F. Durupt et al. the Rhoˆne-Alpes region of France. The dermatologists were tion using a standard procedure.10 Sequences specific for asked to recruit all patients presenting with furuncles or impe- PVL genes (lukS-PV, lukF-PV), ET genes A, B and D (eta, etb, tigo. Patients with folliculitis or other infectious skin diseases etd), staphylococcal enterotoxin A, B, C, D, E, H, K, L, M or with an uncertain clinical diagnosis were excluded. The and O genes (sea–see, seh, sek–sem, seo), the toxic shock syn- dermatologists classified the furuncles as simple when solitary, drome toxin gene (tst), epidermal-cell differentiation inhibi- acute and multiple when several were disseminated in a patient tor (EDIN) genes, staphylococcal leukocidin Luk-DE and with no previous history of furuncles, carbuncles when multiple Luk-M genes (lukE–lukD, lukM), haemolysin genes [gamma furuncles occurred on the same part of the body, and chronic if (hlg), gamma variant (hlgv) and beta (hlb)], and accessory several furuncles had occurred in the previous 6 months. gene regulator alleles (agr-1 to agr-4) were detected by PCR, Furuncles were considered to be epidemic if more than one as previously described.11,12 case occurred within a family. Impetigo was classified as bullous The mecA gene, which codes for methicillin resistance, was if bullae were present, nonbullous if there were crusts but no detected by PCR as described by Murakami et al.13 Amplifica- bullae, and secondary if the impetigo occurred on a pre-existing tion of gyrA was used to confirm the quality of each DNA skin lesion. extract and the absence of PCR inhibitors.14 All PCR products The following data were recorded in each case: clinical pre- were analysed by electrophoresis through 1% agarose gels sentation, number of lesions, demographic characteristics, (Eurobio, Courtaboeuf, France). Isolates were genotyped by underlying conditions (atopic dermatitis, diabetes mellitus, pulsed-field gel electrophoresis (PFGE) after SmaI restriction, as immunosuppressive treatment, or a personal or familial his- previously described.15,16 The PFGE patterns were digitized tory of furuncles or impetigo), risk factors for MRSA (recent and analysed with the Taxotron typing system (Institut Pas- hospitalization, contact with a hospitalized person or health- teur, Paris, France). Strain relatedness was determined accord- care worker in the previous 10 days), and antibiotic or anti- ing to published guidelines.17 Isolates that differed by no septic treatment before admission. The local ethics committee more than three fragments were considered to be subtypes of approved the protocol. a given clonal type.

Specimen collection, bacterial identification and Statistical analysis susceptibility testing Categorical data were compared by using the v2 test, Fisher’s Two bacteriological samples were collected by physicians from exact test and Fisher–Snedecor’s test, implemented with each patient, using cotton wool swabs (Oxoid, Basingstoke, Epi-Info software version 6.0 (Centers for Disease Control and U.K.): one from the skin lesion, and the second on the ant- Prevention, Atlanta, GA, U.S.A.). P <0Æ05 was considered erior nares (to detect S. aureus nasal carriage). Patients with statistically significant. missing or unidentified swabs were excluded. The swabs were placed in Stuart medium (allowing survival of bacteria during Results transport) and mailed to the French Reference Center for Staphylococci. On reception the swabs were plated on CHRO- Patient characteristics and results of skin culture Magar Staph Aureus (CHROMagar, Paris, France), a chromo- genic medium specific for S. aureus. Skin swabs from patients Thirty-five dermatologists working in private practices and with impetigo were also plated on blood agar (bioMe´rieux, nine dermatologists working in seven hospital dermatology Marcy l’E´toile, France) to detect Streptococcus pyogenes. Staphylococcus outpatient clinics collected data and samples from 123 aureus was identified on the basis of colony and microscopic patients. Two patients were excluded because bacteriological morphology, coagulase testing with rabbit plasma (bio- samples were missing (one patient) or unidentified (one Me´rieux) and the Pastorex Staph-Plus kit (Bio-Rad, Marnes- patient). The final analysis therefore involved 121 patients. la-Coquette, France). Streptococcus pyogenes was identified on the The median age was 30Æ2 years and the sex ratio was 1Æ09 basis of colony and microscopic morphology, b-haemolysin (63 males ⁄58 females). Seventy-two patients presented with evidence on blood agar (trypticase-soy agar with 5% horse furuncles, of whom 35 (49%) had chronic furunculosis. blood) and the Streptococcal Grouping kit (Oxoid, Dardilly, The median age of patients presenting with furuncles was France). Antimicrobial susceptibility was tested with the BD 36Æ9 years (range 4–94). Forty-nine patients presented with Phoenix device (Becton Dickinson Diagnostics, Franklin Lakes, impetigo, of whom 28 (57%) had nonbullous impetigo. NJ, U.S.A.). The median age of patients presenting with impetigo was 20Æ8 years (range 7 days–67 years). Skin cultures grew S. aureus in 64 cases (52%), comprising 31 patients with Genotyping and macrorestriction analysis by pulsed furuncles and 33 patients with impetigo (Table 1 and field gel electrophoresis Fig. 1). Streptococcus pyogenes was isolated in two patients with Isolates were grown overnight on brain–heart infusion agar nonbullous impetigo, in association with S. aureus in both or in brain–heart infusion broth. Genomic DNA was used cases. Skin culture never yielded S. pyogenes alone in a patient as the polymerase chain reaction (PCR) target, after extrac- with impetigo.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 S. aureus skin infections, F. Durupt et al. 1163

Table 1 Number of patients with Staphylococcus aureus isolated from a cultures. All 10 isolates from bullous impetigo swabs skin lesion and ⁄or the nose harboured the eta and ⁄or etb genes, compared with 12 of 21 (57%) nonbullous impetigo isolates. There was no association Isolation of Staphylococcus between the type of ET and patient characteristics or clinical aureus presentation. Number From From skin No significant association was found between impetigo or of skin From lesion furuncles and the other toxins tested. Clinical diagnosis patients lesion nose and nose Furuncles 72 31 27 19 Staphylococcus aureus nasal carriage Simple furuncle 25 7 7 2 Acute multiple furuncles 7 5 2 1 Nasal carriage of S. aureus was found in 37 of 64 (58%) patients Chronic furunculosis 35 17 16 14 with culture-confirmed S. aureus skin infection. The nasal Carbuncle 5 2 2 2 carriage rate differed very significantly (P <0Æ007) between Impetigos 49 33 20 18 patients with simple furuncles (two of seven, 29%) and patients Bullous impetigo 13 10 6 4 with chronic furunculosis (14 of 16, 88%). The nasal carriage Nonbullous impetigo 28 21 13 13 Secondary impetigo 8 2 1 1 rates were 40% (4 ⁄10) and 62% (13 ⁄21) in patients with bullous and nonbullous impetigo, respectively (no significant Total 121 64 47 37 difference). The nasal carriage rate was 46% (6 ⁄13) in patients with PVL-positive furuncles and 72% (13 ⁄18) in patients with PVL-negative furuncles (no significant difference). Based on PFGE analysis and toxin expression, the same Toxin detection in skin culture isolates strain was present in the nares and in the skin lesion(s) of all 37 patients who were culture positive at both sites (Fig. 2). Panton–Valentine leukocidin and furuncles

PVL genes were detected in 13 of 31 (42%) S. aureus furuncle Prevalence of methicillin-resistant Staphylococcus aureus isolates and in 0% of impetigo isolates. Six of the 13 PVL- in skin cultures positive isolates (46%) were associated with epidemic furun- culosis vs. two of the 18 PVL-negative isolates (11%). MRSA was isolated by skin culture in four patients (6%), of Epidemic furunculosis was then significantly associated with whom one had impetigo and three had furuncles (Table 2). PVL-positive strains (P <0Æ04) but there was no link between Three of these patients had been in recent contact with health- PVL-positive isolates and patient characteristics (age, sex, care facilities, and their MRSA isolates belonged to the most underlying conditions) or the clinical presentation (number of prevalent MRSA clone (the Lyon clone) spreading in French furuncles, chronicity). hospitals, with an agr allele type 1, no PVL genes, and the enterotoxin A gene.18 PFGE typing of these strains confirmed that they belonged to the Lyon clone,18 which can be Exfoliative toxins and impetigo acquired in the community by patients with specific risk fac- The eta and ⁄or etb genes, coding for ETA and ETB, respectively, tors.19 One patient had no identified risk factor for MRSA were found in 23 of 33 (70%) isolates from impetigo skin acquisition; this patient’s strain was resistant to beta-lactams,

13 PVL+ strains: 2 simple furuncles 2 acute multiple furuncles 8 chronic furunculoses 1 carbuncle 31 cultures from furuncles

18 PVL- strains: 5 simple furuncles 3 acute multiple furuncles 64 cultures 9 chronic furonculoses with S. aureus 1 carbuncle

23 ET+ strains: 121 skin 10 bullous impetigos cultures 12 nonbullous impetigos 33 cultures 1 secondary impetigo Fig 1. Distribution of patients according to from impetigos the isolation of Staphylococcus aureus in skin cultures, the clinical presentation, and the 10 ET- strains: 57 cultures 9 nonbullous impetigos presence of Panton–Valentine leukocidin without S. aureus 1 secondary impetigo (PVL) or exfoliative toxin (ET) genes.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 1164 S. aureus skin infections, F. Durupt et al. Patient 2 Patient 13 Patient 31 Patient 33 Patient 80 Patient 4 Patient 78 Patient 6 Patient 47 Patient

NCTC 8325 N L N L N L NCTC 8325 N L N L NCTC 8325 N L N L NCTC 8325 N L N L NCTC 8325

Fig 2. Comparison of pulsotypes of Staphylococcus aureus strains isolated from the nose (N) and skin lesions (L). NCTC 8325 is the molecular size rerence strain.

Table 2 Epidemiological, phenotypic and genotypic data on methicillin-resistant Staphylococcus aureus (MRSA) strains isolated from skin lesions

Age (years) ⁄ agrb Patient Skin infection sex Risk factors for MRSA Antibiotic resistancea allele seac PVLd MRSA typef 22 Nonbullous impetigo 50 ⁄F Recent contact with hospital Peni, Oxa, Kana, Tobra, Ofloe 1+) HCA-MRSA 27 Carbuncle 94 ⁄M Age, diabetes mellitus, Peni, Oxa, Kana, Tobra, Tetra, 1+) HCA-MRSA chronic dermatosis, Ery, Linco, Ofloe healthcare worker in family 109 Chronic furunculosis 24 ⁄M Severe atopic dermatitis, Peni, Oxa, Tetra, Ery, Lincoe 1+) HCA-MRSA use of topical corticosteroids, recent hospitalization 42 Chronic furunculosis 21 ⁄F None Peni, Oxa, Kana, Tetra, Fusie 3 ) + CA-MRSA

aAntibiotics to which the strain was resistant. bAccessory gene regulator. cStaphylococcal enterotoxin A. dPanton–Valentine leukocidin. ePeni, penicillin G; Oxa, oxacillin; Kana, kanamycin; Tobra, tobramycin; Tetra, tetracycline; Ery, erythromycin; Linco, lincomycin; Oflo, ofloxacin; Fusi, fusidic acid. fHCA-MRSA, healthcare-associated MRSA; CA-MRSA, community-acquired MRSA. + or –, presence or absence of the gene.

kanamycin, tetracycline and fusidic acid. The strain also har- culosis (88%); the nasal S. aureus isolate was always identical to boured the PVL genes and an agr allele type 3. The characteris- the skin isolate; (v) MRSA was isolated in only four cases tics of this strain, and its PFGE type, were similar to those of (6%), three corresponding to hospital strains and one being a the predominant CA-MRSA clone spreading through Europe.3 true CA-MRSA. Since the 1990s, PVL has been classically associated with 20 Discussion furuncles. In the first hospital-based series, Couppie´ et al. and Lina et al. found that, respectively, 86% and 93% of strains We studied 121 patients presenting with community-onset isolated from furuncles produced PVL.21,22 More recently, in a superficial skin infections and found that: (i) the PVL genes multicentre hospital study, Yamasaki et al. found that only were detected in only 42% of S. aureus furuncle isolates 40% of S. aureus isolates from furuncles harboured PVL (mainly in cases of epidemic furunculosis); (ii) bullous impe- genes.23 In our study, 42% of furuncle isolates were PVL posi- tigo was always associated with ET genes, as well as 57% of tive. The striking differences between authors in the frequency strains causing nonbullous impetigo; (iii) S. pyogenes was iso- of PVL genes detection could reflect differences in the studied lated in only two cases of nonbullous impetigo, both of which populations. PVL detection rates would be higher in the most also grew S. aureus; (iv) S. aureus nasal carriage was frequent severe cases (hospital series), whereas most of our patients (58% of patients with culture-confirmed S. aureus skin infec- presented to community practices. PVL detection rates could tion), but the rate was markedly different between patients also vary according to the area of the studied population (e.g. with a simple furuncle (29%) and those with chronic furun- its prevalence is higher in countries with low economic

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 S. aureus skin infections, F. Durupt et al. 1165 level24,25). Finally, the differences observed between authors study the S. aureus nasal carriage rate was about twice as high could also be related with the uncertainty of the clinical diag- as that usually observed in the general population (58% vs. nosis (e.g. furuncle confounded with deep folliculitis), includ- 30%). The nose and skin isolates always belonged to the same ing in our own study as diagnoses were made by 44 different strain in a given patient. These results strongly suggest that physicians. Unlike Yamasaki et al., who found that PVL-posi- the nasal isolate could play a role in the pathophysiology of tive strains were associated with multiple inflammatory furun- skin infections. The nasal carriage rate was significantly higher cles in young patients free of underlying systemic disorders,23 in patients with chronic furunculosis (88%) than in those we found no association between PVL genes and clinical man- with a simple furuncle, whose nasal carriage rate (29%) was ifestations. These results suggest that simple furuncles are not similar to that of the general population. This further supports always associated with PVL-producing strains of S. aureus.In the use of nasal decontamination in patients with chronic contrast, PVL was associated with epidemic furunculosis, furunculosis.9 Mupirocin is probably the best drug for this which suggests that PVL-positive strains are not only more vir- purpose, although its widespread use could lead to the emer- ulent but also more contagious, as witnessed by major out- gence of mupirocin-resistant S. aureus strains.39,40 In France, breaks of furuncles due to PVL-positive S. aureus in jails, sports fusidic acid is also proposed in this indication although its use teams and semiclosed communities (native Americans, in rural bears a high risk of selecting fusidic acid-resistant strains.41 Alaska, etc.).26–29 The epidemic potential of S. aureus PVL-posi- Interestingly, nasal carriage of PVL-positive strains was infre- tive isolates could be linked to the induction by leukocidin of quent in our patients, as in studies of outbreaks of PVL- the expression of cell wall-anchored proteins such as protein positive S. aureus infections.25,27,42 A, theoretically facilitating the adhesion of S. aureus on human We genetically characterized our MRSA isolates and were epithelia.30 thus able to distinguish between hospital MRSA that had Nonbullous impetigo, one of the most common skin dis- spread to the community (healthcare-associated MRSA or eases, can be caused by S. pyogenes alone, by S. aureus alone, or by HCA-MRSA) and true PVL-positive CA-MRSA. Del Giudice et al. the two bacteria in combination. Streptococcus pyogenes was long prospectively studied community-acquired skin infections and considered to be the main cause of nonbullous impetigo, a con- found MRSA in 11% of cases, 8% of the strains corresponding dition formerly called ‘streptococcal impetigo’.31 In the 1990s, to HCA-MRSA and 3% to PVL-positive CA-MRSA.43 PVL-posi- in American and Israeli studies, the importance of S. aureus rose tive CA-MRSA strains are uncommon in Europe,44 accounting dramatically but S. pyogenes was still isolated in about one-third for fewer than 1% of all MRSA isolates collected in France, for of skin cultures, alone or together with S. aureus.32,33 In 2002, a example. Their prevalence was also low in our study: only large Dutch study of 160 children with impetigo showed that one CA-MRSA was isolated, from a patient with chronic S. pyogenes was present in only 8Æ1% of cases, usually in asso- furunculosis. PVL-positive CA-MRSA strains are an emerging ciation with S. aureus.34 In our study, only two patients with threat, because of their global clonal spread. They have nonbullous impetigo had both S. pyogenes and S. aureus, but no become hyperendemic in some parts of the world, and espe- cases of pure streptococcal impetigo were found. These results cially the U.S.A., where they are isolated from about 60% of confirm that the epidemiology of superficial skin infections is patients with S. aureus skin infections.45,46 The other three evolving rapidly and that treatment of impetigo should now MRSA isolates in our study were recovered from patients target S. aureus more than S. pyogenes. who had been in recent contact with healthcare facilities. The association of bullous impetigo with staphylococcal ETA Community-onset skin infections should be strictly monitored, and ETB has been clearly demonstrated.35 ETs act as serine pro- as the epidemiology of S. aureus infections is constantly teases which cleave desmoglein 1 in the superficial epidermis, evolving. leading to blister formation. Our epidemiological results con- firm these laboratory data, as all our cases of bullous impetigo Acknowledgments were associated with ETA and ⁄or ETB. Little is known about the pathophysiology of nonbullous impetigo. In our study, no par- We thank all the dermatologists who collected data and sam- ticular toxin pattern was associated with this condition, but the ples from patients, Christine Courtier, Franc¸oise Forey, Chris- high frequency of ETA and ⁄or ETB genes in strains recovered tine Gardon and Ce´line Spinelli for their technical assistance, from nonbullous impetigo (57%) suggests that bullous and and David Young for editing the manuscript. nonbullous impetigo probably correspond to the same disease. Nonbullous impetigo possibly corresponds to a different quanti- References tative expression of ETs in the epidermis and further studies should be performed to confirm this hypothesis. 1 Gravet A, Couppie´ P, Meunier O et al. Staphylococcus aureus isolated in Staphylococcus aureus is present as a commensal in the anterior cases of impetigo produces both epidermolysin A or B and LukE– nares of about 30% of healthy people.8 It has been shown that LukD in 78% of 131 retrospective and prospective cases. J Clin Microbiol 2001; 39:4349–56. S. aureus nasal carriage is a risk factor for infections in patients 2 Dufour P, Gillet Y, Bes M et al. Community-acquired methicillin- with continuous peritoneal dialysis and for surgical wound resistant Staphylococcus aureus infections in France: emergence of a 36,37 infections. A higher rate of S. aureus nasal carriage has been single clone that produces Panton–Valentine leukocidin. Clin Infect described in patients with S. aureus skin infections.8,38 In our Dis 2002; 35:819–24.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 1166 S. aureus skin infections, F. Durupt et al.

3 Vandenesch F, Naimi T, Enright MC et al. Community-acquired 23 Yamasaki O, Kaneko J, Morizane S et al. The association between methicillin-resistant Staphylococcus aureus carrying Panton–Valentine Staphylococcus aureus strains carrying Panton–Valentine leukocidin leukocidin genes: worldwide emergence. Emerg Infect Dis 2003; genes and the development of deep-seated follicular infection. Clin 9:978–84. Infect Dis 2005; 40:381–5. 4 Zetola N, Francis JS, Nuermberger EL, Bishai WR. Community- 24 Baba Moussa L, Sanni A, Dagnra AY et al. Approche e´pide´miologi- acquired meticillin-resistant Staphylococcus aureus: an emerging threat. que de l’antibiore´sistance et de la production de leucotoxines par Lancet Infect Dis 2005; 5:275–86. les souches de Staphylococcus aureus isole´es en Afrique de l’Ouest. Med 5 Liassine N, Auckenthaler R, Descombes MC et al. Community- Mal Infect 1999; 29:689–96. acquired methicillin-resistant Staphylococcus aureus isolated in Switzer- 25 Issartel B, Tristan A, Lechevallier S et al. Frequent carriage of land contains the Panton–Valentine leukocidin or exfoliative toxin Panton–Valentine leucocidin genes by Staphylococcus aureus isolates genes. J Clin Microbiol 2004; 42:825–8. from surgically drained abscesses. J Clin Microbiol 2005; 43:3203– 6 Yamaguchi T, Yokota Y, Terajima J et al. Clonal association of Staph- 7. ylococcus aureus causing bullous impetigo and the emergence of new 26 Pan ES, Diep BA, Carleton HA et al. Increasing prevalence of methi- methicillin-resistant clonal groups in Kansai district in Japan. J Infect cillin-resistant Staphylococcus aureus infection in California jails. Clin Dis 2002; 185:1511–16. Infect Dis 2003; 37:1384–8. 7 Tulloch LG. Nasal carriage in staphylococcal skin infections. Br Med 27 Kazakova SV, Hageman JC, Matava M et al. A clone of methicillin- J 1954; ii:912–13. resistant Staphylococcus aureus among professional football players. 8 Wertheim HF, Melles DC, Vos MC et al. The role of nasal carriage N Engl J Med 2005; 352:468–75. in Staphylococcus aureus infections. Lancet Infect Dis 2005; 5:751–62. 28 Stemper ME, Shukla SK, Reed KD. Emergence and spread of com- 9 Raz R, Miron D, Colodner R et al. A 1-year trial of nasal mupirocin munity-associated methicillin-resistant Staphylococcus aureus in rural in the prevention of recurrent staphylococcal nasal colonization Wisconsin, 1989 to 1999. J Clin Microbiol 2004; 42:5673–80. and skin infection. Arch Intern Med 1996; 156:1109–12. 29 Baggett HC, Hennessy TW, Leman R et al. An outbreak of commu- 10 Ausubel FM, Brent R, Kingston RE et al. Current Protocols in Molecular nity-onset methicillin-resistant Staphylococcus aureus skin infections Biology. New York: Greene Publishing Associates and Wiley Inter- in southwestern Alaska. Infect Control Hosp Epidemiol 2003; 24:397– science, 1987. 402. 11 Lina G, Boutite F, Tristan A et al. Bacterial competition for human 30 Labandeira-Rey M, Couzon F, Boisset S et al. Staphylococcus aureus Pan- nasal cavity colonization: role of staphylococcal agr alleles. Appl Envi- ton–Valentine leukocidin causes necrotizing pneumonia. Science ron Microbiol 2003; 69:18–23. 2007; 315:1130–3. 12 Jarraud S, Mougel C, Thioulouse J et al. Relationships between 31 Bisno AL, Stevens DL. Streptococcal infections of skin and soft tis- Staphylococcus aureus genetic background, virulence factors, agr groups sues. N Engl J Med 1996; 334:240–5. (alleles), and human disease. Infect Immun 2002; 70:631–41. 32 Demidovich CW, Wittler RR, Ruff ME et al. Impetigo. Current eti- 13 Murakami K, Minamide W, Wada K et al. Identification of methicil- ology and comparison of penicillin, erythromycin, and cephalexin lin-resistant strains of staphylococci by polymerase chain reaction. therapies. Am J Dis Child 1990; 144:1313–15. J Clin Microbiol 1991; 29:2240–4. 33 Dagan R, Bar-David Y. Double-blind study comparing erythro- 14 De Buyser ML, Morvan A, Grimont F et al. Characterization of Staph- mycin and mupirocin for treatment of impetigo in children: ylococcus species by ribosomal RNA gene restriction patterns. J Gen implications of a high prevalence of erythromycin-resistant Staphylo- Microbiol 1989; 135:989–99. coccus aureus strains. Antimicrob Agents Chemother 1992; 36:287– 15 Blanc DS, Struelens MJ, Deplano A et al. Epidemiological validation 90. of pulsed-field gel electrophoresis patterns for methicillin-resistant 34 Koning S, van Suijlekom-Smit LW, Nouwen JL et al. Fusidic acid Staphylococcus aureus. J Clin Microbiol 2001; 39:3442–5. cream in the treatment of impetigo in general practice: double 16 Goering RV, Winters MA. Rapid method for epidemiological evalu- blind randomised placebo controlled trial. BMJ 2002; 324:203– ation of Gram-positive cocci by field inversion gel electrophoresis. 6. J Clin Microbiol 1992; 30:577–80. 35 Amagai M, Matsuyoshi N, Wang ZH et al. Toxin in bullous impe- 17 Tenover FC, Arbeit RD, Goering RV et al. Interpreting chromosomal tigo and staphylococcal scalded-skin syndrome targets desmoglein DNA restriction patterns produced by pulsed-field gel electrophore- 1. Nat Med 2000; 6:1275–7. sis: criteria for bacterial strain typing. J Clin Microbiol 1995; 36 Wertheim HF, Vos MC, Ott A et al. Risk and outcome of nosoco- 33:2233–9. mial Staphylococcus aureus bacteraemia in nasal carriers versus non- 18 Ferry T, Bes M, Dauwalder O et al. Toxin gene content of the Lyon carriers. Lancet 2004; 364:703–5. methicillin-resistant Staphylococcus aureus clone compared with that of 37 Luzar MA, Coles GA, Faller B et al. Staphylococcus aureus nasal carriage other pandemic clones. J Clin Microbiol 2006; 44:2642–4. and infection in patients on continuous ambulatory peritoneal dial- 19 Lescure FX, Locher G, Eveillard M et al. Community-acquired infec- ysis. N Engl J Med 1990; 322:505–9. tion with healthcare-associated methicillin-resistant Staphylococcus aur- 38 Kluytmans J, van Belkum A, Verbrugh H. Nasal carriage of Staphylo- eus: the role of home nursing care. Infect Control Hosp Epidemiol 2006; coccus aureus: epidemiology, underlying mechanisms, and associated 27:1213–18. risks. Clin Microbiol Rev 1997; 10:505–20. 20 Cribier B, Prevost G, Couppie´ P et al. Staphylococcus aureus leukocidin: 39 Hurdle JG, O’Neill AJ, Mody L et al. In vivo transfer of high-level a new virulence factor in cutaneous infections? An epidemiological mupirocin resistance from Staphylococcus epidermidis to methicillin- and experimental study. Dermatology 1992; 185:175–80. resistant Staphylococcus aureus associated with failure of mupirocin 21 Lina G, Piemont Y, Godail-Gamot F et al. Involvement of Panton– prophylaxis. J Antimicrob Chemother 2005; 56:1166–8. Valentine leukocidin-producing Staphylococcus aureus in primary skin 40 Cookson BD. The emergence of mupirocin resistance: a challenge infections and pneumonia. Clin Infect Dis 1999; 29:1128–32. to infection control and antibiotic prescribing practice. J Antimicrob 22 Couppie´ P, Cribier B, Prevost G. Leukocidin from Staphylococcus aureus Chemother 1998; 41:11–18. and cutaneous infections: an epidemiologic study. Arch Dermatol 41 Howden BP, Grayson ML. Dumb and dumber – the potential 1994; 130:1208–9. waste of a useful antistaphylococcal agent: emerging fusidic

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 S. aureus skin infections, F. Durupt et al. 1167

acid resistance in Staphylococcus aureus. Clin Infect Dis 2006; 42:394– 44 Robert J, Etienne J, Bertrand X. Methicillin-resistant Staphylococcus 400. aureus producing Panton–Valentine leukocidin in a retrospective case 42 Ellis MW, Hospenthal DR, Dooley DP et al. Natural history of com- series from 12 French hospital laboratories, 2000–2003. Clin Micro- munity-acquired methicillin-resistant Staphylococcus aureus colonization biol Infect 2005; 11:585–7. and infection in soldiers. Clin Infect Dis 2004; 39:971–9. 45 Kaplan SL, Hulten KG, Gonzalez BE et al. Three-year surveillance 43 Del Giudice P, Blanc V, Durupt F et al. Emergence of two popula- of community-acquired Staphylococcus aureus infections in children. tions of methicillin-resistant Staphylococcus aureus with distinct epi- Clin Infect Dis 2005; 40:1785–91. demiological, clinical and biological features, isolated from patients 46 Moran GJ, Krishnadasan A, Gorwitz RJ et al. Methicillin-resistant with community-acquired skin infections. Br J Dermatol 2006; S. aureus infections among patients in the emergency department. 154:118–24. N Engl J Med 2006; 355:666–74.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1161–1167 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08195.x Increased nuclear b-catenin in suprabasal involved psoriatic epidermis P.J. Hampton, O.K. Ross and N.J. Reynolds Dermatological Sciences, Institute of Cellular Medicine, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH, U.K.

Summary

Correspondence Background Psoriasis is a common inflammatory skin disease characterized by Philip J. Hampton. abnormal keratinocyte proliferation and differentiation, increased angiogenesis E-mail: [email protected] and inflammation. There is evidence that some keratinocyte differentiation events are controlled by changes in cell–cell adhesion. b-catenin is a 94-kDa protein Accepted for publication 30 April 2007 which has a dual function as a component of intercellular adherens junctions and also as a transcription factor as part of the Wnt signalling pathway. b-catenin Key words is not required for keratinocyte proliferation but has been shown to regulate adherens junctions, glycogen synthase kinase 3b, keratinocyte stem cells and hair follicle morphogenesis. glycogen synthase kinase binding protein, Objectives To investigate the distribution and function of b-catenin in involved pso- keratinocyte, transglutaminase 1 riatic epidermis and in epidermal keratinocytes. Conflicts of interest Methods Biopsies were obtained from patients with psoriasis and from normal con- None declared. trols. The distribution of b-catenin was investigated using antibodies to both total and unphosphorylated active b-catenin. Luciferase assays were used to measure transcriptional activation of transglutaminase 1 (TGase 1) and involucrin and to investigate the functional role of b-catenin in interfollicular keratinocytes. Results Increased nuclear b-catenin was seen in lesional suprabasal psoriatic epider- mis compared with uninvolved or normal skin. Increased active unphosphory- lated b-catenin was also detected within the differentiating compartment of involved psoriatic epidermis. Expression of TGase 1 overlapped with b-catenin in suprabasal lesional psoriasis. The TGase 1 promoter was positively regulated by activated b-catenin and by the glycogen synthase kinase binding protein, suggesting that b-catenin and glycogen synthase kinase 3b may regulate TGase 1 expression. Conclusions This is the first report to convincingly demonstrate increased b-catenin in involved psoriasis and to implicate b-catenin in the regulation of TGase 1. This evidence suggests a role for b-catenin signalling in regulating keratinocyte differ- entiation in interfollicular skin in addition to previously reported functions in stem cell fate determination, hair follicle regulation and skin tumorigenesis.

The differentiation of human skin is a complex process involv- Moreover, the recent discovery that filaggrin mutations may ing the activation and expression of numerous proteins and cause atopic eczema highlights the potential primary role of enzymes in a specific order. Psoriasis is a common inflamma- dysregulated keratinocyte biology in inflammatory skin dis- tory skin disease characterized by abnormal keratinocyte pro- ease.4 In lesional (involved) psoriatic epidermis, proteins liferation and differentiation, angiogenesis, immune activation which are expressed in the early stages of keratinocyte differ- and inflammation. Although there is evidence for immune entiation such as involucrin and transglutaminase 1 (TGase 1, activation with lymphocyte infiltration in psoriatic epidermis1 also known as keratinocyte transglutaminase) are expressed there is also evidence for primary keratinocyte defects. Several earlier and in greater amounts.5,6 In contrast, the expression lines of evidence, including the occurrence of a linear form of of proteins which are usually expressed in the later stages of psoriasis along Blaschko’s lines,2 increased proliferation of keratinocyte differentiation such as filaggrin and loricrin are keratinocytes from uninvolved psoriatic skin and abnormalities downregulated in the epidermis of psoriatic plaques.7 There is of calcium capacitance signalling in cultured psoriatic keratin- accumulating evidence that proteins which mediate intercellu- ocytes,3 all point to a primary defect in psoriatic keratinocytes. lar adhesion are involved in differentiation.8–10 One candidate

2007 The Authors 1168 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. 1169 for this is b-catenin, a 94-kDa protein which participates in We demonstrate that GSK-3b regulates TGase 1 transcriptional intercellular adhesion as part of the adherens junctions. In activation and that activated b-catenin activates a TGase 1 adherens junctions, b-catenin links E-cadherin to the actin luciferase reporter. cytoskeleton via a-catenin.11 The binding of b-catenin to adherens junction components is regulated by phosphorylation Materials and methods at specific tyrosine residues by protein tyrosine kinases.12 Increased tyrosine phosphorylation of b-catenin releases it Cell culture from the adherens junction into the cytoplasm. The fate of cytoplasmic b-catenin is determined by the phosphorylation Normal human keratinocytes were cultured from redundant state at N-terminal serine residues. b-catenin phosphorylated neonatal foreskin samples as previously described and, follow- at these sites is degraded whereas unphosphorylated, transcrip- ing informed consent and ethical approval, were used for tionally active b-catenin is translocated to the nucleus.13 Once all in vitro experiments.24 Keratinocytes were maintained in ) there, it combines with members of the Tcf ⁄Lef family of low-calcium (70 lmol L 1) MCDB medium supplemented ) transcription factors to direct gene transcription.14 with transferrin (5 lgmL 1), epidermal growth factor ) ) The largest reservoir of b-catenin is at cell membranes (10 ng mL 1), insulin (5 lgmL 1) and hydrocortisone ) where it forms part of the adherens junctions. The signalling (0Æ18 lgmL 1). role of b-catenin is intimately related to adhesion events. Overexpression of basal layer cadherins in the suprabasal epi- Immunohistochemistry dermis of mice leads to increased cytoplasmic active b-catenin and increased b-catenin transcriptional activation.15 This over- Following ethical approval from The Newcastle and North expression of basal layer cadherins resulted in abnormal differ- Tyneside Health Authority joint ethics committee and entiation.15 It has been demonstrated that b-catenin is not informed consent, skin biopsies were obtained under local required for keratinocyte proliferation.16 The role of b-catenin anaesthetic from adult patients with chronic plaque psoriasis in regulating keratinocyte stem cell differentiation and hair and from healthy volunteers. Punch biopsies (6 mm) were follicle morphogenesis has been extensively reported.17 There taken from non-sun-exposed involved and uninvolved psoriat- are fewer data on the role of b-catenin in differentiation of ic skin (usually lower back or buttock). Uninvolved skin was interfollicular skin. Doglioni et al. studied b-catenin in human skin at least 5 cm from any psoriatic plaques. Patients with skin tumours and commented that b-catenin in a nucleocyto- psoriasis were excluded from the study if they had received plasmic location was occasionally observed in basal cells, in systemic treatment during the last 4 weeks or had used topical scattered differentiated keratinocytes of the upper layers of the treatment apart from emollients during the last 2 weeks. Biop- epidermis and in luminal cells of sebaceous glands,18 although sies from six patients with psoriasis and six normal controls these data were not shown in the figures. were studied for the b-catenin studies and a further six Once released from the cell membrane, b-catenin signalling patients with psoriasis and six normal controls for the TGase 1 activity is regulated by glycogen synthase kinase-3b (GSK-3b), studies. Following informed consent, sections from a piloma- which plays a key role in regulating levels of cytoplasmic trixoma tumour which had been formalin fixed and paraffin b-catenin.19 GSK-3b is a constitutively active kinase found in embedded were used as a positive control for the antibody all tissues and has numerous substrates including metabolic staining. proteins, signalling proteins and structural proteins as well as We used two different antibodies against b-catenin. To transcription factors.20 Constitutive GSK-3b activity is required detect nuclear b-catenin we followed a previously published to phosphorylate free b-catenin and to prevent unphosphory- protocol.25 Punch biopsies (6 mm) were fixed in formalin lated, active b-catenin from translocating to the nucleus.21 and set in paraffin and 5-lm sections were cut. The sections Once phosphorylated, b-catenin is ubiquitinated and targeted were dewaxed using Histoclear (National Diagnostics, Atlanta, to the proteosome for degradation. GSK-3b can be inhibi- GA, U.S.A.) and rehydrated with ethanol and water. Antigen ) ted by N-terminal phosphorylation at serine 9. The func- retrieval was performed by pressure cooking in 10 mmol L 1 tion of GSK-3b can also be inhibited by the endogenous citrate buffer for 1 min. The first antibody was directed protein FRAT1 ⁄glycogen synthase kinase binding protein against total b-catenin (BD Transduction Labs, San Jose, CA, (GSKBP).22,23 This then allows b-catenin to accumulate in the U.S.A.). Sections of hair follicle and pilomatrixoma tumour cytosol and subsequently to translocate to the nucleus. were used as positive controls to confirm that nuclear This study was performed to determine the localization of b-catenin was reliably detected. The second antibody used was b-catenin in psoriasis and to investigate the possible function raised against the unphosphorylated Ser 37 residue, the GSK- of b-catenin in interfollicular skin. We show that b-catenin 3b binding site of b-catenin (antiactive b-catenin monoclonal localizes to the nuclei of suprabasal keratinocytes in psoriatic antibody or anti-ABC; Upstate Bio-technology, Lake Placid, epidermis. We then provide evidence which suggests that NY, U.S.A.). After antigen retrieval, background signal was b-catenin may have a role in differentiation events in inter- blocked with 2% normal goat serum. Sections were incubated follicular keratinocytes. The distribution of nuclear b-catenin with primary antibody for 1 h followed by incubation with in the suprabasal layers was very similar to that of TGase 1. an Oregon green conjugated anti-mouse secondary antibody

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 1170 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al.

(1 : 200; Molecular Probes, Portland, OR, U.S.A.). The nuclei per well using Lipofectamine Plus (Invitrogen, Carlsbad, CA, were counterstained with Toto-3 (1 : 7000; Molecular U.S.A.) according to the manufacturer’s instructions. The trans- Probes). fection reagents were incubated with the cells for 3 h before Immunostaining for TGase 1 and involucrin was performed being replaced with normal medium. Treatments were for 24 h on frozen sections. Six-millimetre punch biopsies were frozen unless stated otherwise. Cell lysis and luciferase assays were in optimal cutting compound and stored at )80 C. A goat performed using Promega Dual Luciferase reagents (Promega) polyclonal anti-TGase 1 antibody (1 : 200; Santa Cruz Bio- and luminescence was measured with a Wallac 96-well plate technology, Santa Cruz, CA, U.S.A.) and a mouse monoclonal luminometer (Perkin Elmer Wallac, Boston, MA, U.S.A.). The anti-involucrin antibody (1 : 200; Novocastra, Newcastle luciferase data presented are combined results of at least upon Tyne, U.K.) were used. Blocking of background signal, three independent experiments. Within each experiment every secondary antibodies, nuclear staining and imaging were the treatment was done in triplicate. same as for the b-catenin antibodies. Images were acquired on a · 63 oil immersion lens using a Statistical evaluation Leica TCS SP II confocal laser scanning microscope (Leica, Mil- ton Keynes, U.K.). Oregon green 488 was excited using the Data are shown as mean ± SEM. Data were analysed using 488-nm line of the argon ion laser and Toto-3 was excited either unpaired t-tests or ANOVA using a generalized linear using the 633-nm helium neon laser. Fluorescence emission model to detect differences at a level of significance of was monitored line by line, using the sequential scanning P <0Æ05, with two-tailed tests. Minitab software was used. mode to minimize any possible crosstalk, and a Z-series of images were captured. The overlay cytofluorograms in Results Figures 1 and 2 were created using the Leica 2D cytofluoro- gram function with pixel intensity thresholds set at 125 for Increased nuclear b-catenin within the suprabasal layers both Oregon green (488) and Toto-3 (633) channels for of involved psoriatic epidermis Figure 2 and at 75 for Figure 1. All images shown are mid z stack except in Figure 2 where maximum projections are To test whether b-catenin signalling was dysregulated in pso- shown to improve image clarity. The images were sized and riasis, sections of involved and uninvolved psoriatic skin and text added using Adobe PhotoshopTM. normal skin were immunostained with an antibody recogniz- ing total b-catenin including unphosphorylated, active, b-cate- nin. Membrane-bound total b-catenin was observed at all Cell counting levels of involved and uninvolved psoriatic skin consistent Biopsies were taken from normal skin from volunteers with its role as a structural component of adherens junctions (n = 6) and from involved and uninvolved skin of six separate (Fig. 1a). In normal and uninvolved psoriatic interfollicular patients with psoriasis (n = 6), all of which were immuno- skin, very little nuclear b-catenin was seen. In contrast, exten- stained with the antibody raised against total b-catenin (BD sive nuclear b-catenin was detected in the suprabasal layers of Transduction Labs). The numbers of b-catenin-positive nuclei involved psoriatic skin just below the stratum corneum and total nuclei in the upper half of the suprabasal region (Fig. 1a,c). In addition to the increased nuclear b-catenin in were counted in four individual fields from each biopsy speci- suprabasal involved psoriasis, there was also a reduction in the men. The size of the counted areas was determined using LCS amount of membranous b-catenin in the same region. When Leica image analysis software (Leica). expressed as a percentage of total nuclei there was a signifi- cantly greater number of nuclear b-catenin-positive cells in the upper suprabasal layers of involved epidermis compared Luciferase assays with similar regions of uninvolved epidermis or normal skin The TGase 1 luciferase reporter containing the full-length (Fig. 1b). human TGase 1 promoter in the poLUC vector was a kind gift of R.H. Rice (UC Davis, CA, U.S.A.).26 The pRL-TK Renilla Increased nuclear b-catenin in involved suprabasal skin luciferase (Promega, Madison, WI, U.S.A.) was used as an detected with an antibody specific for unphosphorylated internal control for transfection efficiency. Wild-type (WT) active b-catenin Xenopus GSK-3b, dominant negative GSK-3b and GSKBP were a kind gift of D. Kimelman (Seattle, WA, U.S.A.).27 The WT An antibody which recognized active unphosphorylated b-cate- and constitutively active (S37A) b-catenin vectors in pcDNA3 nin demonstrated the presence of nuclear activated b-catenin in were kind gifts of R. Herrell (Georgetown, Washington DC, involved suprabasal psoriatic epidermis (Fig. 2), consistent with U.S.A.).21 The S37A mutation prevents phosphorylation by the findings using the anti-total-b-catenin antibody. Nuclear GSK-3b at that site. membrane and extensive cytoplasmic staining for activated Approximately 50 000 keratinocytes were seeded into each b-catenin was also seen in the upper epidermis. In uninvolved well of a 12-well plate. One microgram of reporter vector, or normal skin, unphosphorylated, active b-catenin was seen in 0Æ05 lg of pRL-TK (Renilla) and 1 lg of vector were transfected a small number of nuclei. There was prominent staining of the

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. 1171

(a)

(b) (c)

Fig 1. Nuclear b-catenin is found in uninvolved and involved suprabasal psoriatic epidermis. (a) Following antigen retrieval, paraffin-fixed sections of involved and uninvolved psoriatic skin and normal skin were immunostained with an antibody recognizing both phosphorylated and unphosphorylated b-catenin. Clear increased nuclear b-catenin was seen in the suprabasal layers of involved psoriatic epidermis in six of six samples examined. Extracellular membranous b-catenin was seen in all samples but was reduced in upper suprabasal involved psoriasis where nuclear b-catenin was seen. In the upper panels the position of the epidermal surface is indicated by a dotted line. The scale bar indicates 40 lm. (b) Nuclei stained positively for b-catenin were quantified in suprabasal epidermis and expressed as a percentage of total nuclei. There were significantly more nuclear b-catenin-positive cells in involved psoriasis than in uninvolved or normal skin when expressed as a percentage of total nuclei (P <0Æ005). Data are expressed as mean ± SEM (n = 6). Statistical analysis was by t-test. (c) An enlargement, as indicated, from the top left panel showing involved psoriatic epidermis stained with an antibody against total b-catenin. Mid z images are shown. The scale bar indicates 40 lm.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 1172 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al.

(a) (b) (c) (d)

(e) (f) (g) (h)

Fig 2. Unphosphorylated active b-catenin is found in the nuclei of involved suprabasal psoriatic epidermis. Following an antigen retrieval method, sections of involved and uninvolved psoriatic skin and normal skin were immunostained with an antibody recognizing unphosphorylated active b-catenin (a) and (f). Maximum projections of confocal microscope images are shown. The DNA was labelled by Toto-3 to show the nuclei (b) and (f). The merge images show the colocalization of activated b-catenin and the nuclei (c, g). The 2D cytofluorogram overlay images show only the areas of colocalization of Toto-3 and b-catenin (d, h). Nuclear unphosphorylated active b-catenin was seen in the suprabasal layers of involved psoriatic epidermis just below the stratum corneum (a). Some staining of the nuclear membrane was also seen. There was prominent staining of intercellular activated b-catenin but this was less than with the antibody against total b-catenin shown in Figure 1. In uninvolved psoriatic epidermis occasional nuclear unphosphorylated b-catenin was seen in the uppermost layers of suprabasal epidermis just below the stratum corneum (e). Normal skin showed similar changes to uninvolved skin (data not shown).

intercellular junctions for active, unphosphorylated b-catenin in expression of both these proteins in involved psoriatic epider- the upper layers of the suprabasal epidermis in involved epider- mis was increased in the upper suprabasal layers just below mis which was much greater than in the uninvolved or normal the stratum corneum (not shown). These data suggested the skin. The presence of activated nuclear b-catenin in involved su- possibility of an interaction between b-catenin and TGase 1. prabasal psoriatic epidermis raised the possibility that b-catenin may have a functional role in directing gene expression at this Transglutaminase 1 promoter activity is increased level of the epidermis where numerous differentiation events by inhibition of glycogen synthase kinase 3b direct formation of the cornified envelope. or by overexpression of active b-catenin

GSK-3b decreases the transcriptional activity of b-catenin by Transglutaminase 1 is expressed in the same region phosphorylating free, cytoplasmic b-catenin. We first con- of suprabasal epidermis as b-catenin firmed that b-catenin signalling was active in our cultured The presence of nuclear b-catenin in involved suprabasal epi- keratinocytes and was regulated by GSK-3b. For this, we used dermis led us to ask whether or not b-catenin was directing the b-catenin-sensitive Topflash luciferase reporter. Stimulation expression of genes involved in terminal differentiation. with the differentiation stimulus 12-O-tetradecanoyl-phorbol- Immunostaining of involved psoriasis with an anti-TGase 1 13-acetate (TPA) activated Topflash as did inhibition of GSK- antibody showed staining confined to the suprabasal layer at a 3b by overexpression of GSKBP (data not shown). Inhibition similar level to where nuclear b-catenin was seen (Fig. 3a), as of endogenous GSK-3b by GSKBP induced a 3-fold activation previously described.5 Expression of TGase 1 in uninvolved of a TGase 1 luciferase reporter (P <0Æ005; Fig. 4a). Our skin was confined to the stratum corneum (Fig. 3a). Immuno- immunohistochemistry data indicated that b-catenin localized staining with an anti-involucrin antibody showed a broader to the nucleus in suprabasal regions where TGase 1 was also distribution of involucrin protein expression in the epidermis, found. We therefore asked if b-catenin could mediate the as previously reported28 (Fig. 3b). Light microscopy (trans- transcriptional activation of TGase 1. In normal human kerati- mission) images were obtained for sections stained for both nocytes we coexpressed either the TGase 1 luciferase reporter total b-catenin and TGase 1. These images showed that the along with either WT b-catenin or an activated b-catenin

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. 1173

(a) Anti-transglutaminase 1 Toto-3 Merge Uninvolved psoriasis Uninvolved

(b) Anti-involucrin Toto-3 Merge Uninvolved psoriasisUninvolved psoriasis Involved Involved psoriasis Involved

Fig 3. Expression of the early differentiation markers transglutaminase 1 (TGase 1) and involucrin is increased in involved psoriatic skin. (a) The increased expression of TGase 1 is very similar to the distribution of nuclear b-catenin in involved psoriasis. (b) The band of increased expression of involucrin in involved psoriasis is wider than that of TGase 1, with involucrin-positive cells seen just above the basal layer in some areas. Mid z images shown. The scale bars indicate 40 lm.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 1174 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al.

The role of b-catenin has been extensively studied with (a) respect to epidermal stem cell and hair follicle biology.29 Also b-catenin has been shown, in vitro, to regulate transforming growth factor (TGF)-b-induced fibroblast proliferation but to have no effect on keratinocyte proliferation.16 Little is known about the expression of b-catenin in normal or psoriatic inter- follicular epidermis or about the role of b-catenin in differen- tiation of interfollicular epidermis. In this study, we observed strong nuclear localization of b-catenin in suprabasal involved psoriasis as previously sug- gested but incompletely reported by Tsuji et al.30 We also observed a marked decrease in membranous b-catenin and increased cytoplasmic b-catenin at the same level as the nuclear b-catenin. In the uninvolved psoriasis samples there (b) was some nuclear b-catenin but this was much less than in involved psoriasis. There was even less nuclear b-catenin in normal skin. In the paper by Tsuji et al. nuclear b-catenin was seen in the upper spinous to granular layers. In the image shown of normal epidermis there does appear to be more nuclear b-catenin than seen in our study. They used an anti- gen retrieval technique and the same antibody as used by us. Reasons for the difference are not clear; however, in their paper no quantification is provided, introducing the possibility of sampling bias. In this study we did cell counts from four different fields for each of six samples, giving 24 datasets in total. Stojadinovic et al., using an antigen retrieval technique and the same anti b-catenin antibody, did not see any nuclear Fig 4. Activation of the transglutaminase 1 (TGase 1) reporter is b-catenin in normal epidermis and showed images similar to 31 regulated by glycogen synthase kinase-3b (GSK-3b) and b-catenin. those in this paper. Munne et al. have shown that an antigen (a) Coexpression of the keratinocyte TGase 1 reporter with the retrieval technique is needed to detect nuclear b-catenin glycogen synthase kinase binding protein (GSKBP) demonstrated reliably.25 The results from papers which have not done this significant activation of the TGase 1 promoter (P <0Æ005). must be regarded as unreliable. Overexpression of a dominant negative GSK-3b also induced a b-catenin has a dual function; in cell adhesion as a com- significant increase in activation of the TGase 1 promoter (data not ponent of adherens junctions and also as a transcription factor. shown). Data are shown as mean ± SEM with n = triplicate data Movement of b-catenin between the cell membrane and the from three separate experiments. Statistical analysis was by t-test. cytoplasm is dependent on phosphorylation of tyrosine resi- (b) b-catenin activated the TGase 1 promoter. Cotransfection of the dues which control binding to actin and E-cadherin. Our find- activated b-catenin mutant, S37A, with a TGase 1 luciferase reporter induced activation of the reporter (P =0Æ031). Data are shown as ings suggest that in involved suprabasal psoriatic epidermis an mean ± SEM with n = triplicate data from four separate experiments. excess amount of free unphosphorylated b-catenin enters the Statistical analysis was by ANOVA with subset analysis. WT, wild-type. cytoplasm and subsequently translocates to the nucleus. In normal skin and uninvolved psoriatic epidermis any transcrip- tionally active b-catenin released into the cytoplasm would mutant, S37A. We found that WT b-catenin induced a modest most likely be rapidly targeted for ubiquitination to prevent increase in TGase 1 transcriptional activity which was not abnormal signalling events. This is consistent with our results significant. In contrast, a significant 2-fold increase was ob- showing little nuclear b-catenin in uninvolved or normal served in cells transfected with the S37A b-catenin mutant suprabasal epidermis. In involved psoriatic epidermis, (P =0Æ031; Fig. 4b). These data suggest that TGase 1 may be b-catenin released from the intercellular junctions may not be a transcriptional target of the Wnt ⁄b-catenin signalling path- targeted for ubiquitination in the usual way, leading to way. Interestingly, there was no significant activation of an increased nuclear b-catenin, which may contribute to the involucrin luciferase reporter by WT or activated b-catenin abnormal differentiation phenotype observed. mutant S37A (data not shown). The two anti-b-catenin antibodies produced slightly differ- ent patterns of immunostaining. The antibody against total Discussion b-catenin used here has been widely used for immunostain- ing of tissue sections whereas the anti-activated b-catenin In psoriasis, epidermal differentiation is markedly impaired antibody has only been reported for immunostaining of cells. but the molecular basis for this disruption is not understood. To confirm that both antibodies were working correctly we

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. 1175 immunostained sections from pilomatrixoma tumours and and significantly extend this observation through quantifica- showed prominent nuclear b-catenin consistent with previous tion and functional analysis. reports (data not shown).32 The anti-total b-catenin antibody An association between b-catenin and hyperproliferative (Fig. 1) stained the intercellullar junctions strongly through- skin has been noted before in mice overexpressing b-catenin out most of the epidermis but this intercellular staining in the skin. These studies noted epidermal thickening but became much less prominent in the upper suprabasal layers did not investigate the distribution of nuclear b-catenin in at the level where nuclear b-catenin was seen. The antiactive the interfollicular skin or whether differentiation was b-catenin antibody also stained intercellular junctions but to disordered.34,35 These mice had generalized thickening of the a lesser extent than the anti-total b-catenin. In the upper skin, paws, eyes, lips and eyelids. Gat et al. studied mice suprabasal layers the anti-activated b-catenin antibody showed expressing a truncated b-catenin in skin and also noted a lot of cytoplasmic staining and also nuclear staining. There marked thickening of paws and ears but did not investigate was definite nuclear b-catenin detected by the anti-activated the localization of nuclear b-catenin in interfollicular epider- b-catenin antibody although this was less prominent than mis.34 A role for b-catenin in epidermal proliferation seems with the anti-total b-catenin antibody. The reasons for this unlikely as several papers have reported no role for b-catenin are not clear but technical differences are possible as this is in keratinocyte proliferation. Cheon et al. studied the effect of the first report to use the anti-activated b-catenin antibody b-catenin on wound size.16 They showed that b-catenin defi- on tissue sections. ciency had no effect on keratinocyte proliferation, assessed by Expression of the early differentiation markers involucrin 5-bromodeoxyuridine incorporation. On the other hand, and TGase 1 is increased in involved psoriatic skin and has b-catenin was required for TGF-b-dependent fibroblast prolif- been studied previously.5,33 In Figure 3 we confirmed that eration. Stojadinovic et al. reported increased epidermal nuclear expression of both proteins is increased in involved psoriatic b-catenin in wound edges and inhibition of keratinocyte skin compared with uninvolved and that involucrin has a migration by b-catenin using scratch migration assays.31 They wider pattern of expression than TGase 1 which is confined to suggested that activated nuclear b-catenin was a marker of the upper suprabasal layers. TGase 1 is a 92-kDa protein impaired healing. These reports are consistent with our data which is involved in the cross-linking of proteins and the for- showing that b-catenin promotes differentiation. In psoriasis mation of the cornified envelope during keratinocyte differen- the increased expression of suprabasal b-catenin may be a tiation. In normal epidermis, TGase 1 is found in the stratum compensatory response to the proliferative signals or to the corneum. The molecular mechanism leading to increased incomplete differentiation programme. Whether the increased TGase 1 expression in involved psoriasis has not been previ- suprabasal nuclear b-catenin observed in psoriasis also occurs ously determined. It has been hypothesized that the increase in other hyperproliferative skin diseases such as eczema or ich- in TGase 1 in psoriasis is simply a function of the increased thyosis or reflects abnormal epidermal differentiation remains turnover of the epidermis.7 Our studies revealed that in to be seen. involved psoriasis TGase 1 is expressed in a similar location to In a resting cell, b-catenin is phosphorylated by GSK-3b. nuclear b-catenin. Furthermore, in cultured keratinocytes, GSK-3b activity is inhibited by various stimuli including stim- overexpression of an active b-catenin mutant resulted in ulation of cell-surface receptors by wnt ligands.19 Decreased increased transactivation of the TGase 1 promoter. The modest GSK-3b activity allows increased active unphosphorylated increases in Tgase 1 promoter activity upon expression of con- b-catenin to accumulate and translocate to the nucleus where stitutively active b-catenin were very reproducible, and statis- it binds to elements of the Tcf ⁄Lef family to activate transcrip- tically significant. This raises the possibility that b-catenin may tion of target genes. GSK-3b is also inhibited by endogenous be responsible for the increased expression of TGase 1 in pso- proteins such as GSKBP. GSKBP was first described in Xenopus riasis. To our knowledge, this is the first demonstration of an and the mammalian homologue is called FRAT1.36–38 We effect of b-catenin on TGase 1 promoter activity. These exper- found that inhibition of GSK-3b using GSKBP activated a iments were conducted using proliferating keratinocytes. As TGase 1 reporter identifying a novel regulatory pathway. An b-catenin may be promoting differentiation in keratinocytes, increase in TGase 1 reporter activity was also seen with over- it would be interesting to repeat the luciferase experiments expression of a dominant negative GSK-3b (data not shown). with coadministration of TPA or increased calcium concentra- Having shown that inhibition of GSK-3b activated a TGase tion to see if larger responses were elicited. Also, further work 1 luciferase reporter, we investigated the effect of b-catenin is required to determine if the transcriptional activation is a on transglutaminase transcriptional activity. We showed that result of direct binding of b-catenin to the TGase 1 promoter activated b-catenin significantly activated a TGase 1 luciferase or if it is mediated by other b-catenin targets. reporter (Fig. 4). There has been one previous publication in which b-catenin In conclusion, we have demonstrated an increase in nuclear in psoriasis has been investigated. Increased nuclear b-catenin b-catenin in interfollicular involved psoriatic human skin and was reported in the upper spinous layers of involved psoriai- have provided evidence to suggest a role for nuclear b-catenin sis30 although no images of psoriatic epidermis were pre- in the positive regulation of TGase 1 in interfollicular kerati- sented and the primary focus of the paper was on normal skin nocytes. These data add to the reported functions for b-cate- and hair follicles. The data presented in this report confirm nin which include stem cell lineage determination, hair

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 1176 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. growth and hair follicle cycling17 and are the first attempt to 16 Cheon SS, Wei Q, Gurung A et al. Beta-catenin regulates wound explain the molecular mechanism underlying increased TGase size and mediates the effect of TGF-beta in cutaneous healing. 1 expression in involved psoriasis. FASEB J 2006; 20:692–701. 17 Huelsken J, Vogel R, Erdmann B et al. Beta-catenin controls hair follicle morphogenesis and stem cell differentiation in the skin. Acknowledgments Cell 2001; 105:533–45. 18 Doglioni C, Piccinin S, Demontis S et al. Alterations of beta-catenin This work was funded by a Wellcome Trust Research Training pathway in non-melanoma skin tumors: loss of alpha-ABC nuclear Fellowship awarded to P.J.H. and a Wellcome Trust Research reactivity correlates with the presence of beta-catenin gene muta- Leave Fellowship to N.J.R. O.K.R. was supported by The Psori- tion. Am J Pathol 2003; 163:2277–87. asis Association. We thank Prof. John Matthews for statistical 19 Kim L, Kimmel AR. GSK3, a master switch regulating cell-fate specification and tumorigenesis. Curr Opin Genet Dev 2000; 10:508– advice and Carole Todd for technical assistance. The b-catenin 14. vectors were a kind gift from R. Herrell (Georgetown, Wash- 20 Jope RS, Johnson GV. The glamour and gloom of glycogen 26,39 ington DC, U.S.A.). The GSK-3b vectors were a gift from synthase kinase-3. Trends Biochem Sci 2004; 29:95–102. 40 Dr B. Eldar-Finkelman and the transglutaminase reporter was 21 Orford K, Crockett C, Jensen JP et al. Serine phosphorylation-regu- a kind gift from Dr R.H. Rice (UC Davis, CA, U.S.A.).26 lated ubiquitination and degradation of beta-catenin. J Biol Chem 1997; 272:24735–8. 22 Franca-Koh J, Yeo M, Fraser E et al. The regulation of glycogen References synthase kinase-3 nuclear export by Frat ⁄GBP. J Biol Chem 2002; 277:43844–8. 1 Lebwohl M. Psoriasis. Lancet 2003; 361:1197–204. 23 Farr GH III, Ferkey DM, Yost C et al. Interaction among GSK-3, 2 Ghorpade A. Linear naevoid psoriasis along lines of Blaschko. J Eur GBP, axin, and APC in Xenopus axis specification. J Cell Biol 2000; Acad Dermatol Venereol 2004; 18:726–7. 148:691–702. 3 Karvonen SL, Korkiamaki T, Yla-Outinen H et al. Psoriasis and 24 Todd C, Reynolds NJ. Up-regulation of p21WAF1 by phorbol ester altered calcium metabolism: downregulated capacitative calcium and calcium in human keratinocytes through a protein kinase influx and defective calcium-mediated cell signaling in cultured C-dependent pathway. Am J Pathol 1998; 153:39–45. psoriatic keratinocytes. J Invest Dermatol 2000; 114:693–700. 25 Munne A, Fabre M, Marinoso ML et al. Nuclear beta-catenin in 4 Palmer CN, Irvine AD, Terron-Kwiatkowski A et al. Common loss- colorectal tumors: to freeze or not to freeze? Colon Cancer Team of-function variants of the epidermal barrier protein filaggrin are a at IMAS. J Histochem Cytochem 1999; 47:1089–94. major predisposing factor for atopic dermatitis. Nat Genet 2006; 26 Mariniello L, Qin Q, Jessen BA, Rice RH. Keratinocyte transgluta- 38:441–6. minase promoter analysis. Identification of a functional response 5 Schroeder WT, Thacher SM, Stewart-Galetka S et al. Type I keratino- element. J Biol Chem 1995; 270:31358–63. cyte transglutaminase: expression in human skin and psoriasis. 27 Pierce SB, Kimelman D. Regulation of Spemann organizer forma- J Invest Dermatol 1992; 99:27–34. tion by the intracellular kinase Xgsk-3. Development 1995; 121:755– 6 Eckert RL, Sturniolo MT, Broome AM et al. Transglutaminase func- 65. tion in epidermis. J Invest Dermatol 2005; 124:481–92. 28 Gerritsen MJ, Elbers ME, de Jong EM et al. Recruitment of cycling 7 Iizuka H, Takahashi H, Honma M et al. Unique keratinization pro- epidermal cells and expression of filaggrin, involucrin and tenascin cess in psoriasis: late differentiation markers are abolished because in the margin of the active psoriatic plaque, in the uninvolved skin of the premature cell death. J Dermatol 2004; 31:271–6. of psoriatic patients and in the normal healthy skin. J Dermatol Sci 8 Zhu AJ, Watt FM. Expression of a dominant negative cadherin 1997; 14:179–88. mutant inhibits proliferation and stimulates terminal differentiation 29 Alonso L, Fuchs E. Stem cells in the skin: waste not, Wnt not. Genes of human epidermal keratinocytes. J Cell Sci 1996; 109:3013–23. Dev 2003; 17:1189–200. 9 Nelson WJ, Nusse R. Convergence of Wnt, beta-catenin, and cad- 30 Tsuji H, Ishida-Yamamoto A, Takahashi H et al. Nuclear localization herin pathways. Science 2004; 303:1483–7. of beta-catenin in the hair matrix cells and differentiated keratino- 10 Hines MD, Jin HC, Wheelock MJ et al. Inhibition of cadherin func- cytes. J Dermatol Sci 2001; 27:170–7. tion differentially affects markers of terminal differentiation in cul- 31 Stojadinovic O, Brem H, Vouthounis C et al. Molecular pathogene- tured human keratinocytes. J Cell Sci 1999; 112:4569–79. sis of chronic wounds: the role of beta-catenin and c-myc in the 11 Yap AS, Brieher WM, Gumbiner BM. Molecular and functional inhibition of epithelialization and wound healing. Am J Pathol 2005; analysis of cadherin-based adherens junctions. Annu Rev Cell Dev Biol 167:59–69. 1997; 13:119–46. 32 Chan EF, Gat U, McNiff JM et al. A common human skin tumour is 12 Miravet S, Piedra J, Castano J et al. Tyrosine phosphorylation of caused by activating mutations in beta-catenin. Nat Genet 1999; plakoglobin causes contrary effects on its association with desmo- 21:410–13. somes and adherens junction components and modulates beta- 33 Watanabe S, Wagatsuma K, Ichikawa E et al. Abnormal distribution catenin-mediated transcription. Mol Cell Biol 2003; 23:7391–402. of epidermal protein antigens in psoriatic epidermis. J Dermatol 13 Reya T, Clevers H. Wnt signalling in stem cells and cancer. Nature 1991; 18:143–51. 2005; 434:843–50. 34 Gat U, DasGupta R, Degenstein L et al. De novo hair follicle morpho- 14 Hu P, O’Keefe EJ, Rubenstein DS. Tyrosine phosphorylation of genesis and hair tumors in mice expressing a truncated beta- human keratinocyte beta-catenin and plakoglobin reversibly regu- catenin in skin. Cell 1998; 95:605–14. lates their binding to E-cadherin and alpha-catenin. J Invest Dermatol 35 Gounari F, Signoretti S, Bronson R et al. Stabilization of beta- 2001; 117:1059–67. catenin induces lesions reminiscent of prostatic intraepithelial 15 Hardman MJ, Liu K, Avilion AA et al. Desmosomal cadherin mis- neoplasia, but terminal squamous transdifferentiation of other expression alters beta-catenin stability and epidermal differentia- secretory epithelia. Oncogene 2002; 21:4099–107. tion. Mol Cell Biol 2005; 25:969–78.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 Increased nuclear b-catenin in psoriatic epidermis, P.J. Hampton et al. 1177

36 Freemantle SJ, Portland HB, Ewings K et al. Characterization and tis- 39 Orford K, Orford CC, Byers SW. Exogenous expression of beta- sue-specific expression of human GSK-3-binding proteins FRAT1 catenin regulates contact inhibition, anchorage-independent growth, and FRAT2. Gene 2002; 291:17–27. anoikis, and radiation-induced cell cycle arrest. J Cell Biol 1999; 37 Ferkey DM, Kimelman D. Glycogen synthase kinase-3 beta muta- 146:855–68. genesis identifies a common binding domain for GBP and axin. 40 Eldar-Finkelman H, Argast GM, Foord O et al. Expression and char- J Biol Chem 2002; 277:16147–52. acterization of glycogen synthase kinase-3 mutants and their effect 38 Li L, Yuan H, Weaver CD et al. Axin and Frat1 interact with on glycogen synthase activity in intact cells. Proc Natl Acad Sci USA Dvl and GSK, bridging Dvl to GSK in Wnt-mediated regulation of 1996; 93:10228–33. LEF-1. EMBO J 1999; 18:4233–40.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1168–1177 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08193.x Cutaneous Malassezia flora in atopic dermatitis differs between adults and children Y. Takahata,* T. Sugita, H. Kato,§ A. Nishikawa,§ M. Hiruma and M. Muto* *Department of Dermatology, Yamaguchi University Graduate School of Medicine, Ube, Yamaguchi, Japan Departments of Microbiology and §Immunobiology, Meiji Pharmaceutical University, 2-522-1 Noshio, Kiyose, Tokyo 204-8588, Japan Department of Dermatology and Allergology, Juntendo University Nerima Hospital, Nerima-ku, Tokyo, Japan

Summary

Correspondence Background Malassezia species are suspected to be involved in the development of Takashi Sugita. skin lesions in atopic dermatitis (AD) when the response of adult AD to anti- E-mail: [email protected] inflammatory treatments is poor. However, a comparative analysis of Malassezia flora between adults and children with AD has not been performed. Accepted for publication 8 June 2007 Objectives To compare the cutaneous Malassezia flora between adults and children with AD. Key words Methods Scale samples were collected from skin lesions of 58 patients with AD in age, atopic dermatitis, Malassezia, quantitative the head and neck regions (28 males and 30 females; 31 children and 27 adults), analysis, real-time polymerase chain reaction and fungal DNA was extracted from the samples directly. The number and identi- Conflicts of interest ties of the Malassezia species were analysed with high accuracy using a polymerase None declared. chain reaction-based culture-independent method. The in vivo level of anti- Malassezia IgE antibody was also assayed. Results Malassezia restricta was the predominant species in the children with AD, while both M. restricta and M. globosa predominated in the adults. The adults showed increased sensitization in terms of anti-Malassezia-specific IgE responses in the sera to both M. globosa and M. restricta in comparison with the children. Conclusions The cutaneous Malassezia flora differs significantly between the two age groups.

Atopic dermatitis (AD) is a chronic, fluctuating, inflammatory involved in exacerbating AD.8,12,13 However, the results of skin disease that rarely presents in adulthood; it commonly culture-based analyses of cutaneous Malassezia flora in patients occurs within the first 2 years of life. The worldwide preva- with AD have differed among studies. These discrepancies lence of AD is 10–20% in children and 1–3% in adults.1 are likely to have occurred due to different isolation tech- Atopic dermatitis is a multifactorial disease in which both niques and media and the varying growth characteristics of hereditary and environmental factors play a role. For example, each species. Consequently, Sugita et al.14,15 have recently AD patients with inherited disruptions in the skin barrier developed a molecular-based culture-independent method function can experience induction of IgE production and acti- for analysing cutaneous Malassezia. This method has enabled vation of Th2 cells and an exacerbation of the symptoms of highly accurate analyses. Malassezia species are also associated AD after exposure to various environmental factors.2–7 There- with conditions other than AD, such as seborrhoeic derma- fore, Malassezia species in patients with AD with these features titis, pityriasis versicolor and psoriasis. The major species serve not only as normal skin flora but also as an exacerbating (M. globosa and M. restricta) are common, regardless of the factor. Recently, the number of patients with AD, particularly skin disease,15–18 although the relative quantities of the spe- adults, who respond poorly to anti-inflammatory treatment cies differ with different diseases.14,19 Malassezia have been has been increasing.8 The treatment with antifungal agents of found to colonize the head and neck regions to a greater patients who are affected mainly in the head and neck regions extent than the trunk and limbs.14 has been found to decrease Malassezia colonization and the The effects of ageing on the cutaneous Malassezia flora have severity of the AD lesions, which suggests that Malassezia species not been accurately investigated. In the present study, we play a significant role in AD.9–11 compared the Malassezia flora of two different age groups of Currently, 11 species are recognized in the genus Mala- patients with AD using a polymerase chain reaction (PCR)- ssezia. Several studies have examined which species are based culture-independent method.

2007 The Authors 1178 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1178–1182 Malassezia flora differs between adults and children, Y. Takahata et al. 1179

14 Materials and methods method of Sugita et al. Amplification and detection were per- formed with an ABI PRISM 7500 sequence detection system (Applied Biosystems, Foster City, CA, U.S.A.). Patients and sample collection

Fifty-eight individuals from two different age groups were Detection of Malassezia-specific IgE antibodies sampled in this study. The two groups consisted of children under 15 years of age (mean ± SD 9Æ7±1Æ5 years; 16 boys The levels of specific IgE antibodies against M. globosa and and 15 girls) and adults over 16 years of age (mean ± SD M. restricta were assessed using an enzyme-linked immunosor- 36Æ1±9Æ5 years; 12 men and 15 women). The patients were bent assay according to the method of Kato et al.21 Reactivity all outpatients at Juntendo University Nerima Hospital, Japan, was defined in terms of antibody titres being positive ) ) and had been diagnosed with AD according to the criteria of (‡ 0Æ35 IU mL 1) or negative (< 0Æ35 IU mL 1). Hanifin and Rajka.20 Scale samples were collected from skin lesions by applying Results OpSiteTM transparent dressings (3 · 7 cm; Smith and Nephew Medical, Hull, U.K.) to the neck. As scale samples were Cutaneous Malassezia flora in the different age groups obtained only from the neck, the severity of the AD was not defined with the SCORAD index, but instead was defined Both M. globosa and M. restricta were detected in all cases in the according to the surface of the involved skin, i.e. grade 1, two age groups (Table 1). The other seven Malassezia species scale only without inflammation; grade 2, slight erythema and were detected across the ranges of 0–26% in the children and scale; grade 3, erythema, scale, with a few papules and scratch 0–40% in the adults. Malassezia slooffiae and M. japonica were not marks; grade 4, erythema with swelling, oedema, infiltration detected in either group. The cutaneous Malassezia flora of the and lichenification. Patients had not applied topical corticoster- adults was slightly more diverse than that of the children. oids or antifungal agents for at least the preceding 4 weeks. Samples from the adults contained 3Æ2±1Æ3 species (mean ± The study was approved by the Institutional Review Board of SD), compared with 2Æ8±1Æ1 species in the samples from the Juntendo University Nerima Hospital, and written informed children. consent was obtained from each patient. Quantitative analysis of Malassezia species DNA extraction in the two age groups

The collected OpSite dressings were placed in lysing solution Malassezia colonization was evaluated based on plasmid copy ) ) [100 mmol L 1 Tris–HCl pH 8Æ0, 30 mmol L 1 ethylenedi- number in a real-time PCR assay. Colonization with Malassezia aminetetraacetic acid (EDTA) pH 8Æ0, 0Æ5% sodium dodecyl species was 2Æ4-fold higher in the adults than in the children sulphate] and incubated for 15 min at 100 C. The suspension (mean ± SD 58 827 ± 58 145 vs. 24 262 ± 43 196 copies of ) was extracted with phenol–chloroform–isoamyl alcohol plasmid DNA cm 2; P <0Æ05) (Fig. 1a). (25 : 24 : 1, v ⁄v ⁄v) and then with chloroform–isoamyl alco- Malassezia globosa and M. restricta colonized all of the patients hol (24 : 1, v ⁄v), and the DNA was precipitated by the with AD and accounted for approximately 85Æ5% of the Mala- ) addition of ethanol and 3 mol L 1 sodium acetate, with ssezia isolates in the children and 70Æ6% in the adults (Fig. 1b). Ethatinmate (Nippon Gene, Toyama, Japan) as a precipitation The ratios differed between the two groups: M. restricta coloni- activator. The DNA pellet was resuspended in 30 lLofTE zation was 3Æ5-fold higher than M. globosa colonization in the ) ) buffer (10 mmol L 1 Tris–HCl pH 8Æ0, 1 mmol L 1 EDTA pH 8Æ0) and stored at )20 C until use. Table 1 Detection of Malassezia DNA in children and adults with atopic dermatitis

Detection of Malassezia species from scales Children (n = 23), Adults (n = 20), Eleven species are recognized in the genus Malassezia.Of Malassezia species n (%) n (%) these, M. nana and M. pachydermatis are specific to nonhuman M. globosa 23 (100) 20 (100) animals and therefore were excluded from this study. The M. restricta 23 (100) 20 (100) M. dermatis 3 (13) 7 (35) nine human-associated Malassezia species that colonized the M. sympodialis 2 (9) 4 (20) scales were detected by nested PCR using the method of M. furfur 6 (26) 2 (10) 17 Morishita et al. M. obtusa 2 (9) 3 (15) M. yamatoensis 6 (26) 8 (40) M. slooffiae 00 Quantitative analysis of Malassezia in the scales M. japonica 00 Among the Malassezia flora, two major cutaneous species, Number of species 2Æ2±1Æ13Æ2±1Æ3 M. restricta and M. globosa, were further quantified using a uni- detected (mean ± SD) versal or species-specific TaqMan probe according to the

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1178–1182 1180 Malassezia flora differs between adults and children, Y. Takahata et al.

(a) P > 0·05 ) ) 2 106 P = 0·03 –1 P > 0·05

30 M. globosa M. restricta 20 105

IgE antibody (IU mL 10

<0·35

4 Malassezia Number of plasmid copy (per cm 10 Children group Adult group Children group Adult group

(under 15 years) (over 16 years) Anti- (under 15 years) (over 16 years) N = 23 N = 20 N = 20 N = 17

P = 0·031 Fig 2. Levels of specific IgE antibodies against each Malassezia species in children and adults. (b) P = 0·095 100 M. globosa P >0Æ05; Fig. 2). There was no correlation between the sever- 80 M. restricta ity of disease and the level of IgE antibodies.

60 Discussion

The prevalence of AD has increased continuously over the last 40 30 years, especially in Western industrialized countries, where Distribution (%) it occurs in up to 20–37% of the population.22 It has been 20 suggested that this increasing prevalence results mainly from environmental factors, such as increased air pollution, indoor 0 exposure to house dust mite antigens in less-ventilated mod- Children group Adult group 8 (under 15 years) (over 16 years) ern homes, and dietary changes. The relationship between N = 23 N = 20 decreased exposure to microbial antigens associated with a Western lifestyle and the increasing severity and prevalence of Fig 1. Distribution of Malassezia in lesional skin of adults and children AD has become known as the ‘hygiene hypothesis’.23 It has with atopic dermatitis. (a) Amounts of all Malassezia species. The also been suggested that microorganisms play a role in ) plasmid copy number cm 2 is shown. (b) Distributions of the two AD pathogenesis. major species, M. restricta and M. globosa. Bacterial colonization with Staphylococcus aureus is the most common skin infection in AD (> 90% of patients compared children, whereas M. restricta colonization was 1Æ5-fold higher with 5% of normal individuals) and occurs on both lesional than that of M. globosa in the adults. Quantitative analysis and, to a lesser extent, nonlesional AD skin.24,25 Furthermore, showed that M. restricta was the dominant species in the chil- a causal relationship appears to exist between the number of dren. There was no correlation between disease severity and bacteria present on the skin and the severity of disease in the level of Malassezia colonization. patients with AD. Recently, several researchers have investigated the cutaneous Malassezia microflora, using culture-based methods.26–28 How- Anti-Malassezia IgE antibodies in the two age groups ever, the results have differed across studies. Sugita et al.14,15 In this experiment, 17 serum samples from adults and 20 have developed a PCR-based, culture-independent method to serum samples from children were tested. Specific IgE anti- analyse the cutaneous Malassezia microflora with a high level of bodies against M. globosa and M. restricta were detected (> 0Æ35 accuracy. Using this method, M. restricta and M. globosa have ) IU mL 1) in seven (41%) and seven (41%), respectively, been found to be the most common species in adult patients of the 17 adults with AD, while eight (40%) and four with AD (87Æ5% and 93Æ8%, respectively), while the other (20%), respectively, of the 20 children with AD showed posi- species are detected in fewer than 40% of cases.15 The num- tive reactions. The levels of IgE antibodies against M. globosa bers of M. restricta and M. globosa on the skin of patients with and M. restricta in the adults (mean ± SD 8Æ8±24Æ4 and AD are similar.14 ) 10Æ6±25Æ9IUmL 1, respectively) were higher than those in In the present study, we quantified all the Malassezia species, ) the children (2Æ3±6Æ5 and 0Æ6±1Æ7IUmL 1, respectively; including the two major species (M. restricta and M. globosa), in

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1178–1182 Malassezia flora differs between adults and children, Y. Takahata et al. 1181

AD scales of lesional skin using a highly accurate real-time 2 Cox HE, Moffatt MF, Faux JA et al. Association of atopic dermatitis PCR assay and compared the Malassezia flora between the two to the beta subunit of the high affinity immunoglobulin E receptor. different age groups. Malassezia species were found to colo- Br J Dermatol 1998; 138:182–7. 3 Kawashima T, Noguchi E, Arinami T et al. Linkage and association nize adults to a greater extent than children, as shown in of an interleukin 4 gene polymorphism with atopic dermatitis in Figure 1(a). This may reflect the fact that, compared with Japanese families. J Med 1998; 35:502–4. 29 adults, children have very little sebum on their skin. Of the 4 He JQ, Chan-Yeung M, Becker AB et al. Genetic variants of the IL13 total Malassezia colonization, M. restricta was the dominant and IL4 genes and atopic diseases in at-risk children. Genes Immun species in children, whereas the numbers of M. restricta and 2003; 4:385–9. M. globosa were similar in the adults. Their relative composi- 5 Weidinger S, Klopp N, Rummler L et al. Association of NOD1 poly- tions differed quantitatively between the two age groups, and morphisms with atopic eczema and related phenotypes. J Allergy Clin Immunol 2005; 116:177–84. this difference was probably due to different lipid classes or 30,31 6 Kabesch M, Peters W, Carr D et al. Association between polymor- compositions. phisms in caspase recruitment domain containing protein 15 and The higher anti-Malassezia IgE antibody levels among adults allergy in two German populations. J Allergy Clin Immunol 2003; compared with children with AD may reflect cumulative 111:813–17. exposure.32 In the present study, we divided the patients into 7 Ahmad-Nejad P, Mrabet-Dahbi S, Breuer K et al. The toll-like recep- two groups, children (under 15 years of age) and adults (over tor 2 R753Q polymorphism defines a subgroup of patients with 16 years of age); the latter group showed a higher sensitiza- atopic dermatitis having severe phenotype. J Allergy Clin Immunol 2004; 113:565–7. tion to the antigens of both M. globosa and M. restricta than the 8 Baker BS. The role of microorganisms in atopic dermatitis. Clin Exp former. However, when we divided the patients into children Immunol 2006; 144:1–9. (under 12 years of age) and adults (over 13 years of age), 9 Clemmensen OJ, Hjorth N. Treatment of dermatitis of the head specific IgE antibody against M. restricta was significantly associ- and neck with ketoconazole in patients with type 1 sensitivity to ated with adults with AD (P <0Æ05). Pityrosporum orbiculare. Semin Dermatol 1983; 2:26–9. A Th2 cytokine profile predominates at birth. Microbial 10 Back O, Scheynius A, Johansson SGO. Ketoconazole in atopic antigens induce immune deviations from a Th2 to a Th1 dermatitis: therapeutic response is correlated with decrease in serum IgE. Arch Dermatol Res 1995; 287:448–51. type profile, and adults maintain a Th1 ⁄Th2 balance.8 It has 11 Nikkels AF, Pierard GE. Framing the future of antifungals in atopic been suggested that exposure to microorganisms early in dermatitis. Dermatology 2003; 206:398–400. life is necessary for patients with AD to acquire immuno- 12 Faergemann J. Atopic dermatitis and fungi. Clin Microbiol 2002; competence. However, compared with children, this tends 15:545–63. to become an exacerbating factor in adult patients with AD. 13 Aspres N, Anderson C. Malassezia yeasts in the pathogenesis of atopic Malassezia species were found to colonize the head and neck dermatitis. Australas J Dermatol 2004; 45:199–207. region most heavily.14 Given this observation, we suggest 14 Sugita T, Tajima M, Tsubuku H et al. Quantitative analysis of cuta- neous Malassezia in atopic dermatitis patients using real-time PCR. that antifungal agents may improve lesions in adult pati- Microbiol Immunol 2006; 50:549–52. ents with AD who are affected mainly in the head and 15 Sugita T, Suto H, Unno T et al. Molecular analysis of Malassezia on neck regions and who respond poorly to anti-inflammatory the skin of atopic dermatitis patients and healthy subjects. J Clin 9–11 treatments. Microbiol 2001; 39:3486–90. In the present study, we evaluated the changes in Malassezia 16 Amaya M, Tajima M, Sugita T et al. Molecular analysis of Mala- flora between different age groups of AD patients using a ssezia on the skin of psoriatic patients. Jpn J Med Mycol 2003; PCR-based, culture-independent method with high accuracy. 44:81. 17 Morishita N, Sei Y, Sugita T. Molecular analysis of Malassezia micro- Our results show that M. restricta is the predominant species in flora from patients with pityriasis versicolor. Mycopathologia 2006; children with AD, whereas both M. restricta and M. globosa pre- 161:61–5. dominate in adults. However, further studies are needed to 18 Tajima M. Malassezia species in patients with seborrheic dermatitis clarify the relationship between AD and Malassezia yeasts. and atopic dermatitis. Jpn J Med Mycol 2005; 46:163–7. 19 Takahata Y, Sugita T, Hiruma M et al. Quantitative analysis of Mala- ssezia in the scale of patients with psoriasis using a real-time poly- Acknowledgments merase chain reaction assay. Br J Dermatol 2007; 157:670–3. 20 Hanifin J, Rajka G. Diagnostic features of atopic dermatitis. Acta This study was supported in part by research grants from the Derm Venereol (Stockh) 1980; 92(Suppl.):42–7. Japan Society for the Promotion of Science (no. 16590127); 21 Kato H, Sugita T, Ishibashi Y et al. Detection and quantification of the Ministry of Education, Culture, Sports, Science and Tech- specific IgE antibodies against eight Malassezia species in sera of nology of Japan for an Open Research Center Project; and the patients with atopic dermatitis by using an enzyme-linked immuno- IFO Foundation (to T.S.). sorbent assay. Microbiol Immunol 2006; 50:851–6. 22 Ring J, Kramer U, Schafer T et al. Why are allergies increasing? Curr Opin Immunol 2001; 13:701–8. References 23 Strachan DP. Hay fever, hygiene, and household size. BMJ 1989; 299:1259–60. 1 Schultz Larsen F, Diepgen T, Svensson A. The occurrence of atopic 24 Leyden JJ, Marples RR, Kligman AM. Staphylococcus aureus in the dermatitis in north Europe: an international questionnaire study. lesions of atopic dermatitis. Br J Dermatol 1974; 90:525–30. J Am Acad Dermatol 1996; 34:760–4.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1178–1182 1182 Malassezia flora differs between adults and children, Y. Takahata et al.

25 Ring J, Abeck D, Neuber K. Atopic eczema: role of microorganisms 29 Downing DT, Steward ME, Strauss JS. Changes in sebum on the skin surface. Allergy 1992; 47:265–9. secretion and the sebaceous gland. Clin Geriatr Med 1989; 5:109– 26 Nakabayashi A, Sei Y, Guillot J. Identification of Malassezia species 14. isolated from patients with seborrhoeic dermatitis, atopic derma- 30 Sansone-Bazzano G, Cummings B, Seeler AK et al. Differences in titis, pityriasis versicolor and normal subjects. Med Mycol 2000; the lipid constituents of sebum from pre-pubertal and pubertal 38:337–41. subjects. Br J Dermatol 1980; 103:131–7. 27 Guputa AK, Kohli Y. Prevalence of Malassezia species on various 31 Pochi PE, Strauss JS, Downing DT. Age-related changes in seba- body sites in clinically healthy subjects representing different age ceous gland activity. J Invest Dermatol 1979; 73:108–11. groups. Med Mycol 2004; 42:35–42. 32 Scalabrin DM, Bavbek S, Perzanowski MS et al. Use of specific IgE 28 Lee YW, Yim SM, Lim SH et al. Quantitative investigation on the in assessing the relevance of fungal and dust mite allergens to ato- distribution of Malassezia species on healthy human skin in Korea. pic dermatitis: a comparison with asthmatic and nonasthmatic con- Mycoses 2006; 49:405–10. trol subjects. J Allergy Clin Immunol 1999; 104:1273–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1178–1182 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08203.x Reduction of immunosuppression for transplant-associated skin cancer: thresholds and risks C.C. Otley, M.D. Griffin,* M.R. Charlton, B.S. Edwards, M. Neuburg§ and T. Stasko– for the Reduction of Immunosuppression Task Force of the International Transplant Skin Cancer Collaborative Department of Dermatology, *Division of Nephrology and Hypertension, Division of Gastroenterology and Hepatology, and Division of Cardiovascular Diseases, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, U.S.A. §Department of Dermatology, Medical College of Wisconsin, Milwaukee, WI, U.S.A. –The Vanderbilt Clinic, Nashville, TN, U.S.A.

Summary

Correspondence Background Although evidence supports the efficacy of reducing immunosuppres- Clark C. Otley. sion for transplant-associated skin cancer, clinical thresholds for and risks associ- E-mail: [email protected] ated with reduction are not well defined. Objectives In this study, experienced transplant physicians were surveyed regarding Accepted for publication 20 November 2006 appropriate thresholds for consideration of reduction of immunosuppression and the likelihood of rejection and allograft compromise associated with various lev- Key words els of reduction. allograft compromise, immunosuppression, skin Patients and methods Fifty-two transplant physicians reviewed 13 hypothetical patient cancer, transplant scenarios with graduated morbidity and mortality risk and provided opinions on Conflicts of interest the degree of reduction of immunosuppression that was warranted and the risks None declared. associated with various degrees of reduction. Results Renal, liver and cardiac transplant physicians generally concurred on the level of reduction of immunosuppression warranted by various degrees of skin cancer. As morbidity and mortality from skin cancer increased, physicians were more likely to accept risk to allograft function from more aggressive reduction. Conclusions Reduction of immunosuppression is considered a reasonable adjuvant strategy in recipients of solid organ transplants who have substantial morbidity and mortality risk from skin cancer. Physicians are willing to accept an increased risk of allograft compromise when confronted by severe or extensive skin cancer. Further research is needed to define the precise correlation among levels of reduction of immunosuppression, therapeutic efficacy, and concomitant risks.

Intense systemic immunosuppression is necessary for the pres- particularly oral retinoids, which are effective for reducing the ervation of solid organ allograft transplants. Skin cancer is the incidence of new primary skin cancers.7 most common malignancy developing in the context of trans- The rationale behind reduction of immunosuppression as plant-associated immunosuppression.1,2 Most skin cancers are an adjuvant therapeutic strategy in patients with severe trans- readily managed with common minor surgical procedures, plant-associated skin cancer was recently reviewed.8 Multiple which provide a high cure rate, low morbidity and little risk indirect lines of evidence indicate that reduction of immuno- of metastasis. However, a small proportion of patients who suppression may be efficacious.9–22 Although this evidence have had transplantation experience either high numbers of may be criticized as indirect and incomplete, the fact remains skin cancers or individual skin cancers with considerable risk that clinical decisions regarding reduction of immunosuppres- of mortality.3–6 In patients experiencing severe transplant-asso- sion are being made daily, regardless of the quality of existing ciated skin cancer, reduction of immunosuppression has been evidence. The thresholds for number of skin cancers and the proposed as a reasonable adjuvant therapeutic strategy to risk for metastasis which might warrant reduction of immuno- reduce the risk of further skin cancers or of metastasis and suppression, however, have not been well defined. Addition- death. Other strategies to minimize the risk of new cancers or ally, the risks of allograft rejection and compromise associated adverse metastatic outcomes in these patients include adminis- with reduction of immunosuppression are poorly under- tration of topical or systemic chemopreventive medications, stood.23 A recent consensus survey of expert dermatologists in

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 1183 1184 Transplant-associated skin cancer, C.C. Otley et al. the field of transplant-associated malignancy outlined appro- recorded, and mean and mode responses were determined for priate thresholds of skin cancer for considering reduction of each scenario. immunosuppression.24 The objectives of the survey described The skin cancer scenarios presented were grouped into two herein were to quantify the opinions of transplant physicians basic categories: morbidity due to numerous skin cancers and regarding thresholds for the number of skin cancers and the mortality due to metastatic skin cancer. For scenarios with risk for metastasis which might warrant consideration of both morbidity and mortality concerns, both effects were con- reduction of immunosuppression, and to estimate the risks of sidered. The estimated risks of mortality associated with the allograft rejection and compromise associated with specific various skin cancer scenarios were based, when possible, on scenarios for reduction of immunosuppression. data from the skin cancer literature, standardized to a 3-year period to allow comparison.25–27 Melanoma survival data were Materials and methods derived from validated multivariate data from immunocompe- tent populations.25 Unfortunately, data on survival from mela- The International Transplant-Skin Cancer Collaborative (ITSCC; noma in immunosuppressed patients are insufficient to serve http://www.ITSCC.org accessed 13 September 2007) is a non- as the basis for comparison.26 For skin cancer scenarios with profit medical collaborative composed of 250 dermatologists, less reliable prognostic data available, the best univariate data dermatological surgeons, transplant physicians, oncologists available in the literature in combination with expert dermato- and basic scientists dedicated to the alleviation of skin cancer logical experience were used to derive estimates of risk.27 The in recipients of organ transplants. The Reduction of Immuno- skin cancer scenarios included numerical tumour burden, suppression Task Force of ITSCC recently conducted an expert mortality risks, and the quality-of-life considerations experi- consensus conference of dermatologists regarding appropriate enced by transplant recipients. Reduction of immunosuppres- thresholds for reduction of immunosuppression in transplant- sion levels were stratified into four ascending levels: (i) none; associated skin cancer. The format of that survey was used as (ii) mild; (iii) moderate; and (iv) severe. The potential the basis for the current study, modified after consultation adverse outcomes to allografts resulting from reduction of with specialists in transplant nephrology, transplant hepatology immunosuppression also were categorized into four corre- and transplant cardiology to reflect organ-specific differences. sponding levels: (i) none; (ii) mild; (iii) moderate; and Additionally, various hypothetical scenarios for reduction of (iv) severe. In our risk stratification system, increasingly immunosuppression were created in an attempt to quantify aggressive reduction of immunosuppression correlated with perceptions of risk to allograft function and survival associated progressively higher risks of allograft rejection and of adverse with various levels of reduction of immunosuppression. outcomes (Table 1). Different allograft types (kidney, heart, Members of the task force invited 52 transplant physicians liver) were considered individually because of differences in with whom they collaborated to complete the survey. The sur- their immunogenicity, clinical outcomes associated with acute veys were transmitted to participating transplant physicians by and chronic rejection, and potential for alternative life support electronic mail. An introductory letter was attached which therapies if allograft failure occurs. The survey focused on explained the background of the project and provided explan- reduction of immunosuppression as an adjuvant therapy, atory details for guidance in completing the survey. Among which would be used in addition to standard management of the transplant physicians, 21 were specialists in transplant the tumours, particularly with surgery. nephrology and renal transplant surgery, 18 were specialists in The transplant physicians were also surveyed about the like- transplant hepatology and liver transplant surgery, and 13 lihood of rejection and allograft compromise with various sce- were specialists in transplant cardiology and heart transplant narios for reduction of immunosuppression which were surgery. Physicians and surgeons provided answers pertinent individualized for each organ. only to the allograft of their expertise. Institutions represented by the task force members included Mayo Clinic, Medical Col- Results lege of Wisconsin, Vanderbilt University, University of Wash- ington, University of Pennsylvania, Lahey Clinic, and Wake The mean and mode opinions for all scenarios were identical. Forest Medical Center. The responses of the participants were Table 2 outlines the responses of the transplant physicians

Table 1 Levels of reduction of immunosuppression and associated risks to allograft

Level of reduction of Level of risk immunosuppression to allograft Examples of potential risks to allograft None None No allograft dysfunction Mild Mild Risk of reversible allograft rejection or dysfunction requiring medical treatment Moderate Moderate Risk of partial permanent allograft dysfunction from rejection Severe Severe Risk of allograft failure with potential for death (liver and heart); need to resume dialysis, undergo retransplantation, or potential for death (kidney)

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 Transplant-associated skin cancer, C.C. Otley et al. 1185

Table 2 Survey results on reduction of immunosuppression for specific skin cancer scenarios

Level of reduction of immunosuppression to considera

Skin cancer scenario Kidney allograft Heart allograft Liver allograft 1. No history of actinic keratosis or skin cancer Noneb Noneb Noneb 2. History of actinic keratosis (no risk of mortality; marker for increased Noneb None Noneb risk of skin cancer in future) 3. History of £ 1 NMSC per year (negligible risk of mortality, £ 1 minor None None-mildc Mild surgical procedure per year; patients handle this with ease; warning sign of possible future skin cancers) 4. History of 2–5 NMSCs per year (0Æ5% risk of mortality over 3 years, Mildb Mildb Mild minor-moderate surgical procedure 2–5 times per year; patients can usually handle this, but it starts to bother them; likelihood of numerous future skin cancers) 5. History of 6–10 NMSCs per year (1% risk of mortality over 3 years, Moderate Moderate Moderate minor-moderate surgical procedure 6–10 times per year; patients can usually handle this, but it bothers them; high likelihood of numerous future skin cancers) 6. History of 11–25 NMSCs per year (2% risk of mortality over 3 years, Moderateb Moderate Moderate minor-moderate surgical procedure 11–25 times per year; this level of morbidity causes moderate distress and moderate disfigurement; depression may begin; high likelihood of severe future skin cancers) 7. History of > 25 NMSCs per year (5% risk of mortality over 3 years, Moderate Moderate Moderated moderate-severe surgical procedure > 25 times per year; this level of morbidity causes severe distress and disfigurement; patients question whether transplant was worth it; depression is common; high likelihood of severe and possibly life-threatening future skin cancers) 8. Individual high-risk skin cancer – 1% mortality over 3 years Moderate Moderated Mild (average-risk SCC; cutaneous and oral KS; stage IA melanomae) 9. Individual high-risk skin cancer – 5% mortality over 3 years Moderated Moderated Moderated (moderate-risk SCC; stage IB melanomae) 10. Individual high-risk skin cancer – 10% mortality over 3 years Severed Moderate Moderate (high-risk SCC; early Merkel cell carcinoma; stage IIA melanomae) 11. Individual high-risk skin cancer – 25% mortality over 3 years Severe Moderate-severec Moderated (very high-risk SCC; stage IIB melanomae) 12. Individual high-risk skin cancer – 50% mortality over 3 years Severeb Severeb Severe (metastatic SCC; stage IIC ⁄III melanomae; aggressive Merkel cell carcinoma; visceral KS) 13. Individual high-risk skin cancer – 90% mortality over 3 years Severeb Severeb Severe (untreatable metastatic SCC; stage IV melanomae; metastatic Merkel cell carcinoma)

KS, Kaposi sarcoma; NMSC, nonmelanoma skin cancer; SCC, squamous cell carcinoma. Estimates of mortality risk are derived from data in immunocompetent patients; risk may be higher in immunosuppressed patients. aAppropriate level of reduction of immunosuppression should be individualized on the basis of specific patient- and tumour-related data. bAgreement of 75% or more of respondents (strong consensus). cOpinions were evenly split, and both answers are listed. dNo choice received majority (> 50%) of votes (controversial). eMelanoma staging derived from the American Joint Commission for Cancer.

regarding the level of reduction of immunosuppression and more are designated in Table 2, as are items for which no the associated potential risk to the allograft that would be war- answer received a majority opinion. For skin cancer scenarios ranted for various skin cancer scenarios. Scenarios 1–7 primar- 12 and 13 in a renal transplant recipient, several respondents ily addressed morbidity related to an increasing number of thought that, although the prognosis was poor, reduction of lower-risk skin cancers and their effect on reduced quality of immunosuppression and resumption of dialysis were less pref- life for the transplant recipients. Many of these are common erable than maintaining immunosuppression with its possible scenarios among transplant recipients. Scenarios 8–13 outlined adverse effect on longevity. the risks of mortality from individual high-risk skin cancers. Tables 3–5 outline the participants’ opinions regarding the These scenarios are less common, but some transplant recipi- level of risk for allograft compromise and rejection associated ents do face life-threatening cancers, and death is a well- with various strategies for reduction of immunosuppression. described outcome. Items for which consensus was 75% or With the exception of scenarios in which minimal reduction

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 1186 Transplant-associated skin cancer, C.C. Otley et al.

Table 3 Level of risk of allograft rejection or compromise with various scenarios for reduction of immunosuppression: renal allograft

Reduction of immunosuppression scenario Most likely risk to allografta Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: none Noneb Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 50% None Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% and reduce calcineurin inhibitor 25% Mild Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% and reduce calcineurin inhibitor 75% Moderate Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: complete cessation of calcineurin inhibitor and antiproliferative; increase prednisone to 40 mgc Severeb Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: switch from calcineurin inhibitor to sirolimus Mild

aSee Table 1 for examples of allograft risks. bAgreement of 75% or more of respondents (strong consensus). cTwo respondents indicated that this was not a realistic scenario.

Table 4 Level of risk of allograft rejection or compromise with various scenarios for reduction of immunosuppression: cardiac allograft

Reduction of immunosuppression scenario Most likely risk to allografta Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: none Noneb Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 50% Mildb Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% Mild Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% and reduce calcineurin inhibitor 25% Moderate Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% and reduce calcineurin inhibitor 50% Severe Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: reduce antiproliferative 100% and reduce calcineurin inhibitor 75% Severe Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: complete cessation of calcineurin inhibitor and antiproliferative; increase prednisone as indicated Severeb Baseline regimen: calcineurin inhibitor, full dose; antiproliferative, full dose; prednisone, 5 mg Reduction: switch from calcineurin inhibitor to sirolimus Mild

aSee Table 1 for examples of allograft risks. bAgreement of 75% or more of respondents (strong consensus). of immunosuppression was considered, there was substantial high risk of metastasis. For these patients, consideration of variability among the participants in terms of estimations of reduction of immunosuppression as an adjuvant therapeutic allograft risk associated with specific scenarios for reduction of strategy may be reasonable. A recent review outlined the path- immunosuppression. Of note, 41 of the respondents indicated ogenic rationale for and evidence supporting reduction of that development of clinical guidelines for the reduction of immunosuppression for transplant-associated skin cancer.8 immunosuppression would be helpful, whereas four indicated A recent expert consensus survey outlined dermatologists’ that it would not be helpful. opinions regarding the appropriate reduction of immunosup- pression for transplant-associated skin cancer. However, there Discussion is no discussion of these concepts in the transplant literature. Although the data supporting reduction of immunosuppression Skin cancer is the most common malignancy after solid organ as an effective adjuvant therapeutic strategy are less than defini- transplantation. Fortunately, standard treatments of skin cancer tive, clinicians face these scenarios daily, and organization of are very effective for curing primary skin cancers in most these concepts and the presentation of expert opinions may be patients. Despite the proactive use of chemopreventive strate- helpful to clinicians making clinical decisions. Several respon- gies, including oral retinoids and topical dermatological dents emphasized that the management of transplant recipients agents, some organ transplant recipients experience very high with skin cancer is best accomplished through a multidisciplin- numbers of skin cancers and ⁄or individual skin cancers with a ary effort coordinated by the primary transplant team. In many

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 Transplant-associated skin cancer, C.C. Otley et al. 1187

Table 5 Level of risk of allograft rejection or compromise with various scenarios for Reduction of immunosuppression scenario Most likely risk reduction of immunosuppression: liver to allografta allograft Baseline regimen: calcineurin inhibitor, full dose Reduction: none Noneb Baseline regimen: calcineurin inhibitor, full dose Reduction: reduce calcineurin inhibitor 25% Mild Baseline regimen: calcineurin inhibitor, full dose Reduction: reduce calcineurin inhibitor 50% Moderate Baseline regimen: calcineurin inhibitor, full dose Reduction: reduce calcineurin inhibitor 75%, add Mild low-dose antiproliferative or low-dose sirolimus Baseline regimen: calcineurin inhibitor, full dose Reduction: reduce calcineurin inhibitor 100%, add Mildc low-dose antiproliferative or low-dose sirolimus Baseline regimen: calcineurin inhibitor, full dose Reduction: complete withdrawal of immunosuppression Severeb Baseline regimen: calcineurin inhibitor, full dose Reduction: switch from calcineurin inhibitor to sirolimus Mild

aSee Table 1 for examples of allograft risks. bAgreement of 75% or more of respondents (strong consensus). cNo choice received majority (> 50%) of votes (controversial). medical centres, dermatologists are becoming integrated within Our survey also examined the perceived risks of reduction the transplant centres in order to assist in the management of of immunosuppression based on various scenarios for reduc- these patients and to quantify the risks associated with severe tion of immunosuppression. The results indicate that trans- skin cancer on quality of life and mortality. plant physicians vary considerably in their opinions regarding Several important findings of this survey merit emphasis. the risks associated with specific degrees of reduction of First, all respondents thought that reduction of immunosup- immunosuppression. The transplant literature reflects a similar pression was a reasonable adjuvant therapeutic strategy in trans- variability of experience with rejection associated with reduc- plant patients confronted by severe or life-threatening skin tion of immunosuppression.10,28,29 cancer. Second, assuming that maintenance immunosuppres- In addition to the aforementioned points, important qualifi- sion is already at a level characterized by allograft stability with- cations merit emphasis. The results of this survey may be helpful out excess adverse effects, respondents were unanimous that for physicians to conceptualize the issues surrounding reduction patients without any skin cancer or precancer did not require of immunosuppression; the results should not be interpreted reduction of immunosuppression. Third, transplant physicians as direct advice for patient management. Many respondents are comfortable with increasing levels of reduction of immuno- emphasized that individualization of patient care is optimal, suppression when managing patients with increasing numbers because the risks of reduction of immunosuppression vary con- of skin cancer and increasing risk of mortality. Finally, for siderably between patients based on factors such as age, allograft patients with extreme numbers of skin cancers or with high type, medication levels, HLA match, prior history of rejection, risks of metastatic disease, transplant physicians seem willing to time after transplantation and other comorbidities. We acknowl- consider substantial reduction of immunosuppression despite edge limitations to our findings, including the fact that only one real potential risks to allograft function. Interestingly, the prospective study supports the idea that reduction of immuno- approaches to reduction of immunosuppression were very suppression may be an effective adjuvant strategy for decreasing similar between the different types of allograft specialists, with adverse outcomes of skin cancer in organ transplant recipients.12 only minor differences of opinion regarding levels of reduction However, substantial evidence supports that reduction of immu- of immunosuppression. Surprisingly, the transplant physicians nosuppression diminishes the risk of skin cancer in transplant seem to be comfortable with slightly more aggressive levels of recipients.8 We also acknowledge that the prognostic data pre- reduction of immunosuppression than were dermatologists in a sented were derived from immunocompetent populations and recent survey. These results suggest that consensus guidelines that the degree to which immunosuppression may worsen the for reduction of immunosuppression in organ transplant recipi- prognosis of skin cancer has not been quantified. Uncontrolled ents with skin cancer could be developed. data indicated that skin cancer in immunosuppressed patients Several physicians commented that for scenarios 12 and 13, may be associated with a worse prognosis.26 involving patients with a high predicted risk of mortality from The increasing use of sirolimus as an immunosuppressant severe skin cancer, some patients, when mortality seemed with its additional potential for antineoplastic properties was inevitable, might prefer to maintain their allograft, despite not considered in this survey. Conversion from calcineurin- a possible worse prognosis with maintenance of immuno- based regimens to sirolimus-based regimens is increasingly suppression, in order to avoid resumption of dialysis. popular on the basis of early preliminary data suggesting that

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 1188 Transplant-associated skin cancer, C.C. Otley et al. sirolimus-based regimens may be associated with lower rates 14 Moloney FJ, Kelly PO, Kay EW et al. Maintenance versus reduc- of skin cancer and also reports of regression of skin cancers tion of immunosuppression in renal transplant recipients with after conversion to a sirolimus-based regimen.30–32 aggressive squamous cell carcinoma. Dermatol Surg 2004; 30:674– 8. Continued research into the risks of skin cancer and the 15 Jensen P, Hansen S, Moller B et al. Skin cancer in kidney and heart precise outcomes of severe skin cancer in transplant recipients transplant recipients and different long-term immunosuppressive is necessary for a more data-driven approach to this common therapy regimens. J Am Acad Dermatol 1999; 40:177–86. clinical question. In addition, future investigations should 16 Glover MT, Deeks JJ, Raftery MJ et al. Immunosuppression and risk address the effectiveness of reduction of immunosuppression of non-melanoma skin cancer in renal transplant recipients. Lancet for reducing subsequent skin cancers, the risk of mortality 1997; 349:398. from high-risk skin cancers and the risks of reducing levels of 17 Kehinde EO, Petermann A, Morgan JD et al. Triple therapy and incidence of de novo cancer in renal transplant recipients. Br J Surg immunosuppression on a solid organ allograft. 1994; 81:985–6. 18 Shuttleworth D, Marks R, Griffin PJ, Salaman JR. Epidermal dyspla- Acknowledgments sia and cyclosporine therapy in renal transplant patients: a compar- ison with azathioprine. Br J Dermatol 1989; 120:551–4. Editing, proofreading, and reference verification were pro- 19 Hiesse C, Rieu P, Kriaa F et al. Malignancy after renal transplant- vided by the Section of Scientific Publications, Mayo Clinic, ation: analysis of incidence and risk factors in 1 700 patients fol- Rochester, MN, U.S.A. lowed during a 25-year period. Transplant Proc 1997; 29:831–3. 20 Ramsay HM, Fryer AA, Reece S et al. Clinical risk factors associated with nonmelanoma skin cancer in renal transplant recipients. Am J References Kidney Dis 2000; 36:167–76. 21 Fortina AB, Caforio AL, Piaserico S et al. Skin cancer in heart trans- 1 Berg D, Otley CC. Skin cancer in organ transplant recipients: epide- plant recipients: frequency and risk factor analysis. J Heart Lung miology, pathogenesis, and management. J Am Acad Dermatol 2002; Transplant 2000; 19:249–55. 47:1–17. 22 Gjersvik P, Hansen S, Moller B et al. Are heart transplant recipients 2 Euvrard S, Kanitakis J, Claudy A. Skin cancers after organ trans- more likely to develop skin cancer than kidney transplant recipi- plantation. N Engl J Med 2003; 348:1681–91. ents? Transpl Int 2000; 13 (Suppl. 1):S380–1. 3 Martinez JC, Otley CC, Stasko T et al. Defining the clinical course of 23 Kirk AD, Mannon RB, Swanson SJ, Hale DA. Strategies for mini- metastatic skin cancer in organ transplant recipients: a multicenter mizing immunosuppression in kidney transplantation. Transpl Int collaborative study. Arch Dermatol 2003; 139:301–6. 2005; 18:2–14. 4 Veness MJ, Quinn DI, Ong CS et al. Aggressive cutaneous malignan- 24 Otley CC, Berg D, Ulrich C et al. Reduction of Immunosuppres- cies following cardiothoracic transplantation: the Australian experi- sion Task Force of the International Transplant Skin Cancer ence. Cancer 1999; 85:1758–64. Collaborative and The Skin Care in Organ Transplant Patients 5 Sheil AG, Disney AP, Mathew TH, Amiss N. De novo malignancy Europe. Reduction of immunosuppression for transplant-associ- emerges as a major cause of morbidity and late failure in renal ated skin cancer: expert consensus survey. Br J Dermatol 2006; transplantation. Transplant Proc 1993; 25:1383–4. 154:395–400. 6 Ong CS, Keogh AM, Kossard S et al. Skin cancer in Australian heart 25 Balch CM, Soong SJ, Gershenwald JE et al. Prognostic factors analy- transplant recipients. J Am Acad Dermatol 1999; 40:27–34. sis of 17,600 melanoma patients: validation of the American Joint 7 Bavinck JN, Tieben LM, Van der Woude FJ et al. Prevention of skin Committee on Cancer melanoma staging system. J Clin Oncol 2001; cancer and reduction of keratotic skin lesions during acitretin ther- 19:3622–34. apy in renal transplant recipients: a double-blind, placebo-con- 26 Penn I. Malignant melanoma in organ allograft recipients. Transplan- trolled study. J Clin Oncol 1995; 13:1933–8. tation 1996; 61:274–8. 8 Otley CC, Maragh SL. Reduction of immunosuppression for trans- 27 Rowe DE, Carroll RJ, Day CL Jr. Prognostic factors for local recur- plant-associated skin cancer: rationale and evidence of efficacy. Der- rence, metastasis, and survival rates in squamous cell carcinoma of matol Surg 2005; 31:163–8. the skin, ear, and lip: implications for treatment modality selec- 9 Duman S, Toz H, Asci G et al. Successful treatment of post-trans- tion. J Am Acad Dermatol 1992; 26:976–90. plant Kaposi’s sarcoma by reduction of immunosuppression. Nephrol 28 Penn I. Post-transplant malignancy: the role of immunosuppres- Dial Transplant 2002; 17:892–6. sion. Drug Saf 2000; 23:101–13. 10 Tsai DE, Hardy CL, Tomaszewski JE et al. Reduction in immunosup- 29 Lessan-Pezeshki M, Einollahi B, Khatami MR, Mahdavi M. Kidney pression as initial therapy for posttransplant lymphoproliferative transplantation and Kaposi’s sarcoma: review of 2050 recipients. disorder: analysis of prognostic variables and long-term follow-up Transplant Proc 2001; 33:2818. of 42 adult patients. Transplantation 2001; 71:1076–88. 30 Kahan BD, Knight R, Schoenberg L et al. Ten years of sirolimus 11 Soulillou JP, Giral M. Controlling the incidence of infection and therapy for human renal transplantation: the University of Texas at malignancy by modifying immunosuppression. Transplantation 2001; Houston experience. Transplant Proc 2003; 35 (Suppl.):25S–34S. 72 (Suppl. 12):S89–93. 31 Mathew T, Kries H, Friend P. Two-year incidence of malignancy 12 Dantal J, Hourmant M, Cantarovich D et al. Effect of long-term in sirolimus-treated renal transplant recipients: results from five immunosuppression in kidney-graft recipients on cancer incidence: multicenter studies. Clin Transplant 2004; 18:446–9. randomised comparison of two cyclosporin regimens. Lancet 1998; 32 Kauffman HM, Cherikh WS, Cheng Y et al. Maintenance immuno- 351:623–8. suppression with target-of-rapamycin inhibitors is associated with 13 Otley CC, Coldiron BM, Stasko T, Goldman GD. Decreased skin a reduced incidence of de novo malignancies. Transplantation 2005; cancer after cessation of therapy with transplant-associated immuno- 80:883–9. suppressants. Arch Dermatol 2001; 137:459–63.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1183–1188 CLINICAL AND LABORATORY INVESTIGATIONS DOI 10.1111/j.1365-2133.2007.08235.x Morphoea: a manifestation of infection with Borrelia species? K. Eisendle, T. Grabner and B. Zelger Department of Dermatology and Venereology, Innsbruck Medical University, Anichstraße 35, 6020 Innsbruck, Austria

Summary

Correspondence Background Morphoea or localized scleroderma is a cutaneous inflammatory disease Klaus Eisendle. with still unknown aetiology. Borrelia burgdorferi as causative agent has been dis- E-mail: [email protected] cussed controversially. Objectives To assess the evidence for infection with B. burgdorferi in patients with Accepted for publication 28 July 2007 morphoea by focus-floating microscopy (FFM). Methods Using standard histological equipment, tissue sections stained with a poly- Key words clonal B. burgdorferi antibody were simultaneously scanned through in two planes: aetiology, Borrelia burgdorferi, focus-floating horizontally as in routine cytology, and vertically by focusing through the thick- microscopy, immunohistochemistry, localized ness of the section, i.e. FFM. Part of the material was also investigated with a scleroderma, morphoea Borrelia-specific polymerase chain reaction (PCR). Conflicts of interest Results One hundred and twenty-two cases of morphoea and 68 controls (58 neg- None declared. ative and 10 positive by PCR) were investigated for the presence of Borrelia within tissue specimens. Using FFM Borrelia was detected in 84 cases (68Æ9%) of mor- phoea and in all positive controls, but was absent in all negative controls. Borrelia was significantly more frequent in early inflammatory-rich (75%) than late inflammatory-poor (53%) cases (P =0Æ018). What seemed to be vital micro- organisms were mostly found close to the active border, while degenerated forms were more common in fibrosclerotic parts. The presence of B lymphocytes determined by CD20 staining proved to be a good positive predictor of the microorganism (correlation 0Æ85, P <0Æ001). Borrelia-specific DNA was detected in only one of 30 cases of morphoea analysed by PCR. Conclusions FFM is a reliable and highly sensitive method to detect Borrelia in tissue sections. The frequent detection of this microorganism in morphoea points to a specific involvement of B. burgdorferi or other similar strains in the development of or as a trigger of this disease.

Morphoea is an inflammatory connective tissue disorder of Borrelia in tissue specimens. These conflicting results at least in unknown aetiology. The involvement of Borrelia burgdorferi as a part reflect the difficulties of the various techniques used to causative agent was first proposed by Aberer et al. in 1985.1 prove ⁄document the participation of Borrelia in the disease pro- Since then conflicting results have been obtained by different cess. Serological techniques are unsatisfactory, with false-nega- studies using serological, immunohistochemical, culture and tive (20–80%) and false-positive results (20–50%) in classical polymerase chain reaction (PCR) approaches.2–14 Borrelia has manifestations of borreliosis, such as erythema chronicum frequently been detected in European and Asian patients, but migrans (ECM), Borrelia lymphocytoma (BL) and acrodermatitis not in cases from the U.S.A. or Scotland.15–18 Studies report- chronica atrophicans (ACA). Negative serology does not ing a positive association between B. burgdorferi infection and exclude previous infection with B. burgdorferi and positive serol- morphoea found evidence of the organism in 26–100% of ogy may represent the endemic background.20–22 Histological, cases (see Table 1); on the other hand there are at least 10 histochemical and immunohistochemical detection of micro- reports where no positive cases could be identified. The differ- organisms has turned out to be difficult, frequently unreliable ent studies concerning the detection of Borrelia in morphoea and almost always extremely time consuming.22–25 Cultures have been reviewed by Weide et al.19 and Goodlad et al.13 with specific media can detect Borrelia in all clinical forms, but One major difficulty in assessing the association between these techniques are not generally available and are unreliable morphoea and borreliosis is the challenge of reliably detecting with <50% sensitivity for classical borreliosis,22,26 so negative

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 1189 1190 Detection of Borrelia in morphoea, K. Eisendle et al.

Table 1 Literature with evidence for borrelial Author Country Year Method Results aetiology of morphoea. Results are given as number positive ⁄total tested Aberer and Stanek2 Austria 1987 Immunoperoxidase 7 ⁄21 Aberer et al.3 Austria 1987 Culture 1 ⁄4 Weber et al.58 Germany 1988 Culture 1 ⁄1 Aberer et al.4 Austria 1988 Immunoperoxidase 3 ⁄9 Ross et al.6 Puerto Rico 1990 Silver stain 10 ⁄25 Aberer et al.5 Austria 1991 Culture 1 ⁄11 Schempp et al.7 Germany 1993 PCR 9 ⁄9 Schempp et al.59 Germany 1993 PCR ⁄immunohistochemistry 1 ⁄1 Weidenthaler Germany 1994 PCR 1 ⁄1 et al.46 Granter et al.38 U.S.A. 1994 PCR 1 ⁄1 Trevisan et al.9 Italy 1996 PCR 6 ⁄10 Fujiwara et al.17 Japan 1997 PCR 2 ⁄5 Germany 1997 PCR 3 ⁄4 Breier et al.47 Austria 1999 Culture 1 ⁄1 O¨ zkan et al.12 Turkey 2000 PCR 3 ⁄10 Hercogova41 Czech Republic 2002 PCR 1 ⁄1 This study Germany ⁄Austria 2006 PCR 1 ⁄30 FFM 84 ⁄122

PCR, polymerase chain reaction; FFM, focus-floating microscopy.

cultures may be attributable to the fastidiousness of the organ- logical correlation including photographic documentation in ism in culture. The initial enthusiasm for molecular techniques many instances. Serological results were present for only a gave way to a more realistic evaluation of these methods as it minority of patients and could not be utilized because of the became clear that sensitivity varies (30–90%) according to a priori high endemic background of positive borrelial serology Borrelia strains, source material (fresh, frozen or paraffin mate- in our geographical area.20,21 rial) and the primers applied.22,27–32 All current detection As negative controls served 58 mostly inflammatory skin methods seem to lack sufficient sensitivity for the detection of lesions, which were negative with a Borrelia-specific PCR. As borrelial species, so that even the classical cutaneous borrelial positive controls served 10 cases of clinically and histologically infections remain a diagnosis based on circumstantial evidence characteristic borrelial infections with positive PCR detection combining clinicopathological and laboratory information and of B. burgdorferi (three cases of ECM, three cases of BL, four response to therapy. cases of ACA). As most of our archival paraffin material from We recently developed a highly sensitive immunohisto- Innsbruck had been fixed in inadequately buffered formalin, chemical procedure that proved to be more sensitive than PCR we performed a Borrelia-specific PCR on the 30 cases from in the detection of Borrelia in classical cutaneous borreliosis Friedrichshafen; in 24 of these cases there was enough paraffin (98% vs. 45%) and nearly equally specific (99Æ4% vs. material left to perform FFM. 100%).33 We named this procedure focus-floating microscopy (FFM). Cases where abundant Borrelia were detected by FFM Immunohistochemistry were usually positive by PCR; where fewer organisms were seen with FFM, specimens were more likely to be negative Serial sections were performed for haematoxylin and eosin with PCR. Using this new technique we tried to assess the (H&E) staining and immunohistochemistry from paraffin- evidence for infection with B. burgdorferi in patients with embedded, formalin-fixed tissue blocks as previously morphoea. described.33 Briefly, we used a polyclonal rabbit antibody (Acris BP1002, derived from immunization with whole-cell Materials and methods B. burgdorferi preparations of strain B31 ATCC 35210 reacting with 83-kDa, 41-kDa ⁄flagellin, 32-kDa ⁄OspB and 31-kDa ⁄OspA antigens and their fragments in Western blots, with cross-reac- Patients tion to Treponema pallidum, B. hermsii and B. parkeri) at a dilution We searched the files of the Dermatohistopathological Labora- of 1 : 2000, autoclave antigen retrieval for 30 min in a tory in Innsbruck, Austria, from 1975 to 2006 and in part the sodium citrate buffer (pH 6Æ0–6Æ1) and an incubation time files from the Dermatohistopathological Private Laboratory in of 30 min at 37 C. Further steps followed the Ventana-KIT Friedrichshafen, Germany from 2005 to 2006 and retrieved (Ventana Medical Systems, Munich, Germany) method as 128 cases of morphoea (Innsbruck n = 98, Friedrichshafen routinely used for immunohistochemical analysis in our labo- n = 30). Diagnosis was well established by exact clinicopatho- ratory with a biotinylated second antibody and a streptavidin-

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 Detection of Borrelia in morphoea, K. Eisendle et al. 1191 biotin horseradish peroxidase complex as a third layer. As a Focus-floating microscopy final reaction product we used 3-amino-9-ethylcarbazole (AEC), whose bright red colour proved superior to the brown FFM (Fig. 1a–c) is a modified immunohistochemical tech- colour of diaminobenzidine. The counterstain was omitted to nique which combines several strategies to detect minuscule enable easier recognition of Borrelia. organisms in tissue sections. FFM scans through the sections For technical reasons, the procedure differed slightly for in two planes: horizontally as in routine cytology, and, simul- specimens from each laboratory. In specimens from Innsbruck taneously, vertically by focusing through the thickness of the we first examined all immunohistochemical stains for the cut (usually 3–4 lm) at a magnification of ·200 to ·400.33 presence of Borrelia independently (K.E., T.G. and B.Z.), includ- This holoscopic approach allows detection of B. burgdorferi ing morphoea cases, positive and negative controls. Absence (diameter 0Æ2 lm compared with 2 lm for collagen bundles) of counterstains guaranteed that these sections were evaluated which passes through the section at various angles and accord- in a blinded fashion, without knowledge of or clues for the ingly may appear as undulated, comma-like to dot forms. In correct pathological diagnosis. There was excellent inter- addition, omission of counterstain as well as bright illumina- observer reliability between the different investigators. In rare tion of the scanning field prove to be helpful as the bright red occasions of divergent evaluation, the subtle presence of Bor- colour of the AEC-stained microorganisms best contrasts with relia had been missed by one of the investigators. After FFM the faint yellow colour of unstained collagen bundles as well had been used to decide if Borrelia was present or not, H&E as other tissue structures. sections were correlated; in cases positive for Borrelia, serial sections allowed the exact localization of microorganisms in Polymerase chain reaction relation to the disease process. The cases from Friedrichshafen (one H&E section of each case and the respective paraffin For molecular identification of B. burgdorferi, DNA was prepared block) were sent to Innsbruck without information (no clini- from paraffin-embedded tissue. After deparaffinization with cal, histological, serological or PCR data). We first examined xylene and ethanol and digestion with 0Æ6 mg proteinase K immunohistochemical stains performed in our laboratory for for 16 h the remaining DNA was purified by adsorption chro- the presence of Borrelia independently (K.E., T.G. and B.Z). matography (QIAamp DNA Mini Kit; Qiagen GmbH, Hilden, Using the H&E sections, we then tried to establish a histo- Germany) and the concentration of the sample was adjusted to ) logical diagnosis. Results were sent to our colleagues in Fried- 10 mg L 1. Nested PCR was performed in volumes of 25 lL ) richshafen, who then unblinded the study. In our hands silver with 50 ng DNA, 100 pmol of each primer, 10 mmol L 1 )1 )1 techniques over the last decades never proved to be successful Tris–HCl, pH 9Æ0, 50 mmol L KCl, 1Æ5 mmol L MgCl2, ) for a reliable detection of microorganisms in a routine labora- 200 mmol L 1 of each deoxyribonucleotide triphosphate and tory procedure and thus were not performed in this study. 1Æ5UTaq polymerase, and the samples were subjected to the As we had previously observed the presence of B lympho- following conditions: for the first PCR 30 s at 94 C, 30 s at cytes in all forms of borreliosis, but not in its simulants (e.g. 53 C, 30 s at 72 C for 40 cycles, for the second PCR 30 s at lupus erythematosus tumidus), we additionally performed an 94 C, 30 s at 58 C, 30 s at 72 C for 45 cycles in a PTC immunohistochemical stain for CD20 (L26, 1 : 100; Dako, 200 thermocycler (MJ Research, Inc., Watertown, MA, Glostrup, Denmark) in 38 cases followed by the Ventana-KIT U.S.A.). For amplification the following primers specific for method. the B. burgdorferi 23S rRNA gene were used: for the first PCR:

(a) (b) (c)

Fig 1. Focus-floating microscopy of a control case of erythema chronicum migrans. High-power magnification scanning through different levels of the tissue section shows no microorganisms in the first plane (a), undulated and crossed, slightly granular ‘vanishing’ to delicate forms in the second (b), and plump to delicate, undulated, vibrio-like and dot forms in the third plane (c). Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original magnification · 1000.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 1192 Detection of Borrelia in morphoea, K. Eisendle et al.

Bor-1: 5¢-AGAAGTGCTGGAGTCGA-3¢, Bor-2: 5¢-TAGTGCTCT- Table 2 Detection of Borrelia by focus-floating microscopy (FFM) and ACCTCTATTAA-3¢; for the second PCR: Bor-3: 5¢-GCGAAAG- polymerase chain reaction (PCR) in the different subgroups and CGAGTCTTAAAAGG-3¢, Bor-4: 5¢-ACTAAAATAAGGCTGAACT- activity stages of morphoea. Results are given as number positive ⁄total tested (percentage) TAAAT-3¢.34 After separation on a 2% agarose gel (50 mA for 30 min) and staining with ethidium bromide the PCR product of 219 bp was visualized under ultraviolet radiation Borrelia-positive cases (302 nm). FFM PCR Classification of morphoea35 Statistical analysis Plaque morphoea 66 ⁄93 (71Æ0) 1 ⁄26 (3Æ8) Generalized morphoea 4 ⁄9 (44Æ4) 0 ⁄1(0Æ0) Data were analysed statistically using SPSS statistical software Linear morphoea 5 ⁄7 (71Æ4) NA (SPSS for Windows, version 12.0; SPSS Inc., Chicago, IL, Deep morphoea 9 ⁄13 (69Æ2) 0 ⁄3(0Æ0) U.S.A.). Statistical comparisons were performed using the Stage (activity) of morphoea Æ Æ two-tailed Pearson’s v2 test. The association between the Inflammatory-rich (‘active’) 66 ⁄88 (75 0) 1 ⁄28 (3 6) Inflammatory-poor 18 ⁄34 (52Æ9) 0 ⁄2(0Æ0) immunohistochemical data (CD20 and FFM) was assessed (‘inactive ⁄burned out’) using Cramer’s V ⁄Phi for nominal data. P <0Æ05 was consid- Total 84 ⁄122 (68Æ9) 1 ⁄30 (3Æ3) ered statistically significant. Controls Negativea 0 ⁄58 (0Æ0) 0 ⁄58 (0Æ0) Positiveb 10 ⁄10 (100Æ0) 10 ⁄10 (100Æ0) Results NA, not available. aIncluding normal skin, atopic and stasis derma- The median age of the patients was 53 years (mean 48Æ1, titis, lupus erythematosus (tumidus ⁄subacute cutaneous), acne, range 4–86); 48 (37Æ5%) were male and 80 (62Æ5%) female. eczema, ganglion, dermatofibroma, erythema multiforme, Wells Biopsies were taken mainly from the trunk (n = 55, 43Æ0%) syndrome, prurigo, infectious granulomas (lepra ⁄leishmaniasis), and lower extremities (n = 43, 33Æ6%); 17 (13Æ3%) were mycosis fungoides, drug eruption, foreign body reaction due to from the upper extremities, five (3Æ9%) from the head and ruptured follicular cyst, insect bite, dermatofibroma, folliculitis neck area, one case (0Æ8%) was perianal and in seven cases decalvans, (hypertrophic) scar, keloid, lichen nitidus, pityriasis (5Æ5%) we could not ascertain the localization. According to rubra pilaris, urticaria, Dupuytren’s contracture, scabies, rosacea, progressive systemic sclerosis. bIncluding three cases of erythema the classification of morphoea by Peterson et al.,35 99 (77Æ3%) chronicum migrans, three cases of Borrelia lymphocytoma and four cases were of the plaque type (age range 4–86 years, median cases of acrodermatitis chronica atrophicans. 54), nine (7Æ0%) generalized (age range 14–75 years, median 39), seven (5Æ5%) linear (age range 21–57, median 32), including two cases of coup de sabre, and 13 (10Æ2%) cases vibrio-like to dot forms and rarely clusters and colonies represented deep morphoea (age range 17–57 years, median (Figs 2–6). Generally we found lower numbers of Borrelia in 41), including two cases of eosinophilic fasciitis Shul- our morphoea series than in the classical forms of Borrelia man (Table 2). According to the presence and number of infections such as ECM, BL and ACA. The number of spiro- inflammatory cells we additionally divided our cases into chaetes within sections ranged from a single spirochaete inflammatory-rich (‘active’, n = 94, 73Æ3%) forms, clinically (·400 magnification, Fig. 5) to multiple microorganisms in corresponding to active disease with a lilac ring, and inflam- one high-power field. Detection of Borrelia microorganisms matory-poor (‘inactive’, n = 34, 26Æ6%) forms, clinically showed characteristic findings: the spirochaetes were seen out- without an active border. Cases were estimated as inflamma- side, close to or at the periphery of the inflammatory process tory-rich when lymphocytes and plasma cells were seen easily (Fig. 2). Within the inflammatory and early fibrotic centre, at scanning magnification and fibrocytes and fibroblasts were degenerative products of Borrelia such as swollen, granular frequent and prominent between collagen bundles. In contrast, ‘vanishing’ or clumped material could be found (Fig. 4). A inflammatory-poor cases showed no or only scattered lympho- faint red, diffuse staining of this area proved to be a good clue cytes, rarely plasma cells, and a moderate number of fibro- for detection of degenerative Borrelia within, and what seemed cytes between collagen bundles, which were fibrotic to to be vital forms around or close to the nidus of inflammation occasionally sclerotic. In such instances the epidermis was (Fig. 4). Spirochaetes or their degenerative products were often atrophic with loss of rete ridges and showed a tendency frequently located along or in between collagen bundles to moderate basal hyperpigmentation. In still older stages (collagenotropism) – partially to completely hidden if not fibrocytes further decreased or nearly disappeared and collagen visualized in the correct section plane (Fig. 3). bundles became discohesive, recognizable as part of the pro- Table 2 reveals the results of FFM of archival material from cess only by loss of the normal pattern of tissue arrangement Innsbruck and Friedrichshafen, partially correlated with PCR. forming wavy arrangements of collagen fascicles (‘burned out Of a total of 122 morphoea cases, 84 were positive with FFM morphoea’; atrophoderma Pasini & Pierini). (68Æ9%). Borrelia could be found in all subgroups of morphoea; All forms of Borrelia, as described in detail by Aberer et al.25 for plaque-type, linear and deep morphoea the percentages were seen: single and paired bacilli, very long and undulated, were similar at about 70%, while for generalized morphoea

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 Detection of Borrelia in morphoea, K. Eisendle et al. 1193

(a)

(b)

Fig 2. Focus-floating microscopy of early (c) plaque morphoea. (a–d) Series of magnifications allows identification of a cluster of very long and undulated, plump, slightly granular ‘vanishing’ and delicate microorganisms. Arrows indicate corresponding areas in (a–c). Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original (d) magnification · 1000.

(a) (b)

(c) (d)

Fig 3. Focus-floating microscopy. (a–d) High-power magnifications from Fig. 2, revealing the various aspects of microorganisms at different section planes. Note milky cloud, best seen in (a) and (b). Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original magnification · 1000.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 1194 Detection of Borrelia in morphoea, K. Eisendle et al.

(a) (b)

(c) (d)

Fig 4. Focus-floating microscopy of another case of early plaque morphoea. (a–d) High-power magnifications of four different areas with plentiful microorganisms in net-like colony and cluster forms. Microorganisms are very long and undulated; many are delicate, while others are more plump (b) or granular ‘vanishing’ (c). Note reddish veil overlying some areas with spirochaetes. Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original magnification · 1000. we found a lower, although not statistically significantly with its active border; in late stages red-purple-brown (v2 =2Æ7, P =0Æ44), percentage of positive cases (44%). We macules of burned-out morphoea may mimic ACA. Remark- detected Borrelia significantly more often (v2 =5Æ6, P =0Æ018) ably, these clinical entities can occur at the same time in the in inflammatory-rich (75%) than in inflammatory-poor (53%) same patient, and coexistent morphoea has been described in morphoea. The B. burgdorferi antibody showed no cross-reactions patients with ECM, BL, ACA and Lyme arthritis.36–41 More- with tissue structures. Tissue artefacts occasionally observed over, there are descriptions of new patterns in borreliosis included black or brownish dots. All 58 controls from well- such as interstitial granulomatous dermatitis, which clinically defined inflammatory disorders other than borreliosis (includ- most closely resemble morphoea.42 There are also similarities ing two cases of progressive systemic sclerosis) remained in the histopathological findings (variable infiltrates of lym- negative, both by immunohistochemistry and by PCR analysis. phocytes, macrophages and plasma cells; slight vacuolar Table 3 shows the results of the CD20 staining compared degeneration of the stratum basale in late stages; an increase with the presence of Borrelia. All cases where Borrelia could be of fibrocytes ⁄fibroblasts with variable sclerosis) in ECM and detected were positive for CD20 (correlation 0Æ85, P <0Æ001) ACA on the one hand and in morphoea on the other.43 A and, conversely, Borrelia could not be detected in CD20- bacterial aetiology is further suggested because some cases of negative cases. morphoea respond well to antibiotic therapy,2,5,44,45 such as 3,46,47 penicillin and ceftriaxone. D-penicillamine, discussed in Discussion older text books as inhibiting cross-connections between col- lagen fibres, was likely to have been effective because of Some patients with morphoea show clinical similarities to its metal-chelating activity which deprived the bacteria of classical borreliosis. Early stages of morphoea with a promi- essential trace elements such as manganese, zinc and nent lilac ring are not always easy to differentiate from ECM magnesium.48

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 Detection of Borrelia in morphoea, K. Eisendle et al. 1195

(a) (c)

(b)

Fig 5. Focus-floating microscopy of late plaque morphoea. (a–b) Scanning magnification from the centre of the biopsy (arrow A indicates corresponding structures) reveals a tiny structure (arrow B) which on high-power magnification (c) shows a plump, undulated and crossed part of (degenerating) microorganism. Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original magnification: (a) · 100, (b) · 400, (c) · 1000.

(a) (b)

Fig 6. Two further examples of focus-floating microscopy. (a) Generalized morphoea with a nidus of delicate to plump spirochaetes. Note numerous dots around nidus which vertically pass through this section. (b) Linear morphoea with vibrio-like appearance of the spirochaete, which passes through this section at an oblique angle. Immunohistochemistry for Borrelia burgdorferi, Acris BP 1002, no counterstain; original magnification · 1000.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 1196 Detection of Borrelia in morphoea, K. Eisendle et al.

Table 3 Comparison between detection of Borrelia by focus-floating and B. afzelii VS461, but also newer Borrelia species have been microscopy (FFM) and immunohistochemical staining for CD20 identified. The pathogenic significance of these species, such (L26). Data are given as cases tested (percentage of total); correlation as B. valaisiana, B. hermsii, B. turicatae, B. duttonii, B. parkeri and 0Æ85 (P <0Æ001) most recently B. spielmanii, is not yet fully answered. While B. burgdorferi sensu stricto is the only well-established cause of FFM Lyme disease in the U.S.A., B. afzelii, B. garninii and probably CD20 Negative Positive Total B. valaisiana additionally cause ‘Lyme disease’ in Europe and Asia. Relapsing fever borreliosis caused by B. hermsii, B. turicatae Negative 18 (47) 0 (0) 18 (47) Positive 3 (8) 17 (45) 20 (53) and B. duttonii and ECM by B. spielmanii have been des- cribed.51,52 The study by van Dam et al. suggests that different Total 21 (55) 17 (45) 38 (100) B. burgdorferi genotypes have different pathogenic potentials.53 This is well documented for the classical borrelial manifesta- tions, so ACA rarely occurs in the U.S.A. but is commonly seen in Europe where B. afzelii and B. garinii are more preva- In the present study we detected Borrelia in more than 68% lent.54 Maybe subspecies variations dictate the clinical manifes- of all morphoea cases, with a significantly higher percentage tations that follow infections, with only certain strains (P =0Æ018) in active (75Æ0%) than in inactive morphoea possessing the characteristics required to initiate the develop- (52Æ9%). This might reflect intentional or coincidental antibi- ment of morphoea.13,17 Thus, another explanation for the otic exposure in longer existing cases and ⁄or the natural moderate results by PCR might be that these techniques use course of disease with repression of the microorganisms by primers highly specific for known human pathogenic strains, the immune system. The presence of B lymphocytes as deter- while FFM uses immunohistochemistry with a less specific mined by positive staining for CD20 proved to be a good polyclonal antibody that probably detects most different borre- diagnostic predictor for the presence of Borrelia, with a positive lial species. We did not have the opportunity to examine cases correlation of 0Æ85 (P <0Æ001). This is not surprising as the from the U.S.A., but the detection of Borrelia within those spec- presence of B lymphocytes reflects an immune response imens might give a hint towards a new subgroup of Borrelia in against bacterial infections such as borreliosis. Further, perivas- the U.S.A. cular B cells represent a diagnostic tool for classical borreliosis In summary, borreliosis is a vector-transmitted disease such as ECM. We speculate that the chronic persistence of acti- whose causative agents, B. burgdorferi and variants, share col- vated B lymphocytes reflects the frustrated attempt of the lagenotropism, whereas other spirochaetes or spiral-shaped immune system to overcome the infection and could contrib- bacteria are epitheliotropic and endotheliotropic (e.g. T. palli- ute to the pathophysiological development of morphoea. The dum) or mucotropic (e.g. Helicobacter pylori). Fibronectin-binding infection hypothesis might also explain the linear pattern of proteins of B. burgdorferi promote bacterial attachment to glycos- morphoea where neurotropic Borrelia species could coinfect aminoglycans which are most prominent in connective tissue and partially damage peripheral nerves and thereby cause the between collagen bundles.55,56 While the immune system can development of linear morphoea in the area supplied by the cope with the organisms in ECM and BL comparatively damaged nerve. The low percentage of detection of borrelial quickly, in ACA and morphoea the situation seems more com- DNA with PCR in our study, with one positive case in 30 plicated. The low level of microorganisms probably indicates (3Æ3%), indicates the problematic nature of this technique for that the disease is not only due to the effect of the infectious reliable detection of Borrelia in tissue specimens. agent, but also reflects the challenge for the immune system The reason for the inconsistent results in PCR studies [posi- due to the location of the microorganisms and ⁄or even a com- tive (Table 1) or negative8,10,11,13,15,49] could be the low promised immune reaction in patients with morphoea them- number of microorganisms found in the tissue, i.e. below the selves, where Borrelia might trigger a subsequent autoimmune detection threshold for this technique. Other explanations reaction.57 This could also explain why not all patients benefit include previous antibiotic treatment, old stage of disease, from antibiotic therapy. FFM is a reliable and highly sensitive wrong biopsy site (e.g. from negative sclerotic area), or method to detect Borrelia in tissue sections of routinely forma- wrong fixation of tissue specimens leading to DNA cross-link- lin-fixed, paraffin-embedded material. ing, e.g. with inadequately buffered formalin. Further – except 8 15 for the studies of Ranki et al., Dillon et al. and Wienecke Acknowledgments et al.10 – most other negative PCR studies are lacking appropri- ate positive controls in terms of detection of borrelial DNA in The authors thank Heinz Kutzner, MD and Gabriele Palmedo, tissue specimens from classical borrelioses such as ECM, BL PhD, Dermatohistological Private Laboratory Friedrichshafen, and ACA.11,49,50 Thus, the reliability of the DNA extraction Germany, for contributing blinded cases of morphoea includ- method for small amounts of DNA or the PCR technique used ing paraffin material and molecular data. Walter H.C. in these studies remains somewhat debatable. Burgdorf, MD, Tutzing, Germany helped with manuscript There is still another explanation for negative PCR results: preparation. We also thank Birgit Moser, Margit Abenthung B. burgdorferi sensu lato includes B. burgdorferi sensu stricto, B. garinii and Nadja Greier for their excellent technical assistance.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 Detection of Borrelia in morphoea, K. Eisendle et al. 1197

References 21 Plorer A, Sepp N, Schmutzhard E et al. Effects of adequate versus inadequate treatment of cutaneous manifestations of Lyme borreli- 1 Aberer E, Neumann R, Stanek G. Is localised scleroderma a Borrelia osis on the incidence of late complications and late serologic status. infection? Lancet 1985; ii:278. J Invest Dermatol 1993; 100:103–9. 2 Aberer E, Stanek G. Histological evidence for spirochetal origin of 22 Aguero-Rosenfeld ME, Wang G, Schwartz I, Wormser GP. Diagno- morphea and lichen sclerosus et atrophicans. Infection 1987; 16:25– sis of Lyme borreliosis. Clin Microbiol Rev 2005; 18:484–509. 6. 23 de Koning J, Bosma RB, Hoogkamp-Korstanje JA. Demonstration 3 Aberer E, Stanek G, Ertl M, Neumann R. Evidence for spirochetal of spirochetes in patients with Lyme disease with a modified silver origin of circumscribed scleroderma (morphea). Acta Derm Venereol stain. J Med Microbiol 1987; 23:261–7. (Stockh) 1987; 67:225–31. 24 Aberer E, Duray PH. Morphology of Borrelia burgdorferi: structural 4 Aberer E, Kollegger H, Kristoferitsch W, Stanek G. Neuroborreliosis patterns of cultured borreliae in relation to staining methods. J Clin in morphea and lichen sclerosus et atrophicans. J Am Acad Dermatol Microbiol 1991; 29:764–72. 1988; 19:820–5. 25 Aberer E, Kersten A, Klade H et al. Heterogeneity of Borrelia burgdorferi 5 Aberer E, Klade H, Stanek G, Gebhart W. Borrelia burgdorferi and dif- in the skin. Am J Dermatopathol 1996; 18:571–9. ferent types of morphea. Dermatologica 1991; 182:145–54. 26 Berger BW, Kaplan MH, Rothenberg IR, Barbour AG. Isolation and 6 Ross SA, Sanchez JL, Taboas JO. Spirochetal forms in the dermal characterization of the Lyme disease spirochete from the skin of lesions of morphea and lichen sclerosus et atrophicans. Am J Dermato- patients with erythema chronicum migrans. J Am Acad Dermatol pathol 1990; 12:357–62. 1985; 13:444–9. 7 Schempp C, Bocklage H, Lange R et al. Further evidence for Borrelia 27 Melchers W, Meis J, Rosa P et al. Amplification of Borrelia burgdorferi burgdorferi infection in morphea and lichen sclerosus et atrophicus DNA in skin biopsies from patients with Lyme disease. J Clin Micro- confirmed by DNA amplification. J Invest Dermatol 1993; 100:717–20. biol 1991; 29:2401–6. 8 Ranki A, Aavik E, Peterson P et al. Successful amplification of DNA 28 Guy EC, Stanek G. Detection of Borrelia burgdorferi in patients with specific for Finnish Borrelia burgdorferi isolates in erythema chronicum Lyme disease by the polymerase chain reaction. J Clin Pathol 1991; migrans but not in circumscribed scleroderma lesions. J Invest 44:610–11. Dermatol 1994; 102:339–45. 29 Wienecke R, Neubert U, Volkenandt M. Molecular detection of 9 Trevisan G, Stinco G, Nobile C. Detection of Borrelia burgdorferi in Borrelia burgdorferi in formalin-fixed paraffin-embedded lesions of skin biopsies from patients with morphea by polymerase chain Lyme disease. J Cutan Pathol 1993; 20:385–8. reaction. J Eur Acad Dermatol Venereol 1996; 6:15–19. 30 Moter SE, Hofmann H, Wallich R et al. Detection of Borrelia burgdorferi 10 Wienecke R, Schlupen EM, Zochling N et al. No evidence for Borrelia sensu lato in lesional skin of patients with erythema chronicum mig- burgdorferi-specific DNA in lesions of localized scleroderma. J Invest rans and acrodermatitis chronica atrophicans by ospA-specific PCR. Dermatol 1995; 104:23–6. J Clin Microbiol 1994; 32:2980–8. 11 Weide B, Schittek B, Klyscz T et al. Morphoea is neither associated 31 von Stedingk LV, Olsson I, Hanson HS et al. Polymerase chain reac- with features of Borrelia burgdorferi infection, nor is this agent detect- tion for detection of Borrelia burgdorferi DNA in skin lesions of early able in lesional skin by polymerase chain reaction. Br J Dermatol and late Lyme borreliosis. Eur J Clin Microbiol Infect Dis 1995; 14:1–5. 2000; 143:780–5. 32 Brettschneider S, Bruckbauer H, Klugbauer N, Hofmann H. Diag- 12 O¨ zkan S, Atabey N, Fetil E et al. Evidence for Borrelia burgdorferi in nostic value of PCR for detection of Borrelia burgdorferi in skin biopsy morphea and lichen sclerosus. Int J Dermatol 2000; 39:278–83. and urine samples from patients with skin borreliosis. J Clin Micro- 13 Goodlad JR, Davidson MM, Gordon P et al. Morphoea and Borrelia biol 1998; 36:2658–65. burgdorferi: results from the Scottish Highlands in the context of the 33 Eisendle K, Grabner T, Zelger B. Focus-floating microscopy: ‘gold world literature. Mol Pathol 2002; 55:374–8. standard’ for cutaneous borreliosis? Am J Clin Pathol 2007; 14 Palacios R, Torres A, Trujillo R. IgG antibody reactivity to Borrelia 127:213–22. burgdorferi sensu stricto antigens in patients with morphea in Colombia. 34 Schwartz JJ, Gazumyan A, Schwartz I. rRNA gene organization in Int J Dermatol 2003; 42:882–6. the Lyme disease spirochete, Borrelia burgdorferi. J Bacteriol 1992; 15 Dillon WI, Saed GM, Fivenson DP. Borrelia burgdorferi DNA is 174:3757–65. undetectable by polymerase chain reaction in skin lesions of mor- 35 Peterson LS, Nelson AM, Su WP. Classification of morphea (local- phea, scleroderma, or lichen sclerosus et atrophicus of patients ized scleroderma). Mayo Clin Proc 1995; 70:1068–76. from North America. J Am Acad Dermatol 1995; 33:617–20. 36 Asbrink E, Hovmark A, Olsson I. Clinical manifestations of acro- 16 Fan W, Leonardi CL, Penneys NS. Absence of Borrelia burgdorferi in dermatitis chronica atrophicans in 50 Swedish patients. Zentralbl patients with localized scleroderma (morphea). J Am Acad Dermatol Bakteriol Mikrobiol Hyg [A] 1986; 263:253–61. 1995; 33:682–4. 37 Buchner SA. [Morphea – a tick transmitted borreliosis of the skin? 17 Fujiwara H, Fujiwara K, Hashimoto K. Detection of Borrelia burgdorferi A contribution to the pathogenesis of circumscript scleroderma]. DNA (B. garinii or B. afzelii) in morphea and lichen sclerosus et Z Hautkr 1989; 64:661–4, 667–9. atrophicans tissue of German and Japanese but not of US patients. 38 Granter SR, Barnhill RL, Hewins ME, Duray PH. Identification of Arch Dermatol 1997; 133:41–4. Borrelia burgdorferi in diffuse fasciitis with peripheral eosinophilia: 18 Colome´-Grimmer MI, Payne DA, Tyring SK, Sa´nchez RL. Borrelia borrelial fasciitis. JAMA 1994; 272:1283–5. burgdorferi DNA and Borrelia hermsii DNA are not associated with 39 Duray PH, Asbrink E, Weber K. The cutaneous manifestations of morphea or lichen sclerosus et atrophicus in the southwestern human Lyme disease: a widening spectrum. Adv Dermatol 1989; United States. Arch Dermatol 1997; 133:1174. 4:255–75. 19 Weide B, Walz T, Garbe C. Is morphoea caused by Borrelia burgdorferi? 40 Asbrink E, Hovmark A, Olsson I. Lymphadenosis benigna cutis sol- A review. Br J Dermatol 2000; 142:636–44. itaria – Borrelia lymphocytoma in Sweden. Zentralbl Bakeriol Mikrobiol 20 Schmutzhard E, Stanek G, Pletschette M et al. Infections following Hyg [A] 1989; 263 (Suppl. 18):159–63. tickbites. Tick borne encephalitis and Lyme borreliosis – a prospec- 41 Hercogova J. Borrelia burgdorferi: a protagonist in Lyme disease, a tive epidemiological study from Tyrol. Infection 1988; 16:269–72. bystander in morphoea? J Eur Acad Dermatol Venereol 2002; 16:98–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 1198 Detection of Borrelia in morphoea, K. Eisendle et al.

42 Moreno C, Kutzner H, Palmedo G et al. Interstitial granulomatous atrophicus: a PCR study of thirty-five cases. J Cutan Pathol 1996; dermatitis with histiocytic pseudorosettes: a new histopathologic 23:350–8. pattern in cutaneous borreliosis. Detection of Borrelia burgdorferi DNA 51 Foldvari G, Farkas R, Lakos A. Borrelia spielmanii erythema migrans, sequences by a highly sensitive PCR–ELISA. J Am Acad Dermatol Hungary. Emerg Infect Dis 2005; 11:1794–5. 2003; 48:376–84. 52 Rebaudet S, Parola P. Epidemiology of relapsing fever borreliosis in 43 Aberer E, Klade H, Hobisch G. A clinical, histological, and immuno- Europe. FEMS Immunol Med Microbiol 2006; 48:11–15. histochemical comparison of acrodermatitis chronica atrophicans 53 van Dam AP, Kuiper H, Vos K et al. Different genospecies of Borrelia and morphea. Am J Dermatopathol 1991; 13:334–41. burgdorferi are associated with distinct clinical manifestations of 44 Buchner SA, Rufli T. [Morphoea profunda]. Hautarzt 1990; 41:155– Lyme borreliosis. Clin Infect Dis 1993; 17:708–17. 7. 54 Nadelman RB, Wormser GP. Lyme borreliosis. Lancet 1998; 45 Abele DC, Bedingfield RB, Chandler FW, Given KS. Progressive 352:557–65. facial hemiatrophy (Parry–Romberg syndrome) and borreliosis. 55 Parveen N, Caimano M, Radolf JD, Leong JM. Adaptation of the J Am Acad Dermatol 1990; 22:531–3. Lyme disease spirochaete to the mammalian host environment 46 Weidenthaler B, Roux M, Moter SE et al. Sclerodermiform skin results in enhanced glycosaminoglycan and host cell binding. Mol changes in Borrelia burgdorferi infection. Diagnostic use of polymerase Microbiol 2003; 47:1433–44. chain reaction. Hautarzt 1994; 45:171–5. 56 Fischer JR, LeBlanc KT, Leong JM. Fibronectin binding protein 47 Breier FH, Aberer E, Stanek G et al. Isolation of Borrelia afzelii from BBK32 of the Lyme disease spirochete promotes bacterial attach- circumscribed scleroderma. Br J Dermatol 1999; 140:925–30. ment to glycosaminoglycans. Infect Immun 2006; 74:435–41. 48 Posey JE, Gherardini FC. Lack of a role for iron in the Lyme disease 57 Behar SM, Porcelli SA. Mechanisms of autoimmune disease induc- pathogen. Science 2000; 288:1651–3. tion. The role of the immune response to microbial pathogens. 49 Alonso-Llamazares J, Persing DH, Anda P et al. No evidence for Arthritis Rheum 1995; 38:458–76. Borrelia burgdorferi infection in lesions of morphea and lichen sclero- 58 Weber K, Preac-Mursic V, Reimers CD. Spirochetes isolated from sus et atrophicus in Spain. A prospective study and literature two patients with morphea. Infection 1988; 16:25–6. review. Acta Derm Venereol (Stockh) 1997; 77:299–304. 59 Schempp C, Bocklage H, Owsianowski M et al. [In vivo and in vitro 50 De Vito JR, Merogi AJ, Vo T et al. Role of Borrelia burgdorferi in detection of Borrelia infection in morphea-like skin changes with the pathogenesis of morphea ⁄scleroderma and lichen sclerosus et negative Borrelia serology]. Hautarzt 1993; 44:14–18.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1189–1198 CONTACT DERMATITIS AND ALLERGY DOI 10.1111/j.1365-2133.2007.08252.x Filaggrin null alleles are not associated with hand eczema or contact allergy A. Lerbaek,* H. Bisgaard, T. Agner, K. Ohm Kyvik,§ C.N.A. Palmer– and T. Menne´ *National Allergy Research Centre, Gentofte Hospital, University of Copenhagen, Ledreborg Alle´ 40, 1, 2820 Gentofte, Denmark Department of Dermatology, Gentofte Hospital, University of Copenhagen, Denmark Danish Paediatric Asthma Centre, Gentofte Hospital, University of Copenhagen, Denmark §The Danish Twin Registry, Epidemiology, University of Southern Denmark, Odense, Denmark –Population Pharmacogenetics Group, Biomedical Research Centre, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, U.K.

Summary

Correspondence Background The filaggrin protein is a key component of stratum corneum and Anne Lerbaek. homo- or heterozygotes for the filaggrin variant alleles R501X and 2282del4 E-mail: [email protected] have varying degrees of impaired skin barrier. The variant alleles have repeatedly been shown to be associated with atopic dermatitis. Any possible association with Accepted for publication 31 July 2007 hand eczema or contact allergy are unexplored. Objectives To explore associations between the variant alleles, hand eczema, contact Key words allergy and atopic dermatitis. atopic dermatitis, contact allergy, filaggrin, Methods In total, 183 adult individuals participated in a clinical examination of the hand eczema hands, patch testing and filaggrin genotyping. Children without any evidence of Conflicts of interest atopic dermatitis from the Copenhagen Prospective Study on Asthma in Child- 2 None declared. hood (COPSAC) study were used as controls. The v test was used for compar- ison of allele frequencies. Results The majority (73%) had hand eczema, 25% had contact allergy and 14% had a diagnosis of atopic dermatitis. The association between atopic dermatitis and the filaggrin variant alleles was confirmed (odds ratio 3Æ5, P =0Æ015). Allele frequencies in individuals with hand eczema or contact allergy were not statistic- ally significantly increased. Conclusion There is no association between the variant alleles and hand eczema or contact allergy.

Filament aggregating protein (filaggrin) is an essential com- A number of studies have recently established a strong asso- ponent in the terminal differentiation of the epidermis and ciation between the filaggrin variant alleles and atopic derma- formation of the stratum corneum. Profilaggrin, the precursor titis,3 in particular, atopic dermatitis associated with asthma of filaggrin, accumulates in the keratohyalin granules formed and allergic rhinitis,3,4 IgE sensitization,5 early onset6 and per- in the granular layers of the epidermis. During cornification sistence into adulthood.7 keratin filaments in keratinocytes are aggregated by filaggrin, Hand eczema is a frequent, often chronic relapsing disease, resulting in a flattening of the keratinocytes and eventually with a heterogeneous aetiology, including irritant and allergic formation of the cornified cell envelope, which is crucial for contact dermatitis, atopic dermatitis, mixed forms and minor the skin barrier function.1 The skin barrier provides protection groups with vesicular and hyperkeratotic hand eczema. Atopic against water loss and penetration of chemicals, and infectious dermatitis is one of the main risk factors for hand eczema.8 and allergenic agents. Genetic risk factors significantly influence the risk of develop- Homozygotes or compound heterozygotes for the two loss- ing hand eczema,9 even in the absence of atopic derma- of-function mutations (null alleles) R501X and 2282del4 in titis.10,11 Genetic markers for hand eczema have not yet been the gene encoding filaggrin have a complete loss of filaggrin identified and any possible association with the filaggrin vari- products and present clinically with ichthyosis vulgaris, char- ant alleles is unexplored. An impaired skin barrier facilitates acterized by dry scaly skin.2 The mode of inheritance is semi- the penetrance of allergens, but whether or not the presence dominant with variable penetrance; heterozygotes may present of mutations in the gene encoding filaggrin is associated with with a mild form of ichthyosis vulgaris or without symptoms. an increased frequency of contact allergy is unknown.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 1199 1200 Filaggrin null alleles, hand eczema and contact allergy, A. Lerbaek et al.

The primary aim of this study was to investigate whether for participation in compliance with the principles of the Hel- any relationship exists between the filaggrin variants R501X sinki Declaration. and 2282del4 and hand eczema. Secondly, any associations between the variant alleles and contact allergy and atopic Hand eczema and atopic dermatitis dermatitis were explored. A diagnosis of hand eczema was based on a positive answer Materials and methods to a question on self-reported hand eczema (Have you ever had hand eczema?),13 given either in 1997–1998 or at the current examination. Fifteen individuals with self-reported Study population hand eczema in 1997–1998 denied hand eczema at the In 1996, a cohort of 6666 same-sex twin individuals born present examination. Some of them had other diagnoses between 1953 and 1976 and living on Zealand or its neigh- such as psoriasis and polymorphic light eruption on the bouring islands was drawn from the Danish Twin Registry hands. In case of present hand eczema, symptoms (scaling, and received a short questionnaire on hand eczema.9 A total erythema, vesicles, papules, fissures and oedema) were of 5610 twin individuals responded. Twin pairs where one or recorded. The U.K. Working Party’s Diagnostic Criteria were both twin individuals had self-reported hand eczema or used to define whether or not participants had ever had reported symptoms of hand eczema in the questionnaire and atopic dermatitis.14 lived within 60 km of Copenhagen were invited to a clinical examination and patch test in 1997–1998.12 A total of 1076 Filaggrin genotyping twin individuals participated. In 2005, all twin pairs where one or both had self-reported hand eczema in 1997–1998 Venous blood samples or mouth swabs were collected from were identified (659 individuals). Addresses were available for the twin individuals and kept at )80 C. DNA was prepared 605 twin individuals and they were invited by mail to a sec- from blood samples and mouth swabs using QIAamp 96 DNA ond clinical examination and patch test in 2005–2006 procedures (Qiagen GmbH, Hilden, Germany). Genotyping (Fig. 1). All individuals gave their written, informed consent for R501X and 2282del4 was performed by TaqMan allelic

6666 twin individuals received a questionnaire on HE in 1996

5610 twin individuals responded

659 twin individuals participated in clinical Inclusion criteria: self-reported HE examination and patch test in 1997–1998 or co-twin with self-reported HE. Address within 60 km of Copenhagen

605 twin individuals invited to clinical examination and patch test in 2005–2006

274 twin individuals volunteered to participate in 2005–2006

Successful DNA genotyping in 263 twin individuals (80 pairs and 103 single twins)

183 twin individuals selected for analysis One individual randomly selected – 33 with HE including 26 with AD from each twin pair. All single twin and 37 with CA individuals selected – 50 without HE including 8 with CA Fig 1. Flow diagram illustrating recruitment of the study population. HE, hand eczema; AD, atopic dermatitis; CA, contact allergy.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 Filaggrin null alleles, hand eczema and contact allergy, A. Lerbaek et al. 1201 discrimination assays (Applied Biosystems, Foster City, CA, frequency in women and men was 32% and 11%, respec- U.S.A.) as previously described.3,15 tively. Nickel was the primary allergen responsible for sensiti- zation (data not shown). A total of 14% had current or previous history of atopic dermatitis. All twin individuals with Patch testing atopic dermatitis reported at least one episode of hand eczema. The ready-to-use True Test system panel 1 and 2 (Mekos The frequency of atopic dermatitis in the hand eczema group Laboratories A ⁄S, Hillerod, Denmark) was sent by mail to all was 20%. participants. Patches were placed on the back by the partici- The overall allele frequencies of R501X and 2282del4 in pants 3 days in advance of the scheduled examination and the twin cohort were 3% for both variants (yielding carrier removed after 2 days. Reading of the patches was done on frequencies of 7%). As there were no compound heterozyg- day 3 according to the International Contact Dermatitis otes the combined carrier frequency was 13%. There were no Research Group guidelines.16 homozygotes in the twin cohort. Highest allele frequencies were found in the twin subgroups with atopic dermatitis and contact allergy, 23% and 16%, respectively. Allele frequencies Control groups in the twin cohort, the twin subgroups and the COPSAC At first, allele frequencies in twin individuals with hand group are shown in Table 1. eczema were compared with twin individuals without hand No association between the phenotype with hand eczema eczema, and likewise twin individuals with and without con- and the two variant alleles was found. Also, no association tact allergy and atopic dermatitis were compared. As the num- between contact allergy (positive patch test) and the variant ber of twin individuals without hand eczema and contact alleles could be demonstrated. The increased combined carrier allergy was limited, allele frequencies in the twin subgroups frequency in individuals with atopic dermatitis did not reach with hand eczema and contact allergy were also compared statistical significance when compared with twin individuals with a group of 189 children without atopic dermatitis without atopic dermatitis. Statistical results from the compari- (91 male and 98 female), all born to Danish mothers with sons are displayed in Table 2. asthma. The children are currently being followed from birth Allele frequencies in the twin subgroups with hand eczema in a prospective longitudinal follow-up study [the Copenhagen or contact allergy were not statistically significantly different Prospective Study on Asthma in Childhood (COPSAC) from allele frequencies in the COPSAC subgroup of children study].3,17 Atopic dermatitis in the COPSAC study was defined without atopic dermatitis. Comparison of the twin subgroup using the criteria of Hanifin and Rajka.18 Finally, in an analysis with atopic dermatitis with the COPSAC subgroup reached restricted to the subgroup of twins with hand eczema, allele statistical significance (see Table 2 for details). frequencies in the subgroup with and without atopic derma- In the subanalysis restricted to twin individuals with hand titis were compared. eczema, comparison of the combined allele frequency in indi- viduals with atopic dermatitis (23%) with the subgroup with- out atopic dermatitis (10%) was borderline statistically Statistical analysis significant [odds ratio (OR) 2Æ6, 95% confidence interval When data on both twin individuals in a twin pair were avail- 0Æ87–7Æ91, v2 =3Æ072, P =0Æ080]. able, one twin individual was randomly excluded from the Drop-out analysis of the 274 twin individuals participating analysis, thus leaving 183 twin individuals for analysis in the present study vs. those where one or both twin individ- (Fig. 1). Allele frequencies were compared in subgroups of uals had self-reported hand eczema in 1997–1998 (659 twin twins and in the COPSAC subgroup using the v2 test. Both individuals) revealed no statistically significant difference variants were in Hardy–Weinberg equilibrium in the twin regarding sex, zygosity, hand eczema status in 1997–1998, cohort, the twin subgroups and in the COPSAC subgroup. The co-twins hand eczema status in 1997–1998, patch-test status v2 test was used in the drop-out analysis. SPSS version 11.0 or atopic dermatitis status (data not shown). Age was the only was used for statistical analyses. statistically significant factor influencing willingness to partici- pate. When subdividing the twin individuals into three Results groups, 35% from the youngest age group and 41% in the middle group participated, whereas 50% from the oldest age A total of 274 twin individuals participated. DNA genotyping group volunteered for the study (P =0Æ009). was successful in 263 individuals and 183 twin individuals (70 monozygotic, 103 dizygotic and 10 with unknown zygos- Discussion ity) were selected for analysis (see Statistical analysis). The fol- lowing descriptive data apply to the 183 individuals. The The influence of genetic factors independent of atopic derma- mean age was 41 years (SD 6Æ6) and the male to female ratio titis on the risk of hand eczema was recently confirmed.10 was 64 : 119. The majority (73%) had had hand eczema and Many candidate genes for atopic dermatitis have been pro- in this group, clinical signs of hand eczema were found in posed and investigated, but the filaggrin null alleles, R501X 41%. At least one positive patch test was detected in 25%; the and 2282del4, were the first to be successfully replicated in a

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 1202 Filaggrin null alleles, hand eczema and contact allergy, A. Lerbaek et al.

Table 1 Frequency of filaggrin null alleles in the twin cohort and dependent on hand eczema (HE), contact allergy (CA) and atopic dermatitis (AD) status as well as in the Copenhagen Prospective Study on Asthma in Childhood (COPSAC) cohort

All HE No HE CA No CA AD No AD HE and AD HE and no AD COPSACa Genotype (n = 183) (n = 133) (n = 50) (n = 45) (n = 136) (n = 26) (n = 157) (n = 26) (n = 107) (n = 189) R501X AA 171 (93Æ4) 126 (94Æ7) 45 (90Æ0) 42 (93Æ3) 127 (93Æ4) 23 (88Æ5) 148 (94Æ3) 23 (88Æ5) 103 (96Æ3) 182 (96Æ3) Aa 12 (6Æ6) 7 (5Æ3) 5 (10Æ0) 3 (6Æ7) 9 (6Æ6) 3 (11Æ5) 9 (5Æ7) 3 (11Æ5) 4 (3Æ7) 7 (3Æ7) aa0 000 00 00 0 0 2282del4 AA 171 (93Æ4) 123 (92Æ5) 48 (96Æ0) 41 (91Æ1) 128 (94Æ1) 23 (88Æ5) 148 (94Æ3) 23 (88Æ5) 100 (93Æ5) 180 (95Æ2) Aa 12 (6Æ6) 10 (7Æ5) 2 (4Æ0) 4 (8Æ9) 8 (5Æ9) 3 (11Æ5) 9 (5Æ7) 3 (11Æ5) 7 (6Æ5) 9 (4Æ8) aa0 000 00 00 0 0 Combined frequency AA 159 (86Æ9) 116 (87Æ2) 43 (86Æ0) 38 (84Æ4) 119 (87Æ5) 20 (76Æ9) 139 (88Æ5) 20 (76Æ9) 96 (89Æ7) 174 (92Æ1) Aa 24 (13Æ1) 17 (12Æ8) 7 (14Æ0) 7 (15Æ6) 17 (12Æ5) 6 (23Æ1) 18 (11Æ5) 6 (23Æ1) 11 (10Æ3) 14 (7Æ4) aa0 000 00 00 0 1(0Æ5)

Numbers in parentheses are percentages. AA, homozygous genotype for no R501X ⁄2282del4; Aa, heterozygous genotype for R501X ⁄2282del4; aa, homozygous genotype for R501X ⁄2282del4; aCOPSAC: from the COPSAC study.17 Children without atopic dermatitis.

Table 2 Comparison of allele frequencies in different subgroups with the v2 test

v2 OR (95% CI) P-value Comparison of individuals within the twin group (n) Twins with hand eczema (133) vs. twins without hand eczema (50) 0Æ047 0Æ9(0Æ35–2Æ32) 0Æ828 Twins with contact allergy (45) vs. twins without contact allergy (136) 0Æ274 1Æ3(0Æ50–3Æ34) 0Æ600 Twins with atopic dermatitis (26) vs. twins without atopic dermatitis (157) 2Æ640 2Æ3(0Æ82–6Æ53) 0Æ104 Comparison of individuals in twin subgroups and individuals in COPSAC (n) Twins with hand eczema (133) vs. COPSAC (189) 2Æ048 1Æ7(0Æ82–3Æ54) 0Æ152 Twins with contact allergy (45) vs. COPSAC (189) 2Æ477 2Æ1(0Æ82–5Æ60) 0Æ116 Twins with atopic dermatitis (26) vs. COPSAC (189) 5Æ945 3Æ5(1Æ21–9Æ99) 0Æ015

OR, odds ratio; CI, confidence interval; COPSAC, Copenhagen Prospective Study on Asthma in Childhood: from the COPSAC study.17 Children without atopic dermatitis.

number of studies.19 The epidermal defect caused by the vari- was small, limiting conclusions. Comparison with a subgroup ants could possibly be of aetiological importance in other skin of children without atopic dermatitis from a high-risk cohort diseases, characterized by a compromised skin barrier or of children born to mothers with asthma (COPSAC) also failed where a compromised skin barrier can be a trigger factor. to find any statistically significant difference. We investigated a possible association between the variant A defective skin barrier facilitates induction and elicitation alleles and the phenotype with hand eczema. In this popula- of contact allergy. Kligman found increased rates of sensitiza- tion-based, but selected group of individuals, with a high tion after pretreatment with an irritant or after combined prevalence of hand eczema and atopic dermatitis, we found an exposure.21 Also, pretreatment of the skin with an irritant or overall combined carrier frequency of 13%. This is higher combined exposure to an irritant and an allergen lowers the than other reported frequencies in the background population threshold for elicitation of contact allergy or increases the of between 8Æ8% and 9Æ6%.3,6,7 Two studies found consider- patch-test response.22,23 The impaired skin barrier caused by ably lower combined carrier frequencies in the control groups. the variant alleles could possibly increase the risk of contact Weidinger et al.20 reported 6Æ1% and Marenholz et al.4 found allergy. We did not find any statistically significant association only 5Æ1% carrying the variant alleles; however, in the latter between the variant alleles and contact allergy in the analysis study, a ‘hyper-normal’ control group was selected, as none of the total twin cohort or when compared with the COPSAC of them had any allergy. group. The twin subgroup with hand eczema did not exhibit a The allele frequencies in twin individuals with hand eczema higher mutation frequency than the group without hand or contact allergy were increased compared with the COPSAC eczema; however, the (control) group without hand eczema cohort, 13% and 16% vs. 7%, respectively. The negative

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 Filaggrin null alleles, hand eczema and contact allergy, A. Lerbaek et al. 1203 outcome of the association analyses may be due to insufficient A well-defined phenotype is essential in genetic association power, and using a larger population-based control group studies. In this study the diagnosis of hand eczema was based may yield a different result. Thus a possible association on a question on self-reported hand eczema. A similar ques- between the variant alleles and hand eczema or contact allergy tion has been validated and has shown high specificity, but cannot be entirely excluded. Three of the seven individ- less sensitivity.27 In the present context a high specificity is uals with a variant allele and contact allergy also had atopic preferable. Furthermore the diagnosis could be confirmed in dermatitis. 41% of cases due to visible signs of hand eczema at the exam- Contact allergy is diagnosed by means of patch testing. ination. Some limitations apply to this study. Errors in the patch-test The COPSAC subgroup of children without atopic dermatitis procedure cannot be entirely excluded, as participants was chosen as a secondary control group due to the limited applied and removed the patches themselves. Secondly, as number of twin individuals without hand eczema and due to the reading was done only on day 3, late reactions may availability. A more suitable control group would be an adult have been missed. It has been demonstrated that between group without hand eczema and contact allergy, matched on 3% and 8% of reactions become positive on day 6 or sex and atopy status. 7.24,25 Furthermore, individuals were only tested with 20 Individuals with hand eczema comprise a very heteroge- allergens, including the most frequent sensitizers. A negative neous group both regarding aetiology, severity and prognosis. test is obviously not a proof of absence of contact allergy. Finding a single candidate gene influencing all subtypes may However, these standard allergens have been shown to turn out to be a difficult task, even though the inflammatory detect 77–95% of all contact allergies in departments spe- processes may be similar. Two cytokine gene polymorphisms cialized in contact dermatitis.26 have been identified as being of importance for the develop- In the twin subgroup with atopic dermatitis the combined ment of allergic contact dermatitis: IL16-295 and TNFA- carrier frequency was 23%; however, this was not statistically 308.28,29 The latter was also present in increased frequency in significantly different from the frequency in individuals with- individuals with a low irritation threshold and may thus also out atopic dermatitis, even though the frequency in this group be a marker of increased risk of irritant contact dermatitis.30 was much smaller (11%). A likely explanation is a lack of In conclusion, no association between the filaggrin null power, since the group with atopic dermatitis comprised only alleles and hand eczema or contact allergy overall could be 26 individuals. We confirmed the association between atopic demonstrated; however, insufficient power is a consideration. dermatitis and the filaggrin variant alleles when comparing The association between atopic dermatitis and the filaggrin with the COPSAC subgroup (OR 3Æ5, P =0Æ015). In the sub- variant alleles was confirmed in the comparison of the individ- analysis including only individuals with hand eczema, the uals with atopic dermatitis and the COPSAC subgroup without comparison of subgroups with and without atopic dermatitis atopic dermatitis. In the subanalysis, including only individu- was borderline statistically significant. Thus the filaggrin null als with hand eczema, an almost statistically significant alleles could be a potential genetic marker for increased risk of increased frequency in individuals with atopic dermatitis was atopic hand eczema (in individuals with atopic dermatitis). It seen. This suggests that the filaggrin null alleles in future has previously been shown that the risk of hand eczema could be used as a potential marker for increased risk of hand increases with the severity of atopic dermatitis.8 This will need eczema in patients with atopic dermatitis. However, this further investigation in a study including individuals with hypothesis needs further investigation. atopic dermatitis but without hand eczema. In contrast to many of the previous hospital-based studies Acknowledgments on the association between atopic dermatitis and the filaggrin variant alleles, this study was population-based. Thus, our We thank Aage Vølund PhD for assistance with the statistical group with atopic dermatitis may represent a greater spectrum analyses and Simon Lee BSc for his assistance in the geno- of disease severity including phenotypes with milder symp- typing assays. The study was supported by research grants toms, compared with earlier studies. Weidinger et al. actually from Aage Bangs Foundation and King Christian the 10th demonstrated that the 2282del4 mutation and the combined Foundation. genotype were statistically significantly associated with a more 20 severe phenotype (SCORAD > 31). References The U.K. Working Party’s Diagnostic Criteria have been thoroughly validated, although mostly, but not entirely in 1 Candi E, Schmidt R, Melino G. The cornified envelope: a model of children. In this study the possibility of recall bias is consider- cell death in the skin. Nat Rev Mol Cell Biol 2005; 6:328–40. able. Possibly, mild cases with symptoms restricted to early 2 Smith FJ, Irvine AD, Terron-Kwiatkowski A et al. Loss-of-function mutations in the gene encoding filaggrin cause ichthyosis vulgaris. childhood have been forgotten and thus missed. Such a mis- Nat Genet 2006; 38:337–42. classification could change the results in both directions 3 Palmer CN, Irvine AD, Terron-Kwiatkowski A et al. Common loss- depending on whether or not those in question were carriers of-function variants of the epidermal barrier protein filaggrin are a of the variant alleles and whether or not they had hand major predisposing factor for atopic dermatitis. Nat Genet 2006; eczema. 38:441–6.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 1204 Filaggrin null alleles, hand eczema and contact allergy, A. Lerbaek et al.

4 Marenholz I, Nickel R, Ruschendorf F et al. Filaggrin loss-of-func- 17 Bisgaard H. The Copenhagen Prospective Study on Asthma in tion mutations predispose to phenotypes involved in the atopic Childhood (COPSAC): design, rationale, and baseline data from a march. J Allergy Clin Immunol 2006; 118:866–71. longitudinal birth cohort study. Ann Allergy Asthma Immunol 2004; 5 Weidinger S, Illig T, Baurecht H et al. Loss-of-function variations 93:381–9. within the filaggrin gene predispose for atopic dermatitis with 18 Halkjaer LB, Loland L, Buchvald FF et al. Development of atopic allergic sensitizations. J Allergy Clin Immunol 2006; 118:214–19. dermatitis during the first 3 years of life: the Copenhagen Prospec- 6 Stemmler S, Parwez Q, Petrasch-Parwez E et al. Two common loss- tive Study on Asthma in Childhood cohort study in high-risk of-function mutations within the filaggrin gene predispose for children. Arch Dermatol 2006; 142:561–6. early onset of atopic dermatitis. J Invest Dermatol 2006; 127:722–4. 19 Morar N, Willis-Owen SA, Moffatt MF et al. The genetics of atopic 7 Barker JN, Palmer CN, Zhao Y et al. Null mutations in the filaggrin dermatitis. J Allergy Clin Immunol 2006; 118:24–34. gene (FLG) determine major susceptibility to early-onset atopic 20 Weidinger S, Rodriguez E, Stahl C et al. Filaggrin mutations dermatitis that persists into adulthood. J Invest Dermatol 2006; strongly predispose to early-onset and extrinsic atopic dermatitis. 127:564–7. J Invest Dermatol 2006; 127:724–6. 8 Rystedt I. Hand eczema and long-term prognosis in atopic derma- 21 Kligman AM. The identification of contact allergens by human titis. Acta Derm Venereol Suppl (Stockh) 1985; 117:1–59. assay. II. Factors influencing the induction and measurement of 9 Bryld LE, Agner T, Kyvik KO et al. Hand eczema in twins: a ques- allergic contact dermatitis. J Invest Dermatol 1966; 47:375–92. tionnaire investigation. Br J Dermatol 2000; 142:298–305. 22 McLelland J, Shuster S, Matthews JN. ‘Irritants’ increase the 10 Lerbaek A, Kyvik KO, Mortensen J et al. Heritability of hand eczema response to an allergen in allergic contact dermatitis. Arch Dermatol is not explained by comorbidity with atopic dermatitis. J Invest 1991; 127:1016–19. Dermatol 2007; 127:1632–40. 23 Pedersen LK, Haslund P, Johansen JD et al. Influence of a detergent 11 Lerbaek A, Kyvik KO, Ravn H et al. Incidence of hand eczema in a on skin response to methyldibromo glutaronitrile in sensitized population-based twin cohort – genetic and environmental risk individuals. Contact Dermatitis 2004; 50:1–5. factors. Br J Dermatol 2007; 157:552–7. 24 Saino M, Rivara GP, Guarrera M. Reading patch tests on day 7. 12 Bryld LE, Hindsberger C, Kyvik KO et al. Risk factors influencing Contact Dermatitis 1995; 32:312–13. the development of hand eczema in a population-based twin sam- 25 Jonker MJ, Bruynzeel DP. The outcome of an additional patch-test ple. Br J Dermatol 2003; 149:1214–20. reading on days 6 or 7. Contact Dermatitis 2000; 42:330–5. 13 Susitaival P, Flyvholm MA, Meding B et al. Nordic Occupational 26 Menne T, Dooms-Goossens A, Wahlberg JE et al. How large a pro- Skin Questionnaire (NOSQ-2002): a new tool for surveying portion of contact sensitivities are diagnosed with the European occupational skin diseases and exposure. Contact Dermatitis 2003; standard series? Contact Dermatitis 1992; 26:201–2. 49:70–6. 27 Meding B, Barregard L. Validity of self-reports of hand eczema. 14 Williams HC, Burney PG, Hay RJ et al. The U.K. Working Party’s Contact Dermatitis 2001; 45:99–103. Diagnostic Criteria for Atopic Dermatitis. I. Derivation of a mini- 28 Reich K, Westphal G, Konig IR et al. Association of allergic contact mum set of discriminators for atopic dermatitis. Br J Dermatol 1994; dermatitis with a promoter polymorphism in the IL16 gene. 131:383–96. J Allergy Clin Immunol 2003; 112:1191–4. 15 Sandilands A, Terron-Kwiatkowski A, Hull PR et al. Comprehensive 29 Westphal GA, Schnuch A, Moessner R et al. Cytokine gene poly- analysis of the gene encoding filaggrin uncovers prevalent and rare morphisms in allergic contact dermatitis. Contact Dermatitis 2003; mutations in ichthyosis vulgaris and atopic eczema. Nat Genet 2007; 48:93–8. 39:650–4. 30 Allen MH, Wakelin SH, Holloway D et al. Association of TNFA gene 16 Frosch PJ, Menne T, Lepoittevin JP. Contact Dermatitis, 4th edn. polymorphism at position -308 with susceptibility to irritant con- Berlin: Springer Verlag, 2006. tact dermatitis. Immunogenetics 2000; 51:201–5.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1199–1204 DERMATOPATHOLOGY DOI 10.1111/j.1365-2133.2007.08239.x The applicability and prognostic value of the new TNM classification system for primary cutaneous lymphomas other than mycosis fungoides and Se´zary syndrome: results on a large cohort of primary cutaneous B-cell lymphomas and comparison with the system used by the Dutch Cutaneous Lymphoma Group N.J. Senff and R. Willemze Department of Dermatology, Leiden University Medical Centre, PO Box 9600, 2300 RC Leiden, the Netherlands

Summary

Correspondence Background Recently, a consensus proposal was published for a TNM classification Nancy Senff. system for all primary cutaneous lymphomas other than mycosis fungoides and E-mail: [email protected] Se´zary syndrome, meant to document extent of disease in a consistent manner. The applicability and the prognostic significance of this system have not been Accepted for publication 16 August 2007 investigated thus far. Objectives To test the applicability and prognostic relevance of the proposed TNM Key words classification system on a cohort of primary cutaneous B-cell lymphomas (CBCL). B cell, cutaneous lymphoma, disease extent, Methods The study group included 71 primary cutaneous marginal zone lympho- prognosis, TNM staging mas (PCMZL), 171 primary cutaneous follicle centre lymphomas (PCFCL) and 58 Conflicts of interest primary cutaneous diffuse large B-cell lymphomas, leg type (PCLBCL, LT). As None declared. only patients with primary cutaneous lymphoma were included (T1–3, N0M0), only the T-rating was scored. The results were compared with the scoring as applied by the Dutch Cutaneous Lymphoma Group. Results The system was easily applicable to all cases. In PCMZL and PCFCL no cor- relation was found between T-score and survival (5-year disease-specific survival: T1, 100% and 98%; T2, 94% and 93%; T3, 100% and 88%, respectively). In PCLBCL, LT there was a clear, although statistically not significant, association between increasing T-score and reduced survival (5-year disease-specific survival: T1, 75%; T2, 49%; T3, 0%; P =0Æ077). Comparing the TNM system with the Dutch Cutaneous Lymphoma Group system, there was a discrepancy in the classi- fication of 20 cases. Conclusions The new TNM system is a useful tool to document disease extent in patients with CBCL and provides prognostic information in the group of patients with PCLBCL, LT.

Recently, representatives of the International Society for Cuta- neous involvement, the N-classification describes lymph node neous Lymphomas and the Cutaneous Lymphoma Group of the involvement and the M-classification is used to describe if European Organization for Research and Treatment of Cancer extracutaneous nonlymph node disease is present. As, by defi- (EORTC) published a consensus proposal for a TNM classifica- nition, in all primary cutaneous lymphomas, extracutaneous tion system applicable to all primary cutaneous lymphomas disease (lymph node or visceral) is absent at the time of diag- other than mycosis fungoides and Se´zary syndrome.1 This pro- nosis, the N- and M-scores are used only to document disease posed TNM system is primarily meant to document extent of extent at the time of relapse or disease progression. A detailed disease in a consistent manner, thereby facilitating comparison description of the different categories is given in Table 1. of studies at different institutes. In the proposed system the Studies investigating the applicability of this TNM system T-classification reflects the extent ⁄distribution of primary cuta- have not been performed thus far. Moreover, whether this

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 1205 1206 TNM classification and prognostic value in CBCL, N.J. Senff and R. Willemze

Table 1 Description of different scores in the TNM classification1 addition, these results were compared with the scoring of dis- ease extent as used by the Dutch Cutaneous Lymphoma Classification Description Group. T T1 Solitary skin involvement categories: Materials and methods T1a: a solitary lesion £5 cm diameter T1b: a solitary lesion >5 cm diameter T2 Regional skin involvement categories: Patient selection Multiple lesions limited to one body region or The study group is a retrospective cohort of 300 patients with two contiguous body regions T2a: all disease encompassed in a £15 cm CBCL reclassified according to the criteria of the WHO–EORTC 3 diameter circular area classification. In all cases, gender, age at diagnosis, site, size T2b: all disease encompassed in a >15 and extent of skin lesions (see below), status at last follow-up and £30 cm diameter circular area and duration of follow-up were recorded as part of a recent T2c: all disease encompassed in a >30 cm study.2 In addition, in approximately 50% of cases clinical diameter circular area photographs were available for evaluation. This group com- T3 Generalized skin involvement: prised 71 PCMZL, 171 PCFCL and 58 PCLBCL, LT. T3a: multiple lesions involving two noncontiguous body regions T3b: multiple lesions involving three or Assessment of disease extent more body regions In all cases extent of skin lesions at the time of diagnosis was N 1 N0 No clinical or pathological lymph node scored using the proposed TNM system. As only patients involvement with primary cutaneous lymphoma were included (T1–3, N1 Involvement of one peripheral lymph N0M0), only the T-rating was scored (see Table 1 and node region that drains an area of current Fig. 1). In cases where no measurements were recorded in the or prior skin involvement database of the Dutch Cutaneous Lymphoma Group and clini- N2 Involvement of two or more peripheral lymph cal photographs were not available, measurements of skin node regions2 or involvement of any lesions were obtained from patients’ charts or, in cases where lymph node region that does not drain an area of current or prior skin involvement the patient had undergone radiotherapy, from the files of the N3 Involvement of central lymph nodes radiotherapy department. The results were compared with the scoring of the Dutch Cutaneous Lymphoma Group, used in M our previous study.2 In this Dutch Cutaneous Lymphoma M0 No evidence of extracutaneous nonlymph node disease Group system, extent of involved skin is defined as solitary M1 Extracutaneous nonlymph node disease present when it concerns a single tumour, as localized when the lesion consists of multiple skin lesions that can be irradiated within one radiation field, and as multifocal for separate lesions in adjacent body regions that cannot be irradiated TNM system provides prognostically relevant information for within one radiation field or for lesions involving multiple the different types of primary cutaneous lymphoma and thus nonadjacent body regions. In contrast to the TNM system, no may serve as an additional guide in the appropriate manage- further subdivision is made on the basis of the size of the skin ment of these lymphomas remains to be elucidated. lesions or the affected areas. In the present study we applied the proposed TNM system on a large cohort of 300 primary cutaneous B-cell lymphomas Statistical analysis (CBCL), recently reclassified according to the new World Health Organization (WHO)–EORTC classification.2,3 This Statistical calculations were performed using SPSS 12.0.1 (SPSS classification distinguishes three main types of CBCL: primary Inc., Chicago, IL, U.S.A.). Disease-specific survival was calcu- cutaneous marginal zone lymphoma (PCMZL) and primary lated from the date of diagnosis until death from lymphoma cutaneous follicle centre lymphoma (PCFCL), both indolent or last follow-up without event. Survival curves were esti- types of CBCL, and primary cutaneous diffuse large B-cell mated using the method of Kaplan and Meier and statistical lymphoma, leg type (PCLBCL, LT), which represents a more comparison between curves was done by log-rank testing. aggressive type of CBCL. As these groups have been redefined 4,5 as compared with the formerly used classifications, we have Results summarized the main features of these three groups, as found in a recent study,2 in Table 2. The median follow-up for the whole group was 56 months The aims of this retrospective cohort analysis were to test (range 2–336). The results of scoring extent of disease by the clinical applicability of the proposed TNM system and to T-score and by the Dutch Cutaneous Lymphoma Group evaluate its prognostic relevance for the group of CBCL. In system, as well as the corresponding 5-year disease-specific

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 TNM classification and prognostic value in CBCL, N.J. Senff and R. Willemze 1207

Table 2 Clinical characteristics of the three main groups of primary cutaneous B-cell lymphoma2

PCMZL PCFCL PCLBCL, LT Median age, years (range) 53 (23–87) 58 (21–89) 78 (42–92) Male ⁄female ratio 2Æ11Æ80Æ5 Clinical presentation Solitary or multiple papules, Solitary or grouped tumours or Solitary or multiple nodules and ⁄or plaques or nodules preferentially plaques preferentially on the tumours most often on the leg(s) on the trunk (55%) or head (44%) or the trunk (54%); (88%); lesions at sites other than extremities (67%) lesions on the leg(s) uncommon the leg(s) uncommon (6%) Relapse rate 57% 30% 69% Extracutaneous progression 8Æ5% 10Æ5% 47% Prognosis 5-year DSS 98% 95% 50% 5-year OS 94% 87% 37% Preferred treatment Nonaggressive local treatment Radiotherapy Multiagent chemotherapy or radiotherapy

PCMZL, primary cutaneous marginal zone lymphoma; PCFCL, primary cutaneous follicle centre lymphoma; PCLBCL, LT, primary cutaneous diffuse large B-cell lymphoma, leg type; DSS, disease-specific survival; OS, overall survival.

Fig 1. Body regions as defined in the proposed TNM system for the designation of T-classification. Left and right extremities are assessed as separate body regions.1 survival rates, are summarized in Table 3. Because of the clear multifocal according to the Dutch Cutaneous Lymphoma descriptions of the main T-score groups and the different sub- Group system, but as T2 using the TNM system. Representa- groups (see Table 1), combined with the strictly defined body tive examples of the different stages are presented in Figure 2. regions (see Fig. 1), this system could be easily applied to all The 5-year disease-specific survival for T1, T2 and T3 were cases. 100%, 94% and 100%, respectively, and 100%, 92% and The study included 71 patients with PCMZL (48 men and 100% for solitary, localized and multifocal disease according 23 women) with a median age of 53 years (range 23–87). to the Dutch Cutaneous Lymphoma Group system. These Using the TNM system most patients were classified as T3 results clearly show that extent of disease, either by TNM or (n = 28; 39%), followed by T2 (n = 25; 35%) and T1 Dutch Cutaneous Lymphoma Group system, has no prognostic (n = 18; 25%). According to the Dutch Cutaneous Lymphoma significance in this group. Group system 36 patients (51%) had multifocal disease, 17 Patients with PCFCL (n = 171; 109 men and 62 women) patients (24%) had localized disease, while 18 patients (25%) had a median age of 58 years (range 21–89). Using the TNM had presented with a solitary lesion. Eight cases with multiple system, most patients presented with stage T1 (n = 68; 40%) separate lesions at adjacent body regions were classified as or T2 (n = 88; 51%). Only 15 patients (9%) were classified

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 1208 N lsicto n rgotcvlei BL ..SnfadR Willemze R. and Senff N.J. CBCL, in value prognostic and classification TNM ora Compilation Journal Table 3 Scoring of disease extent in 300 cutaneous B-cell lymphomas according to the TNM system and the Dutch Cutaneous Lymphoma Group (DCLG) system

PCMZL (n = 71) PCFCL (n = 171) PCLBCL, LT (n = 58)

TNM no. of 5-year DCLG no. of 5-year TNM no. of 5-year DCLG no. of 5-year TNM no. of 5-year DCLG no. of 5-year patients (%) DSS patients (%) DSS patients (%) DSS patients (%) DSS patients (%) DSS patients (%) DSS

07BiihAscaino Dermatologists of Association British 2007 Solitary Solitary Solitary

T1 18 (25%) 100% 18 (25%) 100% 68 (40%) 98% 68 (40%) 98% 14 (24%) 75% 14 (24%) 75% T1a 15 (21%) 100% 53 (31%) 98% 8 (14%) 86% T1b 3 (4%) 100% 15 (9%) 100% 6 (10%) 56% Localized Localized Localized

T2 25 (35%) 94% 17 (24%) 92% 88 (51%) 93% 76 (44%) 95% 33 (57%) 49% 33 (57%) 49% T2a 7 (10%) 75% 44 (26%) 91% 18 (31%) 67% T2b 7 (10%) 100% 27 (16%) 96% 9 (16%) 48% T2c 11 (15%) 100% 17 (10%) 92% 6 (10%) 0% Multifocal Multifocal Multifocal

T3 28 (39%) 100% 36 (51%) 100% 15 (9%) 88% 27 (16%) 85% 11 (19%) 0% 11 (19%) 0%

• T3a 3 (4%) 100% 4 (2%) 100% 5 (9%) 20% rts ora fDermatology of Journal British T3b 25 (35%) 100% 11 (6%) 83% 6 (10%) 0%

PCMZL, primary cutaneous marginal zone lymphoma; PCFCL, primary cutaneous follicle centre lymphoma; PCLBCL, LT, primary cutaneous diffuse large B-cell lymphoma, leg type; DSS, disease- specific survival. 2007 157, 07TeAuthors The 2007 pp1205–1211 TNM classification and prognostic value in CBCL, N.J. Senff and R. Willemze 1209

(a) (b)

Fig 2. Examples of different T-scores in primary cutaneous marginal zone lymphoma. (a) T2a, Dutch Cutaneous Lymphoma Group: localized; (b) T2c, Dutch Cutaneous Lymphoma Group: multifocal. as T3 at the time of diagnosis (see Table 3). According to the The 5-year disease-specific survival for T1, T2 and T3 was Dutch Cutaneous Lymphoma Group system 68 patients (40%) 98%, 93% and 88% and 98%, 95% and 85% for solitary, had presented with a solitary skin lesion, 76 patients (44%) localized and multifocal according to the Dutch Cutaneous with localized skin lesions and 27 patients (16%) with multi- Lymphoma Group system, respectively. Although there was a focal skin lesions. Twelve cases with multiple separate lesions tendency towards reduced survival with increasing T-score, at adjacent body regions were classified as multifocal accord- these differences were not significant (P =0Æ560). Moreover, ing to the Dutch Cutaneous Lymphoma Group system, but as no significant differences were found between subgroups T2 using the TNM system. Representative examples of the dif- within the different T-categories (T1a vs. T1b; T2a vs. T2b vs. ferent stages are presented in Figure 3. T2c; T3a vs. T3b; see Table 3).

(a) (b)

(c) (d)

Fig 3. Examples of different T-scores in primary cutaneous follicle centre lymphoma. (a) T1b, Dutch Cutaneous Lymphoma Group: solitary; (b) T2b, Dutch Cutaneous Lymphoma Group: localized; (c, d) T3b, Dutch Cutaneous Lymphoma Group: multifocal.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 1210 TNM classification and prognostic value in CBCL, N.J. Senff and R. Willemze

(a) (b)

(c)

Fig 4. Examples of different T-scores in primary cutaneous diffuse large B-cell lymphoma, leg type. (a) T1a; Dutch Cutaneous Lymphoma Group: solitary; (b) T2a, Dutch Cutaneous Lymphoma Group: localized; (c) T3a, Dutch Cutaneous Lymphoma Group: multifocal.

Patients with PCLBCL, LT (n = 58; 20 men and 38 women) the first time how this system is applied to a large group of had a median age of 78 years (range 42–92). In this group primary cutaneous lymphomas. the numbers according to the T-score corresponded entirely Our results show that the TNM system can easily be applied with the numbers in the extent categories as used by the to the three commonest groups of CBCL. In both indolent Dutch Cutaneous Lymphoma Group: T1 ⁄solitary, n =14 entities, PCMZL and PCFCL, the system does not appear to (24%); T2 ⁄localized, n = 33 (57%); T3 ⁄multifocal, n =11 have prognostic significance, but in PCLBCL, LT, the group (19%) (see Fig. 4). Although statistically not significant with an intermediate prognosis, increasing T-score seems to (P =0Æ077), a clear correlation was seen between extent and be associated with worse survival. In addition, subdivision survival (5-year disease-specific survival of 75%, 49% and 0%, based on the size of the skin lesion or the affected body area respectively). Moreover, the classification into subgroups or, in the case of T3 classification, the number of involved within the main T-scores provided additional prognostic value. body regions, provides additional prognostic information in For instance, patients in the T2 group showed a decreased sur- this group of PCLBCL, LT. vival with increasing size of the affected area (5-year disease- Consistently, in two recent studies on PCLBCL, LT, patients specific survival for involved area of £15 cm, 15–30 cm and presenting with a solitary tumour had a better prognosis than >30 cm was 67%, 48% and 0%, respectively; see Table 3). patients presenting with multiple tumours on one or both legs.6,7 However, in a study of 40 PCLBCL, LT by Kodama 8 Discussion et al. no difference in survival was found between patients with solitary or multiple tumours. Zinzani et al.7 reported This study investigates the applicability and prognostic signifi- a significantly higher overall survival for PCMZL ⁄PCFCL pati- cance of the newly proposed TNM system and describes for ents with a single skin lesion compared with those with

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 TNM classification and prognostic value in CBCL, N.J. Senff and R. Willemze 1211 regional ⁄disseminated disease. As the authors do not describe References this relation for PCMZL and PCFCL separately, comparison 1 Kim YH, Willemze R, Pimpinelli N et al. TNM classification system with the results of the present study is impossible. for primary cutaneous lymphomas other than mycosis fungoides Comparing the TNM system with the Dutch Cutaneous and Se´zary syndrome: a proposal of the International Society for Lymphoma Group system, there was a discrepancy in the Cutaneous Lymphomas (ISCL) and the Cutaneous Lymphoma Task classification of 20 cases (eight PCMZL and 12 PCFCL) with Force of the European Organization of Research and Treatment of multiple separate lesions at two contiguous body regions Cancer (EORTC). Blood 2007; 110:479–84. (see Fig. 2b). As there is no anatomical or biological relation 2 Senff NJ, Hoefnagel JJ, Jansen PM et al. Reclassification of 300 pri- between such separate lesions, in the Dutch Cutaneous Lym- mary cutaneous B-cell lymphomas according to the new WHO– EORTC classification for cutaneous lymphomas: comparison with phoma Group system such lesions are classified as multifocal previous classifications and identification of prognostic markers. or generalized disease, and not as regional disease. However, J Clin Oncol 2007; 25:1581–7. as the T3 category in the TNM system is defined as involve- 3 Willemze R, Jaffe ES, Burg G et al. WHO–EORTC classification for ment of two noncontiguous body regions or three or more cutaneous lymphomas. Blood 2005; 105:3768–85. body regions, such cases are not classified as T3, but as T2, 4 Jaffe ES, Harris NL, Stein H, Vardiman JW (eds). World Health Organiza- which denotes regional skin involvement. Classification of tion Classification of Tumours: Pathology and Genetics of Tumours of Haematopoietic such cases as T2 is of minor clinical importance in PCMZL and Lymphoid Tissues. Lyon: IARC Press, 2001. 5 Willemze R, Kerl H, Sterry W et al. EORTC classification for primary and PCFCL, as both groups represent indolent types of CBCL, cutaneous lymphomas: a proposal from the Cutaneous Lymphoma which can be treated with nonaggressive therapies, irrespective Study Group of the European Organization for Research and Treat- of the extent of skin lesions. However, in other types of ment of Cancer. Blood 1997; 90:354–71. lymphoma, classifying separate lesions at adjacent body sites 6 Grange F, Bekkenk MW, Wechsler J et al. Prognostic factors in pri- as T2 might have important therapeutic consequences. mary cutaneous large B-cell lymphomas: a European multicenter In conclusion, our results show that the proposed TNM study. J Clin Oncol 2001; 19:3602–10. system provides the clinician with a useful tool to document the 7 Zinzani PL, Quaglino P, Pimpinelli N et al. Prognostic factors in pri- mary cutaneous B-cell lymphoma: the Italian Study Group for Cuta- disease extent in patients with CBCL in a consistent man- neous Lymphomas. J Clin Oncol 2006; 24:1376–82. ner. In cases of PCLBCL, LT it also provides prognostic informa- 8 Kodama K, Massone C, Chott A et al. Primary cutaneous large tion. The applicability of the definitions for T2c and T3a ⁄b B-cell lymphomas: clinicopathologic features, classification, and should be evaluated in cohorts of other types of primary cutane- prognostic factors in a large series of patients. Blood 2005; ous lymphomas, to avoid undesirable therapeutic consequences. 106:2491–7.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1205–1211 DERMATOPATHOLOGY DOI 10.1111/j.1365-2133.2007.08246.x Involvement of E-cadherin, b-catenin, Cdc42 and CXCR4 in the progression and prognosis of cutaneous melanoma M.G. Tucci, G. Lucarini,* D. Brancorsini, A. Zizzi,* A. Pugnaloni,* A. Giacchetti, G. Ricotti and G. Biagini* U.O. Dermatologia, INRCA-IRCCS, Via della Montagnola 164, Ancona, Italy *Istologia-Dipartimento di Patologia Molecolare e Terapie Innovative, Universita` Politecnica delle Marche, Via Tronto 10 ⁄A, 60020 Torrette, Ancona, Italy Dipartimento di Neuroscienze, Istituto di Anatomia Patologica, Azienda Ospedaliero-Universitaria, Ospedali Riuniti Lancisi-Salesi-Umberto I, Via Conca 71, 60020 Torrette, Ancona, Italy

Summary

Correspondence Background A key event in cancer metastasis is the migration of tumour cells from Guendalina Lucarini. their original location to a secondary site. The development of melanoma may be E-mail: [email protected] viewed as a consequence of the disruption of homeostatic mechanisms in the skin of the original site. Accepted for publication 25 July 2007 Objectives To investigate whether dysregulation of cell motility (Cdc42 expression), escaping the control of cell–cell and cell–matrix interactions (E-cadherin, b-cate- Key words nin expression), enhances melanoma progression, and whether chemokine recep- b-catenin, Cdc42, cutaneous melanoma, tors (CXCR4) mediate cell migration and activation during invasion and CXCR4, E-cadherin metastasis development. b Conflicts of interest Methods The immunohistochemical expression of Cdc42, E-cadherin, -catenin None declared. and CXCR4 was investigated in 30 patients with surgically treated nodular mela- noma, 18 alive and disease free and 12 with a fatal outcome due to metastatic disease. Results E-cadherin expression was significantly reduced (P <0Æ05) and cytoplasmic b-catenin was increased in the patients who had died compared with disease-free individuals, while membrane expression of b-catenin was similar in the two groups. Patients with fatal outcome had increased Cdc42 (P <0Æ01) and CXCR4 (P <0Æ05). In this group a positive correlation was found between melanocytic Cdc42 expression and Breslow thickness (r =0Æ598, P <0Æ05) and between CXCR4 expression and Breslow thickness (r =0Æ583, P <0Æ05). Conclusions Findings suggest that primary cutaneous melanoma with a high Breslow thickness is characterized by tumour cells with high motility and inva- sion ability, in line with the hypothesis that low E-cadherin levels and over- expression of Cdc42 and CXCR4 could be prognostic markers of poor outcome.

Cancer progression is a multistep process in which tumour primary melanoma cells do have metastatic potential. Active cells migrate from a localized primary tumour mass to an tumour cell motility, which favours invasion and metastasis, is invasive secondary metastasis. We investigated whether dys- thus involved in the first two phases.1,2 The final step is regulation of cell motility (Cdc42 expression), escaping the metastasis to distant organ sites. control of cell–cell and cell–matrix interactions (E-cadherin Melanocyte homeostasis in the skin is a key factor in this and b-catenin), enhances melanoma progression, and whether process; cancer is generally viewed as a result of the disrup- chemokine receptors (CXCR4) mediate cell migration and acti- tion of homeostatic mechanisms, which determine whether vation during invasion and metastasis development. The clini- cells remain quiescent, proliferate, differentiate, or die. The cal and pathological features of melanoma suggest that the development of melanoma may be seen as a consequence of disease often progresses in three distinct growth phases: radial, this disruption. Adhesion molecules play a crucial part. vertical and metastatic. In the radial phase, individual tumour There is compelling evidence to suggest that cadherins play cells from lesions invade the dermis but have no capacity to a major role in epithelial cell–cell adhesion.3 E-cadherin, a metastasize; in the vertical phase the large clusters of invading transmembrane glycoprotein localized mainly in the adherens

2007 The Authors 1212 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1212–1216 E-cadherin, b-catenin, Cdc42 and CXCR4 in melanoma, M.G. Tucci et al. 1213 junction,4 is the major adhesion molecule between intraepi- and duration of follow-up. Patients had been treated according dermal melanocytes and keratinocytes. The proliferation and to World Health Organization guidelines. None had a family differentiation of melanocytes that have lost E-cadherin is not history of melanoma. The median age of the 30 patients was regulated by keratinocytes.5,6 The E-cadherin cytoplasmic 54 years (range 24–70). Lesion sites were head and neck domain directly binds b-catenin, a multifunctional protein that (n = 2), trunk (n = 10), arms (n = 8) and legs (n = 10). controls a number of cell activities in both the membrane and Established criteria were used for melanoma classification.14,15 the nucleus.7 Loss of b-catenin membrane expression has been There were nine patients with superficial spreading melanoma described in several tumour types, and although the under- and NM and 21 patients with NM. Melanomas were divided lying molecular mechanisms are unclear, it is believed to be into those from patients who were still alive at the time of the associated with impaired catenin–cadherin interactions and to study and those from patients who had died of the disease. result in increased cell invasiveness.8 The first group included 18 subjects (median follow-up The recognition that invasiveness is a separable and inde- 5 years, range 4–8) without local recurrences, lymph node pendent cell property from growth has led investigators to and ⁄or distant metastases after wide surgical excision of the focus on the key regulatory events that favour tumour inva- primary melanoma. The Breslow thickness was 1–2 mm in siveness.9 Dysregulation of motility has an important role in five patients, 2–4 mm in nine and > 4 mm in four.15 Three promoting cell invasion and the development of metastases. patients had Clark level III melanomas and 15 were Clark level Cdc42, a protein of the Rho family, regulates cytoskeletal IV. Among the 12 patients in the second group (median rearrangement – specifically the formation of filopodia – and follow-up 2 years, range 1–4) the Breslow thickness was if activated binds to and activates p21 kinase, which in turn 1–2 mm in four, 2–4 mm in six and > 4 mm in two. One targets cytoskeletal movements. Jung et al.10 demonstrated that patient had Clark level III, six had Clark level IV and five had Cdc42 mediates the autotaxin-induced motility of A2058 mel- Clark level V melanomas. After surgical treatment of recur- anoma cells. In a recent study, Goteri et al.11 demonstrated a rences and in-transit or regional lymph node metastases, these role for Cdc42 in human cell migration, finding it to be patients developed distant metastases (four liver, five lung, highly expressed in the eutopic endometrium of women with and three liver and lung) that were treated with palliative ovarian endometrioma and thus contributing to the develop- chemotherapy, and eventually died. ment of ovarian endometriosis. Cancer cells with metastatic potential also express specific Immunohistochemistry sets of chemokine receptors that govern localization and tissue-specific migration.12 High levels of the chemokine The diagnoses of melanoma were verified on haematoxylin receptors CXCR4, CCR7 and CCR10 were observed by Muller and eosin sections. For immunohistochemistry, conventional et al.13 in melanoma cell lines. Moreover, Murakami et al.12 6-lm histological sections were cut from the blocks with a found that CXCR4 overexpression in B16 melanoma cells microtome and mounted on slides pretreated with poly- enhanced cell motility and migration and was thus an impor- L-lysine (Sigma Chemicals, St Louis, MO, U.S.A.). The slides tant factor in invasion and metastasis. were then deparaffinized and rehydrated in a gradient of etha- In this immunohistochemical study we investigated the nol and xylene. Sections were incubated with TUF solution involvement of E-cadherin, b-catenin and Cdc42 in favouring (Target Unmasking Fluid; Alexis Corporation, La¨ufelfingen, the escape of melanoma cells from the epidermal keratinocyte Switzerland) at 90 C for 10 min to unmask the antigens. microenvironment and the expression and possible prognostic After washes in H2O and then in Tris buffer solution, value of CXCR4 in lesions from two groups of patients with they were incubated with the following monoclonal anti- surgically treated primary nodular melanoma, 18 of whom bodies: anti-Cdc42 (dilution 1 : 100); anti-E-cadherin (dilu- were disease free and 12 of whom had died of metastatic tion 1 : 100), anti-b-catenin (dilution 1 : 100) (all from disease. Santa Cruz Biotechnology, Inc., Santa Cruz, CA, U.S.A.); and anti-CXCR4 (dilution 1 : 100; R&D Systems, Minneapolis, Materials and methods MN, U.S.A.). Incubation was carried out overnight in a humidified atmosphere at 4 C. The antibodies were linked to an alkaline phosphatase–antialkaline phosphatase-catalysed Clinical data reaction and washed with phosphate-buffered saline between The records of all patients diagnosed with primary malignant steps. For antibody visualization, sections were incubated with melanoma of the skin from 1998 to 2002 were retrieved from Fast Red substrate (Dakocytomation, Glostrup, Denmark), the archives of the Dermatology Unit of INRCA, Ancona which yields a red end-product at the site of the target (Italy). Only cases diagnosed as malignant melanoma of the antigen. Finally, sections were counterstained with Mayer’s nodular type (NM) with Clark14 level of invasion III or IV fol- haematoxylin, dehydrated and mounted in glycerol-gelatin lowed up until July 2006 were included. There were 30 (Dakocytomation). Each series included appropriate negative patients, 12 men and 18 women. Data included age at diagno- control sections without the primary antibody. sis, sex, tumour site, surgical treatment, recurrences, manage- Immunostaining was evaluated by light microscopy by ment of lymph node and distant metastases, final outcome scanning the entire melanocytic lesion: all positive epidermal

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1212–1216 1214 E-cadherin, b-catenin, Cdc42 and CXCR4 in melanoma, M.G. Tucci et al. and dermal melanocytes were counted and their number was Discussion expressed as a proportion of positive cells. Melanoma progression is a multistep process in which cell–cell and cell–matrix interactions, cell motility and chemoattraction Statistical analysis play an important role in determining organ-selective meta- Expression of Cdc42, E-cadherin, b-catenin and CXCR4 was stasis.1 compared in melanoma cells and correlated with the following Melanoma cells probably escape the control of keratinocytes parameters: Breslow thickness measured on histological slides, through (i) downregulation of receptors important for com- presence of metastases and survival. Data were summarized by munication with keratinocytes, (ii) upregulation of receptors mean and range for continuous variables. Differences between and signalling molecules important for interactions between means in different categories were analysed by Student’s t-test. melanoma cells, and (iii) loss of anchorage to the basement Correlations between continuous variables were analysed using membrane due to an altered expression of cell–matrix adhe- Spearman’s rank correlation test. sion molecules. Loss of E-cadherin, a protein that mediates cell adhesion, has been implicated in tumour progression and Results metastasis. E-cadherin is the major melanocyte adhesion mole- cule to interact with keratinocytes.5 Its cytoplasmic domain The results (means ± SD) are summarized in Table 1. E-cad- binds b-catenin directly, which is essential for the establish- herin immunoreactivity was observed in melanoma cells in ment of cell–cell adhesion. both the epidermis and dermis, mainly as membrane staining. These data reveal a significant loss of E-cadherin expression Melanocytes from patients with metastases had significantly in metastatic melanoma, in line with Sanders et al.,3 who reduced E-cadherin expression (Fig. 1a) compared with those described reduced membrane E-cadherin in the vertical without disease (Fig. 1b) (25Æ83 ± 9Æ17 vs. 35Æ00 ± 11Æ83; growth phase after the establishment of metastatic disease, as P <0Æ05, Student’s t-test). well as aberrant production of b-catenin in cell cytoplasm in Cytoplasmic b-catenin was increased (Fig. 1c) (58Æ83 ± metastatic malignant melanoma. However, our data also docu- 22Æ29) in patients who went on to die of the disease ment a decrease in b-catenin membrane expression in the compared with those who were alive and disease free same patients, even though the reduction was not significant. (Fig. 1d) (45Æ0±31Æ30), but the difference was not signifi- Pecina-Slaus et al.16 found that b-catenin expression showed cant. Membrane b-catenin expression was similar in the two a significant correlation with Clark level, as Clark level IV and groups. V patients had significantly less b-catenin protein than level II Cdc42 expression was diffusely higher in melanoma cells and III patients. The reason that our data are not statistically from deceased patients (Fig. 1e) (60Æ00 ± 10Æ95) than in significant may be that all our patients had high Clark levels. those from disease-free patients (Fig. 1f) (31Æ83 ± 16Æ69) These findings support a role for the disruption of melano- (P <0Æ01, Student’s t-test). Spearman’s test showed a positive cyte–keratinocyte communication in metastatic melanoma.17 correlation between melanocytic Cdc42 expression and Bre- Alterations in membrane expression of E-cadherin and slow thickness (r =0Æ598, P <0Æ05) in patients with meta- b-catenin thus appear to be crucial to progression, allowing static disease (Fig. 2). melanoma cells to escape the control of keratinocytes. In fact, Cytoplasmic CXCR4 staining was more intense in melano- melanoma cells expressing E-cadherin introduced into skin mas from patients with metastases (Fig. 1g) (79Æ17 ± 10Æ21) reconstructs remain confined to the epidermis and show very than in those from disease-free individuals (Fig.1h) (60Æ00 ± little ability to invade the dermis.18 14Æ49), and the difference was significant (P <0Æ05, Student’s El-Bahrawy et al.19 demonstrated that b-catenin is aberrantly t-test). produced in the cell cytoplasm in metastatic malignant mela- A direct correlation was also found in patients with meta- noma, compromising adhesive function. Normal adhesive static disease (Fig. 3) between melanocytic CXCR4 expression linkages (i.e. E-cadherin ⁄b-catenin complexes) maintain epi- and Breslow thickness (r =0Æ583, P <0Æ05; Spearman’s test). thelial cell polarity, which is critical for segregation of apically

Table 1 Immunohistochemical expression of Cdc42, E-cadherin, b-catenin and CXCR4 in melanoma patients with and without metastases

Cdc42 E-cadherin b-catenin CXCR4 Patients Membrane Cytoplasm Total With metastases 60Æ00 ± 10Æ95 25Æ83 ± 9Æ17 45Æ83 ± 20Æ10 58Æ83 ± 22Æ29 52Æ0±12Æ81 79Æ17 ± 10Æ21 Without metastases 31Æ83 ± 16Æ69 35Æ00 ± 11Æ83 47Æ50 ± 39Æ94 45Æ0±31Æ30 42Æ50 ± 28Æ59 60Æ00 ± 14Æ49 P-value 0Æ006 0Æ044 0Æ929 0Æ399 0Æ457 0Æ024

Results are expressed as percentage of positive cells among total cells counted. Values are means ± SD. Differences were analysed by Student’s t-test. Significant P-values are shown in bold.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1212–1216 E-cadherin, b-catenin, Cdc42 and CXCR4 in melanoma, M.G. Tucci et al. 1215

(a) (b) (c)

(d) (e) (f)

(g) (h)

Fig 1. (a) Faint immunohistochemical expression of E-cadherin in melanocytes of patients with metastases; (b) evident E-cadherin membrane and cytoplasmic expression in patients without disease; (c) increased cytoplasmic expression of b-catenin in patients with metastases compared with disease-free patients (d); (e) strong Cdc42 expression in patients with metastases; (f) moderate expression of Cdc42 in melanocytes of patients without metastases; (g) stronger CXCR4 immunostaining in patients with metastases compared with patients without disease (h). Immunoperoxidase staining; original magnification: (c, h) ·100; (a, b, d, g) ·200; (e, f) ·400.

Fig 2. Spearman’s test showing a positive correlation between Fig 3. The Spearman test showed a positive correlation between melanocytic Cdc42 expression and Breslow thickness in patients with melanocytic CXCR4 expression and Breslow thickness in patients with metastases (r =0Æ598, P <0Æ05). metastases (r =0Æ583, P <0Æ05).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1212–1216 1216 E-cadherin, b-catenin, Cdc42 and CXCR4 in melanoma, M.G. Tucci et al. secreted growth factors and basolateral receptors. It has been 3 Sanders DS, Blessing K, Hassan GA et al. Alterations in cadherin and postulated that loss of segregation leads to increased cell catenin expression during the biological progression of melano- motility, favouring invasion.9 Cdc42 and CXCR4 play a key cytic tumours. Mol Pathol 1999; 52:151–7. 4 Silye R, Karayiannakis AJ, Syrigos KN et al. E-cadherin ⁄catenin complex role among motility-inducing factors.10,13 Cdc42 expression in benign and malignant melanocytic lesions. JPathol1998; 186:350–5. was significantly increased in melanomas from patients who 5 Tang A, Eller MS, Hara M et al. E-cadherin is the major mediator of died of the disease compared with the disease-free group. human melanocyte adhesion to keratinocytes in vitro. J Cell Sci 1994; Cdc42 controls the formation of filopodia participating in 107:983–92. chemoattractant-induced motility, suggesting an important role 6 Hsu MY, Meier FE, Nesbit M et al. E-cadherin expression in mela- for this protein in boosting cell motility.20 Our data confirm noma cells restores keratinocyte-mediated growth control and those obtained in vivo by Jung et al.,10 who showed that Cdc42 down-regulates expression of invasion-related adhesion receptors. Am J Pathol 2000; 156:1515–25. mediates the autotaxin-induced motility of A2058 melanoma 7 Peifer M, Polakis P. Wnt signaling in oncogenesis and embryo- cells. As regards CXCR4 expression, our findings agree with genesis: a look outside the nucleus. Science 2000; 287:1606–9. 21 those of Scala et al., who were the first to provide evidence 8 Oyama T, Kanai Y, Ochiai A et al. A truncated b-catenin disrupts that the level of CXCR4 expression might be a prognostic mar- the interaction between E-cadherin and a-catenin: a cause of loss ker in primary cutaneous malignant melanoma. Overexpression of intercellular adhesiveness in human cancer cell lines. Cancer Res of CXCR4 dramatically enhanced the metastatic accumulation 1994; 54:6282–7. of B16 melanoma cells in mouse lung.12 Expression of func- 9 Kassis J, Lauffenburger DA, Turner T, Wells A. Tumor invasion as dysregulated cell motility. Semin Cancer Biol 2001; 11:105–17. tional CXCR4 receptors in melanoma cells indicates that it 10 Jung ID, Lee J, Yun SY et al. Cdc42 and Rac1 are necessary for might contribute to cell motility during invasion as well as to autotaxin-induced tumor cell motility in A2058 melanoma cells. 22 the regulation of cell proliferation and survival. CXCR4 was FEBS Lett 2002; 532:351–6. significantly increased in primary cutaneous melanomas from 11 Goteri G, Ciavattini A, Lucarini G et al. Expression of motility-related patients who developed distant metastases compared with dis- molecule Cdc42 in endometrial tissue of women with adenomyosis ease-free patients, but was not significantly different among and ovarian endometriomata. Fertil Steril 2006; 86:559–65. the former (five patients with lung metastases, four with liver 12 Murakami T, Maki W, Cardones AR et al. Expression of CXC chemokine receptor-4 enhances the pulmonary metastatic potential metastases and three with liver and lung involvement). 21 of murine B16 melanoma cells. Cancer Res 2002; 62:7328–34. Scala et al. described CXCR4 expression in almost 60% of 13 Muller A, Homey B, Soto H et al. Involvement of chemokine recep- melanoma-involved lymph nodes, and suggested that CXCR4 tors in breast cancer metastasis. Nature 2001; 410:50–6. expression in tumour cells correlates with an unfavourable 14 Clark WH Jr, From L, Bernardino EA, Mihm MC. The histogenesis prognosis. Murakami et al.,23 using a mouse model of mela- and biologic behavior of primary human malignant melanomas of noma, suggested that CXCR4 plays a role mainly in lung the skin. Cancer Res 1969; 29:705–27. metastases, while others24,25 showed the involvement of other 15 Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg 1970; 172:902–8. chemokine receptors such as CCR7 and CXCR3 in lymph node 16 Pecina-Slaus N, Zigmund M, Kusec V et al. E-cadherin and b-catenin metastasis. expression patterns in malignant melanoma assessed by image anal- We found a significant positive correlation between Cdc42 ysis. J Cutan Pathol 2007; 34:239–46. expression and Breslow thickness and between CXCR4 expres- 17 Haass NK, Herlyn M. Normal human melanocyte homeostasis as a sion and Breslow thickness in primary melanomas from paradigm for understanding melanoma. J Investig Dermatol Symp Proc patients with metastases. The findings confirm that primary 2005; 10:153–63. cutaneous melanoma with a high Breslow thickness is charac- 18 Hsu M, Andl T, Li G et al. Cadherin repertoire determines partner- specific gap junctional communication during melanoma progres- terized by tumour cells with high motility and invasion capac- sion. J Cell Sci 2000; 113:1535–42. ity. During invasion and dissemination, tumour cell migration 19 El-Bahrawy M, El-Masry N, Alison M et al. Expression of b-catenin through tissues involves dynamic regulation of cell adhesion in basal cell carcinoma. Br J Dermatol 2003; 148:964–70. associated with morphological alterations that must include 20 Eisenmann KM, McCarthy JB, Simpson MA et al. Melanoma chon- changes in adhesion molecules, the organization of the actin droitin sulphate proteoglycan regulates cell spreading through cas cytoskeleton and chemoattracting chemokines. Cdc42, Ack-1 and p130 . Nat Cell Biol 1999; 1:507–13. Although the sample size does not permit us to draw any 21 Scala S, Ottaviano A, Ascierto PA et al. Expression of CXCR4 predicts poor prognosis in patients with malignant melanoma. conclusions, these findings are consistent with the prognostic Clin Cancer Res 2005; 11:1835–41. role of a poor outcome associated with loss of E-cadherin 22 Robledo MM, Bartolome` RA, Longo N et al. Expression of func- expression and increased Cdc42 and CXCR4 expression, and tional chemokine receptors CXCR3 and CXCR4 on human mela- identify them as potential targets for therapeutic intervention. noma cells. J Biol Chem 2001; 276:45098–105. 23 Murakami T, Cardones AR, Hwang ST. Chemokine receptors and melanoma metastasis. J Dermatol Sci 2004; 36:71–8. References 24 Takeuchi H, Fujimoto A, Tanaka M et al. CCL21 chemokine regu- 1 Meier F, Satyamoorthy K, Nesbit M et al. Molecular events in mela- lates chemokine receptor CCR7 bearing malignant melanoma cells. noma development and progression. Front Biosci 1988; 3:D1005–10. Clin Cancer Res 2004; 10:2351–8. 2 Clark WH. Tumour progression and the nature of cancer. Br J Cancer 25 Kawada K, Sonoshita M, Sakashita H et al. Pivotal role of CXCR3 in mel- 1991; 64:631–44. anoma cell metastasis to lymph nodes. Cancer Res 2004; 64:4010–17.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1212–1216 EPIDEMIOLOGY AND HEALTH SERVICES RESEARCH DOI 10.1111/j.1365-2133.2007.08215.x Environmental factors, parental atopy and atopic eczema in primary-school children: a cross-sectional study in Taiwan Y-L. Lee, C-W. Li,* F-C. Sung, H-S. Yu, H-M. Sheu§ and Y.L. Guo– Departments of Occupational and Environmental Medicine, *Internal Medicine and §Dermatology, National Cheng Kung University, Tainan, Taiwan Institute of Environmental Health, China Medical University, Taichung, Taiwan Department of Dermatology, Kaohsiung Medical University, Kaohsiung, Taiwan –Department of Environmental and Occupational Medicine, College of Medicine, National Taiwan University (NTU) and NTU Hospital, 1 Sec 1, Jen-Ai Road, Taipei 100, Taiwan

Summary

Correspondence Background Parental atopy and environmental exposure are recognized risk factors Yueliang Leon Guo. for atopic eczema (AE) in childhood. However, the relative contributions of spe- E-mail: [email protected] cific risk factors and the overall contributions of hereditary and environmental exposure remain unexplored. Accepted for publication 1 July 2007 Objectives To identify risk factors, estimate the population attributable risk (PAR) of environmental exposure, and compare the AE data for boys vs. girls in Key words primary-school children. atopic eczema, children, environmental factors, Methods During a February to June 2001 cross-sectional, Taiwan-based question- parental atopy, population attributable risk naire survey, we investigated 23 980 children from 22 primary schools, all Conflicts of interest located within 1 km of an air-monitoring station. None declared. Results The 12-month prevalence of AE was reported as 6Æ1% in boys and 4Æ9% in girls. In both sexes, the risk of AE was strongly associated with parental atopy and perceived ambient air pollution. The presence of cockroaches [odds ratio (OR) 1Æ18, 95% confidence interval (CI) 1Æ00–1Æ40] and visible mould on walls at home (OR 1Æ46, 95% CI 1Æ22–1Æ70) were also significantly related to AE for girls; however, only visible mould on walls (and not the presence of cockroaches) at home was related to AE for boys (OR 1Æ40, 95% CI 1Æ18–1Æ66). While mutually adjusted models were applied, we found adjusted ORs and PARs were similar in boys and girls in hereditary and outdoor environmental factors. The PAR of indoor environmental factors was higher in girls (8Æ4%) than in boys (5Æ5%). There was no interaction between parental atopy and environmental factors. Conclusions Parental atopy contributed more to AE than indoor or outdoor environ- mental factors. Girls may be more susceptible to indoor environmental factors than boys.

Atopic eczema (AE) is now the most common inflammatory tional exposure, environmental tobacco smoke (ETS), indoor ⁄ skin disease in children,1,2 and recently the prevalence of outdoor air pollution, heating systems, aeroallergens and childhood AE has increased substantially in many countries.3–7 climate).9–20 Both hereditary and environmental factors are This increase has been too rapid to be accounted for by believed to contribute to the relationship.21 However, epi- changes in gene frequencies. It is also unlikely to be accounted demiological evidence concerning different effects in boys and for totally by changes in either clinical diagnostic patterns or girls in relationships between environmental factors and AE increased recognition of associated symptoms by the general was insufficient and thought to warrant further investigation. population.8 It does, however, suggest a role for environmen- To date, factors contributing to childhood AE have not been tal factors in the aetiology of this evolving epidemic.9–11 clearly documented in Taiwan. In this study, the relationship Many factors are proven to be associated with AE, including between AE and selected risk factors in a population-based personal factors (smoking habits, genetics, age, sex, nutritional sample of Taiwanese school children between the ages of 6 status, number of siblings, lifestyle, allergy status, family and 12 years was investigated. The population attributable risk history and occupation) and environmental stimuli (house dust- (PAR) of each factor was estimated and compared for boys mite, animal danders, moulds, cockroach infestation, occupa- and girls. We also tested the hypothesis that the joint effects

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 1217 1218 Environment, parental atopy and childhood eczema, Y-L. Lee et al. of genetic predisposition and environmental factors on the risk perception of ambient air pollution level was also considered of AE are greater than expected on the basis of their indepen- as an outdoor factor. Parental atopy was a measure of genetic dent effects. predisposition and defined by reports of the father or mother of the child ever having been diagnosed with AE, asthma or Materials and methods allergic rhinitis. To adjust for possible confounding, we also included host-related variables: the child’s age and sex, mater- nal smoking during pregnancy, the number of siblings at Population and study design home and the educational level of the household head. Unfor- The International Study of Asthma and Allergies in Childhood tunately, neither blood samplings nor skin tests could be per- (ISAAC) is a multinational collaborative project developed to formed in this large, nationwide study. investigate variations in childhood atopic diseases at the popu- lation level.22 Between February and June 2001, we modified Statistical analysis the ISAAC protocol and conducted a national, cross-sectional, school-based survey of primary-school children. Classroom Previously reported analyses of ecological outcomes have dem- incentives but not individual incentives were used to encour- onstrated a larger intercity variation than would be predicted age participation. The study protocol was approved by the by interindividual variation.28,29 We used two-stage methods Respiratory Health Screening Steering Committee of the Tai- to correct for any excess between-site variability. In the first wan Department of Health and the Institutional Review Board step, a logistic regression model was used to control for indi- at our university hospital, and it complied with the principles vidual-level confounders. In the second step, the community- outlined in the Helsinki Declaration.23 Parents of the school specific adjusted prevalences of perceived air pollution levels children consented to provide information by questionnaire. were regressed against the community-specific air pollutants; Whether or not a child was deemed to suffer from AE was the regression used weights inversely proportional to the sum determined by positive responses to the questions: ‘Has your of the between-site and within-site variances. child ever had an itchy rash which was coming and going for Bivariate logistic models with community clustering were at least 6 months?’ and ‘Has your child had this itchy rash at performed to determine associations with AE. All risk factors any time in the past 12 months?’ If both answers were ‘yes’, were categorized into three subgroups of factors—hereditary, the parent would be further asked: ‘Has this itchy rash at any indoor environmental and outdoor environmental factors— time affected any of the following places: the folds of the and we then developed multiple logistic regression models to elbows, behind the knees, in front of the ankles, under the assess the relative effectiveness of each on AE. Odds ratios buttocks, or around the neck, ears or eyes?’ In our study, chil- (ORs) with 95% confidence intervals (CIs) were calculated dren who were reported to suffer from a skin rash in the pre- after adjustment for potential confounders. PARs were also cal- vious year occurring at specific locations were defined as culated to estimate the contribution of various risk factors for having AE. The ISAAC questions for symptoms of AE used in AE. The PAR represents the preventable AE cases if children the present study have been validated in different parts of the were not exposed to specific agents or risk factors. PAR was world.24–26 calculated using the formula P(R ) 1)/[P(R ) 1) + 1], where In order to compare outdoor air pollution data with ques- P is the prevalence of the exposure and R is the relative risk tionnaire results, the study population was limited to children due to the exposure.30 attending schools located within 1 km of Taiwan Environmen- We assessed potential effect modification by parental atopy tal Protection Agency (EPA) air-monitoring stations. Complete by comparing effect estimates for children with and without monitoring data for the air pollutants sulphur dioxide (SO2), atopic parents. Individual and joint effects of environmental nitrogen oxides (NOx), ozone (O3), carbon monoxide (CO) factor and parental atopy on AE were estimated using indica- and particles with an aerodynamic diameter of 10 lm or less tor variables created for each category, omitting the hypothe-

(PM10) were available from the Taiwan EPA. Twenty-two of sized low–low risk category. Estimates for each of the three the 2604 primary schools in Taiwan’s 22 counties were inves- exposure categories with the reference group were derived tigated. Stratified sampling by grade was applied in each from the same logistic regression model, after adjustment for school. We believe our study population based on 22 different confounders. Statistical significance was set at P <0Æ05 based areas covering diverse parts of Taiwan to be representative of on a two-sided calculation. Taiwanese primary-school children.27,28 Results Genetic and environmental determinants Our study surveyed 23 980 children from 22 primary schools. Literature was reviewed on the causes of childhood AE in The total response rate was 88Æ8% (10 951 boys and 10 340 order to identify the hereditary and environmental risk factors. girls and their parents). All subjects were between 6 and Our focus was residential factors affected by climate and not 12 years old. Overall, 6Æ1% of boys and 4Æ9% of girls were directly related to human behaviour, including cockroaches, reported to have had AE in the previous year. Younger sub- water damage or visible mould on walls at home. Parental jects, higher parental education level, and maternal smoking

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 Environment, parental atopy and childhood eczema, Y-L. Lee et al. 1219 during pregnancy were found to be associated with the occur- Discussion rence of childhood AE (Table 1). In the year 2000 in Taiwan, there were approximately 1Æ04 million boys and 0Æ95 million The present large questionnaire survey for AE among 6–12- girls between the ages of 6 and 12 years, and these propor- year-old school children in Taiwan linked to local air-moni- tions suggest that nationwide about 109 990 children of this toring data demonstrated that outdoor air pollutants were age group are affected by AE. significantly associated with parentally perceived ambient air After adjustment for parental education level, the effects of pollution. Although our study was cross-sectional, we analysed each outdoor air pollutant on parentally perceived ambient air data using a case–control study method, which was very effi- pollution levels were assessed separately, and also expressed as cient compared with a cohort study yielding a similar amount ORs for a change by 1 SD (Table 2). In the regression model of information. Both parental atopy and environmental expos- where P-values were calculated, statistically significant associa- ure increased the risk of childhood AE but did not show sig- tions were found for SO2, CO, PM10 and NOx in prevalences nificant interactive effects. In addition, it showed girls to be of any ambient air pollution and moderate to severe ambient more susceptible to indoor factors than boys. We also found air pollution in the 22 target communities. Relatively weak approximately 7400 excess cases of AE attributable to indoor and nonsignificant associations were noted for O3. Levels of factors, and 11 420 attributable to outdoor ambient air all outdoor air pollutants had relatively stronger predictive pollution. effects on the prevalence of perceived moderate to severe air Questionnaires have been widely used to assess the preva- pollution than on the prevalence of perceived mild to severe lence of chronic illness such as atopic eczema. By using ISAAC air pollution level (Table 2). questions, researchers have had good results in predicting AE After adjustment for host factors such as age, parental edu- diagnosed by dermatologists in the U.K.,24 Germany25 and cation level and maternal smoking during pregnancy, we Ethiopia.26 We used typical symptoms of AE during the past found parental AE, parental asthma ⁄allergic rhinitis, and 12 months as the main outcome measurement in determining parentally reported ambient air pollution were strongly related risk factors. The overall prevalences in our study were 6Æ1% in to AE in both sexes (Table 3). Girls who lived in homes with boys and 4Æ9% in girls, lower than that reported in other cockroaches present were 1Æ18 times (95% CI 1Æ00–1Æ40) countries, such as Australia (16Æ3%),31 Singapore (20Æ8%)32 more likely to develop AE. The presence of visible mould on and Germany (10Æ5%).33 The causality of these substantial dif- walls at home was also significantly related to AE with OR ferences is beyond the scope of this report, but some research- 1Æ40 (95% CI 1Æ18–1Æ66) for boys and OR 1Æ46 (95% CI ers hypothesize that a Western life style is responsible:14,15 1Æ22–1Æ70) for girls. However, water damage at home showed the higher the level of Westernization, the higher the preva- positive but not statistically significant effects for both sexes. lence of AE. When mutually adjusted models were applied, we found In our data, younger subjects seem to have a higher preva- adjusted ORs and PARs were similar in boys and girls in lence of AE, a finding consistent with a previous Australian hereditary and outdoor environmental factors (Table 3). Stron- study.31 We also found parental education level to be associ- ger association between indoor factors—defined as visible ated with childhood AE (Table 1): the better educated the par- cockroaches, water damage or visible mould on walls at ents, the more concerned they were for the health of their home—and AE was noted in girls (OR 1Æ18, 95% CI 0Æ96– children, and hence the more likely these children were to be 1Æ45) compared with boys (OR 1Æ11, 95% CI 0Æ89–1Æ38). The reported as having AE.16,17 Moreover, maternal smoking dur- PAR of indoor environmental factors was higher in girls ing pregnancy exposes the fetus to more allergens,34 which (8Æ4%) than in boys (5Æ5%). For all the hereditary and envi- would contribute to the occurrence of atopic diseases in later ronmental factors we identified, the total PARs were 50Æ7% in life. Because these factors were potential confounders in risk the girls and 49Æ8% in the boys of our population. factor analyses, they were controlled as covariates in the fol- Of the estimated 109 990 cases of atopic eczema in 6–12- lowing regression models. Number of siblings at home was year-old Taiwanese primary-school children, we estimated that not associated with the prevalence of AE in our study more than 36 370 excess cases of AE were attributable to (Table 1). Some studies showed that it was not the number of hereditary factors—defined as parental AE, asthma or allergic siblings that mattered, but was the birth order that really had rhinitis. Exposure to ambient air pollution accounted for effects on the occurrence of AE in children.9,17,35 Therefore, approximately 11 420 excess cases. Indoor factors accounted we did not consider number of siblings as a confounder in for 7400 excess cases (3489 boys and 3911 girls). further analyses. In order to elaborate the potential effect modification, we ETS and incense burning at home showed negative effects systematically conducted stratified analyses in categories of to the occurrence of AE (Table 1), a finding consistent with parental atopy. The stratum-specific relations were relatively recent international studies.14,36 One possible explanation consistent and no significant interaction was found between could be that ETS and incense use might be reduced by fami- environmental factors and parental atopy on AE in both sexes lies with children with AE. Exposure to tobacco or incense (Table 4). This indicated that the environmental exposure and might also provide protective effects for atopic diseases parental atopy both have independent effects on the preva- through selection mechanisms, as shown in a cross-sectional lence of AE in childhood. study.36 Unlike tobacco or incense exposures, the indoor

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 1220 Environment, parental atopy and childhood eczema, Y-L. Lee et al.

Table 1 Prevalence of atopic eczema among % with primary-school children and association with Risk factor % of subjects N AE OR 95% CI potential risk factors Sex Boys 51Æ4 10 951 6Æ11Æ00 Girls 48Æ6 10 340 4Æ90Æ79 0Æ70–0Æ89 Age (years) 6–7 21Æ7 4610 5Æ71Æ00 816Æ4 3487 6Æ31Æ10 0Æ91–1Æ32 916Æ4 3494 5Æ20Æ90 0Æ74–1Æ09 10 17Æ3 3689 6Æ21Æ08 0Æ90–1Æ30 11 16Æ4 3492 4Æ90Æ84 0Æ69–1Æ03 12 11Æ8 2519 4Æ60Æ80 0Æ64–0Æ99 Parental education level (years) <9 17Æ1 3637 4Æ01Æ00 9–11 46Æ2 9837 4Æ71Æ17 0Æ97–1Æ42 ‡ 12 36Æ7 7817 7Æ31Æ88 1Æ57–2Æ28 Mother smoking during pregnancya No 97Æ9 20 733 5Æ51Æ00 Yes 2Æ1 448 8Æ31Æ56 1Æ09–2Æ16 Number of siblingsa 011Æ7 2023 6Æ01Æ00 140Æ3 6967 6Æ11Æ03 0Æ84–1Æ27 ‡ 248Æ0 8304 4Æ70Æ78 0Æ63–0Æ96 AE in fathersa No 97Æ1 20 324 5Æ01Æ00 Yes 2Æ9 597 22Æ55Æ51 4Æ48–6Æ72 AE in mothera No 97Æ5 20 418 5Æ11Æ00 Yes 2Æ5 528 21Æ45Æ06 4Æ06–6Æ26 Asthma ⁄AR in fathera No 82Æ2 17 199 4Æ91Æ00 Yes 17Æ8 3722 8Æ41Æ81 1Æ58–2Æ07 Asthma ⁄AR in mothera No 84Æ8 17 752 4Æ81Æ00 Yes 15Æ2 3194 9Æ72Æ14 1Æ87–2Æ45 ETS at homea No 45Æ9 9674 5Æ91Æ00 Yes 54Æ1 11 421 5Æ20Æ88 0Æ79–0Æ99 Incense burning at homea No 54Æ5 11 346 6Æ01Æ00 Yes 45Æ5 9458 5Æ10Æ85 0Æ75–0Æ95 Cockroaches seen monthly at homea 020Æ5 4309 5Æ11Æ00 1–2 44Æ2 9289 5Æ51Æ07 0Æ91–1Æ26 ‡ 335Æ2 7397 5Æ91Æ17 0Æ99–1Æ38 Water damage at homea No 93Æ3 19 769 5Æ51Æ00 Yes 6Æ7 1423 5Æ61Æ01 0Æ79–1Æ26 Number of walls with visible mould at homea 074Æ8 15 536 5Æ01Æ00 117Æ4 3605 6Æ61Æ34 1Æ15–1Æ55 ‡ 27Æ8 1625 8Æ31Æ70 1Æ40–2Æ05 Perceived ambient air pollution levela No 30Æ6 6386 4Æ81Æ00 Mild 63Æ0 13 145 5Æ81Æ22 1Æ06–1Æ40 Moderate to severe 6Æ5 1350 7Æ31Æ56 1Æ23–1Æ97

OR, odds ratio; CI, confidence interval; AE, atopic eczema; AR, allergic rhinitis; ETS, envi- ronmental tobacco smoke. aNumbers of subjects do not add up to total N because of mis- sing data. Some percentages do not total 100 because of rounding. ORs are crude odds ratios for each risk factor.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 Environment, parental atopy and childhood eczema, Y-L. Lee et al. 1221

Table 2 Odds ratios (ORs) and 95% confidence intervals (CIs) for the relationship Mild to severe air pollution Moderate to severe air pollution between each outdoor air pollutant and prevalence rates of perceived ambient air OR 95% CI P-value OR 95% CI P-value pollution level in 22 communities SO2 1Æ42 1Æ16–1Æ73 0Æ002 1Æ62 1Æ24–2Æ13 0Æ001 CO 1Æ37 1Æ14–1Æ65 0Æ002 1Æ56 1Æ21–2Æ02 0Æ002

O3 1Æ05 0Æ81–1Æ35 0Æ72 1Æ11 0Æ79–1Æ55 0Æ54 PM10 1Æ34 1Æ10–1Æ63 0Æ006 1Æ50 1Æ14–1Æ99 0Æ006 NOx 1Æ42 1Æ21–1Æ66 < 0Æ001 1Æ56 1Æ23–1Æ99 < 0Æ001

Prevalence rates are adjusted for parental education level and results are obtained from sin- gle pollutant model. ORs are calculated by using those reporting no ambient air pollution

as reference group, and expressed for a change in each pollutant by 1 SD (SO2,2Æ25 ppb; )3 CO, 158 ppb; O3,3Æ44 ppb; PM10,17Æ9 lgm and NOx, 8Æ60 ppb).

Table 3 Odds ratios (ORs) with 95% confidence intervals (CIs), mutually adjusted ORs and population attributable risks for parental atopy and environmental factors associated with atopic eczema in primary-school children

Boys Girls

Prevalence OR 95% PAR aOR 95% PAR Prevalence OR 95% PAR aOR 95% PAR (%) CI (%) CI (%) (%) CI (%) CI (%) Hereditary factors AE in father 3Æ06Æ20 4Æ75–8Æ04 10Æ82Æ74Æ38 3Æ12–6Æ03 7Æ4 AE in mother 2Æ75Æ35 4Æ00–7Æ07 8Æ52Æ44Æ09 2Æ85–5Æ74 6Æ0 Asthma ⁄AR in father 18Æ01Æ73 1Æ44–2Æ07 11Æ317Æ61Æ69 1Æ37–2Æ07 10Æ5 Asthma ⁄AR in mother 15Æ51Æ98 1Æ64–2Æ37 12Æ715Æ02Æ22 1Æ80–2Æ73 14Æ9 Any parental atopy 32Æ52Æ65 2Æ26–3Æ11 33Æ92Æ60 2Æ21–3Æ05 33Æ231Æ62Æ63 2Æ19–3Æ16 33Æ12Æ61 2Æ17–3Æ14 32Æ9 Indoor environmental factors Cockroaches 79Æ41Æ08 0Æ89–1Æ33 6Æ079Æ51Æ18 1Æ00–1Æ40 12Æ5 Water damage 6Æ61Æ10 0Æ79–1Æ48 0Æ76Æ91Æ01 0Æ74–1Æ39 0Æ1 Visible mould 25Æ51Æ40 1Æ18–1Æ66 9Æ124Æ91Æ46 1Æ22–1Æ70 10Æ1 Any indoor factor 83Æ41Æ11 0Æ89–1Æ38 8Æ41Æ07 0Æ86–1Æ34 5Æ583Æ61Æ18 0Æ96–1Æ45 13Æ11Æ11 0Æ90–1Æ39 8Æ4 Outdoor environmental factor Perceived ambient 69Æ41Æ26 1Æ06–1Æ52 15Æ21Æ18 0Æ98–1Æ42 11Æ169Æ41Æ25 1Æ02–1Æ54 14Æ81Æ15 0Æ93–1Æ41 9Æ4 air pollution

All ORs are adjusted for age, parental education level and maternal smoking during pregnancy. The risk factors are not mutually exclusive and PARs are not additive in this table. PAR, population attributable risk; aOR, mutually adjusted odds ratio; AE, atopic eczema; AR, allergic rhinitis.

environmental factors we chose, including cockroaches, water a parent carrying any atopic disease were also found to have a damage and visible mould on walls at home, would not be higher probability of developing AE in later life than those easily changed by human behaviours and would not show who lived at home with the selected environmental factors. In negative effects on AE. the mutually adjusted models, hereditary factors—defined as A family history of allergic diseases was associated with an parental AE and asthma ⁄allergic rhinitis—also possessed the increased risk of AE, suggesting that genetic factors play a cen- highest attributable risks, which were consistent with 36 370 tral role in the development of childhood AE.9,17 Genetic excess cases annually in Taiwan. markers could also increase susceptibility of children to the The ecological exposure assessment had many advantages in effects of environmental factors.9 A recent German cross-sec- our study. The density of primary schools in Taiwan is very tional study17 revealed that paternal and maternal histories of high, and almost all the surveyed children attended schools AE were equally strong determinants of AE in children. Besides within 1 km of their homes. Monitoring stations located near AE in parents, maternal and paternal atopy were also found to the schools were therefore also likely to be near the children’s be significantly associated with childhood AE in the U.K.9 Our homes, and thus provided good indicators for both school study demonstrated that paternal and maternal AE were stron- and home exposure. A Taiwanese study has suggested that ger risk factors for childhood AE in both sexes than were parental ranking of the air pollution level was a good pre- asthma and allergic rhinitis in parents (Table 3). Children with dictor for childhood asthma,27,37 which also demonstrated an

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 1222 Environment, parental atopy and childhood eczema, Y-L. Lee et al.

apparent dose–response relationship. In this study, we also proved that the parentally perceived ambient air pollution level was associated with outdoor air pollutants and the associ- 48 07 49 59 43 59 Æ Æ Æ Æ Æ Æ -value for

P interaction ation was stronger if a more severe air pollution level was reported (Table 2). Limited studies have been conducted on the effect of air pollution on AE and the results were still 79 50 0 89 47 0 84 42 0 16 47 0 65 48 0 39 0 98 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ inconclusive,14,19 but in our study, outdoor air pollution was 63–1 79–1 86–1 91–1 80–1 21–2 52–1 91–1 73–1 67–2 89–1 86–2 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ found to play an important role in the occurrence of AE in childhood (Table 3). The greater the ambient air pollution perceived by parents, the higher the risks of AE in children. 10 0 08 0 13 0 28 0 06 0 62 1 91 0 22 0 03 0 20 0 27 0 52 0 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Indoor dampness markers, such as visible mould and water damage, increase the risks of AE in childhood.17,20 A German study demonstrated that water-related damage at home was associated with the amount of house dust-mite antigen in the dust vacuumed from the children’s mattresses.38 In Finland, children exposed to mould in a school building were found 21 31 41 71 31 91 60 81 71 31 71 21 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ AE (%) without factor OR 95% CI to have higher risks of IgE elevation.39 Our results also showed significant associations between AE and visible mould at home after adjustment for covariates in both sexes. Although boys had a higher prevalence of AE than girls, when 43 68 28 42 78 62 48 42 58 67 42 32 we estimated the effect of any indoor environmental factor on Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ AE (%) with factor childhood AE, a stronger relationship was noted in girls than in boys (Table 3). The mechanism of such female-led suscep- tibility is not well understood. One possible explanation con- sidered was that girls might have more extensive exposures 88 8 37 9 25 8 83 7 84 8 70 8 Æ Æ Æ Æ Æ Æ -value for

P interaction because they are relatively inactive and spend more time at home and, therefore, are more influenced by the indoor envi- 74 3 40 0 75 0 37 3 38 0 54 4 62 0 72 3 50 0 78 0 43 3 33ronment. 3 Sex-differences in the pathogenesis of AE might be Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ partly due to difference in lifestyle, or differences in skin mor- 79–1 10–1 77–1 83–1 90–1 66–1 02–1 80–1 71–1 10–4 76–1 02–3 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ phology and physiology. Our data did not show any significant interactive effect between parental atopy and environmental factors on the prev- 05 0 39 1 02 0 06 0 19 0 06 0 32 1 08 0 14 0 13 1 03 0 75 1 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ alence of AE (Table 4). In a recent questionnaire survey of Finnish adolescents, Kilpelainen et al. also found that no inter- active effects existed between indoor dampness and parental atopic diseases.20 Parental atopy would not modify the effects of environmental factors on the risk of AE in childhood. 21 51 01 01 91 31 41 01 01 32 01 41 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ It is difficult to target preventative efforts on childhood AE. AE (%) without factor OR 95% CI Environmental factors showed a relatively small but substantial effect on childhood AE in our study. Only 10Æ4% boys and 8Æ5% girls with parental atopy were reported as having AE. If such efforts were to target only families with a history of 510 89 14 510 73 610 53 510 54 65 14 22 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ atopy, then only 9Æ5% children could potentially benefit. BoysAE (%) with factor Girls However, it seems easier to eliminate such exposures on a national scale than to attempt to counter hereditary factors. Additional research is necessary to prove that the elimination Parental atopy No 4 Yes 10 No 4 No 4 No 4 No 4 No 4 of indoor ⁄outdoor environmental exposures will result in lower rates of childhood AE. Our study has some limitations. This is a cross-sectional study with relatively weak inference in causal relationship. The questionnaire-based assessment based on symptom report- ing is not as precise as doctors’ diagnoses. However, in the presence of a true association, misclassification of AE that was Prevalence (%) of atopic eczema (AE) and association with environmental factors in primary-school children, stratified by parental atopy random with respect to other study variables would weaken

pollution the observed association rather than lead to false-positive Environmental factor Cockroaches Yes 10 Visible mould Yes 12 Perceived ambient air Water damage Yes 10 Any indoor factor Yes 10 Any environmental factor Yes 10 OR, odds ratio; CI, confidence interval. All ORs are adjusted for age, parental education level and maternal smoking during pregnancy.

Table 4 results. In addition, this report is a prevalence study, rather

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 Environment, parental atopy and childhood eczema, Y-L. Lee et al. 1223 than an incidence study. A small number of families relocated, 2 Williams HC. On the definition and epidemiology of atopic derma- and some of the factors we studied might have affected the titis. Dermatol Clin 1995; 13:649–57. prevalence of AE through effects on disease duration rather 3 Grize L, Gassner M, Wuthrich B et al. Trends in prevalence of asthma, allergic rhinitis and atopic dermatitis in 5–7-year-old Swiss than disease incidence. However, our findings are interesting children from 1992 to 2001. Allergy 2006; 61:556–62. and real, regardless of whether or not the observed associa- 4 Galassi C, De Sario M, Biggeri A et al. Changes in prevalence of tions were caused by effects arising from incidence or dura- asthma and allergies among children and adolescents in Italy: tion. In fact, if factors were found to be associated with the 1994–2002. Pediatrics 2006; 117:34–42. prevalence of AE, then they are of major interest in them- 5 Wang XS, Tan TN, Shek LP et al. The prevalence of asthma and selves, irrespective of whether the aetiological mechanism allergies in Singapore; data from two ISAAC surveys seven years involved the increase in disease incidence or in duration. apart. Arch Dis Child 2004; 89:423–6. 6 Selnes A, Bolle R, Holt J et al. Cumulative incidence of asthma and Because we were unable to measure personal environ- allergy in north-Norwegian schoolchildren in 1985 and 1995. Pedi- mental exposures or sensitization to various allergens, such atr Allergy Immunol 2002; 13:58–63. as dust mites, fungi or cockroaches, we might have under- 7 Lee YL, Li CW, Sung FC et al. Increasing prevalence of atopic estimated the effects of these indoor factors on childhood eczema in Taiwanese adolescents from 1995 to 2001. Clin Exp AE. Outdoor environmental factors were measured only by Allergy 2007; 37:543–51. perceived air pollution level, which must be an imprecise 8 Hopkin J. Genetics of atopy. Pediatr Allergy Immunol 1995; 6:139–44. way of measuring the influence of environmental exposure. 9 Harris JM, Cullinan P, Williams HC et al. Environmental association with eczema in early life. Br J Dermatol 2001; 144:795–802. Another potential source of bias was in the interpretation of 10 McNally NJ, Williams HC, Phillips DR. Atopic eczema and the parental history for atopy as an indicator of a genetic pre- home environment. Br J Dermatol 2001; 145:730–6. disposition to childhood AE. Although the importance of 11 McNally NJ, Williams HC, Phillips DR et al. Is there a geograph- parental history as a predictor of disease has been demon- ical variation in eczema prevalence in the U.K.? Evidence from strated,9,17 not every child in the family inherits the allergic the 1958 British birth cohort study. Br J Dermatol 2000; 142: tendency. Ecological confounders such as urbanization and 712–20. socialization actually could exist in data analysis and there 12 Annesi-Maesano I, Oryszczyn MP, Raherison C et al. Increased prev- alence of asthma and allied diseases among active adolescent might be incomplete adjustment and residual confounding. tobacco smokers after controlling for passive smoking exposure. However, more complete personal risk factors are very diffi- A cause for concern? Clin Exp Allergy 2004; 34:1017–23. cult to obtain in such a large-scale survey. Investigators 13 Mitchell EA, Stewart AW, ISAAC Phase One Study Group. Interna- decided not to try to obtain more personal information as tional Study of Asthma and Allergy in Childhood. The ecological it would have reduced the participation rate and introduced relationship of tobacco smoking to the prevalence of symptoms of greater bias into the study. asthma and other atopic diseases in children: the International In conclusion, we identified a number of hereditary and Study of Asthma and Allergies in Childhood (ISAAC). Eur J Epidemiol 2001; 17:667–73. environmental factors associated with AE in 6–12-year-old pri- 14 Dotterud LK, Odland JO, Falk ES. Atopic disease among school- mary-school children in Taiwan. Parental atopy contributed children in Nikel, Russia, an Arctic area with heavy air pollution. more to childhood AE than environmental factors. Exposure to Acta Derm Venereol (Stockh) 2001; 81:198–201. environmental factors increased the risk of AE in children 15 Diepgen TL. Atopic dermtatitis: the role of environmental and regardless of the coexisting hereditary factors. Girls may be social factors, the European experience. J Am Acad Dermatol 2001; more susceptible to indoor environmental factors than boys. 45:S44–8. The present findings suggest that public health policies for 16 Heinrich J, Popescu MA, Wjst M et al. Atopy in children and paren- tal social class. Am J Public Health 1998; 88:1319–24. eliminating certain environmental factors are needed, which 17 Zutavern A, Hirschw T, Leupoldz W et al. Atopic dermatitis, extrin- could contribute not only to children’s health but also to sic atopic dermatitis and the hygiene hypothesis: results from a medical costs in Taiwan. cross-sectional study. Clin Exp Allergy 2005; 35:1301–8. 18 Schafer T, Heinrich J, Wjst M et al. Indoor risk factors for atopic eczema in school children from East Germany. Environ Res 1999; Acknowledgments 81:151–8. The authors thank the field workers, teachers and other school 19 Dotterud LK, Odland JO, Falk ES. Atopic diseases among adults in the two geographically related Arctic areas Nikel, Russia and staff who supported data collection, and all the parents and Sor-Varanger, Norway: possible effects of indoor and outdoor air children who participated in this study. This study was par- pollution. J Eur Acad Dermatol Venereol 2000; 14:107–11. tially supported by grant DOH90-TD-1138 from Department 20 Kilpelainen M, Terho EO, Helenius H et al. Home dampness, cur- of Health in Taiwan. rent allergic diseases, and respiratory infections among young adults. Thorax 2001; 56:462–7. 21 Kluken H, Wienker T, Bieber T. Atopic eczema ⁄dermatitis syn- References drome—a genetically complex disease. New advances in discover- 1 International Study of Asthma and Allergies in Childhood Steering ing the genetic contribution. Allergy 2003; 58:5–12. Committee. Worldwide variation in prevalence of symptoms of 22 Asher MI, Keil U, Anderson HR et al. International Study of Asthma asthma, allergic rhinoconjunctivitis, and atopic eczema: ISAAC. Lan- and Allergies in Childhood (ISAAC): rationale and methods. cet 1998; 351:1225–32. Eur Respir J 1995; 8:483–91.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 1224 Environment, parental atopy and childhood eczema, Y-L. Lee et al.

23 41st World Medical Assembly. Declaration of Helsinki: recommen- 31 Marks B, Kilkenny M, Plunkett A et al. The prevalence of common dations guiding physicians in biomedical research involving human skin conditions in Australian school students: 2. Atopic dermatitis. subjects. Bull Pan Am Health Organ 1990; 24:606–9. Br J Dermatol 1999; 140:468–73. 24 Williams HC, Burney PGJ, Pembroke AC et al. Validation of the 32 Tay YK, Kong KH, Khoo L et al. The prevalence and descriptive epi- U.K. diagnostic criteria for atopic dermatitis in a population set- demiology of atopic dermatitis in Singapore school children. Br J ting. U.K. Diagnostic Criteria for Atopic Dermatitis Working Party. Dermatol 2002; 146:101–6. Br J Dermatol 1996; 135:12–17. 33 Werner S, Buser K, Kapp A et al. The incidence of atopic dermatitis 25 Kramer U, Schafer T, Behrendt H et al. The influence of cultural in school entrants is associated with individual life-style factors but and educational factors on the validity of symptom and diagnosis not with local environmental factors in Hannover, Germany. Br J questions for atopic eczema. Br J Dermatol 1998; 139:1040–6. Dermatol 2002; 147:95–104. 26 Haileamlak A, Lewis SA, Britton J et al. Validation of the Interna- 34 van Bever HP. Early events in atopy. Eur J Pediatr 2002; 161:542–6. tional Study of Asthma and Allergies in Children (ISAAC) and U.K. 35 Strachan DP. Allergy and family size: a riddle worth solving. criteria for atopic eczema in Ethiopian children. Br J Dermatol 2005; Clin Exp Allergy 1997; 27:235–6. 152:735–41. 36 Hjern A, Hedberg A, Haglund B et al. Does tobacco smoke prevent 27 Lee YL, Lin YC, Hsiue TR et al. Indoor ⁄outdoor environmental atopic disorders? A study of two generations of Swedish residents. exposures, parental atopy, and physician-diagnosed asthma in Clin Exp Allergy 2001; 31:908–14. Taiwanese schoolchildren. Pediatrics 2003; 112:e389–95. 37 Lin RS, Sung FC, Huang SL et al. Role of urbanization and air pollu- 28 Lee YL, Shaw CK, Su HJ et al. Climate, traffic-related air pollutants, tion in adolescent asthma: a mass screening in Taiwan. J Formos Med and allergic rhinitis prevalence in middle-school children in Assoc 2001; 100:649–55. Taiwan. Eur Respir J 2003; 21:964–70. 38 Nicolai T, Illi S, von Mutius E. Effect of dampness at home in 29 Dockery DW, Cunningham J, Damokosh AI et al. Health effects of childhood on bronchial hyperreactivity in adolescence. Thorax acid aerosols on North American children: respiratory symptoms. 1998; 53:1035–40. Environ Health Persp 1996; 104:500–5. 39 Savilahti R, Uitti J, Roto P et al. Increased prevalence of atopy 30 Coughlin S, Benichou J, Weed D. Attributable risk estimation in among children exposed to mold in a school building. Allergy case–control studies. Epidemiol Rev 1994; 16:51–64. 2001; 56:175–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1217–1224 PAEDIATRIC DERMATOLOGY DOI 10.1111/j.1365-2133.2007.08254.x Novel EBP gene mutations in Conradi–Hu¨nermann–Happle syndrome P.M. Steijlen, M. van Geel, M. Vreeburg,* D. Marcus-Soekarman,* L.J.M. Spaapen,* F.C.M. Castelijns,à M. Willemsen§ and M.A.M. van Steensel Departments of Dermatology, *Clinical Genetics and Clinical Chemistry, University Hospital Maastricht, PO Box 5800, 6202 AZ Maastricht, the Netherlands àDepartment of Dermatology, Elkerlyck Hospital Helmond, the Netherlands §Department of Pediatric Neurology, University Medical Center Nijmegen, the Netherlands

Summary

Correspondence Background Conradi–Hu¨nermann–Happle syndrome [X-linked dominant chondro- Maurice A.M. van Steensel. dysplasia punctata type 2 (CDPX2); MIM no. 302960] is an X-linked dominant E-mail: [email protected] disorder of cholesterol metabolism that causes a wide spectrum of skeletal abnor- malities and linear ichthyosiform skin lesions. Mosaicism is probably responsible Accepted for publication 29 July 2007 for the variability of the phenotype. Objectives To describe new mutations in patients with variable manifestations of Key words the disease. CDPX2, Conradi–Hu¨nermann–Happle, EBP, Methods We studied three patients with CDPX2. We performed mutation analysis mosaicism, X-inactivation of the EBP (formerly known as CDPX2) gene and gas chromatography–mass spec- Conflicts of interest troscopy on serum of two patients. None declared. Results We found two novel (3GfiT and 419-422delTTCT) and one known mutation in the EBP gene. We demonstrated the presence of increased levels of dehydrocholesterol and 8(9)-cholestenol in the two patients with new mutations, confirming the diagnosis of CDPX2 and strongly suggesting that the mutations are indeed pathogenic. One patient had a very mild phenotype, presenting with linear alopecia and a mild symmetrical epiphyseal dysplasia. X-inactivation studies in peripheral blood of all patients showed skewing in only the most severely affected patient. Conclusions The strong phenotypic variability in our patients suggests that there is no clear genotype–phenotype correlation.

X-linked dominant chondrodysplasia punctata type 2 [CDPX2 of epiphyses can be observed in children suffering from (Conradi–Hu¨nermann–Happle syndrome); MIM no. 302960] CDPX2 but also in several other disorders including the auto- is caused by mutations in the EBP (emopamil-binding protein) somal dominant, recessive and X-linked recessive forms of gene coding for 3b-hydroxysteroid-D8,D7-isomerase, a central chondrodysplasia punctata. Mental retardation has also been component of cholesterol biosynthesis that catalyses the con- reported. As expected, CDPX2 shares characteristics with other version of 8(9)-cholestenol into lathosterol (5-a-cholest-7- disorders associated with defects in the cholesterol synthesis en-3-b-ol to 5-a-cholest-8-en-3-b-ol).1 Dysfunction of the pathways, notably CHILD (congenital hemidysplasia with ich- enzyme leads to increased levels of 8-dehydrocholesterol and, thyosiform erythroderma and limb defects; MIM no. 308050) typically, to the presence of 8(9)-cholestenol in serum. These and Smith–Lemli–Opitz syndromes (MIM no. 270400). Sev- cholesterol precursors provide an indirect indication of the eral different mutations have been described in patients with metabolic defect and can thus be used to confirm functional phenotypes that show considerable variation in severity. The relevance of any EBP mutation found.2,3 Directly after birth, variation is not a function of the particular mutation but is patients can be erythrodermic and show linear hyperkeratotic thought to result from genetic mosaicism caused by X-inacti- skin lesions. During the first few weeks of life, these develop vation.5–7 There has been a report of a single gene mosaicism into linear and whorled atrophic and hypopigmented skin in an affected boy,8 confirming the notion that CDPX2 can lesions. Linear alopecia, sectoral cataracts and skeletal abnor- only exist in a mosaic state. In conflict with this assertion, a malities such as rhizomelic limb shortening, short stature, ver- nonmosaic mutation (L18P) was described in a boy with what tebral anomalies and facial defects are also found.4 Stippling was called a ‘severe atypical phenotype’.9 However, this case

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1225–1229 1225 1226 Novel EBP gene mutations in Conradi–Hu¨nermann–Happle syndrome, P.M. Steijlen et al. is doubtful as no stippling of the epiphyses or shortening of tubular bones was seen. The facial phenotype is in our view not typical either, an observation shared by the authors them- selves. 8(9)-Cholestenol was increased in the patient and his mother, although not to an extreme degree. Thus, we agree with Happle10 that L18P is a hypomorphic allele not associ- ated with CDPX2. We report three female patients with CDPX2, two of whom have novel mutations in EBP. Their phenotypes range from very mild to quite severe. A clear genotype–phenotype correl- ation does not seem to be present. X-inactivation patterns did not correlate with disease severity.

Materials and methods

Informed consent was obtained from the parents in all cases. DNA was isolated from peripheral blood leucocytes as previ- ously described.11 The EBP gene has four coding exons that were amplified using the primers and conditions listed in Table 1. The polymerase chain reaction (PCR) products were subjected to direct sequencing using the PCR primers. Sequences were assembled using the Phred-Phrap-Consed soft- ware package.12–14 Mutation M1I was screened in 100 healthy controls by NlaIII (New England Biolabs, Frankfurt am Main, Germany) restriction analysis, while mutation 419-422del- Fig 1. Patient 1 with clear asymmetrical rhizomelic shortening of the TTCT was tested on controls by sizing of a fluorescently extremities. labelled PCR fragment on a ABI377 DNA sequencer (Applied Biosystems, Foster City, CA, U.S.A.). haemangioma on the upper back. At birth, she had been ery- For gas chromatography–mass spectrometric (GC-MS) anal- throdermic and an ‘ichthyosis’ had been diagnosed at the ysis, fresh plasma was obtained and stored at )80 C prior to time. In the weeks after birth, the erythroderma disappeared, analysis. Denaturing high-performance liquid chromatography as did most of the scaling. The haemangioma was not visible (DHPLC) analysis was performed in patients 2 and 3 as previ- at birth but first appeared a few days afterwards, growing rap- ously described.2 idly. When we examined her at age 3 months, we saw atro- X-inactivation studies were performed on peripheral blood phic hypopigmented linear skin lesions following the lines of leucocytes in all three patients. X-inactivation at the androgen Blaschko. Fine white scaling was present in several lesions. receptor locus (HUMARA) was examined as previously The right leg appeared to be slightly shorter than the left and described.15 the left upper arm shorter than the right, with the shortening limited to the femur and humerus (rhizomelic). X-ray exami- nation of the skeleton confirmed the clinical impression of Patients rhizomelic shortening (Fig. 1). The haemangioma on the back Patient 1 is a 6-year-old girl of Dutch descent who presented was ulcerating and infected with Staphylococcus aureus. It was for evaluation of linear scaling skin lesions and an infected treated with oral amoxicillin, with excellent clinical result. Upon revision several months later, the shortening of the right femur was more pronounced. The haemangioma was no Table 1 EBP polymerase chain reaction (PCR) and sequence primers longer ulcerated and had regressed. We diagnosed the patient with CDPX2. Primer PCR product Patient 2, a 12-year-old girl of Iraqi descent, presented to name Sequence (5¢fi3¢) size (bp) our department for the evaluation of life-long patchy alopecia EBP2Fa TGGTAAATTCGGTCCATTTACATTTC 533 of the scalp. She was short for her age but had otherwise EBP2R GGTAAACATGCAAGCCGATCC developed normally. No other family members were affected. EBP2F TGCCTATACACACGCAGCCATC Sequence primer The parents denied consanguinity. An older sister was not EBP3F GGAAGCAGACGCATGGGAAG 521 affected. According to the parents, the patient had been born EBP4R TGGAAGGGCACCGTTGAGAG with normal skin. EBP5F TGATGACTTGGAAAGTGCTTTGG 616 Upon examination, we noted subtle atrophic linear skin EBP5R TCATAGCGCATCTCTGGGGTTAG lesions mostly on the extremities following a type I Blaschko pattern (Fig. 2). On the scalp, a linear hairless lesion was

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1225–1229 Novel EBP gene mutations in Conradi–Hu¨nermann–Happle syndrome, P.M. Steijlen et al. 1227

(a) (b)

Fig 2. Upper arms of patient 2 with subtle hypopigmented streaks following the lines of Blaschko.

present extending from the vertex to the occiput. There were no overt asymmetries or dysmorphic features. Her height was 150 cm. Mid-parental height was used to predict normal height, as growth curves for Middle Eastern children are not available; it was calculated to be 1.65 m. X-ray examination of both upper legs showed small, slightly deformed, capital epiphyses, an observation which the radiologist deemed con- sistent with having had ‘epiphyseal dysplasia’ in the past. Based on the skin abnormalities and the skeletal abnormalities we considered a diagnosis of CDPX2. Patient 3 was born with erythroderma and linear ichthyosi- form skin lesions (Fig. 3). The consultant dermatologist sus- pected CDPX2 immediately and referred the patient for confirmation of the diagnosis. Upon examination, we saw lin- ear hyperkeratotic, slightly erythematous, skin lesions on the thorax and both arms. The nasal bridge appeared slightly sunken. Other abnormalities were not evident at the time, although clinical revision after 2 months showed that the patient was developing a pronounced scoliosis. A skeletal survey showed several abnormalities consistent with the diagnosis of CDPX2. In particular, there were tho- racic hemivertebrae (Fig. 4) and epiphyseal stippling with asymmetric rhizomelic shortening of the lower extremities. A magnetic resonance image of the brain was performed but did not show clear abnormalities.

Mutation analysis

In patient 1, mutation analysis by direct sequencing of the EBP gene showed a 184CfiT mutation in exon 2, leading to a Fig 3. Neonatal presentation of X-linked dominant chondrodysplasia R62W substitution in EBP. This mutation has been previously punctata type 2 with erythroderma and linear ichthyosiform skin 16 described. lesions in patient 3.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1225–1229 1228 Novel EBP gene mutations in Conradi–Hu¨nermann–Happle syndrome, P.M. Steijlen et al.

explained by assuming functional mosaicism.7 The other patients, 1 and 3, had more typical manifestations of CDPX2. Surprisingly, but in line with previous findings,18 we found no evidence for skewing of X-inactivation in two of our patients (patients 1 and 2). In patient 2, with her mild phenotype, it might be expected that she should show skewed X-inactivation. That we found no evidence for this assumption can be explained in two ways. Either the pattern of X-inactivation strongly varies between tissues, or the extent of mosaicism has no influence on the severity of the phenotype. The latter expla- nation is untenable in view of previous reports on X-linked dominant disorders.7 With regard to the former, although X-inactivation patterns in different tissues are correlated some- what, they vary strongly.19 Therefore, we think that our findings reflect tissue heterogeneity in the pattern of X-inacti- vation. Confirming this is quite difficult, as affected skin cells tend to disappear over time as a result of secondary selection processes and the sampling of other tissues is for obvious ethi- Fig 4. A severe left-convex scoliosis caused by thoracic hemivertebrae cal reasons often not possible. Thus, absence of skewing in in patient 3 (arrow). peripheral blood leucocytes does not equal absence of an X-linked dominant disorder. Its presence is not correlated with Patient 2 had a heterozygous G to T transversion of the the severity of the disease. From a purely clinical point of third nucleotide of the initiation codon (ATGfiATT). Patient view, the presentation of patient 2 is most interesting. She 3 was found to have a 4-bp deletion (419-422delTTCT) in came to our attention because of the linear alopecia that was the EBP gene, leading to a frameshift and premature stop in considered as cosmetically unacceptable. The other skin lesions the protein (F140fsX166). The latter two mutations have not had never been noticed by the parents, not even directly after been described previously. None of the mutations were birth although that may seem unlikely in retrospect. However, detected in 100 unrelated healthy controls. the skin lesions that we found were quite subtle and could eas- Gas chromatographic examination of fresh plasma was ily be overlooked. Hence, it seems appropriate to suggest that performed in patients 2 and 3, showing increased levels of the presence of linear alopecia should alert clinicians to the 8-dehydrocholesterol as well as the presence of 8(9)-choleste- possibility that it is caused by an X-linked dominant ectoder- nol, thus confirming the functional relevance of the mutations mal dysplasia. In the presence of atrophic skin lesions, the found. X-inactivation studies were performed in all three diagnoses to be entertained are CDPX2, incontinentia pigmenti, patients and showed significant skewing only in patient 3, focal dermal hypoplasia and MIDAS ⁄MLS (microphthalmia, where more than 90% use of one allele was demonstrated. dermal aplasia and sclerocornea ⁄microphthalmia and linear skin defects).7 A careful history and physical examination will Discussion usually allow a distinction, but chromatography and mutation analysis are required for a definitive diagnosis. In three patients with widely varying manifestations of CDPX2 we have found two novel and one known mutation in the EBP Acknowledgments gene. By GC-MS analysis of plasma sterols2 in the two patients with novel mutations, we confirmed the diagnosis. The abnor- M. van S. is supported by grants from Barrier Therapeutics NV, mal sterol levels also strongly suggest functional relevance of the University Hospital Maastricht and the Netherlands Organi- the new mutations. The one patient in whom GC-MS was not zation for Scientific Research (ZONMW grant 907-00-202). performed had a classical phenotype with a known mutation and thus, chromatography was not deemed necessary. The sur- References prisingly mild phenotype in the patient with the start codon mutation can in our view best be explained by functional 1 Derry JM, Gormally E, Means GD et al. Mutations in a delta 8-delta mosaicism.5,7 Alternatively, an alternative in-frame initiation 7 sterol isomerase in the tattered mouse and X-linked dominant codon may explain the mild phenotype.17 However, none are chondrodysplasia punctata. Nat Genet 1999; 22:286–90. 2 Has C, Seedorf U, Kannenberg F et al. Gas chromatography–mass available upstream of EBP exon 1 and a putative downstream spectrometry and molecular genetic studies in families with the alternative initiation codon in exon 4 would eliminate more Conradi–Hunermann–Happle syndrome. J Invest Dermatol 2002; than half of the EBP protein, excluding this explanation for the 118:851–8. mild phenotype. If the nature of the mutations were important, 3 DiPreta EA, Smith KJ, Skelton H. Cholesterol metabolism defect abolition of the translation start site should be expected to lead associated with Conradi–Hunerman–Happle syndrome. Int J Dermatol to a severe phenotype. We suggest that this observation can be 2000; 39:846–50.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1225–1229 Novel EBP gene mutations in Conradi–Hu¨nermann–Happle syndrome, P.M. Steijlen et al. 1229

4 Herman GE. Disorders of cholesterol biosynthesis: prototypic meta- 12 Gordon D, Abajian C, Green P. Consed: a graphical tool for bolic malformation syndromes. Hum Mol Genet 2003; 12(Spec No 1): sequence finishing. Genome Res 1998; 8:195–202. R75–88. 13 Ewing B, Green P. Base-calling of automated sequencer traces using 5 Feldmeyer L, Mevorah B, Grzeschik KH et al. Clinical variation in phred. II. Error probabilities. Genome Res 1998; 8:186–94. X-linked dominant chondrodysplasia punctata (X-linked dominant 14 Ewing B, Hillier L, Wendl MC et al. Base-calling of automated ichthyosis). Br J Dermatol 2006; 154:766–9. sequencer traces using phred. I. Accuracy assessment. Genome Res 6 Happle R, Matthiass HH, Macher E. Sex-linked chondrodysplasia 1998; 8:175–85. punctata? Clin Genet 1977; 11:73–6. 15 Allen RC, Zoghbi HY, Moseley AB et al. Methylation of HpaII and 7 Happle R. X-chromosome inactivation: role in skin disease expres- HhaI sites near the polymorphic CAG repeat in the human andro- sion. Acta Paediatr Suppl 2006; 95:16–23. gen-receptor gene correlates with X chromosome inactivation. Am J 8 Aughton DJ, Kelley RI, Metzenberg A et al. X-linked dominant Hum Genet 1992; 51:1229–39. chondrodysplasia punctata (CDPX2) caused by single gene mosai- 16 Herman GE, Kelley RI, Pureza V et al. Characterization of mutations cism in a male. Am J Med Genet 2003; 116:255–60. in 22 females with X-linked dominant chondrodysplasia punctata 9 Milunsky JM, Maher TA, Metzenberg AB. Molecular, biochemical, (Happle syndrome). Genet Med 2002; 4:434–8. and phenotypic analysis of a hemizygous male with a severe atypical 17 Puel A, Reichenbach J, Bustamante J et al. The NEMO mutation cre- phenotype for X-linked dominant Conradi–Hunermann–Happle ating the most-upstream premature stop codon is hypomorphic syndrome and a mutation in EBP. Am J Med Genet A 2003; 116:249– because of a reinitiation of translation. Am J Hum Genet 2006; 54. 78:691–701. 10 Happle R. Hypomorphic alleles within the EBP gene cause a pheno- 18 Ikegawa S, Ohashi H, Ogata T et al. Novel and recurrent EBP muta- type quite different from Conradi–Hunermann–Happle syndrome. tions in X-linked dominant chondrodysplasia punctata. Am J Med [Comment, Letter.] Am J Med Genet A 2003; 122:279–80. Genet 2000; 94:300–5. 11 Miller SA, Dykes DD, Polesky HF. A simple salting out procedure 19 Sharp A, Robinson D, Jacobs P. Age- and tissue-specific variation for extracting DNA from human nucleated cells. Nucleic Acids Res of X chromosome inactivation ratios in normal women. Hum Genet 1988; 16:1215. 2000; 107:343–9.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1225–1229 PHOTOBIOLOGY DOI 10.1111/j.1365-2133.2007.08209.x Melanocortin 1 receptor (MC1R) genotype influences erythemal sensitivity to psoralen–ultraviolet A photochemotherapy G. Smith, M.J.V. Wilkie, Y.Y. Deeni, P.M. Farr,* J. Ferguson, C.R. Wolf and S.H. Ibbotson Biomedical Research Centre, Ninewells Hospital and Medical School, Dundee DD1 9SY, U.K. *Department of Dermatology, Royal Victoria Infirmary, Newcastle upon Tyne NE1 4LP, U.K. Photobiology Unit, Ninewells Hospital and Medical School, Dundee DD1 9SY, U.K.

Summary

Correspondence Background The melanocortin 1 receptor (MC1R) is a highly polymorphic G Gillian Smith. protein-coupled receptor. Inheritance of various MC1R alleles has been associated E-mail: [email protected] with a red hair ⁄fair skin phenotype, increased incidence of skin cancer and altered sensitivity to ultraviolet (UV) radiation. Accepted for publication 27 June 2007 Objectives To investigate whether MC1R genotype influences erythemal sensitivity to psoralen–UVA photochemotherapy (PUVA) in patients with psoriasis and Key words other common skin diseases. erythema, genotype, MC1R, psoralen–ultraviolet Methods Patients (n = 111) about to start PUVA were recruited to the study. Ery- A photochemotherapy, ultraviolet radiation themal responses were assessed visually at 72 h and 96 h following PUVA by Conflicts of interest assessment of the minimal phototoxic dose (MPD). MC1R genotype was deter- None declared. mined by direct sequencing. Results Inheritance of the MC1R Arg151Cys allele was associated with a red hair phenotype (odds ratio 25Æ19, P =0Æ0004). In contrast, inheritance of the Val60 Leu and Arg163Gln SNPs was associated with increased PUVA erythemal sensitiv- ity (reduced MPD) 72 h following treatment in all patients (n = 111; Val60Leu 2 2 v =5Æ764, P =0Æ016; Arg163Gln v =5Æ469, P =0Æ019) and in a subset of 2 patients with psoriasis (n = 55; Val60Leu v =4Æ534, P =0Æ033; Arg163Gln v2 =7Æ298, P =0Æ007). Inheritance of two or more MC1R SNPs was also associ- ated with increased PUVA erythemal sensitivity (reduced MPD) in both patient groups (n = 111; v2 =8Æ166, P =0Æ017; n = 55; v2 =10Æ303, P =0Æ016). Conclusions Our data demonstrate that MC1R genotype influences PUVA erythemal sensitivity in patients with psoriasis and other common skin diseases.

The melanocortin 1 receptor (MC1R) is a seven transmem- most MC1R SNPs are rare or do not yet have a well-characterized brane G protein-coupled receptor, which activates G protein- phenotype, several variant MC1R alleles have been described dependent cyclic AMP (cAMP)-mediated signal transduction which have reduced function.4–6 Inheritance of these alleles, following stimulation by a variety of physiological and most notably alleles containing the Arg151Cys, Arg160Trp or external stimuli, including a-melanocyte stimulating hormone Asp294His SNPs, has been associated with a red hair and fair skin (a-MSH) and ultraviolet (UV) radiation.1 In human skin, phenotype (the RHC phenotype), where reduced MC1R activity MC1R is predominantly expressed in melanocytes, where it results in the production of relatively more red ⁄yellow pheo- regulates melanin production via cAMP-mediated activation of melanin than black ⁄brown eumelanin.7 Pheomelanin is less tyrosinase, the rate-limiting step in melanin biosynthesis.2 photoprotective than eumelanin and inheritance of the RHC Following synthesis in melanosomes, melanin migrates to MC1R alleles has been associated with increased susceptibility to neighbouring keratinocytes, where it forms a photoprotective melanoma8,9 and nonmelanoma skin cancers.10 Consistent with barrier.3 these epidemiological studies, recent data suggest that func- MC1R is a highly polymorphic gene, for which more than 60 tional MC1R protects cells from UV radiation-induced apoptosis single nucleotide polymorphisms (SNPs) resulting in noncon- and photodamage, while loss of MC1R function reduces the servative amino acid substitutions have been described.1 While antiapoptotic effect of a-MSH.11,12

2007 The Authors 1230 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1230–1234 MC1R genotype influences PUVA sensitivity, G. Smith et al. 1231

Although there is now compelling evidence that MC1R therapies again after PUVA, were documented in the Dundee genotype influences sensitivity to UV radiation in healthy vol- patients with psoriasis (n = 55). unteers, the influence of MC1R genotype on patient sensitivity to the commonly used psoralen–UVA photochemotherapy MC1R sequencing analysis (PUVA) has not been examined. PUVA has quite distinct clini- cal (e.g. erythemal sensitivity) and cellular effects compared Due to the number of polymorphisms in MC1R, direct with UV radiation, and it therefore cannot be assumed that sequencing was performed to assess MC1R genotype. Two the influence of MC1R genotype on PUVA erythemal sensitivity overlapping polymerase chain reaction (PCR) fragments, rep- will be similar to the effects on UV radiation sensitivity. In resenting the entire MC1R coding sequence, were amplified this study, we investigated the influence of individuality in using the MC1R-specific primers MC1RF1 5¢-tgt aaa acg acg MC1R expression on sensitivity to PUVA in patients with psori- gcc agt GCT GGC TGC CAA CCA GAC-3¢ and MC1RR1 5¢-cag asis and a variety of other common skin diseases. gaa aca gct atg acc GAA GAC CAC GAG GCA CAG-3¢, where lower case sequences represent M13 sequencing tags, and Patients and methods MC1RF2 5¢-tgt aaa acg acg gcc agt CTT CCT GGG CGC CAT C-3¢ and MC1RR2 5¢-cag gaa aca gct atg acc GGT CCG CGC TTC AAC ACT-3¢. PCR was performed in a 25-lL reaction vol- Clinical recruitment ume in the presence of 1 · Expand High Fidelity buffer, )1 )1 Patients (n = 111) about to start PUVA were recruited to the 2 mmol L MgCl2,0Æ77 mmol L dimethylsulphoxide, 0Æ2 ) ) study in Dundee (n = 91) and Newcastle (n = 20), following mmol L 1 deoxyribonucleotide triphosphates, 0Æ45 lmol L 1 approval by the Tayside Committee on Medical Research Eth- primers and 0Æ6 units Expand High Fidelity enzyme mix ics and Newcastle & North Tyneside Local Research Ethics (Roche, Lewes, U.K.). PCR was performed with an initial Committee. Written informed consent was obtained from all 5-min denaturation cycle at 95 C, followed by 34 PCR cycles study participants. Details of natural hair colour at age (denaturation: 95 C, 20 s; primer annealing: 55 C, 20 s; 21 years, eye colour, freckling tendency, skin phototype13 primer extension: 72 C, 45 s) and a final 5-min extension and diagnosis were documented. Two hours after administra- and annealing cycle. PCR products were visualized on 2% ethi- ) tion of oral methoxsalen (8-MOP; 25 mg m 2),14 a 10-mL dium bromide-stained agarose gels using a UV transillumina- venous blood sample was taken for genotyping and stored tor and purified prior to sequencing using ExoSAP-IT in ethylenediaminetetraacetic acid (EDTA) at )20 C. Geno- reagent.17 Direct sequencing of purified MC1R PCR products mic DNA was extracted from 200 lL of whole blood, was performed using M13 primers by the DNA Analysis Facil- using a QIAamp 96 spin blood kit (Qiagen, Crawley, U.K.) ity, Department of Molecular and Cellular Pathology, Nine- according to the manufacturer’s instructions, and stored wells Hospital, Dundee. MC1R polymorphisms were identified ) ) in 10 mmol L 1 Tris–HCl, 1 mmol L 1 EDTA at 4 C. UVA from comparison of individual MC1R sequences with the MC1R phototesting was also performed 2 h after 8-MOP ingestion consensus sequence (NCBI accession no. NM_002386). using a geometric dose series from a standard UVA fluores- Predicted MC1R haplotypes were constructed using SimHap cent lamp of the type used for PUVA (peak of fluorescence software.18 emission 352 nm). Statistical analysis Assessment of erythemal response to Correlations of erythemal sensitivity with MC1R genotype were psoralen–ultraviolet A photochemotherapy performed using Kruskal–Wallis tests for three or more dis- Erythemal responses were assessed visually at 72 h (and add- crete variables. Correlations of two discrete variables (e.g. itionally at 96 h in the Dundee patients15), by assessment of MC1R genotype with hair colour) were analysed using 2 · 2 the minimal phototoxic dose (MPD), defined as the UVA dose contingency tables and are presented as odds ratios (ORs) required to elicit just perceptible erythema in psoralen-sensi- with associated 95% confidence intervals (CIs), determined by tized skin. Fisher’s exact test. Statistical significance was taken as P <0Æ05. Psoralen–ultraviolet A photochemotherapy treatment and response Results

A PUVA regimen which employed a 70% MPD start dose 96h MC1R genotypes in patients undergoing and 20% (10% if erythema developed) increments at each psoralen–ultraviolet A photochemotherapy treatment was used. Patients were treated twice weekly until clearance ⁄minimal residual activity (MRA). Number of treat- In order to investigate the extent of genetically determined ments to clearance, reduction in psoriasis severity scoring sequence variation in MC1R in our study population (SEI: scaling, erythema, induration16) by 12 treatments and (n = 111), we used direct sequencing to analyse the entire the time to relapse, defined by the need to use antipsoriatic MC1R coding sequence. Fourteen MC1R SNPs were identified

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1230–1234 1232 MC1R genotype influences PUVA sensitivity, G. Smith et al.

Table 1 MC1R single nucleotide polymorphism (SNP) frequencies colours (OR 1Æ25, 95% CI 0Æ51–3Æ11, P =0Æ65). No associa- in patients undergoing psoralen–ultraviolet A photochemotherapy tions with hair colour, skin type, eye colour or freckling ten- (n = 111) dency were seen with the other MC1R SNPs examined, although this finding may be influenced by limited population MC1R SNP SNP frequency (%) size. Inheritance of two or more MC1R SNPs was also signifi- 2 Val60Leu 25 ⁄222 (11Æ3) cantly associated with red (v trend 8Æ322, P =0Æ004) or red ⁄ Asp Glu 1 ⁄222 (<1) 2 84 blonde hair (v trend 3Æ998, P =0Æ046), but not with eye col- Val92Met 21 ⁄222 (9Æ5) our, freckling tendency or skin type (Table 2). Arg142His 1 ⁄222 (<1) Arg151Cys 32 ⁄222 (14Æ4) Ile155Thr 2 ⁄222 (<1) Correlation of MC1R genotype with erythemal sensitivity Arg160Trp 19 ⁄222 (8Æ6) Arg163Gln 11 ⁄222 (4Æ9) Before correlating MC1R genotype with erythemal sensitivity Gln233Gln Silent 4 ⁄222 (1Æ8) in patients undergoing PUVA, we investigated whether indi- Val265Val Silent 1 ⁄222 (<1) vidual MC1R SNPs were most frequently inherited indepen- Asn279Ser 1 ⁄222 (<1) dently or in combination. MC1R SNPs which did not result in Asp His 9 ⁄222 (4Æ1) 294 amino acid changes, or which had an allele frequency <1%, Thr Thr Silent 1 ⁄222 (<1) 314 were excluded from this and any further analysis as it was not Ser316Ser Silent 1 ⁄222 (<1) felt that these SNPs would significantly influence MC1R pheno- type. Predicted haplotypes resulting from the inheritance

of the common MC1R SNPs Val60Leu, Val92Met, Arg151Cys,

(Table 1), one of which (Val265Val) has not previously Arg160Trp, Arg163Gln and Asp294His were constructed as been reported. Ten SNPs leading to amino acid substitutions described in Patients and methods. The most common haplo- were identified (Val60Leu, Asp84Glu, Val92Met, Arg142His, type, present in 51 of 111 patients (46%), was the MC1R con-

Arg151Cys, Ile155Thr, Arg160Trp, Arg163Gln, Asn279Ser, sensus reference sequence (NCBI accession no. NM_002386), Asp294His). The most common MC1R SNP in our patient while each of the next six most common haplotypes contained group was Arg151Cys (allele frequency 14Æ4%), while SNPs only a single MC1R SNP. Together, these haplotypes accounted

Asp84Glu, Arg142His and Asn279Ser were each found in only a for MC1R sequence variation in more than 92% of our study single individual. patients and we therefore chose initially to analyse the influ- ence of individual MC1R SNPs on PUVA sensitivity. In order to determine whether MC1R genotype influenced Correlation of MC1R genotype with red hair, fair skin erythemal sensitivity, we correlated inheritance of each of the and freckling six most common mis-sense MC1R SNPs with PUVA erythemal We correlated inheritance of each of the six most common sensitivity, assessed by visual MPD readings at 72 and 96 h. mis-sense MC1R SNPs with the previously described RHC phe- Inheritance of the MC1R Val60Leu and Arg163Gln SNPs was notype (Table 2). Only the MC1R Arg151Cys allele was associ- associated with increased erythemal sensitivity (reduced MPD) ated with red hair (OR 25Æ19, 95% CI 2Æ92–217Æ39, at 72 h, in the whole patient group (n = 111) and the subset

P =0Æ0004), an association which was also apparent compar- of patients with psoriasis (n = 55), while only the Arg163Gln ing red or blonde hair with all other hair colours (OR 3Æ69, SNP was associated with PUVA erythemal sensitivity at 96 h

95% CI 1Æ40–9Æ72, P =0Æ01). The MC1R Arg151Cys allele was (Table 3). 2 also associated with skin type (v trend 4Æ821, P =0Æ03) and Previous studies relating MC1R genotype to UV radiation freckling (OR 2Æ89, 95% CI 1Æ04–7Æ99, P =0Æ04), but not erythemal sensitivity have considered the phenotypic con- with eye colour, comparing blue eyes with all other eye sequences of inheriting multiple SNP combinations. We

Table 2 Arg151Cys and multiple MC1R single nucleotide polymorphisms (SNPs) influence the RHC phenotype

PUVA patients

(n = 111) Val60Leu Val92Met Arg151Cys Arg160Trp Arg163Gln Asp294His Two or more SNPs Red hair 1Æ800a (NS) 1Æ822a (NS) 25Æ190a (0Æ0004)1Æ364a (NS) 0Æ759a (NS) 0Æ188a (NS) 8Æ322b (0Æ004) Red ⁄blonde hair 0Æ710a (NS) 1Æ424a (NS) 3Æ69a (0Æ010)2Æ639a (NS) 1Æ373a (NS) 0Æ701a (NS) 3Æ998b (0Æ046) Blue eyes 0Æ914a (NS) 0Æ407a (NS) 1Æ25a (NS) 1Æ545a (NS) 0Æ313a (NS) 0Æ817a (NS) 1Æ437b (NS) Freckles 0Æ965a (NS) 1Æ364a (NS) 2Æ89a (0Æ041)0Æ467a (NS) 1Æ647a (NS) 0Æ476a (NS) 1Æ186b (NS) Skin type 2Æ178b (NS) 3Æ944b (NS) 4Æ821b (0Æ027)1Æ864b (NS) 1Æ995b (NS) 2Æ815b (NS) 11Æ426b (NS)

a b 2 PUVA, psoralen–ultraviolet A photochemotherapy. Odds ratio; v trend statistic. P-values are given in parentheses. Significant values (P <0Æ05) are given in bold. NS, not significant (P >0Æ05).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1230–1234 MC1R genotype influences PUVA sensitivity, G. Smith et al. 1233

Table 3 Val60Leu, Arg163Gln and multiple MC1R single nucleotide have been described which have been associated with constitu- polymorphisms (SNPs) influence erythemal sensitivity tive pigmentation (the RHC phenotype), altered UV radiation sensitivity and skin cancer incidence.1 PUVA is commonly used Whole patient Patients with to treat common skin diseases including psoriasis, but has quite group (n = 111) psoriasis (n = 55) distinct clinical and cellular effects compared with UV radia- 19 v2 statistic P-value v2 statistic P-value tion. The influence of MC1R genotype on sensitivity to PUVA has not previously been investigated. In this study, we have Visual MPD at 72 h shown that MC1R genotype influences PUVA erythemal sensitiv- Val60Leu 5Æ764 0Æ016 4Æ534 0Æ033 ity in patients with psoriasis and other common skin diseases. Val92Met 0Æ398 NS 11Æ682 NS Arg151Cys 1Æ965 NS 1Æ617 NS Consistent with previous literature, we demonstrated that Arg160Trp 2Æ443 NS 1Æ307 NS inheritance of the MC1R Arg151Cys SNP correlated with the 4–6 Arg163Gln 5Æ469 0Æ019 7Æ298 0Æ007 RHC phenotype in our patient group. Although the MC1R Asp His 0Æ087 NS 0Æ596 NS 294 Arg151Cys SNP has been most consistently associated with the Æ Æ Æ Æ Two or more SNPs 8 166 0 017 10 303 0 016 RHC phenotype, similar phenotypic associations have been Visual MPD at 96 h 1 reported for the Arg160Trp and Asp294His SNPs. In contrast to Val60Leu 1Æ091 NS 1Æ081 NS these observations, we did not find associations between inheri- Val92Met 0Æ165 NS 2Æ026 NS Arg151Cys 2Æ800 NS 0Æ611 NS tance of these SNPs and the RHC phenotype in our patient Arg160Trp 2Æ977 NS 1Æ136 NS group. The reason for this discrepancy is not clear, although it Arg163Gln 5Æ221 0Æ022 5Æ387 0Æ020 may simply reflect differences in the relative frequencies of the Asp294His 0Æ781 NS 1Æ690 NS various MC1R alleles. In addition, recruitment to several previous Æ Æ Two or more SNPs 3 035 NS 4 654 NS studies described populations with an over-representation of MPD, minimal phototoxic dose. Significant values (P <0Æ05) are redheads, whereas our study group represented patients referred given in bold. NS, not significant (P >0Æ05). for PUVA with unselected hair and skin types. While inheritance of the MC1R Arg151Cys SNP was clearly associated with the RHC phenotype in our patient group, it therefore additionally investigated whether inheritance of two did not appear to influence PUVA erythemal sensitivity. In or more common mis-sense MC1R SNPs influenced PUVA contrast, we demonstrated that inheritance of either the MC1R erythemal sensitivity and demonstrated an association with Val60Leu or Arg163Gln SNPs was associated with increased ery- increased erythemal sensitivity (reduced MPD) at 72 h in all themal sensitivity, as assessed by MPD, in both the whole patients (n = 111; v2 =8Æ166, P =0Æ017) and in the subset patient group and in the subgroup of patients with psoriasis. 2 of patients with psoriasis (n = 55; v =10Æ303, P =0Æ016) The phenotypes associated with the MC1R Val60Leu or

(Table 3). Arg163Gln SNPs are less well characterized than Arg151Cys, although both have been associated with reduced receptor function.20,21 The Val Leu and Arg Gln SNPs have previ- Correlation of MC1R genotype with treatment response 60 163 ously been described as ‘weak’ RHC alleles, where ORs for the To investigate whether MC1R genotype influenced PUVA treat- RHC phenotype are an order of magnitude lower than for the 1 ment response in patients with psoriasis, we correlated inheri- Arg151Cys SNP. This observation, together with our own tance of each of the most common mis-sense MC1R SNPs with data, suggests that MC1R genotype independently influences clearance ⁄MRA, percentage reduction in SEI scores assessed RHC phenotype and PUVA erythemal sensitivity. Consistent over 12 PUVA treatments, total treatment numbers to clear- with this hypothesis, Healy et al. have previously demonstrated ance and time to relapse. No significant associations were seen that the MC1R genotype is a determinant of sun sensitivity in comparing the inheritance of single or multiple MC1R SNPs individuals without red hair.22 (data not shown), suggesting that the RHC phenotype and We did not find any associations between MC1R genotype acute PUVA erythemal sensitivity are not predictive markers of and PUVA treatment outcomes in patients with psoriasis, treatment response. assessed according to percentage reduction in SEI scores over 12 treatments, achievement of clearance ⁄MRA, total treatment Discussion numbers to clearance or time to relapse. As high-dose PUVA has been consistently associated with significantly increased Genetically determined variation in the expression and activity skin cancer risk,23,24 however, it will be of future interest to of the highly polymorphic MC1R gene has been shown to influ- investigate whether cancer risk in patients exposed to PUVA is ence susceptibility to melanoma8,9 and nonmelanoma skin can- modified by MC1R genotype. cers.10 One of the best characterized functions of MC1R is the regulation of pigmentation, where MC1R activates tyrosinase to Acknowledgments direct melanin biosynthesis, resulting in the generation of a photoprotective melanin ‘barrier’, relatively resistant to chal- We thank the Specialist Registrars, Photobiology Technicians lenge by UV radiation.1 Various reduced activity MC1R SNPs and Research Nurse Susan Yule in the Photobiology Unit,

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1230–1234 1234 MC1R genotype influences PUVA sensitivity, G. Smith et al.

Ninewells Hospital and Medical School for help with patient 12 Kadekaro AL, Kavanagh R, Kanto H et al. Alpha-melanocortin and recruitment. We acknowledge financial support from the Med- endothelin-1 activate antiapoptotic pathways and reduce DNA ical Research Council (grant reference G0000281). damage in human melanocytes. Cancer Res 2005; 65:4292–9. 13 Fitzpatrick TB. The validity and practicality of sun-reactive skin types I through VI. Arch Dermatol 1988; 124:869–71. References 14 Mclelland J, Fisher C, Farr PM et al. The relationship between plasma psoralen concentration and psoralen-UVA erythema. Br J 1 Garcia-Borron JC, Sanchez-Laorden BL, Jimenez-Cervantes C. Mela- Dermatol 1991; 124:585–90. nocortin-1 receptor structure and functional regulation. Pigment Cell 15 Ibbotson SH, Farr PM. The time-course of psoralen ultraviolet A Res 2005; 18:393–410. (PUVA) erythema. J Invest Dermatol 1999; 113:346–50. 2 Busca R, Ballotti R. Cyclic AMP a key messenger in the regulation 16 Cameron H, Dawe RS, Yule S et al. A randomized, observer-blinded of skin pigmentation. Pigment Cell Res 2000; 13:60–9. trial of twice vs. three times weekly narrowband ultraviolet B 3 Kadekaro AL, Kavanagh RJ, Wakamatsu K et al. Cutaneous photobi- phototherapy for chronic plaque psoriasis. Br J Dermatol 2002; ology. The melanocyte vs. the sun: who will win the final round? 147:973–8. Pigment Cell Res 2003; 16:434–47. 17 Dugan KA, Lawrence HS, Hares DR et al. An improved method for 4 Box NF, Wyeth JR, O’Gorman LE et al. Characterization of melano- post-PCR purification for mtDNA sequence analysis. J Forensic Sci cyte stimulating hormone receptor variant alleles in twins with red 2002; 47:811–18. hair. Hum Mol Genet 1997; 6:1891–7. 18 McCaskie PA, Carter KW, McCaskie SR, Palmer LJ. The effect of 5 Healy E, Jordan SA, Budd PS et al. Functional variation of MC1R missing data on linkage disequilibrium mapping and haplotype alleles from red-haired individuals. Hum Mol Genet 2001; 10:2397– association analysis in the GAW14 simulated datasets. BMC Genet 402. 2005; 6 (Suppl. 1):S151. 6 Valverde P, Healy E, Jackson I et al. Variants of the melanocyte- 19 Man I, Dawe RS, Ferguson J, Ibbotson SH. An intraindividual study stimulating hormone receptor gene are associated with red hair of the characteristics of erythema induced by bath and oral meth- and fair skin in humans. Nat Genet 1995; 11:328–30. oxsalen photochemotherapy and narrowband ultraviolet B. Photo- 7 Rees J. Plenty new under the sun. J Invest Dermatol 2006; 126:1691– chem Photobiol 2003; 78:55–60. 2. 20 Ringholm A, Klovins J, Rudzish R et al. Pharmacological character- 8 Palmer JS, Duffy DL, Box NF et al. Melanocortin-1 receptor poly- ization of loss of function mutations of the human melanocortin 1 morphisms and risk of melanoma: is the association explained receptor that are associated with red hair. J Invest Dermatol 2004; solely by pigmentation phenotype? Am J Hum Genet 2000; 66:176– 123:917–23. 86. 21 Schioth HB, Phillips SR, Rudzish R et al. Loss of function mutations 9 Kennedy C, ter Huurne J, Berkhout M et al. Melanocortin 1 recep- of the human melanocortin 1 receptor are common and are associ- tor (MC1R) gene variants are associated with an increased risk for ated with red hair. Biochem Biophys Res Commun 1999; 260:488–91. cutaneous melanoma which is largely independent of skin type 22 Healy E, Flannagan N, Ray A et al. Melanocortin-1-receptor gene and hair colour. J Invest Dermatol 2001; 117:294–300. and sun sensitivity in individuals without red hair. Lancet 2000; 10 Bastiaens MT, ter Huurne JA, Kielich C et al. Melanocortin-1 recep- 355:1072–3. tor gene variants determine the risk of nonmelanoma skin cancer 23 Stern RS, Laird N. The carcinogenic risk of treatments for severe independently of fair skin and red hair. Am J Hum Genet 2001; psoriasis. Photochemotherapy follow-up study. Cancer 1994; 68:884–94. 73:2759–64. 11 Bohm M, Wolff I, Scholzen TE et al. Alpha-melanocyte-stimulating 24 Lever LR, Farr PM. Skin cancers or premalignant lesions occur in hormone protects from ultraviolet radiation-induced apoptosis and half of high-dose PUVA patients. Br J Dermatol 1994; 131:215–19. DNA damage. J Biol Chem 2005; 280:5795–802.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1230–1234 THERAPEUTICS DOI 10.1111/j.1365-2133.2007.08250.x Isotretinoin therapy and the incidence of acne relapse: a nested case–control study L. Azoulay,* D. Oraichi and A. Be´rard* *Faculty of Pharmacy, University of Montreal, Montreal, Quebec, Canada Research Center, CHU Sainte-Justine, Centre de Recherche´, 3175, Chemin de la Coˆte-Ste-Catherine, Montreal, Quebec, Canada H3T 1C5

Summary

Correspondence Background Previous studies on predictors of acne relapse in patients treated with Anick Be´rard. isotretinoin had either small sample sizes, short follow-up periods, or lacked E-mail: [email protected] population-based data. Objectives To identify and quantify predictors of acne relapse, and predictors of Accepted for publication 2 August 2007 receiving a second isotretinoin treatment. Methods Using the Re´gie de l’Assurance Maladie du Que´bec (RAMQ) and Quebec’s Key words hospital discharge (Med-E´cho) administrative databases, a population-based acne, isotretinoin, predictors, relapse cohort of 17 351 first-time isotretinoin users was assembled between 1984 and 2003. A nested case–control analysis was performed to determine predictors of Conflicts of interest acne relapse (as defined by receiving an antiacne medication). A second nested Laurent Azoulay was the recipient of a doctoral bursary from the Fonds de la recherche en sante´ du case–control analysis was performed to determine predictors of receiving a sec- Que´bec (FRSQ). Dr Anick Be´rard is the recipient ond isotretinoin treatment. The index date of cases was the calendar date of dis- of a career award from the CIHR ⁄Canadian pensing an antiacne medication (isotretinoin or other). Five controls were Foundation for Health, and is on the endowment matched to each case on follow-up time. Rate ratios were estimated using condi- research Chair of the Famille Louis-Boivin on tional logistic regression. ‘Medications, Pregnancy and Lactation’ at the Faculty of Pharmacy of the University of Results A total of 7100 (41%) subjects experienced an acne relapse. These were Montreal. matched to 35 500 controls. Being male, under 16 years of age and living in an urban area, and receiving isotretinoin cumulative doses greater than 2450 mg and an isotretinoin treatment longer than 121 days were statistically associated (P <0Æ05) with acne relapse. The publishing of the different Canadian acne guidelines had no impact on the incidence of acne relapse (P >0Æ05). A total of 4443 (26%) subjects required a second isotretinoin treatment. These were matched to 22 215 controls. There was a greater probability of receiving a sec- ond isotretinoin treatment after the publishing of the Canadian acne guidelines (P <0Æ05). Conclusion A relatively high rate of subjects experienced an acne relapse after an isotretinoin treatment.

Since its introduction in the market more than 20 years ago, those who will not. As such, determining predictors of experi- isotretinoin has revolutionized the treatment of severe nodular encing an acne relapse may be of great prognostic value to acne. However, despite its remarkable effectiveness, isotretino- physicians who treat patients with acne. This is particularly in has important side-effects. It is a teratogen if taken during important in countries such as Canada which have no restric- the first trimester of pregnancy.1 In Canada, a pregnancy pre- tions on which physicians may prescribe isotretinoin. Since its vention programme was implemented in 1988, but failed to introduction in the U.S.A. in 1982 and in Canada the follow- prevent isotretinoin-exposed pregnancies.2 Other side-effects ing year, several observational studies have examined predic- include mucocutaneous, ophthalmological, neuromuscular and tors of experiencing an acne relapse after being treated gastrointestinal problems.3 In addition, an association between initially with isotretinoin. The relapse rates varied between isotretinoin and depression has recently been found (L. Azou- studies, ranging from 5Æ6% to 65Æ4%.4–15 There are several lay et al., article in press). possible explanations for this large discrepancy. Firstly, Given isotretinoin’s safety profile, it is important to deter- these studies had small sample sizes, ranging between 52 and mine which patients will benefit most from this treatment vs. 299 patients.4–15 Secondly, some studies had short follow-up

2007 The Authors 1240 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 Isotretinoin therapy and acne relapse, L. Azoulay et al. 1241 periods, and thus may have underestimated the relapse rate. defined as filling consecutive prescriptions with fewer than Finally, relapse was not defined consistently across the differ- 30 days between renewals. The subjects were aged between ent studies. 13 and 45 years on the first day of the first isotretinoin pre- Given the limitations described above and lack of popula- scription within the study period. They were required to be tion-based data on the subject, we sought to: (i) determine continuously insured by the RAMQ drug plan for at least the characteristics of subjects initially treated with isotretinoin 12 months before and at least 6 months after the first day of experiencing an acne relapse as defined by requiring further the first isotretinoin prescription. Based on the criteria above, treatment with an antiacne medication (either a second isotre- we identified 17 351 first-time isotretinoin users. A detailed tinoin treatment or another antiacne medication); (ii) identify description of the cohort has been described previously.21 and quantify predictors of experiencing an acne relapse; and (iii) identify and quantify predictors of receiving a second iso- Study design tretinoin treatment. Two nested case–control analyses were conducted using a SAS 22 Methods incidence density program that was adapted for the purposes of the present study. The first nested case–control analysis determined predictors of receiving an antiacne medication Data sources (isotretinoin or other) after being initially treated with isotre- Data were obtained from the Re´gie de l’Assurance Maladie du tinoin. The second nested case–control analysis determined Que´bec (RAMQ) and Quebec’s hospital discharge (Med-E´cho) predictors of receiving a second isotretinoin treatment. With administrative databases. All Quebec residents are covered by incidence density sampling, a subject could serve as a control the RAMQ for medical services. Prior to 1 January 1997, the for more than one case. Moreover, cases could serve as con- RAMQ drug plan covered those who were 65 years and older, trols before they became cases. For both nested case–control and welfare recipients and their children. After 1 January analyses, controls were matched to cases on the time since the 1997, the RAMQ drug plan was changed to also include end of their first isotretinoin treatment. Five controls were workers and their spouses ⁄children who do not have access to randomly matched to each case. a private medication insurance programme. The RAMQ drug plan covers approximately 50% of Quebec residents.16 Follow-up The medical and pharmaceutical databases of the RAMQ were linked by a unique patient identification number. The The date of entry in the cohort for the first nested case– medical claims database includes information on the date and control study started the day after the end of the first iso- type of services received, and diagnoses are classified accord- tretinoin treatment. This is because it was possible to ing to the International Classification of Diseases, 9th revision receive an antiacne medication at any time after the first (ICD-9).17 The pharmaceutical claims database contains infor- isotretinoin treatment. As for the second nested case–control mation on the date medications were dispensed, formulations, study, the date of entry in the cohort for all subjects started doses, duration of prescriptions and quantities dispensed. only 30 days after the end of their first isotretinoin treat- Medications prescribed during hospitalizations are not ment. This is because an isotretinoin treatment was defined included in the database. Both the medical and pharmaceutical as filling consecutive prescriptions with fewer than 30 days claims databases have been validated and shown to be accurate between renewals. Therefore by design, there were no iso- for certain diagnoses.18,19 tretinoin prescriptions in the 30 days after the end of the The unique patient identification number was also used to first treatment. This was done to avoid immortal bias,23,24 link the RAMQ to the Med-E´cho administrative databases. where the person-time at which subjects were not at risk of Med-E´cho contains hospitalization data on all Quebec resi- receiving an isotretinoin prescription was excluded. Subjects dents. These data include patient demographic information, in both nested case–control analyses were followed until physician characteristics, admission diagnostic, length of stay, they became cases, were no longer covered by the RAMQ as well as all services received during the hospitalization. Med- drug plan, or until 30 June 2003, whichever came first. ical diagnoses recorded in Med-E´cho have been shown to be valid and precise.20 Cases and controls The study protocol was approved by the CHU Sainte-Justine Ethics Committee and the Commission d’acce`sa` l’information Cases for the first nested case–control analysis consisted of du Que´bec. subjects who experienced an acne relapse and required further treatment with an antiacne medication (isotretinoin or other). For these cases, the index date was the calendar date of dis- Study cohort pensing the antiacne medication (isotretinoin or other). Eligi- We identified all Quebec residents who received a first isotre- ble controls were subjects who were still present in the cohort tinoin treatment between 1 January 1984 and 30 June 2003 at the time of the cases’ index date and had not received any within the RAMQ drug plan. An isotretinoin treatment was antiacne medications (isotretinoin or other) prior to that date.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 1242 Isotretinoin therapy and acne relapse, L. Azoulay et al.

The antiacne medications considered were isotretinoin, treti- Guidelines published in 1995.25 These guidelines recom- noin, tazarotene, benzoyl peroxide, topical clindamycin, topi- mended up to three courses of systemic antibiotics before cal erythromycin, systemic doxycycline, systemic tetracycline, considering isotretinoin. The third was the Canadian Consen- systemic minocycline, systemic erythromycin, systemic azi- sus Guidelines on Treatment of Acne and Prevention of Scar- thromycin, and acne-specific oral contraceptives (Diane-35). ring published in 2000. These guidelines recommended that Adapalene and benzoyl peroxide ⁄erythromycin combination isotretinoin be prescribed primarily to patients with scarring were not considered since they are not reimbursed by the acne, regardless of severity.26,27 RAMQ drug plan. Given that systemic antibiotics are pre- scribed for a variety of conditions, it was necessary to ascer- Statistical analysis tain that those considered as acne treatments were truly for acne. Therefore, an acne diagnosis (ICD-9 code: 706.1) must Descriptive statistics were used to summarize the characteris- have been present in the database within 30 days of the dis- tics of the study population. Kaplan–Meier survival curves pensing of the systemic antibiotic to be considered an antiacne were constructed to model the time from the end of the first treatment. In the event of a missing acne diagnosis, we used isotretinoin treatment until receipt of an antiacne medication dermatological procedure codes recorded within 30 days of (isotretinoin or other). Rate ratios (RR) together with 95% the dispensing of the medication. confidence intervals (CI) were estimated using conditional Cases for the second nested case–control analysis consisted logistic regression. Crude and adjusted RRs were calculated for of subjects who experienced an acne relapse severe enough to both nested case–control analyses. All analyses were conducted require a second isotretinoin treatment. As such, their index using SAS version 8.2 (SAS Institute Inc, Cary, NC, U.S.A.). date was the calendar date of the dispensing of the second iso- tretinoin treatment. Eligible controls were subjects who were Results still present in the cohort at the time of the cases’ index date and had not received any isotretinoin prescriptions prior to Predictors of receiving an antiacne medication that date. Of the 17 351 first-time isotretinoin users in the cohort, a total of 7100 (40Æ9%, 95% CI 40Æ4–41Æ4) subjects (cases) Potential predictors required further treatment with an antiacne medication (iso- Potential predictors for both nested case–control analyses were tretinoin or other). There was a total of 36 911 person-years assessed in four different time periods. The time periods were: of follow-up, where the rate of receiving an antiacne medica- (i) at the index date; (ii) between the end of the first isotre- tion was 192 ⁄1000 persons per year. Half of the cohort tinoin treatment and index date (follow-up period); (iii) dur- received an antiacne medication within 58Æ8 months ing the first isotretinoin treatment; and (iv) in the 12 months (4Æ9 years) after the end of their isotretinoin treatment prior to the first isotretinoin treatment. The predictors consid- (Fig. 1). ered in the four time periods related to subject sociodemo- A total of 35 500 controls were matched to the 7100 cases. graphic information (age, gender, place of residence, adherent The characteristics of the cases and controls were similar of the RAMQ drug plan ⁄welfare recipient), healthcare and (Table 1). Isotretinoin was dispensed to 4011 (56Æ5%) sub- medication utilization (visits to the physician, emergency jects, whereas 3089 (43Æ5%) required other antiacne medica- department visits, all-cause hospitalizations, and number of tions (Table 2). Out of the latter group, 432 (14%) received different types of medications) as well as prescriber informa- isotretinoin during the remainder of their follow-up. tion (specialty, gender of prescriber and whether isotretinoin was prescribed by more than two physicians during the first treatment). Other predictors for requiring further treatment 1 with an antiacne medication (isotretinoin or other) inclu- ded median duration of the first isotretinoin treatment 0·8 (‡ 121 days) and isotretinoin cumulative dose. The latter was 0·6 entered as quartiles in the models. A cumulative dose of 2450 mg was considered equivalent to a 2 months’ supply of 0·4 isotretinoin, assuming an average dosage of 46 mg daily.21 Since our cohort spanned a 20-year period, we also 0·2 adjusted the models on the calendar time periods of the differ- ent Canadian programmes and guidelines promulgated over

Proportion of patients in the cohort 0 the years. We considered the following three programmes and 0 50 100 150 200 250 guidelines. The first was the Pregnancy Prevention Program Time (months) (PPP) implemented in October 1988, whose goal was to prevent isotretinoin-exposed pregnancies in women of child- Fig 1. Kaplan–Meier curve of first-time isotretinoin users requiring bearing age. The second was the Canadian Acne Treatment further treatment with an antiacne medication (isotretinoin or other).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 Isotretinoin therapy and acne relapse, L. Azoulay et al. 1243

Table 1 Characteristics of cases and controls for the antiacne Predictors of receiving a second isotretinoin treatment medication analysis Of the 17 351 first-time isotretinoin users in the cohort, 12 908 (74Æ4%) had one treatment, 2997 (17Æ3%) had two Cases Controls Characteristics (n = 7100) (n = 35 500) treatments, and 1446 (8Æ3%) had three or more isotretinoin treatments during the follow-up period. Thus a total of 4443 Age (years), mean (SD) 25Æ2(8Æ1) 25Æ5(7Æ9) Males, n (%) 3770 (53Æ1) 17 833 (50Æ2) (25Æ6%, 95% CI 25Æ1–26Æ0) had two or more isotretinoin Welfare recipients, n (%) 3993 (56Æ2) 16 982 (52Æ2) treatments during the study period. There was a total of Urban dwellers, n (%) 5732 (80Æ7) 26 678 (78Æ2) 48 802 person-years of follow-up, where the rate of receiving a second isotretinoin treatment was 91 ⁄1000 persons per year. Nearly all subjects who received a second isotretinoin treat- ment did so within 25Æ4 months after the end of their first Table 2 Type of antiacne medications dispensed (n = 7100) treatment (Fig. 2). There were 4443 cases and 22 215 controls in the nested n (%) case–control analysis. The characteristics of cases and controls Systemic retinoids were similar (Table 4). Table 5 presents crude and adjusted Isotretinoin 4011 (56Æ5) RRs of predictors of receiving a second isotretinoin treatment. Systemic antibiotics In adjusted analyses, subjects under 16 years of age, as well as Minocycline 678 (9Æ6) males were more likely to receive a second isotretinoin treat- Tetracycline 264 (3Æ7) ment. The chances of receiving a second isotretinoin treatment Doxycycline 185 (2Æ6) Erythromycin 136 (1Æ9) increased after the implementation of the PPP and the publish- Clindamycin 94 (1Æ3) ing of the different guidelines. Subjects who had received iso- Azithromycin 90 (1Æ3) tretinoin cumulative doses greater than 2450 mg compared Antimicrobial and hormonal agents with those who received cumulative doses less than 2450 mg Benzoyl peroxide 406 (5Æ7) during their first treatment were less likely to receive a second Acne-specific oral contraceptives 176 (2Æ5) isotretinoin treatment. Likewise, subjects who had an isotretin- Topical retinoids oin treatment lasting ‡ 121 days were less likely to receive a Tretinoin 468 (6Æ6) Tazarotene 0 (0) second isotretinoin treatment. Topical antibiotics Erythromycin 366 (5Æ2) Discussion Clindamycin 1 (0Æ01) Treatment combinations In a cohort of first-time isotretinoin users, 41% of subjects Systemic antibiotic plus topical antibiotic 90 (1Æ3) experienced an acne relapse necessitating further treatment Systemic antibiotic plus topical retinoid or 78 (1Æ1) with an antiacne medication (isotretinoin or other). Twenty- benzoyl peroxide Two topical agents 46 (0Æ6) six per cent of subjects experienced a relapse severe enough to Two systemic antibiotics 6 (0Æ1) receive a second isotretinoin treatment at some point during Acne-specific oral contraceptives plus systemic 5(0Æ1) their follow-up. Younger age, male gender, living in an urban antibiotic area, and several healthcare utilization variables were associ- ated with receiving an antiacne medication (isotretinoin or other). Guidelines published over the years had no impact on the incidence of acne relapse. However, subjects were more Table 3 displays crude and adjusted RRs of receiving an likely to receive isotretinoin after the publishing of these antiacne medication (isotretinoin or other) after being ini- guidelines. tially treated with isotretinoin. Male subjects and those Forty-one per cent of the cohort experienced an acne relapse under 16 years of age were more likely to receive an anti- necessitating further treatment with an antiacne medication. acne medication. Subjects living in urban areas were also This figure is similar to those published previously.12–14 more likely to receive an antiacne medication. The imple- Male subjects and those under the age of 16 were more likely mentation of the PPP and publishing of other guidelines to require further treatment with an antiacne medication. were not statistically associated with receiving an antiacne Compared with subjects living in rural areas, those living in medication. Subjects who had received isotretinoin cumula- urban areas were more likely to require further treatment. tive doses greater than 2450 mg compared with those who A possible explanation for this finding is that subjects living received cumulative doses less than 2450 mg during their in urban areas had a greater access to physicians than those first treatment were less likely to receive another antiacne living in rural areas, and were thus more likely to consult a medication. Those whose first isotretinoin treatment lasted physician and receive an antiacne medication. Subjects who ‡ 121 days were also less likely to receive another antiacne visited an emergency department or who were hospitalized medication. after the end of their isotretinoin treatment were less likely to

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 1244 Isotretinoin therapy and acne relapse, L. Azoulay et al.

Table 3 Predictors of receiving an antiacne Crude rate ratio Adjusted rate ratioa treatment after being treated initially with (95% CI) (95% CI) isotretinoin At index date <16 years of age 1Æ31 (1Æ19–1Æ44) 1Æ40 (1Æ26–1Æ54) Male gender 0Æ89 (0Æ85–0Æ94) 1Æ14 (1Æ07–1Æ20) Urban dweller 1Æ16 (1Æ09–1Æ24) 1Æ09 (1Æ01–1Æ16) Welfare recipient 1Æ15 (1Æ09–1Æ21) 1Æ01 (0Æ94–1Æ09) Time periods 1 January 1984–30 September 30 1988 1Æ00 (Reference) 1Æ00 (Reference) 1 October 1988–31 December 1994 1Æ44 (1Æ30–1Æ58) 1Æ07 (0Æ97–1Æ19) 1 January 1995–31 December 1999 1Æ22 (1Æ11–1Æ33) 1Æ02 (0Æ92–1Æ13) 1 January 2000–30 June 2003 1Æ14 (1Æ05–1Æ24) 1Æ04 (0Æ94–1Æ16) Between the end of the first treatment and index date Number of different types of medications 1Æ09 (1Æ09–1Æ10) 1Æ08 (1Æ07–1Æ09) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 14Æ00 (3Æ71–4Æ32) 3Æ83 (3Æ53–4Æ15) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ13 (1Æ06–1Æ22) 0Æ80 (0Æ74–0Æ87) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ46 (1Æ33–1Æ61) 0Æ86 (0Æ76–0Æ96) During the first isotretinoin treatment Number of different types of medications 1Æ04 (1Æ02–1Æ05) 0Æ98 (0Æ96–0Æ99) At least one antiacne medication 1Æ82 (1Æ66–1Æ99) 1Æ71 (1Æ55–1Æ89) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 10Æ93 (0Æ86–1Æ01) 0Æ89 (0Æ82–0Æ97) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 10Æ96 (0Æ89–1Æ05) 1Æ00 (0Æ92–1Æ10) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ09 (0Æ93–1Æ26) 0Æ91 (0Æ77–1Æ07) Age of treating physician (years) 1Æ00 (1Æ00–1Æ01) 0Æ99 (0Æ99–1Æ00) Treated by male physician 1Æ03 (0Æ97–1Æ10) 1Æ03 (0Æ96–1Æ10) Treated by dermatologist 1Æ03 (0Æ98–1Æ08) 1Æ09 (1Æ03–1Æ16) ‡ 2 isotretinoin prescribers 1Æ00 (0Æ90–1Æ10) 1Æ09 (0Æ98–1Æ22) Cumulative dose < 2450 mg 1Æ00 (Reference) 1Æ00 (Reference) 2450–4840 mg 0Æ76 (0Æ71–0Æ81) 0Æ82 (0Æ76–0Æ88) 4840–7584 mg 0Æ62 (0Æ58–0Æ67) 0Æ75 (0Æ68–0Æ83) ‡ 7584 mg 0Æ52 (0Æ49–0Æ56) 0Æ63 (0Æ57–0Æ70) Treatment duration ‡ 121 days 0Æ65 (0Æ61–0Æ68) 0Æ83 (0Æ76–0Æ89) 12 months prior to the first isotretinoin treatment Number of different types of medications 1Æ00 (1Æ00–1Æ00) 1Æ00 (1Æ00–1Æ00) At least one antiacne medication 1Æ34 (1Æ27–1Æ41) 1Æ36 (1Æ29–1Æ44) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ27 (1Æ13–1Æ44) 0Æ86 (0Æ76–0Æ99) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ05 (0Æ99–1Æ11) 0Æ98 (0Æ92–1Æ05) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ11 (1Æ03–1Æ20) 0Æ94 (0Æ86–1Æ02)

aAdjusted for the covariates in the table. CI, confidence interval.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 Isotretinoin therapy and acne relapse, L. Azoulay et al. 1245

meant to increase physician awareness of isotretinoin’s 1 potential side-effects so as to restrict its use to patients who 0·9 most need it. It appears that such guidelines had no impact on prescribing patterns. This could be due to the increase in 0·8 prescribing of isotretinoin to subjects with mild or moderate acne,29 or the fact that it is increasingly being used as a 21,30,31 0·7 first-line agent. There are a number of differences between the present 0·6 study and the ones published previously. Firstly, the present study is the largest to date with a sample size exceeding

Proportion of patients in the cohort 0·5 17 000 first-time isotretinoin users. Previous studies had smal- 4–15 0 50 100 150 200 250 ler sample sizes. Secondly, the majority of previous studies Time (months) recruited patients from specialized settings such as dermatol- ogy clinics and hospitals.4,5,7–9,12,13,15 Although the motiva- Fig 2. Kaplan–Meier curve of first-time isotretinoin users requiring a tion for using patients from these settings was to facilitate second isotretinoin treatment. recruitment, it does limit the generalizability of their results. Although the subjects included in the present study came from restricted socioeconomic backgrounds, they were treated by Table 4 Characteristics of cases and controls for the isotretinoin analysis physicians who also treat patients from higher socioeconomic backgrounds. Thus, all residents of the province of Quebec have access to the same medical care regardless of their socio- Cases Controls Characteristics (n = 4443) (n = 22 215) economic status. As a result, it is unlikely that subjects from lower socioeconomic backgrounds would be treated differ- Age (years), mean (SD) 24Æ8(7Æ9) 25Æ3(7Æ9) Males, n (%) 2235 (50Æ3) 10 792 (48Æ6) ently from those subjects from higher socioeconomic back- Adherents, n (%) 2518 (56Æ7) 12 287 (55Æ3) grounds given the fact that both groups were reimbursed for Urban dwellers, n (%) 3534 (79Æ5) 17 383 (78Æ9) their medications. Thus, we believe our cohort is representa- tive of the acne population. Thirdly, we were able to follow subjects for up to 20 years since the end of their first isotre- tinoin treatment. This long-term follow-up is essential for an experience an acne relapse. It is possible that these subjects accurate detection of relapses. Fourthly, administrative data- had other serious medical conditions that may have prevented bases contain complete information on healthcare services them from receiving an antiacne medication. Furthermore, received and medications dispensed. Thus, we were able to subjects who had isotretinoin treatment durations longer than control for important variables that are seldom collected in the median (121 days) were less likely to experience an acne field studies. In addition, we considered predictors of receiv- relapse, as well as those who had cumulative doses greater ing an antiacne medication in four different time periods, than 2450 mg. This is consistent with other studies that have from 12 months prior to the first isotretinoin treatment to the investigated cumulative dose on the incidence of acne index date. relapse.6,10 Despite the strengths mentioned above, the present study Interestingly, the acne guidelines published over the years does have some limitations inherent in the use of adminis- had no impact on the incidence of acne relapse. This suggests trative databases. Administrative databases provide informa- one of two possibilities. The first is that these guidelines may tion only on those subjects who filled their prescriptions. not have been adequate in preventing relapses in subjects ini- Therefore, the number of subjects who experienced a relapse tially treated with isotretinoin. As such, future research should and did not consult a physician is unknown. The number of determine the effectiveness of such guidelines on patient out- subjects who were prescribed an antiacne medication but did comes. The second is that physicians did not fully adhere to not fill them at the pharmacy is also unknown, as is the these guidelines. If true, it may be due to lack of awareness, number of subjects who used over-the-counter (OTC) prod- lack of familiarity with new guidelines, disagreement, or iner- ucts to treat their acne. Although these are possibilities, they tia of previous practice.28 are unlikely for the following two reasons. Firstly, it is rea- Twenty-six per cent of the cohort required a second iso- sonable to assume that subjects who were previously treated tretinoin treatment at some point during the follow-up. This for their acne would seek further treatment had they experi- estimate is similar to those published previously.8,10 Most enced an acne relapse. Not consulting a physician for an were dispensed isotretinoin within the first 2 years after the acne relapse may be indicative that the relapse was not end of their first treatment. In contrast to the first analysis, severe enough to warrant further intervention. Secondly, all subjects were more likely to receive a second isotretinoin subjects included in this study were covered for their medi- treatment in the years after the implementation of the PPP cations by the RAMQ drug plan. Thus, the cost of the medi- and other guidelines. The PPP and the other guidelines were cations is not likely to have been a major factor for not

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 1246 Isotretinoin therapy and acne relapse, L. Azoulay et al.

Table 5 Predictors of receiving a second Crude rate ratio Adjusted rate ratioa isotretinoin treatment (95% CI) (95% CI) At index date < 16 years of age 1Æ30 (1Æ16–1Æ46) 1Æ26 (1Æ12–1Æ43) Male gender 1Æ07 (1Æ01–1Æ14) 1Æ32 (1Æ22–1Æ41) Urban dweller 1Æ04 (0Æ96–1Æ12) 0Æ94 (0Æ86–1Æ02) Welfare recipient 1Æ14 (1Æ06–1Æ21) 1Æ08 (0Æ99–1Æ17) Time periods 1 January 1984–30 September 1988 1Æ00 (Reference) 1Æ00 (Reference) 1 October 1988–31 December 1994 1Æ16 (1Æ03–1Æ31) 1Æ10 (0Æ97–1Æ24) 1 January 1995–31 December 1999 1Æ10 (0Æ99–1Æ22) 1Æ22 (1Æ09–1Æ37) 1 January 2000–30 June 2003 1Æ06 (0Æ96–1Æ17) 1Æ29 (1Æ15–1Æ46) Between the end of the first treatment and index date Number of different types of medications 1Æ09 (1Æ08–1Æ09) 1Æ11 (1Æ10–1Æ12) At least one antiacne medication 2Æ06 (1Æ87–2Æ27) 1Æ51 (1Æ36–1Æ68) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 12Æ75 (2Æ52–3Æ01) 2Æ71 (2Æ46–2Æ98) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ05 (0Æ96–1Æ15) 0Æ78 (0Æ70–0Æ86) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ13 (1Æ00–1Æ29) 0Æ76 (0Æ65–0Æ88) During the first isotretinoin treatment Number of different types of medications 1Æ00 (0Æ98–1Æ01) 0Æ95 (0Æ93–0Æ98) At least one antiacne medication 1Æ08 (0Æ95–1Æ22) 0Æ94 (0Æ82–1Æ08) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 10Æ74 (0Æ67–0Æ81) 0Æ81 (0Æ73–0Æ89) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 10Æ92 (0Æ83–1Æ02) 1Æ03 (0Æ91–1Æ15) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 10Æ99 (0Æ82–1Æ21) 1Æ04 (0Æ84–1Æ29) Age of treating physician (years) 1Æ01 (1Æ00–1Æ01) 1Æ00 (1Æ00–1Æ00) Treated by male physician 1Æ15 (1Æ06–1Æ24) 1Æ20 (1Æ10–1Æ31) Treated by dermatologist 1Æ00 (0Æ93–1Æ06) 1Æ08 (1Æ00–1Æ16) ‡ 2 isotretinoin prescribers 0Æ88 (0Æ77–1Æ00) 1Æ09 (0Æ94–1Æ26) Cumulative dose < 2450 mg 1Æ00 (Reference) 1Æ00 (Reference) 2450–4840 mg 0Æ71 (0Æ65–0Æ77) 0Æ85 (0Æ77–0Æ93) 4840–7584 mg 0Æ54 (0Æ50–0Æ59) 0Æ80 (0Æ71–0Æ90) ‡ 7584 mg 0Æ41 (0Æ37–0Æ45) 0Æ62 (0Æ54–0Æ71) Treatment duration ‡ 121 days 0Æ52 (0Æ48–0Æ55) 0Æ69 (0Æ62–0Æ76) 12 months prior to the first isotretinoin treatment Number of different types of medications 1Æ00 (1Æ00–1Æ00) 0Æ99 (0Æ99–0Æ99) At least one antiacne medication 1Æ13 (1Æ06–1Æ21) 1Æ15 (1Æ07–1Æ24) Visits to the physician None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ14 (0Æ99–1Æ32) 0Æ90 (0Æ77–1Æ06) Visits to the emergency department None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ05 (0Æ98–1Æ13) 0Æ98 (0Æ90–1Æ06) Hospitalizations None 1Æ00 (Reference) 1Æ00 (Reference) ‡ 11Æ02 (0Æ92–1Æ12) 0Æ95 (0Æ85–1Æ06)

aAdjusted for the covariates in the table. CI, confidence interval.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 Isotretinoin therapy and acne relapse, L. Azoulay et al. 1247 getting antiacne medications, had they been prescribed. The 6 Layton AM, Knaggs H, Taylor J et al. Isotretinoin for acne vulgaris – use of OTC products is also likely to have been minimal. 10 years later: a safe and successful treatment. Br J Dermatol 1993; Because of their insurance status, subjects were more likely 129:292–6. 7 Lehucher-Ceyrac D, Weber-Buisset MJ. Isotretinoin and acne in to get a prescription for products also available OTC (such as practice: a prospective analysis of 188 cases over 9 years. Dermatology benzoyl peroxide). 1993; 186:123–8. Another limitation was that acne relapse was defined as 8 Stainforth JM, Layton AM, Taylor JP et al. Isotretinoin for the receiving either a topical or systemic antiacne medication. The treatment of acne vulgaris: which factors may predict the inclusion of topical agents in this definition could have need for more than one course? Br J Dermatol 1993; 129:297– included subjects with mild acne who would not have been 301. defined as relapsers in a clinical study. However, topical agents 9 Shahidullah M, Tham SN, Goh CL. Isotretinoin therapy in acne vul- garis: a 10-year retrospective study in Singapore. Int J Dermatol accounted for less than 20% of all antiacne medications pre- 1994; 33:60–3. scribed. Furthermore, the majority of these agents are also 10 White GM, Chen W, Yao J et al. Recurrence rates after the first prescribed for moderate acne (e.g. tretinoin, benzoyl per- course of isotretinoin. Arch Dermatol 1998; 134:376–8. oxide). With regards to antiacne medications not reimbursed 11 Ng PP, Goh CL. Treatment outcome of acne vulgaris with oral iso- by the RAMQ drug plan (e.g. adapalene), the exclusion of tretinoin in 89 patients. Int J Dermatol 1999; 38:213–16. these agents is not likely to have had an impact on the acne 12 Lehucher-Ceyrac D, de La SP, Chastang C et al. Predictive factors for relapse rate. Although such agents may be covered by some failure of isotretinoin treatment in acne patients: results from a cohort of 237 patients. Dermatology 1999; 198:278–83. private insurance programmes, physicians will typically pre- 13 Al-Mutairi N, Manchanda Y, Nour-Eldin O et al. Isotretinoin in scribe medications they know are covered by their patient’s acne vulgaris: a prospective analysis of 160 cases from Kuwait. insurance programme. Thus, it is likely that physicians would J Drugs Dermatol 2005; 4:369–73. replace an antiacne medication not covered by a given insur- 14 Haryati I, Jacinto SS. Profile of acne patients in the Philippines ance programme by one that is. Other limitations of this study requiring a second course of oral isotretinoin. Int J Dermatol 2005; were that we were not able to adjust directly for acne severity, 44:999–1001. or the site of acne. This is because the ICD-9 classification 15 Quereux G, Volteau C, N’Guyen JM et al. Prospective study of risk factors of relapse after treatment of acne with oral isotretinoin. does not specify the severity or the location of the acne. How- Dermatology 2006; 212:168–76. ever, we did use antiacne medication dispensing information 16 Statistiques annuelles. Re´gie de l’assurance maladie du Que´bec. Quebec: in time periods before the index date, cumulative dose and Government of Quebec, 1997. length of the first isotretinoin treatment as proxies for acne 17 World Health Organization. International Classification of Diseases, 9th severity. Revision (ICD-9). Geneva: World Health Organization, 1977. Isotretinoin is an effective medication associated with long- 18 Tamblyn R, Lavoie G, Petrella L et al. The use of prescription claims term remissions in patients with acne. However, it remains databases in pharmacoepidemiological research: the accuracy and comprehensiveness of the prescription claims database in Quebec. that 41% of patients may experience an acne relapse necessi- J Clin Epidemiol 1995; 48:999–1009. tating further treatment with antiacne medications. Several 19 Tamblyn R, Reid T, Mayo N et al. Using medical services claims to patient characteristics have been found to be associated with assess injuries in the elderly: sensitivity of diagnostic and proced- acne relapses. These data could be of prognostic value to clini- ure codes for injury ascertainment. J Clin Epidemiol 2000; 53:183– cians who treat patients with acne. 94. 20 Levy AR, Mayo NE, Grimard G. Rates of transcervical and pertro- chanteric hip fractures in the province of Quebec, Canada, 1981– Acknowledgments 1992. Am J Epidemiol 1995; 142:428–36. 21 Azoulay L, Oraichi D, Berard A. Patterns and utilization of isotre- This study was supported by the Canadian Institutes of Health tinoin for acne from 1984 to 2003: is there need for concern? Research (CIHR grant number: IHD – 67337). Eur J Clin Pharmacol 2006; 62:667–74. 22 Richardson DB. An incidence density sampling program for nested case–control analyses. Occup Environ Med 2004; 61:e59. References 23 Rothman KJ, Greenland S. Modern Epidemiology, 2nd edn. Philadelphia, 1 Mitchell AA, Van Bennekom CM, Louik C. A pregnancy-prevention PA: Lippincott-Raven, 1998. program in women of childbearing age receiving isotretinoin. 24 Suissa S. Effectiveness of inhaled corticosteroids in chronic obstruc- N Engl J Med 1995; 333:101–6. tive pulmonary disease: immortal time bias in observational stud- 2Be´rard A, Azoulay L, Koren G et al. Isotretinoin, pregnancies, abor- ies. Am J Respir Crit Care Med 2003; 168:49–53. tions and birth defects: a population-based perspective. Br J Clin 25 Ho V, Schachter D, Miller R. Acne management for the 90s: Pharmacol 2007; 63:196–205. current treatment guidelines. Can J Diagnosis 1995; 12 (Suppl.):1– 3 Ellis CN, Krach KJ. Uses and complications of isotretinoin therapy. 25. J Am Acad Dermatol 2001; 45:S150–7. 26 Tan JK. Perspectives on isotretinoin and the Canadian Consen- 4 Harms M, Masouye I, Radeff B. The relapses of cystic acne after sus Guidelines on treatment of acne. Skin Therapy Lett 2000; 6:1– isotretinoin treatment are age-related: a long-term follow-up study. 4. Dermatologica 1986; 172:148–53. 27 Madden WS, Landells ID, Poulin Y et al. Treatment of acne vulgaris 5 Chivot M, Midoun H. Isotretinoin and acne – a study of relapses. and prevention of acne scarring: Canadian consensus guidelines. Dermatologica 1990; 180:240–3. J Cutan Med Surg 2000; 4 (Suppl. 1):S2–13.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 1248 Isotretinoin therapy and acne relapse, L. Azoulay et al.

28 Cabana MD, Rand CS, Powe NR et al. Why don’t physicians follow 30 Wert S. Identification and management of oral isotretinoin use clinical practice guidelines? A framework for improvement. JAMA inconsistent with product labeling. Manag Care Interface 2003; 1999; 282:1458–65. 16:41–3. 29 Wysowski DK, Swann J, Vega A. Use of isotretinoin (Accutane) in 31 Chen K, White TJ, Juzba M, Chang E. Oral isotretinoin: an analysis the United States: rapid increase from 1992 through 2000. JAm of its utilization in a managed care organization. J Manag Care Pharm Acad Dermatol 2002; 46:505–9. 2002; 8:272–7.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1240–1248 CONCISE COMMUNICATION DOI 10.1111/j.1365-2133.2007.08190.x Psoriasis patients show signs of insulin resistance S. Boehncke, D. Thaci,* H. Beschmann,* R.J. Ludwig,* H. Ackermann, K. Badenhoop and W-H. Boehncke* Department of Internal Medicine, Section for Endocrinology, Metabolism and Diabetology, *Department of Dermatology and Department of Biomathematics, Johann Wolfgang Goethe-University, Theodor-Stern-Kai 7, D-60590 Frankfurt am Main, Germany

Summary

Correspondence Background Recent observations suggest that psoriasis is a risk factor for the devel- Wolf-Henning Boehncke. opment of coronary artery calcification which in turn represents an indicator for E-mail: [email protected] atherosclerosis. Objective To clarify a possible pathogenetic link between psoriasis and atherosclero- Accepted for publication 5 June 2007 sis, we studied the metabolic state of patients with psoriasis. Methods Thirty-nine consecutive patients with moderate-to-severe plaque-type pso- Key words riasis were enrolled in the study. Detailed information was obtained on the adipokines, atherosclerosis, coronary heart disease, patients’ clinical picture and history of psoriasis, smoking habits and medication. insulin resistance, psoriasis, resistin The body mass index (BMI) of the patients was calculated. Laboratory investiga- Conflicts of interest tions focused on values for inflammation, lipid profile and multiple cytokines. None declared. The intima–media thickness of the carotid artery was measured by ultrasound, and an oral glucose tolerance test was performed to calculate the homeostasis model assessment of insulin resistance (HOMA). Results Numerous well-recognized correlations such as between BMI and HOMA (P <0Æ02) as well as BMI and vessel wall thickness (P <0Æ05) were successfully reproduced, thus confirming consistency of our dataset. With regard to psoriasis, we observed a significant correlation between the Psoriasis Area and Severity Index (PASI) score and insulin secretion. Moreover, the PASI score was signifi- cantly correlated with serum resistin levels—a cytokine known to be increased in insulin resistance. Conclusions Taken together, several measurements indicative of insulin resistance were found to be significantly correlated with the PASI score. The concept of insulin resistance as a consequence of chronic inflammation and possible patho- genetic cause for comorbidities known to be associated with psoriasis is sup- ported by these data. Our findings validate further studies on larger cohorts as well as interventional studies.

Psoriasis is a chronic inflammatory skin disease characterized myocardial infarction,5 and a case–control study showed sub- by red scaly plaques, which may occur on any site of the stantially elevated levels of coronary artery calcification as an body; prevalence is about 2–3% in Caucasians.1 Approximately early indicator for developing coronary artery disease.6 20% of the patients suffer from widespread disease necessitat- A well-established pathway links atherosclerosis to obesity ing photo- or systemic therapies, but health-related quality of via overproduction of tumour necrosis factor (TNF)-a, which life is substantially reduced even in less severely affected indi- also contributes to insulin resistance and the development of viduals. Numerous diseases have been found to be associated type 2 diabetes mellitus.7 A recent study documented the risk with psoriasis, including diabetes mellitus and cardiovascular of psoriasis to be directly related to the body mass index disease.2 (BMI).8 We therefore investigated the metabolic state in Inflammation in psoriasis results in a T-helper (Th) 1 lym- patients with psoriasis in an attempt to identify insulin resis- phocyte cytokine milieu, and signs of systemic inflammation, tance as a possible link between the comorbidities observed. such as increased C-reactive protein levels or platelet activa- 3 tion. These factors seem to play a major role in the develop- Patients and methods ment of atherosclerosis and ultimately myocardial infarction.4 In line with this hypothesis, a recent population-based cohort We studied 39 consecutive patients referred to our dermatol- study identified psoriasis as an independent risk factor for ogy unit for treatment of moderate-to-severe plaque-type

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1249–1251 1249 1250 Psoriasis and insulin resistance, S. Boehncke et al. psoriasis. Detailed information was obtained on the patients’ BMI > 30. A pathological HOMA (> 2Æ5) was observed in 12 clinical picture and history of psoriasis, smoking habits and of the 39 patients. medication. The current disease severity was quantified using Initially, the technique to measure vessel wall thickness the Psoriasis Area and Severity Index (PASI) which determines was successfully validated by establishing a firm correlation the affected body surface area together with erythema, infiltra- between vessel wall thickness and age (r=0Æ632, P <0Æ01, tion and scaling. The BMI of the patients was also calculated. two-sided). Laboratory investigations focused on measurements for inflam- Subsequent analyses initially focused on confirmation of mation, lipid profile and serum levels of multiple cytokines, well-recognized correlations between obesity and insulin resis- the latter measured by enzyme-linked immunosorbent assay. tance as well as atherosclerosis as a consequence (Table 1). In addition, the intima–media thickness of the carotid artery Among the most important correlations reconfirmed in our was measured by ultrasound (‘leading edge’ method). Finally, dataset were those between BMI and HOMA (r=0Æ595, an oral glucose tolerance test was performed to calculate the P <0Æ01) as well as BMI and vessel wall thickness (r=0Æ370, homeostasis model assessment of insulin resistance (HOMA). P <0Æ05). These correlations together with the validated mea- Statistical analyses were performed using linear regression, surements for vessel wall thickness are proof of the consist- multiple regression with the PASI score as a response variable ency of our observations. and, where adequate, the Wilcoxon–Mann–Whitney U-test. With regard to psoriasis, we observed a significant correl- ation between the PASI score and insulin secretion, particularly Results at t = 30 min, in the oral glucose tolerance test (r=0Æ355, P <0Æ05) (Table 1). Analyses of inflammatory mediators As in most psoriasis studies, our cohort was found to be over- focused on adiponectin, leptin, resistin and TNF-a. These weight to obese, with 12 patients having a BMI of 20–24Æ9, cytokines belong to the so-called adipokines, referring to their 14 patients with a BMI of 25–29Æ9 and 13 patients with a role in metabolic regulation.7 In contrast to adiponectin and

Table 1 Correlations between parameters describing psoriasis severity and insulin resistance

Vessel wall Insulin at Glucose at PASI BMI thickness t = 30 min t = 30 min HOMA Resistin PASI Correlation (Pearson) 1 )0Æ009 )0Æ043 0Æ355* 0Æ055 0Æ191 0Æ318* Significance (two-sided) 0Æ957 0Æ799 0Æ029 0Æ743 0Æ249 0Æ048 n 39 39 37 38 38 38 39 BMI Correlation (Pearson) )0Æ009 1 0Æ370* 0Æ255 0Æ274 0Æ595** 0Æ144 Significance (two-sided) 0Æ957 0Æ024 0Æ122 0Æ096 <0Æ01 0Æ382 n 39 39 37 38 38 38 39 Vessel wall thickness Correlation (Pearson) )0Æ043 0Æ370* 1 0Æ075 0Æ240 0Æ292 )0Æ078 Significance (two-sided) 0Æ799 0Æ024 0Æ660 0Æ153 0Æ079 0Æ644 n 37 37 37 37 37 37 37 Insulin at t = 30 min Correlation (Pearson) 0Æ355* 0Æ255 0Æ075 1 0Æ077 0Æ595** 0Æ044 Significance (two-sided) 0Æ029 0Æ122 0Æ660 0Æ645 <0Æ001 0Æ792 n 38 38 37 38 38 38 38 Glucose at t = 30 min Correlation (Pearson) 0Æ055 0Æ274 0Æ240 0Æ077 1 0Æ395* 0Æ359* Significance (two-sided) 0Æ743 0Æ096 0Æ153 0Æ645 0Æ014 0Æ027 n 38 38 37 38 38 38 38 HOMA Correlation (Pearson) 0Æ191 0Æ595** 0Æ292 0Æ595** 0Æ395* 1 0Æ176 Significance (two-sided) 0Æ249 <0Æ01 0Æ079 <0Æ001 0Æ014 0Æ291 n 38 38 37 38 38 38 38 Resistin Correlation (Pearson) 0Æ318* 0Æ144 )0Æ078 0Æ044 0Æ359* 0Æ176 1 Significance (two-sided) 0Æ048 0Æ382 0Æ644 0Æ792 0Æ027 0Æ291 n 39 39 37 38 38 38 39

PASI, Psoriasis Area and Severity Index score; BMI, body mass index; HOMA, homeostasis model assessment of insulin resistance. *Statisti- cally significant at the P <0Æ05 level (two-sided); **statistically significant at the P <0Æ01 level (two-sided).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1249–1251 Psoriasis and insulin resistance, S. Boehncke et al. 1251 leptin, we were able to establish a statistically significant posi- Clinical trials are needed to evaluate therapeutic strategies that tive correlation between the PASI score and serum resistin interfere with common pathways of inflammation induced by levels (r=0Æ318, P <0Æ05) (Table 1), an adipokine related to obesity on the one hand and psoriasis on the other hand. As insulin resistance.9 part of good clinical practice, physicians should encourage patients with psoriasis to minimize their cardiovascular risk by Discussion addressing their modifiable risk factors.

Psoriasis is regarded an independent risk factor for increased References cardiovascular morbidity in general and myocardial infarction in particular.5 These observations support the concept of Th1 1 Scho¨n MP, Boehncke W-H. Psoriasis. N Engl J Med 2005; lymphocyte-mediated diseases causing atherosclerosis and cor- 352:1899–912. onary artery disease. Psoriasis is the most common Th1 dis- 2 Henseler T, Christophers E. Disease concomitance in psoriasis. JAm Acad Dermatol 1995; 32:982–6. ease. Other diseases of comparable immunopathogenesis such 3 Ludwig RJ, Schultz JE, Boehncke W-H et al. Activated, not resting, as rheumatoid arthritis have also been shown to confer an platelets increase leukocyte rolling in murine skin utilizing a dis- increased risk for myocardial infarction, when adjusting for tinct set of adhesion molecules. J Invest Dermatol 2004; 122:830–6. other known risk factors, such as diabetes.10 The exact mecha- 4 Hansson GL. Inflammation, atherosclerosis, and coronary artery nism by which these diseases predispose a patient to cardio- disease. N Engl J Med 2005; 352:1685–95. vascular disease is unclear, but may involve various cellular 5 Gelfand JM, Neimann AL, Shin DB et al. Risk of myocardial infarc- and humoral inflammatory mediators, such as TNF-a.11 Block- tion in patients with psoriasis. JAMA 2006; 296:1735–41. 6 Ludwig RJ, Herzog C, Rostock A et al. Psoriasis: a possible risk fac- ade of TNF-a has also been established as an effective strategy 1 tor for development of coronary artery calcification. Br J Dermatol for treatment of psoriasis. In addition, overproduction of 2007; 156:271–6. TNF-a regularly occurs in obesity where the source is the adi- 7 Wellen KE, Hotamisligil GS. Inflammation, stress, and diabetes. pose tissue.12 Obesity is linked to atherosclerosis as well as J Clin Invest 2005; 115:1111–19. insulin resistance.7 A recent study documented the risk of pso- 8 Naldi L, Chatenoud L, Linder D et al. Cigarette smoking, body riasis to be directly related to the BMI.8 mass index, and stressful life events as risk factors for psoriasis: In this study, we found the metabolic state in psoriasis to results from an Italian case control study. J Invest Dermatol 2005; 125:61–7. be shifted towards insulin resistance, which favours athero- 9 Steppan CM, Bailey ST, Bhat S et al. The hormone resistin links sclerosis. Interestingly, we observed a significant correlation obesity to diabetes. Nature 2001; 409:307–12. between the adipokine resistin and the PASI score as a mea- 10 Maradit-Kremers H, Nicola PJ, Crowson CS et al. Cardiovascular sure for clinical disease severity. Recently, a similar correlation death in rheumatoid arthritis: a population-based study. Arthritis between disease activity and serum resistin levels was Rheum 2005; 52:722–32. described in rheumatoid arthritis.13 Moreover, we were able 11 Popa C, Netea MG, Radstake T et al. Influence of anti-tumor necro- to detect clearly the known robust impact of obesity on insu- sis factor therapy on cardiovascular risk factors in patients with active rheumatoid arthritis. Ann Rheum Dis 2005; 64:303–5. lin resistance. These observations support the concept of syn- 12 Klein J, Permana PA, Owecki M et al. What are subcutaneous ergistic effects from the chronic state of inflammation caused adipocytes really good for? Exp Dermatol 2007; 16:45–70. by obesity and the chronic systemic Th1 lymphocyte-mediated 13 Senolt L, Housa D, Vernerova Z et al. Resistin is abundantly present inflammation characteristic for psoriasis, which has recently in rheumatoid arthritis synovial tissue, synovial fluid, and elevated been put forward by Hamminga et al.14 Clinical evidence for serum resistin reflects disease activity. Ann Rheum Dis 2007; 66:458– the relevance of insulin resistance in psoriasis comes from a 63. pilot study which showed efficacy of pioglitazone, used as an 14 Hamminga EA, van der Lely AJ, Neumann HAM, Thio HB. Chronic inflammation in psoriasis and obesity: implications for therapy. insulin-sensitizer in type 2 diabetes mellitus.15 Med Hypoth 2006; 67:768–73. Our data are novel since they provide insights into the 15 Shafiq N, Malhotra S, Pandhi P et al. Pilot trial: pioglitazone versus pathophysiology which leads to atherosclerosis and ultimately placebo in patients with plaque psoriasis. Int J Dermatol 2005; an increased risk for myocardial infarction in patients with 44:328–33. psoriasis, thus complementing recent epidemiological studies.5

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1249–1251 CASE REPORT DOI 10.1111/j.1365-2133.2007.08159.x Unusual molecular findings in Kindler syndrome K. Arita,* V. Wessagowit,* A.C. Inamadar,§ A. Palit,§ H. Fassihi,* J.E. Lai-Cheong,* C. Pourreyron,– A.P. South– and J.A. McGrath* *Genetic Skin Disease Group, St John’s Institute of Dermatology, Division of Genetics and Molecular Medicine, The Guy’s, King’s College and St Thomas’ School of Medicine, St Thomas’ Hospital, Lambeth Palace Road, London SE1 7EH, U.K. Department of Dermatology, Hokkaido University Graduate School of Medicine, Sapporo, Japan Institute of Dermatology, Bangkok, Thailand §Department of Dermatology, Venereology & Leprosy, BLDEA’s SBMP Medical College, Hospital & Research Centre, Bijapur, Karnataka, India –Centre for Cutaneous Research, Bart’s and The London, Queen Mary’s School of Medicine and Dentistry, London, U.K.

Summary

Correspondence Kindler syndrome (KS) is a rare inherited skin disorder with blistering and poi- Ken Arita. kiloderma as its main clinical features. It is caused by loss-of-function mutations E-mail: [email protected] in the C20orf42 (KIND1) gene which encodes kindlin-1, an actin cytoskeleton– focal contact-associated protein which is predominantly expressed in keratino- Accepted for publication 24 June 2007 cytes. We investigated the molecular basis of KS in a 16-year-old Indian boy who had additional clinical findings, including scleroatrophic changes of the Key words hands and feet, pseudoainhum and early onset of squamous cell carcinoma on Kindler syndrome, pseudoainhum, scleroatrophy, his foot. Immunostaining for kindlin-1 in the patient’s skin was completely splice-site mutation, squamous cell carcinoma absent and sequencing of C20orf42 (KIND1) genomic DNA showed a homozygous fi Conflicts of interest splice-site mutation at the -6 position, IVS9-6T A. Amplification and sequencing None declared. of cDNA from the skin revealed aberrant splicing with either deletion of exon 10 or deletion of exons 9, 10 and 11, both of which involve loss of the pleckstrin homology domain of kindlin-1 that is thought to play a role in cytoskeletal attachment and integrin-mediated cell signalling. Pathogenic splice-site mutations at the -6 position are unusual and have rarely been reported for any genetic dis- order. Collectively, these findings extend the spectrum of clinical and molecular abnormalities in this rare genodermatosis.

Kindler syndrome (KS; OMIM 173650) is an autosomal Case report recessive disorder characterized by a combination of acral blistering, poikiloderma, skin atrophy and photosensitivity, The proband was a 16-year-old Indian boy who presented although other clinical features, such as nail dystrophy and with a rapidly growing fungating skin tumour on the right palmoplantar keratoderma, have also been reported.1,2 The lower leg. The lesion had been present for 10 years as a causative gene for Kindler syndrome is C20orf42 (KIND1) hyperkeratotic plaque but had expanded dramatically and encoding kindlin-1, a protein implicated in anchorage of the ulcerated over the preceding 2 months. His skin was normal actin cytoskeleton to integrin-associated platforms.3–7 Here at birth but after the age of 1 year he developed speckled we describe a 16-year-old boy with Kindler syndrome who hyper- and hypopigmentation on sun-exposed areas, acral ves- has additional clinical findings, including scleroatrophic icles and sclerosis of the skin over the hands and feet. Over changes of the hands and feet, pseudoainhum and aggressive the next 5 years there was progressive tapering of the fingers squamous cell carcinoma (SCC) on the leg. The molecular with flexion contractures, nail dystrophy and loss of the distal pathology is unusual, a homozygous C20orf42 (KIND1) splice- phalanges. In addition, from the age of 5 years he developed site mutation in the -6 position, IVS9-6TfiA, that results in asymptomatic hyperkeratotic plaques on the feet which gradu- aberrant splicing and very low levels of C20orf42 (KIND1) ally enlarged. At presentation, examination revealed an ulcer- mRNA and kindlin-1 protein. Collectively, these findings ated tumour, 25 · 15 cm, on his right ankle (Fig. 1a). extend the spectrum of clinical and molecular abnormalities Enlarged matted lymph nodes were palpable in the popliteal in this rare genodermatosis. and inguinal regions. Histopathology of the tumour showed a

2007 The Authors 1252 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1252–1256 Unusual molecular findings in Kindler syndrome, K. Arita et al. 1253

(a) (b) (c)

(d) (e) (f)

Fig 1. Clinical features of the proband. (a) Squamous cell carcinoma on his right ankle; (b) facial poikiloderma; (c) scleroatrophy of the digits, loss of distal part of the left middle finger [shown in more detail in (d)], and marked skin atrophy; (e) development of pseudoainhum on both index fingers, loss of palmar creases, and crusted plaques and skin atrophy on the wrists; (f) verrucous keratotic plaque on the right sole. well-differentiated squamous cell carcinoma (SCC; not epidermis of normal human skin (Fig. 2a), supporting a diag- shown). General examination revealed poikiloderma on the nosis of Kindler syndrome.7 No differences in collagen VII face and neck (Fig. 1b). He had erosive gingivitis and second- immunolabelling intensity between the proband and control ary dental caries. His hands and feet showed scleroatrophy, were seen (data not shown). Next, we sequenced C20orf42 nail dystrophy and palmoplantar keratoderma (Fig. 1c–e). (KIND1) in the proband’s genomic DNA using primers and Pseudoainhum was observed on almost all fingers and the dis- conditions detailed elsewhere.3,4 Sequencing disclosed a tal phalanges of the middle finger on his left hand had been homozygous TfiA transversion at nucleotide c.1140–6 (Gen- lost (Fig. 1d). A verrucous hyperkeratotic plaque was present Bank no. NM_017671) within intron 9 of the C20orf42 on the right sole (Fig. 1f). (KIND1) gene (Fig. 2b). The two affected siblings were also His parents were first cousins and he was the youngest of homozygous for this mutation, IVS9-6TfiA, and both parents three siblings. His 24-year-old brother and 20-year-old sister were heterozygous (data not shown). No other potentially were also affected but neither had features of acral sclerosis, pathogenic homozygous or heterozygous sequence variants in pseudoainhum, nail dystrophy or skin nodules, although the C20orf42 (KIND1) were identified. To assess whether this muta- older brother had more extensive generalized poikiloderma. tion affects RNA splicing, we performed reverse transcriptase– Clinically, the differential diagnoses in the proband included polymerase chain reaction (RT–PCR) of C20orf42 (KIND1) Huriez syndrome (OMIM 181600) and Weary syndrome, using cDNA extracted from primary cultured dermal fibro- although the pattern of inheritance was more in keeping with blasts (proband and control) obtained from punch biopsies an autosomal recessive disorder, such as recessive dystrophic (attempts to culture the patient’s keratinocytes were unsuc- epidermolysis bullosa (OMIM 226600) or Kindler syndrome cessful). Fibroblasts were cultured to subconfluence in Dul- or, perhaps less likely, a congenital poikiloderma or DNA becco’s Modified Eagle’s Medium (Gibco BRL, Grand Island, repair disorder. Local ethical approval from Guy’s and St Tho- NY, U.S.A.) supplemented with 10% fetal bovine serum (ICN mas’ Hospital was obtained for this study which was carried Biomedicals, Basingstoke, U.K.) and 1% penicillin ⁄streptomy- out in accordance with the principles of the Helsinki Accord. cin (Sigma-Aldrich, St Louis, MO, U.S.A.), and RNA was To make a diagnosis, following informed consent, we per- extracted using RNeasy kits (Qiagen, Crawley, U.K.) with formed immunostaining of the proband’s skin for kindlin-1 subsequent cDNA synthesis using RT (Qiagen). Primers target- using a rabbit polyclonal antikindlin-1 antibody (kindly sup- ing the 3¢ end of the cDNA (forward 5¢-TCAAACAGTGGAA- plied by Dr M. Beckerle, Salt Lake City, Utah, U.S.A.), as well TGTAAACTGG-3¢; reverse 5¢-TACATGTGGGCACGTTAGG-3¢) as a mouse monoclonal anticollagen VII antibody, LH7Æ2 were used to assess C20orf42 (KIND1) expression. For PCR, we (a gift from Dr I. Leigh, Royal London Hospital, U.K.), as used Taq DNA polymerase (Qiagen) and the PCR conditions described previously.5,8 Kindlin-1 staining was completely were denaturation for 5 min at 94 C, followed by 45 cycles absent in contrast to the staining of kindlin-1 in the lower of 30 s at 94 C, 30 s at 56 C and 45 s at 72 C, with a

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1252–1256 1254 Unusual molecular findings in Kindler syndrome, K. Arita et al.

(a)

NHS Patient Dapi Patient (b) (c) KIND1 GAPDH c.1140-6T>A IVS9 Exon 10 mw C P C P

(d)

mw C P Exon 9 Exon 11

Patient’s cDNA sequence (361-bp)

486-bp Exon 8 Exon 12

Patient’s cDNA sequence (204-bp)

Fig 2. Skin biopsy and C20orf42 (KIND1) gene analysis supports a diagnosis of Kindler syndrome. (a) Immunostaining for kindlin-1 in normal human skin (NHS) and the proband’s skin. No staining of kindlin-1 is observed in the proband in contrast to the positive labelling of kindlin-1 within the lower epidermis of NHS. Dapi nuclear staining of the proband’s skin is shown for reference. (b) Sequencing of the proband’s genomic DNA reveals a homozygous splice-site mutation (NM_017671 c.1140–6TfiA; IVS9-6TfiA); this sequence variant was not detected in 100 ethnically matched control chromosomes. (c) Reverse transcriptase–polymerase chain reaction (RT–PCR) of cDNA using primers targeting the 3¢ end of C20orf42 (KIND1) did not yield any detectable product from the proband’s cDNA (P) in contrast to the bright band from control fibroblast cDNA (C). GAPDH primers are shown for control amplification and loading. (d) RT–PCR using 10 times the amount of proband cDNA template with primers spanning exon 8 to exon 12 detected abnormal splicing of mutated C20orf42 (KIND1): compared with a control RT–PCR of control (C) band of 486 bp, RT–PCR in the proband (P) yielded two separate smaller bands of 361 bp and 204 bp and no normal band of 486 bp. Sequencing of the 361-bp band showed skipping of exon 10, whereas the product of 204 bp revealed skipping of exons 9, 10 and 11. Both aberrant splice variants lack the pleckstrin domain of kindlin-1. single final extension reaction at 72 C for 7 min. RT–PCR exons 9, 10 and 11 in the other (Fig. 2d). Deletion of exon using the proband’s cDNA showed no detectable band in con- 10 causes a frameshift and a stop codon 28 amino acids trast to a bright band from the control cDNA (Fig. 2c). Next, downstream in exon 11, while the deletion of exons 9, 10 we designed cDNA primer pairs spanning from exon 8 to and 11 results in in-frame deletion of 68 amino acids. In both exon 12 of C20orf42 (KIND1) (forward 5¢-AGCAAACTGTCG- abnormally spliced transcripts there is loss of the pleckstrin TTGTCTGC-3¢; reverse 5¢-TGAGGACCTCTGGCTGGTAG) and homology domain of kindlin-1, which is thought to have a repeated the PCR using increasing amounts of cDNA template role in cytoskeletal attachment and cell signalling.9 from the proband (the PCR conditions were as above but with an annealing temperature of 55 C and amplifying for Discussion 50 cycles). Using 10 times the original cDNA, two faint, but smaller, bands were seen on agarose gel electrophoresis Splice-site mutations at the -6 position of acceptor sites are a (Fig. 2d). Sequencing of PCR products extracted from the gel rare cause of genetic disease. Indeed, when Krawczak et al.10 showed deletion of exon 10 in one band and deletion of analysed 478 splicing mutations in 38 selected genes they

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1252–1256 Unusual molecular findings in Kindler syndrome, K. Arita et al. 1255 found that only two mutations involved the -6 position. In an antibody against kindlin-1 is very helpful in rapidly estab- contrast, the vast majority of acceptor splice-site mutations lishing a correct diagnosis. Finally, unusual splice-site muta- (~80%) occurred within the -1 or -2 positions (i.e. affecting tions, here at the -6 position of intron 9, need to be fully the obligate AG sequence).10 Predicting the consequences of investigated to examine their potential consequences on gene splice-site mutations at the -6 position is also difficult. Nota- expression and to assess their disease relevance. bly, in silico analysis of the impact of IVS9-6TfiAinC20orf42 (KIND1) on splicing reveals that the information in bits (Ri) Acknowledgments for the acceptor splice site is reduced from 9Æ0to6Æ3 bits, making the site leaky but not abolishing the acceptor splice Funding for this study was kindly provided by the Charitable site completely (see reference 11 for web-based link used to Foundation of Guy’s and St Thomas’, the Dystrophic Epiderm- calculate and interpret these values). However, this mutation olysis Bullosa Research Association (DebRA, U.K.) and the Bar- also reduces the lariat branchpoint Ri from 2Æ9to1Æ4 bits, bara Ward Children’s Foundation. thus abolishing the lariat branchpoint in intron 9. The effect of this mutation, therefore, is predicted to result in skipping References of exon 10 by preventing the spliceosome from removing intron 9 (which then recognizes the next natural lariat 1 Ashton GH. Kindler syndrome. Clin Exp Dermatol 2004; 29:116– branchpoint and acceptor splice site in intron 10). Indeed, 21. similar base substitutions in mammalian branchpoint 2 Fischer IA, Kazandjieva J, Vassileva S, Dourmishev A. Kindler syn- drome: a case report and proposal for clinical diagnostic criteria. sequences have been shown to reduce substantially the effi- Acta Dermatovenerol Alp Panonica Adriat 2005; 14:61–7. ciency of pre-mRNA splicing in vitro and to alter acceptor 3 Ashton GH, McLean WH, South AP et al. Recurrent mutations in 12 splice-site selection in vivo. Human lariat branchpoint muta- kindlin-1, a novel keratinocyte focal contact protein, in the autoso- tions leading to a skin phenotype, however, have rarely been mal recessive skin fragility and photosensitivity disorder, Kindler published, with only three reports in cases of xeroderma syndrome. J Invest Dermatol 2004; 122:78–83. pigmentosum,13 nail patella syndrome14 and type II Ehlers– 4 Siegel DH, Ashton GH, Penagos HG et al. Loss of kindlin-1, a Danlos syndrome.15 human homolog of the Caenorhabditis elegans actin–extracellular– matrix linker protein UNC-112, causes Kindler syndrome. Am J Our case also has several noteworthy clinical features, Hum Genet 2003; 73:174–87. including the degree of scleroatrophy of the hands and feet, 5 Kloeker S, Major MB, Calderwood DA et al. The Kindler syndrome the pseudoainhum and the development of fulminant SCC. protein is regulated by transforming growth factor-beta and Scleroatrophy and pseudoainhum are not well-recognized fea- involved in integrin-mediated adhesion. J Biol Chem 2004; tures of Kindler syndrome, although one description of acral 279:6824–33. pseudoscleroderma16 and five reports of variable degrees of 6 White SJ, Mclean WHI. Kindler surprise: mutations in a novel pseudoainhum,2,17–20 including one case with loss of digits, actin-associated protein cause Kindler syndrome. J Dermatol Sci 2005; 38:169–75. similar to our case,19 have been reported in the literature. 7 Herz C, Aumailley M, Schulte C et al. Kindlin-1 is a phosphoprotein The occurrence of cutaneous SCC in a teenager with Kindler involved in regulation of polarity, proliferation and motility of epi- syndrome is also somewhat unusual. When this complication dermal keratinocytes. J Biol Chem 2006; 281:36082–90. occurs in Kindler syndrome, affected individuals are usually 8 Heagerty AH, Kennedy AR, Leigh IM et al. Identification of an epi- older: 43 years3 or 57 years,21 although a 34-year-old patient dermal basement membrane defect in recessive forms of dystrophic with SCC of the hard palate has also been reported.22 The epidermolysis bullosa by LH7:2 monoclonal antibody: use in diag- site of the cutaneous SCCs reported has been on the hands, nosis. Br J Dermatol 1986; 115:125–31. 9 Maffucci T, Falasca M. Specificity in pleckstrin homology (PH) feet and lip suggesting that chronic inflammation and ⁄or domain membrane targeting: a role for a phosphoinositide-protein ultraviolet radiation exposure may be relevant, although the co-operative mechanism. FEBS Lett 2001; 506:173–9. direct causal association between loss of kindlin-1 and the 10 Krawczak M, Thomas NST, Hundrieser B et al. Single base-pair sub- increased risk of SCC is not known. Nor is it clear why stitutions in exon–intron junctions of human genes: nature, distri- the proband we are reporting developed SCC at such a rela- bution, and consequences for mRNA splicing. Hum Mutat 2007; tively young age. The development of SCC in teenagers is 28:150–8. well recognized in disorders such as recessive dystrophic epi- 11 Laboratory of Human Molecular Genetics and Genomic Disorders. Automated Splice Site Analyses. https://splice.cmh.edu/ (accessed dermolysis bullosa but this is by far the youngest patient 20 June 2007). with Kindler syndrome to develop SCC. Moreover, the 12 Reed R, Maniatis T. The role of the mammalian branchpoint tumour, despite its well-differentiated histology, behaved sequence in pre-mRNA splicing. Genes Dev 1988; 2:1268–76. very aggressively and the affected individual died within a 13 Khan SG, Metin A, Gozukara E et al. Two essential splice lariat few months of his presentation. branchpoint sequences in one intron in a xeroderma pigmentosum In summary, this report has identified unusual clinical and DNA repair gene: mutations result in reduced XPC mRNA lev- molecular pathology in a case of Kindler syndrome. Firstly, it els that correlate with cancer risk. Hum Mol Genet 2004; 13:343– 52. should be recognized that the spectrum of clinical abnormali- 14 Hamlington JD, Clough MV, Dunston JA, McIntosh I. Deletion of a ties in this disorder includes scleroatrophy, pseudoainhum and branch-point consensus sequence in the LMX1B gene causes exon early-onset SCC. Secondly, skin immunohistochemistry using

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1252–1256 1256 Unusual molecular findings in Kindler syndrome, K. Arita et al.

skipping in a family with nail patella syndrome. Eur J Hum Genet 19 Krunic AL, Ljiljana M, Novak A et al. Hereditary bullous acrokera- 2000; 8:311–14. totic poikiloderma of Weary–Kindler associated with pseudo- 15 Burrows NP, Nicholls AC, Richards AJ et al. A point mutation in an ainhum and sclerotic bands. Int J Dermatol 1997; 36:529–33. intronic branch site results in aberrant splicing of COL5A1 and in 20 Al Aboud K, Al Hawsawi K, Al Aboud D, Al Githami A. Ehlers–Danlos syndrome type II in two British families. Am J Hum Kindler syndrome in a Saudi kindred. Clin Exp Dermatol 2002; 27: Genet 1998; 63:390–8. 673–6. 16 Khopkar U, Raj S, Wadhwa SL. Weary–Kindler syndrome with 21 Emanuel PO, Rudikoff D, Phelps RG. Aggressive squamous cell car- multiple seborrheic keratoses. Int J Dermatol 1993; 32:44–5. cinoma in kindler syndrome. Skinmed 2006; 5:305–7. 17 van der Lugt L. Dermatopathia pigmentosa reticularis hyperkerato- 22 Lotem M, Raben M, Zeltser R et al. Kindler syndrome compli- tica et mutilans. Dermatologica 1970; 140:294–302. cated by squamous cell carcinoma of the hard palate: successful 18 Verret JL, Avenel M, Larre`gue M, Panigel-Nguyen C. Syndrome de treatment with high-dose radiation therapy and granulo- Kindler. Un cas avec etude ultrastructurale. Ann Dermatol Venereol cyte-macrophage colony-stimulating factor. Br J Dermatol 2001; 1984; 111:259–69. 144: 1284–6.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1252–1256 CASE REPORT DOI 10.1111/j.1365-2133.2007.08210.x A sporadic case of early-onset sarcoidosis resembling Blau syndrome due to the recurrent R334W missense mutation on the NOD2 gene P. Coto-Segura, S. Mallo-Garcia, M. Costa-Romero,* J.I. Arostegui, J. Yague, E. Ramos-Polo* and J. Santos-Juanes Departments of Dermatology and *Pediatrics, Asturias Central University Hospital, Asturias, Spain Department of Immunology, Hospital Clinic i Provincial de Barcelona, Barcelona, Spain

Summary

Correspondence Sarcoidosis is a multisystem granulomatous disorder characterized by the infiltra- P. Coto-Segura. tion of noncaseating granulomata in the affected tissues. We report here the clin- E-mail: [email protected]; [email protected] ical case of a Caucasian Spanish patient suffering from sporadic early-onset sarcoidosis (EOS) with simultaneous cutaneous and articular symptoms. NOD2 Accepted for publication 7 May 2007 (nucleotide-binding oligomerization domain; previously known as CARD15, cas- pase recruitment domain) gene mutational analysis revealed the presence of the Key words recurrent R334W missense mutation. As in previously reported EOS cases, our Blau syndrome, early-onset sarcoidosis, patient was initially misdiagnosed with dermatitis. NOD2 ⁄CARD15

Conflicts of interest None declared.

Sarcoidosis is a multisystem granulomatous disorder character- We report herein a sporadic clinical case of a Caucasian ized by the infiltration of noncaseating granulomata in the af- Spanish patient suffering from EOS with simultaneous cutane- fected tissues. From a clinical point of view, two different ous and articular symptoms, in whom a NOD2 gene muta- forms can be differentiated. The adult-type sarcoidosis, which tional analysis revealed the presence of the recurrent R334W affects children older than 4 years and adult patients, is a missense mutation. The cutaneous manifestations strongly chronic disease, with clinical self-limited exacerbations, char- resembled an atopic rash and the patient was initially misdiag- acterized by lung involvement, hilar adenopathies and systemic nosed with atopic dermatitis. To the best of our knowledge it symptoms.1 Its aetiology is unknown, although monocyte is the first case with the simultaneous appearance of nodules activation and Th1 response have been previously reported. and scaly dermatitis in EOS. Early-onset sarcoidosis (EOS; MIM 609464) usually begins before 4 years of age, and is a persistent inflammatory disease Case report characterized by skin rash, chronic symmetric polyarthritis, and the possible appearance of recurrent uveitis and campto- A 2-year-old white girl was admitted to our Dermatology dactyly.2,3 Interestingly enough, EOS patients show neither Department for a nonpruritic skin rash on the face, and sub- clinical nor radiological lung involvement (Table 1). All the cutaneous nodules over the joints which she had had since clinical symptoms of EOS patients resemble the main symp- she was 12 months old. She was born after an unremarkable toms of Blau syndrome (BS; MIM 186580), a dominantly pregnancy, and no consanguinity was detected in her family. inherited autoinflammatory disorder, associated with muta- This patient had previously been diagnosed as suffering from tions on the nucleotide-binding oligomerization domain gene atopic dermatitis, having been successfully treated with topic NOD2 (previously known as CARD15, caspase recruitment and oral steroids. Her medical history was otherwise normal domain).4,5 Since the description of BS, it has been widely except for an atrial septal defect, which had closed spontane- discussed as to whether or not EOS and BS are the same dis- ously. At 17 months she presented again with a facial skin ease.6 The recent identification of NOD2 gene mutations rash, which was simultaneously accompanied by a symmetric among different EOS patients, some of which are identical to polyarthritis affecting her elbow, wrist, knee, ankle and proxi- those mutations previously identified in BS patients, has estab- mal interphalangeal joints. The articular swelling increased lished its molecular basis, and has clearly differentiated EOS slowly and the patient refused to walk, not because of from the adult-type sarcoidosis.7–10 pain, but because of limited joint mobility. A dermatological

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1257–1259 1257 1258 A sporadic case of early-onset sarcoidosis, P. Coto-Segura et al.

Table 1 Differences between early-onset sarcoidosis (EOS), Blau syndrome (BS) and adult-type sarcoidosis

EOS BS Adult-type sarcoidosis Age at onset (years) <4 <4 >4 Pulmonary disease Rare Rare Frequent NOD2 mutation Described Described Not described Inheritance Sporadic Dominant Sporadic

Fig 2. Symmetrical polyarthritis and swelling affecting the proximal interphalangeal joints.

Fig 1. Nonpruritic, slightly scaly, erythematous macules on the face of the patient. Atopic dermatitis was initially misdiagnosed. examination revealed the presence of nonpruritic, slightly scaly, more erythematous than tan-coloured macules on her cheeks including the nasal–labial area. Coexisting subcutaneous hard nodules were found over the affected joints and under normal skin (Figs 1, 2). No abnormalities were detected in repeated ophthalmological examinations. No familial history Fig 3. Well-circumscribed, histiocytic granulomas with necrosis in the of similar skin or articular disease, either pulmonary, ocular papillary dermis and few lymphocytes at the periphery (haematoxylin or inflammatory bowel disease was detected. and eosin; original magnification · 100). Skin biopsy specimens taken from a nodule on the back of the hand demonstrated the presence of well-circumscribed, histiocytic granulomas without necrosis in the papillary dermis (PCR). Bidirectional sequencing of PCR products was per- and few lymphocytes at the periphery. Periodic acid–Schiff formed in an ABI 3100 automatic sequencer (Applied Biosys- and acid-fast Bacillus staining were negative. No polarizable tems, Foster City, CA, U.S.A.). Mutational analysis of the particles were detected in the examination of biopsy speci- NOD2 gene revealed the presence of a heterozygous C-to-T mens under polarized light (Fig. 3). transition at position 1000 of the complementary DNA (EMBL Chest radiograph, abdominal ultrasound and electrocardio- accession number: AJ303140), which provokes the sub- gram showed no pathological findings. Radiographs of the stitution of arginine to tryptophan at 334 amino acid on the involved joints revealed swollen soft tissue, with no bone NACHT domain of nucleotide-binding oligomerization abnormalities. Haematological, microbiological and immuno- domain NOD2 protein. This genetic variant, called R334W, logical analyses performed were either normal or negative. A was absent among 115 Spanish Caucasian control chromo- maintained acute-phase response was observed (C-reactive somes. ) protein: 5Æ0mgL 1). Calcium blood and urine levels were The presence of clinical characteristics resembling BS and also normal. A tuberculin test was negative. EOS, as well as its typical histological hallmarks, and the Mutational analysis of systemic autoinflammatory disease- absence of familial history of the disease suggested that we associated genes was performed having obtained the written should perform the EOS diagnosis, which is confirmed by informed consent of the parents. All coding regions and mutational analysis of the NOD2 gene. Oral prednisone ) intronic flanking boundaries of the NOD2 gene were amplified treatment, 1 mg kg 1, was started, resulting in a dramatic from genomic DNA by specific polymerase chain reaction improvement of both cutaneous and articular symptoms.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1257–1259 A sporadic case of early-onset sarcoidosis, P. Coto-Segura et al. 1259

10 Discussion suffering from atopic dermatitis. To the best of our knowl- edge this is the first patient presenting the combination of Adult-type sarcoidosis and EOS are systemic granulomatous active rash and coexisting nodules in EOS. Furthermore, given diseases characterized by the presence of noncaseating granu- that the rash is more erythematous than a tan-coloured dirty lomata in the affected tissues.1,2 EOS is a rare disease, starting scaly rash and the fact that it affects the nasal–labial area in children under 4 years old, and is characterized by the tet- would be useful for dermatologists in order to diagnose this rad of skin, joint and ⁄or eye involvement and the appearance condition correctly. of camptodactyly. Unlike adult-type sarcoidosis, EOS patients do not typically present either clinical or radiological pulmo- References nary disease. The course of the disease is progressive, and this causes both the appearance of inflammatory and iatrogenic 1 Newman LS, Rose CS, Maier LA. Sarcoidosis. N Engl J Med 1997; complications increasing its morbidity.11 336:1224–34. In 1985, Dr Edward Blau reported a family with a domi- 2 Becker ML, Rose CD. Caspase recruitment domain 15 mutations and rheumatic diseases. Curr Opin Rheum 2005; 17:579–85. nantly inherited condition, with similar symptoms to those 3 Brydges S, Kastner DL. The systemic autoinflammatory diseases: 4 previously detected in EOS patients. Since then, several atypi- inborn errors of the innate immune system. Curr Top Microbiol cal manifestations, such as recurrent fever, malignant hyper- Immunol 2006; 305:127–60. tension, renal granulomatous infiltration and cranial 4 Blau EB. Familial granulomatous arthritis, iritis, and rash. J Pediatr neuropathies have been reported.12–15 An interesting discus- 1985; 107:689–93. sion was then started concerning whether EOS and BS were 5 Miceli-Richard C, Lesage S, Rybojad M et al. CARD15 mutations in the same entity. Blau syndrome. Nat Genet 2001; 29:19–20. 6 Miller JJ III. Early-onset ‘sarcoidosis’ and ‘familial granuloma- In 2001, mutations on the NOD2 gene in BS were identified, tous arthritis (arteritis)’: the same disease. J Pediatr 1986; 109:387– 5 thus establishing its genetic basis. In 2004–2005, different 8. authors identified NOD2 gene mutations in EOS patients, and 7 Kanazawa N, Matsushima S, Kambe N et al. Presence of a sporadic some of these mutations are the same as those that have previ- case of systemic granulomatosis syndrome with a CARD15 muta- ously been identified in BS patients. Interestingly enough, an tion. J Invest Dermatol 2004; 122:851–2. important mutational hot-spot at amino acid arginine 334 of 8 Priori R, Bombardieri M, Spinelli FR et al. Sporadic Blau syndrome the NOD2 protein has been described, which occurs in the with a double CARD15 mutation. Report of a case with lifelong follow-up. Sarcoidosis Vasc Diffuse Lung Dis 2004; 21:228–31. majority of BS and EOS patients.5,10,11,15 At this point, we 9 Rose CD, Doyle TM, McIlvain-Simpson G et al. Blau syndrome must highlight the fact that our patient harbours one of these mutation of CARD15 ⁄NOD2 in sporadic early onset granulomatous recurrent mutations which affects amino acid 334 (R334W), arthritis. J Rheumatol 2005; 32:373–5. and this has been previously identified in both BS and EOS 10 Kanazawa N, Okafuji I, Kambe N et al. Early-onset sarcoidosis and patients. CARD15 mutations with constitutive nuclear factor-jB activation: The absence of familial history of the disease and the fact common genetic etiology with Blau syndrome. Blood 2005; that all BS ⁄EOS-associated mutations show full penetrance, 105:1195–7. 11 Rose CD, Wouters CH, Meiorin S et al. Pediatric granulomatous supports the hypothesis that our patient is a sporadic case arthritis. An International Registry. Arthritis Rheum 2006; 54:3337– probably due to a de novo NOD2 mutation, although the non- 44. availability of parent samples to perform genetic analyses pre- 12 Rotenstein D, Gibbas DL, Majmudar B, Chastain EA. Familial gran- vents us from being strictly categorical. ulomatous arteritis with polyarthritis of juvenile onset. N Engl J Med The NOD2 gene encodes the NOD2 protein, a cytosolic 1982; 306:86–90. receptor for pathogen-associated molecular patterns, involved 13 Jabs DA, Houk JL, Bias WB, Arnett FC. Familial granulomatous in the innate immune response.16 All BS ⁄EOS-associated muta- synovitis, uveitis, and cranial neuropathies. Am J Med 1985; 78:801–4. tions show full penetrance, and are located on the NACHT 14 Gross KR, Malleson PN, Culham G et al. Vasculopathy with renal domain of the NOD2 protein, which is involved in self-oligo- artery stenosis in a child with sarcoidosis. J Pediatr 1986; 108:724– 10 merization. A recent study found six variants of NOD2 show- 6. ing increased basal nuclear factor-jB activity sharing a 15 Wang X, Kuivaniemi H, Bonavita G et al. CARD15 mutations in common genetic aetiology with BS. familial granulomatosis syndromes: a study of the original Blau In conclusion, we report herein a sporadic case of EOS due syndrome kindred and other families with large-vessel arteritis and to the recurrent missense mutation R334W on the NOD2 gene, cranial neuropathy. Arthritis Rheum 2002; 46:3041–5. 16 Inohara N, Nun˜ez G. NODs: intracellular proteins involved in which probably occurs as a de novo mutation. As in previously inflammation and apoptosis. Nat Rev Immunol 2003; 3:371–82. reported EOS cases, our patient was initially misdiagnosed as

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1257–1259 Gene Corner

COL7A1 mutational analysis in Korean patients with dystrophic epidermolysis bullosa

DOI: 10.1111/j.1365-2133.2007.08191.x fied as previously reported.3 Polymerase chain reaction (PCR) products were subsequently sequenced. In all mutations other Dystrophic epidermolysis bullosa (DEB) is an inherited blister- than nonsense mutations, 100 control alleles were studied to ing skin disorder, characterized by mucocutaneous blistering, rule out the possibility that the putative mutation might be a scarring and nail dystrophy following minor trauma. DEB is frequent polymorphism. caused by mutations in the COL7A1 gene and occurs either as To assess the potential pathogenicity of the mutations found an autosomal dominant (DDEB, MIM 131750) or as a reces- in this study, information theory splice site analysis was car- sive (RDEB, MIM 226600) trait. The phenotypic variability ried out via the world wide web interface at http://splice. results from the different types of mutations in COL7A1 and cmh.edu/.4 their positions within the gene. In most cases, RDEB has a 1 more severe clinical presentation. The most severe type of Results and discussion DEB, the recessive Hallopeau–Siemens variant (HS-RDEB), is caused by the presence of mutations that lead to premature There are various reports of ethnicity-related or population- termination codons (PTCs) in both alleles, whereas the autoso- based mutational analyses of COL7A1, including studies in the mal dominant cases are frequently caused by heterozygous British, Mexican, Italian, Japanese, central European and U.S. glycine substitutions (GSs) within the collagenous triple helix. populations.2,5–10 This is the first report of a mutational analy- To date, more than 500 different mutations of COL7A1 have sis in Korean patients with DEB. The results demonstrated 30 been reported.2 In this study, mutational analysis was per- pathogenic COL7A1 mutations among a total of 33 alleles, formed in 18 distinct Korean families with DEB, and a com- which included 14 previously identified COL7A1 mutations putational study of each mutation was carried out. and 10 novel mutations (Table 1). No such mutations were found in the 50 unrelated controls. Among the 30 pathogenic Cases and methods mutations, we found 13 distinct PTC-causing mutations, which appear to be fairly evenly distributed throughout the Eighteen unrelated Korean families with DEB were enrolled in COL7A1 gene. We found five GSs and five alternative splicing this study (14 recessive, four dominant). DEB was first diag- mutations. All GS mutations were located in the collagenous nosed clinically, and the diagnosis was later confirmed by domain. Despite complete sequencing of all exonic sequences, immunofluorescence antigen mapping and electron micros- including exon–intron borders, we could not detect patho- copy if the diagnosis was not definitive based on clinical find- genic mutations in three of the patients with RDEB. It is possi- ings. Four DEB families were diagnosed as having DDEB, ble that, in these cases, the mutations could reside in the based on family history and detection of a previously pub- intron or outside the coding regions, such as the promoter lished mutation associated with families with DDEB. RDEB region, which were not analysed in this study. cases were subclassified based on the following clinical fea- Three of the four patients with DDEB had GS mutations in tures: (i) a severe mutilating phenotype with extensive ero- exon 73 (two G2043R and one G2034R). Our data agree with sions and blistering since birth, pseudosyndactyly and joint other published reports that describe G2043R as a recur- contractures (HS-RDEB); (ii) mild, more localized involvement rent mutation in the DDEB population.2,9,11 The other GS, and lack of mutilating pseudosyndactyly (moderate-to-severe G2034R, has also been reported in DDEB cases.9,12 Interest- RDEB); (iii) mild blisters and scarring limited to trauma- ingly, the substitution of glycine with glutamic acid exposed sites (mild or mitis RDEB); and (iv) transient bullous (G2034E)5 or tryptophan (G2034W)13 at the same amino dermolysis of the newborn (TBDN, MIM 131705). All patients acid position has also been reported in DDEB cases. A muta- gave their written informed consent, and the experiments tion at the same codon, G2034E, was found one of the RDEB were approved by the Institutional Review Board at Yongdong patients in this study (patient 15); however, a DEB phenotype Severance Hospital which adheres to the Helsinki Guidelines. was not observed in the patient’s parents. Similar examples Genomic DNA was extracted from peripheral blood lym- were reported at amino acid positions 2028 and 2623. The phocytes of patients and their families. COL7A1 segments, mutation G2028R has been detected in a patient with DDEB including all 118 exons and exon–intron borders, were ampli- pruriginosa and a family with toenail dystrophy without skin

2007 The Authors 1260 Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1260–1264 COL7A1 mutational analysis in Korean patients with DEB, S-W. Oh et al. 1261

Table 1 Phenotypes and genotypes in Korean patients with dystrophic epidermolysis bullosa

Patient Phenotype Mutation Exon Effect 1 DDEB G2043R (6127GfiA)2,11,13,23–29 73 GS 2 Q2300R (6899AfiG)30 87 AS 3 G2043R (6127GfiA)2,11,13,23–29 73 GS 4 G2034R (6100GfiA)12,31–33 73 GS 5 TBDN G798R (2392GfiA) 6246del27 18 75 AS In-frame deletion 6 RDEB, mitis G1694C (5080GfiT) 682+1GfiA2,29,34–36 55 Intron 5 GS AS 7a G114V (341GfiT)19 R1933X (5797CfiT)19,36,37 3 70 AS PTC 8 Moderate-to-severe RDEB Q1211X (3631CfiT)38 Not determined 27 – PTC – 9 242delCA Not determined 2 – PTC – 10 R1730X (5188CfiT)37 Not determined 58 – PTC – 11 G2204S (6610GfiA)39 R669X (2005CfiT)2,40,41 82 15 GS PTC 12 R669X (2005CfiT)2,40,41 E2857X (8569GfiT)6,23,41–44 15 116 PTC PTC 13 2645del4 E2857X (8569GfiT)6,23,41–44 20 116 PTC PTC 14a G2204S (6610GfiA)39 S1689X (5066CfiG)39 82 55 GS PTC 15 HS-RDEB G2034E (6101GfiA)5 E2857X (8569GfiT)6,23,41–44 73 116 GS PTC 16 G798R (2392GfiA) Q1286X (3856CfiT) 18 31 AS PTC 2621ins5 (GCTTC) 20 PTC 17 R578X (1732CfiT)10,23,45–47 2063delC 13 16 PTC PTC 18 8291delGA 1094)1GfiC 111 Intron 8 PTC AS

aPreviously reported patients (patient 719 and patient 1439). Novel mutations are underlined. DDEB, dominant dystrophic epidermolysis bull- osa; TBDN, transient bullous dermolysis of the newborn; RDEB, recessive dystrophic epidermolysis bullosa; HS-RDEB, Hallopeau–Siemens RDEB; AS, alternative splicing; PTC, premature termination codon; GS, glycine substitution.

fragility, while the mutation G2028A has been implicated in a usually with marked improvement or even resolution in the family with classic DDEB.14,15 Similar reports showed the first few months to years of life. In previous reports of TBDN, mutations G2623S16 and G2623C17 in families with RDEB and the immunofluorescence finding was granular intracellular the pretibial form of DDEB, respectively. Patients with DDEB retention of type VII collagen and decreased anchoring usually harbour GS mutations within the collagenous region fibrils.20–22 The patient in our study had a history of localized of type VII collagen. One of the patients with DDEB in our blistering on her lower leg at birth, which resolved over a study, however, has the mutation Q2300R, which is a non-GS short period of time with scarring. However, we were unable mutation within the collagenous region. So far, only 14 muta- to carry out microscopic or immunofluorescence studies tions other than GS have been reported in the literature. because her parents refused consent for a skin biopsy. There Of these, 4084)1GfiC, 4120)1GfiC, 4669GfiC (G1557R), have been only three reports identifying the pathogenic muta- 5772+1GfiT, 6619GfiC (G2207R), 6899AfiG (Q2300R), tions involved in TBDN. One case involves a patient with 6900+1GfiT, 6900+4AfiG and 8045AfiG (K2682R) were compound heterozygosity for two glycine substitution muta- missense mutations, which all resulted in cryptic splicing. tions, G1519D and G2251E.20 Christiano et al.21 identified a There is one reported missense mutation (V760M) on the dominant acceptor splice site mutation in intron 35, NC-1 domain. The others are gross deletion mutations 4120GfiC, which resulted in in-frame skipping of exon 36. 6847del27, 6863del16, 6081del28 and 8068del17insGA that A dominant heterozygous glycine substitution mutation, are associated with non-GS DDEB.2,18 G1522E, has also been reported.22 The patient with TBDN Among 14 patients with RDEB, seven were classified as hav- described here was a compound heterozygote for a 2392GfiA ing moderate-to-severe RDEB and four as having HS-RDEB. (G798R) splicing site mutation and a 6246del27 in-frame PTC-causing mutations were common in both groups (82% in deletion mutation. We could not detect any common features moderate-to-severe RDEB; 67% in HS-RDEB). There were related to these genotypes in our patient with TBDN, although two patients with mitis-type RDEB, who showed G1694C ⁄ the splicing site mutation in this case could have ameliorated 682+1GfiA and R1933X ⁄341GfiT (G114V) combination the phenotype. Hammami-Hauasli et al.20 also proposed mutations. The 341GfiT mutation was previously reported to that the transitory blistering in TBDN has possibly more to do be a splicing mutation that results in deletion of 87 bp in the with the quantity rather than the quality of the collagen. noncollagenous domain.19 In general, splicing mutations We found five different recurrent mutations, R669X, generally result in a milder phenotype.19 2392GfiA (G798R), G2043R, G2204S and E2857X, which This study included one patient (patient 5, Fig. 1a,b) who account for approximately 30% of the DEB alleles in Korean was suspicious for a rare variant of DEB, TBDN, characterized patients. In particular, E2857X has been reported in a Japanese by transient blistering which manifests at birth or soon after, study on patients with DEB as an ethnic-specific recurrent

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1260–1264 1262 COL7A1 mutational analysis in Korean patients with DEB, S-W. Oh et al.

(a) (b)

Fig 1. The clinical features of the patient with transient bullous dermolysis of the newborn. There are only scars on the lower right leg, without any further blisters or erosions. (a) At 2 months after birth; (b) at 22 months after birth. mutation.6,23 Another recurrent mutation, R578X, has been ever, interpretation of this mutation is limited due to the small reported to be an ethnic-specific recurrent mutation, exclusive number of patients. to British patients.10 Murata et al.23 also demonstrated the The effect of a specific mutation on the splicing can be pre- absence of the recurrent mutation R578X in 42 non-British dicted by information theory. Changes in the affinity of a pro- patients, mainly Asian. However, the mutation studies on our tein for its binding site, such as splicing machinery, can be patients demonstrated that the R578X mutation is not exclu- estimated from the individual information content (Ri) of the sive to British patients. Among the five recurrent mutations in natural and variant sequences.4 Computational analysis identi- our study, G798R is a novel mutation, which seems to be an fied five splicing site mutations that have changed the Ri ethnic-specific mutation in Korean patients with DEB. How- enough to make a cryptic splicing site or to abolish the

(a) (b) Fig 2. (a) Reverse transcription–polymerase chain reaction analysis of the mutation 2392GfiA reveals one abnormal transcript of 181 bp in the sample from the patient (p) and the mother (m), in addition to the 298-bp band present in the control sample (c, p, m). (b) The mutation 2392GfiA (arrow) induces cryptic splicing 52 bp upstream from the normal splice site, which results in an (c) (d) out-of-frame shift and, consequently, a premature termination codon (PTC) in exon 19. Even if it does undergo normal splicing, there is an additional downstream PTC-forming mutation, 2621insGCTTC, in patient 16 who has a severe phenotype. (c) Direct sequencing of the 181-bp lower band reveals the expected cryptic splicing in exon 18, which results in a PTC in exon 19. (e) (f) (d) Direct sequencing of the 298-bp upper band reveals normally spliced sequences that include unaffected normal cDNA sequence and mutated sequences with normal splicing (arrow). (e, f) The mutation 1094)1GfiC results in two transcripts, one normal at 421 bp and one abnormal at 274 bp (p), whereas the transcript of the control reveals only a normal 421-bp transcript (c).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1260–1264 COL7A1 mutational analysis in Korean patients with DEB, S-W. Oh et al. 1263 original splicing site. Among them, the mutations 341GfiT mutation in recessive dystrophic epidermolysis bullosa. Hum Mutat (G114V), 682+1GfiA and 6899AfiG were previously 1997; 10:408–14. reported as splicing site mutations, and the mutations 4 Nalla VK, Rogan PK. Automated splicing mutation analysis by information theory. Hum Mutat 2005; 25:334–42. 1094)1GfiC and 2392GfiA are novel splicing site mutations. ) 5 Kern JS, Kohlhase J, Bruckner-Tuderman L et al. Expanding In the case of 1094 1GfiC, the mutation is predicted to abol- the COL7A1 mutation database: novel and recurrent mutations ish the natural acceptor site of exon 9 (Ri decreases from 9Æ5 and unusual genotype–phenotype constellations in 41 patients to 1Æ9 bits). Consequently, exon 9 was spliced out in our with dystrophic epidermolysis bullosa. J Invest Dermatol 2006; study (Fig. 2e,f). The mutation 2392GfiA also induces cryptic 126:1006–12. splicing. The Ri of the natural donor site of exon 18 (8Æ04) 6 Sawamura D, Goto M, Yasukawa K et al. Genetic studies of 20 Japa- remained unchanged, whereas the Ri of the cryptic donor site nese families of dystrophic epidermolysis bullosa. J Hum Genet 2005; 50:543–6. located 52 bases upstream of the natural site increased from 7 Gardella R, Castiglia D, Posteraro P et al. Genotype–phenotype cor- 5Æ5to8Æ0 bits. As a consequence, 52 nucleotides were deleted relation in Italian patients with dystrophic epidermolysis bullosa. from exon 18, causing a frameshift which resulted in a PTC J Invest Dermatol 2002; 119:1456–62. in exon 19 (Fig. 2a–d). The mutation 2392GfiA was detected 8 Salas-Alanis JC, Amaya-Guerra M, McGrath JA. The molecular basis in two phenotypes, TBDN (patient 5) and HS-RDEB (patient of dystrophic epidermolysis bullosa in Mexico. Int J Dermatol 2000; 16). The presence of another downstream PTC-forming muta- 39:436–42. tion on the same allele in exon 20 (2621ins5) of patient 16, 9 Jarvikallio A, Pulkkinen L, Uitto J. Molecular basis of dystrophic epidermolysis bullosa: mutations in the type VII collagen gene in addition to the PTC-forming mutation Q1296X on the (COL7A1). Hum Mutat 1997; 10:338–47. other allele, may result in the formation of no functional pro- 10 Mellerio JE, Dunnill MG, Allison W et al. Recurrent mutations in tein and result in a severe phenotype in that patient. In con- the type VII collagen gene (COL7A1) in patients with recessive trast, in patient 5 the splicing mutation 2392GfiA in addition dystrophic epidermolysis bullosa. J Invest Dermatol 1997; 109:246– to an in-frame deletion (6246del27) on the other allele results 9. in some partially functioning protein possibly from either 11 Mellerio JE, Salas-Alanis JC, Talamantes ML et al. A recurrent glycine allele and a milder TBDN phenotype (Fig. 2). In total, we substitution mutation, G2043R, in the type VII collagen gene (COL7A1) in dominant dystrophic epidermolysis bullosa. Br J Dermatol found five splicing site mutations, which comprise 20% (six 1998; 139:730–7. alleles) of the 30 total pathogenic mutations. This result is 12 Hammami-Hauasli N, Schumann H, Raghunath M et al. Some, but 2 similar to previous reports (about 17%). not all, glycine substitution mutations in COL7A1 result in intracel- In conclusion, mutational analysis of 18 distinct Korean lular accumulation of collagen VII, loss of anchoring fibrils, and DEB families revealed 30 pathogenic COL7A1 mutations among skin blistering. J Biol Chem 1998; 273:19228–34. a total of 33 alleles. In this study we used computational anal- 13 Rouan F, Pulkkinen L, Jonkman MF et al. Novel and de novo glycine ysis to predict cryptic splicing and confirmed the mis-splicing substitution mutations in the type VII collagen gene (COL7A1) in dys- trophic epidermolysis bullosa: implications for genetic counseling. caused by the novel mutations G798R and 1094)1GfiC using J Invest Dermatol 1998; 111:1210–13. reverse transcription–PCR. This is the first comprehensive 14 Nakamura H, Sawamura D, Goto M et al. The G2028R glycine sub- mutational analysis of Korean patients with DEB and expands stitution mutation in COL7A1 leads to marked inter-familiar clini- the genotype ⁄phenotype correlation database as well as identi- cal heterogeneity in dominant dystrophic epidermolysis bullosa. fying 10 novel and 14 previously identified mutations in the J Dermatol Sci 2004; 34:195–200. COL7A1 gene. 15 Murata T, Masunaga T, Shimizu H et al. Glycine substitution muta- tions by different amino acids in the same codon of COL7A1 lead to heterogeneous clinical phenotypes of dominant dystrophic epi- Department of Dermatology and S-W. OH dermolysis bullosa. Arch Dermatol Res 2000; 292:477–81. Cutaneous Biology Research Institute, J.S. LEE 16 Sawamura D, Mochitomi Y, Kanzaki T et al. Glycine substitution Yonsei University College of Medicine, M.Y. KIM mutations by different amino acids at the same codon in COL7A1 Yongdong Severance Hospital, 146-92 Dogok-dong, S-C. KIM cause different modes of dystrophic epidermolysis bullosa inheri- Kangnam-gu, Seoul 135-720, Korea tance. Br J Dermatol 2006; 155:834–7. Correspondence: Soo-Chan Kim. 17 Christiano AM, Lee JY, Chen WJ et al. Pretibial epidermolysis E-mail: [email protected] bullosa: genetic linkage to COL7A1 and identification of a glycine- to-cysteine substitution in the triple-helical domain of type VII collagen. Hum Mol Genet 1995; 4:1579–83. References 18 Sawamura D, Nizeki H, Miyagawa S et al. A novel indel COL7A1 mutation 8068del17insGA causes dominant dystrophic epidermoly- 1 Bruckner-Tuderman L, Hopfner B, Hammami-Hauasli N. Biology sis bullosa. Br J Dermatol 2006; 154:995–7. of anchoring fibrils: lessons from dystrophic epidermolysis bullosa. 19 Wessagowit V, Kim SC, Oh SW, McGrath JA. Genotype–pheno- Matrix Biol 1999; 18:43–54. type correlation in recessive dystrophic epidermolysis bullosa: 2 Varki R, Sadowski S, Uitto J et al. Epidermolysis bullosa. II. Type when missense doesn’t make sense. J Invest Dermatol 2005; VII collagen mutations and phenotype–genotype correlations in the 124:863–6. dystrophic subtypes. J Med Genet 2007; 44:181–92. 20 Hammami-Hauasli N, Raghunath M, Kuster W et al. Transient 3 Christiano AM, Hoffman GG, Zhang X et al. Strategy for identifica- bullous dermolysis of the newborn associated with compound tion of sequence variants in COL7A1 and a novel 2-bp deletion

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1260–1264 1264 COL7A1 mutational analysis in Korean patients with DEB, S-W. Oh et al.

heterozygosity for recessive and dominant COL7A1 mutations. 35 Lin GT, Chen SK, Liu CS et al. Splice site mutation in the type VII J Invest Dermatol 1998; 111:1214–19. collagen gene (COL7A1) in a Taiwanese family with recessive 21 Christiano AM, Fine JD, Uitto J. Genetic basis of dominantly inherited dystrophic epidermolysis bullosa. J Formos Med Assoc 2000; 99:693– transient bullous dermolysis of the newborn: a splice site mutation in 7. the type VII collagen gene. J Invest Dermatol 1997; 109:811–14. 36 Csikos M, Szocs HI, Laszik A et al. High frequency of the 425AfiG 22 Fassihi H, Diba VC, Wessagowit V et al. Transient bullous dermoly- splice-site mutation and novel mutations of the COL7A1 gene in sis of the newborn in three generations. Br J Dermatol 2005; central Europe: significance for future mutation detection strategies 153:1058–63. in dystrophic epidermolysis bullosa. Br J Dermatol 2005; 152:879– 23 Murata T, Masunaga T, Ishiko A et al. Differences in recurrent 86. COL7A1 mutations in dystrophic epidermolysis bullosa: ethnic- 37 Whittock NV, Ashton GH, Mohammedi R et al. Comparative muta- specific and worldwide recurrent mutations. Arch Dermatol Res 2004; tion detection screening of the type VII collagen gene (COL7A1) 295:442–7. using the protein truncation test, fluorescent chemical cleavage of 24 Christiano AM, Morricone A, Paradisi M et al. A glycine-to-arginine mismatch, and conformation sensitive gel electrophoresis. J Invest substitution in the triple-helical domain of type VII collagen in a Dermatol 1999; 113:673–86. family with dominant dystrophic epidermolysis bullosa. J Invest Der- 38 Pulkkinen L, Uitto J. Mutation analysis and molecular genetics of matol 1995; 104:438–40. epidermolysis bullosa. Matrix Biol 1999; 18:29–42. 25 Winberg JO, Hammami-Hauasli N, Nilssen O et al. Modulation of 39 Kim J, Kim SC, Yasukawa K et al. Compound heterozygosity for disease severity of dystrophic epidermolysis bullosa by a splice site premature termination codon and glycine substitution mutations in mutation in combination with a missense mutation in the COL7A1 the COL7A1 gene in Korean siblings with a moderately severe phe- gene. Hum Mol Genet 1997; 6:1125–35. notype of recessive dystrophic epidermolysis bullosa. J Dermatol Sci 26 Cserhalmi-Friedman PB, Karpati S, Horvath A et al. Identification of 2003; 33:180–3. the glycine-to-arginine substitution G2043R in type VII collagen in 40 Cserhalmi-Friedman PB, McGrath JA, Mellerio JE et al. Restoration a family with dominant dystrophic epidermolysis bullosa from of open reading frame resulting from skipping of an exon with Hungary. Exp Dermatol 1997; 6:303–7. an internal deletion in the COL7A1 gene. Lab Invest 1998; 27 Klingberg S, Mortimore R, Parkes J et al. Prenatal diagnosis of domi- 78:1483–92. nant dystrophic epidermolysis bullosa, by COL7A1 molecular analy- 41 Yonei N, Ohtani T, Furukawa F. Recessive dystrophic epidermolysis sis. Prenat Diagn 2000; 20:618–22. bullosa: case of non-Hallopeau–Siemens variant with premature ter- 28 Wessagowit V, Ashton GH, Mohammedi R et al. Three cases of mination codons in both alleles. J Dermatol 2006; 33:802–5. de novo dominant dystrophic epidermolysis bullosa associated with 42 Shimizu H, McGrath JA, Christiano AM et al. Molecular basis of the mutation G2043R in COL7A1. Clin Exp Dermatol 2001; 26:97– recessive dystrophic epidermolysis bullosa: genotype ⁄phenotype 9. correlation in a case of moderate clinical severity. J Invest Dermatol 29 Dang N, Klingberg S, Marr P, Murrell DF. Review of collagen VII 1996; 106:119–24. sequence variants found in Australasian patients with dystrophic 43 Shibusawa Y, Negishi I, Ishikawa O. Compound heterozygosity epidermolysis bullosa reveals nine novel COL7A1 variants. J Dermatol in sibling patients with recessive dystrophic epidermolysis bullosa Sci 2007; 46:169–78. associated with a mild phenotype. Int J Dermatol 2006; 45:302–5. 30 Jiang W, Bu D, Yang Y et al. A novel splice site mutation in colla- 44 Suzuki S, Shimomura Y, Yamamoto Y et al. A case of recessive gen type VII gene in a Chinese family with dominant dystrophic dystrophic epidermolysis bullosa caused by compound heterozy- epidermolysis bullosa pruriginosa. Acta Derm Venereol (Stockh) 2002; gous mutations in the COL7A1 gene. Br J Dermatol 2006; 155:838– 82:187–91. 40. 31 Kon A, Nomura K, Pulkkinen L et al. Novel glycine substitution 45 Dunnill MG, Richards AJ, Milana G et al. A novel homozygous point mutations in COL7A1 reveal that the Pasini and Cockayne–Touraine mutation in the collagen VII gene (COL7A1) in two cousins with variants of dominant dystrophic epidermolysis bullosa are allelic. recessive dystrophic epidermolysis bullosa. Hum Mol Genet 1994; J Invest Dermatol 1997; 109:684–7. 3:1693–4. 32 Chen X, Li G, Zhu X. Study on COL7A1 gene mutation in a epi- 46 Mohammedi R, Mellerio JE, Ashton GH et al. A recurrent COL7A1 dermolysis bullosa pruriginosa family. Zhonghua Yi Xue Za Zhi 2000; mutation, R2814X, in British patients with recessive dystrophic epi- 80:869–71. dermolysis bullosa. Clin Exp Dermatol 1999; 24:37–9. 33 Mecklenbeck S, Hammami-Hauasli N, Hopfner B et al. Clustering of 47 Pfendner EG, Nakano A, Pulkkinen L et al. Prenatal diagnosis for COL7A1 mutations in exon 73: implications for mutation analysis in epidermolysis bullosa: a study of 144 consecutive pregnancies at dystrophic epidermolysis bullosa. J Invest Dermatol 1999; 112:398– risk. Prenat Diagn 2003; 23:447–56. 400. 34 Hovnanian A, Rochat A, Bodemer C et al. Characterization of 18 Accepted for publication: 27 June 2007 new mutations in COL7A1 in recessive dystrophic epidermolysis bullosa provides evidence for distinct molecular mechanisms under- Key words: COL7A1, dystrophic epidermolysis bullosa, Korean, mutation lying defective anchoring fibril formation. Am J Hum Genet 1997; Conflicts of interest: none declared. 61:599–610.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1260–1264 Gene Corner

A novel homozygous mutation of the EVER1 ⁄TMC6 gene in a Japanese patient with epidermodysplasia verruciformis

DOI: 10.1111/j.1365-2133.2007.08206.x (a) Epidermodysplasia verruciformis (EV; OMIM 226400) is a rare, autosomal recessive genetic disease characterized by abnormal susceptibility to disease-specific human papilloma- viruses (HPVs). Patients with EV have usually presented with skin lesions resembling disseminated flat warts or pityriasis versicolor since childhood. Some of the lesions may progress to skin cancers late in life. Recently, genetic susceptibility loci for EV were mapped to chromosomes 2p21–24 and 17q25.1 Two different genes, EVER1 ⁄TMC6 and EVER2 ⁄ TMC8, were identified in chromosome 17q25.2 We report a woman with HPV type 12-induced EV who carried a homo- zygous CfiT transition at nucleotide position 220 within (b) exon 4 of EVER1 that led to a premature translation termination.

Case and methods

A 57-year-old woman presented with a 46-year history of multiple, scaly erythematous plaques on her neck, trunk and extremities (Fig. 1a). There was consanguinity between the patient’s parents. Of her five siblings, a brother had died soon after birth, and two sisters had died of uncertain causes at the ages of 3 and 5 years. The remaining two siblings and her two daughters have been in good health. Lesional skin biopsy showed vacuolated keratinocytes with abundant basophilic cytoplasm (Fig. 1b). HPV type 12, one of the EV-related HPV Fig 1. (a) Scaly reddish-brown plaques were found on the chest and types, was detected in DNA samples extracted from the scales neck. (b) Swollen irregular-shaped keratinocytes which have by using the consensus primer pair, EV3 and EV4, for the basophilic cytoplasm (haematoxylin and eosin). polymerase chain reaction (PCR) amplification of the E7 open reading frame of EV-associated HPVs.3 Laboratory tests showed a normal full blood count with normal white blood cell dif- and EVER2 were amplified by AccuPrime Taq DNA Polymer- ferentiation. No abnormality was found in blood chemistry ase System (Invitrogen Corp., Carlsbad, CA, U.S.A.) using tests except for an elevated level of lactate dehydrogenase gene-specific primers.2 The PCR products were run on 2% ) (393 IU L 1, normal 120–240). The CD4 ⁄CD8 ratio of circu- agarose gels, purified by QIAquick Gel Extraction Kit (Qia- lating lymphocytes was within normal range (1Æ0; normal gen), and analysed by direct sequencing (ABI Prism 3100 0Æ9–3Æ2). Values of thymidine uptake by lymphocytes stimu- Genetic Analyzer; Applied BioSystems, Foster City, CA, lated with phytohaemagglutinin and concanavalin A were U.S.A.). comparable with those of normal subjects. After informed consent, genomic DNA was extracted from Results and discussion a blood sample using the FlexiGene DNA Extraction Kit (Qiagen, Tokyo, Japan). Pooled DNA samples were used for A homozygous CfiT transition at nucleotide position 220 control studies. With the patient’s DNA, all exons of EVER1 was identified within exon 4 of the EVER1 gene (Fig. 2a). This

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1265–1266 1265 1266 Novel homozygous mutation of EVER1 ⁄TMC6 in a Japanese patient with EV, S. Aochi et al.

(a) (b)

Fig 2. (a) Mutational analysis of EVER1. A homozygous 220CfiT (Q74X) mutation was identified in exon 4 of the EVER1 gene. (b) Restriction fragment length polymorphism analysis of polymerase chain reaction (PCR) fragment amplified from exons 3–4 of EVER1 gene using the BfaI restriction enzyme. This (c) enzyme fully cleaved the PCR product derived from the patient (lane P1, precleaved; lane P2, postcleaved). The PCR product from the wild-type allele was not cleaved (lane C1, precleaved; lane C2, postcleaved). (c) Summary of EVER1 mutations in epidermodysplasia verruciformis, modified from Ramoz et al.2 mutation shifted a CAG (coding for glutamine at amino acid References position 74) triplet to a TAG (Q74X). As the CfiT transition led to a BfaI restriction site, we performed restriction fragment 1 Ramoz N, Taı¨eb A, Rueda LA et al. Evidence for a nonallelic hetero- geneity of epidermodysplasia verruciformis with two susceptibility length polymorphism analysis of the PCR fragment of 565 bp, loci mapped to chromosome regions 2p21-p24 and 17q25. J Invest encompassing exons 3 and 4 of the EVER1 gene. The PCR Dermatol 2000; 114:1148–53. products obtained from the patient’s DNA were cleaved by 2 Ramoz N, Rueda LA, Bouadjar B et al. Mutations in two adjacent BfaI to form 401-bp and 164-bp fragments (Fig. 2b), whereas novel genes are associated with epidermodysplasia verruciformis. no cleavage was observed in those from 54 healthy control Nat Genet 2002; 32:579–81. blood samples (108 alleles). 3 Adachi A, Kiyono T, Hayashi Y et al. Detection of human papillo- Two single nucleotide polymorphisms (SNPs), CfiT transi- mavirus (HPV) type 47 DNA in malignant lesions from epiderm- odysplasia verruciformis by protocols for precise typing of related tion at nucleotide position 373 and CfiT transition at nucleo- HPV DNAs. J Clin Microbiol 1996; 34:369–75. tide position 457, were detected. Both have already been 4 Lutzner MA. Epidermodysplasia verruciformis: an autosomal reces- registered in the SNP database at the National Center for Bio- sive disease characterized by viral warts and skin cancer. A model technology Information (Bethesda, MD, U.S.A.) (rs2748427 for viral oncogenesis. Bull Cancer 1978; 65:169–82. and rs12449858). 5 Androphy EJ, Dvoretzky I, Lowy DR. X-linked inheritance of Most patients with EV have autosomal recessive inheri- epidermodysplasia verruciformis: genetic and virologic studies of a tance,4 while a few patients have been reported having kindred. Arch Dermatol 1985; 121:864–8. 6 McKusick VA. Mendelian Inheritance in Man. A Catalog of Human Genes and X-linked recessive or autosomal dominant,5,6 indicating the Genetic Disorders. Baltimore: Johns Hopkins University Press, 1998. heterogeneity of EV. To date, seven mutations of the EVER1 7 Tate G, Suzuki T, Kishimoto K, Mitsuya T. Novel mutations of (Fig. 2c) and three of the EVER2 genes have been identi- EVER1 ⁄TMC6 gene in a Japanese patient with epidermodysplasia 2,7–10 fied, but no accumulation of specific mutation sites has verruciformis. J Hum Genet 2004; 49:223–5. been found. The present case had a novel mutation at nucleo- 8 Sun XK, Chen JF, Xu AE. A homozygous nonsense mutation in the tide position 220 in exon 4 of EVER1. This mutation is the EVER2 gene leads to epidermodysplasia verruciformis. Clin Exp third found in Japanese patients.7 Dermatol 2005; 30:573–4. 9 Zuo YG, Ma D, Zhang Y et al. Identification of a novel mutation The genes associated with EV, EVER1 and EVER2, belong to 11 and a genetic polymorphism of EVER1 gene in two families with the TMC (transmembrane channel-like) gene family and are epidermodysplasia verruciformis. J Dermatol Sci 2006; 44:153–9. now termed TMC6 and TMC8. The EVER proteins localize in 10 Gober MD, Rady PL, He E et al. Novel homozygous frameshift the endoplasmic reticulum with features of integral mem- mutation of EVER1 gene in an epidermodysplasia verruciformis brane proteins,2 but the function of EVER proteins in the patient. J Invest Dermatol 2007; 127:817–20. development of persistent HPV infections remains to be 11 Keresztes G, Mutai H, Heller S. TMC and EVER genes belong to a discovered. larger novel family, the TMC gene family encoding transmembrane proteins. BMC Genomics 2003; 4:24–35.

Departments of Dermatology and S. AOCHI Accepted for publication: 4 July 2007 *Pediatrics, Okayama University, G. NAKANISHI Graduate School of Medicine, N. SUZUKI Key words: epidermodysplasia verruciformis, EVER1, homozygous mutation, Dentistry and Pharmaceutical Sciences, N. SETSU nonsense mutation 2-5-1 Shikata-cho, Okayama City 700-8558, Japan D. SUZUKI Conflicts of interest: none declared. Correspondence: G. Nakanishi K. AYA* S.A. and G.N. contributed equally to this study. E-mail: [email protected] K. IWATSUKI

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1265–1266 Correspondence

A case of primary anetoderma in an infant matory skin diseases and did not show any symptom associ- ated with the development of the skin lesions. He was born DOI: 10.1111/j.1365-2133.2007.08199.x after a full-term single pregnancy and had no problems during the perinatal period. The family history was not contributory. SIR, Anetoderma is a rare cutaneous disorder characterized by On clinical examination, the otherwise healthy-looking boy a localized depression or outpouching of the skin caused by presented with 40–50 atrophic and sac-like round to oval- laxity and weakening of the dermal connective tissue as a shaped whitish papules on the trunk, buttocks and both result of focal loss of elastic fibres.1–3 It is sometimes not asso- extremities. The lesions were 2–20 mm in diameter and were ciated with any underlying disease [primary (idiopathic) ane- well circumscribed, presenting either as a depression below toderma] or it can be related to many kinds of dermatoses the level of normal skin or as a sac-like protrusion. Routine (secondary anetoderma).1–3 In the past, cases of primary ane- laboratory examinations were within normal limits. There toderma were classified into the Jadassohn–Pelizzari type, fol- were no neurological, orthopaedic or ophthalmological abnor- lowing erythema or urticaria, and the Schweninger–Buzzi type malities. with no preceding inflammatory skin lesions.1–3 Currently, A biopsy was performed from a typical looking lesion on this classification is becoming less important because, whether his back. Histopathological examination revealed a mild peri- or not one can recognize this feature, it is only an individual vascular infiltrate with lymphocytes in the mid-dermis variable, not a disease-defining feature.1–3 Anetoderma occurs (Fig. 2a). The elastic stain demonstrated a decreased number mainly in middle-aged women, often in children and rarely in of elastic fibres in the superficial and the mid-dermis infants.1,2,4 (Fig. 2b). A 12-month period of treatment with clobetasone ) We report an interesting case of primary anetoderma, the butyrate (0Æ5mgg 1) cream showed no effective response Schweninger–Buzzi type, which occurred in a 1-month-old and, during that time, we noticed about 10–15 newly devel- infant. To our knowledge, this is the youngest reported age of oped skin lesions. We are still observing the patient every onset of a case that is not congenital. 2–3 months. A 5-month-old boy presented with increasing numbers of Anetoderma, which was first described by Jadassohn in multiple atrophic and sac-like whitish papules on his trunk 1892, is characterized clinically by small, atrophic papules and both extremities, which started to develop only 1 month with flaccid skin, inward herniation at palpation and histo- after birth (Fig. 1a,b). He had no history of preceding inflam- pathological loss of elastic tissue. This rare disorder occurs

(a) (b)

Fig 1. (a) Multiple atrophic and sac-like whitish papules are noted on the posterior trunk. (b) Close-up view of the left buttock.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1267 1268 Correspondence

(a) (b)

Fig 2. (a) A mild perivascular lymphocytic infiltrate is shown in the mid-dermis (haematoxylin and eosin stain; original magnification · 200). (b) Loss of elastic fibre is seen in the superficial and the mid-dermis (elastic stain; original magnification · 200).

mainly in women aged 20–40 years, but is occasionally The diagnosis of anetoderma is based on an association of reported in younger and older patients of both sexes.1–3 Also, clinical features, including localized areas of flaccid skin with anetoderma in premature infants has been described in some loss of dermal substance, a characteristic herniation phenome- cases, and it may have been related to the use of cutaneous non and the histological loss of elastic tissue in the mid-der- monitoring leads or adhesives.1,2 Predilection sites for aneto- mis.1–4 There is no known effective treatment. Several derma are the trunk and proximal portions of the extremities anecdotal reports of treatment with aspirin, aminocaproic acid, and, less commonly, on the neck and face.1,2 The number of phenytoin, dapsone, vitamin E, niacin, penicillin, colchicine lesions varies widely, from fewer than five to 100 or more.1,2 and hydroxychloroquine have shown variable short-term The lesions remain unchanged throughout life and new results; however, long-term outcome was generally unsuccess- lesions often continue to develop for many years.1,2,5 ful.1–4 Only prompt treatment of any preceding inflammatory We searched Medline using the combination of terms ‘pri- conditions or surgical excision of cosmetically unacceptable mary’, ‘idiopathic’, ‘anetoderma’ and ‘Schweninger–Buzzi lesions has any benefits.1,2 type’ and reviewed all the listed articles and abstracts to find a Although isolated and perhaps coincidental, numerous total of 10 cases reported worldwide in childhood (Table 1) abnormalities have been reported in patients with aneto- that satisfied both conditions of ‘primary anetoderma’ and ‘no derma.3,4 These include mainly abnormalities of the eyes, preceding inflammation’.3–10 The mean age at onset of the bones, endocrine organs and heart, which suggests that aneto- reported cases was 8Æ7 years. Other cases described as aneto- derma is at least sometimes part of a more generalized dis- derma of prematurity and congenital or familiar anetoderma order.3,4 The diagnosis of primary anetoderma can be were excluded from our research. established only by excluding the presence of any of the

Table 1 Primary anetoderma in childhood: review of the literature

Age Inflammatory Systemic Case Reference Sex (years) Localization of anetoderma skin lesions diseases 1 Venencie et al.3 M1Æ5 Face, neck, upper trunk, shoulders, arms, legs )) 2 F 11 Face, neck, trunk, shoulders, arms, buttock, legs )) 3F7Æ5 Face, neck, trunk, shoulders, arms, legs )) 4 M 8 Neck, forearms )) 5 M 13 Face, neck, trunk, arms, forearms, thighs )) 6 Karrer et al.4 M 7 Neck )) 7 M 9 Upper trunk, both extremities )) 8 Aghaei et al.5 M 14 Both extremities )) 9 Woo et al.9 F 6 Trunk )) 10 Ponnighaus et al.10 M10 )))

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1269 diseases known to be generalized or associated with secondary (a) atrophy.3,8 Therefore, it will be necessary to observe our patient for the future development of any of these abnor- malities.8 Herein, we report a case of what is, to our knowledge, pri- mary anetoderma in an infant with the earliest known age of onset.

Department of Dermatology, H-J. YU Hanyang University Guri Hospital, H. SHIN Hanyang University College of Medicine, M-S. KANG Guri-Si, Kyonggi-do 471-701, Korea J-S. KIM Correspondence: J-S. Kim. E-mail: [email protected] (b)

References

1 Maari C, Powell J. Atrophies of connective tissue. In: Dermatology (Bolognia JL, Jorizzo J, Rapini RP, eds), Vol. 2. London: Mosby, 2003; 1540–2. 2 Burgdorf W. Anetoderma and other atrophic disorders of the skin. In: Fitzpatrick’s Dermatology in General Medicine (Freedberg IM, Eisen AZ, Wolff K et al., eds), 6th edn, Vol. 1. New York: McGraw-Hill, 2003; 1027–9. 3 Venencie P, Winkelmann R, Moore B. Anetoderma: clinical findings, associations, and long-term follow-up evaluations. Arch Dermatol 1984; 120:1032–9. 4 Karrer S, Szeimies R, Stolz W et al. Primary anetoderma in children: Fig 1. Clinical appearance (a) before and (b) after treatment with report of two cases and literature review. Pediatr Dermatol 1996; tissue adhesive. 13:382–5. 5 Aghaei S, Sodaifi M, Aslani FS et al. An unusual presentation of anetoderma: a case report. BMC Dermatol 2004; 19:4–9. 6 Ricci RM, Meffert JJ, McCollough ML. Primary anetoderma. Cutis area with clobetasol propionate ointment, Fucibet cream 1998; 62:101–3. (betamethasone valerate + fusidic acid; Leo, Princes Risbor- 7 Romani J, Perez F, Llobet M et al. Anetoderma associated with anti- ough, U.K.), Tri-Adcortyl cream (triamcinolone acetonide, phospholipid antibodies: case report and review of the literature. gramicidin, neomycin + nystatin; Bristol-Myers Squibb, J Eur Acad Dermatol Venereol 2001; 15:175–8. Uxbridge, U.K.), Trimovate cream (clobetasone butyrate, 8 Sparsa A, Piette JC, Wechsler B et al. Anetoderma and its prothrom- botic abnormalities. J Am Acad Dermatol 2003; 49:1008–12. oxytetracycline + nystatin; GlaxoSmithKline, Uxbridge, U.K.), 9 Woo HJ, Park CJ, Yi JY. A case of primary anetoderma. Korean J tacrolimus 0Æ1% ointment, pimecrolimus 1% cream, benzoin Dermatol 1999; 37:951–3. tincture compound, calcitriol ointment and Dovobet ointment 10 Ponnighaus JM, Jaeger G, Baum HP. Anetoderma Schwenninger– (betamethasone dipropionate + calcipotriol; Leo) had not Buzzi in a dark-skinned patient. Hautarzt 2001; 52:950–1 (in German). provided any prolonged benefit. Duoderm (ConvaTec, Key words: infant, primary anetoderma, Schweninger–Buzzi type Uxbridge, U.K.) dressings locally and Cavilon (3M, Lough- borough, U.K.) skin barrier film had also been unhelpful. Her Conflicts of interest: none declared. psoriasis had been recalcitrant to systemic treatment with methotrexate, mycophenolate mofetil, Fumaderm (mono- ethylfumarate + dimethylfumarate; Fumedica, Herne, Ger- many) and pioglitazone. Hydroxycarbamide and ciclosporin had been tried but were discontinued due to the development Successful treatment of severe psoriatic natal of side-effects. She has subsequently been commenced on cleft fissuring with tissue adhesive etanercept. In an attempt to offer symptomatic relief for the fissuring, DOI: 10.1111/j.1365-2133.2007.08186.x LiquiBand tissue adhesive (MedLogic Ltd, Plymouth, U.K.) 0Æ5 g was applied to the central fissure with yellow soft paraf- SIR, A 43-year-old woman with chronic plaque and flexural pso- fin to the peripheral margins, and repeated twice weekly over riasis had been particularly troubled for 7 years by fissuring of a period of 5 weeks. This resulted in resolution of the fissure the natal cleft (Fig. 1a). This area frequently bled and caused (Fig. 1b) which has remained healed now for 7 months, severe pain and discomfort on sitting. Topical treatments to this despite her psoriasis elsewhere continuing to be extensive.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1270 Correspondence

To our knowledge, this is the first report of tissue adhesive loss. The peripheral blood leucocyte count increased to ) ) being an effective treatment for psoriatic skin fissuring. As this 20Æ3 · 109 L 1, containing 2Æ4 · 109 L 1 eosinophils and ) can be a severely disabling complication of flexural psoriasis, 7Æ9 · 109 L 1 atypical lymphocytes. Although the dominant we propose that this treatment should be considered in T-cell clone was increased in number, nonclonal CD4+ cells patients not responding to conventional therapies. bearing various TCR Vb chains were also expanded, resulting in a marked increase of the CD4 ⁄CD8 ratio to 97. Flow cyto- Department of Dermatology, V.H. SMITH metric analysis demonstrated a decrease of IFN-c+ ⁄IL-4) cells Clatterbridge Hospital, Bebington, Clatterbridge, S.K. JONES in the CD4+ fraction (Th1) to 0Æ1% (reference value 11Æ0– Wirral, Merseyside CH63 4JY, U.K. 44Æ0%), as compared with 2Æ4% IFN-c– ⁄IL-4+ cells (Th2) E-mail: [email protected] (reference value < 5%). In January 2004, the patient began to complain of lum- Key words: psoriatic fissure, tissue adhesive bago and leg pain due to spondylitis. Around that time, the Conflicts of interest: none declared. erythroderma and alopecia improved gradually, associated with disappearance of the dominant T-cell clone and normal- ization of the CD4 ⁄CD8 ratio, eosinophil counts and soluble IL-2 receptor levels. Based on X-ray results, treatment with isoniazid, rifampicin and pyrazinamide was started under a Recovery from Se´zary syndrome following tentative diagnosis of tuberculous spondylitis. In March Mycobacterium avium spondylitis 2005, improvement of erythroderma enabled us to obtain tissue samples from the spine, from which Mycobacterium avium DOI: 10.1111/j.1365-2133.2007.08187.x was cultured. Th1-type cells increased to 19Æ2%, resulting in a Th1 ⁄Th2 ratio of 5Æ2. The patient has been free from ery- SIR,Se´zary syndrome is a leukaemic variant of erythrodermic throderma for 2 years, although the M. avium spondylitis has cutaneous T-cell lymphoma characterized by the expansion of persisted. 1,2 malignant Th2 cell clones. Previous investigators have Our patient’s clinical manifestations became aggravated reported that superantigens produced by Staphylococcus aureus upon S. aureus infection, associated with the expansion of may exacerbate Se´zary syndrome with a Th2-dominant cyto- nonclonal CD4+ T cells incapable of producing IFN-c. After 3 kine shift; in contrast, the malignant Th2 clones may dis- M. avium spondylitis, however, the cutaneous symptoms appear from the peripheral blood in association with an improved with a Th1 cytokine shift, followed by disappear- increase in interferon (IFN)-c production and decreased pro- ance of the expanded CD4+ T cells including the malignant 4 duction of interleukin (IL)-4. clone. Although the dominant clone became negative by In April 2002, a 65-year-old woman was referred to us Southern blot, the residual T-cell clone was detected by because of a 2-year history of erythroderma and hair loss. more sensitive, polymerase chain reaction-based TCR-c gene Physical examination revealed generalized desquamative analysis. dermatitis, alopecia on the scalp, and swollen lymph nodes in the neck, axillae and inguinal areas. Skin biopsy showed Department of Dermatology, A. YAMADA dense infiltration of lymphocytes in the upper dermis with Okayama University Graduate School of Medicine, O. YAMASAKI epidermotropism, in which atypical lymphocytes with a Dentistry, and Pharmaceutical Sciences, K. ASAGOE convoluted nucleus were present. Lymph node biopsy 2-5-1 Shikata-cho, Okayama 700-8558, Japan K. TSUJI revealed features of dermatopathic lymphadenitis. Laboratory *Department of Orthopedics, National Hospital T. HAMADA 9 )1 results revealed a white blood cell count of 10Æ5 · 10 L , Organization Minami-Okayama Medical Center, Y. OTA* containing 22Æ0% atypical lymphocytes with a CD3+, Okayama, Japan K. IWATSUKI ) CD4+, CD5+, CD7 phenotype. The CD4 ⁄CD8 ratio Correspondence: Keiji Iwatsuki increased to 21Æ5 with a clonal expansion of T cells con- E-mail: [email protected] firmed by Southern blot analysis of T-cell receptor (TCR) b genes. Serum levels of lactate dehydrogenase and soluble ) IL-2 receptors were increased to 781 and 1958 U L 1 (normal References 190–650), respectively. Serology for human T-cell leukaemia 1 Vowels BR, Lessin SR, Cassin M et al. Th2 cytokine mRNA expression virus-1 was negative. The patient was diagnosed as having in skin in cutaneous T-cell lymphoma. J Invest Dermatol 1994; Se´zary syndrome, and was treated with photochemotherapy 103:669–73. and prednisone. 2 Dummer R, Heald PW, Nestle FO et al. Se´zary syndrome T-cell In May and October 2003, the patient had two episodes of clones display T-helper 2 cytokines and express the accessory factor- 1 (interferon-c receptor b-chain). Blood 1996; 88:1383–9. severe cellulitis caused by methicillin-resistant S. aureus carrying 3 Jackow CM, Cather JC, Arne V et al. Association of erythrodermic cu- 5 the Panton–Valentine leukocidin gene as previously described. taneous T-cell lymphoma, superantigen-positive Staphylococcus aureus, After the staphylococcal infections, the desquamative der- and oligoclonal T-cell receptor Vb gene expression. Blood 1997; matitis became worse, accompanied by complete scalp hair 89:32–40.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1271

4 Yoo EK, Cassin M, Lessin SR, Rook AH. Complete molecular remis- the efficacy of this treatment in this disease and observed sion during biologic response modifier therapy for Se´zary syndrome important infectious complications in two patients, one of is associated with enhanced helper T type 1 cytokine produc- them associated with late-onset neutropenia (LON).1 In our tion and natural killer cell activity. J Am Acad Dermatol 2001; 45:208– opinion, this is very important. Even though the occurrence 16. 5 Yamasaki O, Morizane S, Akiyama H, Iwatsuki K. Tendon destruc- of early and delayed-onset cytopenia has been described, par- 2–5 tion induced by Panton–Valentine leukocidin-positive Staphylococcus ticularly in patients with oncohaematological diseases, to aureus in a patient with Se´zary syndrome. Br J Dermatol 2005; our knowledge LON has not been reported previously in 152:586–7. patients with autoimmune diseases who have received ritux- 6,7 Key words: interferon-c, Mycobacterium infection, Se´zary syndrome, imab. We have recently treated a patient with PV who also staphylococcal infection, Th2 cytokine developed LON after treatment with rituximab. This patient had persistent, extensive cutaneous bullae for 1 year which Conflicts of interest: none declared. had failed to respond to oral corticosteroids, azathioprine, intravenous immunoglobulin and plasmapheresis. For this rea- son we started adjuvant treatment with rituximab (weekly ) infusions of 375 mg m 2 for 4 weeks) together with azathio- prine 150 mg daily and prednisone 20 mg daily. Clinical Late-onset neutropenia following rituximab improvement was spectacular, achieving complete remission treatment in patients with autoimmune within 5 months following rituximab and a progressive reduc- diseases tion of steroids to deflazacort 15 mg daily (equivalent to pred- nisone 10 mg daily). On day 191 after the last dose of DOI: 10.1111/j.1365-2133.2007.08189.x rituximab, the patient experienced severe neutropenia with an ) absolute neutrophil count of 0Æ361 · 109 L 1, along with an 1 SIR, In the May 2007 issue of this Journal, Goh et al. reported episode of fever of an unknown origin that resolved with the results of rituximab therapy as adjuvant treatment for cefepime and granulocyte-colony stimulating factor within pemphigus vulgaris (PV) in five patients. The authors analysed 5 days.

Table 1 Characteristics of the patients

Sex ⁄age Indication for rituximab Follow-up Late-onset Patient (years) treatment Concomitant treatment (days) neutropenia 1M⁄41 Dermatomyositis Azathioprine, hydroxychloroquine, prednisone 327 No 2F⁄27 Microscopic polyangiitis Mycophenolate mofetil, prednisone 616 No 3M⁄69 Wegener’s granulomatosis Prednisone, co-trimoxazole, cyclophosphamide 692 No 4M⁄48 Thrombotic thrombocytopenic Prednisone 758 No purpura 5M⁄23 Systemic lupus erythematosus Methotrexate, mepacrine, prednisone 607 No 6F⁄27 Pemphigus vulgaris Azathioprine, prednisone 225 Yes 7M⁄35 Systemic lupus erythematosus Deflazacort, hydroxychloroquine, aspirin 806 No 8M⁄46 Wegener’s granulomatosis Methotrexate, prednisone, co-trimoxazole 504 No 9F⁄49 Dermatomyositis and rheumatoid Prednisone, hydroxychloroquine, intravenous 486 No arthritis immunoglobulin, cyclophosphamide 10 M ⁄35 Systemic lupus erythematosus Deflazacort, hydroxychloroquine, aspirin 522 No 11 M ⁄66 Microscopic polyangiitis Prednisone, cyclophosphamide, co-trimoxazole 218 No 12 F ⁄37 Wegener’s granulomatosis Prednisone, methotrexate 574 No 13 F ⁄65 Type II acquired angio-oedema Co-trimoxazole, methotrexate 228 No 14 F ⁄51 Idiopathic thrombopenic purpura Dexamethasone 551 No 15 F ⁄74 Idiopathic thrombopenic purpura Prednisone, azathioprine 686 No 16 F ⁄82 Idiopathic thrombopenic purpura Prednisone 238 No 17 M ⁄69 Wegener’s granulomatosis Prednisone, cyclophosphamide 118 No 18 F ⁄30 Systemic lupus erythematosus Hydroxychloroquine, prednisone, mepacrine, 141 No azathioprine 19 F ⁄30 Systemic lupus erythematosus Prednisone, hydroxychloroquine, methotrexate 148 No 20 M ⁄69 Idiopathic thrombopenic purpura Prednisone 98 No 21 M ⁄68 Dermatomyositis Prednisone, azathioprine, aspirin 79 No 22 M ⁄34 Wegener’s granulomatosis Prednisone 85 No 23 F ⁄34 CREST syndrome and Methotrexate, prednisone, aspirin 60 No rheumatoid arthritis

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1272 Correspondence

The aforementioned study aimed retrospectively to evaluate The immunosuppressive therapy probably adds little toxicity the development of LON in other patients with different auto- when compared with chemotherapy. Although rituximab has immune diseases who had been treated with rituximab. We shown some efficacy in the treatment of PV, our case, reviewed the medical histories of consecutive patients who together with the one observed by Goh et al.,1 might suggest had completed treatment with rituximab in our Autoimmune that patients with PV are more likely to develop LON than Diseases Unit from February 2005 to November 2006 to iden- those with other autoimmune diseases. This small case series tify episodes of LON. Except for the patients with rheumatoid does not provide enough evidence to support this conclusion. arthritis who received two doses of 1 g 2 weeks apart, all the Nevertheless, it raises several questions that larger case series ) patients received weekly infusions of rituximab 375 mg m 2 and studies may answer. In the meantime, we think that clini- for 4 weeks. The episodes of neutropenia were classified as cians should be aware of the potential risk and the need to delayed onset when the neutropenia developed 30 days after monitor patients closely for this complication. the last dose of rituximab. We defined LON as neutropenia of 9 )1 £ 1Æ0 · 10 L according to the National Cancer Institute Autoimmune Diseases Unit, R. RIOS-FERNA´ NDEZ Common Toxicity Criteria. Generally, a full blood count with Department of Internal Medicine M.T. GUTIERREZ-SALMERO´ N * differential was carried out every 2–4 weeks during the and *Department of Dermatology, J.-L. CALLEJAS-RUBIO follow-up of the patients. Hospital Clı´nico San Cecilio, M. FERNA´ NDEZ-PUGNAIRE* In total, 23 patients were identified. The characteristics of Avenida Dr Olo´riz 16, N. ORTEGO-CENTENO these patients and primary treatment are as shown in Table 1. 18012 Granada, Spain Age ranged from 23 to 82 years. Twelve patients were men. E-mail: [email protected] Rituximab was prescribed in all patients because of refractory activity of their autoimmune diseases. Sixteen were treated References with rituximab in combination with immunosuppressants: azathioprine in five patients, methotrexate in six, cyclophos- 1 Goh MSY, McCormack C, Dinh HV et al. Rituximab in the adjuvant phamide in four and mycophenolate mofetil in one. All treatment of pemphigus vulgaris: a prospective open-label pilot patients had a normal polymorphonuclear cell count before study in five patients. Br J Dermatol 2007; 156:990–6. 2 Cattaneo C, Spedini P, Casari S et al. Delayed-onset peripheral blood rituximab treatment was started. With a median follow-up of cytopenia after rituximab: frequency and risk factor assessment in 327 days, no patient developed early neutropenia and only a consecutive series of 77 treatments. Leuk Lymphoma 2006; the patient with PV, who was also receiving azathioprine 47:1013–17. (150 mg daily), experienced severe delayed neutropenia. 3 Nitta E, Izutsu K, Sato T et al. A high incidence of late-onset Rituximab is a chimaeric murine-human monoclonal anti- neutropenia following rituximab-containing chemotherapy as a body directed against the CD20 antigen, which is increasingly primary treatment of CD20-positive B-cell lymphoma: a single- being used to treat severe autoimmune diseases.6 It is gen- institution study. Ann Oncol 2006; 18:364–9. 4 van Oers MH, Klasa R, Marcus RE et al. Rituximab maintenance erally well tolerated and most of its side-effects have been improves clinical outcome of relapsed ⁄resistant follicular non- described during the first infusion, typically including brief, Hodgkin lymphoma in patients both with and without rituximab 8,9 moderate fever and chills. Recently, several cases of LON during induction: results of a prospective randomized phase 3 have been described in relation to its use.10 Chaiwatanatorn intergroup trial. Blood 2006; 108:3295–301. et al.11 were the first to establish a possible link between LON 5 Forstpointner R, Unterhalt M, Dreyling M et al. Maintenance therapy and rituximab. LON develops 1–6 months after rituximab with rituximab leads to a significant prolongation of response dura- treatment and has usually been reported following rituximab- tion after salvage therapy with a combination of rituximab, fludara- bine, cyclophosphamide, and mitoxantrone (R-FCM) in patients based chemotherapy for haematological disorders such as with recurring and refractory follicular and mantle cell lymphomas: 3,12 B-lymphocytic malignancies. The natural history and inci- results of a prospective randomized study of the German Low Grade dence of LON have yet to be adequately defined. Its mecha- Lymphoma Study Group (GLSG). Blood 2006; 108:4003–8. nism is also unknown, although some researchers have 6 Gottenberg JE, Guillevin L, Lambotte O et al. Tolerance and short hypothesized an immune-mediated mechanism secondary to a term efficacy of rituximab in 43 patients with systemic autoim- transient production of autoantibodies,13 suppression of neu- mune diseases. Ann Rheum Dis 2005; 64:913–20. trophils by large granular lymphocytes,14 and an immune dys- 7 Larrar S, Guitton C, Willems M et al. Severe hematological side effects following rituximab therapy in children. Haematologica 2006; regulation during B-cell recovery as potential aetiologies.11,13 91(Suppl. 8):ECR36. We reduced the dose of azathioprine to 50 mg daily in our 8 Stasi R, Pagano A, Stipa E et al. Rituximab chimeric anti-CD20 patient after the episode of neutropenia. Although azathioprine monoclonal antibody treatment for adults with chronic idiopathic could have been implicated in the neutropenia, rituximab is a thrombocytopenic purpura. Blood 2001; 98:952–7. more likely cause, as there had not been previous episodes 9 Thachil J, Mukherje K, Woodcock B. Rituximab-induced haemor- prior to rituximab while the patient was on treatment with rhagic thrombocytopenia in a patient with hairy cell leukaemia. predisolone and azathioprine. Br J Haematol 2006; 135:273–4. 10 Lemieux B, Tartas S, Traulle C et al. Rituximab-related late-onset neu- Our findings suggest that LON seems to be less frequent tropenia after autologous stem cell transplantation for aggressive when rituximab is combined with less aggressive immunosup- non-Hodgkin’s lymphoma. Bone Marrow Transplant 2004; 33:921–3. pressive therapy in patients with autoimmune diseases.2,3,11

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1273

) 11 Chaiwatanatorn K, Lee N, Grigg A et al. Delayed-onset neutropenia lesterolaemia (267 mg dL 1; normal 100–200), hypertriglyce- ) associated with rituximab therapy. Br J Haematol 2003; 121:913–18. ridaemia (449 mg dL 1; normal 100–180) and weakness; we 12 Micallef IN, Kahl BS, Maurer MJ et al. A pilot study of epratuzumab decided to continue treatment and after 3 months of therapy and rituximab in combination with cyclophosphamide, doxorubi- the patient developed a severe weakness and a remarkable cin, vincristine, and prednisone chemotherapy in patients with pre- )1 viously untreated, diffuse large B-cell lymphoma. Cancer 2006; hypertriglyceridaemia (689 mg dL ) while total serum choles- )1 107:2826–32. terol levels remained unchanged (264 mg dL ). Because of 13 Voog E, Morschhauser F, Solal-Celigny P. Neutropenia in patients these important alterations we decided to discontinue therapy treated with rituximab. N Engl J Med 2003; 348:2691–4. with the biological drug and after 4 weeks the TG level was ) ) 14 Terrier B, Ittah M, Tourneur L et al. Late-onset neutropenia follow- 220 mg dL 1, cholesterol level was 224 mg dL 1 and weakness ing rituximab results from a hematopoietic lineage competition was notably improved. due to an excessive BAFF-induced B-cell recovery. Haematologica TNF-a is a multifunctional regulatory cytokine that is pro- 2007; 92(Suppl. 3):ECR10. duced by many inflammatory cells and adipocytes; it plays a Key words: autoimmune diseases, pemphigus vulgaris, rituximab pivotal role in the pathogenesis of severe inflammatory diseases Conflicts of interest: none declared. such as psoriasis, psoriatic arthritis and rheumatoid arthritis. Moreover, TNF-a acts on lipoprotein metabolism with a very complex mechanism, which increases both low-density lipo- protein cholesterol and high-density lipoprotein cholesterol (HDL-C) and also serum levels of TGs, supporting in this way Hypertriglyceridaemia during treatment with a proatherogenic lipid profile.2 Many studies have shown a rise adalimumab in psoriatic arthritis in cardiovascular risk in patients with inflammatory diseases in which there is a notable rise in plasma and articular level of DOI: 10.1111/j.1365-2133.2007.08188.x TNF-a.3 TNF-a blockers should induce an antiatherogenic modification of lipid profile, with decreases in TG concentra- SIR, Adalimumab, the first fully human antitumour necrosis fac- tion and increases in serum levels of total cholesterol and tor (TNF)-a antibody developed to date, has been shown to be HDL-C during treatment, as demonstrated in several studies.4 useful in the treatment of inflammatory diseases such as However, some authors observed a proatherogenic modifica- rheumatoid arthritis, Crohn disease and psoriatic arthritis, and it tion of lipid profile during treatment with anti-TNF-a anti- has also demonstrated a safety profile similar to other anti-TNF-a bodies.5,6 In our case we observed an increase of TGs after agents. There are currently no available data about triglyceride 8 weeks of treatment (following administration of the fourth (TG) alterations in patients during treatment with adalimumab. dose), peaking after 3 months of therapy. To explain a similar We report a 39-year-old man with a 1-year history of oligo- observation Dahlqvist et al.5 supposed that systemic administra- articular asymmetrical type psoriatic arthritis. Minimal plaque tion of TNF-a-neutralizing antibodies does not inhibit local psoriasis developed 2 months before the articular involvement; action of this cytokine, so that a TNF-a blockage could lead to the patient had no comorbidities, in particular, no features of a local upregulation of the TNF system in adipose tissue, metabolic syndrome. There was no family history of hyper- through increased synthesis of TNF in the adipocytes. We do triglyceridaemia or hypercholesterolaemia. Given the lack of not agree with this hypothesis as the excellent results observed efficacy of previous therapies with disease-modifying antirheu- with the treatment underline that TNF-a should be inactivated matic drugs, the patient was considered as ‘high need’ for bio- by antibodies also at a local level. We suppose that TNF-a is logic therapy. Baseline Psoriasis Area and Severity Index (PASI) only one of a very complex network of proinflammatory mole- was 8 and Ritchie index1 was 6. Before starting treatment, com- cules that play a role in the lipid profile and that probably, plete laboratory tests were performed, including chest X-ray, once blocked, it promotes the upregulation of other cytokines full blood count, urea, creatinine, transaminases, cholesterol, that lead to an increase of synthesis of TGs; probably the pri- TGs, antinuclear antibodies, antidouble-stranded DNA anti- mer of this ‘side’ mechanism is genetically regulated. This is bodies, antiextractable nuclear antigens, antiphospholipid anti- the first case of a remarkable increase in plasma levels of TGs bodies (lupus anticoagulant and anticardiolipin), tumour during treatment with adalimumab and, although it is only a markers, blood culture and urine culture; no significant altera- single report, it could suggest the need for a very strict moni- tions were found. A Mantoux test was negative. Anticitrullin toring of the lipid profile during anti-TNF-a treatment. antibody and rheumatoid factor were normal while C-reactive )1 protein was 13 mg L . Hence we started systemic therapy with Institute of Dermatology, G. STINCO adalimumab 40 mg every 2 weeks administered subcutaneously Department of Clinical and F. PICCIRILLO as monotherapy. Blood tests were repeated every 4 weeks. After Experimental Pathology and Medicine, P. PATRONE the first 4 weeks of treatment an improvement of skin condition University of Udine School of Medicine, and articular performance (PASI 4 and Ritchie index 2) was Ospedale San Michele, Piazza Rodolone 1, observed, without any side-effects. After 8 weeks of treatment 33013 Gemona del Friuli, Udine, Italy (PASI 3.2 and Ritchie index 2) the patient developed hypercho- E-mail: [email protected]

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1274 Correspondence

References (a)

1 Ritchie DM, Boyle JA, McInnes JM et al. Clinical studies with an articular index for the assessment of joint tenderness in patients with rheumatoid arthritis. Q J Med 1968; 37:393–406. 2 Khovidhunkit W, Memon RA, Feingold KR, Grunfeld C. Infection and inflammation-induced proatherogenic changes of lipoproteins. J Infect Dis 2000; 18:S462–72. 3 Peters MJ, van der Horst-Bruinsma IE, Dijkmans BA, Nurmohamed MT. Cardiovascular risk profile of patients with spondylarthro- pathies, particularly ankylosing spondylitis and psoriatic arthritis. Semin Arthritis Rheum 2004; 34:585–92. 4 Spanakis E, Sidiropoulos P, Papadakis J et al. Modest but sustained increase of serum high density lipoprotein cholesterol levels in patients with inflammatory arthritides treated with infliximab. J Rheumatol 2006; 33:2440–6. 5 Dahlqvist SR, Engstrand S, Berglin E, Johnson O. Conversion towards an atherogenic lipid profile in rheumatoid arthritis patients (b) during long-term infliximab therapy. Scand J Rheumatol 2006; 35:107– 11. 6 Cauza E, Cauza K, Hanusch-Enserer U et al. Intravenous anti TNF- alpha antibody therapy leads to elevated triglyceride and reduced HDL-cholesterol levels in patients with rheumatoid and psoriatic arthritis. Wien Klin Wochenschr 2002; 114:1004–7.

Key words: adalimumab, antitumour necrosis factor-a, hypertriglyceridaemia, psoriatic arthritis

Conflicts of interest: none declared.

Fig 1. Lower leg lesion (a) at presentation and (b) after 5 months of treatment with adalimumab 40 mg bimonthly.

Adalimumab for treatment of pyoderma gangrenosum discontinued when the patient developed a systemic reaction at the second dose. Additionally, a previous regimen of ciclo- DOI: 10.1111/j.1365-2133.2007.08212.x sporin had been ineffective. Therapy with subcutaneous adalimumab 40 mg bimonthly SIR, A 61-year-old woman was referred to the Center for Skin was initiated, along with oral prednisone 20 mg daily for and Related Musculoskeletal Diseases, Brigham and Women’s 4 weeks and intermittent use of intralesional triamcinolone ) Hospital (Boston, MA, U.S.A.), with extremely painful ulcerat- 5mgmL 1. Etanercept and mycophenolate mofetil were ing lesions on both pretibial areas (Fig. 1a). She had a 20-year discontinued. history of inflammatory bowel disease (IBD) and associated After three injections of adalimumab, most of the ulcers inflammatory arthritis, and at least a 5-year history of pyo- were largely healed, and the pain had improved substan- derma gangrenosum (PG). tially. Adalimumab treatment was continued at 40 mg For management of her IBD, arthritis and PG, the patient bimonthly, and the prednisone dose was reduced to 15 mg was taking subcutaneous etanercept 100 mg weekly, oral daily for 1 week and then to 10 mg daily for the subse- mycophenolate mofetil 1Æ5 g daily, and intermittent courses quent week. of antibiotics and of prednisone. She had received bioengi- Two months after initiation of adalimumab therapy, the neered skin grafts, which she dressed with 0Æ01% becaplermin lesions continued to heal, the erythema surrounding the ulcers (platelet-derived growth factor topical gel). was substantially reduced, and there were no new lesions. The Since the patient began taking etanercept, her arthritis and prednisone dose was tapered over a 10-week period and then IBD were quiescent. However, the PG lesions progressed in discontinued. number and severity. The wound healing regimen aided the Five months after beginning therapy with adalimumab, the healing of existing ulcers, but did not prevent the develop- patient was completely healed on both lower extremities ment of new flares. In the past she had been treated with (Fig. 1b), and her IBD remained quiescent. She continued to ) infliximab at up to 10 mg kg 1 for her arthritis and PG, with receive adalimumab 40 mg bimonthly and had not had any rapid improvement of the PG lesions, but infliximab use was further flares when she was last seen in our clinic, 18 months

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1275 after beginning treatment. Her legs were clear except for crib- University of Pittsburgh School of Medicine, R.G. POMERANTZ riform scars from old lesions. Pittsburgh, PA, U.S.A. M.E. HUSNI* Current treatments for PG include topical, intralesional and *Department of Rheumatic and Immunologic Diseases, E. MODY systemic corticosteroids, nonsteroidal immunomodulators, Cleveland Clinic Foundation, Cleveland, OH, U.S.A. A.A. QURESHI local wound care regimens, and biologic agents such as the Departments of Rheumatology and tumour necrosis factor (TNF)-a antagonists. Our patient had Dermatology, Brigham and Women’s Hospital, failed treatment with ciclosporin, mycophenolate mofetil and Boston, MA 02115, U.S.A. etanercept, and could not take infliximab due to an adverse Correspondence: Abrar A. Qureshi. reaction. An alternative steroid-sparing agent was needed to E-mail: [email protected] treat this refractory case of PG. Adalimumab is a human monoclonal antibody against References TNF-a, an important proinflammatory cytokine involved in adhesion molecule expression, keratinocyte proliferation, and 1 Trent JT, Kerdel FA. Tumor necrosis factor alpha inhibitors for the the recruitment of leucocytes to sites of inflammation. Because treatment of dermatologic diseases. Dermatol Nurs 2005; 17:97– 107. of its role as a mediator of inflammatory processes and its over- 2 Brooklyn TN, Dunnill MG, Shetty A et al. Infliximab for the treat- 1 a production in many chronic inflammatory diseases, TNF- has ment of pyoderma gangrenosum: a randomised, double blind, been implicated as a therapeutic target for the management of placebo controlled trial. Gut 2006; 55:505–9. a variety of dermatological and systemic autoimmune diseases. 3 Roy DB, Conte ET, Cohen DJ. The treatment of pyoderma gangre- Along with infliximab and etanercept, adalimumab is one of nosum using etanercept. J Am Acad Dermatol 2006; 54 (3 Suppl. three currently available TNF-a antagonists. 2):S128–34. The significant role of TNF-a in PG pathophysiology is 4 Hubbard VG, Friedmann AC, Goldsmith P. Systemic pyoderma gangrenosum responding to infliximab and adalimumab. Br J demonstrated by the success of anti-TNF-a therapy with 2,3 Dermatol 2005; 152:1059–61. infliximab and etanercept in treating PG. Additionally, a few 5 Fonder MA, Cummins DL, Ehst BD et al. Adalimumab therapy for individual cases of PG effectively treated with adalimumab, recalcitrant pyoderma gangrenosum. J Burns Wounds 2006; 5:e8. following the failure of multiple other immunomodulating 6 Heffernan MP, Anadkat MJ, Smith DI. Adalimumab treatment for treatment regimens, have been described.4–6 pyoderma gangrenosum. Arch Dermatol 2007; 143:306–8. Adalimumab is currently indicated for the treatment of 7 Khanna D, McMahon M, Furst DE. Safety of tumour necrosis rheumatoid arthritis and psoriatic arthritis. However, it may factor-alpha antagonists. Drug Saf 2004; 27:307–24. 8 Scheinfeld N. A comprehensive review and evaluation of the side also be a useful off-label therapy for a number of other dis- effects of the tumor necrosis factor alpha blockers etanercept, eases, including PG. In contrast to the fusion protein etaner- infliximab and adalimumab. J Dermatolog Treat 2004; 15:280–94. cept, the monoclonal antibodies adalimumab and infliximab 9 Desai SB, Furst DE. Problems encountered during anti-tumour necro- are indicated for IBD, a disease that simultaneously affects a sis factor therapy. Best Pract Res Clin Rheumatol 2006; 20:757–90. substantial proportion of patients with PG. Hence, the mono- 10 Youdim A, Vasiliauskas EA, Targan SR et al. A pilot study of adal- clonal antibodies may be reasonable monotherapy options in imumab in infliximab-allergic patients. Inflamm Bowel Dis 2004; selected patients with PG and IBD. 10:333–8. Anti-TNF-a therapy has been associated with risks of infec- Key words: adalimumab, pyoderma gangrenosum tion and malignancy, and thus careful selection and screening Conflicts of interest: E.M. is a speaker for Amgen, Abbott, Genen- of patients as well as close follow-up is recommended. The tech and Centocor; A.A.Q. is a consultant and speaker for Abbott, most widely reported serious complication of anti-TNF-a ther- Amgen and Genentech, and has a limited consulting relationship with 7 apy is reactivation of latent tuberculosis. Use of anti-TNF-a Centocor. agents is also associated with an increased risk of other bacte- rial infections,8 the development of which may require inter- ruption or cessation of therapy. There is an association between use of anti-TNF-a agents and increased incidence of Effects of etanercept therapy on fatigue and lymphoma, but it is unclear whether the increased risk is symptoms of depression in subjects treated for due to these medications or to the pathology underlying moderate to severe plaque psoriasis for up to their use.9 96 weeks Adalimumab may be an alternative for use in severe PG that has failed conventional therapy or in patients with adverse DOI: 10.1111/j.1365-2133.2007.08205.x reactions or lack of response to infliximab. Adalimumab was well tolerated by eight patients who had previously discontin- SIR, Fatigue and depression have been associated with psoria- ued infliximab use due to allergic reactions or intolerance.10 sis.1,2 Proinflammatory cytokines such as tumour necrosis fac- Furthermore, subcutaneous adalimumab offers the conve- tor (TNF), implicated in the pathogenesis of psoriasis, also nience of self-administration. Clinical trials are needed to eval- have been linked to symptoms of fatigue3 and depression.4 uate the efficacy and safety of adalimumab as a treatment Etanercept is a soluble TNF receptor-Fc fusion protein option for PG. that has U.S. Food and Drug Administration approval for the

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1276 Correspondence treatment of moderate to severe plaque psoriasis. Results have Double Open label been reported previously of a 12-week, randomized, double- (a) blind blind trial of etanercept or placebo in psoriasis.5 Herein, we 6 5·5 report the longer-term results from this study. After week 12, 5 591 subjects (95% of enrolled subjects) entered a second por- 4·5 tion of the study in which all received open-label etanercept 4 50 mg twice weekly. Of these, 233 subjects (77%) from the 3·5 3 etanercept ⁄etanercept group and 231 subjects (81%) from the 2·5 Placebo/etanercept placebo ⁄etanercept group completed 84 weeks of open-label 2 Etanercept/etanercept etanercept. 1·5 1 Fatigue was assessed using the Functional Assessment of Mean change in FACIT-F score 0·5 6 Chronic Illness Therapy–Fatigue (FACIT-F) questionnaire; sub- 0 Baseline 12 24 36 48 60 72 84 96 jects whose score improved at least 3 units from baseline were Week classified as FACIT-F responders.7 Symptoms of depression were Double Open label assessed using the self-administered Beck Depression Inventory (b) blind (BDI)8 and the Hamilton rating scale for depression (Ham-D),9 5·5 administered by trained interviewers. As in trials of antidepres- 5 4·5 sants,10 subjects were designated as Ham-D and BDI responders 4 if an improvement of ‡ 50% from baseline was observed. 3·5 Per the protocol eligibility criteria, patients were excluded 3 from the study if they had evidence of psychiatric disease or 2·5 other comorbidities that might interfere with their participa- 2 1·5 Placebo/etanercept tion. As previously reported, at baseline, the majority of sub- Etanercept/etanercept Mean change in BDI score 1 jects had no or minimal symptoms of depression as measured 0·5 by the Ham-D and BDI, respectively, thus limiting our ability 0 Baseline 12 24 36 48 60 72 84 96 to demonstrate a reduction in symptoms of depression in Week patients with psoriasis. Fatigue, however, was more pro- nounced at baseline than in the general population.5 Fig 1. Mean changes in (a) Functional Assessment of Chronic In previously reported results of the 12-week double-blind Illness Therapy–Fatigue (FACIT-F) score and (b) Beck Depression period, subjects who received etanercept had significantly Inventory (BDI) score from baseline over time. Data represent mean greater improvements in FACIT-F scores and there were more improvements over time in subjects receiving etanercept during responders, compared with placebo-treated subjects.5 the double-blind and open-label portions of the study (etanercept ⁄ During the open-label period, after 12 weeks of active ther- etanercept subjects; boxes) and subjects receiving placebo during the apy (week 24), the placebo ⁄etanercept group achieved a mean double-blind portion and etanercept during the open-label portion of improvement of 5Æ0 units from baseline, and more than half the trial (placebo ⁄etanercept subjects; diamonds). Error bars represent of subjects in both groups were classified as FACIT-F respond- SEM. The last-observation-carried-forward method was used to impute ers (58% etanercept ⁄etanercept group vs. 52% placebo ⁄etaner- missing data. Analyses included subjects who received at least one dose of study drug. Treatment group comparisons were made using a cept group). Furthermore, mean FACIT-F improvements from two-sided van Elteren’s test stratified by prior psoriasis therapy as the baseline (5Æ3 units vs. 5Æ4 units, respectively; Fig. 1a) and primary analysis. improvements in responder rates (59% vs. 55%, respectively) were sustained up to week 96. were sustained in both groups up to week 96 (4Æ1 units Mean BDI scores also were significantly improved with vs. 4Æ4 units, respectively; Fig. 1b). Also sustained were etanercept at week 12 of the double-blind period (3Æ9 units) improvements in BDI responder rates (57% for both groups) compared with the placebo ⁄etanercept groups (2Æ1 units; and improvements in the percentage of subjects with mini- P =0Æ0001). Additionally, the proportion of BDI responders mal symptoms of depression (84% vs. 86%, respectively). in the etanercept ⁄etanercept group was significantly greater In another measurement of depression, mean Ham-D scores (55% vs. 39%, respectively; P <0Æ0001), as was the propor- were significantly improved with etanercept during the double- tion of subjects with minimal symptoms of depression (84% blind period, compared with placebo (1Æ5 units etanercept ⁄eta- vs. 75%, respectively; P =0Æ0129). nercept group vs. 0Æ4 unit placebo ⁄etanercept group; P = At week 24 of the open-label period, mean BDI improve- 0Æ0012 at week 12). Also, the proportion of Ham-D responders ments were comparable (4Æ0 units etanercept ⁄etanercept was significantly greater (43% vs. 32%, respectively; P = group vs. 4Æ5 units placebo ⁄etanercept group). The percent- 0Æ0048), and the proportion of subjects with no symptoms of age of BDI responders in each group increased to similar lev- depression was greater (84% vs. 77%, respectively; P =0Æ0415). els (58% vs. 55%, respectively), and similar proportions of At week 24 of the open-label period, the placebo ⁄etaner- subjects in both groups had minimal symptoms of depres- cept group had a mean improvement of 1Æ7 units from base- sion (85% vs. 88%, respectively). Mean BDI improvements line in the Ham-D, similar to the improvement achieved by

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1277 subjects in the etanercept group in the double-blind period. References The percentage of Ham-D responders in each group also in- 1 Evers AWM, Lu Y, Duller P et al. Common burden of chronic creased to similar levels (45% etanercept ⁄etanercept group skin diseases? Contributors to psychological distress in adults with vs. 46% placebo ⁄etanercept group), and similar proportions psoriasis and atopic dermatitis. Br J Dermatol 2005; 152:1275– of subjects in both groups had no symptoms of depression 81. (86% vs. 87%, respectively). Consistent with the observations 2 Devrimci-Ozguven H, Kundakci N, Kumbasar H, Boyvat A. in the FACIT-F and BDI measures, mean Ham-D improve- The depression, anxiety, life satisfaction and affective expression ments from baseline were sustained in both groups up to levels in psoriasis patients. J Eur Acad Dermatol Venereol 2000; 14:267– week 96 (1Æ9 units vs. 2Æ1 units, respectively). Also sustained 71. 3 Heesen C, Nawrath L, Reich C et al. Fatigue in multiple sclerosis: were improvements in Ham-D responder rates (51% vs. an example of cytokine mediated sickness behaviour? J Neurol Neuro- 48%, respectively) and improvements in the percentage of surg Psychiatry 2006; 77:34–9. subjects with no symptoms of depression as measured by 4 Illman J, Corringham R, Robinson D Jr et al. Are inflammatory the Ham-D (87% vs. 88%, respectively). cytokines the common link between cancer-associated cachexia and Although the limitations of this study have been acknowl- depression? J Support Oncol 2005; 3:37–50. edged, the authors believe the long-term results reported 5 Tyring S, Gottlieb A, Papp K et al. Etanercept and clinical outcomes, herein underscore the value of assessing fatigue and symptoms fatigue, and depression in psoriasis: double-blind placebo- controlled randomised phase III trial. Lancet 2006; 367:29–35. of depression in patients with psoriasis. In this phase III clini- 6 Yellen SB, Cella DF, Webster K et al. Measuring fatigue and other cal trial, improvements in FACIT-F, BDI and Ham-D explor- anemia-related symptoms with the Functional Assessment of Can- atory endpoints were sustained for up to 96 weeks. cer Therapy (FACT) measurement system. J Pain Symptom Manage 1997; 13:63–74. 7 Cella D, Eton DT, Lai JS et al. Combining anchor and distribution- Acknowledgments based methods to derive minimal clinically important differences Immunex, Stephen Tyring, and other members of the Etaner- on the Functional Assessment of Cancer Therapy (FACT) anemia and fatigue scales. J Pain Symptom Manage 2002; 24:547–61. cept Psoriasis Study Group designed the study. Research was 8 Beck A, Steer R. Beck Depression Inventory Manual. San Antonio: Harcourt funded by Immunex Corporation, a wholly owned subsidiary Brace, 1993. of Amgen Inc., and by Wyeth Pharmaceuticals. Investigators 9 Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry in the study group collected the data, which were analysed by 1960; 23:56–62. Meleana Dunn of Amgen Inc. The academic investigators had 10 Iosifescu DV, Nierenberg AA, Mischoulon D et al. An open study of full access to the data; the complete data set was held at the triiodothyronine augmentation of selective serotonin reuptake central data-processing facility at Amgen. The authors made inhibitors in treatment-resistant major depressive disorder. J Clin Psychiatry 2005; 66:1038–42. decisions about submission of this paper for publication in collaboration with Amgen. Financial support for the prepara- Key words: Beck Depression Inventory, depression, fatigue, Functional Assessment of tion of the manuscript was provided by Amgen Inc. The Chronic Illness Therapy–Fatigue scale, Hamilton rating scale for depression authors thank Edward Mancini for providing medical writing Conflicts of interest: see Acknowledgments. assistance on behalf of Amgen Inc. Dr Krishnan received fund- ing from Amgen, Inc. for his part in the design, conduct and Trial registration: This study is registered with ClinicalTrials.gov with analysis of this study. He has served as a consultant in the past the identifier NCT00111449. to Wyeth, Abbott, Johnson and Johnson, BristolMyersSquibb, Roche and Pfizer.

Department of Psychiatry and Behavioral Sciences, R. KRISHNAN Box 3950, Duke University Medical Center, D. CELLA* Cutaneous Langerhans cell histiocytosis in Durham, NC 27710, U.S.A. C. LEONARDI an elderly man successfully treated with *Center on Outcomes, Research and Education, K. PAPP narrowband ultraviolet B Evanston Northwestern Healthcare, A.B. GOTTLIEB§ Evanston, IL 60201, U.S.A. M. DUNN– DOI: 10.1111/j.1365-2133.2007.08204.x Department of Dermatology, C.F. CHIOU– St Louis University School of Medicine, V. PATEL– SIR, Langerhans cell histiocytosis (LCH) is characterized by St Louis, MO 63117-1206, U.S.A. A. JAHREIS– infiltration and proliferation of pathogenic Langerhans cells in Probity Medical Research, various organs. LCH can occur in patients of any age, but it Waterloo, Ontario N2J 1C4, Canada has two peak ages of occurrence. The first peak occurs at a §Department of Dermatology, mean age of 18 months, and the second occurs between the Tufts–New England Medical Center, ages of 30 and 50 years.1,2 LCH in adults, especially in sub- Boston, MA 02111-1533, U.S.A. jects aged over 70 years, is very rare. –Amgen Inc., Thousand Oaks, CA 91320-1799, U.S.A. Narrowband ultraviolet (UV) B is a newly developed E-mail: [email protected] therapeutic modality that utilizes a very narrow wavelength

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1278 Correspondence band of UVB (peak at 311 ± 2 nm). Narrowband UVB has Upon examination, slightly scaly erythematous papules of the advantage over conventional broadband UVB therapy 5–10 mm in diameter were scattered on his abdomen and that it has a reduced risk of burning and contains fewer extremities (Fig. 1). No eruptions were found on the scalp or carcinogenic wavelengths. It has been shown to be effective in the groin. in early-stage mycosis fungoides3 and other inflammatory Skin biopsy showed a nest of aggregated large epithelioid skin diseases. We report the first case of successful remis- cells both in the upper dermis and in the epidermis (Fig. 2a), sion of cutaneous LCH in an elderly man by narrowband with parakeratosis in the overlying stratum corneum. Immuno- UVB irradiation. histochemically, the cells stained positive with anti-S100 A 72-year-old Japanese man was referred to our outpatient (Fig. 2b) and anti-CD1a (Fig. 2c) antibodies but negative with clinic with a 5-month history of an itchy eruption on his anti-HMB-45, anti-CD45RO and anti-CD54 antibodies. A diag- thigh. He had visited a physician who prescribed an oral anti- nosis of LCH was established. Blood tests showed mild leuco- ) histamine and topical corticosteroid, but no acceptable remis- cytosis (9Æ57 · 109 L 1) and mild anaemia (haemoglobin ) sion was achieved. The patient had worked in an office and 11Æ8gdL 1), and his serum chemistry, including protein frac- had no history of smoking. None of his family members had tions, was unremarkable. Computed tomography excluded experienced similar skin symptoms. hepatosplenomegaly and lymph node swelling. Skeletal X-ray revealed no defect. These data suggested that the patient had a single-system disease. A series of narrowband UVB irradiations was employed for the treatment. Total-body narrowband UVB irradiation (Waldmann K5001, Villingen-Schwenningen, Germany) was ) initiated at 0Æ4Jcm 2, and the dose was gradually increased ) ) to 1Æ0Jcm 2. Eleven sessions of irradiation (9Æ3Jcm 2 total) brought almost complete clearance of the papules as well as the itch. To ensure remission of the papules, a biopsy was taken from a healed pigmented macule adjacent to the original biopsy site. The numbers of CD1a+ and S100+ cells were dramatically reduced both in the upper dermis and in the epidermis (data not shown). No recur- rence has been observed 12 months after the termination of UVB irradiation. The pathogenesis of LCH is still unknown. An analysis by human androgen receptor assay revealed clonality of LCH cells 4 Fig 1. Clinical appearance before treatment. Slightly desquamated in at least some cases; however, this observation does not papules were scattered on the thigh. support the notion that the disease is malignant. LCH affects

(a) (b) (c)

Fig 2. Photomicrographs of the biopsy specimen taken before ultraviolet B irradiation. (a) Nests of tumour cells with large pale cytoplasm are seen in the upper dermis and epidermis (haematoxylin and eosin). (b) The cells were S100 positive. (c) The cells were CD1a positive. Original magnification · 200.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1279 various organs such as lung, bone, pituitary gland and skin, 6 Steen AE, Steen KH, Bauer R, Bieber T. Successful treatment of and 30–69% of patients have multisystem involvement.1,2 The cutaneous Langerhans cell histiocytosis with low-dose metho- prognosis of LCH depends on the age of the patient and the trexate. Br J Dermatol 2001; 145:137–40. 7 Lichtenwald DJ, Jakubovic HR, Rosenthal D. Primary cutaneous organs involved. A single-system disease limited to skin repre- 1 Langerhans cell histiocytosis in an adult. Arch Dermatol 1991; sents only 4–7% of cases of LCH. LCH limited to skin usually 127:1545–8. has a good prognosis, but the management of LCH is difficult 8 Sakai H, Ibe M, Takahashi H et al. Satisfactory remission achieved 5 because its response to therapy is inconsistent. Various thera- by PUVA therapy in Langerhans cell hisiocytosis in an elderly pies have been applied in the treatment of cutaneous LCH in patient. J Dermatol 1996; 23:42–6. adults. These include oral steroids, methotrexate,6 total-body 9 Iwatsuki K, Tsugiki M, Yoshizawa N et al. The effect of photothera- electron beam irradiation together with vinblastine administra- pies on cutaneous lesions of histiocytosis X in the elderly. Cancer 1986; 57:1931–6. tion,7 topical imiquimod,5 and phototherapy.8–10 10 Neumann C, Kolde G, Bonsmann G. Histiocytosis X in an elderly Six patients with adult cutaneous LCH have been treated by patient. Ultrastructure and immunocytochemistry after PUVA psoralen–UVA (PUVA) therapy and described in the literature. photochemotherapy. Br J Dermatol 1988; 119:385–91. One patient had a remission without recurrence, four had Key words: elderly, Langerhans cell histiocytosis, narrowband ultraviolet B recurrence with response to repeated irradiation,8–10 and one had only a partial response.7 Narrowband UVB has not been Conflicts of interest: none declared. used previously for the treatment of cutaneous LCH; it suc- cessfully provided remission in our patient. As shown in mycosis fungoides, narrowband UVB has antitumour efficacy 3 comparable with PUVA. As UV exposure of skin results in a Pyoderma gangrenosum and interleukin 8 profound depletion of Langerhans cells, the pathogenic LCH cells may also be sensitive to the UV radiation. With easier DOI: 10.1111/j.1365-2133.2007.08202.x and safe administration, narrowband UVB can be considered as the first choice of skin-directed therapy for LCH. Narrow- SIR, Interleukin (IL)-8 is a potent chemotactic polypeptide for band UVB may also be used as a part of combination therapy neutrophils.1 We have previously demonstrated that high levels with oral or topical steroids or topical imiquimod. of IL-8 are immunohistochemically detectable in dermal fibro- As far as we know, this is the first reported case of remis- blasts from ulcers of patients with pyoderma gangrenosum.2 sion of adult cutaneous LCH induced by narrowband UVB. Furthermore, overexpression of IL-8 using an adenovirus vector The patient has been disease free for 12 months, but follow- in the skin of severe combined immunodeficiency mice has up is necessary because of the possible recurrence. been shown to result in the development of skin ulcers resem- bling pyoderma gangrenosum.2 To clarify the level and mode *Division of Dermatology, S. IMAFUKU* (constitutive or transient) of IL-8 production in this condition, Kitakyushu Medical Center, Kitakyushu, Japan S. SHIBATA* fibroblasts from a patient with pyoderma gangrenosum were Department of Dermatology, A. TASHIRO* isolated and cultured, and IL-8 levels were measured in the School of Medicine, Fukuoka University, M. FURUE supernatants. Serum IL-8 levels were also measured. 45-1 Nanakuma, Fukuoka City 814-0180, Japan An 81-year-old woman presented with a 4-year history of Department of Dermatology, diarrhoea. A painful skin ulcer had developed in her right Graduate School of Medical Sciences, breast 2 months prior to examination and had enlarged rap- Kyushu University, Fukuoka, Japan idly. In addition, loss of appetite and appearance of bloody E-mail: [email protected] stools began 1 month prior to examination. Ulcerative colitis was diagnosed in the Department of Internal Medicine at Hig- References ashi Kobe Hospital and she was admitted. She had not received any pharmacotherapies prior to admission and was 1 Arico M, Girschikofsky M, Genereau T et al. Langerhans cell histio- treated using mesalazine alone (2250 mg daily for the first cytosis in adults: report from the International Registry of the Histiocyte Society. Eur J Cancer 2003; 39:2341–8. 3 weeks and 1500 mg daily thereafter). The patient was 2 Howarth DM, Gilchrist GS, Mullan BP et al. Langerhans cell histio- referred to the Department of Dermatology in the same hos- cytosis: diagnosis, natural history, management, and outcome. pital for diagnosis of the skin ulcer. The ulcer measured Cancer 1999; 85:2278–90. 12 · 10 cm with an erythematous, raised edge on the right 3 Diederen PV, van Weelden H, Sanders CJ et al. Narrowband UVB breast (Fig. 1a). The ulcer base was irregular and displayed a and psoralen–UVA in the treatment of early-stage mycosis fungo- granular appearance. After obtaining informed consent, skin ides: a retrospective study. J Am Acad Dermatol 2003; 48:215–19. biopsies were taken from both the edge of the ulcer and un- 4 Yu RC, Chu C, Buluwela L, Chu AC. Clonal proliferation of Langer- hans cells in Langerhans cell histiocytosis. Lancet 1994; 343:767–8. involved skin of the left upper arm for the purposes of 5 Taverna JA, Stefanato CM, Wax FD, Demierre MF. Adult cutaneous diagnosis and cultivation of fibroblasts. Biopsy of the ulcer Langerhans cell histiocytosis responsive to topical imiquimod. JAm showed nonspecific changes with a dermal infiltrate compris- Acad Dermatol 2006; 54:911–13. ing neutrophils, and to a lesser extent, histiocytes and

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1280 Correspondence

Table 2 Serum levels of interleukin (IL)-8 after the start of mesalazine (a) therapy

Time after start of therapy (days) 15 21 40 60 ) IL-8 (pg mL 1)a 224 188 UD UD

aIL-8 protein content in the serum was measured using an enzyme-linked immunosorbent assay kit (Biosource, Camarillo, CA, U.S.A.) according to the manufacturer’s instruction. UD, under the level of detection.

Fibroblasts from the pyoderma gangrenosum ulcer and uninvolved skin were isolated and cultured as described pre- viously.3 IL-8 levels in supernatants of cultured fibroblasts (b) were measured by enzyme-linked immunosorbent assay, as described previously,2 at several time points after almost com- plete elimination of contamination by other types of cells. Fibroblasts from the ulcer produced high levels of IL-8 in the early days of primary culture (Table 1). However, IL-8 pro- duction decreased rapidly and was undetectable by 27 days after the start of primary culture. IL-8 levels in the super- natants of fibroblasts from uninvolved skin were below the level of detection at all time points up to 41 days after the start of primary culture. Serum IL-8 concentrations were also measured in the patient at several time points after initiating mesalazine therapy (Table 2). Serum IL-8 levels were elevated in the early days of treatment, but fell below detectable levels by 40 days after the start of treatment. One of the characteristics of pyoderma gangrenosum is Fig 1. (a) Clinical appearance of pyoderma gangrenosum ulcer. 4 (b) Histopathological findings. Nonspecific dermal infiltration of massive infiltration of neutrophils in the dermis. Certain fac- neutrophils is seen, along with histiocytes and mononuclear tors that induce neutrophil accumulation may be mediators of lymphocytic cells to a lesser extent. Neutrophils are also observed in this effect. IL-8 could represent one such factor and our previ- 2 blood vessels. ous report suggested that IL-8 produced by dermal fibroblasts plays a part in the pathogenesis of this disease. The present study showed that fibroblasts from the ulcer of pyoderma gan- mononuclear lymphocytic cells (Fig. 1b). Neutrophils were grenosum produced high levels of IL-8 in vitro, whereas fibro- also identified in blood vessels. Extravasation of erythrocytes blasts from uninvolved skin did not. Production of IL-8 by was present, but no evidence of vasculitis was apparent. fibroblasts from the skin lesion decreased rapidly as cultivation Biopsy of uninvolved skin showed normal skin with no infil- proceeded, indicating that production of the chemokine by tration of neutrophils (data not shown). Clinical and histo- fibroblasts is not constitutive. In addition, IL-8 production pathological findings led to a diagnosis of pyoderma from uninvolved skin was undetectable, showing that IL-8 is gangrenosum and the skin lesion was treated using topical sul- produced locally in fibroblasts in the skin lesion. IL-8 produc- fadiazine ointment. Mesalazine therapy led to rapid improve- tion is not constitutive in normal fibroblasts in vitro, similar to ments in both ulcerative colitis and pyoderma gangrenosum other types of normal cells,1 and is induced following stimula- and the skin ulcer was almost eradicated leaving scar tissue tion by proinflammatory cytokines such as IL-1 or tumour after 6 weeks of mesalazine therapy. necrosis factor a.5 Increased production of IL-8 in fibroblasts

Table 1 Interleukin (IL)-8 levels in cultured fibroblasts from a biopsy specimen from the edge of the ulcer

Time after primary culture (days) 13 16 21 25 34 41 IL-8 (pg per 1 · 106 cells in 24 h) in the supernatanta 47 35 2Æ7UDUDUD

aIL-8 protein content in the culture medium was measured using an enzyme-linked immunosorbent assay kit (Biosource, Camarillo, CA, U.S.A.) according to the manufacturer’s instruction. UD, under the level of detection.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1281 from pyoderma gangrenosum may be induced by local stimu- 10 Volin MV, Campbell PL, Connors MA et al. The effect of sulfasal- lation with such proinflammatory cytokines. azine on rheumatoid arthritic synovial tissue chemokine produc- Concentrations of IL-8 were elevated in serum from the tion. Exp Mol Pathol 2002; 73:84–92. patient in the initial days of treatment, returning to undetectable Key words: fibroblasts, interleukin 8, mesalazine, pyoderma gangrenosum, ulcerative levels after the skin lesion had healed. Saito et al.6 also reported a colitis case of pyoderma gangrenosum with increased levels of serum Conflicts of interest: none declared. IL-8. At present, pyoderma gangrenosum is a diagnosis of exclusion7 and there is much room for diagnostic error.4,7 Serum IL-8 levels have also been shown to be high in some patients with ulcerative colitis.8 Serum IL-8 levels in the present patient may thus reflect IL-8 from both pyoderma gangrenosum Recurrent KIND1 (C20orf42) gene mutation, and ulcerative colitis lesions. Measuring serum IL-8 levels may c.676insC, in a Brazilian pedigree with Kindler be helpful for diagnosis if the levels can be shown to be high in syndrome the serum of patients with pyoderma gangrenosum. One of the mechanisms of action for mesalazine is down- DOI: 10.1111/j.1365-2133.2007.08219.x regulation of IL-8 production from cells.9,10 This fact and our previous2 and present studies strongly suggest the involvement SIR, Kindler syndrome (KS; OMIM 173650) is a rare autoso- of IL-8 in the pathogenesis of pyoderma gangrenosum. Elucida- mal recessive skin and mucous membrane disorder associated tion of the mechanisms underlying high production of IL-8 in with trauma-induced blisters and erosions, poikiloderma and the fibroblasts of patients with pyoderma gangrenosum will be 1 variable degrees of photosensitivity. In 2003, the KS gene helpful in understanding the pathophysiology of this disease. KIND1, also known as C20orf42 was identified for this geno- dermatosis through genetic linkage and candidate gene analy- Division of Dermatology, M. OKA 2,3 sis. Loss-of-function mutations were found in KIND1 Department of Clinical Molecular Medicine, encoding a 677 amino acid protein called kindlin-1 which is Kobe University Graduate School of Medicine, involved in linking the actin cytoskeleton to focal contacts, Kobe 650-0017, Japan 2–5 predominantly in basal keratinocytes. Previously, our group E-mail: [email protected] reported the mutation c.676insC in six Pakistani families, and this frameshift was shown to have occurred on the same allelic References background in five of the six families described. In one fam- ily, however, the mutated allele displayed an alternative haplo- 1 Oppenheim JJ, Zachariae CO, Mukaida N et al. Properties of type with six out of the nine intragenic polymorphisms being the novel proinflammatory supergene ‘intercrine’ cytokine family. 6 Annu Rev Immunol 1991; 9:617–48. different. 2 Oka M, Berking C, Nesbit M et al. Interleukin-8 overexpression is In this report, we describe the identification of the same present in pyoderma gangrenosum ulcers and leads to ulcer forma- homozygous KIND1 mutation, c.676insC, in a Brazilian pedi- tion in human skin xenografts. Lab Invest 2000; 80:595–604. gree. Interestingly, the Brazilian patients were found to have a 3 Berking C, Takemoto R, Scaider H et al. Transforming growth different haplotype from the reported Pakistani haplotype. b factor- 1 increases survival of human melanoma through stroma Moreover, two new different KIND1 mutant haplotypes were remodeling. Cancer Res 2001; 61:8306–16. demonstrated in the affected Brazilian individuals indicating 4 Su WP, Davis MD, Weenig RH et al. Pyoderma gangrenosum: clini- copathologic correlation and proposed diagnostic criteria. Int J two separate origins for the c.676insC mutant alleles in one Dermatol 2004; 43:790–800. pedigree. Of note, in the Brazilian pedigree, we also report 5 Mielke V, Bauman JGJ, Sticherling M et al. Detection of neutrophil- the very unusual finding of intragenic nonpathogenic hetero- activating peptide NAP ⁄IL-8 and NAP ⁄IL-8 mRNA in human zygous single nucleotide polymorphisms (SNPs) alongside recombinant IL-1a- and human recombinant tumor necrosis fac- the same homozygous pathogenic KIND1 mutation in two a tor- -stimulated human dermal fibroblasts. An immunocytochemi- branches of the same family. cal and fluorescent in situ hybridization study. J Immunol 1990; The affected individuals are fourth and fifth generation 144:153–61. 6 Saito S, Yasui K, Hosoda W et al. CD30+ anaplastic large cell descendants of immigrants who came to Porto Alegre (Brazil) lymphoma complicated by pyoderma gangrenosum with increased in the early 20th century from Treviso (Italy). Part of the ped- levels of serum cytokines. Eur J Haematol 2006; 77:251–4. igree (illustrated in Figure 1a and referred to as family 1) is 7 Weenig RH, Davis MDP, Dahl PR et al. Skin ulcers misdiagnosed as nonconsanguineous and in generation IV contains three unaf- pyoderma gangrenosum. N Engl J Med 2002; 347:1412–18. fected individuals and seven siblings with KS, two of whom 8 Jones SC, Evans SW, Lobo AJ et al. Serum interleukin-8 in inflam- died prematurely from skin fragility ⁄infection. The part of the matory bowel disease. J Gastroenterol Hepatol 1993; 8:508–12. pedigree referred to as family 2 is consanguineous and in gen- 9 Volin MV, Harlow LA, Woods JM et al. Treatment with sulfasal- azine or sulfapyridine, but not 5-aminosalicyclic acid, inhibits eration V contains one individual with KS. The clinical features basic fibroblast growth factor-induced endothelial cell chemotaxis. in all affected subjects comprised trauma-induced blistering, Arthritis Rheum 1999; 42:1927–35. photosensitivity and poikiloderma as well as periodontal

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1282 Correspondence

(a)

(b)

Fig 1. (a) Brazilian pedigree with Kindler syndrome. In family 1, generation IV contains seven affected siblings, two of whom are deceased. In family 2, consanguinity is present and there is one individual with Kindler syndrome in generation V. Asterisks indicate genomic DNA available for KIND1 sequencing ⁄haplotype analysis. (b) Clinical features of Kindler syndrome in this Brazilian family. Poikiloderma is evident on the face, neck and hands and ulceration is present on the lower lip. inflammation and severe oesophageal stenosis requiring and the Guy’s and St Thomas’ Local Ethics Committee and in frequent dilatations (Fig. 1b). compliance with the Helsinki Accord, blood samples were Following ethical approval from both the Human Research taken for DNA extraction, polymerase chain reaction (PCR) Ethics Committee of the Hospital de Clinicas de Porto Alegre amplification and gene sequencing of KIND1. Briefly, PCR

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1283

(a)

(b)

Fig 2. (a) Sequencing of genomic DNA from affected individuals in family 1 and family 2 reveals homozygous insertion of a cytosine at position 676 in exon 5 of the KIND1 gene, c.676insC (arrows). In control DNA, there are seven instead of eight cytosine nucleotides. (b) Sequencing of two single nucleotide polymorphisms (SNPs) within intron 4 of the KIND1 gene (IVS4+8T ⁄C and IVS4+34C ⁄T) reveals heterozygosity for both polymorphisms in individuals from family 1 but nucleotide homozygosity in the one affected individual in family 2. These particular SNPs are not pathogenic but the finding of heterozygous intragenic SNPs in individuals harbouring a homozygous pathogenic mutation within the same gene is a very unusual finding. amplification of genomic DNA was carried out using 14 pairs IVS4+34C ⁄T, IVS4-25A ⁄G and IVS4-17A ⁄C). Additionally, we of primers spanning the coding exons and flanking introns of have analysed three further SNPs in the affected individuals the KIND1 gene, as described elsewhere.3,6 The PCR products (IVS5-17C ⁄T ⁄A, c.772T ⁄C and c.1695T ⁄C) (NM_017671.4). were purified using a QIAquick PCR Purification Kit (Qiagen, Intragenic haplotype analysis revealed that the Brazilian pedi- Crawley, U.K.) and sequenced in an ABI 310 genetic analyser gree arose on a different genetic background compared with (Applied Biosystems, Warrington, U.K.). the two Pakistani haplotypes (Table 1). Surprisingly, however, Mutational analysis disclosed a homozygous single nucleo- haplotype analysis of the Brazilian families showed that the tide insertion mutation, c.676insC, in exon 5 of KIND1 in all mutant alleles were different in families 1 and 2 (Fig. 2b and affected individuals screened for KS (Fig.2a). Because this Table 1). In consanguineous family 2, the affected individ- particular frameshift mutation has been reported previously in ual was homozygous for all polymorphic markers, but in several Pakistani families,6,7 we carried out KIND1 haplotype nonconsanguineous family 1, affected individuals were het- analysis on several members of the Brazilian pedigree (III-1, erozygous for two nonpathogenic SNPs in intron 4, IV-4, IV-6 and V-1). We also expanded the haplotype analysis notwithstanding that the mutation c.676insC was present on previously described for the affected Pakistani patients with both KIND1 alleles. Genomic analysis of DNA from the mother the homozygous c.676insC, typing for several intragenic in family 1 (who is also of Italian descent) showed that she (exonic and intronic) polymorphisms, as reported previous- was heterozygous for the c.676insC mutation (data not ly.6 Sequencing of genomic DNA of affected individuals from shown) and her haplotype was similar to that of her affected both Brazilian and Pakistani backgrounds has identified the offspring (Table 1). previously reported nine nonpathogenic SNPs (IVS1-11T ⁄G, In conclusion, the KIND1 mutation c.676insC has now c.114T ⁄G, IVS2+20C ⁄T, IVS2-4G ⁄A, c.479T ⁄C, IVS4+8T ⁄C, been shown to have arisen on four different haplotypes. This

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1284 Correspondence

Table 1 Haplotype analysis for single nucleotide polymorphisms (SNPs) in the KIND1 gene associated with the mutation c.676insC. In the Brazilian pedigree, haplotyping reveals two additional mutant alleles which are both different from the Pakistani haplotypes. Of note, the individuals with Kindler syndrome in family 1 are homozygous for c.676insC but heterozygous for two SNPs in intron 4. Haplotype analysis of the mother from Brazilian family 1 is similar to that of her offspring

Affected individuals Affected individual Mother of affected individuals KIND1 Exon (·), Pakistani Pakistani polymorphism Intron (IVS) haplotype 1 haplotype 2 Brazilian family 1 Brazilian family 2 Brazilian family 1 IVS1-11T ⁄G IVS2 G G G G G c.114T ⁄C ·2C C C C C IVS2+20C ⁄T IVS2 C C C C C IVS2)4G ⁄A IVS3 A G A A A c.479T ⁄C ·4T C T T T IVS4+8T ⁄C IVS4 C T T ⁄CTT⁄C IVS4+34C ⁄T IVS4 T C C ⁄TCC⁄T IVS4)25A ⁄G IVS5 A G A A A IVS4)17A ⁄C IVS5 A C A A A IVS5)17C ⁄T ⁄A IVS6 C T A A A c.772T ⁄C · 6C T T T T c.1695T ⁄C · 13 T C T T T mutation represents the most common recurrent pathogenic 3 Siegel DH, Ashton GH, Penagos HG et al. Loss of kindlin-1, a human KIND1 gene mutation in patients with KS and is probably due homolog of the Caenorhabditis elegans actin–extracellular–matrix linker to slipped mispairing in a run of seven consecutive cytosine protein UNC-112, causes Kindler syndrome. Am J Hum Genet 2003; 73:174–87. residues during DNA replication. Interfamilial and intrafamilial 4 Kloeker S, Major MB, Calderwood DA et al. The Kindler syndrome variability in the clinical phenotype is evident for individuals protein is regulated by transforming growth factor-beta and with this mutation, although such variability is not uncom- involved in integrin-mediated adhesion. J Biol Chem 2004; 8 mon in KS. In addition, the occurrence in a single pedigree 279:6824–33. of the same mutation that causes a very rare disease arising on 5 Herz C, Aumailley M, Schulte C et al. Kindlin-1 is a phosphoprotein two different mutant alleles is extremely unusual and serves as involved in regulation of polarity, proliferation, and motility of epi- a useful reminder to avoid assumptions of consanguinity dermal keratinocytes. J Biol Chem 2006; 281:36082–90. 6 Ashton GH, McLean WH, South AP et al. Recurrent mutations in when homozygous mutations are delineated, especially for kindlin-1, a novel keratinocyte focal contact protein, in the auto- uncommon inherited diseases. somal recessive skin fragility and photosensitivity disorder, Kindler syndrome. J Invest Dermatol 2004; 122:78–83. Acknowledgments 7 Thomson MA, Ashton GH, McGrath JA et al. Retrospective diagnosis of Kindler syndrome in a 37-year-old man. Clin Exp Dermatol 2006; We thank all subjects for their participation in the study. Sup- 31:45–7. port for the project was kindly provided by the Dystrophic 8 Lai-Cheong JE, Liu L, Sethuraman G et al. Five new homozygous mutations in KIND1 gene in Kindler syndrome. J Invest Dermatol 2007; Epidermolysis Bullosa Research Association (DebRA U.K.) and 127:2268–70. the Barbara Ward Children’s Foundation. J.E.L-C. is supported by a Wellcome Trust Clinical Training Fellowship. Key words: blister, genetic, haplotype, kindlin-1, Kindler syndrome Conflicts of interest: none declared. Service of Dermatology, Universidade B.C.F. MARTIGNAGO Federal do Rio Grande do Sul, Hospital de J.E. LAI-CHEONG* Clı´nicas de Porto Alegre, Brazil L. LIU* *Genetic Skin Disease Group, St John’s Institute of J.A. MC G RATH* Dermatology, Division of Genetics and T.F. CESTARI Transient macular erythema with eosinophilia Molecular Medicine, King’s College London, London, U.K. in a patient carrying the FIP1L1-PDGFRA fusion E-mail: [email protected] gene

References DOI: 10.1111/j.1365-2133.2007.08220.x

1 Kindler T. Congenital poikiloderma with traumatic bulla formation SIR, Hypereosinophilic syndrome (HES) is a heterogeneous and progressive cutaneous atrophy. Br J Dermatol 1954; 66:104–11. group of haematological disorders, characterized by eosino- 2 Jobard F, Bouadjar B, Caux F et al. Identification of mutations in a )1 new gene encoding a FERM family protein with a pleckstrin philia (> 1500 cells lL ) that lasts for more than 6 months, homology domain in Kindler syndrome. Hum Mol Genet 2003; with no sign of reactive eosinophilia from other causes (such 12:925–35. as parasitic infections), and evidence of end-organ damage.1

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1285

Recently, the FIP1L1-PDGFRA (F ⁄P) fusion gene was identified an association between cutaneous manifestations and the F ⁄P in some patients with HES.2 According to recent World Health fusion gene has rarely been reported. Organization criteria, HES with F ⁄P fusion genes can be classi- A 58-year-old woman presented with a 10-day history of fied as chronic eosinophilic leukaemia (CEL)3 or F ⁄P-positive fever and skin eruption on her trunk and extremities. The CEL ⁄HES. While classical HES has several skin manifestations,4 eruptions were well-demarcated, dark-red erythematous

(a) (b)

(c) (d)

Fig 1. The patient’s clinical and histological features. (a) Round or oval erythematous lesions of various sizes were distributed on the trunk and extremities (10 days after the onset of the disease). A perivascular or diffuse cellular infiltrate was observed in the entire dermis (b) and subcutaneous fat (c). The cellular infiltrate consisted of numerous eosinophils with occasional degranulation and lymphocytes (d).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1286 Correspondence lesions with slight peripheral elevation and occasional central abnormality, no cellular atypia or blast cells were noted. In pigmentation with scaling (Fig. 1a). No lymphadenopathy was the present case, the aetiopathology of the skin eruptions noted. Laboratory findings showed a significant blood was not fully defined. HES has been known to exhibit a vari- ) hypereosinophilia (16Æ1 · 103 lL 1) and increased levels of ety of skin manifestations.4 In addition, there has been one ) aspartate aminotransferase (68 IU L 1), alanine aminotrans- report of a patient with HES ⁄CEL with the F ⁄P gene who ) ) ferase (91 IU L 1), c-glutamyl transpeptidase (73 IU L 1), had eosinophilic cellulitis.5 Thus, it can be postulated that ) antileucoproteinase (766 IU L 1), lactic dehydrogenase (488 the macular eruptions resulted from tissue damage caused by ) ) IU L 1), C-reactive protein (4Æ8mgdL 1), IgM (320 invasive abnormal eosinophils carrying the gene abnormality. ) ) mg dL 1) and serum interleukin (IL)-5 (9Æ7pgmL 1, normal A recent report showed that the transplantation of F ⁄P-trans- ) <7Æ8pgmL 1). Despite careful examinations, including echo- duced haematopoietic stem cells into mice resulted in the cardiography, electrocardiogram, endoscopy and computed development of chronic myelogenous leukaemia-like features, tomography, neither cardiac, gastrointestinal, nor neurovascu- which did not resemble HES.6 Transgenic T-cell IL-5 over- lar involvement was observed. Histologically, a diffuse cellular expression was required for hypereosinophilia and HES-like infiltrate, consisting of a significant number of eosinophils and organ eosinophil involvement. These findings suggest that the to a lesser extent of lymphocytes, was present in the entire der- F ⁄P fusion gene may not be sufficient for the development of mis and subcutaneous fat lobules (Fig. 1b–d). Within 2 weeks, hypereosinophilia or HES symptoms. Notably, the present case the patient’s skin eruptions and laboratory data improved spon- had a marked blood hypereosinophilia during the skin erup- taneously without any other specific medications. The patient tions associated with an increased serum IL-5; the eosinophil had been taking several dietary supplements; however, a chal- count then declined after the skin eruptions improved. This lenge test by peroral administration of these drugs was nega- indicates that unknown events and factors stimulated eosinophil tive. Parasitic infections were excluded by faeces examination production in addition to the presence of the F ⁄P gene. In this and serum antibody titres. In contrast, elevated serum IgG respect, it is interesting to note the increased serum IgG levels titres, but not IgM titres, against herpes simplex virus (HSV) for VZV and ⁄or HSV. Viral reactivation might occur in the and varicella zoster virus (VZV) were observed 3 weeks after patient during the course of the disease. It was postulated that the onset of the disease (enzyme immunoassay index: HSV the development of eosinophilia and the marked eosinophil 36Æ2–70Æ4 and VZV 11Æ7–39Æ2). There were no changes in the infiltration of the skin might be provoked by virus reactivation. antibody titres against cytomegalovirus and Mycoplasma pneumo- The screening test for the F ⁄P fusion gene should be carried niae. Although blood eosinophil counts returned to normal lev- out in patients with eosinophilic skin eruptions of unknown els after the skin symptoms improved, intermittent mild aetiology, irrespective of their blood eosinophilia level and the increases in the eosinophil counts, ranging from 256 to degree of cellular atypia. ) 828 cells lL 1, were noted. On bone-aspiration biopsy, an increased number of eosinophil lineage cells was noted, but Acknowledgments the number of blast cells was not excessive. However, on poly- merase chain reaction analysis, the F ⁄P fusion gene was We thank Dr Ayako Arai, Department of Hematology, Internal detected (Fig. 2); this was further confirmed by sequencing Medicine, Tokyo Medical and Dental University, for the analysis (data not shown). Since the classical diagnostic criteria FIP1L1-PDGFRA fusion gene analysis. for HES1 were not fulfilled, the patient was diagnosed as having a type of CEL based on the F ⁄P gene abnormality. Department of Dermatology, H. YAHARA We have presented a case of transient macular erythema Graduate School, Tokyo Medical and T. SATOH in a patient carrying the F ⁄P fusion gene. Despite the gene Dental University, 1-5-45 Yushima, T. HASHIMOTO Bunkyo-ku, Tokyo 113-8519, Japan H. YOKOZEKI Correspondence: Takahiro Satoh. BM PB E-mail: [email protected]

References

1 Chusid MJ, Dale DC, West BC et al. The hypereosinophilic syn- drome: analysis of fourteen cases with review of the literature. Medi- cine (Baltimore) 1975; 54:1–27. 2 Cools J, DeAngelo DJ, Gotlib J et al. A tyrosine kinase created by fusion of the PDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hypereosinophilic syndrome. N Engl J Med FIP1L1-PDGFRA 2003; 348:1201–14. 3 Gotlib J, Cools J, Malone JM III et al. The FIP1L1-PDGFRalpha fusion tyrosine kinase in hypereosinophilic syndrome and chronic eosino- Fig 2. Polymerase chain reaction analysis for the FIP1L1-PDGFRA fusion philic leukemia: implications for diagnosis, classification, and man- gene. BM, bone marrow cells; PB, peripheral blood cells. agement. Blood 2004; 103:2879–91.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1287

4 Kazmierowski JA, Chusid MJ, Parrillo JE et al. Dermatologic manifes- Balzer and Fouquet1 in 1903 and was named by Hubler et al.2 tations of the hypereosinophilic syndrome. Arch Dermatol 1978; in 1978. To date only 28 cases have been reported in the lit- 114:531–5. erature.3–6 It tends to appear in middle-aged women (median 5 Davis RF, Dusanjh P, Majid A et al. Eosinophilic cellulitis as a pre- age 43 years; male ⁄female ratio 9 : 19). In most previous senting feature of chronic eosinophilic leukaemia, secondary to a deletion on chromosome 4q12 creating the FIP1L1-PDGFRA fusion reports, the lesions were small and distributed in the periau- 3,6 gene. Br J Dermatol 2006; 155:1087–9. ricular area. MEP is extremely rare, and its clinicohisto- 6 Yamada Y, Rothenberg ME, Lee AW et al. The FIP1L1-PDGFRA fusion pathological definition and treatment have not been gene cooperates with IL-5 to induce murine hypereosinophilic syn- established. We report a case of MEP with unusually wide- drome (HES) ⁄chronic eosinophilic leukemia (CEL)-like disease. Blood spread lesions of the face, and describe its immunohistochem- 2006; 107:4071–9. ical features and the effectiveness of oral etretinate. Key words: eosinophilic leukaemia, erythema multiforme, FIP1L1-PDGFRA, A 44-year-old, otherwise healthy Japanese man found sev- fusion gene, herpes virus, hypereosinophilic syndrome eral tiny white papules without any symptoms on the right temple and right submandibular area, 9 months before he Conflicts of interest: none declared. consulted our hospital. The white papules increased in number to form plaques. There were no apparent causative factors: no episodes of trauma, chemical peeling, topical agents or other skin diseases. The patient had tried various antibiotic creams, A case of milia en plaque successfully treated corticosteroid creams, etretinate creams and oral antibiotics, with oral etretinate without benefit. On examination, there were numerous whitish, globoid DOI: 10.1111/j.1365-2133.2007.08223.x papules up to 1 mm in diameter, superimposed on an ery- thematous ⁄brownish, indurated plaque covering the entire SIR, Milia en plaque (MEP) is an uncommon variant of pri- right cheek (Fig.1a). On the glabella, the left temple and the mary milia, characterized by numerous tiny grouped milia left side of the chin there were tiny whitish papules densely overlying an erythematous plaque. MEP was first described by located on brownish-pigmented macules. Bilateral eyebrows

(a) (b)

(c)

Fig 1. (a) On the right cheek there were many pearly, whitish, globoid papules, superimposed on erythematous ⁄brownish indurated plaques. (b) Many cysts, lined by stratified squamous epithelium and filled with keratin in laminated layers, were located immediately beneath the epidermis and within the dermis (haematoxylin and eosin; original magnification · 40). (c) Before treatment, expression of cytokeratin 6 was seen in the cyst walls of milia, and in all the layers of the epidermis inside the lesions (original magnification · 400).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1288 Correspondence

(a) (b)

(c)

Fig 2. (a) On the right cheek, the number of milia had remarkably decreased, and the brownish plaques had lightened to the normal skin colour 1 year after the start of treatment with oral etretinate. (b) After treatment, the number and size of keratin-filled cysts had remarkably decreased and infiltration of mononuclear cells around cysts had decreased (haematoxylin and eosin; original magnification · 40). (c) The expression of cytokeratin 6 was notably decreased in all the layers of the lesional epidermis (original magnification · 100). were sparse and whitish papules were also scattered. No walls and the suprabasal layers were positively stained. The abnormalities were found on laboratory studies. expression of proliferating cell nuclear antigen (PCNA), a pro- Biopsy specimens were taken from the lesion of the right liferation marker, was increased in the outer layer of the cyst cheek. Haematoxylin and eosin staining showed many cysts, walls and the epidermis inside the lesion. These immunohisto- lined by stratified squamous epithelium, filled with keratin in chemical studies indicated that the basal layer of the cyst walls laminated layers, immediately beneath the epidermis and had actively proliferated and that the inner layer of the cyst within the dermis (Fig. 1b). Some cysts contained vellus hairs. walls was ectopically expressing inflammatory keratins and Around these cysts, intense infiltration of lymphocytes with actively keratinizing. The epidermis overlying the cysts was monocytes and eosinophils was seen. Lymphocytes were poly- also activated. clonal and had no atypia. Basal pigmentation was found. We started treatment with oral etretinate 50 mg daily. After Results of the lupus band test were negative. MEP was diag- 1 month of treatment, the number of milia started to nosed. Previous cases were reported in patients with pseudo- decrease, and the brownish plaque lightened to the normal xanthoma elasticum4 or with discoid lupus erythematosus.5 skin colour. Three months after the start of treatment, the Some authors suggested MEP to be a rare variant of mycosis dose of etretinate was decreased. The milia are now well con- fungoides3,7 or follicular mucinosis.8 In our case, however, histo- trolled with oral etretinate 10 mg on alternate days (Fig.2a). pathological study indicated no evidence of those diseases. After treatment with oral etretinate for 14 months, another On immunohistochemistry, expression of cytokeratins 6 biopsy of the right cheek was performed. Haematoxylin and and 16 was seen in the cyst walls of the milia, and in all the eosin staining showed that the number and the size of keratin- layers of the lesional epidermis (Fig. 1c), while normal inter- filled cysts had remarkably decreased. Also, infiltration of follicular epidermis was negative. Cytokeratin 19, which is mononuclear cells around cysts was sparse (Fig. 2b). The expressed in the sweat gland apparatus, was negative in the expression of cytokeratin 6 was notably decreased in all the cyst walls. In areas of normal skin, granular layers and the layers of the epidermis (Fig. 2c). The expression of involucrin upper part of the hair follicles exhibited positive staining for was limited to the upper layers with no expression in the supra- involucrin. In the lesional skin, the inner layer of the cyst basal layer, and had decreased in the cyst walls. The number

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1289 of cells positive for PCNA had decreased slightly in the outer tissue diseases concurrently or consecutively. We describe two layer of the cyst walls and the basal layer of the epidermis. Japanese patients with an overlap syndrome consisting of limi- Etretinate is a retinoid. These are known to suppress kerati- ted systemic sclerosis (lSSc) and discoid lupus erythematosus nization in various skin disorders, such as psoriasis, Darier dis- (DLE). The coexistence of these two disorders has rarely been ease and cystic acne, and to induce peeling of the keratinizing reported. It is well known that many autoimmune diseases are layers of the skin. Retinoids suppress the expression of kerati- associated with certain human leucocyte antigen (HLA) types. nizing markers such as involucrin and cytokeratins 6 and 16,9 Our present patients had some common clinical and labora- by directly downregulating their promoter activities. Retinoids tory findings, including anticentromere antibody (ACA) positi- also have anti-inflammatory effects.10 These mechanisms may vity and common HLA typing (Table 1). explain the effectiveness of etretinate in this case. This is the Patient 1, a 42-year-old Japanese woman, had had erythem- first case of MEP to be treated successfully with oral etretinate. atous skin lesions on her forehead for 5 years. Raynaud phe- nomenon and sclerodactyly developed in the last 3 years Department of Dermatology, N. ISHIURA before she was admitted to our clinic. Physical examination Graduate School of Medicine, M. KOMINE revealed multiple reddish scaly erythematous plaques on her University of Tokyo, 7-3-1 Hongo, T. KADONO forehead. The surface was slightly atrophic (Fig.1a). Sclero- Bunkyo, Tokyo 113-8655, Japan K. KIKUCHI dactyly, petechiae and nail fold bleeding were observed E-mail: [email protected] K. TAMAKI on her fingers and hands (Fig. 1b). Skin biopsy from her forehead lesion revealed hyperkeratosis, follicular plugging, References liquefaction degeneration and inflammatory infiltrates in the dermis. Direct immunofluorescence (DIF) studies showed con- 1 Balzer F, Fouquet C. Milium confluent retro-auriculaire bilateral. tinuous granular deposition of IgG, IgM and C3 along the Bull Soc Fr Derm Syph 1903; 14:361–2. dermoepidermal junction. Seroimmunological examination at 2 Hubler WR Jr, Rudolph AH, Kelleher RM. Milia en plaque. Cutis a titre of 1:1280 revealed positive fluorescent antinuclear anti- 1978; 22:67–70. bodies with a discrete-speckled pattern, and ACA. Antitopo- 3 Losada-Campa A, de la Torre-Fraga C, Cruces-Prado M. Milia en plaque. Br J Dermatol 1996; 134:970–2. isomerase I, rheumatoid factor, anti-SSA, anti-SSB, anti-Sm 4 Cho SH, Cho BK, Kim CW. Milia en plaque associated with and anti-RNP antibodies were all normal or negative. There pseudoxanthoma elasticum. J Cutan Pathol 1997; 24:61–3. was no evidence of calcinosis cutis or lung, renal or oesopha- 5 Boehm I, Schupp G, Bauer R. Milia en plaque arising in discoid geal involvement. lupus erythematosus. Br J Dermatol 1997; 137:649–51. Patient 2, a 59-year-old Japanese man, presented with a 6 Stefanidou MP, Panayotides JG, Tosca AD. Milia en plaque: a 10-year history of multiple scaly erythematous lesions on his case report and review of the literature. Dermatol Surg 2002; 28:291–5. upper extremities, and more recently developed lesions on his 7 Leverkus M, Rose C, Brocker EB et al. Follicular cutaneous T-cell lymphoma: beneficial effect of isotretinoin for persisting cysts and face and back. He had Raynaud phenomenon and sclerodactyly comedones. Br J Dermatol 2005; 152:193–4. which had developed in the 2 months prior to his admission 8 Gibson LE, Muller SA, Leiferman KM et al. Follicular mucinosis: clin- to our clinic. Telangiectasia appeared in several fingertips ical and histopathologic study. J Am Acad Dermatol 1989; 20:441–6. approximately 1 month before admission. There was evidence 9 Tomic-Canic M, Freedberg IM, Blumenberg M. Codominant regu- of mild oedema in his fingers, extending to the dorsum of his lation of keratin gene expression by cell surface receptors and hand. Physical examination revealed atrophic and slightly nuclear receptors. Exp Cell Res 1996; 224:96–102. depressed erythematous patches on his cheek and extremities. 10 Thacher SM, Vasudevan J, Chandraratna RA. Therapeutic applications for ligands of retinoid receptors. Curr Pharm Des 2000; 6:25–58. Swelling, stiffness and telangiectasia were observed in his fin- gers and hands. Skin biopsy from the sclerotic lesion on the Key words: etretinate, cytokeratin, involucrin, milia en plaque, proliferating cell extensor forearm showed thickened collagen bundles and atro- nuclear antigen phic sweat glands in the whole dermis. Histopathological find- Conflicts of interest: none declared. ings in the erythematous upper right arm lesions showed keratin plugs, liquefaction degeneration of the basal layer, vac- uolar degeneration of the basal cell layer with marked melanin incontinence and mild to severe inflammatory cell infiltration into the dermis. DIF on the lesional skin showed granular IgG, Limited cutaneous systemic sclerosis IgM and C3 deposits along the basement membrane zone. associated with discoid lupus erythematosus Laboratory studies revealed a positive antinuclear antibody in two Japanese patients with anticentromere (1:640), positive ACA, but no antitopoisomerase I antibody. antibodies He did not have calcinosis cutis or lung, renal or oesophageal involvement. DOI: 10.1111/j.1365-2133.2007.08224.x HLA types common to both our patients were A*2402, B*0702, Cw*0702, DQA1*0101, DQB1*050101 and SIR, Overlap syndromes are defined as clinical entities that sat- DRB1*0101 (Table 1). The frequencies of HLA-A*2402, isfy each of the diagnostic criteria of two different connective B*0702 and DRB1*0101 among HLA antigens were low in

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1290 Correspondence

Table 1 Common human leucocyte antigen (HLA) typing in our patients

HLA A B Cw DQA DQB DRB Patient 1 A*2402 B*0702 Cw*0102 DQA1*0101 DQB1*0401 DRB1*0101 B*5401 Cw*0702 DQA1*0303 DQB1*050101 DRB1*0410 Patient 2 A*2402 B*0702 Cw*0702 DQA1*0101 DQB1*050101 DRB1*0101 B*5201 Cw*1202 DQA1*0103 DQB1*060101 DRB1*1501

Common haplotypes are shown in bold.

B*07, Cw*07, DQB1*05 and DRB1*01. Although the con- (a) comitant presence of lSSc and DLE might be coincidental, we speculate that it could represent an expression of increased susceptibility conferred by certain HLA types. If the two dis- eases share a common genetic background, this commonality could be related to HLA phenotypes. DLE in our patients preceded the development of lSSc with ACA. Although SSc and DLE are not rare diseases, there have been only a few reports of these conditions occurring concur- rently.5–7 ACA is regarded as a distinct feature in SSc, particu- larly lSSc. However, several reports have emphasized that ACA is not specific to SSc alone. More recently, ACA has been detected in sera from patients with autoimmune rheumatic disorders including primary Sjo¨gren syndrome, rheumatoid arthritis, systemic lupus erythematosus and DLE.8–10 Relatively (b) little is known about the specific target of ACA in the various rheumatic disorders. Our cases may indicate that specific HLA types contribute to the development of coexistent lSSc and DLE. The mechanism of ACA production and how this relates to the development of the clinical syndromes with which it is associated is unknown, although HLA molecules are likely to play a role. Gene–gene and ⁄or gene–environment interactions presumably modulate disease onset and ⁄or the clinical course.

Department of Dermatology, St Marianna University T. KAWAKAMI School of Medicine, 2-16-1 Sugao, Miyamae-ku, K. KAWASAKI Kawasaki, Kanagawa 216-8511, Japan Y. SOMA E-mail: [email protected]

Fig 1. Patient 1. (a) Reddish atrophic plaques with a scaly surface on References the forehead. (b) Nail fold bleeding on the fingers. 1 Reveille JD, Fischbach M, McNearney T et al. Systemic sclerosis in 3 US ethnic groups: a comparison of clinical, sociodemographic, serologic, and immunogenetic determinants. Semin Arthritis Rheum healthy Japanese controls (35Æ6%, 5Æ2% and 4Æ76%, respec- 2001; 30:332–46. 2 Genth E, Mierau R, Genetzky P et al. Immunogenetic associations of tively). HLA-DQB1*05, DRB1*01 and DRB1*04 are associated scleroderma-related antinuclear antibodies. Arthritis Rheum 1990; 1,2 with the presence of ACA. In Japanese individuals, the HLA- 33:657–65. DRB1*0101, *0405 and *1302 alleles are associated with high 3 Kuwana M, Okano Y, Kaburaki J, Inoko H. HLA class II genes asso- ACA titres.3 It seems likely that ACA is influenced by the pres- ciated with anticentromere antibody in Japanese patients with sys- ence of these HLA types. On the other hand, DLE appears temic sclerosis (scleroderma). Ann Rheum Dis 1995; 54:983–7. to be associated with the HLA-A*03, B*07, Cw*07 and 4 Millard TP, Kondeatis E, Vaughan RW et al. Polymorphic light DRB1*15 haplotype.4 The biological significance of these vari- eruption and the HLA DRB1*0301 extended haplotype are inde- pendent risk factors for cutaneous lupus erythematosus. Lupus ous associations, some of which have previously been 2001; 10:473–9. reported, are far from clear, and the combined data suggest a 5 Asherson RA, Angus H, Mathews JA et al. The progressive systemic complex role for the major histocompatibility complex in the sclerosis ⁄systemic lupus overlap: an unusual clinical progression. development of DLE. Both our patients were positive for HLA- Ann Rheum Dis 1991; 50:323–7.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1291

6 Sasaki T, Nakajima H. Systemic sclerosis (scleroderma) associated Staging examinations were done in 2005 and 2007. White with discoid lupus erythematosus. Dermatology 1993; 187:178–81. blood cell counts were between 21Æ0 · 109 and 28Æ2 · ) 7 Hayakawa K, Nagashima M. Systemic sclerosis associated with dis- 109 L 1, with 80–86Æ7% lymphocytes. Flow cytometry of seminated discoid lupus erythematosus. Int J Dermatol 1993; peripheral blood showed 80% lymphocytes; 70% of the 32:440–1. 8 Chan HL, Lee YS, Hong HS, Kuo TT. Anticentromere antibodies lymphoid cells were CD19– CD5– CD23+, characteristic for (ACA): clinical distribution and disease specificity. Clin Exp Dermatol B-CLL. Zap expression of 22Æ5% and low CD38 expression 1994; 19:298–302. indicated a favourable prognosis. Fluorescent in situ hybridiza- 9 Aragone MG, Balestrero S, Parodi A, Rebora A. Anti-centromere tion (FISH) analysis did not reveal the presence of 17p deletion antibodies in 2 patients with discoid lupus erythematosus and no or 12 trisomy; however, biallelic 13q14 deletion was shown signs of systemic sclerosis. Acta Derm Venereol (Stockh) 1999; 79:165. in 80% of the cells. Polymerase chain reaction analysis of the 10 Katano K, Kawano M, Koni I et al. Clinical and laboratory features peripheral blood detected polyclonal TCR-c gene rearrange- of anticentromere antibody positive primary Sjo¨gren’s syndrome. J Rheumatol 2001; 28:2238–44. ment. Bone marrow showed leukaemic infiltration with small, Key words: anticentromere, antibody, discoid lupus erythematosus, human leucocyte mature CD79a+ CD30) lymphocytes. FISH reaction showed antigen, limited cutaneous systemic sclerosis biallelic deletion of chromosome 13. Flow cytometric analysis Conflicts of interest: none declared. showed 67% lymphoid cells: 87% of them were CD19+, CD5+, CD23+, j positive. The presence of ALCL or MF was not detected in the bone marrow. Abdominal ultrasound revealed enlarged spleen with 75 mm diameter and hepatomegaly. Slightly enlarged reactive Coexistence of primary cutaneous anaplastic lymph nodes were detected by computed tomographic scan in large cell lymphoma and mycosis fungoides the right axillary region as well as 5–13-mm lymph nodes in in a patient with B-cell chronic lymphocytic cervical, axillary, mediastinal and hilar regions. leukaemia B-CLL Rai stage II was diagnosed with good prognostic parameters, and no specific treatment was initiated. Because of DOI: 10.1111/j.1365-2133.2007.08226.x ALCL the patient received X-ray irradiation (0Æ8-Gy fractions, 11 fractions per session, 20 sessions in total) which resulted SIR, Concordant lymphomas are histologically distinguishable in complete healing of the ulcer. Because of early-stage MF lymphoid neoplasms involving different anatomical sites. These the patient was treated with topical steroid with almost com- lymphomas are rarely seen and coexistence of more than two plete response. lymphomas is exceptional. We report a case of primary cutane- Association of cutaneous lymphoproliferations with other ous CD30+ anaplastic large cell lymphoma (ALCL) followed by types of Hodgkin and non-Hodgkin lymphomas is rare: the true mycosis fungoides (MF) one and a half years later in a patient incidence is difficult to assess. Second malignancies are frequent with B-cell chronic lymphocytic leukaemia (B-CLL). complications in patients with CLL;1 however, the association of We describe a 75-year-old man in whom Rai stage II B-CLL B-CLL and cutaneous T-cell lymphoma (CTCL) is uncommon: with good prognostic parameters was diagnosed in 1998; no only a few cases, including composite lymphomas, have been treatment was necessitated. Skin symptoms started in August published.2,3 Kantor et al. investigated 544 patients with CTCL 2005 as firm nodules that coalesced into a 20 x 20 cm large and found only five with second lymphomas.4 Barzilai et al. anal- ulcer below the right axillary region (Fig.1a). ysed 398 patients with MF and found 11 patients with associ- Histological examination showed a dense, monomorphic ated B-cell lymphoma.5 In contrast, Hallermann et al. stated that dermal infiltrate composed of anaplastic cells larger than the frequency of a second lymphoid neoplasm in patients with 30 lm with variable nuclei, irregular shape and with multiple, cutaneous lymphoma is underestimated.6 medium-sized and prominent nucleoli (Fig. 1b). Cells showed Multiple CD30+ lymphoproliferations – ALCL, lymphoma- strong positivity for CD30 (Fig. 1c), weak positivity for CD4 toid papulosis and Hodgkin disease – can occur together.7 and intense cytoplasmic TIA-1 staining. Anaplastic lymphoma Assaf et al. described a patient who presented initially with kinase reaction, epithelial membrane antigen, CD3, CD5, CD7, T-prolymphocytic leukaemia, and subsequently with MF, lym- CD8, CD79a and B-cell markers were negative. A T-cell recep- phomatoid papulosis and ALCL: in this case an identical tor (TCR)-c gene rearrangement study did not reveal clonal monoclonal gene rearrangement with the same chromosomal gene rearrangement. The diagnosis of ALCL was established. abnormalities was shown in all different clinical entities.8 One and a half years later, in January 2007, brownish, well- In our case the same TCR clonal rearrangement was not demarcated plaques developed on his trunk, involving < 10% proven and ALCL preceded MF, as in a patient described by of the body surface. Histology and immunohistochemistry Kang et al.9 True coexistence of MF and ALCL, and not large showed Pautrier microabscesses in the epidermis and a CD4+, cell transformation of MF, is a rarity. Lee et al. reported a simi- CD30– dermal infiltrate of small lymphoid cells without a lar case of MF with preceding ALCL; however, the same TCR CD30+ large cell component (Fig.2). TCR-c gene rearrange- gene rearrangement was shown in both processes.10 In our ment analysis revealed clonal rearrangement at 180 bp. patient three different malignant lymphoproliferative processes

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1292 Correspondence

(a) (b)

(c)

Fig 1. (a) Ulcerated lesion below the right axillary region. (b) Strong and diffuse dermal infiltrate composed of large anaplastic cells (haematoxylin and eosin; original magnification · 40). (c) Strong and diffuse cytoplasmic CD30 positivity (original magnification · 40).

factors and malignancy-prone genotype, or they may originate from different or unrelated precursors.

Department of Dermatology, M. MARSCHALKO´ Venereology and Dermato-oncology, J. CSOMOR* *First Department of Pathology and N. ERO} S Experimental Cancer Research and A´ .SZIGETI First Department of Internal Medicine, J. HA´ RSING Faculty of Medicine, J. SZAKONYI Semmelweis University, M. DE´ SAKNAI Ma´ria u. 41, H-1085 A. MATOLCSY* Budapest, Hungary J. DEMETER E-mail: [email protected] S. KA´ RPA´ TI

References Fig 2. Epidermotropic lymphoid cell infiltrate with a Pautrier microabscess (haematoxylin and eosin; original magnification · 10). 1 Wiernik PH. Second neoplasms in patients with chronic lympho- cytic leukemia. Curr Treat Options Oncol 2004; 5:215–23. were diagnosed: first, B-CLL; then, after 7 years, cutaneous 2 Harland CC, Whittaker SJ, Ng YL et al. Coexistent cutaneous lymph- ALCL; then, 1Æ5 years later, a third lymphoproliferation, MF. oma and B-cell chronic lymphocytic leukaemia. Br J Dermatol 1992; According to the clinical follow-up these associations did not 127:519–23. 3 Hull PR, Saxena A. Mycosis fungoides and chronic lymphocytic cause an unfavourable course. leukaemia – composite T-cell and B-cell lymphomas presenting in The reason for the possible association between different the skin. Br J Dermatol 2000; 143:439–44. types of lymphomas remains unclear: the explanations include 4 Kantor AF, Curtis RE, Vonderheid EC et al. Risk of second malig- lymphoma-related immunosuppression, shared aetiological nancy after cutaneous T cell lymphoma. Cancer 1989; 63:1612–15.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1293

5 Barzilai A, Trau H, David M et al. Mycosis fungoides associated Tuberculin skin tests (TSTs) have been in widespread use with B-cell malignancies. Br J Dermatol 2006; 155:379–86. for the detection of asymptomatic infection, currently termed 6 Hallermann C, Kjell MK, Tiemann M et al. High frequency of pri- latent tuberculosis infection (LTBI). There is growing interest mary cutaneous lymphomas associated with lymphoproliferative in other diagnostic tests for LTBI that might be simpler to disorders of different lineage. Ann Hematol 2007; 86:509–15. 7 van den Berg A, Maggio E, Rust R et al. Clonal relation in a case interpret. As a result two ex vivo interferon (IFN)-c release of CLL, ALCL, and Hodgkin’s composite lymphoma. Blood 2002; assays (IGRAs) are now commercially available. We report the 100:1425–9. usefulness of the QuantiFERON -TB Gold (QFT-TB) test 8 Assaf C, Hummel M, Dippel E et al. Common clonal T-cell origin (Cellestis, Carnegie, Vic., Australia) in the confirmation of in a patient with T-prolymphocytic leukaemia and associated cuta- LTBI in a patient with erythema induratum. neous T-cell lymphomas. Br J Haematol 2003; 120:488–91. A 14-year-old boy was referred by the paediatricians with a 9 Kang SK, Chang SE, Choi JH et al. Coexistence of CD30-positive 4-month history of tender, hot painful lumps on both legs anaplastic large cell lymphoma and mycosis fungoides. Clin Exp Dermatol 2002; 27:212–15. (Fig.1a). One lesion had ulcerated and was discharging. There 10 Lee MW, Chi DH, Choi JH et al. A case of mycosis fungoides after was no history of night sweats or weight loss. He had been CD30 positive anaplastic large cell lymphoma. J Dermatol 2000; born in Sierra Leone, having moved to the U.K. 6 years ago. 27:458–61. There was a past medical history of epilepsy with a hemipare- sis and learning difficulties following a prolonged seizure in Key words: anaplastic large cell lymphoma, B-cell chronic lymphocytic leukaemia, concordant lymphoma, mycosis fungoides early life. He was on sodium valproate and lamotrigine. On examination, there were poorly defined, tender, indu- Conflicts of interest: none declared. rated plaques of 2–4 cm with hyperpigmentation and a scaly surface. There was no lymphadenopathy. A clinical diagnosis of erythema induratum was made. This was confirmed histo- logically with lobular inflammation of the subcutaneous fat Usefulness of the QuantiFERON test with noncaseating granulomas in the deep dermis and subcu- in the confirmation of latent tuberculosis taneous fat. Some of the vessels showed features of acute vas- in association with erythema induratum culitis with fibrinoid necrosis of the vessel walls (Fig. 1b). Chest X-ray was normal, and early morning sputum, urine DOI: 10.1111/j.1365-2133.2007.08227.x and gastric aspirates were consistently negative for acid-fast bacilli and culture of mycobacteria. A Mantoux test was nega- SIR, There is a wide variety of clinical presentations of cutane- tive; however, a positive QFT-TB test led us to a diagnosis of ous tuberculosis. Erythema induratum of Bazin is considered LTBI. He was treated with rifampicin, isoniazid and pyrazina- to be a tuberculide, a hypersensitivity response to Mycobacterium mide with successful resolution of his skin lesions. tuberculosis, in which typically there is a positive tuberculin test, The tuberculides represent a group of disorders that classi- evidence of manifest or past tuberculosis and a positive cally were associated with a focus of internal tuberculosis. response to antituberculous therapy. There is ongoing debate They are considered immune reactions within the skin due to about the pathogenesis of tuberculides and their association haematogenous dissemination of M. tuberculosis or its antigens with M. tuberculosis,1 which is thought to be a type IV, cell- from a primary source, in an individual with strong antituber- mediated response to an antigenic stimulus presumed to be culous cell-mediated immunity which evolves into a granu- mycobacterial antigens. The use of polymerase chain reaction lomatous inflammatory response. to detect mycobacterial DNA in skin lesions has recently Until recently, TSTs were in widespread use for detecting strengthened this association.2 LTBI. However, tuberculin tests are limited by difficulties with

(a) (b)

Fig 1. (a) Clinical appearance at presentation. (b) Photomicrograph showing lobular panniculitis with noncaseating granulomas in the deep dermis (haematoxylin and eosin; high-power view).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1294 Correspondence administration and in interpreting the results, particularly in and recommendations for research. Ann Intern Med 2007; 146:340– the context of previous bacille Calmette-Gue´rin (BCG) vaccina- 54. tion.3 Of newly diagnosed patients with pulmonary tuberculo- 5 Dinnes J, Deeks J, Kunst H et al. A systematic review of rapid diag- nostic tests for the detection of tuberculosis infection. Health Technol sis, 10–25% will have false-negative skin tests. This is due to Assess 2007; 11:1–196. a combination of factors, including inadequate nutrition, the 6 Mazurek GH, Jereb J, Lobue P et al. Guidelines for using the Quanti- inability to react to skin tests due to immunosuppression or FERON-TB Gold test for detecting Mycobacterium tuberculosis infection, the release of specific cytokines liberated in active tuberculosis United States. MMWR Recomm Rep 2005; 54:49–55. which inhibit the delayed hypersensitivity, severe tuberculosis Key words: erythema induratum, QuantiFERON, tuberculosis or concomitant corticosteroid use. A whole-blood IGRA for the detection of LTBI is an in vitro Conflicts of interest: none declared. T cell-based assay. The principal of the assay is that T cells of individuals previously sensitized with tuberculous antigen will produce IFN-c when they re-encounter mycobacterial anti- gens. Thus a high level of IFN-c production is presumed to Respiratory involvement in toxic epidermal be indicative of tuberculosis infection. Early assays used puri- necrolysis portends a poor prognosis that may fied protein derivative as the stimulating antigen; however, not be reflected in SCORTEN newer assays use RD1 antigens specific to M. tuberculosis. These antigens are not shared with BCG or most nontuberculous DOI: 10.1111/j.1365-2133.2007.08222.x mycobacteria. Currently, two methods exist for detecting the IFN-c released by the T cell: an enzyme-linked immunosor- SIR, Toxic epidermal necrolysis (TEN) is a life-threatening bent assay (e.g. QFT-TB) and an enzyme-linked immunospot erosive mucocutaneous condition almost exclusively related to assay (e.g. T-SPOT.TB; Oxford Immunotech, Abingdon, U.K.). drugs.1 The severity-of-illness score, ‘SCORTEN’, uses seven A recent meta-analysis found that the new IGRAs were a factors recorded within the first 24 h of admission to predict useful diagnostic tool with excellent specificity in the diagno- prognosis.2 A point is scored for each of the following: age ) sis of LTBI but concluded that further research was required > 40 years; known malignancy; pulse rate > 120 min 1; epi- to define their use in high-risk populations.4 A systematic dermal detachment > 10% body surface area; urea > 10 mmol ) ) ) review recommended the use of RD1 antigen-based assays in L 1; glucose > 14 mmol L 1; bicarbonate < 20 mmol L 1. Pre- preference to those based on a TST for the diagnosis of LTBI, dicted mortality rises from 3Æ2% for 0–1 points, through to especially in the context of previous BCG vaccination or im- 90% for ‡ 5 points.2 Recent analysis suggests that SCORTEN munocompromise.5 The U.S. Centers for Disease Control and performs well during the first 5 days of admission, and slightly Prevention published guidance for the use and interpretation better on day 3, irrespective of any delay to admission.3 Respi- of the QFT-TB test and recommend that these tests can be ratory tract involvement in TEN predicts a poor prognosis;4 used safely in all circumstances in which the TST is currently however, no discrete markers of respiratory compromise appear used.6 in SCORTEN. We report a case of TEN which led us to question In summary, we found the QFT-TB test to be useful in the whether SCORTEN accurately predicts mortality in patients with confirmation of suspected latent tuberculosis in the manage- severe respiratory involvement. ment of erythema induratum. These new IFN-c assays may pro- A 24-year-old man was admitted with a generalized erythem- vide a reliable and more easily administered alternative to TSTs. atous eruption. Three days previously he had taken a 200 mg ibuprofen tablet for a headache. He had no significant past med- Departments of Dermatology and J. ANGUS ical history, took no regular medication, and had never taken *Histopathology, Queen’s Medical Centre, C. ROBERTS ibuprofen before. The following day he took two further ibu- Nottingham NG7 2UH, U.K. K. KULKARNI* profen tablets, and later developed a sore throat, gritty eyes, E-mail: [email protected] I. LEACH* and a presternal rash. The day prior to admission he became R. MURPHY unwell and the rash became generalized. On admission he had a ) fever (39Æ4 C), tachycardia (pulse 114 min 1), mild tachyp- ) noea (respiratory rate 16 min 1) and slightly low oxygen satu- References rations (SaO2 94% on air). He had facial swelling, bilateral 1 Chuang YH, Kuo TT, Wang CM et al. Simultaneous occurrence of conjunctivitis, and erosions of the buccal mucosa and lips. A papulonecrotic tuberculide and erythema induratum and the identi- generalized tender morbilliform eruption was evident, with fication of Mycobacterium tuberculosis DNA by polymerase chain reaction. flaccid blistering involving more than 30% body surface area Br J Dermatol 1997; 137:276–81. (Fig.1a). Nikolsky’s sign was positive. A diagnosis of TEN was 2 MacGregor RR. Cutaneous tuberculosis. Clin Dermatol 1995; 13:245– made, and confirmed by punch biopsy. Blood tests showed 55. )1 )1 3 Nelson K. Tuberculin testing to detect latent tuberculosis in develop- serum urea 3Æ8 mmol L , glucose 6Æ8 mmol L and bicar- )1 ing countries. Epidemiology 2007; 18:348–9. bonate 28Æ4 mmol L . The patient was transferred to the 4 Menzies D, Pai M, Comstock G. Meta-analysis: new tests for the intensive care unit, and human intravenous immunoglobulin ) diagnosis of latent tuberculosis infection: areas of uncertainty was initiated (1 g kg 1 daily for 3 days).

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1295

Severe Adult Respiratory failure), but was randomized to con- (a) ventional treatment. A systemic inflammatory response syn- drome developed, leading to haemodynamic instability, acute renal failure requiring haemofiltration, and multisystem fail- ure. On day 14 an arterial haemorrhage from his oronaso- pharynx led to an irreversible destabilization of his condition. This fatal case of TEN occurred in a healthy young man who scored a maximum of 2 ⁄7 on the SCORTEN scale on days 1–3 of admission (scoring 1 for body surface area and 1 for tachycardia), predicting a mortality risk of 12Æ1%.2,3 Although he did not die directly from respiratory failure, his early and severe respiratory involvement was the significant factor in his stormy course and eventual demise. We feel that the SCORTEN did not reflect his poor prognosis indicated by his respiratory involvement. A prospective clinical study of 41 patients with TEN showed that 10 (27%) had bronchial necrosis.4 This was associated with significantly higher mortality compared with those with- out respiratory involvement or those with late-onset respira- tory complications such as pulmonary oedema and infection. The authors did not find any association between respiratory involvement and extent of epidermal detachment. Nine of the 10 patients required ventilation, and seven subsequently died. Dyspnoea, hypoxia, increased bronchial secretions, and normal chest X-rays, within the first 3 days of illness indicated early bronchial involvement. Bronchoscopy showed that healing (b) began at about day 15, suggesting that prolonged ventilation may be required, with concomitant risk of pulmonary and systemic complications. SCORTEN is commonly used to aid prognostication, but we would question whether it can accurately predict mortal- ity in patients with severe respiratory involvement. The ori- ginal cohort of 165 patients from which SCORTEN was devised did not include any patients who were ventilated ‘within the first days’.2 It is also unclear whether the second cohort on which the SCORTEN was validated included patients who required early ventilation. Hypoxia was one of eight factors with a statistically significant impact on progno- sis; however, it was not included in the final seven SCOR- TEN factors, as it did not remain a significant prognostic Fig 1. (a) Generalized blistering on a background of morbilliform factor when controlled for by a low bicarbonate and abnor- erythema. (b) Bronchoscopy showing extensive sloughing of mal heart rate. This could simply reflect a lack of severe bronchial mucosa. respiratory involvement in the study population, unless data to the contrary become available. The case presented is a reminder that bronchial epithelium Within 24 h of admission he developed respiratory dis- necrosis in TEN may be detected by the early presence of tress, obtundation, and an arterial blood gas confirmed type I dyspnoea, hypoxia, and increased bronchial secretion. It is respiratory failure with severe hypoxia (pO2 7Æ7 kPa on associated with a poor prognosis, and we would recommend )1 15 L min O2 via nonrebreath bag). He was intubated and further validation of SCORTEN in these patients. ventilated, but his oxygen requirements increased progres- sively, and sequential chest X-rays indicated development of Departments of Dermatology and J.S. HAGUE acute respiratory distress syndrome. Marked respiratory secre- *Anaesthesia and Intensive Care, J.M.R. GOULDING tions developed on day 3 of admission, and bronchoscopy on South Warwickshire General Hospitals NHS Trust, T.M.W. LONG* days 4 and 11 revealed severely inflamed and sloughy airways Warwick Hospital, Warwick CV34 5BW, B.C. GEE (Fig. 1b). He was entered in the CESAR trial (Conventional U.K. ventilation or Extracorporeal membrane oxygenation for E-mail: [email protected]

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1296 Correspondence

References

1 French LE, Prins C. Toxic epidermal necrolysis. In: Dermatology (Bolognia JL, Jorizzo JL, Rapini RP, eds). Philadelphia: Elsevier Mosby, 2003; 323–31. 2 Bastuji-Garin S, Fouchard N, Bertocchi M et al. SCORTEN: a severity- of-illness score for toxic epidermal necrolysis. J Invest Dermatol 2000; 115:149–53. 3 Guegan S, Bastuji-Garin S, Poszepczynska-Guigne E et al. Performance of the SCORTEN during the first five days of hospitalization to pre- dict the prognosis of epidermal necrolysis. J Invest Dermatol 2006; 126:272–6. 4 Lebargy F, Wolkenstein P, Gisselbrecht M et al. Pulmonary complica- tions in toxic epidermal necrolysis: a prospective clinical study. Inten- sive Care Med 1997; 23:1237–44.

Key words: respiratory, SCORTEN, toxic epidermal necrolysis

Conflicts of interest: none declared. Fig 1. Histopathological finding showing spongiform subcorneal pustules without eosinophil exocytosis (haematoxylin and eosin; original magnification · 200).

Hydroxyzine-induced acute generalized Table 1 Patch-test results exanthematous pustulosis Time of exposure

DOI: 10.1111/j.1365-2133.2007.08225.x Ingredient Concentration 48 h 72 h 7 days Cetirizine As is )) ) SIR, Antihistamines are often used to treat allergic and pruritic Levocetirizine As is )) ) disorders, but there are rare reports of drug eruptions with Hydroxyzine 2% aq1,3 IR ++, pustulos P these agents, including several caused by the oral antihist- 5% aq1,3 IR ++, pustulos P 1,3 amine, hydroxyzine.1–4 We report such a case. 10% aq IR ++, pustulos P Æ 4 A 73-year-old woman with psoriasis that had been well 2 5% pet ++ ++, pustulos P Colloidal silica 5% aq1,3 )) ) controlled for 15 years developed a pruritic scalp lesion for As is )) ) which a physician gave her an unknown medication. She Macrocrystalline 5% aq1,3 )) ) developed a rash 2 days later consisting of diffuse areas of ten- cellulose As is )) ) der, oedematous erythema with hundreds of nonfollicular Magnesium stearate As is1,3 )) ) pustules on her trunk and limbs. Microscopically, a biopsy Lactose 20% aq1,3 )) ) specimen from the abdomen showed spongiform subcorneal Pet, petrolatum; aq, aqueous; IR, irritated erythema; P, pigmen- pustules without eosinophil exocytosis and a perivascular lym- tation. phohistiocytic infiltrate without obvious papillary oedema or vasculitis (Fig.1). The patient was given intravenous methyl- prednisolone for 9 days, and the pustules rapidly resolved reactions (++) at 48 h to Atarax tablets and hydroxyzine over several days with extensive desquamation. By the time of 2Æ5% pet. The aqueous formulations produced irritated ery- discharge, she had only residual hyperpigmentation. thema. By 72 h, the reactions to Atarax and to hydroxyzine Forty days later, she was given hydroxyzine hydrochloride in all formulations tested were positive (++) with pustules (Atarax UCB, Brussels, Belgium) for pruritic psoriatic scalp (Fig.2). There was also focal flaring in previously involved lesions. She took one tablet and the next day again developed areas. None of the other ingredients in Atarax, or cetirizine diffuse erythema and pustules on her trunk and extremities, or levocetirizine, induced a positive reaction. Five control sub- associated with 1 day of fever. Her serum albumin was low jects had negative patch tests with Atarax tablets and ) (2Æ8mgdL 1) with a commensurately low serum calcium hydroxyzine (2Æ5% pet.). ) (7Æ9 mmol dL 1), but there was no renal dysfunction. The Hydroxyzine is often used to treat allergic diseases, but this pustules were sterile, with no growth on culture of the con- case is a reminder that serious hypersensitivity reactions are tents. The physician who had initially treated the patient con- possible even with an antihistamine. Acute generalized exan- firmed that the drug she took before the first episode had thematous pustulosis (AGEP) typically has hundreds of widely been hydroxyzine. distributed nonfollicular pustules on an oedematous, erythem- Patch tests were performed 1 month later, using Atarax atous base, predominantly in the intertriginous areas.5 It tablets and each of its ingredients, cetirizine (Zyertex) and resolves quickly, often in less than 15 days. The fact that the levocetirizine (Xyzal)(Table 1). The patient had positive eruption is drug-induced in a particular case can be confirmed

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1297

(a) (b)

(c) (d)

Fig 2. Patch-test results: positive (++) reaction with pustules to Atarax tablet at 48 h (a) and 72 h (b), and to hydroxyzine 2Æ5% pet. at 48 h (c) and 72 h (d). by patch testing. In this case, hydroxyzine in petrolatum 3 Lew BL, Haw CR, Lee MH. Cutaneous drug eruption from cetirizine yielded a more clear-cut reaction than did the various aqueous and hydroxyzine. J Am Acad Dermatol 2004; 50:953–6. preparations, so we suggest using a 2Æ5% pet. preparation if 4 Dalmau J, Serra-Baldrich E, Roe E et al. Skin reaction to hydroxyzine (Atarax) patch test utility. Contact Dermatitis 2006; 54:216–17. patch testing is indicated. Hydroxyzine has been reported to 1–4 5 Sidoroff A, Halevy S, Bavinck JN et al. Acute generalized exanthema- cause a generalized maculopapular eruption, but we were tous pustulosis (AGEP) – a clinical reaction pattern. J Cutan Pathol unable to find any published reports of hydroxyzine-induced 2001; 28:113–19. AGEP. Key words: acute generalized exanthematous pustulosis, hydroxyzine

*Department of Dermatology, Mackay Y-S. TSAI* Conflicts of interest: none declared. Memorial Hospital, 92, Sec 2, Chung-Shan N Rd, M-E. TU * Taipei 10449, Taiwan Y-H. WU * Mackay Medicine, Nursing and Y-C. LIN* Management College and Lee-Ming Institute of Technology, Tapei, Taiwan Malignant melanoma in a woman with Correspondence: Mei-Eng Tu. LEOPARD syndrome: identification of a E-mail: [email protected] germline PTPN11 mutation and a somatic BRAF mutation References

1 Michel M, Dompmartin A, Louvet S et al. Skin reactions to hydroxy- DOI: 10.1111/j.1365-2133.2007.08229.x zine. Contact Dermatitis 1997; 36:147–9. 2 Ash S, Scheman AJ. Systemic contact dermatitis to hydroxyzine. Am J SIR, LEOPARD syndrome (LS) is a congenital developmental Contact Dermat 1997; 8:2–5. disorder and is an acronym for multiple lentigines, electro-

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1298 Correspondence cardiographic conduction abnormalities, ocular hypertelorism, (a) pulmonary stenosis, abnormalities of genitalia, retardation of growth and sensorineural deafness.1 LS is also occasionally associated with malignant lesions such as leukaemia and neuro- blastoma.2,3 These clinical features overlap with those in Noonan syndrome (NS) and, consistent with this, heterozy- gous germline mutations of the gene for PTPN11 have been identified in ~45% of patients with NS and in > 80% of patients with LS.1 Malignant melanoma (MM) is a rare malignancy that usu- ally occurs in older people after several genetic perturbations.4 Indeed, while somatic mutations of BRAF have been identified in roughly two-thirds of MM,5 they are also frequently observed in benign naevi, implying that a further molecular defect (or defects) is required for the development of MM.4 Furthermore, while Clark et al. have suggested that MM pri- (b) marily develops from benign or atypical naevi through the stage of dysplastic naevi,6 Ackerman has postulated that MM can occur directly in the normal melanocytes.7 Here, we report the first case of LS with MM, and discuss a relationship between LS and MM. The patient was a 62-year-old Japanese woman who had numerous naevi of ~1 mm diameter over the whole body except for her palms and soles. The naevi had been present since her infancy and had increased with age. She also had short stature and cardiac abnormality. Thus, she was suspected as having LS. At 60 years of age, she noticed a black and par- tially dark brown macula on the left heel which steadily increased in size to 1Æ8 · 1Æ5 cm. This lesion was diagnosed as MM stage IA by pathological findings of specimens obtained by an incision biopsy, and was surgically excised with a 1-cm Fig 1. Histological findings of the pigmented macula on the left heel, margin. Tumour thickness was 0Æ9 mm, and melanoma cells indicating the development of melanoma cells in the lower epidermis with clear cytoplasm and atypicality were observed histologi- and upper dermis. (a) Low magnification; (b) high magnification cally in the lower epidermis and upper dermis (Fig.1). (haematoxylin and eosin staining). After obtaining informed consent, direct sequencing was per- formed for PTPN11 and BRAF, using genomic DNA extracted from leucocytes and excised normal skin and MM tissues. The primer sequences are available on request. Consequently, a novel heterozygous PTPN11 mutation (Thr468Pro, 1402AfiC) was identified in the three different tissues examined, and a common BRAF mutation (Val600Glu, 1799TfiA; previously designated Val599Glu, 1796TfiA) was detected in the MM tis- sue only (Fig.2). These results indicated that this woman had a germline PTPN11 mutation and a somatic BRAF mutation. The LS phenotype of this patient is explained by the germ- line PTPN11 mutation. In this regard, it is notable that the novel Thr468Pro mutation affects the 468th Thr residue. To date, seven mutations, including two recurrent Tyr279Cys and Thr468Met mutations, have been identified in LS, and all the seven mutations are located at the catalytic cleft of the PTPN11 protein.1,8 Thus, the Thr468Pro mutation in this patient is compatible with such a positional property of PTPN11 muta- tions leading to LS. Fig 2. Electrochromatograms showing a heterozygous germline This patient with LS also had MM. Although this is primarily PTPN11 mutation (Thr468Pro, 1402AfiC) in the genomic DNA of ascribed to the somatic BRAF mutation,4,5 recent studies have melanoma tissue, normal skin tissue and leucocytes, and a somatic suggested that a somatic BRAF mutation alone is insufficient BRAF mutation (Val600Glu, 1799TfiA) in melanoma tissue only.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1299 to cause MM.4 Similarly, although this woman had a PTPN11 7 Ackerman AB. The concept of malignant melanoma in situ. In: Patho- mutation, there has been no previous report describing the biology of Malignant Melanoma (Elder DE, ed.). Basel: Karger, 1997; development of MM in patients with a germline PTPN11 muta- 205–10. 8 Kontaridis MI, Swanson KD, David FS et al. PTPN11 (Shp2) mutations tion.1 In this context, it is noteworthy that both BRAF and in LEOPARD syndrome have dominant negative, not activating, PTPN11 are involved in the mitogen-activated protein kinase effects. J Biol Chem 2006; 281:6785–92. 1,4 (MAPK) cascade relevant to the development of MM. It is 9 Kondoh T, Ishii E, Aoki Y et al. Noonan syndrome with leukaemoid possible, therefore, that the somatic BRAF mutation took place reaction and overproduction of catecholamines: a case report. Eur J in apparently normal but germline PTPN11 mutation-positive Pediatr 2003; 162:548–9. skin tissue, resulting in the development of MM because of a Key words: BRAF, LEOPARD syndrome, malignant melanoma, PTPN11 drastic perturbation of the MAPK signalling. This notion would explain why MM arose from an apparently naevi-free region Conflicts of interest: none declared. at an old age, because such a somatic mutation would occur in both naevi-positive and -negative skin tissues in an age- dependent fashion. One may argue that while the BRAF mutation activates the 4,5 Anonychia, hyponychia and spontaneous MAPK signalling, recent studies have shown that LS-associ- amputation of the distal phalanges as a ated PTPN11 mutations impair catalytic functions and exert consequence of ischaemic necrosis of the dominant negative effects, in contrast to NS- and neoplasia- extremities after umbilical catheterization associated PTPN11 mutants that exert gain-of-function effects with excessive phosphatase activities.1,8 Indeed, the NS- and DOI: 10.1111/j.1365-2133.2007.08228.x neoplasia-associated mutations and the LS-associated mutations are mutually exclusive.1 However, despite the marked differ- SIR, We report here a case of early acquired anonychia, with ence in in vitro functions of the PTPN11 mutants, LS and NS residual ectopic parts of the nail in one toe, and hyponychia, share similar clinical features,1 and are sometimes associated due to ischaemic necrosis of the extremities following umbili- with malignant diseases such as leukaemia and neuroblas- cal catheterization (UC). In one finger and one toe the necro- toma.1–3,9 Thus, NS- and neoplasia-associated PTPN11 mutants sis involved the subcutaneous tissues more extensively, with and LS-associated PTPN11 mutants may have a common func- bone destruction and spontaneous amputation of the distal tional perturbation in vivo that could raise the predisposition to phalanges. malignant lesions. In addition, the novel PTPN11 mutation A 6-month-old girl was referred to us because of nail might have a specific tumorigenic effect. Further studies will abnormalities. Physical examination showed anonychia of six determine whether or not PTPN11 mutation-positive LS is a toes and one finger, with residual ectopic parts of the nail in risk factor for the development of MM. the first toe of the right foot, hyponychia of one toe of the left foot and amputation of the distal phalanges of a finger Department of Dermatology, M. SEISHIMA and a toe (Figs 1, 2). Ogaki Municipal Hospital, Ogaki 503-8502, Japan Y. MIZUTANI Personal history revealed that the condition was a sequela *Department of Endocrinology and Metabolism, Y. SHIBUYA of an ischaemic necrosis of the extremities that occurred National Research Institute for Child Health and C. ARAKAWA immediately after birth during admission for prematurity and Development, Tokyo 157-8535, Japan R. YOSHIDA* respiratory distress to the Neonatal Intensive Care Unit of our E-mail: [email protected]. T. OGATA*

References

1 Gelb BD, Tartaglia M. Noonan syndrome and related disorders: dys- regulated RAS-mitogen activated protein kinase signal transduction. Hum Mol Genet 2006; 15:R220–6. 2 Uc¸arc C, C¸aly´pkan U¨ , Martini S et al. Acute myelomonocytic leuke- mia in a boy with LEOPARD syndrome (PTPN11 gene mutation positive). J Pediatr Hematol Oncol 2006; 28:123–5. 3 Keren B, Hadchouel A, Saba S et al. PTPN11 mutations in patients with LEOPARD syndrome: a French multicentric experience. J Med Genet 2004; 41:e117. 4 Miller AJ, Mihm MC Jr. Melanoma. N Engl J Med 2006; 355:51– 65. 5 Davies H, Bignell GR, Cox C et al. Mutations of the BRAF gene in human cancer. Nature 2002; 417:949–54. 6 Clark WH Jr, Elder DE, Guerry D IV et al. A study of tumor progres- Fig 1. Anonychia of six toes with residual ectopic parts of the nail in sion: the precursor lesions of superficial spreading and nodular mel- the first toe of the right foot and hyponychia of the fourth toe of the anoma. Hum Pathol 1984; 15:1147–65. left foot.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1300 Correspondence

Anonychia may be classified into congenital and acquired, transient and permanent forms. Congenital anonychia can involve all, several, few or even one nail. It may be a sign of many heterogeneous diseases or may be isolated and nonsyndromic, sometimes associated with other malformations, most frequently of the underlying bone.1–3 Acquired anonychia is divided into post-traumatic, postin- fectious, postinflammatory or related to hereditary and acquired blistering disorders. It may be classified into early- onset forms, occurring soon after birth, such as in epidermol- ysis bullosa, and late-onset forms, occurring later in life, such as in inflammatory diseases, infections and after traumas. It may involve only one or a few fingers or toes, more often when it is post-traumatic, or it may involve many or even all the fingers and toes. Anonychia is permanent when complete destruction of the nail matrix occurs. Hyponychia is defined as the permanent presence of part of a nail or the presence of a rudimentary nail. In our patient, anonychia and hyponychia had an early onset and were related to an ischaemic injury. Ischaemia and gangrene of the extremities in newborns may be provoked by many different causes, such as umbilical, radial or ulnar artery cannulation, antiphospholipid syndrome, infections, vascular malformations and vasculitis.3–5 The cases due to artery catheterization are easily diagnosed because they occur immediately after the procedure is started, as in our case, in which signs of ischaemia occurred immediately Fig 2. Close-up magnification of the residual ectopic parts of the nail after UC. in the first toe of the right foot. UC can present many serious complications, even when per- formed by experienced neonatologists.3,6–8 The most important cutaneous complication of UC is skin necrosis, caused by vascu- hospital. The baby underwent positioning of an umbilical lar occlusion due to thrombosis, embolism, vasospasm (with catheter and, almost immediately, a widespread vasospasm secondary thrombosis), aberrant catheter placement, direct vas- was observed with blanching of the abdomen and both arms cular damage by the catheter tip or by the injection of hyper- and legs, in particular at the extremities. This clinical picture tonic solutions.3,6 Different areas, typically the buttocks and the was probably due to the insertion of the catheter in the lower extremities, may be affected, depending on which vessel umbilical artery. The catheter was immediately removed, with was involved, both unilaterally or bilaterally. rapid improvement of the cutaneous signs. However, altered Blanching, cyanosis, vesicle or bullae formation and ery- perfusion and cyanosis persisted at the distal phalanges of all thema are important signs that should alert physicians because the toes of both feet and of the second and third fingers of they may precede skin necrosis in patients with an UC.3,6 In the right hand. For this reason nitroglycerin-releasing plasters such cases immediate removal of the catheter is mandatory. If were applied locally. Three days later the plasters were ischaemia persists, intravenous anticoagulants or topical nitro- replaced by a nitroglycerin cream, twice daily, and an antibi- glycerin may be used and, if gangrene develops, spontaneous otic cream containing fusidic acid, twice daily. After 1 week, amputation is desirable.3,9 Only in a few cases is skin grafting the areas of necrosis became more evident and the nitroglyc- or surgical amputation required.6,10 erin cream was stopped and a collagenase-based cream was To our knowledge, this is the first report in the English lit- started, twice daily. erature of anonychia and hyponychia following ischaemic Spontaneous amputation of the distal phalanges of the third necrosis of the extremities after UC. finger of the right hand and of the first toe of the right foot occurred and was confirmed by an X-ray, which did not show Department of Specialist and Experimental I. NERI any other abnormalities of the fingers or toes. Complete heal- Clinical Medicine, Division of Dermatology, F. SAVOIA ing of the cutaneous lesions was achieved 45 days later but S. Orsola-Malpighi Hospital, University of Bologna, F. GIACOMINI anonychia of seven toes, residual ectopic parts of the nail in via Massarenti 1 CAP 40138, A. PATRIZI the first toe of the right foot and hyponychia of the fourth toe Bologna, Italy of the left foot remained as a sequela of the ischaemic necrosis. E-mail: [email protected]

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Correspondence 1301

References

1 Rigopoulos D, Petropoulou H, Nikopoulou M et al. Total congenital anonychia in two children of the same family. J Eur Acad Dermatol Venereol 2006; 20:894–6. 2 Silverman R, Baran R. Nail and appendageal abnormalities. In: Pedi- atric Dermatology (Schachner LA, Hansen RC, eds), 3rd edn. Philadel- phia: Mosby, 2003; 572–3. 3 Esterly NB. Iatrogenic and traumatic injuries. In: Textbook of Neonatal Dermatology (Eichenfield LF, Frieden IJ, Esterly NB, eds). Philadel- phia: WB Saunders Company, 2001; 108. 4 Green JA, Tonkin MA. Ischaemia of the hand in infants following radial or ulnar artery catheterisation. Hand Surg 1999; 4:151–7. 5 Wollina U, Verma SB. Acute digital gangrene in a newborn. Arch Dermatol 2007; 143:121–2. 6 Cutler VE, Stretcher GS. Cutaneous complications of central umbili- Fig 1. Anatomy of the medial canthus. cal artery catheterization. Arch Dermatol 1977; 113:61–3. 7 Grupo de Hospitales Castrillo. Prospective evaluation of umbilical catheters in newborn infants. The Castrillo Hospital Group. An Esp Pediatr 2000; 53:470–8. 8 Green C, Yohannan MD. Umbilical arterial and venous catheters: placement, use and complications. Neonatal Netw 1998; 17:23–8. 9 Baserga MC, Puri A, Sola A. The use of topical nitroglycerin oint- ment to treat peripheral tissue ischemia secondary to arterial line complications in neonates. J Perinatol 2002; 22:416–19. 10 Letts M, Blastorah B, al-Azzam S. Neonatal gangrene of the extremities. J Pediatr Orthop 1997; 17:397–401.

Key words: anonychia, hyponychia, ischaemic necrosis, umbilical catheterization

Conflicts of interest: none declared.

Basal cell carcinomas of the inner canthus: Fig 2. Topographical division of the medial canthus into three zones. incidence of incomplete excision according to topographical localization of tumours on subregions, justifies a topographical classification of this DOI: 10.1111/j.1365-2133.2007.08221.x area. This allows us to identify the zones that are statistically more likely to host an incomplete neoplasm removal, for a SIR, The inner canthus consists of the lateral nasal wall, the correct therapeutic approach. cutaneous area overlying the medial canthal tendon, the lacri- We proposed an original topographical classification of the mal caruncle, the medial third of superior and inferior eyelids inner canthus region and identified three zones (Fig. 2): zone and the puncta (Fig. 1). The medial canthus is a challenging I corresponds to the lateral nasal wall and overlying medial area to treat because of the functionally important lacrimal canthal tendon; zone II corresponds to the lacrimal carun- drainage and eyelid stabilizing apparatus (medial canthal ten- cle ± zone I; zone III corresponds to the medial upper or don), the relatively unobstructed access down the medial wall medial lower lid including the lacrimal punctum ± of the orbit (leading to potentially sight-threatening disease) zone II ± zone I. and the natural concavity of the region. In a recent large series We assessed the pathological reports of 153 primary nodu- of periocular tumours from an Australian tertiary referral cen- lar BCCs excised by the senior author between 1994 and tre, 92 of 417 (22%) basal cell carcinomas (BCCs) involved 2002. Margins were evaluated on a permanent specimen. Fro- the medial canthus.1 BCC is the most common malignant zen sections, although routinely utilized in case of surgeon’s tumour in the medial canthal area.2 Such neoplasms can dif- doubt on completeness of excision, were not considered for fuse towards conjunctival and orbital planes and show a high the study. The tumour margin aimed for macroscopically1 was tendency to recur.1 Local aggressiveness and risk of recur- 0Æ5 cm in zone I, 0Æ3 cm in zone II and if possible 0Æ3cmin rence3–5 depend on primary localization, in addition to histo- zone III. The surgical reconstruction was performed by means logical subtype.6,7 of direct closure, skin grafts and local or distant flaps. The The observation that surgical excision in the canthal area assistance of an oculoplastic surgeon was sought when the appears incomplete with a significantly higher frequency based excision involved the canalicular region.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1302 Correspondence

Each case was assigned to a specific topographical class, It appears that zones I and II have similar characteristics in according to the above-mentioned three-zone classification. terms of completeness of excision, whereas zone II resembles For each tumour the completeness or incompleteness of surgi- zone III regarding reconstructive options. In a tumour which cal removal was taken into account, as well as the method of is theoretically entirely curable, our data suggest that BCCs in reconstruction. zones I–III have an incomplete excision rate of 29–60%. With Mean age of the patients was 69 years; with 58% men and the knowledge that only between 35 and 67%8–10 of incom- 42% women. With regard to localization, 36% of tumours pletely excised tumours recur, our patients need to be made were in zone I, 41% in zone II and 23% in zone III. aware of the relative risks, costs and benefits of histologically The surgical excision achieved complete clearance in 71% controlled tumour margins in the medial canthal area. of cases for either zone I or II, and in 40% of patients in zone III. Acknowledgments The methods of reconstruction utilized were as follows: in zone I the most used reconstructive techniques were The authors are very grateful to Dr Nicola Kefalas for his assis- direct closure (57%) and local flaps (25%); in zone II skin tance with the drawings. Filippo Boriani had full access to all graft (53%) and local flaps (37%) were the preferred the data in the study and takes responsibility for the integrity options; in zone III skin grafts and local flaps (43% and of the data and the accuracy of the data analysis. 29%, respectively) were by far the most utilized reconstruc- tive methods. Director: Prof. Giovanni Bocchiotti. F. BORIANI Consensus is that the medial canthus is a problematic area Department of Plastic Surgery, F. MARCONI* because of difficulty and low probability of complete exci- Turin University, Turin, Italy sions,1,2 unless Mohs’ micrographic surgery is used. Neverthe- *Villa Laura Hospital, Bologna, Italy less this technique is demanding of time, effort and money E-mail: fi[email protected]; fi[email protected] and not always available. With traditional surgery, incidence of incomplete excisions and recurrences is reported in the lit- References erature as very high.1 Nemet et al.1 reported an incomplete periocular BCC excision rate of 106 out of 417 (25%). Medial 1 Nemet AY, Deckel Y, Martin PA et al. Management of periocular canthal lesions did particularly badly with incomplete excision basal and squamous cell carcinoma: a series of 485 cases. Am J Oph- rates of 25 out of 77 (32%) for nodular and 8 out of 15 thalmol 2006; 142:293–7. 2 Aliseda D, Vazquez J, Munuera JM. Medial canthus tumor surgery: (53%) for morphoeiform BCCs. Choosing the right recon- a prospective study of microscopically controlled excision. Eur J structive option is also difficult in this morphologically com- Ophthalmol 1997; 7:216–22. 1 plex area. 3 Spraul CW, Ahr WM, Lang GK. Clinical and histologic features of The complexity of the canthal area increases in a medial– 141 primary basal cell carcinomas of the periocular region and lateral direction, as eye-related structures get closer. In particu- their rate of recurrence after surgical excision [in German]. Klin lar, what renders the excision and the consequent repair more Monatsbl Augenheilkd 2000; 4:207–14. complex is the presence of lacrimal structures (puncta and 4 Silverman MK, Kopf AW, Grim CM et al. Recurrence rates of trea- ted basal cell carcinomas. Part 1: overview. Dermatol Surg Oncol 1991; drainage system). For this reason it is not precise to consider 17:713–18. the medial canthus as a single homogeneous area, with no 5 Sigurdsson H, Agnarsson BA. Basal cell carcinoma of the eyelid. topographical differences in terms of oncological risk and Risk of recurrence according to adequacy of surgical margins. Acta reconstructive complexity. The medial area contains anatomi- Ophthalmol Scand 1998; 76:477–80. cal structures of which the surgical sacrifice involves relevant 6 Sloane JP. The value of typing basal cell carcinomas in predicting functional and morphological deficits, even after complex recurrence after surgical excision. Br J Dermatol 1977; 96:127– reconstructive attempts. 32. 7 Robinson JK, Fisher SG. Recurrent basal cell carcinoma after The proposed topographical distinction is meant to orient incomplete resection. Arch Dermatol 2000; 136:1318–24. and guide surgeons, especially those who are less experienced 8 Shanoff LB Spira M, Hardy SB. Basal cell carcinoma: a statistical in periorbital oncology and surgery. These colleagues can thus approach to rational management. Ophthalmol Plast Reconstr Surg 1967; assess such neoplasms according to a scale of gravity, and con- 39:619–24. sequently ask for referral to centres of higher specialization or 9 Sarma DP, Griffing CC, Weilbaecher TG. Observations on the proceed with Mohs’ surgery, frozen sections or simply tradi- inadequately excised basal cell carcinomas. J Surg Oncol 1984; 25: tional surgery. 79–80. 10 Hauben DJ, Zirkin H, Mahler D et al. The biologic behavior of basal The proposed topographical classification can also guide the cell carcinoma: analysis of recurrence in excised basal cell carci- reconstructive phase, which may reasonably change according noma: part II. Plast Reconstr Surg 1982; 69:110–16. to the zone. Zone I is in fact often repaired by direct closure, which is less frequently feasible in the other zones where less Key words: basal cell carcinoma, eyelid, medial canthus, periocular area, periorbital surgery, reconstructive surgery skin is available and the defect is usually morphologically more complex. Conflicts of interest: none declared.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 News and Notices 1303

News and Notices Registration: Fondation Rene´ Touraine, Hoˆpital St Louis, Pavillon Bazin, 1, avenue Claude Vellefaux – F-75475 – PARIS DOI: 10.1111/j.1365-2133.2007.08348.x cedex 10 – (France) Te´l.: 33-1-53 72 20 60; Fax: 33-1-53 72 20 61; e-mail: International Board Certification in Dermatopathology [email protected] http://www.icdermpath.org *The program of the scientific meeting. The European Union of Medical Specialists (Union Europe´- enne des Me´decins Spe´cialistes¼UEMS)/Section of Dermato- Venereology and Section of Pathology will organize under the auspices of the International Committee for Dermatopathology the 5th International Board Certifying Examination in Dermatopathology (Diploma in Dermatopathology) on The BSDS Annual Surgery Workshop, Monday 31st March December 8, 2007, in Frankfurt/Main, Germany. 2008–Thursday 3rd April 2008, At the University of Shef- Further details about this examination, including eligibility field criteria, application form and contact informations are posted The BSDS Annual Surgery Workshop is now in its 25th year. at the website http://www.icdermpath.org. The course uses lectures and hands-on practical sessions to Address for correspondence: teach basic/intermediate Dermatological surgical skills and Lorenzo Cerroni, M.D. techniques. The workshop has a large number of expert Department of Dermatology faculty surgeons on hand to teach delegates directly. Medical University of Graz The course is always fully subscribed and early application is Auenbruggerplatz 8 advised. A-8036 Graz/Austria The course organisers are Drs Vindy Ghura and Graham Colver Ph.: +43-316-385-2423 and further details can be found on the BSDS website Fax: +43-316-385-2466 www.bsds.org.uk or via Dr Ghura c/o [email protected] e-mail: [email protected]

88th Annual Meeting of the British Association of Derma- tologists 1st–4th July 2008, Liverpool Fondation Rene Touraine, Pour la Recherche en Dermato- The 88th Annual Meeting of the British Association of Derma- logie tologists will be held at the Arena & Convention Centre *Fellowships 2007: Four fellowships up to €4500 (for (ACC), Liverpool, 1st–4th July 2008, organised by Dr David short exchange periods) and one up to €18000 (for long Shuttleworth, BAD Clinical Vice-President. exchange period) are given towards supporting a period Abstracts of papers and posters should be submitted for consid- spent in a research laboratory or clinical department of eration by the Scientific Committee. Original communications a different country, in order to promote international will be allotted 15 minutes, which must include time for dis- collaboration. cussion. Clinicopathological cases will be allotted 7 minutes Application forms, for either type of grant, may be requested for presentation, again allowing time for discussion. from the office of the Foundation at the following e-mail Online submission will be the only method of abstract sub- address: [email protected] or at the fol- mission available. Full instructions and the submission form lowing address: Fondation Rene´ Touraine, Hoˆpital St Louis, can be accessed via the BAD website http://www.bad.org.uk/ Pavillon Bazin, 1, avenue Claude Vellefaux – F-75475 – PARIS healthcare/annual_meeting/ cedex 10 – (France) The closing date for the receipt of abstracts is Monday 7th Deadline for reception of applications: October 1, 2007 January 2008 and the deadline will be adhered to strictly. Te´l.: 33-1-53 72 20 60; Fax: 33-1-53 72 20 61 Any abstracts received after this date will not be considered. *Scientific Meeting 2007: Friday, November 16, 2007, The deadline for abstract submissions to any of the special Mus´ee des Moulages, Hˆopital St Louis, 1 avenue Claude interest group meetings will be Monday 4th February 2008. Vellefaux, 75010 Paris. Poster submissions on the following topics are strongly Subject: ‘‘Angio- and Lympangio-Genesis in Skin Physiol- encouraged: audit, medical education, and service delivery. ogy and Disease’’ Conference & Event Services, British Association of Dermatolo- Invited speakers: M. Dewerchin, K. Alitalko, M. Detmar, gists, 4 Fitzroy Square, London, W1T 5HQ, UK or e-mail: C. Boshoff, D.M. McDonald, T. Kupper, N. Gale, M. Vikkula [email protected]

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 1304 Corrigenda

Corrigenda

DOI: 10.1111/j.1365-2133.2007.8351.x

CD8+ poikilodermatous mycosis fungoides with a nonaggressive clinical behaviour and a good response to psoralen plus ultraviolet A treatment. Br J Dermatol 2007, Ada and Gu¨ lec¸ In the above mentioned article1 there was an error in the author listing. A. Tu¨lin Gu¨lec¸ should have appeared as AT Gu¨lec¸.

The authors apologise for this error.

References

1 Ada S, Tu¨lin Gu¨lec¸A.CD8+ poikilodermatous mycosis fungoides with a nonaggressive clinical behaviour and a good response to psoralen plus ultraviolet A treatment. Br J Dermatol 2007; 157:1064–66.

DOI: 10.1111/j.1365-2133.2007.8352.x

Lymphocytic infiltration (Jessner–Kanof): lupus erythematosus tumidus or a manifestation of borreliosis? Br J Dermatol 2007, Kaatz et al. In the above mentioned article,1 the description of the used immunohistochemical antibody – ‘A polyclonal borrelia Treponema pallidum antibody (Biocare Medical, Concord, CA, U.S.A.) was used at a dilution of 1:150 and with an autoclave pre-treatment for 20 min in citrate buffer’ was incorrect. The correct text should be: ‘We used a polyclonal rabbit antibody (Acris BP1002, derived from immunization with whole cell B. burgdorferi preparations strain B31 ATCC#35210 reacting with 83 kD, 41 kD ⁄flagellin, 32 kD ⁄OspB and 31 kD ⁄OspA antigens and their fragments in Western blots, with cross reaction to Treponema pallidum, B. hermesii and B. parkeri) at a dilution of 1:2000 after autoclave antigen retrieval for 30 min at 37 C.’ The authors apologises for this error.

References

1 Kaatz M, Zelger B, Norgauer J, Ziemer M. Lymphocytic infiltration (Jessner–Kanof): lupus erythematosus tumidus or a manifestation of borreliosis. Br J Dermatol 2007; 157:403–5.

Erratum

DOI: 10.1111/j.1365-2133.2007.8353.x

Human papillomavirus type 26 infection causing multiple invasive squamous cell carcinomas of the fingernails in an AIDS patient under highly active antiretroviral therapy. Br J Dermatol 2007, Handisurya et al. The above mentioned article1 was mistakenly published as a Case Report. It should in fact have been published as a Concise Communication.

The Journal apologises for this error.

References

1 Handisurya A, Rieger A, Bankier A, Koller A, Salat A, Stingl G, Kirnbauer R. Human papillomavirus type 26 infection causing multiple invasive squamous cell carcinomas of the fingernails in an AIDS patient under highly active antiretroviral therapy. Br J Dermatol 2007; 157:788–94.

2007 The Authors Journal Compilation 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1267–1304 Author index

ABE, R. see YAOSAKA, M. reconstitution inflammatory syndrome in a patient with acquired ABE, Y. see YAOSAKA, M. immune deficiency syndrome, 1032 ABENI, D. see SAMPOGNA, F. AOCHI, S., NAKANISHI, G., SUZUKI, N., SETSU, N., SUZUKI, D., ABERER, W. see GRIMS, R.H. AYA, K. & IWATSUKI, K. A novel homozygous mutation of the ABIDA, O. see MEJRI, K. EVER1 ⁄TMC6 gene in a Japanese patient with epidermodysplasia ACKERMANN, H. see BOEHNCKE, S. verruciformis, 1265 ACQUAVIVA, V. see ANTINORI, S. APARICIO ESPAN˜ OL, G. see HERAS MULERO, C. ADA, S. & TU¨ LIN GU¨ LE¸, A. CD8+ poikilodermatous mycosis ARABATZIS, M., BRUIJNESTEIJN VAN COPPENRAET, L.E.S., KUIJPER, fungoides with a nonaggressive clinical behaviour and a good E.J., DE HOOG, G.S., LAVRIJSEN, A.P.M., TEMPLETON, K., response to psoralen plus ultraviolet A treatment, 1064 VAN DER RAAIJ-HELMER, E.M.H., VELEGRAKI, A., GRA¨SER, AFFLECK, A.G. & VARMA, S. A case of do-it-yourself Mohs’ surgery Y. & SUMMERBELL, R.C. Diagnosis of common dermatophyte using bloodroot obtained from the internet, 1078 infections by a novel multiplex real-time polymerase chain AFFLECK, P. see STRAUSS, R.M. reaction detection ⁄identification scheme, 681 AGAPITOS, E. see MOUZOPOULOS, G. ARACTINGI, S. see RE´GNIER, S. AGIUS, R. see ATHAVALE, P. ARAKAWA, C. see SEISHIMA, M. AGIUS, R. see TURNER, S. ARAKI, E. see NISHIKAWA, M. AGLIANO` , A.M. see GRADILONE, A. ARBAB, E. see GRIMS, R.H. AGNER, T. see LERBAEK, A. ARDIGO` ,M.see VASSALLO, C. AGNER, T. see NOIESEN, E. ARGENZIANO, G., ZALAUDEK, I., FERRARA, G., HOFMANN- AHMAD, K. & ROGERS, S. Development of Crohn disease in a patient WELLENHOF, R. & SOYER, H.P. Proposal of a new classification on etanercept for psoriasis, 396 system for melanocytic naevi, 217 AHMAD, W. see TARIQ, M. ARITA, K., WESSAGOWIT, V., INAMADAR, A.C., PALIT, A., FASSIHI, AHMED, N. see ESCUDIER, M. H., LAI-CHEONG, J.E., POURREYRON, C., SOUTH, A.P. & AKASAKA, T. see WATABE, D. MCGRATH, J.A. Unusual molecular findings in Kindler syndrome, AKITA, N. see NAKANO, H. 1252 AKIYAMA, M. see HOSHINA, D. ARORA, A. see MADKAN, V.K. AKMAN, A. see ALPSOY, E. AROSTEGUI, J.I. see COTO-SEGURA, P. ALESSANDRINO, P.E. see VASSALLO, C. ASAGOE, K. see YAMADA, A. ALOMAR, A., BICHEL, J. & MCRAE, S. Vehicle-controlled, ASAHINA, A., FUJITA, H., FUKUDA, S., KAI, H., YAMAMOTO, M., randomized, double-blind study to assess safety and efficacy of HATTORI, N. & MORI, T. Extensive skin pigmentation caused by imiquimod 5% cream applied once daily 3 days per week in one deposits of metallic particles following total elbow arthroplasty: or two courses of treatment of actinic keratoses on the head, metallosis or not?, 1074 133 ASSAF, C. see GNIADECKI, R. ALOMAR, A. see GARCIA-NAVARRO, X. ATHAVALE, P., SHUM, K.W., CHEN, Y., AGIUS, R., CHERRY, N. & ALPSOY, E., DONMEZ, L., ONDER, M., GUNASTI, S., USTA, A., GAWKRODGER, D.J. ON BEHALF OF EPIDERM. Occupational KARINCAOGLU, Y., KANDI, B., BUYUKKARA, S., KESEROGLU, dermatitis related to chromium and cobalt: experience of O., UZUN, S., TURSEN, U., SEYHAN, M. & AKMAN, A. Clinical dermatologists (EPIDERM) and occupational physicians (OPRA) in features and natural course of Behc¸et’s disease in 661 cases: a the U.K. over an 11-year period (1993–2004), 518 multicentre study, 901 ATHERTON, D. see PONNAMPALAM, J. ALTMEYER, P. see BECHARA, F.G. AURICCHIO, L. see ESPOSITO, G. ALTMEYER, P. see KREUTER, A. AUST, M.C., SPIES, M., KALL, S., GOHRITZ, A., BOORBOOR, P., AMANTEA, A. see CAPITANIO, B. KOLOKYTHAS, P. & VOGT, P.M. Lipomas after blunt soft tissue and the IDI Multipurpose Psoriasis Research on Vital Experiences trauma: are they real? Analysis of 31 cases, 92 (IMPROVE) investigators see SAMPOGNA, F. AYA, K. see AOCHI, S. ANDERS, M. see TORRES, A. AZOULAY, L., ORAICHI, D. & BE´RARD, A. Isotretinoin therapy and ANGUS, J., ROBERTS, C., KULKARNI, K., LEACH, I. & MURPHY, R. the incidence of acne relapse: a nested case–control study, 1240 Usefulness of the QuantiFERON test in the confirmation of latent tuberculosis in association with erythema induratum, 1293 BABILAS, P., KNOBLER, R., HUMMEL, S., GOTTSCHALLER, C., ANHALT, G.J. see CUMMINS, D.L. MAISCH, T., KOLLER, M., LANDTHALER, M. & SZEIMIES, R-M. ANSTEY, A.V. Photodermatology. Henry W. Lim, Herbert Ho¨nigsmann & Variable pulsed light is less painful than light-emitting diodes for John L.M. Hawk (editors). New York: Informa Healthcare, 2007, topical photodynamic therapy of actinic keratosis: a prospective 429 randomized controlled trial, 111 ANTAL, A., ZELGER, B., REIFENBERGER, J., NIEHUES, T., FEYEN, O., BACHELEZ, H. see VIGUIER, M. MEGAHED, M., RUZICKA, T. & HOMEY, B. Multiple eruptive BADENHOOP, K. see BOEHNCKE, S. myxoid dermatofibromas: report of first case and review of BAGOT, M. see GNIADECKI, R. literature, 382 BAJAJ, V. & LANGTRY, J.A.A. Use of oral glycopyrronium bromide in ANTINORI, S., LONGHI, E., BESTETTI, G., PIOLINI, R., ACQUAVIVA, hyperhidrosis, 118 V., FOSCHI, A., TROVATI, S., PARRAVICINI, C., CORBELLINO, M. BALME, B. see BARBERIO, E. & MERONI, L. Post-kala-azar dermal leishmaniasis as an immune BALME, B. see DALLE, S.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1305 1306 Author index

BALME, B. see DEBARBIEUX, S. BENOIT, S., SEITZ, C.S., HAMM, H., VETTER-KAUCZOK, C.S. & BALME, B. see PHAN, A. BRO¨ CKER, E-B. Circumscribed palmar hypokeratosis: partial BAMFORD, M. see HELBLING, I. remission by photodynamic therapy, 804 BANG, R.H. see ZLOTOFF, B.J. BE´RARD, A. see AZOULAY, L. BANKIER, A. see HANDISURYA, A. BERGER, V.W. & GEE, E. On confusing prima facie validity with true BANNO, T. see KAWACHI, Y. validity, 425 BANSAL, S., GOEL, A., SARDANA, K., KUMAR, V. & KHURANA, N. BERGMANN, C. see SEITZ, C.S. Postkala-azar dermal leishmaniasis coexisting with borderline BERK, D.R. & BAYLISS, S.J. Sock-line bands in infancy, 1063 tuberculoid leprosy, 811 DE BERKER, D. see BARAN, R. BANSAL, S. see GROVER, C. BERNARDINI, M.L. see CAMPANATI, A. BAO, S. see FANG, Y. BERROETA, L., CLARK, C., DAWE, R.S., IBBOTSON, S.H. & BARAN, R., SIGURGEIRSSON, B., DE BERKER, D., KAUFMANN, R., FLEMING, C.J. A randomized study of minimal curettage followed LECHA, M., FAERGEMANN, J., KERROUCHE, N. & SIDOU, F. A by topical photodynamic therapy compared with surgical excision multicentre, randomized, controlled study of the efficacy, safety for low-risk nodular basal cell carcinoma, 401 and cost-effectiveness of a combination therapy with amorolfine BERROETA, L., MAN, I., GOUDIE, D.R., WHATLEY, S.D., ELDER, nail lacquer and oral terbinafine compared with oral terbinafine G.H. & IBBOTSON, S.H. Late presentation of erythropoietic alone for the treatment of onychomycosis with matrix protoporphyria: case report and genetic analysis of family involvement, 149 members, 1030 BARBAROT, S., NICOL, C., VOLTEAU, C., LE FORESTIER, D., BES, M. see DURUPT, F. N’GUYEN, J.M., MANSAT, E., WOLKENSTEIN, P. & STALDER, J.F. BESCHMANN, H. see BOEHNCKE, S. Cutaneous lesions in neurofibromatosis 1: confused terminology, BESTETTI, G. see ANTINORI, S. 183 BEYLOT-BARRY, M. see STOKKERMANS-DUBOIS, J. BARBAROT, S. see VOURC’H, M. BE´ZIEAU, S. see KILIC, S.S. BARBERIO, E., THOMAS, L., SKOWRON, F., BALME, B. & DALLE, S. BIAGINI, G. see TUCCI, M.G. Transformed mycosis fungoides: clinicopathological features and BICHEL, J. see ALOMAR, A. outcome, 284 BIEBER, T. see WENZEL, J. BARBIER, N. see MURRELL, D.F. BILKER, W. see MARGOLIS, D.J. BARBUZZA, O. see VACCARO, M. BIRMACHU, W. see TORRES, A. BARKHAM, T. see TAN, H-H. BISGAARD, H. see LERBAEK, A. BARNHILL, R.L. see LUGASSY, C. BISSET, Y. see FINNEN, M.J. BARTKOWIAK, J. see ROBAK, E. BISWAS, A., COOPER, J. & LATIFAJ, B. Metastatic calcinosis cutis BARTRALOT SOLER, R. see HERAS MULERO, C. presenting as bilateral vulval cysts, 622 BASKETTER, D.A. see WHITE, J.M.L. BIZIKOVA, P. see OLIVRY, T. BASSAS FREIXAS, P. see HERAS MULERO, C. BJO¨ RK, J. see EKQVIST, S. BAUER, J.W. see SADLER, E. BLACK, M.M. see ESCUDIER, M. BAUZA´, A., DEL POZO, L.J., ESCALAS, J. & MESTRE, F. Radiation BLANC-AMRANE, V. see DEL GIUDICE, P. recall dermatitis in a patient affected with pheochromocytoma BLAZQUEZ, S. see ZABALLOS, P. after treatment with lanreotide, 1061 BLOMGREN, B. see JOHANNESSON, U. BAYLISS, S.J. see BERK, D.R. BODET CASTILLO, D. see HERAS MULERO, C. BE´ATRIX, O. see DEBARBIEUX, S. BOEHNCKE, S., THACI, D., BESCHMANN, H., LUDWIG, R.J., BECHARA, F.G., SAND, M., TOMI, N.S., ALTMEYER, P. & ACKERMANN, H., BADENHOOP, K. & BOEHNCKE, W-H. Psoriasis HOFFMANN, K. Repeat liposuction-curettage treatment of axillary patients show signs of insulin resistance, 1249 hyperhidrosis is safe and effective, 739 BOEHNCKE, W-H. see BOEHNCKE, S. BECK, M. see ORTEU, C.H. BO¨ ER, A. see RO¨ GLIN, J. BECK, M.H. see TURNER, S. BOERSMA, I.H. see VERFAILLE, C.J. BEDINI, C., NASORRI, F., GIROLOMONI, G., DE PITA`,O.& BOGUNIEWICZ, M. see SPERGEL, J.M. CAVANI, A. Antitumour necrosis factor-a chimeric antibody BO¨ HM, M. see VOLZ, A. (infliximab) inhibits activation of skin-homing CD4+ and BOHM-STARKE, N. see JOHANNESSON, U. CD8+ T lymphocytes and impairs dendritic cell function, BOISSY, C. see DEL GIUDICE, P. 249 BOLLING, M.C., MEKKES, J.R., GOLDSCHMIDT, W.F.M., VAN BEDNAREK, A. see ROBAK, E. NOESEL, C.J.M., JONKMAN, M.F. & PAS, H.H. Acquired BEEK, L. see SERCU, S. palmoplantar keratoderma and immunobullous disease associated BEGON, E. see VIGUIER, M. with antibodies to desmocollin 3, 168 BELL, P.M. see SHAW, J. BOLTON, A. see TURNER, S. BELL, R.R., DUNSTAN, R.W. & KHAN, N.K. Skin wound healing in BONNET-DUQUENNOY, M. see SAUVAIGO, S. the SKH-1 female mouse following inducible nitric oxide synthase BONTE´,F.see SAUVAIGO, S. inhibition, 656 BOO, Y.C. see LEE, J. BELLOSILLO, B. see PARERA, E. BOOHER, S. see COWEN, E.W. BELOUSOVA, I.E., NIKONOVA, S.M., SIMA, R. & KAZAKOV, D.V. BOON, A.P. see STRAUSS, R.M. Granulomatous slack skin with clonal T-cell receptor-c gene BOORBOOR, P. see AUST, M.C. rearrangement in skin and lymph node, 405 BORALEVI, F. see MORICE-PICARD, F. BEN AYED, M. see MEJRI, K. BORDIGNON, V. see CAPITANIO, B. BENAZZO, M. see VASSALLO, C. BORGERS, M. see VERFAILLE, C.J.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1307

BORGIA, F. see VACCARO, M. CAPITANIO, B., SINAGRA, J.L., OTTAVIANI, M., BORDIGNON, V., BORIANI, F. & MARCONI, F. Basal cell carcinomas of the inner AMANTEA, A. & PICARDO, M. ‘Smoker’s acne’: a new clinical canthus: incidence of incomplete excision according to entity?, 1070 topographical localization of tumours, 1301 CAPOMAGI, C. see GOTERI, G. BORNSHIN, S. see HAYDOCK, S.F. CARDER, M. see TURNER, S. BORRONI, G. see VASSALLO, C. CARDONE, C. see MOURMOURAS, V. BORRONI, R.G. see ROSSI, A. CARIOLA, F. see ROSSI, A. BOSCHIERO, L. see GOMEZ LIRA, M. CARLESIMO, M. see ROSSI, A. BOTEY, A. see MORENO-ROMERO, J.A. CARR, R.A. see TAIBJEE, S.M. BOUABDALLAH, K. see STOKKERMANS-DUBOIS, J. CASTELIJNS, F.C.M. see STEIJLEN, P.M. BOURKE, J. see MALIK, M. CASTRO, L.G.M. see TAKAHASHI, M.D.F. BOUWES BAVINCK, J.N. see SIDOROFF, A. CAUMES, E. see DEL GIUDICE, P. BRAATHEN, L.R. see SIMON, D. CAVANI, A. see BEDINI, C. BRANCORSINI, D. see TUCCI, M.G. CELLA, D. see KRISHNAN, R. BRAUN, R.P., GAIDE, O., OLIVIERO, M., KOPF, A.W., FRENCH, L.E., CESTARI, T.F. see MARTIGNAGO, B.C.F. SAURAT, J-H. & RABINOVITZ, H.S. The significance of multiple CHALLACOMBE, S.J. see ESCUDIER, M. blue-grey dots (granularity) for the dermoscopic diagnosis of CHAN, I. see COX, G.A. melanoma, 907 CHAN, K.H.N., TANG, W.Y.M., LAM, W.Y. & LO, K.K. Eruptive vellus BRAUN, R.P. see PERRINAUD, A. hair cysts presenting as bluish-grey facial discoloration masquering BRAZZELLI, V. see VASSALLO, C. as naevus of Ota, 188 BREETVELD, M. see LE DUC, Q. CHAPMAN, A. see SINCLAIR, R. BRO¨ CKER, E-B. see BENOIT, S. CHARAKIDA, A., CHARAKIDA, M. & CHU, A.C. Double-blind, BRO¨ CKER, E-B. see WACHTER, T. randomized, placebo-controlled study of a lotion containing BROCKMEYER, N.H. see KREUTER, A. triethyl citrate and ethyl linoleate in the treatment of acne vulgaris, BROOKLYN, T.N., WILLIAMS, A.M., DUNNILL, M.G.S. & PROBERT, C.S. 569 T-cell receptor repertoire in pyoderma gangrenosum: evidence for CHARAKIDA, M. see CHARAKIDA, A. clonal expansions and trafficking, 960 CHARLTON, M.R. see OTLEY, C.C. BROSTRO¨ M, U. see BURACZEWSKA, I. CHEN, B. see SHAW, J. BROWN, S.J., TALKS, S.J., NEEDHAM, S.J. & TAYLOR, A.E.M. CHEN, G-S. see HU, S.C-S. Pseudoxanthoma elasticum: biopsy of clinically normal skin in the CHEN, Y. see ATHAVALE, P. investigation of patients with angioid streaks, 748 CHENG, H. see ZHU, K.J. BROWN, V.L. see STAVRAKOGLOU, A. CHENG, K.F. see HON, K.L.E. BROWNELL, I. & STROBER, B.E. Folate with methotrexate: big CHERRY, N. see ATHAVALE, P. benefit, questionable cost, 213 CHEVRANT-BRETON, J. see MEYER, N. BRU¨ CHER, J-J., FRANKE, I., ULRICH, J., GOLLNICK, H. & LEVERKUS, CHIMENTI, S. see COSTANZO, A. M. Giant genital variant of folliculosebaceous cystic hamartoma: CHIOU, C.F. see KRISHNAN, R.

successful management by CO2 laser and acitretin therapy, 833 CHIU, C-T. see YANG, C-H. BRUCKNER-TUDERMAN, L. see VOLZ, A. CHIU, H.C. see LIAO, Y.H. BRUIJNESTEIJN VAN COPPENRAET, L.E.S. see ARABATZIS, M. CHIU, H-C. see HONG, J-B. BRUZE, M. see EKQVIST, S. CHO, K.H. see KIM, H.J. BUCKLEY, D.A. see WHITE, J.M.L. CHOI, C. see YUN, S.J. BUCKLEY, D.A. Fragrance ingredient labelling in products on sale in CHOI, J.Y. see OH, S-W. the U.K., 295 CHOWDHURY, M.M.U. see SHAH, D. BULBULIAN, B.J. see TORRES, A. CHU, A.C. see CHARAKIDA, A. BUNKER, C.B. Male genital lichen sclerosus and tacrolimus, 1079 CHUNG, H.J. see GOO, B. BURACZEWSKA, I., BROSTRO¨ M, U. & LODE´N, M. Artificial CHUNG, K.Y. see GOO, B. reduction in transepidermal water loss improves skin barrier CHUNG, W.G. see GOO, B. function, 82 CIANCHINI, G., MASINI, C., LUPI, F., CORONA, R., DE PITA`,O.& BUTLER, P. see LLOYD, M.S. PUDDU, P. Severe persistent pemphigoid gestationis: long-term BUYUKKARA, S. see ALPSOY, E. remission with rituximab, 388 CIGNA, E. see GRADILONE, A. CABANE, J. see FARDET, L. CINN, Y.W. see KIM, H.J. CA´CERES MENDEZ, O.A. see PIEPENBRING, M. CIOCON, D.H. & KIMBALL, A.B. Psoriasis and psoriatic arthritis: CALISTA, D. Treatment of lentigo maligna with topical 1% cidofovir, separate or one and the same?, 850 421 CLARK, A. see LLOYD, M.S. CALLEJAS-RUBIO, J-L. see RIOS-FERNA´NDEZ, R. CLARK, C. see BERROETA, L. CALVIERI, S. see MURRELL, D.F. COEL, M. see VERFAILLE, C.J. CALVIERI, S. see ROSSI, A. COLLINS, P. see MOLONEY, F.J. CAMPANATI, A., GOTERI, G., SIMONETTI, O., GANZETTI, G., CONNICK, R.M. see HAYDOCK, S.F. GIULIODORI, K., STRAMAZZOTTI, D., MORICHETTI, D., CONSTANTINU, M. see ROBAK, E. BERNARDINI, M.L., MANNELLO, B., FABRIS, G. & OFFIDANI, A. CONTI, A. see GISONDI, P. CTACK ⁄CCL27 expression in psoriatic skin and its modification COOK-NORRIS, R.H. see MADKAN, V.K. after administration of etanercept, 1155 COOPER, J. see BISWAS, A.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1308 Author index

CORBELLINO, M. see ANTINORI, S. DIFFEY, B.L. & FARR, P.M. The challenge of follow-up in narrowband CORNELIS, F. see MEJRI, K. ultraviolet B phototherapy, 344 CORONA, R. see CIANCHINI, G. DISCEPOLI, G. see GOTERI, G. CORONA, S. see VASSALLO, C. DODSON, H.I. see WILSON, A.S. COSCI, E. see MOURMOURAS, V. DOLIANITIS, C. see TUXEN, A.J. COSTANZO, A., PAPOUTSAKI, M., MAZZOTTA, A. & CHIMENTI, S. DONMEZ, L. see ALPSOY, E. Consecutive use of different biological therapies in the treatment DOUGHTY, L. see GERAMI, P. of psoriasis, 394 DOUTRE, M.S. see STOKKERMANS-DUBOIS, J. COSTA-ROMERO, M. see COTO-SEGURA, P. DUBERTRET, L. see VIGUIER, M. COTO-SEGURA, P., MALLO-GARCIA, S., COSTA-ROMERO, M., DUBUCQUOI, S. see VERCAMBRE-DARRAS, S. AROSTEGUI, J.I., YAGUE, J., RAMOS-POLO, E. & SANTOS- DUFOUR, J. see MEYER, N. JUANES, J. A sporadic case of early-onset sarcoidosis resembling DUMMER, R. see GNIADECKI, R. Blau syndrome due to the recurrent R334W missense mutation on DUNANT, A. see SIDOROFF, A. the NOD2 gene, 1257 DUNN, M. see KRISHNAN, R. COUNILLON, E. see DEL GIUDICE, P. DUNNILL, M.G.S. see BROOKLYN, T.N. COUTTS, I. see STAVRAKOGLOU, A. DUNSTAN, R.W. see BELL, R.R. COWEN, E.W., LIU, C-W., STEINBERG, S.M., KANG, S., DUPIN, N. see RE´GNIER, S. VONDERHEID, E.C., KWAK, H.S., BOOHER, S., PETRICOIN, E.F., DURU, G. see DEBARBIEUX, S. LIOTTA, L.A., WHITELEY, G. & HWANG, S.T. Differentiation of DURUPT, F., MAYOR, L., BES, M., REVERDY, M-E., VANDENESCH, tumour-stage mycosis fungoides, psoriasis vulgaris and normal F., THOMAS, L. & ETIENNE, J. Prevalence of Staphylococcus aureus controls in a pilot study using serum proteomic analysis, 946 toxins and nasal carriage in furuncles and impetigo, 1161 COWPER, S.E. see MORENO-ROMERO, J.A. DUVIC, M. see GNIADECKI, R. COX, G.A., CHAN, I., LLOYD, J., WITHEROW, R.O. & LEONARD, DYER, M.J.S. see HELBLING, I. J.N. Urachal sinus presenting as periumbilical dermatitis, 419 DYTOC, M.T., KOSSINTSEVA, I. & TING, P.T. First case series on the CRIBIER, B. see LIPSKER, D. use of calcipotriol–betamethasone dipropionate for morphoea, 615 CRUMRINE, D. see DENDA, M. CSOMOR, J. see MARSCHALKO´ ,M. EDWARDS, B.S. see OTLEY, C.C. CUMBER, P. see SHAH, D. VAN EGMOND, S. see SINCLAIR, R. CUMMINS, D.L., ANHALT, G.J., MONAHAN, T. & MEYERLE, J.H. EHRLICH, R. see KHUMALO, N.P. Treatment of pyoderma gangrenosum with intravenous EISENDLE, K., GRABNER, T. & ZELGER, B. Morphoea: a manifestation immunoglobulin, 1235 of infection with Borrelia species?, 1189 CZARNECKI, D. see LY, L. EISENDLE, K., JASCHKE, W. & SEPP, N. Livedo racemosa and digital necrosis in a patient with primary seronegative antiphospholipid DALLE, S., MARROU, K., BALME, B. & THOMAS, L. Neonatal syndrome and fibromuscular dysplasia of peripheral arteries, 389 follicular mucinosis, 609 EKQVIST, S., SVEDMAN, C., MO¨ LLER, H., KEHLER, M., PRIPP, C.M., DALLE, S. see BARBERIO, E. BJO¨ RK, J., GRUVBERGER, B., HOLMSTRO¨ M, E., GUSTAVSSON, DALLE, S. see DEBARBIEUX, S. C.G. & BRUZE, M. High frequency of contact allergy to gold in DALLE, S. see PHAN, A. patients with endovascular coronary stents, 730 DALMAU, J. see GARCIA-NAVARRO, X. ELDER, G.H. see BERROETA, L. DAWE, R.S. see BERROETA, L. ELIAS, P.M. see DENDA, M. DE NISI, M.C. see MOURMOURAS, V. ELLIOTT, F. see STRAUSS, R.M. DE OLIVEIRA, W.R.P. see RADY, P.L. ENGEL, F. see LIPSKER, D. DE PITA`,O.see CIANCHINI, G. ENGLISH, J. see TSAO, H. DEBARBIEUX, S., DURU, G., DALLE, S., BE´ATRIX, O., BALME, B. & ENGLISH, J.S.C. see JOHNSTON, G.A. THOMAS, L. Sentinel lymph node biopsy in melanoma: a ENJOLRAS, O. see RE´GNIER, S. micromorphometric study relating to prognosis and completion ENK, A. see GHOLAM, P. lymph node dissection, 58 ENRI´QUEZ DE SALAMANCA, R. see ME´NDEZ, M. DEENI, Y.Y. see SMITH, G. EPISTOLATO, M.C. see MOURMOURAS, V. DEL GIUDICE, P., CAUMES, E., BOISSY, C., LEDUFF, F., DELAUNAY, EROS,} N. see MARSCHALKO´ ,M. P., BLANC-AMRANE, V., GOFF-LEVAN, S.L., MARTY, P., LE ESCALAS, J. see BAUZA´,A. FICHOUX, Y. & COUNILLON, E. An outbreak of creeping ESCUDIER, M., AHMED, N., SHIRLAW, P., SETTERFIELD, J., eruption in southern France, 824 TAPPUNI, A., BLACK, M.M. & CHALLACOMBE, S.J. A scoring DEL POZO, L.J. see BAUZA´,A. system for mucosal disease severity with special reference to oral DELANEY, S. see LACEY, N. lichen planus, 765 DELAUNAY, P. see DEL GIUDICE, P. ESPINO ESPINOZA, A.A. see PIEPENBRING, M. DEMETER, J. see MARSCHALKO´ ,M. ESPOSITO, G., TADINI, G., PAPARO, F., VIOLA, A., IENO, L., DENDA, M., TSUTSUMI, M., INOUE, K., CRUMRINE, D., FEINGOLD, PENNACCHIA, W., MESSINA, F., GIORDANO, L., PICCIRILLO, A. K.R. & ELIAS, P.M. Potassium channel openers accelerate epidermal & AURICCHIO, L. Transglutaminase 1 deficiency and corneocyte barrier recovery, 888 collapse: an indication for targeted molecular screening in DENDA, M. see TSUTSUMI, M. autosomal recessive congenital ichthyosis, 808 DE´SAKNAI, M. see MARSCHALKO´ ,M. ESTINES-CHARTIER, O. see FAGUER, S. DEVIRGILIIS, V. see ROSSI, A. ETIENNE, J. see DURUPT, F. DIFFEY, B. Sunbeds, beauty and melanoma, 215 EVANS, A. see MULLER, F.M.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1309

FABBI, M. see VASSALLO, C. FUJIMOTO, M. see KUWANO, Y. FABRIS, G. see CAMPANATI, A. FUJISAWA, Y. see KAWACHI, Y. FAERGEMANN, J. see BARAN, R. FUJITA, H. see ASAHINA, A. FAGOT, J-P. see SIDOROFF, A. FUKUDA, S. see ASAHINA, A. FAGUER, S., LAUNAY, F., YSEBAERT, L., MAILHOL, C., ESTINES- FULCO, T.O. see MOURA, D.F. CHARTIER, O., LAMANT, L. & PAUL, C. Acute cutaneous T-cell FUNG, K.P. see HON, K.L.E. lymphoma transformation during treatment with alemtuzumab, FURUE, M. see IMAFUKU, S. 841 FURUE, M. see KIDO, M. FAJARDY, I. see VERCAMBRE-DARRAS, S. FURUTA, J. see KAWACHI, Y. FANG, K. see ZUO, Y-G. FANG, Y., GONG, S-J., XU, Y-H., HAMBLY, B.D. & BAO, S. Impaired GAIDE, O. see BRAUN, R.P. cutaneous wound healing in granulocyte ⁄macrophage colony- GAIDE, O. see PERRINAUD, A. stimulating factor knockout mice, 458 GALLAGHER, P.R. see SPERGEL, J.M. FARDET, L., FLAHAULT, A., KETTANEH, A., TIEV, K.P., GE´NE´REAU, GALLARDO, F. see PARERA, E. T., TOLE´DANO, C., LEBBE´, C. & CABANE, J. Corticosteroid- GAMBICHLER, T. see KREUTER, A. induced clinical adverse events: frequency, risk factors and GANDINI, O. see GRADILONE, A. patient’s opinion, 142 GANZETTI, G. see CAMPANATI, A. FARR, P.M. see DIFFEY, B.L. GARCI´A-BRAVO, M. see ME´NDEZ, M. FARR, P.M. see SMITH, G. GARCIA-NAVARRO, X., PUIG, L., FERNA´NDEZ-FIGUERAS, M.T., FASSIHI, H. see ARITA, K. DALMAU, J., ROE, E. & ALOMAR, A. Bortezomib-associated FAVIER, A. see SAUVAIGO, S. cutaneous vasculitis, 799 FAZIOLI, F. see GOTERI, G. GARCI´A-PATOS, V. see HERAS MULERO, C. FAZZA, B. see MEJRI, K. GARRIDO-ASTRAY, M.C. see ME´NDEZ, M. FEINGOLD, K.R. see DENDA, M. GAWKRODGER, D.J. see ATHAVALE, P. FERGUSON, J. see KERR, A.C. GAZZANIGA, P. see GRADILONE, A. FERGUSON, J. see OLIVER, H. GEE, B.C. see HAGUE, J.S. FERGUSON, J. see SMITH, G. GEE, B.C. see TAIBJEE, S.M. FERNA´NDEZ-FIGUERAS, M.T. see GARCIA-NAVARRO, X. GEE, E. see BERGER, V.W. FERNA´NDEZ-PUGNAIRE, M. see RIOS-FERNA´NDEZ, R. VAN GEEL, M. see STEIJLEN, P.M. FERRARA, G. see ARGENZIANO, G. GE´NE´REAU, T. see FARDET, L. FERREIRA, H. see MOURA, D.F. GENTILE, M. see ROSSI, A. FESTA, C. see RADY, P.L. GERAMI, P., WALLING, H.W., LEWIS, J., DOUGHTY, L. & FEYEN, O. see ANTAL, A. SONTHEIMER, R.D. A systematic review of juvenile-onset clinically FIMIANI, M. see MOURMOURAS, V. amyopathic dermatomyositis, 637 FINCHER, E.F. see KOUBA, D.J. GHOLAM, P., HARTMANN, M. & ENK, A. Arndt–Gottron FINLAY, A.Y. see THOMAS, C.L. scleromyxoedema: successful therapy with intravenous FINNEN, M.J., HENNESSY, A., MCLEAN, S., BISSET, Y., MITCHELL, immunoglobulins, 1058 R., MEGSON, I.L. & WELLER, R. Topical application of acidified GIACCHETTI, A. see GOTERI, G. nitrite to the nail renders it antifungal and causes nitrosation of GIACCHETTI, A. see TUCCI, M.G. cysteine groups in the nail plate, 494 GIACOMINI, F. see NERI, I. FISCHER, L.A., JOHANSEN, J.D. & MENNE´, T. Nickel allergy: GIANNETTI, A. see GISONDI, P. relationship between patch test and repeated open application test GIBBS, S. see LE DUC, Q. thresholds, 723 GILBERT, D. see MEJRI, K. FLAHAULT, A. see FARDET, L. GILCHREST, B.A. see YAAR, M. FLAIG, M.J. & RUPEC, R.A. Cutaneous pseudolymphoma in association GILLITZER, R. see TOKSOY, A. with Leishmania donovani, 1042 GIORDANO, L. see ESPOSITO, G. FLEMING, C.J. see BERROETA, L. GIRAUD, M. see KILIC, S.S. FLEMING, J. see WHITE, J.M.L. GIROLOMONI, G. see BEDINI, C. FOK, T.F. see HON, K.L.E. GIROLOMONI, G. see GISONDI, P. FONTANELLAS, A. see ME´NDEZ, M. GIROLOMONI, G. see GOMEZ LIRA, M. FORNI, A. see GOMEZ LIRA, M. GISONDI, P., TESSARI, G., CONTI, A., PIASERICO, S., SCHIANCHI, S., FO¨ RSTI, A. see WILKENING, S. PESERICO, A., GIANNETTI, A. & GIROLOMONI, G. Prevalence of FOSCHI, A. see ANTINORI, S. metabolic syndrome in patients with psoriasis: a hospital-based FRANK, J. see ME´NDEZ, M. case–control study, 68 FRANK, J. see SEITZ, C.S. GIULIODORI, K. see CAMPANATI, A. FRANKE, I. see BRU¨ CHER, J-J. GLA¨SER, R. see HU¨ GEL, R. FRENCH, L.E. see BRAUN, R.P. GLAZENBURG, E.J. see ORANJE, A.P. FRENCH, L.E. see PERRINAUD, A. GLEN´ ,J.see NEDOSZYTKO, B. FRIEDMANN, P.S. The relationships between exposure dose and GLENDINNING, A. see WHITE, J.M.L. response in induction and elicitation of contact hypersensitivity in GNIADECKI, R., ASSAF, C., BAGOT, M., DUMMER, R., DUVIC, M., humans, 1093 KNOBLER, R., RANKI, A., SCHWANDT, P. & WHITTAKER, S. FUJIKURA, M., OHTSUKA, T. & OYAMADA, Y. Systemic sclerosis The optimal use of bexarotene in cutaneous T-cell lymphoma, in association with peristomal pyoderma gangrenosum, 618 433

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1310 Author index

GOEBELER, M. see TOKSOY, A. GUARNERI, C. & GUARNERI, F. Mondor’s phlebitis after using GOEL, A. see BANSAL, S. tadalafil, 209 GOFF-LEVAN, S.L. see DEL GIUDICE, P. GUARNERI, F. see GUARNERI, C. GOHRITZ, A. see AUST, M.C. GUERRA, A.P. see MAVILIA, L. GOLDSCHMIDT, W.F.M. see BOLLING, M.C. GUILHOU, J.-J. see MOLE`S, J-P. GOLLNICK, H. see BRU¨ CHER, J-J. GUMEDZE, F. see KHUMALO, N.P. GOMEZ LIRA, M., MAZZOLA, S., TESSARI, G., MALERBA, G., GUNASTI, S. see ALPSOY, E. ORTOMBINA, M., NALDI, L., REMUZZI, G., BOSCHIERO, L., GUO, Y.L. see LEE, Y-L. FORNI, A., RUGIU, C., PIASERICO, S., GIROLOMONI, G. & GUPTA, A.K., ZAMAN, M. & SINGH, J. Fast and sensitive detection of TURCO, A. Association of functional gene variants in the Trichophyton rubrum DNA from the nail samples of patients with regulatory regions of COX-2 gene (PTGS2) with nonmelanoma onychomycosis by a double-round polymerase chain reaction- skin cancer after organ transplantation, 49 based assay, 698 GONG, S-J. see FANG, Y. GURZAU, E. see WILKENING, S. GOO, B., CHUNG, H.J., CHUNG, W.G. & CHUNG, K.Y. Intramuscular GUSTAVSSON, C.G. see EKQVIST, S. immunoglobulin for recalcitrant suppurative diseases of the skin: a GUTIERREZ-SALMERO´ N, M.T. see RIOS-FERNA´NDEZ, R. retrospective review of 63 cases, 563 GUTTMAN-YASSKY, E. & KRUEGER, J.G. Psoriasis: evolution of GOODLAD, J.R. see MULLER, F.M. pathogenic concepts and new therapies through phases of GOTERI, G., RUPOLI, S., STRAMAZZOTTI, D., DISCEPOLI, G., translational research, 1103 SCORTECHINI, A.R., GIACCHETTI, A., MORICHETTI, D., TASSETTI, A., PULINI, S., MULATTIERI, S., STRONATI, A. & HAGUE, J.S., GOULDING, J.M.R., LONG, T.M.W. & GEE, B.C. LEONI, P. Coexistence of two discordant B-cell lymphomas in the Respiratory involvement in toxic epidermal necrolysis skin and lymph node: report of a case with primary cutaneous portends a poor prognosis that may not be reflected in SCORTEN, follicle-center lymphoma and nodal mantle-cell lymphoma, 1294 629 HAKAMADA, A. see KUROKAWA, I. GOTERI, G., SIMONETTI, O., RUPOLI, S., PICCININI, G., RUBINI, C., HALEVY, S. see SIDOROFF, A. STRAMAZZOTTI, D., FAZIOLI, F., CAPOMAGI, C., LEONI, P., HAMADA, T., YASUMOTO, S., KARASHIMA, T., ISHII, N., SHIMADA, OFFIDANI, A.M. & LOMUZIO, L. Differences in survivin location H., KAWANO, Y., IMAYAMA, S., MCGRATH, J.A. & and Bcl-2 expression in CD30+ lymphoproliferative disorders of HASHIMOTO, T. Recurrent p.N767S mutation in the ATP2A2 gene the skin compared with systemic anaplastic large cell lymphomas: in a Japanese family with haemorrhagic Darier disease clinically an immunohistochemical study, 41 mimicking epidermolysis bullosa simplex with mottled GOTERI, G. see CAMPANATI, A. pigmentation, 605 GOTO, M. see YASUKAWA, K. HAMADA, T. see YAMADA, A. GOTTLIEB, A.B. see KRISHNAN, R. HAMASAKA, K. see HOSHINA, D. GOTTSCHALLER, C. see BABILAS, P. HAMBLY, B.D. see FANG, Y. GOUDIE, D.R. see BERROETA, L. HAMETNER, R. see SADLER, E. GOULDING, J.M.R. see HAGUE, J.S. HAMILTON, T.A. see HELFRICH, Y.R. GRABNER, T. see EISENDLE, K. HAMM, H. see BENOIT, S. GRADILONE, A., GAZZANIGA, P., CIGNA, E., VASATURO, F., HAMM, H. see SEITZ, C.S. VINCENZI, B., GANDINI, O., SILVESTRI, I., RIBUFFO, D., HAMPTON, P.J., ROSS, O.K. & REYNOLDS, N.J. Increased SCARPA, S., SCUDERI, N. & AGLIANO` , A.M. Fibronectin and nuclear b-catenin in suprabasal involved psoriatic epidermis, laminin expression in sentinel lymph nodes of patients with 1168 malignant melanoma, 398 HANDISURYA, A., RIEGER, A., BANKIER, A., KOLLER, A., SALAT, A., GRA¨SER, Y. see ARABATZIS, M. STINGL, G. & KIRNBAUER, R. Human papillomavirus type 26 GRATTAN, C.E.H. & HUMPHREYS, F. ON BEHALF OF THE BRITISH infection causing multiple invasive squamous cell carcinomas of ASSOCIATION OF DERMATOLOGISTS THERAPY GUIDELINES AND the fingernails in an AIDS patient under highly active antiretroviral SUBCOMMITTEE. Guidelines for evaluation and management of therapy, 788 urticaria in adults and children, 1116 HANN, S., HUGHES, T.M. & STONE, N.M. Flexural allergic contact GREENBLATT, D. see SHETH, N. dermatitis to benzalkonium chloride in antiseptic bath oil, 795 GREENLAND, K.J. see SINCLAIR, R. HARMAN, K.E. see HELBLING, I. GREGORIOU, S., KALOGEROMITROS, D., LARIOS, G., MAKRIS, M. & HARMAN, K.E. see MACBETH, A.E. RIGOPOULOS, D. Impact of a public service advertisement about HARPER, J.I. see MAZEREEUW-HAUTIER, J. onychomycosis on the health behaviour of the Greek population HARPER, J.I. see PURVIS, D.J. with nail disorders, 821 HA´RSING, J. see MARSCHALKO´ ,M. GRIFFIN, M.D. see OTLEY, C.C. HARTMANN, M. see GHOLAM, P. GRIMS, R.H., WEGER, W., REITER, H., ARBAB, E., KRA¨NKE, B. & HASAN, T. see JA¨RVINEN, T.M. ABERER, W. Delayed-type hypersensitivity to low molecular HASHIMOTO, K. see TOHYAMA, M. weight heparins and heparinoids: cross-reactivity does not depend HASHIMOTO, T. see HAMADA, T. on molecular weight, 514 HASHIMOTO, T. see YAHARA, H. GROVER, C., BANSAL, S., NANDA, S., REDDY, B.S.N. & KUMAR, V. HATAMOCHI, A., SASAKI, T., KAWAGUCHI, T., SUZUKI, H. & Combination of surgical avulsion and topical therapy for single YAMAZAKI, S. A novel point mutation in the gene encoding nail onychomycosis: a randomized controlled trial, 364 capillary morphogenesis protein 2 in a Japanese patient with GRUVBERGER, B. see EKQVIST, S. juvenile hyaline fibromatosis, 1037 GUARNERI, B. see VACCARO, M. HATTORI, N. see ASAHINA, A.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1311

HAYASHIDA, S. see KIDO, M. HOSHINA, D., AKIYAMA, M., HAMASAKA, K. & SHIMIZU, H. HAYDOCK, S.F., BORNSHIN, S., WALL, E.C. & CONNICK, R.M. An infantile case of pityriasis lichenoides et varioliformis acuta, 194 Admissions to a U.K. teaching hospital with nonnecrotizing lower HOW, P. see LO, S. limb cellulitis show a marked seasonal variation, 1047 HØYEN, M. see NOIESEN, E. HAZANE-PUCH, F. see SAUVAIGO, S. HRABA´LEK, A. see VA´VROVA´,K. HE, Q. see RADY, P.L. HSU, C.K., LEE, J.Y.Y., YU, C.H., HSU, M.M.L. & WONG, T.W. HEBERT, A.A. see SPERGEL, J.M. Lip verrucous carcinoma in a pregnant woman successfully treated HEDELIN, G. see LIPSKER, D. with carbon dioxide laser surgery, 813 HELBLING, I., WALEWSKA, R., DYER, M.J.S., BAMFORD, M. & HSU, M.M.L. see HSU, C.K. HARMAN, K.E. Erythema annulare centrifugum associated with HU, K-S. see SONG, W-C. chronic lymphocytic leukaemia, 1044 HU, S.C-S., CHEN, G-S., WU, C-S. & LAN, C-C.E. Serum tissue HELFRICH, Y.R., KANG, S., HAMILTON, T.A. & VOORHEES, J.J. Topical polypeptide antigen correlating with clinical course in a patient becocalcidiol for the treatment of psoriasis vulgaris: a randomized, with mycosis fungoides: a potential disease marker?, 423 placebo-controlled, double-blind, multicentre study, 369 HUANG, W-H. see SHAW, J. HELLER, M., SHIN, H.T., ORLOW, S.J. & SCHAFFER, J.V. HU¨ GEL, R., SCHWARZ, T. & GLA¨SER, R. Resistance to Mycophenolate mofetil for severe childhood atopic dermatitis: hydroxychloroquine due to smoking in a patient with lupus experience in 14 patients, 127 erythematosus tumidus, 1081 HEMMINKI, K. see WILKENING, S. HUGHES, C.M. see SHAW, J. HENNESSY, A. see FINNEN, M.J. HUGHES, D.A. see ORTEU, C.H. HERAS MULERO, C., BARTRALOT SOLER, R., RODRI´GUEZ-CANO, L., HUGHES, T.M. see HANN, S. MOLLET SA´NCHEZ, J., PALACIO ALLER, L., APARICIO ESPAN˜ OL, HUH, S. see LEE, J. G., BODET CASTILLO, D., BASSAS FREIXAS, P. & GARCI´A-PATOS, HULTSCH, T. see SPERGEL, J.M. V. Aplasia cutis associated with coarctation of the aorta: could this HUMBERT, P. see VA´VROVA´,K. be an incomplete form of Adams–Oliver syndrome?, 836 HUMMEL, S. see BABILAS, P. HERPHELIN, F. see SERCU, S. HUMPHREYS, F. see GRATTAN, C.E.H. HERRERO, C. see MORENO-ROMERO, J.A. HUSNI, M.E. see MODY, E. HIDE, M. see KAMEYOSHI, Y. HUSNI, M.E. see POMERANTZ, R.G. HILLIGES, M. see JOHANNESSON, U. HUSSEY, L. see TURNER, S. HINTNER, H. see SADLER, E. HUYLEBROECK, D. see SERCU, S. HIRUMA, M. see TAKAHATA, Y. HWANG, S.T. see COWEN, E.W. HISCUTT, E. see WAHIE, S. HYUN, C.G. see LEE, J. HO, M.G. see ZUO, Y-G. HO, V.C. see MURRELL, D.F. IBBOTSON, S.H. see BERROETA, L. HOEDEMAKER, C. see SINCLAIR, R. IBBOTSON, S.H. see SMITH, G. HOFFJAN, S. & STEMMLER, S. On the role of the epidermal IENO, L. see ESPOSITO, G. differentiation complex in ichthyosis vulgaris, atopic dermatitis IIJIMA, M. see TOHYAMA, M. and psoriasis, 441 IKEDA, S. see ZHENG, Y. HOFFMANN, K. see BECHARA, F.G. IMAFUKU, S., SHIBATA, S., TASHIRO, A. & FURUE, M. Cutaneous HOFFSTAD, O. see MARGOLIS, D.J. Langerhans cell histiocytosis in an elderly man successfully treated HOFMANN-WELLENHOF, R. see ARGENZIANO, G. with narrowband ultraviolet B, 1277 HOLLOWOOD, K. see LALLY, A. IMANISHI, Y. see OKA, M. HOLM, E.A., WULF, H.C., THOMASSEN, L. & JEMEC, G.B.E. IMAYAMA, S. see HAMADA, T. Assessment of atopic eczema: clinical scoring and noninvasive INAMADAR, A.C. see ARITA, K. measurements, 674 INOUE, K. see DENDA, M. HOLMSTRO¨ M, E. see EKQVIST, S. ISHII, N. see HAMADA, T. HOLTZ, R. see SHAW, J. ISHIKAWA, O. see NAKAMURA, M. HOLUB, A. see ROBAK, E. ISHIURA, N., KOMINE, M., KADONO, T., KIKUCHI, K. & TAMAKI, K. HOMEY, B. see ANTAL, A. A case of milia en plaque successfully treated with oral etretinate, HON, K.L.E., LEUNG, T.F., NG, P.C., LAM, M.C.A., KAM, W.Y.C., 1287 WONG, K.Y., LEE, K.C.K., SUNG, Y.T., CHENG, K.F., FOK, ISHIURA, N. see KUWANO, Y. T.F., FUNG, K.P. & LEUNG, P.C. Efficacy and tolerability of a ISODA, K-I. see KUROKAWA, I. Chinese herbal medicine concoction for treatment of atopic ITOH, M. see KAWACHI, Y. dermatitis: a randomized, double-blind, placebo-controlled study, IWATSUKI, K. see AOCHI, S. 357 IWATSUKI, K. see YAMADA, A. HON, K-L.E., LAM, M-C.A., WONG, K-Y., LEUNG, T-F. & NG, P-C. Pathophysiology of nocturnal scratching in childhood atopic JACKSON, H.A. see OLIVRY, T. dermatitis: the role of brain-derived neurotrophic factor and JAHREIS, A. see KRISHNAN, R. substance P, 922 JANAWAY, R.C. see WILSON, A.S. HONG, J-B., CHIU, H-C., WANG, S-H. & TSAI, T-F. Recurrence of JANSEN, T. see ORTEU, C.H. classical juvenile pityriasis rubra pilaris in adulthood: report of a JA¨RVINEN, T.M., KANNINEN, P., JESKANEN, L., KOSKENMIES, S., case, 842 PANELIUS, J., HASAN, T., RANKI, A. & SAARIALHO-KERE, U. DE HOOG, G.S. see ARABATZIS, M. Matrix metalloproteinases as mediators of tissue injury in different HORIKAWA, T. see TOHYAMA, M. forms of cutaneous lupus erythematosus, 970

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1312 Author index

JASCHKE, W. see EISENDLE, K. KAWAKAMI, T., KAWASAKI, K., MIZOGUCHI, M. & SOMA, Y.

JAUSSAUD, R. see ORTEU, C.H. Therapeutic effect of lipoprostaglandin E1 on livedoid vasculitis JEANMOUGIN, M. see VIGUIER, M. associated with essential cryoglobulinaemia, 1051 JEFFERIES, D. see WHITE, J.M.L. KAWANO, Y. see HAMADA, T. JEMEC, G.B.E. see HOLM, E.A. KAWASAKI, K. see KAWAKAMI, T. JEMEC, G.B.E. see MOGENSEN, M. KAZAKOV, D.V. see BELOUSOVA, I.E. JESIONEK-KUPNICKA, D. see ROBAK, E. KEHLER, M. see EKQVIST, S. JESKANEN, L. see JA¨RVINEN, T.M. KENDALL, B.R. see MACBETH, A.E. JESSOP, S. see KHUMALO, N.P. VAN DE KERKHOF, P.C.M. The evolution of the psoriatic lesion, 4 JIN, H. see NAKANO, H. KERNOHAN, N. see MULLER, F.M. JIN, J. see TORRES, A. KERR, A.C. & FERGUSON, J. Actinic prurigo deterioration due to JOHANNESSON, U., BLOMGREN, B., HILLIGES, M., RYLANDER, E. & degradation of DermaGard window film, 619 BOHM-STARKE, N. The vulval vestibular mucosa—morphological KERROUCHE, N. see BARAN, R. effects of oral contraceptives and menstrual cycle, 487 KESEROGLU, O. see ALPSOY, E. JOHANSEN, J.D. see FISCHER, L.A. KETTANEH, A. see FARDET, L. JOHNSTON, G.A. & ENGLISH, J.S.C. The alcohol hand rub: a good KHAN, N.K. see BELL, R.R. soap substitute?, 1 KHUMALO, N.P., JESSOP, S., GUMEDZE, F. & EHRLICH, R. JONES, S.K. see SMITH, V.H. Hairdressing and the prevalence of scalp disease in African adults, JONKMAN, M.F. see BOLLING, M.C. 981 JULIA`,M.see MORENO-ROMERO, J.A. KHUMALO, N.P., JESSOP, S., GUMEDZE, F. & EHRLICH, R. JUNG, E. see LEE, J. Hairdressing is associated with scalp disease in African schoolchildren, 106 KAATZ, M., ZELGER, B., NORGAUER, J. & ZIEMER, M. Lymphocytic KHURANA, N. see BANSAL, S. infiltration (Jessner–Kanof): lupus erythematosus tumidus or a KIDO, M., TAKEUCHI, S., HAYASHIDA, S., URABE, K., SAWADA, R. manifestation of borreliosis?, 403 & FURUE, M. Assessment of abnormal blood flow and efficacy of KADONO, T. see ISHIURA, N. treatment in patients with systemic sclerosis using a newly KAI, H. see ASAHINA, A. developed microwireless laser Doppler flowmeter and arm-raising KAJI, H. see OKA, M. test, 690 KALL, S. see AUST, M.C. KIKUCHI, K. see ISHIURA, N. KALLEL-SELLAMI, M. see MEJRI, K. KILIC, S.S., GIRAUD, M., SCHMITT, S., BE´ZIEAU, S. & KU¨ RY, S. A KALOGEROMITROS, D. see GREGORIOU, S. novel mutation of the SLC39A4 gene causing acrodermatitis KAM, W.Y.C. see HON, K.L.E. enteropathica, 386 KAMEYOSHI, Y., TANAKA, T., MIHARA, S., TAKAHAGI, S., NIIMI, N. KIM, H.J., LEE, J.Y., KIM, S.H., SEO, Y.J., LEE, J.H., PARK, J.K., KIM, & HIDE, M. Increasing the dose of cetirizine may lead to better M.H., CINN, Y.W., CHO, K.H. & YOON, T.Y. Stromelysin-3 control of chronic idiopathic urticaria: an open study of 21 patients, expression in the differential diagnosis of dermatofibroma and 803 dermatofibrosarcoma protuberans: comparison with factor XIIIa KAMO, T. see OKA, M. and CD34, 319 KAMPF, G. see LO¨ FFLER, H. KIM, H-J. see SONG, W-C. KANDI, B. see ALPSOY, E. KIM, H-J. see YUN, S.J. KANG, M-S. see YU, H-J. KIM, I-H. see OH, J. KANG, S. see COWEN, E.W. KIM, J-S. see YU, H-J. KANG, S. see HELFRICH, Y.R. KIM, M.H. see KIM, H.J. KANG, T-W. see OH, S-W. KIM, M.Y. see OH, S-W. KANNINEN, P. see JA¨RVINEN, T.M. KIM, M-B. see SEO, S-H. KANNO, H. see WATABE, D. KIM, N. see OH, J. KARASHIMA, T. see HAMADA, T. KIM, S.H. see KIM, H.J. KARINCAOGLU, Y. see ALPSOY, E. KIM, S-C. see OH, S-W. KAROO, R.O.S. see SOWDEN, H.M. KIM, S-J. see YUN, S.J. KA´RPA´TI, S. see MARSCHALKO´ ,M. KIM, Y.C. see OH, S-W. KATAYAMA, I. see MABUCHI, E. KIM, Y-S. see LEE, J. KATO, H. see TAKAHATA, Y. KIMBALL, A.B. see CIOCON, D.H. KAUFMANN, R. see BARAN, R. KIMURA, H. see TOHYAMA, M. KAVANAGH, K. see LACEY, N. KIRBY, B. see MALIK, M. KAWACHI, Y., ITOH, M., FUJISAWA, Y., FURUTA, J., NAKAMURA, KIRNBAUER, R. see HANDISURYA, A. Y., BANNO, T., TAKAHASHI, T. & OTSUKA, F. Epidermal cell KIRSCHNER, R. see PIEPENBRING, M. necrosis with direct epidermal infiltration of Epstein–Barr virus KLEINMAN, H.K. see LUGASSY, C. (EBV)-encoded small nuclear RNA-positive T lymphocytes KLOEPPER, J.E. see TIEDE, S. in a patient with EBV-associated haemophagocytic syndrome, KNOBLER, R. see BABILAS, P. 1053 KNOBLER, R. see GNIADECKI, R. KAWAGUCHI, T. see HATAMOCHI, A. KOH, K-S. see SONG, W-C. KAWAKAMI, T., KAWASAKI, K. & SOMA, Y. Limited cutaneous KOLLER, A. see HANDISURYA, A. systemic sclerosis associated with discoid lupus erythematosus in KOLLER, M. see BABILAS, P. two Japanese patients with anticentromere antibodies, 1289 KOLOKYTHAS, P. see AUST, M.C.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1313

KOMINE, M. see ISHIURA, N. LAM, W.Y. see CHAN, K.H.N. KOPF, A.W. see BRAUN, R.P. LAMANT, L. see FAGUER, S. KOPPOVA, K. see WILKENING, S. LAMY, T. see MEYER, N. KORE-EDA, S. see NISHIWAKI, F. LAN, C-C.E. see HU, S.C-S. KOSKENMIES, S. see JA¨RVINEN, T.M. LANDTHALER, M. see BABILAS, P. KOSSARD, S. see MARTIN, L.K. LANGTRY, J.A.A. see BAJAJ, V. KOSSINTSEVA, I. see DYTOC, M.T. LANGTRY, J.A.A. see RAJAN, N. KOUBA, D.J., YIP, D., FINCHER, E.F. & MOY, R.L. Topical imiquimod LANSCHUETZER, C. see SADLER, E. in the treatment of a long-standing capillary malformation, 1071 LANZERINI, P. see VASSALLO, C. KRA¨NKE, B. see GRIMS, R.H. LARIOS, G. see GREGORIOU, S. KREUTER, A., BROCKMEYER, N.H., ALTMEYER, P., PFISTER, H. & LARSEN, K. see NOIESEN, E. WIELAND, U. Rapid onset of multifocal human papillomavirus LATIFAJ, B. see BISWAS, A. 72-associated oral intraepithelial neoplasia in a human LAUNAY, F. see FAGUER, S. immunodeficiency virus-infected patient, 826 LAVRIJSEN, A.P.M. see ARABATZIS, M. KREUTER, A., WIELAND, U., GAMBICHLER, T., ALTMEYER, P., LAWRENCE, R. see MARTIN, L.K. PFISTER, H., TENNER-RACZ, K., RACZ, P., POTTHOFF, A. & LAZZARINO, M. see VASSALLO, C. BROCKMEYER, N.H. FOR THE GERMAN NETWORK OF LE DANFF, C. see RE´GNIER, S. COMPETENCE HIV ⁄AIDS. p16ink4a expression decreases during LE DUC, Q., BREETVELD, M., MIDDELKOOP, E., SCHEPER, R.J., imiquimod treatment of anal intraepithelial neoplasia in human ULRICH, M.M.W. & GIBBS, S. A cytotoxic analysis of antiseptic immunodeficiency virus-infected men and correlates with the medication on skin substitutes and autograft, 33 decline of lesional high-risk human papillomavirus DNA load, 523 LE FICHOUX, Y. see DEL GIUDICE, P. KRISHNAN, R., CELLA, D., LEONARDI, C., PAPP, K., GOTTLIEB, A.B., LE FORESTIER, D. see BARBAROT, S. DUNN, M., CHIOU, C.F., PATEL, V. & JAHREIS, A. Effects of LEACH, I. see ANGUS, J. etanercept therapy on fatigue and symptoms of depression in LEBBE´,C.see FARDET, L. subjects treated for moderate to severe plaque psoriasis for up to LECHA, M. see BARAN, R. 96 weeks, 1275 LEDUFF, F. see DEL GIUDICE, P. KRUEGER, J.G. see GUTTMAN-YASSKY, E. LEE, A-R. see SHAW, J. KUIJPER, E.J. see ARABATZIS, M. LEE, J., JUNG, E., LEE, J., HUH, S., BOO, Y.C., HYUN, C.G., KIM, KULKARNI, K. see ANGUS, J. Y-S. & PARK, D. Mechanisms of melanogenesis inhibition by KUMAR, R. see WILKENING, S. 2,5-dimethyl-4-hydroxy-3(2H)-furanone, 242 KUMAR, V. see BANSAL, S. LEE, J. see LEE, J. KUMAR, V. see GROVER, C. LEE, J. see TORRES, A. KURFU¨ RST, R. see SAUVAIGO, S. LEE, J.H. see KIM, H.J. KUROKAWA, I., UMEDA, K., NISHIMURA, K., YAMANAKA, K-I., LEE, J.S. see OH, S-W. HAKAMADA, A., ISODA, K-I., TSUBURA, A. & MIZUTANI, H. LEE, J.Y. see KIM, H.J. Filaggrin expression and the pathogenesis of epidermal cysts, LEE, J.Y.Y. see HSU, C.K. 415 LEE, J-B. see YUN, S.J. KUROSE, A. see WATABE, D. LEE, K.C.K. see HON, K.L.E. KU¨ RY, S. see KILIC, S.S. LEE, S-C. see YUN, S.J. KU¨ STER, R.M. see ROTT, S. LEE, Y-L., LI, C-W., SUNG, F-C., YU, H-S., SHEU, H-M. & GUO, Y.L. KUWANO, Y., FUJIMOTO, M., WATANABE, R., ISHIURA, N., Environmental factors, parental atopy and atopic eczema in NAKASHIMA, H., OHNO, Y., YANO, S., YAZAWA, N., OKOCHI, primary-school children: a cross-sectional study in Taiwan, 1217 H. & TAMAKI, K. Serum chemokine profiles in patients with LEISNER, R.I. see MAZEREEUW-HAUTIER, J. alopecia areata, 466 LEONARD, J.N. see COX, G.A. KWAK, H.S. see COWEN, E.W. LEONARDI, C. see KRISHNAN, R. KWON, K-S. see SEO, S-H. LEONI, P. see GOTERI, G. KYVIK, K.O. see LERBAEK, A. LEPREUX, S. see MORICE-PICARD, F. LERBAEK, A., BISGAARD, H., AGNER, T., OHM KYVIK, K., PALMER, LABRE`ZE, C. see MORICE-PICARD, F. C.N.A. & MENNE´, T. Filaggrin null alleles are not associated with LACEY, N., DELANEY, S., KAVANAGH, K. & POWELL, F.C. Mite- hand eczema or contact allergy, 1199 related bacterial antigens stimulate inflammatory cells in rosacea, LERBAEK, A., KYVIK, K.O., RAVN, H., MENNE´, T. & AGNER, T. 474 Incidence of hand eczema in a population-based twin cohort: LACHMANN, N. see SAUVAIGO, S. genetic and environmental risk factors, 552 LACOMBE, D. see MORICE-PICARD, F. LEUNG, P.C. see HON, K.L.E. LAGAN, K.M. see SHAW, J. LEUNG, T.F. see HON, K.L.E. LAI-CHEONG, J.E. see ARITA, K. LEUNG, T-F. see HON, K-L.E. LAI-CHEONG, J.E. see MARTIGNAGO, B.C.F. LEVERKUS, M. see BRU¨ CHER, J-J. LAIMER, M. see SADLER, E. LEW, W. see OH, S-W. LALLY, A., HOLLOWOOD, K., WHITTAKER, S. & TURNER, R. LEWIS, J. see GERAMI, P. Central nervous system involvement in stage 1b mycosis fungoides, LEWIS-JONES, S. see MULLER, F.M. 815 LI, C-W. see LEE, Y-L. LAM, M.C.A. see HON, K.L.E. LIAO, Y.H., CHIU, H.C., TSENG, Y.S. & TSAI, T.F. Comparison ) LAM, M-C.A. see HON, K-L.E. of cutaneous tolerance and efficacy of calcitriol 3 lgg 1

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1314 Author index

) ointment and tacrolimus 0Æ3mgg 1 ointment in chronic plaque MCPARTLAND, J.L. see WEST, E.A. psoriasis involving facial or genitofemoral areas: a double-blind, MCRAE, S. see ALOMAR, A. randomized controlled trial, 1005 MADKAN, V.K., COOK-NORRIS, R.H., STEADMAN, M.C., ARORA, LIDOVE, O. see ORTEU, C.H. A. & TYRING, S.K. The oncogenic potential of human LIEKENS, J. see SERCU, S. papillomaviruses: a review on the role of host genetics and LIN, A.H. see ZHU, K.J. environmental cofactors, 228 LIN, Y-C. see TSAI, Y-S. MAIBACH, H.I. see LO¨ FFLER, H. LINES, S. see TURNER, S. MAILHOL, C. see FAGUER, S. LIOTTA, L.A. see COWEN, E.W. MAISCH, T. see BABILAS, P. LIPSKER, D., ENGEL, F., CRIBIER, B., VELTEN, M. & HEDELIN, G. MAKNI, S. see MEJRI, K. Trends in melanoma epidemiology suggest three different types of MAKRIS, M. see GREGORIOU, S. melanoma, 338 MALAGNINO, V. see MOURMOURAS, V. LIU, C. see SHAW, J. MALARA, G. see MAVILIA, L. LIU, C-W. see COWEN, E.W. MALERBA, G. see GOMEZ LIRA, M. LIU, L. see MARTIGNAGO, B.C.F. MALIK, M., TOBIN, A.-M., SHANAHAN, F., O’MORAIN, C., KIRBY, LIU, Y-H. see ZUO, Y-G. B. & BOURKE, J. Steroid allergy in patients with inflammatory LLOYD, J. see COX, G.A. bowel disease, 967 LLOYD, M.S., CLARK, A., PARIKH, A. & BUTLER, P. Designing a MALLO-GARCIA, S. see COTO-SEGURA, P. validated patient information website: a quality-controlled MALVEHY, J. see ZABALLOS, P. information portal illustrated by skin cancer, 1048 MAN, I. see BERROETA, L. LO RE, M. see MAVILIA, L. MANNELLO, B. see CAMPANATI, A. LO, K.K. see CHAN, K.H.N. MANSAT, E. see BARBAROT, S. LO, S., HOW, P. & MOSS, A.L.M. Plexiform schwannoma mimicking MAO, X., ORCHARD, G., RUSSELL-JONES, R. & WHITTAKER, S. haemangioma: pitfalls in clinical diagnosis and histological Abnormal activator protein 1 transcription factor expression in interpretation, 838 CD30-positive cutaneous large-cell lymphomas, 914 LOCATELLI, F. see VASSALLO, C. MARCONI, F. see BORIANI, F. LODE´N, M. see BURACZEWSKA, I. MARCUS-SOEKARMAN, D. see STEIJLEN, P.M. LO¨ FFLER, H., KAMPF, G., SCHMERMUND, D. & MAIBACH, H.I. MARGHOOB, A.A. see PERRINAUD, A. How irritant is alcohol?, 74 MARGOLIS, D.J., HOFFSTAD, O. & BILKER, W. Association or lack of LOMUZIO, L. see GOTERI, G. association between tetracycline class antibiotics used for acne LONG, T.M.W. see HAGUE, J.S. vulgaris and lupus erythematosus, 540 LONGHI, E. see ANTINORI, S. MARROU, K. see DALLE, S. LU, Z.M. see ZHU, K.J. MARSCHALKO´ , M., CSOMOR, J., EROS,} N., SZIGETI, A´., HA´RSING, J., LUCARINI, G. see TUCCI, M.G. SZAKONYI, J., DE´SAKNAI, M., MATOLCSY, A., DEMETER, J. & LUDWIG, R.J. see BOEHNCKE, S. KA´RPA´TI, S. Coexistence of primary cutaneous anaplastic large cell LUGASSY, C., KLEINMAN, H.K., VERNON, S.E., WELCH, D.R. & lymphoma and mycosis fungoides in a patient with B-cell chronic BARNHILL, R.L. C16 laminin peptide increases angiotropic lymphocytic leukaemia, 1291 extravascular migration of human melanoma cells in a shell-less MARTI´,R.see PARERA, E. chick chorioallantoic membrane assay, 780 MARTIGNAGO, B.C.F., LAI-CHEONG, J.E., LIU, L., MCGRATH, J.A. LUPI, F. see CIANCHINI, G. & CESTARI, T.F. Recurrent KIND1 (C20orf42) gene mutation, LY, L. & CZARNECKI, D. The rapid onset of multiple squamous c.676insC, in a Brazilian pedigree with Kindler syndrome, cell carcinomas during etanercept treatment for psoriasis, 1281 1076 MARTIN, L.K., LAWRENCE, R., KOSSARD, S. & MURRELL, D.F. Cutaneous Mycobacterium neoaurum infection causing scarring alopecia MABUCHI, E., UMEGAKI, N., MUROTA, H., NAKAMURA, T., TAMAI, in an immunocompetent host, 204 K. & KATAYAMA, I. Oral steroid improves bullous pemphigoid- MARTIN, R. see SHAW, J. like clinical manifestations in non-Herlitz junctional epidermolysis MARTY, P. see DEL GIUDICE, P. bullosa with COL17A1 mutation, 596 MASCARO´ JR, J.M. see MORENO-ROMERO, J.A. MACBETH, A.E., KENDALL, B.R., SMITH, A., SALDANHA, G. & MASINI, C. see CIANCHINI, G. HARMAN, K.E. Calcified subcutaneous nodules: a long-term MASMOUDI, H. see MEJRI, K. complication of interferon beta-1a therapy, 624 MATOLCSY, A. see MARSCHALKO´ ,M. MCCORMICK, C. see SPERGEL, J.M. MATSUMOTO, K. see TOHYAMA, M. MCFADDEN, J.P. see WHITE, J.M.L. MATSUMURA, Y. see NISHIWAKI, F. MCGRATH, J.A. see ARITA, K. MAVILIA, L., MALARA, G., MORETTI, G., LO RE, M. & GUERRA, A.P. MCGRATH, J.A. see HAMADA, T. Photodynamic therapy of acne using methyl aminolaevulinate MCGRATH, J.A. see MARTIGNAGO, B.C.F. diluted to 4% together with low doses of red light, 810 MCGRATH, J.A. see SERCU, S. MAYOR, L. see DURUPT, F. MCGRATH, J.A. see SHETH, N. MAZEREEUW-HAUTIER, J., SYED, S., LEISNER, R.I. & HARPER, J.I. VON MACKENSEN, Y.A. & STICHERLING, M. Cold urticaria: tolerance Extensive venous/lymphatic malformations causing life-threatening induction with cold baths, 835 haematological complications, 558 MCLEAN, S. see FINNEN, M.J. MAZZOLA, S. see GOMEZ LIRA, M. MAC-MARY, S. see VA´VROVA´,K. MAZZOTTA, A. see COSTANZO, A. MCNAMEE, R. see TURNER, S. MBAREK, H. see MEJRI, K.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1315

MEGAHED, M. see ANTAL, A. MORA´N-JIME´NEZ, M.J. see ME´NDEZ, M. MEGSON, I.L. see FINNEN, M.J. MORENO-ROMERO, J.A., SEGURA, S., MASCARO´ JR, J.M., COWPER, MEHTA, A.B. see ORTEU, C.H. S.E., JULIA`, M., POCH, E., BOTEY, A. & HERRERO, C. MEIKLEJOHN, D. see MULLER, F.M. Nephrogenic systemic fibrosis: a case series suggesting gadolinium MEJRI, K., KALLEL-SELLAMI, M., PETIT-TEIXEIRA, E., ABIDA, O., as a possible aetiological factor, 783 MBAREK, H., ZITOUNI, M., BEN AYED, M., TEIXEIRA, V.H., MORETTI, G. see MAVILIA, L. MOKNI, M., FAZZA, B., TURKI, H., TRON, F., GILBERT, D., MORI, T. see ASAHINA, A. MASMOUDI, H., CORNELIS, F. & MAKNI, S. PTPN22 R620W MORICE-PICARD, F., BORALEVI, F., LEPREUX, S., LABRE`ZE, C., polymorphism is not associated with pemphigus, 1068 LACOMBE, D. & TAI¨EB, A. Severe linear form of granuloma MEKKES, J.R. see BOLLING, M.C. annulare along Blaschko’s lines preceding the onset of a classical ME´NDEZ, M., POBLETE-GUTIE´RREZ, P., GARCI´A-BRAVO, M., form of granuloma annulare in a child, 1056 WIEDERHOLT, T., MORA´N-JIME´NEZ, M.J., MERK, H.F., MORICHETTI, D. see CAMPANATI, A. GARRIDO-ASTRAY, M.C., FRANK, J., FONTANELLAS, A. & MORICHETTI, D. see GOTERI, G. ENRI´QUEZ DE SALAMANCA, R. Molecular heterogeneity of MORITA, N. see NISHIWAKI, F. familial porphyria cutanea tarda in Spain: characterization of 10 MORIYAMA, Y. see UGAJIN, T. novel mutations in the UROD gene, 501 MORLEY, S. see MULLER, F.M. MENNE´,T.see FISCHER, L.A. MORRISON, L. see ZLOTOFF, B.J. MENNE´,T.see LERBAEK, A. MORTIER, L. see VERCAMBRE-DARRAS, S. MENTER, A. see STERRY, W. MOSELEY, H. see OLIVER, H. MERK, H.F. see ME´NDEZ, M. MOSS, A.L.M. see LO, S. MERONI, L. see ANTINORI, S. MOSS, C. see THOMSON, M.A. MERREGAERT, J. see SERCU, S. MOURA, D.F., TELES, R.M.B., RIBEIRO-CARVALHO, M.M., TELES, MERTENS, J. see VERFAILLE, C.J. R.B., SANTOS, I.M.C.F., FERREIRA, H., FULCO, T.O., NERY, MESSINA, F. see ESPOSITO, G. J.A.C., SAMPAIO, E.P. & SARNO, E.N. Long-term culture of MESTRE, F. see BAUZA´,A. multibacillary leprosy macrophages isolated from skin lesions: a METZE, D. see VOLZ, A. new model to study Mycobacterium leprae–human cell interaction, MEYER, N., DUFOUR, J., LAMY, T. & CHEVRANT-BRETON, J. 273 Cutaneous vasculitis and T-large granular lymphocyte leukaemia MOURMOURAS, V., FIMIANI, M., RUBEGNI, P., EPISTOLATO, M.C., with parallel evolution, 631 MALAGNINO, V., CARDONE, C., COSCI, E., DE NISI, M.C. & MEYERLE, J.H. see CUMMINS, D.L. MIRACCO, C. Evaluation of tumour-infiltrating MIDDELKOOP, E. see LE DUC, Q. CD4+CD25+FOXP3+ regulatory T cells in human cutaneous MIHARA, S. see KAMEYOSHI, Y. benign and atypical naevi, melanomas and melanoma metastases, MIHELARAKIS, I. see MOUZOPOULOS, G. 531 MIKUS, S. see WENZEL, J. MOUZOPOULOS, G., TSOUPAROPOULOS, V., STAMATAKOS, M., MILLER, R.L. see TORRES, A. MIHELARAKIS, I., PASPARAKIS, D. & AGAPITOS, E. Cutaneous MIRACCO, C. see MOURMOURAS, V. mercury deposits after henna dye application in the arm, 394 MITCHELL, R. see FINNEN, M.J. MOY, R.L. see KOUBA, D.J. MITRA, A. see WILLIAMS, C. MROWIETZ, U. see ROTT, S. MITSUHASHI, Y. see NAKANO, H. MULATTIERI, S. see GOTERI, G. MIYACHI, Y. see NISHIKAWA, M. MULLER, F.M., LEWIS-JONES, S., MORLEY, S., KERNOHAN, N., MIYACHI, Y. see NISHIWAKI, F. MEIKLEJOHN, D., GOODLAD, J.R. & EVANS, A. Lymphomatoid MIZOGUCHI, M. see KAWAKAMI, T. granulomatosis complicating other haematological malignancies, MIZUTANI, H. see KUROKAWA, I. 426 MIZUTANI, Y. see SEISHIMA, M. MU¨ LLER, V. see TOKSOY, A. MOCKENHAUPT, M. see SIDOROFF, A. MUNAKATA, T. see NAKANO, H. MODY, E., HUSNI, M.E., SCHUR, P. & QURESHI, A.A. MUNK, M.D. see NOIESEN, E. Multidisciplinary evaluation of patients with psoriasis presenting MURACH, W.M. see WACHTER, T. with musculoskeletal pain: a dermatology: rheumatology clinic MURET, P. see VA´VROVA´,K. experience, 1050 MUROTA, H. see MABUCHI, E. MODY, E. see POMERANTZ, R.G. MURPHY, K.M. see OLIVRY, T. MOGENSEN, M., THOMSEN, J.B., SKOVGAARD, L.T. & JEMEC, G.B.E. MURPHY, R. see ANGUS, J. Nail thickness measurements using optical coherence tomography MURRELL, D.F., CALVIERI, S., ORTONNE, J.P., HO, V.C., and 20-MHz ultrasonography, 894 WEISE-RICCARDI, S., BARBIER, N. & PAUL, C.F. A randomized MOKNI, M. see MEJRI, K. controlled trial of pimecrolimus cream 1% in adolescents and MOLE`S, J-P., TESNIERE, A. & GUILHOU, J.-J. Reverse transcriptase adults with head and neck atopic dermatitis and intolerant of, or activity in human normal and psoriatic skin samples, 482 dependent on, topical corticosteroids, 954 MO¨ LLER, H. see EKQVIST, S. MURRELL, D.F. see MARTIN, L.K. MOLLET SA´NCHEZ, J. see HERAS MULERO, C. MUTO, M. see TAKAHATA, Y. MOLONEY, F.J. & COLLINS, P. Randomized, double-blind, prospective study to compare topical 5-aminolaevulinic acid methylester with NAGAOKA, I. see ZHENG, Y. topical 5-aminolaevulinic acid photodynamic therapy for extensive N’GUYEN, J.M. see BARBAROT, S. scalp actinic keratosis, 87 NAKAJIMA, K. see TOHYAMA, M. MONAHAN, T. see CUMMINS, D.L. NAKAMURA, H. see YASUKAWA, K.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1316 Author index

NAKAMURA, M. & ISHIKAWA, O. Erythrokeratoderma variabilis OFFIDANI, A.M. see GOTERI, G. without GJB3 or GJB4 mutation: a review of Japanese patients, OGATA, T. see SEISHIMA, M. 410 OGAWA, H. see ZHENG, Y. NAKAMURA, T. see MABUCHI, E. OH, C-K. see SEO, S-H. NAKAMURA, Y. see KAWACHI, Y. OH, J., KIM, N., SEO, S. & KIM, I-H. Alteration of extracellular matrix NAKAMURA, Y. see SATO, H. modulators after nonablative laser therapy in skin rejuvenation, NAKANISHI, G. see AOCHI, S. 306 NAKANO, A. see NAKANO, H. OH, S-W., KANG, T-W., KIM, Y.C. & LEW, W. Coccygeal polypoid NAKANO, H., TOYOMAKI, Y., OHASHI, S., NAKANO, A., JIN, H., eccrine naevus, 614 MUNAKATA, T., AKITA, N., TAMAI, K. & MITSUHASHI, Y. OH, S-W., LEE, J.S., KIM, M.Y. & KIM, S-C. COL7A1 mutational Novel COL7A1 mutations in a Japanese family with transient bullous analysis in Korean patients with dystrophic epidermolysis bullosa, dermolysis of the newborn associated with pseudosyndactyly, 1260 179 OH, S-W., LEE, J.S., KIM, M.Y., CHOI, J.Y. & KIM, S-C. Recessive NAKASHIMA, H. see KUWANO, Y. dystrophic epidermolysis bullosa associated with dilated NALDI, L. see GOMEZ LIRA, M. cardiomyopathy, 610 NALDI, L. see SIDOROFF, A. OHASHI, S. see NAKANO, H. NANDA, S. see GROVER, C. OHM KYVIK, K. see LERBAEK, A. NASHAN, D. see VOLZ, A. OHNO, Y. see KUWANO, Y. NASORRI, F. see BEDINI, C. OHTSUKA, T. see FUJIKURA, M. NATARAJAN, S. see WAHIE, S. OKA, M., KAMO, T., SASAKI, E., KAJI, H., NISHIZAWA, H., NEDOSZYTKO, B., SZCZERKOWSKA-DOBOSZ, A., ZABŁOTNA, M., IMANISHI, Y. & NISHIGORI, C. A case of phosphaturic GLEN´ , J., RE˛BAŁA, K. & ROSZKIEWICZ, J. Associations of mesenchymal tumour (mixed connective tissue variant) promoter region polymorphisms in the tumour necrosis factor-a that developed in the subcutaneous tissue of a patient with gene and early-onset psoriasis vulgaris in a northern Polish oncogenic osteomalacia and produced fibroblast growth factor 23, population, 165 198 NEEDHAM, S.J. see BROWN, S.J. OKA, M. Pyoderma gangrenosum and interleukin 8, 1279 NERI, I., SAVOIA, F., GIACOMINI, F. & PATRIZI, A. Anonychia, OKOCHI, H. see KUWANO, Y. hyponychia and spontaneous amputation of the distal phalanges as OKUMURA, K. see ZHENG, Y. a consequence of ischaemic necrosis of the extremities after OLIVER, H., FERGUSON, J. & MOSELEY, H. Quantitative risk umbilical catheterization, 1299 assessment of sunbeds: impact of new high power lamps, 350 NERY, J.A.C. see MOURA, D.F. OLIVIERO, M. see BRAUN, R.P. NEUBURG, M. see OTLEY, C.C. OLIVRY, T., PAPS, J.S., BIZIKOVA, P., MURPHY, K.M., JACKSON, NEWTON-BISHOP, J.A. see STRAUSS, R.M. H.A. & ZEBALA, J. A pilot open trial evaluating the efficacy of NG, P.C. see HON, K.L.E. low-dose aminopterin in the canine homologue of human atopic NG, P-C. see HON, K-L.E. dermatitis, 1040 NG, S.Y. & WILKINSON, J. A salutary case of Fumaderm potentially OMOTO, M. see NISHIWAKI, F. masking the symptoms of bowel cancer and partial bowel ONDER, M. see ALPSOY, E. obstruction, 825 ORAICHI, D. see AZOULAY, L. NICHOLSON, J. see PONNAMPALAM, J. ORANJE, A.P., GLAZENBURG, E.J., WOLKERSTORFER, A. & NICOL, C. see BARBAROT, S. DE WAARD-VAN DER SPEK, F.B. Practical issues on interpretation NIEHUES, T. see ANTAL, A. of scoring atopic dermatitis: the SCORAD index, objective NIIMI, N. see KAMEYOSHI, Y. SCORAD and the three-item severity score, 645 NIKONOVA, S.M. see BELOUSOVA, I.E. ORCHARD, G. see MAO, X. NISHIGORI, C. see OKA, M. ORLOW, S.J. see HELLER, M. NISHIKAWA, A. see TAKAHATA, Y. O’MORAIN, C. see MALIK, M. NISHIKAWA, M., TANIOKA, M., ARAKI, E., YAMAMOTO, T., ORTEGO-CENTENO, N. see RIOS-FERNA´NDEZ, R. SAKURAI, T., MIYACHI, Y. & UTANI, A. Primary essential cutis ORTEU, C.H., JANSEN, T., LIDOVE, O., JAUSSAUD, R., HUGHES, verticis gyrata with hyaluronic acid deposition, 806 D.A., PINTOS-MORELL, G., RAMASWAMI, U., PARINI, R., NISHIMURA, K. see KUROKAWA, I. SUNDER-PLASSMAN, G., BECK, M. & MEHTA, A.B. ON BEHALF NISHIOKA, K. see UGAJIN, T. OF THE FOS INVESTIGATORS. Fabry disease and the skin: data NISHIWAKI, F., MATSUMURA, Y., MORITA, N., KORE-EDA, S., from FOS, the Fabry outcome survey, 331 MIYACHI, Y. & OMOTO, M. Acrodermatitis continua of Hallopeau ORTOMBINA, M. see GOMEZ LIRA, M. due to oral terbinafine, 1073 ORTON, D.I. A clinical assessment of a patch test kit marketed to U.K. NISHIZAWA, H. see OKA, M. hairdressers for detecting hair dye allergy, 1017 NIYONSABA, F. see ZHENG, Y. ORTONNE, J.P. see MURRELL, D.F. VAN NOESEL, C.J.M. see BOLLING, M.C. OTA, Y. see YAMADA, A. NOIESEN, E., MUNK, M.D., LARSEN, K., HØYEN, M. & AGNER, T. OTLEY, C.C., GRIFFIN, M.D., CHARLTON, M.R., EDWARDS, B.S., Use of complementary and alternative treatment for allergic NEUBURG, M. & STASKO FOR THE REDUCTION OF contact dermatitis, 301 IMMUNOSUPPRESSION TASK FORCE OF THE INTERNATIONAL NORGAUER, J. see KAATZ, M. TRANSPLANT SKIN CANCER COLLABORATIVE. Reduction of immunosuppression for transplant-associated skin cancer: ODIN, F. see SAUVAIGO, S. thresholds and risks, 1183 OFFIDANI, A. see CAMPANATI, A. OTSUKA, F. see KAWACHI, Y.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1317

OTSUKA, F. see SATO, H. POMERANTZ, R.G., HUSNI, M.E., MODY, E. & QURESHI, A.A. OTTAVIANI, M. see CAPITANIO, B. Adalimumab for treatment of pyoderma gangrenosum, 1274 OYAMADA, Y. see FUJIKURA, M. PONNAMPALAM, J., NICHOLSON, J. & ATHERTON, D. Leptomeningeal melanoma and multiple cutaneous melanocytic PADILLA, R.S. see ZLOTOFF, B.J. naevi, 397 PALACIO ALLER, L. see HERAS MULERO, C. POTTHOFF, A. see KREUTER, A. PALIT, A. see ARITA, K. POUMAY, Y. see SERCU, S. PALLER, A.S. see SPERGEL, J.M. POURREYRON, C. see ARITA, K. PALMER, C.N.A. see LERBAEK, A. POWELL, F.C. see LACEY, N. PANASITI, V. see ROSSI, A. PRIPP, C.M. see EKQVIST, S. PANELIUS, J. see JA¨RVINEN, T.M. PROBERT, C.S. see BROOKLYN, T.N. PAPARO, F. see ESPOSITO, G. PUDDU, P. see CIANCHINI, G. PAPOUTSAKI, M. see COSTANZO, A. PUGNALONI, A. see TUCCI, M.G. PAPP, K. see KRISHNAN, R. PUIG, L. see GARCIA-NAVARRO, X. PAPS, J.S. see OLIVRY, T. PUIG, S. see ZABALLOS, P. PARERA, E., TOLL, A., GALLARDO, F., BELLOSILLO, B., PUJOL, R.M. PUJOL, R.M. see PARERA, E. & MARTI´, R. Lichen sclerosus et atrophicus-like lesions in mycosis PULINI, S. see GOTERI, G. fungoides, 411 PURVIS, D.J., RAMIREZ, A., ROBERTS, N. & HARPER, J.I. PARIKH, A. see LLOYD, M.S. Gomez–Lopez–Hernandez syndrome: another consideration in PARINI, R. see ORTEU, C.H. focal congenital alopecia, 196 PARK, D. see LEE, J. PARK, J.K. see KIM, H.J. QU, T. see ZUO, Y-G. PARNEIX-SPAKE, A. see SPERGEL, J.M. QURESHI, A.A. see MODY, E. PARRAVICINI, C. see ANTINORI, S. QURESHI, A.A. see POMERANTZ, R.G. PARSLEW, R.A.G. see WEST, E.A. PAS, H.H. see BOLLING, M.C. VAN DER RAAIJ-HELMER, E.M.H. see ARABATZIS, M. PASPARAKIS, D. see MOUZOPOULOS, G. RABINOVITZ, H.S. see BRAUN, R.P. PATEL, V. see KRISHNAN, R. RACZ, P. see KREUTER, A. PATRIZI, A. see NERI, I. RADY, P.L., DE OLIVEIRA, W.R.P., HE, Q., FESTA, C., RIVITTI, E.A., PATRONE, P. see STINCO, G. TUCKER, S.B. & TYRING, S.K. Novel homozygous nonsense PAUL, C. see FAGUER, S. TMC8 mutation detected in patients with epidermodysplasia PAUL, C.F. see MURRELL, D.F. verruciformis from a Brazilian family, 831 PAUS, R. see TIEDE, S. RAGHAVAN, S. see TORRES, A. PENNACCHIA, W. see ESPOSITO, G. RAJAN, N. & LANGTRY, J.A.A. The punch and graft technique: a PERRINAUD, A., GAIDE, O., FRENCH, L.E., SAURAT, J-H., novel method of surgical treatment for chondrodermatitis MARGHOOB, A.A. & BRAUN, R.P. Can automated dermoscopy nodularis helicis, 744 image analysis instruments provide added benefit for the RAMASWAMI, U. see ORTEU, C.H. dermatologist? A study comparing the results of three systems, RAMIREZ, A. see PURVIS, D.J. 926 RAMOS-POLO, E. see COTO-SEGURA, P. PESERICO, A. see GISONDI, P. RANKI, A. see GNIADECKI, R. PETIT-TEIXEIRA, E. see MEJRI, K. RANKI, A. see JA¨RVINEN, T.M. PETRICOIN, E.F. see COWEN, E.W. RAVN, H. see LERBAEK, A. PFISTER, H. see KREUTER, A. RE˛BAŁA, K. see NEDOSZYTKO, B. PHAN, A., TOUZET, S., DALLE, S., RONGER-SAVLE´, S., BALME, B. & REDDY, B.S.N. see GROVER, C. THOMAS, L. Acral lentiginous melanoma: histopathological RE´GNIER, S., DUPIN, N., LE DANFF, C., WASSEF, M., ENJOLRAS, O. prognostic features of 121 cases, 311 & ARACTINGI, S. Endothelial cells in infantile haemangiomas PIASERICO, S. see GISONDI, P. originate from the child and not from the mother (a fluorescence PIASERICO, S. see GOMEZ LIRA, M. in situ hybridization-based study), 158 PICARDO, M. see CAPITANIO, B. REIFENBERGER, J. see ANTAL, A. PICCININI, G. see GOTERI, G. REITER, H. see GRIMS, R.H. PICCIRILLO, A. see ESPOSITO, G. REMUZZI, G. see GOMEZ LIRA, M. PICCIRILLO, F. see STINCO, G. REVERDY, M-E. see DURUPT, F. PIEPENBRING, M., CA´CERES MENDEZ, O.A., ESPINO ESPINOZA, A.A., REYNOLDS, N.J. see HAMPTON, P.J. KIRSCHNER, R. & SCHO¨ FER, H. Chromoblastomycosis caused by RIBEIRO-CARVALHO, M.M. see MOURA, D.F. Chaetomium funicola: a case report from Western Panama, 1025 RIBUFFO, D. see GRADILONE, A. PIETTE, F. see VERCAMBRE-DARRAS, S. RICOTTI, G. see TUCCI, M.G. PINTOS-MORELL, G. see ORTEU, C.H. RIEGER, A. see HANDISURYA, A. PIOLINI, R. see ANTINORI, S. RIGBY, H. see WEST, E.A. DE PITA`,O.see BEDINI, C. RIGOPOULOS, D. see GREGORIOU, S. POBLETE-GUTIE´RREZ, P. see ME´NDEZ, M. RIOS-FERNA´NDEZ, R., GUTIERREZ-SALMERO´ N, M.T., CALLEJAS- POCH, E. see MORENO-ROMERO, J.A. RUBIO, J-L., FERNA´NDEZ-PUGNAIRE, M. & ORTEGO-CENTENO, N. POHLA-GUBO, G. see SADLER, E. Late-onset neutropenia following rituximab treatment in patients POLLARD, A.M. see WILSON, A.S. with autoimmune diseases, 1271

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1318 Author index

RIVITTI, E.A. see RADY, P.L. SARDANA, K. see BANSAL, S. ROBAK, E., JESIONEK-KUPNICKA, D., ROBAK, T., HOLUB, A., SARNO, E.N. see MOURA, D.F. WAWRZYNIAK, E., BARTKOWIAK, J., BEDNAREK, A., SASAKI, E. see OKA, M. CONSTANTINU, M. & URBANSKA-RYS, H. Primary cutaneous SASAKI, T. see HATAMOCHI, A. marginal zone B-cell lymphoma in a patient with chronic SATO, H., NAKAMURA, Y., TAKAHASHI, T. & OTSUKA, F. lymphocytic leukaemia, 591 Concordant lymphoma of cutaneous anaplastic large cell ROBAK, T. see ROBAK, E. lymphoma and systemic B-cell leukaemia, 1060 ROBERTS, C. see ANGUS, J. SATO, T. see UGAJIN, T. ROBERTS, N. see PURVIS, D.J. SATOH, T. see YAHARA, H. RODERO, J. see ZABALLOS, P. SAURAT, J-H. see BRAUN, R.P. RODRI´GUEZ-CANO, L. see HERAS MULERO, C. SAURAT, J-H. see PERRINAUD, A. ROE, E. see GARCIA-NAVARRO, X. SAUVAIGO, S., BONNET-DUQUENNOY, M., ODIN, F., HAZANE- ROGERS, S. see AHMAD, K. PUCH, F., LACHMANN, N., BONTE´, F., KURFU¨ RST, R. & FAVIER, RO¨ GLIN, J. & BO¨ ER, A. Skin manifestations of intravascular lymphoma A. DNA repair capacities of cutaneous fibroblasts: effect of sun mimic inflammatory diseases of the skin, 16 exposure, age and smoking on response to an acute oxidative ROMITI, R. see TAKAHASHI, M.D.F. stress, 26 RONGER-SAVLE´,S.see PHAN, A. SAVOIA, F. see NERI, I. RORTVEIT, G. see RØRTVEIT, S. SAWADA, R. see KIDO, M. RØRTVEIT, S. & RORTVEIT, G. Impetigo in epidemic and nonepidemic SAWAI, T. see WATABE, D. phases: an incidence study over 4 years in a general population, 100 SAWAMURA, D. see YASUKAWA, K. ROSEEUW, D. see VERFAILLE, C.J. SCARPA, S. see GRADILONE, A. ROSS, O.K. see HAMPTON, P.J. SCHAFFER, J.V. see HELLER, M. ROSSI, A., DEVIRGILIIS, V., PANASITI, V., BORRONI, R.G., SCHAFLEITNER, B. see SADLER, E. CARLESIMO, M., GENTILE, M., CARIOLA, F. & CALVIERI, S. SCHAGEN VAN LEEUWEN, J.H. see SLEE, P.H.T.J. Missense mutation in exon 7 of TRPS1 gene in an Italian family SCHEPER, R.J. see LE DUC, Q. with a mild form of trichorhinophalangeal syndrome type I, 1021 SCHIANCHI, S. see GISONDI, P. ROSZKIEWICZ, J. see NEDOSZYTKO, B. SCHMERMUND, D. see LO¨ FFLER, H. ROTT, S., KU¨ STER, R.M. & MROWIETZ, U. Successful treatment of SCHMITT, S. see KILIC, S.S. severe psoriatic arthritis with infliximab in an 11-year-old child SCHO¨ FER, H. see PIEPENBRING, M. suffering from linear psoriasis along lines of Blaschko, 191 SCHO¨ N, M.P. see WACHTER, T. ROUJEAU, J-C. see SIDOROFF, A. SCHUR, P. see MODY, E. ROWBOTTOM, A.W. see SHAH, D. SCHWANDT, P. see GNIADECKI, R. RUBEGNI, P. see MOURMOURAS, V. SCHWARZ, T. see HU¨ GEL, R. RUBINI, C. see GOTERI, G. SCORTECHINI, A.R. see GOTERI, G. RUDNAI, P. see WILKENING, S. SCUDERI, N. see GRADILONE, A. RUGIU, C. see GOMEZ LIRA, M. SEGURA, S. see MORENO-ROMERO, J.A. RUPEC, R.A. see FLAIG, M.J. SEISHIMA, M., MIZUTANI, Y., SHIBUYA, Y., ARAKAWA, C., RUPOLI, S. see GOTERI, G. YOSHIDA, R. & OGATA, T. Malignant melanoma in a woman with RUSSELL-JONES, R. see MAO, X. LEOPARD syndrome: identification of a germline PTPN11 mutation RUSTIN, M.H.A. The safety of tacrolimus ointment for the treatment and a somatic BRAF mutation, 1297 of atopic dermatitis: a review, 861 SEITZ, C.S., VAN STEENSEL, M., FRANK, J., SENDEREK, J., ZERRES, RUZICKA, T. see ANTAL, A. K., HAMM, H. & BERGMANN, C. The Wnt signalling ligand RYLANDER, E. see JOHANNESSON, U. RSPO4, causing inherited anonychia, is not mutated in a patient with congenital nail hypoplasia ⁄aplasia with underlying skeletal SAARIALHO-KERE, U. see JA¨RVINEN, T.M. defects, 801 SADLER, E., SCHAFLEITNER, B., LANSCHUETZER, C., LAIMER, M., SEITZ, C.S. see BENOIT, S. POHLA-GUBO, G., HAMETNER, R., HINTNER, H. & BAUER, J.W. SELDENRIJK, C.A. see SLEE, P.H.T.J. Treatment-resistant classical epidermolysis bullosa acquisita SENDEREK, J. see SEITZ, C.S. responding to rituximab, 417 SENFF, N.J. & WILLEMZE, R. The applicability and prognostic value of SAKURAI, T. see NISHIKAWA, M. the new TNM classification system for primary cutaneous SALAT, A. see HANDISURYA, A. lymphomas other than mycosis fungoides and Se´zary syndrome: SALDANHA, G. see MACBETH, A.E. results on a large cohort of primary cutaneous B-cell lymphomas SALSENCH, E. see ZABALLOS, P. and comparison with the system used by the Dutch Cutaneous SAMPAIO, E.P. see MOURA, D.F. Lymphoma Group, 1205 SAMPOGNA, F., TABOLLI, S., ABENI, D. & THE IDI MULTIPURPOSE SEO, S. see OH, J. PSORIASIS RESEARCH ON VITAL EXPERIENCES (IMPROVE) SEO, S-H., OH, C-K., KWON, K-S. & KIM, M-B. A case of milium- INVESTIGATORS. The impact of changes in clinical severity on like syringoma with focal calcification in Down syndrome, 612 psychiatric morbidity in patients with psoriasis: a follow-up study, SEO, Y.J. see KIM, H.J. 508 SEPP, N. see EISENDLE, K. SAND, M. see BECHARA, F.G. SERCU, S., POUMAY, Y., HERPHELIN, F., LIEKENS, J., BEEK, L., SANDERS, D.S.A. see TAIBJEE, S.M. ZWIJSEN, A., WESSAGOWIT, V., HUYLEBROECK, D., MCGRATH, SANTOS, I.M.C.F. see MOURA, D.F. J.A. & MERREGAERT, J. Functional redundancy of extracellular SANTOS-JUANES, J. see COTO-SEGURA, P. matrix protein 1 in epidermal differentiation, 771

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1319

SETSU, N. see AOCHI, S. SMITH, A. see MACBETH, A.E. SETTERFIELD, J. see ESCUDIER, M. SMITH, A. see TAIBJEE, S.M. SEYHAN, M. see ALPSOY, E. SMITH, G., WILKIE, M.J.V., DEENI, Y.Y., FARR, P.M., FERGUSON, J., SHAH, D., ROWBOTTOM, A.W., THOMAS, C.L., CUMBER, P. & WOLF, C.R. & IBBOTSON, S.H. Melanocortin 1 receptor (MC1R) CHOWDHURY, M.M.U. Hypocomplementaemic urticarial genotype influences erythemal sensitivity to psoralen–ultraviolet A vasculitis associated with non-Hodgkin lymphoma and treatment photochemotherapy, 1230 with intravenous immunoglobulin, 392 SMITH, V.H. & JONES, S.K. Successful treatment of severe psoriatic SHANAHAN, F. see MALIK, M. natal cleft fissuring with tissue adhesive, 1269 SHAW, H.M. & THOMPSON, J.F. Prognostic value of sentinel lymph SOMA, Y. see KAWAKAMI, T. node biopsy, 405 SONG, W-C., HU, K-S., KIM, H-J. & KOH, K-S. A study of the SHAW, J., HUGHES, C.M., LAGAN, K.M. & BELL, P.M. The clinical secretion mechanism of the sebaceous gland using three- effect of topical phenytoin on wound healing: a systematic review, dimensional reconstruction to examine the morphological 997 relationship between the sebaceous gland and the arrector pili SHAW, J., LIU, C., MARTIN, R., CHEN, B., HOLTZ, R., HUANG, muscle in the follicular unit, 325 W-H. & LEE, A-R. Inhibition of tumour necrosis factor-a secretion SONTHEIMER, R.D. see GERAMI, P. from EpiDermTM tissues by a novel small molecule, UTL-5d, 575 SOUTH, A.P. see ARITA, K. SHEAR, N.H. see TOHYAMA, M. SOWDEN, H.M., KAROO, R.O.S. & TOBIN, D.J. Transforming growth SHETH, N., GREENBLATT, D. & MCGRATH, J.A. New KRT10 gene factor-b receptor II is preferentially expressed in the companion mutation underlying the annular variant of bullous congenital layer of the human anagen hair follicle, 161 ichthyosiform erythroderma with clinical worsening during SOYER, H.P. see ARGENZIANO, G. pregnancy, 602 SPAAPEN, L.J.M. see STEIJLEN, P.M. SHEU, H-M. see LEE, Y-L. SPERGEL, J.M., BOGUNIEWICZ, M., PALLER, A.S., HEBERT, A.A., SHIBATA, S. see IMAFUKU, S. GALLAGHER, P.R., MCCORMICK, C., PARNEIX-SPAKE, A. & SHIBUYA, Y. see SEISHIMA, M. HULTSCH, T. Addition of topical pimecrolimus to once-daily SHIMADA, H. see HAMADA, T. mid-potent steroid confers no short-term therapeutic benefit in the SHIMIZU, H. see HOSHINA, D. treatment of severe atopic dermatitis; a randomized controlled SHIMIZU, H. see YAOSAKA, M. trial, 378 SHIMIZU, H. see YASUKAWA, K. SPIES, M. see AUST, M.C. SHIN, H. see YU, H-J. STALDER, J.F. see BARBAROT, S. SHIN, H.T. see HELLER, M. STALDER, J-F. see VOURC’H, M. SHIN, M-G. see YUN, S.J. STAMATAKOS, M. see MOUZOPOULOS, G. SHIRLAW, P. see ESCUDIER, M. STASKO, T. see OTLEY, C.C. SHUM, K.W. see ATHAVALE, P. STAVRAKOGLOU, A., BROWN, V.L. & COUTTS, I. Successful SIDOROFF, A., DUNANT, A., VIBOUD, C., HALEVY, S., BOUWES treatment of primary cutaneous follicle centre lymphoma with BAVINCK, J.N., NALDI, L., MOCKENHAUPT, M., FAGOT, J-P. & topical 5% imiquimod, 620 ROUJEAU, J-C. Risk factors for acute generalized exanthematous STEADMAN, M.C. see MADKAN, V.K. pustulosis (AGEP)—results of a multinational case–control study VAN STEENSEL, M. see SEITZ, C.S. (EuroSCAR), 989 VAN STEENSEL, M.A.M. see SLEE, P.H.T.J. SIDOU, F. see BARAN, R. VAN STEENSEL, M.A.M. see STEIJLEN, P.M. SIGURGEIRSSON, B. see BARAN, R. STEIJLEN, P.M., VAN GEEL, M., VREEBURG, M., MARCUS- SILVESTRI, I. see GRADILONE, A. SOEKARMAN, D., SPAAPEN, L.J.M., CASTELIJNS, F.C.M., SIMA, R. see BELOUSOVA, I.E. WILLEMSEN, M. & VAN STEENSEL, M.A.M. Novel EBP SIMON, D., BRAATHEN, L.R. & SIMON, H-U. Increased gene mutations in Conradi–Hu¨nermann–Happle syndrome, lipopolysaccharide-induced tumour necrosis factor-a, interferon-c 1225 and interleukin-10 production in atopic dermatitis, 583 STEINBERG, S.M. see COWEN, E.W. SIMON, H-U. see SIMON, D. STEMMLER, S. see HOFFJAN, S. SIMONETTI, O. see CAMPANATI, A. STERLING, J.C. see TOMSON, N. SIMONETTI, O. see GOTERI, G. STERRY, W., STROBER, B.E. & MENTER, A. ON BEHALF OF THE SINAGRA, J.L. see CAPITANIO, B. INTERNATIONAL PSORIASIS COUNCIL. Obesity in psoriasis: the SINCLAIR, R., GREENLAND, K.J., VAN EGMOND, S., HOEDEMAKER, metabolic, clinical and therapeutic implications. Report of an C., CHAPMAN, A. & ZAJAC, J.D. Men with Kennedy disease have a interdisciplinary conference and review, 649 reduced risk of androgenetic alopecia, 290 STICHERLING, M. see VON MACKENSEN, Y.A. SINGH, J. see GUPTA, A.K. STINCO, G., PICCIRILLO, F. & PATRONE, P. Hypertriglyceridaemia SKOVGAARD, L.T. see MOGENSEN, M. during treatment with adalimumab in psoriatic arthritis, SKOWRON, F. see BARBERIO, E. 1273 SLADDEN, C.S. see SLADDEN, M.J. STINGL, G. see HANDISURYA, A. SLADDEN, M.J. & SLADDEN, C.S. Maximizing the quality of review STOKKERMANS-DUBOIS, J., BEYLOT-BARRY, M., VERGIER, B., articles, 409 BOUABDALLAH, K. & DOUTRE, M.S. Erythema annulare SLADE, H.B. see TORRES, A. centrifugum revealing chronic lymphocytic leukaemia, 1045 SLEE, P.H.T.J., VAN DER WAAL, R.I.F., SCHAGEN VAN LEEUWEN, STONE, N.M. see HANN, S. J.H., TUPKER, R.A., TIMMER, R., SELDENRIJK, C.A. & VAN STOREY, L. see TORRES, A. STEENSEL, M.A.M. Paraneoplastic hypertrichosis lanuginosa STRAMAZZOTTI, D. see CAMPANATI, A. acquisita: uncommon or overlooked?, 1087 STRAMAZZOTTI, D. see GOTERI, G.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1320 Author index

STRAUSS, R.M., ELLIOTT, F., AFFLECK, P., BOON, A.P. & NEWTON- TEIXEIRA, V.H. see MEJRI, K. BISHOP, J.A. A retrospective study addressed to understanding TELES, R.B. see MOURA, D.F. what predicts severe histological dysplasia ⁄early melanoma in TELES, R.M.B. see MOURA, D.F. excised atypical melanocytic lesions, 758 TEMPLETON, K. see ARABATZIS, M. STREET, A.C. see TUXEN, A.J. TENNER-RACZ, K. see KREUTER, A. STROBER, B.E. see BROWNELL, I. TESNIERE, A. see MOLE`S, J.-P. STROBER, B.E. see STERRY, W. TESSARI, G. see GISONDI, P. STRONATI, A. see GOTERI, G. TESSARI, G. see GOMEZ LIRA, M. SUBCOMMITTEE, A. see GRATTAN, C.E.H. THACI, D. see BOEHNCKE, S. SUGITA, T. see TAKAHATA, Y. THOMAS, C.L. & FINLAY, A.Y. The ‘handprint’ approximates to 1% of SUMMERBELL, R.C. see ARABATZIS, M. the total body surface area whereas the ‘palm minus the fingers’ SUNDER-PLASSMAN, G. see ORTEU, C.H. does not, 1080 SUNG, F-C. see LEE, Y-L. THOMAS, C.L. see SHAH, D. SUNG, Y.T. see HON, K.L.E. THOMAS, L. see BARBERIO, E. SUZUKI, D. see AOCHI, S. THOMAS, L. see DALLE, S. SUZUKI, H. see HATAMOCHI, A. THOMAS, L. see DEBARBIEUX, S. SUZUKI, N. see AOCHI, S. THOMAS, L. see DURUPT, F. SVEDMAN, C. see EKQVIST, S. THOMAS, L. see PHAN, A. SYED, S. see MAZEREEUW-HAUTIER, J. THOMASSEN, L. see HOLM, E.A. SZAKONYI, J. see MARSCHALKO´ ,M. THOMPSON, J.F. see SHAW, H.M. SZCZERKOWSKA-DOBOSZ, A. see NEDOSZYTKO, B. THOMSEN, J.B. see MOGENSEN, M. SZEIMIES, R-M. see BABILAS, P. THOMSON, M.A. & MOSS, C. Pityriasis rubra pilaris in a mother and SZIGETI, A´. see MARSCHALKO´ ,M. two daughters, 202 TIEDE, S., KLOEPPER, J.E., WHITING, D.A. & PAUS, R. The ‘follicular TABOLLI, S. see SAMPOGNA, F. trochanter’: an epithelial compartment of the human hair follicle TADINI, G. see ESPOSITO, G. bulge region in need of further characterization, 1013 TAIBJEE, S.M., GEE, B.C., SANDERS, D.S.A., SMITH, A. & CARR, R.A. TIEV, K.P. see FARDET, L. Lentigo maligna involving the tumour nests and stroma of a TIMMER, R. see SLEE, P.H.T.J. nodular basal cell carcinoma, 184 TING, P.T. see DYTOC, M.T. TAI¨EB, A. see MORICE-PICARD, F. TOBIN, A.-M. see MALIK, M. TAKAHAGI, S. see KAMEYOSHI, Y. TOBIN, D.J. see SOWDEN, H.M. TAKAHASHI, M.D.F., CASTRO, L.G.M. & ROMITI, R. Infliximab, as TOBIN, D.J. see WILSON, A.S. sole or combined therapy, induces rapid clearing of erythrodermic TOHYAMA, M., HASHIMOTO, K., YASUKAWA, M., KIMURA, H., psoriasis, 828 HORIKAWA, T., NAKAJIMA, K., URANO, Y., MATSUMOTO, K., TAKAHASHI, T. see KAWACHI, Y. IIJIMA, M. & SHEAR, N.H. Association of human herpesvirus 6 TAKAHASHI, T. see SATO, H. reactivation with the flaring and severity of drug-induced TAKAHATA, Y., SUGITA, T., HIRUMA, M. & MUTO, M. Quantitative hypersensitivity syndrome, 934 analysis of Malassezia in the scale of patients with psoriasis using a TOKSOY, A., MU¨ LLER, V., GILLITZER, R. & GOEBELER, M. Biphasic real-time polymerase chain reaction assay, 670 expression of stromal cell-derived factor-1 during human wound TAKAHATA, Y., SUGITA, T., KATO, H., NISHIKAWA, A., HIRUMA, healing, 1148 M. & MUTO, M. Cutaneous Malassezia flora in atopic dermatitis TOLE´DANO, C. see FARDET, L. differs between adults and children, 1178 TOLL, A. see PARERA, E. TAKEUCHI, S. see KIDO, M. TOMI, N.S. see BECHARA, F.G. TALKS, S.J. see BROWN, S.J. TOMSON, N. & STERLING, J.C. Hydroxycarbamide: a treatment for TAMAI, K. see MABUCHI, E. lichen sclerosus?, 622 TAMAI, K. see NAKANO, H. VAN TONGEREN, M. see TURNER, S. TAMAKI, K. see ISHIURA, N. TORRES, A., STOREY, L., ANDERS, M., MILLER, R.L., BULBULIAN, TAMAKI, K. see KUWANO, Y. B.J., JIN, J., RAGHAVAN, S., LEE, J., SLADE, H.B. & BIRMACHU, TAN, A.W.H. see TAN, H-H. W. Microarray analysis of aberrant gene expression in actinic TAN, H-H., TAN, A.W.H., BARKHAM, T., YAN, X-Y. & ZHU, M. keratosis: effect of the Toll-like receptor-7 agonist imiquimod, Community-based study of acne vulgaris in adolescents in 1132 Singapore, 547 TOUZET, S. see PHAN, A. TANAKA, T. see KAMEYOSHI, Y. TOYOMAKI, Y. see NAKANO, H. TANG, W.Y.M. see CHAN, K.H.N. TRON, F. see MEJRI, K. TANIOKA, M. see NISHIKAWA, M. TROVATI, S. see ANTINORI, S. TAPPUNI, A. see ESCUDIER, M. TSAI, T.F. see LIAO, Y.H. TARIQ, M., WASIF, N. & AHMAD, W. A novel deletion mutation in TSAI, T-F. see HONG, J-B. the EDAR gene in a Pakistani family with autosomal recessive TSAI, Y-S., TU, M-E., WU, Y-H. & LIN, Y-C. Hydroxyzine-induced hypohidrotic ectodermal dysplasia, 207 acute generalized exanthematous pustulosis, 1296 TASHIRO, A. see IMAFUKU, S. TSAO, H. & ENGLISH, J. ‘Benchside-to-bedside’, 849 TASSETTI, A. see GOTERI, G. TSENG, Y.S. see LIAO, Y.H. TAYLOR, A. see WAHIE, S. TSOUPAROPOULOS, V. see MOUZOPOULOS, G. TAYLOR, A.E.M. see BROWN, S.J. TSUBURA, A. see KUROKAWA, I.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1321

TSUJI, K. see YAMADA, A. VERCAMBRE-DARRAS, S., DUBUCQUOI, S., FAJARDY, I., PIETTE, F. & TSUTSUMI, M. & DENDA, M. Paradoxical effects of b-estradiol on MORTIER, L. Does spontaneous autoimmunity improve survival in epidermal permeability barrier homeostasis, 776 visceral metastatic melanoma?, 413 TSUTSUMI, M. see DENDA, M. VERFAILLE, C.J., COEL, M., BOERSMA, I.H., MERTENS, J., BORGERS, TU, M-E. see TSAI, Y-S. M. & ROSEEUW, D. Oral R115866 in the treatment of moderate TUCCI, M.G., LUCARINI, G., BRANCORSINI, D., ZIZZI, A., to severe facial acne vulgaris: an exploratory study, 122 PUGNALONI, A., GIACCHETTI, A., RICOTTI, G. & BIAGINI, G. VERGIER, B. see STOKKERMANS-DUBOIS, J. Involvement of E-cadherin, b-catenin, Cdc42 and CXCR4 in the VERNON, S.E. see LUGASSY, C. progression and prognosis of cutaneous melanoma, 1212 VEROLA, O. see VIGUIER, M. TUCKER, S.B. see RADY, P.L. VETTER-KAUCZOK, C.S. see BENOIT, S. TU¨ LIN GU¨ LE¸,A.see ADA, S. VIBOUD, C. see SIDOROFF, A. TUPKER, R.A. see SLEE, P.H.T.J. VIGUIER, M., JEANMOUGIN, M., BEGON, E., VEROLA, O., TURCO, A. see GOMEZ LIRA, M. DUBERTRET, L. & BACHELEZ, H. Remission of photosensitivity TURKI, H. see MEJRI, K. following treatment of psoriasis vulgaris with tumour necrosis TURNER, R. see LALLY, A. factor inhibitors, 625 TURNER, S., CARDER, M., VAN TONGEREN, M., MCNAMEE, R., VINCENZI, B. see GRADILONE, A. LINES, S., HUSSEY, L., BOLTON, A., BECK, M.H., WILKINSON, M. VIOLA, A. see ESPOSITO, G. & AGIUS, R. The incidence of occupational skin disease as VIVES, J.M. see ZABALLOS, P. reported to The Health and Occupation Reporting (THOR) VOGT, P.M. see AUST, M.C. network between 2002 and 2005, 713 VOLTEAU, C. see BARBAROT, S. TURSEN, U. see ALPSOY, E. VOLZ, A., METZE, D., BO¨ HM, M., BRUCKNER-TUDERMAN, L. & TU¨ TING, T. see WENZEL, J. NASHAN, D. Idiopathic eruptive macular pigmentation in a TUXEN, A.J., YONG, M.K., STREET, A.C. & DOLIANITIS, C. 7-year-old girl: case report and discussion of differences from Disseminated cryptococcal infection in a patient with severe erythema dyschromicum perstans, 839 psoriasis treated with efalizumab, methotrexate and ciclosporin, VONDERHEID, E.C. see COWEN, E.W. 1067 VOORHEES, J.J. see HELFRICH, Y.R. TYRING, S.K. see MADKAN, V.K. VOURC’H, M., BARBAROT, S. & STALDER, J-F. Ciclosporin improves TYRING, S.K. see RADY, P.L. quality of life in Kimura’s disease, 420 VREEBURG, M. see STEIJLEN, P.M. UGAJIN, T., YAHARA, H., MORIYAMA, Y., SATO, T., NISHIOKA, K. & YOKOZEKI, H. Two siblings with neonatal pemphigus vulgaris VAN DER WAAL, R.I.F. see SLEE, P.H.T.J. associated with mild maternal disease, 192 DE WAARD-VAN DER SPEK, F.B. see ORANJE, A.P. UJIIE, H. see YAOSAKA, M. WACHTER, T., MURACH, W.M., BRO¨ CKER, E-B. & SCHO¨ N, M.P. ULRICH, J. see BRU¨ CHER, J-J. Recalcitrant lithium-induced psoriasis in a suicidal patient ULRICH, M.M.W. see LE DUC, Q. alleviated by tumour necrosis factor-a inhibition, 627 UMEDA, K. see KUROKAWA, I. WAHIE, S., HISCUTT, E., NATARAJAN, S. & TAYLOR, A. Pityriasis UMEGAKI, N. see MABUCHI, E. lichenoides: the differences between children and adults, URABE, K. see KIDO, M. 941 URANO, Y. see TOHYAMA, M. WALEWSKA, R. see HELBLING, I. URBANSKA-RYS, H. see ROBAK, E. WALL, E.C. see HAYDOCK, S.F. USHIO, H. see ZHENG, Y. WALLING, H.W. see GERAMI, P. USTA, A. see ALPSOY, E. WALTON, S. see WILLIAMS, C. UTANI, A. see NISHIKAWA, M. WANG, B. see ZUO, Y-G. UZUN, S. see ALPSOY, E. WANG, S-H. see HONG, J-B. WASIF, N. see TARIQ, M. VACCARO, M., BORGIA, F., BARBUZZA, O. & GUARNERI, B. WASSEF, M. see RE´GNIER, S. Photodistributed eruptive telangiectasia: an uncommon adverse WATABE, D., KANNO, H., YOSHIDA, A., KUROSE, A., AKASAKA, drug reaction to venlafaxine, 822 T. & SAWAI, T. Adhesion of peripheral blood mononuclear cells VANDENESCH, F. see DURUPT, F. and CD4+ T cells from patients with psoriasis to cultured VARMA, S. see AFFLECK, A.G. endothelial cells via the interaction between lymphocyte function- VASATURO, F. see GRADILONE, A. associated antigen type 1 and intercellular adhesion molecule 1, VASSALLO, C., ARDIGO` , M., BRAZZELLI, V., ZECCA, M., LOCATELLI, 259 F., ALESSANDRINO, P.E., LAZZARINO, M., CORONA, S., WATANABE, R. see KUWANO, Y. LANZERINI, P., BENAZZO, M., FABBI, M. & BORRONI, G. WATANABE, T., YOSHIDA, Y. & YAMAMOTO, O. Angiomyo- Bartonella-related pseudomembranous angiomatous papillomatosis fibroblastoma of the vulva with a penile appearance, 189 of the oral cavity associated with allogeneic bone WAWRZYNIAK, E. see ROBAK, E. marrow transplantation and oral graft-versus-host disease, WEGER, W. see GRIMS, R.H. 174 WEISE-RICCARDI, S. see MURRELL, D.F. VA´VROVA´, K., HRABA´LEK, A., MAC-MARY, S., HUMBERT, P. & WELCH, D.R. see LUGASSY, C. MURET, P. Ceramide analogue 14S24 selectively recovers WELLER, R. see FINNEN, M.J. perturbed human skin barrier, 704 WENZEL, J., ZAHN, S., MIKUS, S., WIECHERT, A., BIEBER, T. & VELEGRAKI, A. see ARABATZIS, M. TU¨ TING, T. The expression pattern of interferon-inducible VELTEN, M. see LIPSKER, D. proteins reflects the characteristic histological distribution of

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 1322 Author index

infiltrating immune cells in different cutaneous lupus YAMADA, A., YAMASAKI, O., ASAGOE, K., TSUJI, K., HAMADA, T., erythematosus subsets, 752 OTA, Y. & IWATSUKI, K. Recovery from Se´zary syndrome WESSAGOWIT, V. see ARITA, K. following Mycobacterium avium spondylitis, 1270 WESSAGOWIT, V. see SERCU, S. YAMAMOTO, M. see ASAHINA, A. WEST, E.A., MCPARTLAND, J.L., RIGBY, H. & PARSLEW, R.A.G. Giant YAMAMOTO, O. see WATANABE, T. bathing trunk naevus with lymphadenopathy and unusual YAMAMOTO, T. see NISHIKAWA, M. pathology, 599 YAMANAKA, K-I. see KUROKAWA, I. WHATLEY, S.D. see BERROETA, L. YAMASAKI, O. see YAMADA, A. WHITE, I.R. see WHITE, J.M.L. YAMAZAKI, S. see HATAMOCHI, A. WHITE, J.M.L. & WHITE, I.R. The role of self-tests in the diagnosis of YAN, X-Y. see TAN, H-H. hair dye allergy, 847 YANG, C-H., WU, T-S. & CHIU, C-T. Chronic hepatitis B reactivation: a WHITE, J.M.L., WHITE, I.R., GLENDINNING, A., FLEMING, J., word of caution regarding the use of systemic glucocorticosteroid JEFFERIES, D., BASKETTER, D.A., MCFADDEN, J.P. & BUCKLEY, therapy, 587 D.A. Frequency of allergic contact dermatitis to isoeugenol is YANO, S. see KUWANO, Y. increasing: a review of 3636 patients tested from 2001 to 2005, YAOSAKA, M., ABE, R., UJIIE, H., ABE, Y. & SHIMIZU, H. Unilateral 580 periorbital oedema due to sarcoid infiltration of the eyelid: an WHITELEY, G. see COWEN, E.W. unusual presentation of sarcoidosis with facial nerve palsy and WHITING, D.A. see TIEDE, S. parotid gland enlargement, 200 WHITTAKER, S. see GNIADECKI, R. YASUKAWA, K., SAWAMURA, D., GOTO, M., NAKAMURA, H. & WHITTAKER, S. see LALLY, A. SHIMIZU, H. Histone deacetylase inhibitors preferentially augment WHITTAKER, S. see MAO, X. transient transgene expression in human dermal fibroblasts, WIECHERT, A. see WENZEL, J. 662 WIEDERHOLT, T. see ME´NDEZ, M. YASUKAWA, M. see TOHYAMA, M. WIELAND, U. see KREUTER, A. YASUMOTO, S. see HAMADA, T. WILKENING, S., HEMMINKI, K., RUDNAI, P., GURZAU, E., YAZAWA, N. see KUWANO, Y. KOPPOVA, K., FO¨ RSTI, A. & KUMAR, R. No association between YIP, D. see KOUBA, D.J. MDM2 SNP309 promoter polymorphism and basal cell carcinoma YOKOZEKI, H. see UGAJIN, T. of the skin, 375 YOKOZEKI, H. see YAHARA, H. WILKIE, M.J.V. see SMITH, G. YONG, M.K. see TUXEN, A.J. WILKINSON, J. see NG, S.Y. YOON, T.Y. see KIM, H.J. WILKINSON, M. see TURNER, S. YOSHIDA, A. see WATABE, D. WILLEMSEN, M. see STEIJLEN, P.M. YOSHIDA, R. see SEISHIMA, M. WILLEMZE, R. see SENFF, N.J. YOSHIDA, Y. see WATANABE, T. WILLIAMS, A.M. see BROOKLYN, T.N. YSEBAERT, L. see FAGUER, S. WILLIAMS, C., MITRA, A. & WALTON, S. ‘Turkey ear’: a diagnosis or YU, C.H. see HSU, C.K. a physical sign?, 816 YU, H-J., SHIN, H., KANG, M-S. & KIM, J-S. A case of primary WILSON, A.S., DODSON, H.I., JANAWAY, R.C., POLLARD, A.M. anetoderma in an infant, 1267 & TOBIN, D.J. Selective biodegradation in hair shafts derived YU, H-S. see LEE, Y-L. from archaeological, forensic and experimental contexts, YUN, S.J., SHIN, M-G., CHOI, C., KIM, H-J., LEE, J-B., KIM, S-J., 450 LEE, S-C. & WON, Y.H. Fatal disseminated angioinvasive Fusarium WITHEROW, R.O. see COX, G.A. falciforme infection in a patient with acute myeloid leukaemia, WOLF, C.R. see SMITH, G. 407 WOLKENSTEIN, P. see BARBAROT, S. WOLKERSTORFER, A. see ORANJE, A.P. ZABALLOS, P., BLAZQUEZ, S., PUIG, S., SALSENCH, E., RODERO, J., WON, Y.H. see YUN, S.J. VIVES, J.M. & MALVEHY, J. Dermoscopic pattern of intermediate WONG, K.Y. see HON, K.L.E. stage in seborrhoeic keratosis regressing to lichenoid keratosis: WONG, K-Y. see HON, K-L.E. report of 24 cases, 266 WONG, T.W. see HSU, C.K. ZABŁOTNA, M. see NEDOSZYTKO, B. WU, C-S. see HU, S.C-S. ZAHN, S. see WENZEL, J. WU, T-S. see YANG, C-H. ZAJAC, J.D. see SINCLAIR, R. WU, Y-H. see TSAI, Y-S. ZALAUDEK, I. see ARGENZIANO, G. WULF, H.C. see HOLM, E.A. ZAMAN, M. see GUPTA, A.K. ZEBALA, J. see OLIVRY, T. XU, Y. see ZUO, Y-G. ZECCA, M. see VASSALLO, C. XU, Y-H. see FANG, Y. ZELGER, B. see ANTAL, A. ZELGER, B. see EISENDLE, K. YAAR, M. & GILCHREST, B.A. Photoageing: mechanism, prevention ZELGER, B. see KAATZ, M. and therapy, 874 ZERRES, K. see SEITZ, C.S. YAGUE, J. see COTO-SEGURA, P. ZHENG, Y., NIYONSABA, F., USHIO, H., NAGAOKA, I., IKEDA, S., YAHARA, H., SATOH, T., HASHIMOTO, T. & YOKOZEKI, H. OKUMURA, K. & OGAWA, H. Cathelicidin LL-37 induces the Transient macular erythema with eosinophilia in a patient carrying generation of reactive oxygen species and release of human the FIP1L1-PDGFRA fusion gene, 1284 a-defensins from neutrophils, 1124 YAHARA, H. see UGAJIN, T. ZHOU, Q. see ZHU, K.J.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Author index 1323

ZHU, K.J., ZHOU, Q., LIN, A.H., LU, Z.M. & CHENG, H. The use of ZLOTOFF, B.J., BANG, R.H., PADILLA, R.S. & MORRISON, L. Cutaneous intravenous immunoglobulin in cutaneous and recurrent angiokeratoma and venous malformations in a Hispanic-American perforating intestinal Degos disease (malignant atrophic papulosis), patient with cerebral cavernous malformations, 210 206 ZUO, Y-G., XU, Y., WANG, B., LIU, Y-H., QU, T., FANG, K. & HO, ZHU, M. see TAN, H-H. M.G. A novel mutation of CYLD in a Chinese family with multiple ZIEMER, M. see KAATZ, M. familial trichoepithelioma and no CYLD protein expression in the ZITOUNI, M. see MEJRI, K. tumour tissue, 818 ZIZZI, A. see TUCCI, M.G. ZWIJSEN, A. see SERCU, S.

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1305–1323 Subject index acidified nitrite, topical application to nail, 494 allergic contact dermatitis acitretin with laser therapy in treatment of giant genital variant complementary and alternative treatment, 301 of folliculosebaceous cystic hamartoma (Correspondence), flexural, to benzalkonium chloride in antiseptic bath oil 833 (Case Report), 795 acne to isoeugenol (Concise Communication), 580 community-based study in adolescents in Singapore, alopecia 547 androgenetic, and Kennedy disease, 290 moderate to severe facial, oral R115866 therapy, 122 focal congenital, and Gomez–Lopez–Hernandez syndrome photodynamic therapy (Correspondence), 810 (Correspondence), 196 relapse, and isotretinoin therapy, 1240 scarring, in cutaneous Mycobacterium neoaurum infection smoker’s (Correspondence), 1070 (Correspondence), 204 tetracycline antibiotics and lupus erythematosus, 540 alopecia areata, serum chemokine profiles, 466 treatment with lotion containing triethyl citrate and ethyl a-defensins, cathelicidin LL-37-induced release from linoleate, 569 neutrophils, 1124 acquired palmoplantar keratoderma associated with antibodies 5-aminolaevulinic acid methylester, topical, comparison with to desmocollin-3 (Case Report), 168 photodynamic therapy in scalp actinic keratosis, 87 acral lentiginous melanoma, histopathological prognostic aminopterin, efficacy of low-dose treatment in canine features, 311 homologue of atopic dermatitis (Correspondence), 1040 acrodermatitis continua of Hallopeau due to terbinafine amorolfine, oral terbinafine compared with combination (Correspondence), 1073 of amorolfine nail lacquer and oral terbinafine in onycho- acrodermatitis enteropathica caused by novel mutation of mycosis with matrix involvement, 149 SLC39A4 gene (Gene Corner), 386 anal intraepithelial neoplasia, imiquimod treatment in HIV actinic keratosis infection, correlation of p16ink4a expression with decline of head, safety and efficacy of imiquimod 5% cream, 133 lesional high-risk HPV DNA load, 523 microarray analysis of aberrant gene expression and effect of anaplastic large cell lymphoma, systemic, Bcl-2 expression and imiquimod therapy, 1132 survivin location, 41 scalp, comparison of topical 5-aminolaevulinic acid androgenetic alopecia and Kennedy disease, 290 methylester with photodynamic therapy, 87 anetoderma, primary in infancy (Correspondence), 1267 variable pulsed light compared with light-emitting diodes in angioid streaks in pseudoxanthoma elasticum, biopsy of topical photodynamic therapy, 111 clinically normal skin in investigation of, 748 actinic prurigo, deterioration due to degradation of angiokeratoma, cutaneous, in Hispanic–American patient DermaGard window film (Correspondence), 619 with cerebral cavernous malformations (Correspondence), activator protein-1, abnormal expression in CD30+ cutaneous 210 large-cell lymphoma, 914 angiomatous papillomatosis of the oral cavity, pseudomembra- acute generalized exanthematous pustulosis nous Bartonella-related, association with bone marrow hydroxyzine-induced (Correspondence), 1296 transplantation and oral GVHD (Case Report), 174 risk factors, results of EuroSCAR study, 989 angiomyofibroblastoma of the vulva (Correspondence), acute myeloid leukaemia, fatal Fusarium falciforme infection in 189 (Correspondence), 407 anticentromere antibodies in systemic sclerosis associated adalimumab with discoid lupus erythematosus in Japanese patients in treatment of psoriatic arthritis, hypertriglyceridaemia (Correspondence), 1289 during (Correspondence), 1273 antiphospholipid syndrome with fibromuscular dysplasia of in treatment of pyoderma gangrenosum (Correspondence), peripheral arteries, livedo racemosa with digital necrosis 1274 (Correspondence), 389 Adams–Oliver syndrome, incomplete form (Correspondence), antiseptic medication, cytotoxic analysis on skin substitutes 836 and autograft, 33 adhesion molecules in psoriasis, 259 aortic coarctation with aplasia cutis (Correspondence), 836 AIDS see HIV/AIDS aplasia cutis with aortic coarctation (Correspondence), 836 alcohol as irritant, 74 Arndt–Gottron scleromyxoedema, intravenous immuno- alcohol hand rub as a soap substitute, 1 globulin therapy (Correspondence), 1058 alemtuzumab, transformation of acute cutaneous T-cell arrector pili muscle, morphological relationship with lymphoma during treatment (Correspondence), 841 sebaceous gland in the follicular unit, 325

1324 2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1325 atopic dermatitis Behc¸et’s disease, clinical features and natural course, 901 canine homologue, efficacy of low-dose aminopterin benzalkonium chloride in antiseptic bath oil, flexural allergic (Correspondence), 1040 contact dermatitis due to (Case Report), 795 childhood, role of brain-derived neurotrophic factor and b-catenin substance P in nocturnal scratching, 922 increased expression in suprabasal involved psoriatic clinical scoring and noninvasive measurement in assessment epidermis, 1168 of, 674 involvement in progression and prognosis of cutaneous cutaneous Malassezia flora, difference between adults and melanoma, 1212 children, 1178 b-estradiol, paradoxical effects on epidermal permeability efficacy and tolerability of Chinese herbal medicine, 357 barrier status (Concise Communication), 776 environmental factors and parental atopy in primary-school bexarotene, optimal use in cutaneous T-cell lymphoma children in Taiwan, 1217 (Review), 433 head and neck, trial of pimecrolimus cream in adolescents bloodroot as do-it-yourself Mohs’ surgery (Correspondence), and adults, 954 1078 lipopolysaccharide-induced TNF-a, IFN-c and IL-10 blunt soft tissue trauma, lipoma following, 92 production (Concise Communication), 583 body surface area, ‘handprint’ versus ‘palm minus the fingers’ practical issue on interpretation of scoring (Review), 645 (Correspondence), 1080 role of epidermal differentiation complex (Review), 441 bone marrow transplantation associated with Bartonella-related safety of tacrolimus ointment (Review), 861 pseudomembranous angiomatous papillomatosis of the oral SCORAD index (Review), 645 cavity (Case Report), 174 severe, addition of topical pimecrolimus to once-daily borreliosis (Correspondence), 403 mid-potent steroid (Concise Communication), 378 morphoea as manifestation of infection, 1189 severe childhood, mycophenolate mofetil therapy, 127 bortezomib-associated cutaneous vasculitis (Correspondence), ATP2A2 gene mutation in Japanese family with haemorrhagic 799 Darier disease (Gene Corner), 605 bowel cancer and obstruction, potential masking of symptoms atypical melanocytic lesions, prediction of severe histological by Fumaderm therapy (Correspondence), 825 dysplasia/early melanoma, 758 brain-derived neurotrophic factor, role in nocturnal scratching autograft, cytotoxic analysis of antiseptic medication, 33 in childhood atopic dermatitis, 922 autosomal recessive congenital ichthyosis, transglutaminase 1 bullous congenital ichthyosiform erythroderma, KRT10 gene deficiency and corneocyte collapse as indication for targeted mutation in annular variant with clinical worsening in molecular screening (Correspondence), 808 pregnancy (Gene Corner), 602 autosomal recessive hypohidrotic ectodermal dysplasia, novel bullous dermolysis of the newborn, transient with EDAR gene deletion mutation (Correspondence), 207 pseudosyndactyly, novel COL7A1 mutations (Gene Corner), axillary hyperhidrosis, repeat liposuction–curettage treatment, 179 739 C16 laminin, effect on extravascular migration of human B-cell leukaemia, systemic, and cutaneous anaplastic large-cell melanoma cells in shell-less chick chorioallantoic membrane lymphoma (Correspondence), 1060 assay (Concise Communication), 780 B-cell lymphomas, coexistence of discordant lymphomas in calcinosis cutis, metastatic, presentation as bilateral vulval cysts skin and lymph node (Correspondence), 629 (Correspondence), 622 Bartonella-related pseudomembranous angiomatous papillo- calcipotriol–betamethasone dipropionate in morphoea matosis of the oral cavity associated with bone marrow (Correspondence), 615 transplantation and oral GVHD (Case Report), 174 calcitriol compared with tacrolimus in chronic plaque psoriasis basal cell carcinoma of the facial/genitofemoral area, 1005 of inner canthus, incomplete excision (Correspondence), capillary malformation, long-standing, topical imiquimod 1301 therapy (Correspondence), 1071 lack of association with MDM2 SNP309 promoter polymorph- capillary morphogenesis protein-2 gene mutation in juvenile ism (Concise Communication), 375 hyaline fibromatosis (Gene Corner), 1037 low-risk nodular, minimal curettage and photodynamic cathelicidin LL-37 induction of generation of reactive oxygen therapy compared with surgical excision (Correspondence), species and release of human a-defensins from neutrophils, 401 1124 nodular, lentigo maligna involving tumour nests and stroma Cdc42, involvement in progression and prognosis of (Correspondence), 184 cutaneous melanoma, 1212 Bcl-2 expression in CD30+ lymphoproliferative disorders and cellulitis, non-necrotizing of the lower limb, seasonal systemic anaplastic large cell lymphoma, 41 variations in admission to UK teaching hospital becocalcidiol, topical treatment in psoriasis, 369 (Correspondence), 1047

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 1326 Subject index central nervous system, involvement in stage 1b mycosis corneocyte collapse as indication for targeted molecular fungoides (Correspondence), 815 screening in autosomal recessive congenital ichthyosis ceramide analogues and selective recovery of perturbed human (Correspondence), 808 skin barrier, 704 coronary stenting, contact allergy to gold in patients with, cerebral cavernous malformations with cutaneous angiokera- 730 toma and venous malformations in Hispanic–American corticosteroids patient (Correspondence), 210 addition of pimecrolimus to once-daily mid-potent steroid in cetirizine, better control of chronic idiopathic urticaria with severe atopic dermatitis (Concise Communication), 378 increased dosage (Correspondence), 803 allergy to in patients with inflammatory bowel disease, 967 Chaetomium funicola, chromoblastomycosis caused by (Case clinical adverse effects induced by, 142 Report), 1025 systemic chemokine serum profiles in alopecia areata, 466 effect on clinical manifestations in non-Herlitz junctional childhood epidermolysis bullosa (Case Report), 596 environmental factors and parental atopy in primary-school and reactivation of chronic hepatitis B infection (Concise children with atopic dermatitis in Taiwan, 1217 Communication), 587 guidelines for evaluation and management of urticaria, COX-2 gene, association of functional variants in regulatory 1116 regions with nonmelanoma skin cancer following organ see also infancy transplantation, 49 Chinese herbal medicine, efficacy and tolerability in atopic creeping eruption, outbreak in southern France (Correspon- dermatitis, 357 dence), 824 chondrodermatitis nodularis helicis, punch and graft technique Crohn disease, development in patient on etanercept therapy as novel surgical treatment method, 744 for psoriasis (Correspondence), 396 chromium, occupational dermatitis due to, 518 cryptococcal infection, disseminated, in psoriasis treated with chromoblastomycosis caused by Chaetomium funicola (Case efalizumab, methotrexate and ciclosporin (Correspondence), Report), 1025 1067 chronic lymphocytic leukaemia CTACK/CCL27, cutaneous expression in psoriasis and with erythema annulare centrifugum (Correspondence), modification following etanercept administration, 1155 1044, 1045 cutaneous anaplastic large-cell lymphoma and systemic B-cell with primary cutaneous anaplastic large-cell lymphoma and leukaemia (Correspondence), 1060 mycosis fungoides (Correspondence), 1291 cutaneous large-cell lymphoma, CD30+, abnormal activator with primary cutaneous marginal zone B-cell lymphoma protein-1 transcription factor expression, 914 (Case Report), 591 cutaneous lupus erythematosus ciclosporin expression pattern of interferon-inducible proteins in and disseminated cryptococcal infection in psoriasis subsets, 752 (Correspondence), 1067 matrix metalloproteinases as mediators of tissue injury, and quality of life in Kimura’s disease (Correspondence), 970 420 cutaneous pseudolymphoma associated with leishmaniasis cidofovir, topical 1% in treatment of lentigo maligna (Correspondence), 1042 (Correspondence), 421 cutaneous T-cell lymphoma cobalt, occupational dermatitis due to, 518 acute, transformation during alemtuzumab treatment coccygeal polypoid eccrine naevus (Correspondence), 614 (Correspondence), 841 COL7A1 gene optimal use of bexarotene (Review), 433 mutational analysis in Korean patients with dystrophic cutis verticis gyrata, primary essential with hyaluronic acid epidermolysis bullosa (Gene Corner), 1260 deposition (Correspondence), 806 novel mutations in transient bullous dermolysis of the CXCR4, involvement in progression and prognosis of newborn with pseudosyndactyly (Gene Corner), 179 cutaneous melanoma, 1212 cold urticaria, tolerance induction (Correspondence), 835 CYLD gene mutation in Chinese family with multiple familial complementary and alternative treatment in allergic contact trichoepithelioma (Correspondence), 818 dermatitis, 301 Conradi–Hu¨nermann–Happle syndrome, novel EBP gene Darier disease, haemorrhagic, ATP2A2 gene mutation in mutation, 1225 Japanese family (Gene Corner), 605 contact allergy Degos disease, cutaneous and recurrent perforating intestinal, lack of association with filaggrin null alleles, 1199 use of intravenous immunoglobulin (Correspondence), 206 relationship between exposure dose and response in delayed-type hypersensitivity to low molecular weight induction and elicitation (Review), 1093 heparins and heparinoids, 514 to gold, in patients with endovascular coronary stents, 730 dendritic cells, impairment of function by infliximab, 249

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1327

DermaGard window film, deterioration of actinic prurigo epidermal barrier due to degradation of (Correspondence), 619 acceleration of recovery by potassium channel openers, 888 dermatofibroma paradoxical effects of b-estradiol on status (Concise Commu- multiple eruptive myxoid (Case Report), 382 nication), 776 stromelysin-3 expression in differential diagnosis, 319 epidermal cysts, filaggrin expression and pathogenesis dermatofibrosarcoma protuberans, stromelysin-3 expression in (Correspondence), 415 differential diagnosis, 319 epidermal differentiation, functional redundancy of extracellu- dermatomyositis, juvenile-onset clinically amyopathic lar matrix protein 1 (Concise Communication), 771 (Review), 637 epidermal differentiation complex, role in ichthyosis vulgaris, dermatophyte infections, diagnosis by a novel multiplex atopic dermatitis and psoriasis (Review), 441 real-time PCR detection/identification scheme, 681 epidermodysplasia verruciformis dermoscopy novel homozygous nonsense TMC8 mutation in Brazilian benefits of automated image analysis instruments, 926 family (Correspondence), 831 in intermediate stage of seborrhoeic keratosis regressing to novel mutation of EVER1/TMC6 gene in Japanese patient lichenoid keratosis, 266 (Gene Corner), 1265 significance of granularity in diagnosis of melanoma, epidermolysis bullosa acquisita, treatment-resistant, response 907 to rituximab (Correspondence), 417 desmocollin-3 antibodies associated with acquired palmoplan- EpiDermTM tissues, inhibition of tumour necrosis factor-a tar keratoderma and immunobullous disease (Case Report), secretion by UTL-5d (Concise Communication), 575 168 Epstein Barr virus, epidermal cell necrosis with direct epider- digital necrosis with livedo racemosa in patient with mal infiltration of EBV-encoded small nuclear RNA+ T cells antiphospholipid syndrome and fibromuscular dysplasia in EBV-associated haemophagocytic syndrome (Correspon- of peripheral arteries (Correspondence), 389 dence), 1053 dilated cardiomyopathy with recessive dystrophic epidermoly- eruptive vellus hair cysts mimicking naevus of Ota sis bullosa (Correspondence), 610 (Correspondence), 188 2,5-dimethyl-4-hydroxy-3(2H)-furanone, mechanism of erythema annulare centrifugum with chronic lymphocytic melanogenesis inhibition, 242 leukaemia (Correspondence), 1044, 1045 discoid lupus erythematosus with limited cutaneous systemic erythema dyschromicum perstans, differentiation from idio- sclerosis in Japanese patients with anticentromere antibodies pathic eruptive macular pigmentation in 7-year-old girl (Correspondence), 1289 (Correspondence), 839 Down syndrome, milium-like syringoma with focal calcifica- erythrokeratoderma variabilis, lack of GJB3 or GJB4 mutation tion (Correspondence), 612 (Correspondence), 410 drug-induced hypersensitivity syndrome, association of erythropoietic protoporphyria, late presentation (Case Report), HHV-6 with flaring and severity, 934 1030 drug-induced lupus erythematosus, and use of tetracycline essential cryoglobulinaemia with livedoid vasculitis, therapeu-

antibiotics in acne vulgaris, 540 tic effect of lipoprostaglandin E1 (Correspondence), 1051 dystrophic epidermolysis bullosa, COL7A1 mutational analysis etanercept therapy in psoriasis in Korean patients (Gene Corner), 1260 and cutaneous CTACK/CCL27 expression, 1155 and development of Crohn disease (Correspondence), 396 E-cadherin, involvement in progression and prognosis of effect on fatigue and symptoms of depression (Correspon- cutaneous melanoma, 1212 dence), 1275 ears, turkey ear (Correspondence), 816 rapid onset of multiple squamous cell carcinomas following EBP gene mutation in Conradi–Hu¨nermann–Happle syndrome, (Correspondence), 1076 1225 ethyl linoleate with triethyl citrate in lotion for treatment of EDAR gene mutation in autosomal recessive hypohidrotic acne, 569 ectodermal dysplasia (Correspondence), 207 etretinate in treatment of milia en plaque (Correspondence), efalizumab and disseminated cryptococcal infection in 1287 psoriasis (Correspondence), 1067 EuroSCAR study, risk factors for acute generalized exanthema- elbow, total arthroplasty, skin pigmentation caused by tous pustulosis, 989 metal particle deposition following (Correspondence), EVER1/TMC6 gene, novel mutation in Japanese patient with 1074 epidermodysplasia verruciformis (Gene Corner), 1265 endovascular coronary stenting and contact allergy to gold, extracellular matrix modulators, alteration after nonablative 730 laser therapy in skin rejuvenation, 306 eosinophilia with transient macular erythema in patient extracellular matrix protein 1, functional redundancy in carrying FIP1L1-PDGFRA fusion gene (Correspondence), epidermal differentiation (Concise Communication), 1284 771

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 1328 Subject index

Fabry disease, Fabry outcome survey, 331 granuloma annulare, severe linear form along Blaschko’s lines facial nerve palsy in sarcoidosis (Correspondence), 200 preceding onset of classical form in childhood (Correspon- fibroblast growth factor 23, production by phosphaturic dence), 1056 mesenchymal tumour developing in subcutaneous tissue of granulomatous slack skin with clonal T-cell receptor-c gene patient with oncogenic osteomalacia (Correspondence), 198 rearrangement in skin and lymph node (Correspondence), fibroblasts 405 cutaneous DNA repair capacities, 26 haemangioma effect of sun exposure, age and smoking on response to infantile, origin of endothelial cells (Concise Communica- acute oxidative stress, 26 tion), 158 human dermal, preferential augmentation of transient mimicked by plexiform schwannoma (Correspondence), transgene expression by histone deacetylase inhibitors, 662 838 fibronectin expression in sentinel lymph nodes in melanoma haemophagocytic syndrome, EBV-associated, epidermal cell (Correspondence), 398 necrosis with direct epidermal infiltration of EBV-encoded filaggrin small nuclear RNA+ T cells (Correspondence), 1053 expression and pathogenesis of epidermal cysts (Correspon- hair, selective biodegradation in shafts derived from archaeo- dence), 415 logical, forensic and experimental contexts, 450 null alleles not associated with hand eczema or contact hair dye allergy allergy, 1199 clinical assessment of patch test kit marketed to UK hair- FIP1L1-PDGFRA fusion gene and transient macular erythema dressers (Concise Communication), 1017 with eosinophilia (Correspondence), 1284 role of self-tests in diagnosis, 847 flexural allergic contact dermatitis to benzalkonium chloride in hair follicle antiseptic bath oil (Case Report), 795 epithelial compartment (follicular trochanter) in need of focal congenital alopecia and Gomez–Lopez–Hernandez further characterization (Concise Communication), 1013 syndrome (Correspondence), 196 human anagen, TGF-b receptor II preferential expression in folate with methotrexate (Correspondence), 213 companion layer (Concise Communication), 161 follicular mucinosis, neonatal (Correspondence), 609 hairdressing follicular unit, morphological relationship between sebaceous and scalp disease in African adults, 981 gland and arrector pili muscle, 325 and scalp disease in African schoolchildren, 106 folliculosebaceous cystic hamartoma, giant genital variant, hand eczema management by laser and acitretin therapy (Correspon- incidence in population-based twin cohort, 552 dence), 833 lack of association with filaggrin null alleles, 1199 14S24, and selective recovery of perturbed human skin head, actinic keratosis, safety and efficacy of imiquimod 5% barrier, 704 cream, 133 fragrance ingredient labelling in products for sale in UK, 295 henna dye application, cutaneous mercury deposits associated Fumaderm, potential masking of symptoms of bowel cancer (Correspondence), 394 and obstruction (Correspondence), 825 heparinoids, low molecular weight, delayed-type hypersensi- furuncles, prevalence of Staphylococcus aureus toxins and nasal tivity to, 514 carriage, 1161 heparins, low molecular weight, delayed-type hypersensitivity Fusarium falciforme, fatal infection in acute myeloid leukaemia to, 514 (Correspondence), 407 hepatitis B, chronic, reactivation in systemic corticosteroid therapy (Concise Communication), 587 giant bathing trunk naevus with lymphadenopathy and histone deacetylase inhibitors and preferential augmentation of unusual pathology (Case Report), 599 transient transgene expression in human dermal fibroblasts, giant genital folliculosebaceous cystic hamartoma, manage- 662 ment by laser and acitretin therapy (Correspondence), 833 HIV/AIDS glycopyrronium bromide, oral therapy in hyperhidrosis, 118 imiquimod treatment of anal intraepithelial neoplasia, gold, contact allergy to, in patients with endovascular correlation of p16ink4a expression with decline of lesional coronary stents, 730 high-risk HPV DNA load, 523 Gomez–Lopez–Hernandez syndrome and focal congenital multiple squamous cell carcinomas of the fingernails caused alopecia (Correspondence), 196 by HPV type 26 infection in AIDS patient under HAART graft-versus-host disease, oral, associated with Bartonella-related (Case Report), 788 pseudomembranous angiomatous papillomatosis of the oral post-kala-azar dermal leishmaniasis as an immune reconstitu- cavity (Case Report), 174 tion inflammatory syndrome (Case Report), 1032 granulocyte/macrophage colony-stimulating factor knockout rapid onset of HPV-associated oral intraepithelial neoplasia mice, impaired cutaneous wound healing, 458 disease (Correspondence), 826

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1329 human herpesvirus-6 associated with flaring and severity in infancy drug-induced hypersensitivity syndrome, 934 haemangioma, origin of endothelial cells (Concise Communi- human papillomaviruses (HPVs) cation), 158 oncogenic potential (Review), 228 pityriasis lichenoides et varioliformis acuta (Correspondence), rapid onset of type 72-associated oral intraepithelial neoplasia 194 in HIV infection (Correspondence), 826 primary anetoderma (Correspondence), 1267 type 26 infection causing multiple squamous cell carcinomas sock-line bands (Correspondence), 1063 of the fingernails in AIDS patient under HAART (Case see also childhood Report), 788 inflammatory bowel disease and allergy to corticosteroids, 967 hyaluronic acid deposition in primary essential cutis verticis infliximab gyrata (Correspondence), 806 inhibition of skin-homing CD4+ and CD8+ T cell activation hydroxycarbamide in treatment of lichen sclerosus and impairment of dendritic cell function, 249 (Correspondence), 622 and rapid clearing of erythrodermic psoriasis (Correspon- hydroxychloroquine resistance due to smoking in a patient dence), 828 with lupus erythematosus tumidus (Correspondence), in treatment of severe childhood psoriatic arthritis 1081 (Correspondence), 191 hydroxyzine-induced acute generalized exanthematous inner canthus, basal cell carcinoma, incomplete excision pustulosis (Correspondence), 1296 (Correspondence), 1301 hyperhidrosis insulin resistance and psoriasis (Concise Communication), 1249 axillary, repeat liposuction–curettage treatment, 739 intercellular adhesion molecule 1, interaction with lympho- use of oral glycopyrronium bromide, 118 cyte function-associated antigen type 1, and adhesion of hypertrichosis lanuginosa acquisita, paraneoplastic, 1087 peripheral blood mononuclear cells and CD4+ T cells to hypertriglyceridaemia during adalimumab treatment of cultured endothelial cells, 259 psoriatic arthritis (Correspondence), 1273 interferon-b-1a therapy, calcified subcutaneous nodules as hypocomplementaemic urticarial vasculitis with non-Hodgkin long-term complication (Correspondence), 624 lymphoma, intravenous immunoglobulin treatment interferon-c, lipopolysaccharide-induced production in atopic (Correspondence), 392 dermatitis (Concise Communication), 583 hypohidrotic ectodermal dysplasia, autosomal recessive, novel interferon-inducible protein expression patterns in subsets of EDAR gene deletion mutation (Correspondence), 207 cutaneous lupus erythematosus, 752 interleukin-8 and pyoderma gangrenosum (Correspondence), ichthyosis vulgaris, role of epidermal differentiation complex 1279 (Review), 441 interleukin-10, lipopolysaccharide-induced production in imiquimod atopic dermatitis (Concise Communication), 583 effect on aberrant gene expression in actinic keratosis, intramuscular immunoglobulin in recalcitrant suppurative skin 1132 diseases, 563 safety and efficacy of 5% cream in actinic keratosis of the intravascular lymphoma, skin manifestations mimicking head, 133 inflammatory skin disease (Review), 16 treatment of anal intraepithelial neoplasia in HIV infection, intravenous immunoglobulin correlation of p16ink4a expression with decline of lesional in Arndt–Gottron scleromyxoedema (Correspondence), 1058 high-risk HPV DNA load, 523 in cutaneous and recurrent perforating intestinal Degos treatment of long-standing capillary malformation disease (Correspondence), 206 (Correspondence), 1071 in hypocomplementaemic urticarial vasculitis with treatment of primary cutaneous follicle centre lymphoma non-Hodgkin lymphoma (Correspondence), 392 (Correspondence), 620 in pyoderma gangrenosum, 1235 immunobullous disease associated with antibodies to isoeugenol, frequency of allergic contact dermatitis to desmocollin-3 (Case Report), 168 (Concise Communication), 580 immunoglobulin, intramuscular, in recalcitrant suppurative isotretinoin and incidence of relapse in acne, 1240 skin diseases, 563 see also intravenous immunoglobulin junctional epidermolysis bullosa, non-Herlitz, effect of oral impetigo steroids on clinical manifestations (Case Report), 596 epidemic and nonepidemic phases, incidence study in a juvenile hyaline fibromatosis, novel point mutation in gene encod- general population, 100 ing capillary morphogenesis protein-2 (Gene Corner), 1037 prevalence of Staphylococcus aureus toxins and nasal carriage, 1161 Kennedy disease and androgenetic alopecia, 290 inducible nitric oxide synthase inhibition and wound healing Kimura’s disease, ciclosporin and quality of life (Correspon- in SKH-1 female mouse, 656 dence), 420

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 1330 Subject index

KIND1 gene recurrent mutation in a Brazilian pedigree with liposuction–curettage, repeat treatment in axillary hyperhidro- Kindler syndrome (Correspondence), 1281 sis, 739 Kindler syndrome lips, verrucous carcinoma in pregnancy, laser therapy recurrent KIND1 gene mutation in a Brazilian pedigree (Correspondence), 813 (Correspondence), 1281 livedo racemosa with digital necrosis in patient with anti- unusual molecular findings (Case Report), 1252 phospholipid syndrome and fibromuscular dysplasia of KRT10 gene mutation in annular variant of bullous congenital peripheral arteries (Correspondence), 389 ichthyosiform erythroderma with clinical worsening in livedoid vasculitis with essential cryoglobulinaemia, therapeu-

pregnancy (Gene Corner), 602 tic effect of lipoprostaglandin E1 (Correspondence), 1051 lupus erythematosus, drug-induced, and use of tetracycline laminin expression in sentinel lymph nodes in melanoma antibiotics in acne vulgaris, 540 (Correspondence), 398 lupus erythematosus tumidus (Correspondence), 403 Langerhans cell histiocytosis, cutaneous, in elderly man resistance to hydroxychloroquine in a smoker (Correspon- successfully treated with narrowband UVB phototherapy dence), 1081 (Correspondence), 1277 lymphocyte function-associated antigen type 1, interaction lanreotide treatment of phaeochromocytoma, radiation recall with ICAM-1, and adhesion of peripheral blood mononuc- dermatitis following (Correspondence), 1061 lear cells and CD4+ T cells to cultured endothelial cells, laser therapy 259 with acitretin in treatment of giant genital variant of lymphocytic infiltration, Jessner–Kanof (Correspondence), 403 folliculosebaceous cystic hamartoma (Correspondence), lymphoma, intravascular, skin manifestations mimicking 833 inflammatory skin disease (Review), 16 nonablative in skin rejuvenation, alteration of extracellular lymphomatoid granulomatosis as complication of other matrix modulators after, 306 haematological malignancies (Correspondence), 426 verrucous carcinoma of the lips in pregnancy (Correspon- lymphoproliferative disorders, CD30+, Bcl-2 expression and dence), 813 survivin location, 41 legs, non-necrotizing cellulitis, seasonal variations in admission to UK teaching hospital (Correspondence), macular pigmentation, idiopathic eruptive in 7-year-old girl, 1047 differentiation from erythema dyschromicum perstans leishmaniasis (Correspondence), 839 cutaneous pseudolymphoma associated (Correspondence), Malassezia 1042 cutaneous flora in atopic dermatitis, difference between post-kala-azar dermal adults and children, 1178 as an immune reconstitution inflammatory syndrome in quantitative analysis in scale of patients with psoriasis, 670 AIDS (Case Report), 1032 malignant atrophic papulosis, cutaneous and recurrent with borderline tuberculoid leprosy (Correspondence), 811 perforating intestinal, use of intravenous immunoglobulin lentigo maligna (Correspondence), 206 involvement of tumour nests and stroma of nodular basal cell mantle-cell lymphoma, nodal, coexistence with primary carcinoma (Correspondence), 184 cutaneous follicle centre lymphoma (Correspondence), 629 treatment with topical 1% cidofovir (Correspondence), 421 matrix metalloproteinases as mediators of tissue injury in LEOPARD syndrome with melanoma (Correspondence), 1297 cutaneous lupus erythematosus, 970 leprosy MC1R genotype and erythemal sensitivity to PUVA, 1230 borderline tuberculoid, with post-kala-azar dermal leishma- MDM2 SNP309 promoter polymorphism, lack of association niasis (Correspondence), 811 with basal cell carcinoma (Concise Communication), 375 long-term culture of multibacillary macrophages isolated melanocortin-1 receptor genotype and erythemal sensitivity to from skin lesions, 273 PUVA, 1230 lichen planus, oral cavity, scoring system for disease severity, melanocytic naevi 765 multiple cutaneous, and leptomeningeal melanoma lichen sclerosus (Correspondence), 397 hydroxycarbamide treatment (Correspondence), 622 new classification system (Review), 217 male genital, tacrolimus therapy (Correspondence), 1079 melanogenesis, inhibition by 2,5-dimethyl-4-hydroxy-3(2H)- lichen sclerosus et atrophicus, lesions resembling in mycosis furanone, 242 fungoides (Correspondence), 411 melanoma lipoma following blunt soft tissue trauma, 92 acral lentiginous, histopathological prognostic features, 311 lipoprostaglandin E1, therapeutic effect on livedoid vasculitis beauty and sunbeds, 215 associated with essential cryoglobulinaemia (Correspon- early, prediction in excised atypical melanocytic lesions, dence), 1051 758

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1331

effect of C16 laminin on extravascular migration of human central nervous system involvement in stage 1b disease melanoma cells in shell-less chick chorioallantoic membrane (Correspondence), 815 assay (Concise Communication), 780 differentiation from psoriasis by serum proteomic analysis, evaluation of tumour-infiltrating CD4+CD25+FOXP3+ 946 regulatory T cells, 531 lichen sclerosus et atrophicus-like lesions in (Correspon- fibronectin and laminin expression in sentinel lymph nodes dence), 411 (Correspondence), 398 with primary cutaneous anaplastic large-cell lymphoma involvement of E-cadherin, b-catenin, Cdc42 and CXCR4 in in patient with B-cell chronic lymphocytic leukaemia progression and prognosis, 1212 (Correspondence), 1291 with LEOPARD syndrome (Correspondence), 1297 serum tissue polypeptide antigen correlating with clinical leptomeningeal, and multiple cutaneous melanocytic naevi course (Correspondence), 423 (Correspondence), 397 transformed, 284 metastatic, evaluation of tumour-infiltrating CD4+CD25+FOXP3+ regulatory T cells, 531 naevi, atypical and benign, evaluation of tumour-infiltrating sentinel lymph node biopsy, 58 CD4+CD25+FOXP3+ regulatory T cells, 531 significance of granularity in dermoscopic diagnosis, 907 nails trends in epidemiology, 338 anonychia, hyponychia and spontaneous phalangeal amputa- types, 338 tion as a consequence of ischaemic necrosis of the extremi- visceral metastatic, autoimmunity and survival (Correspon- ties following umbilical catheterization (Correspondence), dence), 413 1299 menstrual cycle, morphological effects on vulval vestibular combination of surgical avulsion and topical therapy in single mucosa, 487 nail onychomycosis, 364 mercury, cutaneous deposits associated with henna dye impact of public service advertisement about onychomycosis application (Correspondence), 394 on health behaviour of Greek population (Correspondence), metabolic syndrome, prevalence in psoriasis, 68 821 methotrexate lack of RSPO4 mutation in congenital nail hypoplasia/ and disseminated cryptococcal infection in psoriasis aplasia with underlying skeletal defects (Correspondence), (Correspondence), 1067 801 with folate (Correspondence), 213 multiple squamous cell carcinomas caused by human milia en plaque, treatment with oral etretinate (Correspon- papillomavirus type 26 infection in AIDS patient under dence), 1287 HAART (Case Report), 788 mites, bacterial antigens related to and stimulation of thickness measurements using optical coherence tomography inflammatory cells in rosacea, 474 and ultrasonography, 894 Mohs’ surgery, do-it-yourself with internet-obtained topical application of acidified nitrite, 494 bloodroot (Correspondence), 1078 narrowband UVB phototherapy Mondor’s phlebitis after use of tadalafil (Correspondence), challenge of follow-up, 344 209 and cutaneous Langerhans cell histiocytosis (Correspon- morphoea dence), 1277 as manifestation of Borrelia infection, 1189 neonates use of calcipotriol–betamethasone dipropionate (Correspon- follicular mucinosis (Correspondence), 609 dence), 615 pemphigus vulgaris in siblings associated with mild maternal mucosal disease, scoring system for severity, 765 disease (Correspondence), 192 multiple eruptive myxoid dermatofibromas (Case Report), 382 nephrogenic systemic fibrosis, case series (Case Report), 783 multiple familial trichoepithelioma, novel mutation of CYLD in neurofibromatosis 1, terminology for cutaneous lesions Chinese family (Correspondence), 818 (Correspondence), 183 Mycobacterium avium spondylitis, recovery from Se´zary syndrome neutropenia, late-onset, following rituximab treatment in following (Correspondence), 1270 autoimmune diseases (Correspondence), 1271 Mycobacterium leprae, model to study interaction with human neutrophils, cathelicidin LL-37-induced generation of reactive cells, 273 oxygen species and release of human a-defensins, 1124 Mycobacterium neoaurum, cutaneous infection causing scarring nickel allergy, relationship between patch test and repeated alopecia (Correspondence), 204 open application test thresholds, 723 mycophenolate mofetil treatment of severe childhood atopic NOD2 gene mutation in sarcoidosis resembling Blau syndrome dermatitis, 127 (Case Report), 1257 mycosis fungoides non-Hodgkin lymphoma with hypocomplementaemic CD8+ poikilodermatous nonaggressive, response to PUVA urticarial vasculitis, intravenous immunoglobulin treatment (Correspondence), 1064 (Correspondence), 392

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 1332 Subject index nonmelanoma skin cancer, association of functional variants periumbilical dermatitis, urachal sinus presenting as in regulatory regions of COX-2 gene following organ (Correspondence), 419 transplantation, 49 phaeochromocytoma, radiation recall dermatitis after lanreo- tide treatment (Correspondence), 1061 obesity and psoriasis, 649 phenytoin, topical, clinical effect on wound healing, 997 occupational dermatitis to chromium and cobalt, 518 phosphaturic mesenchymal tumour developing in subcuta- occupational skin disease, incidence as reported to THOR neous tissue of patient with oncogenic osteomalacia network, 713 (Correspondence), 198 oncogenic osteomalacia, development of phosphaturic photoageing, mechanism, prevention and therapy, 874 mesenchymal tumour in subcutaneous tissue (Correspon- photodynamic therapy dence), 198 acne (Correspondence), 810 onychomycosis comparison with topical 5-aminolaevulinic acid methylester combination of surgical avulsion and topical therapy, 364 in scalp actinic keratosis, 87 detection of Trichophyton rubrum DNA from nail samples, 698 with minimal curettage in low-risk nodular basal cell impact of public service advertisement on health behaviour carcinoma, comparison with surgical excision (Correspon- of Greek population (Correspondence), 821 dence), 401 with matrix involvement, oral terbinafine compared with partial remission of circumscribed palmar hypokeratosis combination of amorolfine nail lacquer and oral terbinafine, (Correspondence), 804 149 variable pulsed light compared with light-emitting diodes in oral cavity actinic keratosis, 111 Bartonella-related pseudomembranous angiomatous papilloma- photosensitivity, remission following tumour necrosis tosis associated with bone marrow transplantation and oral factor inhibitor treatment of psoriasis (Correspondence), GVHD (Case Report), 174 625 lichen planus, scoring system for disease severity, 765 pigmentation caused by deposits of metal particles following oral contraceptives, morphological effects on vulval vestibular total elbow arthroplasty (Correspondence), 1074 mucosa, 487 pimecrolimus oral intraepithelial neoplasia, rapid onset of HPV-associated addition to once-daily mid-potent steroid in severe atopic disease in HIV infection (Correspondence), 826 dermatitis (Concise Communication), 378 osteomalacia, oncogenic, development of phosphaturic trial in adolescents and adults with head and neck atopic mesenchymal tumour in subcutaneous tissue (Correspon- dermatitis intolerant of/dependent on steroids, 954 dence), 198 pityriasis lichenoides, differences between children and adults, 941 p16ink4a expression during imiquimod treatment of anal pityriasis lichenoides et varioliformis acuta in infancy intraepithelial neoplasia in HIV infection, correlation with (Correspondence), 194 decline of lesional high-risk HPV DNA load, 523 pityriasis rubra pilaris palmar hypokeratosis, circumscribed, partial remission with classical juvenile, recurrence in adulthood (Correspondence), photodynamic therapy (Correspondence), 804 842 paraneoplastic hypertrichosis lanuginosa acquisita, 1087 in mother and daughters (Correspondence), 202 parotid gland enlargement in sarcoidosis (Correspondence), plexiform schwannoma mimicking haemangioma (Correspon- 200 dence), 838 patch test polypoid eccrine naevus, coccygeal (Correspondence), 614 clinical assessment of kit marketed to UK hairdressers for porphyria cutanea tarda, familial in Spain, molecular hetero- detection of hair dye allergy (Concise Communication), geneity and novel mutations in UROD gene, 501 1017 potassium channel openers and acceleration of epidermal in nickel allergy, relationship with repeated open application barrier recovery, 888 test thresholds, 723 pregnancy, laser therapy of verrucous carcinoma of the lips patient information website, designing (Correspondence), (Correspondence), 813 1048 primary cutaneous anaplastic large-cell lymphoma with myco- pemphigoid gestationis, severe persistent (Correspondence), sis fungoides in patient with B-cell chronic lymphocytic 388 leukaemia (Correspondence), 1291 pemphigus primary cutaneous B-cell lymphoma, applicability and lack of association with PTPN22 R620W polymorphism prognostic value of new TNM classification system, 1205 (Correspondence), 1068 primary cutaneous follicle centre lymphoma neonatal, in siblings associated with mild maternal disease coexistence with nodal mantle-cell lymphoma (Correspon- (Correspondence), 192 dence), 629 periorbital oedema in sarcoidosis (Correspondence), 200 treatment with topical imiquimod (Correspondence), 620

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1333 primary cutaneous marginal zone B-cell lymphoma with treatment of severe natal cleft fissuring with tissue adhesive chronic lymphocytic leukaemia (Case Report), 591 (Correspondence), 1269 pseudolymphoma, cutaneous, associated with leishmaniasis treatment with tumour necrosis factor inhibitors, remission (Correspondence), 1042 of photosensitivity following (Correspondence), 625 pseudomembranous angiomatous papillomatosis of the oral psoriatic arthritis cavity, Bartonella-related, association with bone marrow hypertriglyceridaemia during adalimumab treatment transplantation and oral GVHD (Case Report), 174 (Correspondence), 1273 pseudosyndactyly with transient bullous dermolysis of the and psoriasis (Review), 850 newborn, novel COL7A1 mutations (Gene Corner), 179 severe childhood, infliximab treatment (Correspondence), pseudoxanthoma elasticum, biopsy of clinically normal skin in 191 investigation of patients with angioid streaks, 748 psychiatric morbidity in psoriasis, impact of changes in psoriasis clinical severity, 508 adhesion of peripheral blood mononuclear cells and CD4+ PTPN22 R620W polymorphism, lack of association with T cells to cultured endothelial cells, 259 pemphigus (Correspondence), 1068 changes in clinical severity and psychiatric morbidity in punch and graft technique as novel surgical treatment method patients, 508 in chondrodermatitis nodularis helicis, 744 chronic plaque of the facial/genitofemoral area, comparison PUVA of tolerance to and efficacy of calcitriol and tacrolimus, erythemal sensitivity to and melanocortin-1 receptor 1005 genotype, 1230 consecutive use of different biological therapies (Correspon- response of CD8+ poikilodermatous nonaggressive mycosis dence), 394 fungoides (Correspondence), 1064 differentiation from mycosis fungoides by serum proteomic pyoderma gangrenosum analysis, 946 adalimumab treatment (Correspondence), 1274 disseminated cryptococcal infection following treatment with and interleukin-8 (Correspondence), 1279 efalizumab, methotrexate and ciclosporin (Correspondence), peristomal, with systemic sclerosis (Correspondence), 618 1067 T-cell receptor repertoire, 960 early-onset, association with promoter region polymorphisms treatment with intravenous immunoglobulin, 1235 in TNF-a gene in northern Polish population (Concise Communication), 165 QuantiFERON test, usefulness in confirmation of latent erythrodermic, rapid clearing with infliximab (Correspon- tuberculosis (Correspondence), 1293 dence), 828 etanercept treatment R115866 in moderate to severe facial acne, 122 CTACK/CCL27 cutaneous expression and its modification radiation recall dermatitis in phaeochromocytoma after following, 1155 lanreotide treatment (Correspondence), 1061 and development of Crohn disease (Correspondence), 396 reactive oxygen species, cathelicidin LL-37-induced generation effects on fatigue and symptoms of depression (Correspon- from neutrophils, 1124 dence), 1275 recessive dystrophic epidermolysis bullosa with dilated rapid onset of multiple squamous cell carcinomas following cardiomyopathy (Correspondence), 610 (Correspondence), 1076 reverse transcriptase activity in normal and psoriatic skin evolution of lesion (Review), 4 samples, 482 evolution of pathogenic concepts and new therapies through review articles, maximizing quality (Correspondence), 409 phases of translational research (Review), 1103 rituximab increased nuclear b-catenin in suprabasal involved epidermis, late-onset neutropenia following treatment in autoimmune 1168 diseases (Correspondence), 1271 and insulin resistance (Concise Communication), 1249 response of treatment-resistant epidermolysis bullosa multidisciplinary evaluation of patients (Correspondence), acquisita (Correspondence), 417 1050 in severe persistent pemphigoid gestationis (Correspon- and obesity, 649 dence), 388 prevalence of metabolic syndrome, 68 rosacea, stimulation of inflammatory cells by mite-related and psoriatic arthritis (Review), 850 bacterial antigens, 474 quantitative analysis of Malassezia in scale, 670 RSPO4, lack of mutation in congenital nail hypoplasia/aplasia recalcitrant lithium-induced, alleviation by tumour necrosis with underlying skeletal defects (Correspondence), 801 factor inhibitors (Correspondence), 627 reverse transcriptase activity in skin samples, 482 sarcoidosis role of epidermal differentiation complex (Review), 441 with facial nerve palsy and parotid gland enlargement topical becocalcidiol treatment, 369 (Correspondence), 200

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 1334 Subject index

resemblance to Blau syndrome, recurrent missense mutation stoma, peristomal pyoderma gangrenosum with systemic in NOD2 gene (Case Report), 1257 sclerosis (Correspondence), 618 scalp stromal cell-derived factor-1 expression during wound actinic keratosis, comparison of topical 5-aminolaevulinic healing, 1148 acid methylester with photodynamic therapy, 87 stromelysin-3 expression in differential diagnosis of disease in African adults associated with hairdressing, 981 dermatofibroma and dermatofibrosarcoma protuberans, disease in African schoolchildren associated with hairdres- 319 sing, 106 substance P, role in nocturnal scratching in childhood atopic scarring alopecia in cutaneous Mycobacterium neoaurum infection dermatitis, 922 (Correspondence), 204 sunbeds schwannoma, plexiform, mimicking haemangioma impact of new high power lamps, 350 (Correspondence), 838 and melanoma, 215 SCORAD index (Review), 645 quantitative risk assessment, 350 SCORTEN, respiratory involvement in toxic epidermal necroly- survivin location in CD30+ lymphoproliferative disorders and sis and poor prognosis not reflected in (Correspondence), systemic anaplastic large cell lymphoma, 41 1294 syringoma, milium-like with focal calcification in Down scratching, nocturnal, in childhood atopic dermatitis, role of syndrome (Correspondence), 612 brain-derived neurotrophic factor and substance P, 922 systemic fibrosis, nephrogenic, case series (Case Report), sebaceous gland 783 morphological relationship with arrector pili muscle in the systemic sclerosis follicular unit, 325 assessment of abnormal blood flow and efficacy of study of secretion mechanism using 3D reconstruction, 325 treatment with laser Doppler flowmeter and arm-raising seborrhoeic keratosis, intermediate stage regressing to test, 690 lichenoid keratosis, dermoscopic pattern, 266 limited cutaneous, associated with discoid lupus erythemato- sentinel lymph node biopsy sus in Japanese patients with anticentromere antibodies melanoma, 58 (Correspondence), 1289 fibronectin and laminin expression (Correspondence), 398 with peristomal pyoderma gangrenosum (Correspondence), prognostic value (Correspondence), 405 618 serum proteomic analysis in differentiation of mycosis fungoides, psoriasis and normal controls, 946 T-cell receptors Se´zary syndrome following Mycobacterium avium spondylitis repertoire in pyoderma gangrenosum, 960 (Correspondence), 1270 TCR-c gene rearrangement in skin and lymph node in skin barrier function granulomatous slack skin (Correspondence), 405 improvement following artificial reduction in transepidermal T cells water loss, 82 infliximab inhibition of skin-homing CD4+ and CD8+ T cell selective recovery with 14S24 ceramide analogue, 704 activation, 249 skin substitutes, cytotoxic analysis of antiseptic medication, tumour-infiltrating CD4+CD25+FOXP3+ regulatory, evalua- 33 tion in benign and atypical naevi, melanoma and melanoma SLC39A4 gene mutation in acrodermatitis enteropathica (Gene metastases, 531 Corner), 386 T-large granular lymphocyte leukaemia, parallel evolution with smoking cutaneous vasculitis (Correspondence), 631 and acne (Correspondence), 1070 tacrolimus and resistance to hydroxychloroquine in lupus erythematosus compared with calcitriol in chronic plaque psoriasis of the tumidus (Correspondence), 1081 facial/genitofemoral area, 1005 soap substitutes, alcohol hand rub, 1 safety of ointment in atopic dermatitis (Review), 861 sock-line bands in infancy (Correspondence), 1063 in treatment of male genital lichen sclerosus (Correspon- spondylitis, Mycobacterium avium, recovery from Se´zary syndrome dence), 1079 following (Correspondence), 1270 tadalafil, Mondor’s phlebitis following use (Correspondence), squamous cell carcinoma 209 multiple, rapid onset following etanercept treatment of Taiwan, environmental factors and parental atopy and atopic psoriasis (Correspondence), 1076 dermatitis in primary-school children, 1217 multiple of the fingernails caused by human papillomavirus telangiectasia, photodistributed eruptive, adverse drug reaction type 26 infection in AIDS patient under HAART (Case to venlafaxine (Correspondence), 822 Report), 788 terbinafine Staphylococcus aureus, prevalence of toxins and nasal carriage in acrodermatitis continua of Hallopeau due to (Correspon- furuncles and impetigo, 1161 dence), 1073

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335 Subject index 1335

oral terbinafine compared with combination of amorolfine tumour necrosis factor inhibitors nail lacquer and oral terbinafine in onychomycosis with in treatment of psoriasis, remission of photosensitivity matrix involvement, 149 following (Correspondence), 625 tetracycline antibiotics in acne vulgaris and development of in treatment of recalcitrant lithium-induced psoriasis drug-induced lupus erythematosus, 540 (Correspondence), 627 The Health and Occupation Reporting (THOR) network, turkey ear (Correspondence), 816 incidence of occupational skin disease reported to, 713 twin studies in hand eczema, 552 tissue polypeptide antigen, serum, correlation with clinical course in mycosis fungoides (Correspondence), 423 umbilical catheterization, anonychia, hyponychia and sponta- TMC8 gene mutation in Brazilian family with epidermodyspla- neous phalangeal amputation as a consequence of ischaemic sia verruciformis (Correspondence), 831 necrosis of the extremities following (Correspondence), total elbow arthroplasty, skin pigmentation caused by 1299 deposits of metal particles following (Correspondence), urachal sinus presenting as periumbilical dermatitis 1074 (Correspondence), 419 toxic epidermal necrolysis, respiratory involvement and poor UROD gene mutations in familial porphyria cutanea tarda in prognosis not reflected in SCORTEN (Correspondence), Spain, 501 1294 urticaria transepidermal water loss, improvement in skin barrier chronic idiopathic, better control with increased cetirizine function following artificial reduction in, 82 dosage (Correspondence), 803 transforming growth factor-b receptor II, preferential expres- guidelines for evaluation and management in adults and sion in companion layer of human anagen hair follicle children, 1116 (Concise Communication), 161 urticarial vasculitis, hypocomplementaemic, with non-Hodgkin transglutaminase 1 deficiency as indication for targeted mole- lymphoma, intravenous immunoglobulin treatment cular screening in autosomal recessive congenital ichthyosis (Correspondence), 392 (Correspondence), 808 UTL-5d, and inhibition of tumour necrosis factor-a secretion transient bullous dermolysis of the newborn with pseudosyn- from EpiDermTM tissues (Concise Communication), 575 dactyly, novel COL7A1 mutations (Gene Corner), 179 transplant patients, skin cancer thresholds and risks for validity, confusing prima facie with true (Correspondence), reduction of immunosuppression, 1183 425 transplant recipients, association of functional variants in vasculitis, cutaneous regulatory regions of COX-2 gene on nonmelanoma skin bortezomib-associated (Correspondence), 799 cancer, 49 parallel evolution with T-large granular lymphocyte trauma, blunt soft tissue, lipoma following, 92 leukaemia (Correspondence), 631 Trichophyton rubrum DNA, detection from nail samples in venlafaxine, photodistributed eruptive telangiectasia as adverse onychomycosis, 698 drug reaction to (Correspondence), 822 trichorhinophalangeal syndrome type I, TRPS1 gene missense venous/lymphatic malformations, extensive, life-threatening mutation in Italian family with mild form (Concise haematological complications, 558 Communication), 1021 verrucous carcinoma of the lips, laser therapy during triethyl citrate with ethyl linoleate in lotion for treatment of pregnancy (Correspondence), 813 acne, 569 vulva TRPS1 gene mutation in Italian family with mild form of angiomyofibroblastoma (Correspondence), 189 trichorhinophalangeal syndrome type I (Concise bilateral cysts as presentation of metastatic calcinosis cutis Communication), 1021 (Correspondence), 622 tuberculosis, latent, usefulness of QuantiFERON test in morphological effects of oral contraceptives and menstrual confirmation of (Correspondence), 1293 cycle on vestibular mucosa, 487 tumour necrosis factor-a gene promoter region polymorphisms in early-onset psoriasis wound healing in northern Polish population (Concise Communication), biphasic expression of stromal cell-derived factor-1, 1148 165 clinical effect of topical phenytoin, 997 inhibition of secretion from EpiDermTM tissues by UTL-5d impaired, in granulocyte/macrophage colony-stimulating (Concise Communication), 575 factor knockout mice, 458 lipopolysaccharide-induced production in atopic dermatitis in SKH-1 female mouse following inducible nitric oxide (Concise Communication), 583 synthase inhibition, 656

2007 British Association of Dermatologists • British Journal of Dermatology 2007 157, pp1324–1335