archives of dermatologylib.ajaums.ac.ir/booklist/593498.pdfsarah h. cash, md; tara t. dever, md;...

SEPTEMBER 2007 VOLUME 143, NUMBER 9 PAGES 1093-1234

ARCHIVES A CENTURY AGO

1108Some Further Observations on the Treatmentof Pigmented Hairy Naevi With Liquid Air;Three Additional Case Reports.

THE CUTTING EDGE

1110Infliximab as a Therapy for IdiopathicHypereosinophilic SyndromeJosephine A. Taverna, MD; Adam Lerner, MD;Lynne Goldberg, MD; Stephen Werth, MD;Marie-France Demierre, MD, FRCPC

STUDIES

1113Clinical Severity of Psoriasis in Last20 Years of PUVA StudyTamar Nijsten, MD, PhD;Caspar W. N. Looman, PhD;Robert S. Stern, MD

1125Clinical, Biochemical, and Genetic Study of11 Patients With Erythropoietic ProtoporphyriaIncluding One With Homozygous DiseaseCarmen Herrero, MD, PhD; Jordi To-Figueras, PhD;Celia Badenas, PhD; Manuel Mendez, PhD;Patricia Serrano, MD;Rafael Enrıquez-Salamanca, MD, PhD;Mario Lecha, MD, PhD

1131Five-Year Follow-up of a Randomized,Prospective Trial of Topical MethylAminolevulinate Photodynamic Therapy vsSurgery for Nodular Basal Cell CarcinomaLesley E. Rhodes, MD, FRCP; Menno A. de Rie, MD;Ragna Leifsdottir, MD; Raymond C. Yu, MD, FRCP;Ingeborg Bachmann, MD; Victoria Goulden, MD, FRCP;Gavin A. E. Wong, MRCP; Marie-Aleth Richard, MD;Alex Anstey, MD, FRCP; Peter Wolf, MD

1139Effect of Cold Air Cooling on the Incidenceof Postinflammatory Hyperpigmentation AfterQ-Switched Nd:YAG Laser Treatment of AcquiredBilateral Nevus of Ota–like MaculesWoraphong Manuskiatti, MD; Sasima Eimpunth, MD;Rungsima Wanitphakdeedecha, MD

1144Primary Cutaneous Diffuse LargeB-Cell Lymphoma, Leg Type:Clinicopathologic Features and PrognosticAnalysis in 60 CasesFlorent Grange, MD, PhD;Marie Beylot-Barry, MD, PhD; Phillipe Courville, MD;Eve Maubec, MD; Martine Bagot, MD, PhD;Beatrice Vergier, MD, PhD; Pierre Souteyrand, MD;Laurent Machet, MD, PhD; Sophie Dalac, MD;Eric Esteve, MD; Isabelle Templier, MD;Emmanuel Delaporte, MD; Marie-Francoise Avril, MD;Caroline Robert, MD, PhD; Stephane Dalle, MD;Liliane Laroche, MD, PhD; Michele Delaunay, MD;Pascal Joly, MD, PhD; Janine Wechsler, MD;Tony Petrella, MD

OBSERVATIONS

1153Familial Acanthosis Nigricans Dueto K650T FGFR3 MutationDavid R. Berk, MD; Elaine B. Spector, PhD;Susan J. Bayliss, MD

1157Eczematoid Graft-vs-Host Disease:A Novel Form of Chronic CutaneousGraft-vs-Host Disease and Its Responseto Psoralen–UV-A TherapyDaniel Creamer, MD;Claire L. Martyn-Simmons, MRCP(England);Genevieve Osborne, MRCP(England);Michelle Kenyon, MSc; Jon R. Salisbury, MD;Stephen Devereux, MD; Antonio Pagliuca, MD;Aloysius Y. Ho, MD; Ghulam J. Mufti, MD;Anthony W. P. du Vivier, MD

1164Epidermolysis Bullosa Nevus:An Exception to the Clinical and DermoscopicCriteria for MelanomaSarah H. Cash, MD; Tara T. Dever, MD;Patrice Hyde, MD; Jason B. Lee, MD

1168Herpes Gestationis in a Mother and Newborn:Immunoclinical Perspectives Based on a WeeklyFollow-up of the Enzyme-Linked ImmunosorbentAssay Index of a Bullous Pemphigoid AntigenNoncollagenous DomainYumi Aoyama, MD; Kanako Asai, MD; Kana Hioki, MD;Michinori Funato, MD; Naomi Kondo, MD;Yasuo Kitajima, MD

ARCHIVESOF

DERMATOLOGY

(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), SEP 2007 WWW.ARCHDERMATOL.COM1097

http://www.archdermatol.com

EVIDENCE-BASED DERMATOLOGY: STUDY

1175Balancing the Benefits and Risksof Drug Treatment: A Stated-Preference, DiscreteChoice Experiment With Patients With PsoriasisElizabeth M. Seston, PhD; Darren M. Ashcroft, PhD;Christopher E. M. Griffiths, MD, FRCP

EVIDENCE-BASED DERMATOLOGY:RESEARCH COMMENTARY

1180Use of Antibiotic Ointment After CleanCutaneous SurgeryMichael Bigby, MD

EDITORIAL

1185Evidence-Based Dermatology Section Welcomesa New Feature: Critically Appraised TopicMichael Bigby, MD

EVIDENCE-BASED DERMATOLOGY:CRITICALLY APPRAISED TOPIC

1187What Is the Chance of a Normal Pregnancyin a Woman Whose Fetus Has BeenExposed to Isotretinoin?Michael J. Sladden, MAE, MRCP(UK);Karen E. Harman, MD, MRCP

ON THE HORIZON

1189Accessible Evidence-Based Medicine:Critically Appraised TopicsDavid A. Barzilai, MD, PhD;Martin A. Weinstock, MD, PhD

EDITORIALS

1191Long-term Treatment for Severe Psoriasis:We’re Halfway There, With a Long Way to GoJoel M. Gelfand, MD, MSCE

1194Dermatology and the Human Genome:An Epidemiologic ApproachMarta Gwinn, MD, MPH; Muin J. Khoury, MD, PhD

OFF-CENTER FOLD

1201Hyperpigmented Keratotic NodulesSer Ling Chua, MRCP; Kusum Kulkarni, MRCPath;Eric Saihan, FRCP

1201Verrucous Papules and Plaquesin a Pediatric PatientMercedes E. Gonzalez, MD; Fiona P. Blanco, MD;Maria C. Garzon, MD

1201Diffuse Nodules in a Woman With Renal FailureErin B. Lesesky, MD; Michelle T. Pelle, MD;Terence C. O’Grady, MD

1201Tender Nodules on the Palms and SolesLisa Esler-Brauer, MD; Ilene Rothman, MD

CORRESPONDENCE: RESEARCH LETTERS

1209A Novel Missense Mutation in the CYLD Genein a Spanish Family With MultipleFamilial TrichoepitheliomaAgustın Espana, MD; Fermin Garcıa-Amigot, PhD;Leyre Aguado, MD; Jesus Garcıa-Foncillas, MD

1210A Novel PTPN11 Gene Mutation in a PatientWith LEOPARD SyndromeAurelie Du-Thanh, MD; Helène Cave, PharmD, PhD;Didier Bessis, MD; Carine Puso, MD;Jean-Jacques Guilhou, MD;Olivier Dereure, MD, PhD

CORRESPONDENCE: COMMENTS AND OPINIONS

1211Sponsorship of Graduate Medical Education:One Successful ModelWilliam D. James, MD

1213Narrowband UV-B Phototherapyfor Extragenital Lichen SclerosusAlexander Kreuter, MD; Thilo Gambichler, MD

1213Exclusively Benign Dermoscopic Patternin a Patient With Acral MelanomaRalph P. Braun, MD; Olivier Gaide, MD;Andreas M. Skaria, MD; Alfred W. Kopf, MD;Jean-Hilaire Saurat, MD; Ashfaq A. Marghoob, MD

1215In ReplyToshiaki Saida, MD, PhD

1216Evidence Insufficient to RecommendMelanoma Surveillance FollowingPhototherapy for JaundiceThomas B. Newman, MD, MPH;M. Jeffrey Maisels, MB BCh

1216In ReplyVincent Descamps, MD, PhD

1217Allergic Contact Dermatitis:Another Adverse Effect of Over-the-counterTopical HydrocortisoneKalman L. Watsky, MD;Erin M. Warshaw, MD, MS

1217In ReplyWilliam H. Eaglstein, MD;Scott M. Ravis

ARCHIVESOF

DERMATOLOGY



CORRESPONDENCE: VIGNETTES

1218Successful Treatment of Pityriasis VersicolorWith 5-Aminolevulinic AcidPhotodynamic TherapyYoung Jin Kim, MD;You Chan Kim, MD, PhD

1220Leukemia Cutis: A Presenting Signin Acute Promyelocytic LeukemiaTania R. Markowski, BA; Donna Bilu Martin, MD;Grace F. Kao, MD; Linda Lutz, MD;April Deng, MD, PhD; Anthony A. Gaspari, MD

1221Necrolytic Migratory Erythema: The OutermostMarker for Glucagonoma SyndromeMary L. Gantcheva, MD, PhD;Valentina K. Broshtilova, MD;Adriana I. Lalova, MD, PhD

1222Lymphomatoid Granulomatosis Inducedby Imatinib-TreatmentAmir S. Yazdi, MD; Gisela Metzler, MD;Susanne Weyrauch, MD; Mark Berneburg, MD;Michael Bitzer, MD;Hans-Konrad Muller-Hermelink, MD;Martin Rocken, MD

1223Focal Acne During Topical TacrolimusTherapy for VitiligoLucio Bakos, MD, PhD;Renato Marchiori Bakos, MD, MSc

1224Pneumocystis carinii Pneumonia in InfantTreated With Oral Steroids for HemangiomaMandi L. Maronn, MD; Timothy Corden, MD;Beth A. Drolet, MD

1225Clinical and Dermoscopic Featuresof Agminated Blue NevusMaria A. Pizzichetta, MD; H. Peter Soyer, MD;Cesare Massone, MD; Lorenzo Cerroni, MD

SKINSIGHT

1234Subungual ExostosisMarıa Elena Sanchez-Castellanos, MD;Cecilia Sandoval-Tress, MD;Patricia Ramırez-Barcena, MD

CORRECTION

1186Error in Byline in: Breaking Strengthof Barbed Polypropylene Sutures: Rater-Blinded,Controlled Comparison With NonbarbedSutures of Various CalibersRashid [M.] Rashid, MD, PhD; Mark Sartori, BS;Lucile E. White, MD; Mark T. Villa, MD;Simon S. Yoo, MD; Murad Alam, MD, MSCI

REGULAR DEPARTMENTS

1107This Month in Archives of Dermatology

1113Online CME Article

1121Archives Feature

1136Call for Papers

1162Announcement

1182Archives Web Quiz Winner

1229Classified Advertising

1232Journal Advertiser Index

1233Contact Information

Information for authors and readers and the subject andauthor indexes are available on our Web site:http://www.archdermatol.com.

ARCHIVESOF

DERMATOLOGY



Mission Statement: The Archives of Dermatology publishes information concerning the skin, its diseases, and their treatment. Its mission is to explicate the structureand function of the skin and its diseases and the art of using this information to deliver optimal medical and surgical care to the patient. We attempt to enhance theunderstanding of cutaneous pathophysiology and improve the clinician’s ability to diagnose and treat skin disorders. This journal has a particular interest in publish-ing clinical and laboratory studies that reveal new information pertinent to the interests and needs of the medical dermatologist, dermatologic surgeon, and all thoseconcerned with state-of-the-art care of cutaneous disease. We believe that knowledge derived from well-designed clinical trials and studies of cost-effectiveness areespecially important for improving the practice of dermatology. Studies that increase the understanding of the outcome of treatment or the means by which the bur-den of dermatologic disease can be measured and reduced to promote the health of patients with skin disease will receive special priority. The Archives regularlypublishes reports on clinical investigations, editorials, and reviews. It also features reports and discussions on clinicopathologic correlations; clinical disorders ofunique didactic value; pharmacologic, medical and surgical therapeutics; and ethical, moral, socioeconomic, and political issues.

Editor: June K. Robinson, MD Chicago, IllinoisAssociate Editor: Jeffrey P. Callen, MD Louisville, KentuckyEditorial Assistant: Mark D. Matthews Chicago, IllinoisEditorial Office: Archives of Dermatology

132 E Delaware Pl#5806Chicago, IL 60611Telephone (312) 943-7759; Fax (312) 943-7752e-mail: [email protected]

Section Editors

Archives a Century Ago: Mark Bernhardt, MD Ft Lauderdale, FloridaThe Cutting Edge: George J. Hruza, MD St Louis, MissouriAssistant Section Editors: Michael P. Heffernan, MD Dayton, Ohio

Christie Ammirati, MD Hershey, PennsylvaniaEvidence-Based Dermatology: Michael Bigby, MD

Boston, MassachusettsAssistant Section Editors: Damiano Abeni, MD, MPH Rome, Italy

Rosamaria Corona, DSc, MD Rome, ItalyUrbà González, MD, PhD Barcelona, SpainAbrar A. Qureshi, MD, MPH Boston, MassachusettsHywel Williams, MSc, PhD, FRCP Nottingham, England

Feature Editor: Michael J. Sladden, MAE, MRCP(UK)Launceston, Tasmania

Off-Center Fold: Michael E. Ming, MD, MSCE Philadelphia, PennsylvaniaAssistant Section Editors: Carrie Ann R. Cusack, MD

Philadelphia, PennsylvaniaSenait W. Dyson, MD Irvine, CaliforniaJacqueline M. Junkins-Hopkins, MD Philadelphia, PennsylvaniaVincent Liu, MD Iowa City, IowaKarla S. Rosenman, MD New York, New York

On the Horizon: Gary S. Wood, MD Madison, WisconsinAssistant Section Editors: Craig A. Elmets, MD Birmingham, Alabama

Molly A. Hinshaw, MD Madison, WisconsinJay C. Klemme, MD, MPH Wooster, OhioMark R. Pittelkow, MD Rochester, MinnesotaMartin A. Weinstock, MD, PhD Providence, Rhode IslandDavid T. Woodley, MD Los Angeles, California

Web Editor: Ashish C. Bhatia, MD Naperville, IllinoisCME Editor: Andrew D. Samel, MD Attleboro, MassachusettsskINsight: James M. Grichnik, MD, PhD Durham, North CarolinaThis Month: Robin L. Travers, MD Boston, Massachusetts

Editorial Board

Murad Alam, MD Chicago, IllinoisAshish C. Bhatia, MD Naperville, IllinoisMichael Bigby, MD Boston, MassachusettsMary-Margaret Chren, MD San Francisco, CaliforniaRobert P. Dellavalle, MD, PhD, MSPH Denver, ColoradoAlan B. Fleischer, Jr, MD Winston-Salem, North CarolinaJames M. Grichnik, MD, PhD Durham, North CarolinaThomas D. Horn, MD Little Rock, ArkansasGeorge J. Hruza, MD St Louis, MissouriMoise L. Levy, MD Houston, TexasMichael E. Ming, MD, MSCE Philadelphia, PennsylvaniaRonald L. Moy, MD Los Angeles, CaliforniaAnthony E. Oro, MD, PhD Stanford, CaliforniaDarrell S. Rigel, MD New York, New YorkLawrence A. Schachner, MD Miami, FloridaKathryn Schwarzenberger, MD Burlington, VermontMaria L. Chanco Turner, MD Bethesda, MarylandGary S. Wood, MD Madison, Wisconsin

Surgical Advisory Board

Murad Alam, MD, Chairman Chicago, IllinoisLeslie Baumann, MD Miami, FloridaJeffrey S. Dover, MD, FRCP Boston, MassachusettsScott W. Fosko, MD St Louis, MissouriDee Anna Glaser, MD St Louis, MissouriGlenn D. Goldman, MD Burlington, VermontSeth L. Matarasso, MD San Francisco, California

International Advisory Committee

Alvaro E. Acosta, MD Bogota, ColombiaJan Nico Bouwes Bavinck, MD Leiden, the NetherlandsReuven Bergman, MD Haifa, IsraelFrancisco G. Bravo, MD Lima, PeruLorenzo Cerroni, MD Graz, AustriaYahya Dowlati, MD, PhD Tehran, IranReinhard Dummer, MD Zurich, SwitzerlandJames Ferguson, MD Dundee, ScotlandMiguel Ruben Guarda, MD Santiago, ChileJohn L. M. Hawk, MD London, EnglandJana Hercogova, MD Prague, Czech RepublicChung-Hong Hu, MD, FACP Taipei, TaiwanAbdul-Ghani M. Kibbi, FACP, MD Beirut, LebanonAndreas D. Katsambas, MD Athens, GreeceSteven Kossard, MD Sydney, AustraliaMargarita M. Larralde, MD Buenos Aires, ArgentinaSungnack Lee, MD Suwon, KoreaThomas A. Luger, MD Munster, GermanyGillian Murphy, MD Dublin, IrelandOumeish Y. Oumeish, MD Amman, JordanFezal Ozdemir, MD Izmir, TurkeyJaime Piquero-Martin, MD Caracas, VenezuelaNiwat Polnikorn, MD, FRCP(T) Bangkok, ThailandMarcia Ramos-e-Silva, MD, PhD Rio de Janeiro, BrazilJean E. Revuz, MD Creteil, FranceLuis Requena, MD Madrid, SpainHiroshi Shimizu, MD Sapporo, JapanMohsen Soliman, MD Cairo, EgyptT. Thirumoorthy, MD SingaporeAntonella Tosti, MD Bologna, ItalyHans Christian Wulf, MD Copenhagen, DenmarkWen-Yuan Zhu, MD Nanjing, China

Publication StaffOffices: 515 N State St, Chicago, IL 60610Catherine D. DeAngelis, MD, MPH

Editor in ChiefScientific Publications & Multimedia Applications

Cheryl Iverson: Managing EditorEditorial Processing Department

Department Director: Paula GlitmanManuscript Editing Manager: Lauren B. FischerSenior Manuscript Editors: Dee Egger, John McFadden,

Donald S. SmithManuscript Editor: Karen L. BoydFreelance Manuscript Editing Manager: Diane L. CannonFreelance Manuscript Editing Coordinators: Cheryl Tobey Davidow,

Vickey Golden, Connie MannoFreelance Copy Editors and Copyreaders:

Catherine A. Alden, Lisa M. Cluver, Mary Coerver-Connolly,Allison Esposito, GeorgeMary Gardner, Robin Husayko, Laura King,Mary A. Kingzette, Nicole C. Netter, Beena Rao, Caroline Woods

Editorial Systems Managers: Paul Frank, Monica MungleSystems Programmer/Internet Specialist: J.D. NeffEditorial Assistant: Wanda HillScientific Publications & Multimedia Applications

Executive Assistant to the Editor: Marla A. HallScientific Online Resources

Director: Margaret Winker, MDElectronic Media Department Director: Michelle KurzynskiAssociate Web Editors: Christopher Hastings, Peter KimWeb Administrative Assistant: Reuben Rios

Periodic Publishing & Business StaffSenior VP & Publisher: Robert A. Musacchio, PhD

Advertising Sales & Marketing

Division Director: Jeffery J. BonistalliManager: Kathy RussellStaff: Mirna MonalSales Representatives: Peter G. Messina, Patricia QuiltyAssistant Publisher, Classified Advertising/

Circulation Promotion: Geoffrey FlickManager: Thalia MossStaff: Tem Harrier, Kim Jenkins, Allison KruseDirector, Marketing Support Services: Mark ThornburyStaff: Nancy Bell, Tricia Leonardis, Dennis SammartinoDirector, Reprints: Mark KuhnsStaff: Susan Lovenberg

Journal Sales & Online Business

Division Director: Richard NewmanDirector, Worldwide Sales: Vida DamijonaitisCirculation Sales & Marketing:Managers: Silvana Benzoni, Saskia Bolore, Jeremy NielsenStaff: Rebecca Haas, Christine HearneDirector, Marketing and Product Development:

Michael T. ClarkeStaff: Millette Jackson-Bates, Sara Nicholson

Publishing Operations

Division Director: Vanessa HaydenDirectors: Karen Adams-Taylor, Sue Sherrill

Managers: David Antos, Karen Branham,Suzanne J. Bukovsky, Debbie Camp, Susan Chiara,Diane Darnell, Susana Eloy de Vitale, Karl Elvin,Mary Ellen Johnston, Michael McGraw, Chris Meyer,Phil O’Leary, Ryan Patella, Susan Price, Stacy Tucker

Advertising and Production Services: Karen Allbright,Bernadette Bauer, Rhonda Bailey Brown, Anna Goins,Terri Hanley, Euwanaha Joiner-Smith, Brian McCartney,Sean O’Donnell, Noemi Olmo, Cindy Rendlen,Denise Steinhauser, Ellen Yoshihara

Circulation Processing: Julie Burton, Michael Deegan,Rosa Eng, Donovan Hanson, Felicia Kyles,Lawrence Majkszak, Kathryn Osten

Composition and New Media: Gail Barrett,Regina Brownlee, Brenda Chandler-Grayer, Scott Curl,Amy Evers, Teri Hutchison, Mary Ann Kuranda,Joshua Lampinen, Patricia E. Panek,La’son Diggs Sledge, George Thomson

Graphics: Joe Amft, Fred Furtner, Carolyn Hall,Maria Kowalkowski, Ted Piwowar, JoAnne Weiskopf

Proofreading: Carol Joy Farrell, Daniel James,Judith A. Literskis, Teresa H. Omiotek,Barbara Wojtowicz, Agnes A. Zarkadas

AMA Executives

Michael D. Maves, MD, MBAExecutive Vice President, Chief Executive Officer

Bernard L. HengesbaughChief Operating Officer

Robert A. Musacchio, PhDSenior Vice President, Publishing & Business Services

All articles published, including editorials, letters, and book reviews, represent the opinions of the authorsand do not reflect the policy of the American Medical Association, the Editorial Board, or the institutionwith which the author is affiliated, unless this is clearly specified.

Copyright 2007 by the American Medical Association. All rights reserved. Reproduction without permis-sion is prohibited.

ARCHIVESOF

DERMATOLOGY



THIS MONTH IN ARCHIVES OF DERMATOLOGY

SECTION EDITOR: ROBIN L. TRAVERS, MD

Effect of Cold Air Coolingon the Incidenceof PIH After Q-SwitchedNd:YAG Laser Treatmentof Acquired Bilateral Nevusof Ota–like Macules

P ostinflammatoryhyperpigmen-tation (PIH) is very likely the

most common adverse effect of lasertreatment in dark-skinned individu-als. Treatment of PIH is difficult, thusemphasizing the need for preventionstrategies. Acquired bilateral nevus ofOta–like macules (ABNOMs) areclinically characterized by blue-brown macules on periocular and na-sal skin, and Q-switched lasers rep-resent an excellent therapeuticmodality. In this randomized, con-trolled, split-face study, Manuskiattiet al sought to evaluate the benefit ofepidermalcoolingonreducingtheoc-currence of PIH after Q-switchedNd:YAG laser treatment of ABNOMs.Unexpectedly,coldaircoolingwasas-sociated with an increased risk of PIHafter laser treatment, through an un-known mechanism.

See page 1139

Herpes Gestationis in aMother and Newborn

H erpesgestationis(HG)isarare,autoimmune, bullous disease

that affects women in the second orthird trimester of pregnancy. Immu-nologically,HGischaracterizedbylin-ear deposition of C3 with or withoutassociated IgG at the basement mem-brane zone on direct immunofluo-rescence and by the presence of se-rum antibodies directed against the180-kDa bullous pemphigoid anti-gen (BP180). In this case report,Aoyama et al describe a mother andneonate with cutaneous manifesta-tions of HG. The BP180 enzyme-linked immunosorbent assay in-dexes were observed in both patientsand clearly demonstrated that the ve-sicular lesions in the neonate werecaused by transplacental passage ofpathogenic BP180 antibodies.

See page 1168

Eczematoid Graft-vs-Host Disease

G raft-vs-host disease (GVHD) is a multisystem disease initiated by allo-geneic T lymphocytes that recognize foreign tissue antigens in the host.

In the chronic form, GVHD demonstrates cutaneous manifestations that maybe classified as lichenoid or sclerodermatous in nature. In this case series ofpatients who developed cutaneous GVHD after hematopoietic stem cell trans-plantation, Creamer et alidentified a novel, distinctgroup of 10 who were char-acterizedbypersistent,wide-spread, chroniceczematousdermatitis. Histopathologi-cally, these eczematous fea-tures strikingly coexistedwithchangesofGVHD.Thefactors provoking the ex-pression of an eczematousphenotype in these patientsremain unclear.

See page 1157

Clinical, Biochemical, and Genetic Study of 11 PatientsWith Erythropoietic Protoporphyria Including OneWith Homozygous Disease

E rythropoietic protoporphyria (EPP) is an inherited disorder of porphy-rin metabolism caused by a partial deficiency of the ferrochelatase gene

(FECH). Clinical expression of the disease commonly results from the coin-heritance of a null FECH mutation plus a wild-type low-expressed allele, whichfurther reduces ferrochelatase activity to below a critical threshold of about 35%.The low-expression allele IVS3-48C is relatively common among European popu-lations, such that overt EPP is typically inherited in such a way that it re-sembles an autosomal dominant disease with low clinical penetrance. In thiscase series, Herrero et al confirm the common view that clinically overt EPP ismost typically associated with inheritance of the low-expression IVS3-48C al-lele in trans to a severe FECH mutation. One rare homozygous mutation wasidentified in a patient showing early signs of liver involvement.

See page 1125

Balancing the Benefits and Risks of Drug Treatment

P soriasis is a common, chronic, inflammatory, hyperproliferative skin dis-ease. Therapeutic options for psoriasis vary widely in terms of efficacy and

toxic effects. In this discrete choice experiment, Seston et al explore the extentto which different attributes of a treatment affect a patient’s choice of therapyand the hierarchical importance of these attributes. Most respondents were will-ing to trade among different attributes of a treatment to achieve an improve-ment in their psoriasis with minimal adverse effects. For example, patients in-dicated that they would wait longer for therapy to work if the chance of a severeadverse effect could be reduced. Long-term risks (such as skin cancer or liverdamage) were prioritized over short-term risks (such as skin irritation). Onlyone-quarter of the sample patients were unwilling to trade among attributes andalways chose the therapy with a particular level of their chosen attribute, mostoften no risk of skin cancer.

See page 1175

A B



ARCHIVES A CENTURY AGO

SECTION EDITOR: MARK BERNHARDT, MD

THE JOURNAL OFCUTANEOUS DISEASES

VOL. XXV. SEPTEMBER, 1907. NO. 9.

SOME FURTHER OBSERVATIONS ON THE TREATMENTOF PIGMENTED HAIRY NAEVI WITH LIQUID AIR;

THREE ADDITIONAL CASE REPORTS.

By WILLIAM B. TRIMBLE, M. D.

Everyone is more or less familiar with this new remedial agent, yet it may be well to quote heresome of its physical characteristics. Liquid air is the common air we breathe, very much reduced intemperature. It liquefies at 312º F. below zero. We can readily understand how cold it must be whenwe are told that it is about 400º colder than the ordinary atmospheric air.

J Cutan Dis.September 1907;25(9);409-413.

The bridge at midnight trembles,The country doctor rambles,Bankers’ nieces seek perfection,Expecting all the gifts that wise men bring.The wind howls like a hammer,The night blows cold . . .

Love Minus Zero/No LimitBob Dylan, 1965

Temperature, °C Occurrence

�116 At this temperature, Hell’s lakes of sulfuric brimstone freeze over0 Temperature of water when Dr John Gorrie created the first artificial ice

(Apalachicola, Florida, July 14, 1850)a

−2 Coldest temperature recorded in West Palm Beach, Florida (January 19, 1977)−4 to −20 Temperature of ethyl chloride spray, depending on delivery system−20 to −30 Your tongue sticks to a metal pole, stupid!−39 Temperature at which the bulb of mercury in a thermometer (aka, the witch’s tit)

freezes−57 Temperature of over-the-counter wart treatment−70 Coldest temperature recorded in contiguous 48 states (Rogers Pass, Montana,

January 20, 1954)−75 Exposed skin freezes within 30 seconds−78 Surface temperature of a block of dry ice−89 Coldest air temperature recorded on earth (Vostok, Antarctica, July 21, 1983)−196 Temperature of liquid nitrogen in which Ted Williams’ head is preserved−230 Average surface temperature on the dark side of the moon−235 Lowest temperature in our solar system (Neptune’s moon Triton)−271 Temperature of intergalactic deep space−272.000000083 Bose-Einstein condensate first formed (Boulder, Colorado, June 5, 1995)−273 Absolute zero (Sorry, Bob. That is the limit.)

aBernhardt M. Selections: Morbus Maculosus Werlhofu. Arch Dermatol. 1992;128(9):1298.



THE CUTTING EDGE

SECTION EDITOR: GEORGE J. HRUZA, MD; ASSISTANT SECTION EDITORS: MICHAEL P. HEFFERNAN, MD; CHRISTIE AMMIRATI, MD

Infliximab as a Therapy for IdiopathicHypereosinophilic SyndromeJosephine A. Taverna, MD; Adam Lerner, MD; Lynne Goldberg, MD; Stephen Werth, MD; Marie-France Demierre, MD, FRCPC;Boston University School of Medicine (Drs Taverna, Lerner, Goldberg, and Demierre) and Tufts University Schoolof Medicine (Dr Werth), Boston, Massachusetts

The Cutting Edge: Challenges in Medical and Surgical Therapeutics

REPORT OF A CASE

A 77-year-old woman from Guyana presented with a3-year history of a generalized, pruritic eruption associ-ated with chronic chills, fatigue, generalized malaise, anda 14-kg weight loss over a 9-month period (Figure 1).Her medical history was significant for hypertension anddiabetes, and her medications included glucophage, aten-olol, and lisinopril. She had no personal or family his-tory of atopy. Her physical examination findings were re-markable for exfoliative erythroderma and keratodermaassociated with generalized lymphadenopathy (Figure 1).Complete blood cell counts were normal except for a per-sistently elevated absolute eosinophil count (1500-3000/µL) for more than 8 months. Positron emission tomo-graphic scans detected enhanced fludeoxyglucose F 18uptake bilaterally in the supraclavicular, axillary (larg-est node measuring 2.5 cm), and inguinal nodes (larg-est node measuring 2 cm). A left inguinal lymph nodebiopsy specimen revealed dermatopathic lymphadenitisand no evidence of a lymphoproliferative disorder. His-topathologic examination of a skin sample revealed para-keratosis, hypogranulosis, psoriasiform epidermal hy-perplasia, mild spongiosis, a mild superficial perivascularlymphocytic infiltrate with occasional plasma cells and

rare eosinophils, and numerous melanophages. Flow cy-tometry and skin and bone marrow biopsy specimens re-vealed no evidence of a clonal T- or B-cell population onimmunophenotyping or molecular genetic analysis, andFIP1L1-PDGFRA gene fusion was not detected. Exten-sive studies were performed to rule out infectious, aller-gic, vasculitic, rheumatic, and malignant causes of eo-sinophilia. A transthoracic echocardiogram, which wasobtained to rule out heart failure or mural thrombus inthe setting of chronic hypereosinophilia, demonstratedno abnormalities.

THERAPEUTIC CHALLENGE

The pruritic eruption was initially thought to be causedby the use of a newly prescribed medication (lisinopril).Unfortunately, erythroderma and keratoderma per-sisted long after therapy with the angiotensin-converting enzyme inhibitor was discontinued. In the fol-lowing 3 years, the patient was treated unsuccessfully witha variety of therapeutic modalities, including systemic cor-ticosteroids, psoralen–UV-A, isotretinoin, bexarotene, andextracorporeal photopheresis. Two years after her ini-tial presentation, the patient required inpatient hospi-talization for intense pruritus and failure to thrive, and

A B

Figure 1. Exfoliative erythroderma (A) and keratoderma (B) before treatment with infliximab.



the decision was made to initiate treatment targeting T-cell activation. After 3 monthly cycles of denileukin difti-tox infusions (18 µg/kg/d) in combination with oral bex-arotene, there was no improvement in erythroderma orpruritus. A 1-month course of oral imatinib mesylatetherapy (400 mg/d) was not helpful. A trial of chemo-therapy with 2 cycles of fludarabine–mitoxantrone–dexamethasone (fludarabine phosphate, 25 mg/m2 days1-3; mitoxantrone, 10 mg/m2 day 1; and dexametha-sone, 20 mg/d days 1-5), resulted in a transient decreasein the patient’s absolute eosinophil count but no clini-cal improvement. An effective treatment was needed toameliorate her skin lesions and severe pruritus.

SOLUTION

Intravenous infusions of infliximab (5 mg/kg) were fol-lowed by a sharp decrease in the peripheral blood eo-sinophil count to 700/µL (Figure2), with complete reso-lution of erythroderma, keratoderma, and pruritus after3 monthly cycles of therapy (Figure 3). For mainte-nance, the patient received 5 additional monthly cyclesof infliximab that were well tolerated, without reportedadverse effects or evidence of recurrent disease. Inflixi-mab therapy was subsequently discontinued, and the pa-

tient’s skin remained clear at the 1-month follow-up visit.Her debilitating skin condition, which was associated withhypereosinophilia, had responded quickly and dramati-cally to infliximab therapy alone, resulting in improvedsleep and functional status.

COMMENT

Hypereosinophilic syndrome (HES) is a heterogeneousgroup of conditions that are characterized by blood hy-pereosinophilia and end-organ dysfunction due to eo-sinophilic infiltration and toxic mediator release.1-3 End-organ damage commonly affects the skin, heart, lungs,and nervous system. Idiopathic HES (IHES), which is adiagnosis of exclusion, is defined as blood eosinophiliaexceeding an eosinophil count of 1500/µL for more than6 consecutive months, of unknown etiology, with evi-dence of end-organ damage.1-3 The distinctions betweenmyeloid and lymphoid lineages of IHES are based on dif-ferences in molecular pathogenesis, clinical presenta-tion, complications, and prognosis and are necessary toguide patient treatment.4 Lymphocytic HES (l-IHES) ischaracterized by a nonmalignant expansion of a T-cellpopulation that produces eosinophilopoetic cytokines (ie,interleukin 5). It carries a good prognosis, with pro-longed survival, and may respond to glucocorticoidtherapy or to treatment with human monoclonal anti–interleukin 5 antibody.4,5 Cutaneous lesions are com-mon in l-HES and vary from pruritic erythematous mac-ules, papules, plaques, or nodules to urticaria andangioedema.4 Myeloproliferative HES (m-IHES) is char-acterized by clonally derived eosinophils, carries a poorprognosis, and may respond to imatinib mesylate (a ty-rosine kinase inhibitor) therapy, particularly in the pres-ence of FIP1L1-PDGFRA gene fusion.4,6 If left untreated,patients with m-IHES can undergo a rapidly fatal courseowing to congestive heart failure or to the developmentof acute leukemic disease.4 Other therapies for m-IHESreported in the literature include hydroxyurea, inter-feron alfa,7 cyclosporine, and allogenic stem cell trans-plantation.8 Based on the results of clinical and molecu-lar analysis, our patient had l-IHES: she presented withcutaneous end-organ damage in the absence of heart in-

3500

3000

2500

2000

1500

1000

500

0Apr

2005Jun

2005Aug2005

Oct2005

Dec2005

Feb2006

Apr2006

Jun2006

Month

Eosi

noph

ils, µ

L

Series 1

Feb2005

Figure 2. Eosinophil count in a patient with idiopathic hypereosinophilicsyndrome. Infliximab therapy was initiated in late November 2005, with asharp decrease in absolute eosinophil counts. Normalization of eosinophilcount was achieved after 3 monthly doses of infliximab, and the effect wasmaintained after 7 monthly doses.

A B

Figure 3. Full response to infliximab following 7 monthly infusions, with complete resolution of erythroderma (A) and keratoderma (B).



volvement, failed to improve with imatinib mesylatetherapy, and did not harbor the FIP1L1-PDGFRA fusiongene. We report the first case (to our knowledge) of l-IHESthat was responsive to infliximab therapy.

Infliximab is a chimeric, human-mouse antibody for tu-mor necrosis factor (TNF) � that has been shown to be ef-fective in the treatment of Crohn disease, rheumatoid ar-thritis, severe psoriasis (ie, recalcitrant erythrodermicpsoriasis),9-11 and pityriasis rubra pilaris.12 Tumor necro-sis factor � is a proinflammatory cytokine that contrib-utes to lymphocyte and eosinophil recruitment by up-regulating vascular cell adhesion molecule 1 (VCAM-1),endothelial leukocyte adhesion molecule 1 (ELAM-1), andintercellular adhesion molecule 1 (ICAM-1) on activatedendothelial cells.13-15 The precise mechanism by which in-fliximab therapy could induce remission in an erythroder-mic patient with HES is not fully understood but could bethe result of inhibition of TNF-�–induced eosinophil-dependent recruitment, toxic effects, and leukocyte infil-tration. It has been proposed that through a positive feed-back loop of TNF interacting with TNF receptors onkeratinocytes, “secondary” keratinocyte cytokines are re-leased, including ICAM-1.16 Also, Costa et al17 have shownthat blood eosinophils isolated from hypereosinophilic pa-tients can also represent a potential source of TNF-�. Byblocking TNF-� expression from primary sources in theskin, TNF-� inhibitors could break the cycle of cytokinefeedback by down-regulating VCAM-1, ICAM-1, andELAM-1 expressions on dermal endothelial cells, therebypreventing eosinophils from homing to skin and subse-quent toxic mediator release.

Several adverse effects of infliximab therapy have beenreported. In a recent meta-analysis, Bongartz et al18 showedan increased risk of serious infection and a dose-dependent increased risk of malignancy in patients withrheumatoid arthritis treated with anti–TNF antibodytherapy. For this reason, the decision to initiate TNF-�therapy requires careful thought, andpatientsmustbemoni-tored closely during treatment. In our patient, the clinicalresponse resulted in significantly improved functioning andsymptom relief. The efficacy of a treatment suggested by asingle report may serve to provide clues about the patho-genesis of the disease but will need to be subjected to fur-ther study before the treatment is widely used.

Accepted for Publication: September 28, 2006.Correspondence: Marie-France Demierre, MD, FRCPC,Boston University School of Medicine, Skin Oncology Pro-gram, Boston Medical Center, 720 Harrison Ave,DOB801A, Boston, MA 02118 ([email protected]).Author Contributions: Dr Demierre had full access toall the data in the study and takes responsibility for theintegrity of the data and the accuracy of the data analy-sis. Study Concept and Design: Taverna, Lerner, andDemierre. Acquisition of Data: Taverna, Lerner, Werth,and Demierre. Analysis and Interpretation of Data: Taverna,Lerner, Goldberg, and Demierre. Drafting of the Manu-script: Taverna. Critical Revision of the Manuscript for Im-portant Intellectual Content: Taverna, Lerner, Goldberg,Werth, and Demierre. Obtained funding: Demierre. Ad-ministrative, Technical, and Material Support: Taverna,

Goldberg, and Demierre. Study Supervision: Taverna,Goldberg, Werth, and Demierre.Financial Disclosure: Dr Demierre has acted as a speakerfor Schering, Ligand, Therakos, Merck, and Gloucester;has received research support from Ligand, Therakos,Merck, Gloucester, Genmab, Curagen, and Novartis; hasbeen a consultant for Ligand and Novartis; and has servedon the advisory board of Gloucester.Previous Presentation: This work was presented as a caseseries at the Annual Fall meeting of the MassachusettsAcademy of Dermatology; September 17, 2006; New-port, Rhode Island.Additional Contributions: We thank Robert E. Tigelaar,MD, for thoughtful discussions and Swarne Adikari fortechnical photographic assistance.

REFERENCES

1. Chusid MJ, Dale DC, West BC, Wolff SM. The hypereosinophilic syndrome: analy-sis of fourteen cases with review of the literature. Medicine (Baltimore). 1975;54(1):1-27.

2. Brito-Babapulle F. The eosinophilias, including the idiopathic hypereosinophilicsyndrome. Br J Haematol. 2003;121(2):203-223.

3. Tefferi A. Blood eosinophilias: a new paradigm in disease classification, diagno-sis, and treatment. Mayo Clin Proc. 2005;80(1):75-83.

4. Roufosse F, Cogan E, Goldman M. Recent advances in pathogenesis and man-agement of hypereosinophilic syndromes. Allergy. 2004;59(7):673-689.

5. Sutton SA, Ass’ad AH, Rothenberg ME. Anti–IL-5 and hypereosinophilic syndromes.Clin Immunol. 2005;115(1):51-60.

6. Cools J, DeAngelo DJ, Gotlib J, et al. A tyrosine kinase created by fusion of thePDGFRA and FIP1L1 genes as a therapeutic target of imatinib in idiopathic hy-pereosinophilic syndrome. N Engl J Med. 2003;348(13):1201-1214.

7. Yoon TY, Ahn GB, Chang SH. Complete remission of hypereosinophilic syn-drome after interferon-alpha therapy: report of a case and literature review.J Dermatol. 2000;27(2):110-115.

8. Ueno NT, Anagnostopoulos A, Rondon G, et al. Successful non-myeloablativeallogeneic transplantation for treatment of idiopathic hypereosinophilic syndrome.Br J Haematol. 2002;119(1):131-134.

9. Fiehn C, Andrassy K. Case number 29: hitting three with one strike: rapid im-provement of psoriatic arthritis, psoriatic erythroderma, and secondary renal amy-loidosis by treatment with infliximab (Remicade). Ann Rheum Dis. 2004;63(3):232.

10. O’Quinn RP, Miller JL. The effectiveness of tumor necrosis factor alpha anti-body (infliximab) in treating recalcitrant psoriasis: a report of two cases. ArchDermatol. 2002;138(5):644-648.

11. Winterfield LS, Menter A, Gordon K, Gottlieb A. Psoriasis treatment: current andemerging directed therapies. Ann Rheum Dis. 2005;64(suppl 2):ii87-ii92.

12. Liao WC, Mutasim DF. Infliximab for the treatment of adult-onset pityriasis ru-bra pilaris. Arch Dermatol. 2005;141(4):423-425.

13. Bochner BS, Luscinskas FW, Gimbrone MA, Newman W. Adhesion of humanbasophils, eosinophils and neutrophils to interleukin 1–activated human vascu-lar endothelial cells: contributions of endothelial cell adhesion molecules. J ExpMed. 1991;173(6):1553-1557.

14. Kyan-Aung U, Haskard DO, Poston RN, Thornhill MH, Lee TH. Endothelial leu-kocyte adhesion molecule-1 and intercellular adhesion molecule-1 mediate theadhesion of eosinophils to endothelial cells in vitro and are expressed by endo-thelium in allergic cutaneous inflammation in vivo. J Immunol. 1991;146(2):521-528.

15. Weller PF, Rand TH, Goelz SE, Chi-Rosso G, Lobb RR. Human eosinophil ad-herence to vascular endothelium mediated by binding to vascular cell adhesionmolecule 1 and endothelial leukocyte adhesion molecule 1. Proc Natl Acad SciU S A. 1991;88(16):7430-7433.

16. Williams IR, Kupper TS. Immunity at the surface: homeostatic mechanisms ofthe skin immune system. Life Sci. 1996;58(18):1485-1507.

17. Costa JJ, Matossian K, Resnick MB, et al. Human eosinophils can express thecytokines tumor necrosis factor alpha and macrophage inflammatory protein-1alpha. J Clin Invest. 1993;91(6):2673-2684.

18. Bongartz T, Sutton AJ, Sweeting MJ, Buchan I, Matteson EL, Montori V. Anti-TNF antibody therapy in rheumatoid arthritis and the risk of serious infectionsand malignancies: systematic review and meta-analysis of rare harmful effectsin randomized controlled trials. JAMA. 2006;295(19):2275-2285.



STUDY

Clinical Severity of Psoriasisin Last 20 Years of PUVA StudyTamar Nijsten, MD, PhD; Caspar W. N. Looman, PhD; Robert S. Stern, MD

Objective: To assess the severity of psoriasis over time.

Design: We analyzed the results of structured derma-tologic examinations administered over a 20-year pe-riod beginning 10 years after study enrollment.

Setting: The PUVA [psoralen–UV-A] Follow-up Study,which is a prospective cohort study.

Patients: The analyses were restricted to 815 patients(83.2% of those eligible) who underwent at least 2 of 4possible examinations between 1985 and 2005.

Main Outcome Measure: A 4-point physician globalassessment (PGA).

Results: The distribution of the PGA levels in the studygroup did not change significantly over time, except thatin 2005 more patients had no psoriasis compared with

patients who underwent examinations in the previousstudy years (9.6% vs �5.1%, P� .03). The PGA levelchanged more than 1 level between examinations in only14% of patients. Multistate Markov models estimated thatpatients had a likelihood of about 80% to remain at thesame PGA level 1 year later. After 10 years, this likeli-hood varied between 19% and 53%, depending on thePGA level. Except for patients who were clear of diseaseat baseline, on average patients had about 1 year with-out psoriasis over 20 years. On average, individuals withmoderate to severe disease remained at these levels for11 or more years.

Conclusion: Three decades after a large and diverse groupof patients sought a cure for their psoriasis, consistentcontrol of their psoriasis often had not been achieved.

Arch Dermatol. 2007;143(9):1113-1121

P S O R I A S I S I S A C O M M O N

chronic skin condition thataffects about 2% of the peoplein the United States.1 Al-though it is widely accepted

that psoriasis waxes and wanes and var-ies substantially in severity among pa-tients, the severity of psoriasis over a long

period in a single person or group has notbeen well documented. In the 1960s and1970s, several cross-sectional studies basedon self-reports and without clearly de-fined outcomes estimated that 40% of pa-tients with psoriasis had experienced a“complete remission” at least once in theirlifetime.2-4 In clinical trials with a maxi-mum of 16 weeks of follow-up, the meanrelative change of the psoriasis area andseverity index ranged from −20% to 50%among almost 500 placebo-treated pa-tients.5 A recent study using the GeneralPractice Research Database found that theprevalence of psoriasis, as reported as anindependent diagnosis, declined in pa-

tients older than 70 years, suggesting thatpsoriasis may go into remission in el-derly patients.6

The severity of psoriasis in a given in-dividual at any point reflects both endog-enous (underlying disease activity) and ex-ogenous factors, including treatment. Theshort- to intermediate-term effects of thera-pies on disease severity have been exam-ined in hundreds of clinical trials, but theyare unlikely to be representative of the pso-riasis severity over a long period amongindividuals who have this disease for anaverage of 40 or more years. Also, per-sons entering clinical trials are likely tohave worse psoriasis than their usual state.Patients who are receiving active treat-ment are likely to have less disease, andthere is a discrepancy between the effi-cacy of treatments observed in clinical trialsand their effectiveness in daily life.

The PUVA [psoralen–UV-A] Fol-low-up Study is a long-term prospectivecohort safety study, which has included

For editorial commentsee page 1191

CME available online atwww.archdermatol.com

Author Affiliations:Departments of Dermatology(Dr Nijsten) and Public Health(Dr Looman), Erasmus MedicalCenter, Rotterdam,the Netherlands; andDepartment of Dermatology,Beth Israel Deaconess MedicalCenter, Harvard MedicalSchool, Boston, Massachusetts(Dr Stern).



study-sponsored structured dermatologic examina-tions, including an assessment of psoriasis severity usinga physician global assessment (PGA) scale administeredover nearly 30 years.7 In contrast to earlier examina-tions in the first study decade, by 1985 the timing of theseexaminations was not coordinated with therapeutic de-cisions and/or disease status. Therefore, the results of thelast 4 examinations were used in this study. The pri-mary objective of the present study was to analyze changeof psoriasis severity among participants in a prospectivecohort study over the last 2 decades.

METHODS

THE PUVA FOLLOW-UP STUDY

The PUVA Follow-up Study is a long-term safety and efficacystudy that has prospectively studied 1380 patients with mod-erate to severe psoriasis who were first treated with PUVA in1975 and 1976 at 16 university centers in the United States. In1977, a total of 1380 of 1450 patients who participated in theoriginal trial enrolled in the PUVA Follow-up Study. By 2005,22 cycles of follow-up interviews and 9 cycles of physical ex-aminations had been completed. The Declaration of Helsinkiprotocols were followed, and at time of enrollment all patientsprovided written informed consent. A more detailed descrip-tion of this cohort study has been published previously.8,9

GLOBAL PSORIASIS SEVERITY

During 5 periods of the cohort study, from entry (1975-1976)to 1983, members were invited to come to a participating cen-ter for an in-person follow-up interview (which assessed treat-ment exposures and health-related outcomes since last inter-view) and a structured dermatologic examination, both of whichwere performed annually by a local investigator (Figure 1).These examinations were very likely to coincide with medicalcare. By 1985, a coordinating center conducted telephone in-terviews, with the exception of 1 center, which interviewed itsown patients. At the time of the telephone interview in desig-nated years, the interviewer helped to arrange the patient’s clini-cal examination, which, when possible, was performed at a par-ticipating center or, alternatively, at the patient’s currentdermatology department and was performed relatively inde-pendently of clinical status or care. Also, by 1985, PUVA use

had decreased substantially (Figure 1). In this report, the der-matologic examinations analyzed were performed during 4 pe-riods (1985-1986, 1990-1991, 2000-2001, and 2003-2005). Thestudy provided instructions and standardized forms to all ex-amining physicians for scoring several clinical cutaneous signs.Each clinical examination included a 4-point assessment of PGA(clear, mild, moderate, and severe).

STUDY POPULATION

Cohort members were included in this analysis if they had un-dergone 2 or more PGA assessments between 1985 and 2005.To be eligible, patients must have undergone a second exami-nation after 1985; therefore, they had to have received fol-low-up at least until 1991. In 1991, a total of 979 patients werestill participating in the study, 341 had died, 37 were unavail-able for follow-up, and 23 had withdrawn from the study(Figure 1). A comparison of the 979 cohort members who werefollowed up in 1991 and the 815 patients who were eligible toenter this study demonstrated that the study patients were sig-nificantly more likely to be male (66.6% vs 61.7%, P=.03), butmen and women were of comparable age (mean [SD], 56.7 [12.9]years vs 55.9 [15.9] years, P=.16). Up to 1986, the cumulativeexposures to PUVA, UV-B, and methotrexate were compa-rable between the study patients and those not included (mean[SD], 188.5 [137.6] PUVA treatments vs 203.3 [163.1] PUVAtreatments, P=.12; 248.6 [351.1] UV-B treatments vs 288.5[458.5] UV-B treatments, P=.31; and 15.6 [31.2] months ofmethotrexate therapy vs 17.4 [32.9] months of methotrexatetherapy, P=.27, respectively).

At time of the enrolment in the PUVA study, the 815 studypatients were significantly younger and more likely to be fe-male than the other 565 cohort members (mean age, 41.5years vs 47.1 years [P� .001] and 64.2% women vs 56.4%women [P=.005], respectively). At time of the first PGA as-sessment (1977, N=1192), study participants were signifi-cantly more likely to have more moderate to severe diseasecompared with the remaining patients (18.9% vs 32.9%[P� .001], respectively).

STATISTICAL ANALYSIS

We used the �2 test to determine the statistical significance ofdifference in the distribution of categorical variables and the ttest for continuous variables. For paired comparisons, the Wil-coxon signed rank test and the McNemar test were used as ap-propriate. To test for statistical differences between multiple

Periodical clinical examinations likely not to be related to seeking care

3 5 19 15 37 425164 92 89

1975Start

2005Start

1977-1982 1986 1991 2001

PUVAapproval

Annual clinicalexaminations likely to be related to seeking care

Patients receivingPUVA therapy, %∗

Figure 1. Time line of the clinical examinations for the entire PUVA [psoralen–UV-A] Follow-up Study and for the study period. Asterisk indicates the percentage ofpatients who received 10 or more PUVA treatments for that year.



groups, 2-way analysis of variance or the Kruskal-Wallis testwas used as appropriate. To assess the intraindividual variabil-ity of the PGA scorings, we calculated turnover tables for allintervals and pooled them. We applied multistate Markov mod-eling, which describes the process in which an individual movesthrough a series of states in continuous time.10 The multivar-iate statistical modeling package for the statistical program Rwas used to estimate a transition intensity matrix (hazard ma-trix: hazard ratios with 95% confidence intervals [CIs] [datanot shown]), which were subsequently transformed to propor-tions (without 95% CIs because of software limitations).11 First,multistate Markov modeling was used to estimate the transitrates between levels of PGA after 1, 5, and 10 years. We chosea model that allowed changes only to subsequent PGA stagesbut used a likelihood ratio test to compare it with a model thatallowed all possible changes. Second, Markov models esti-mated the length of stay of the total study period (20 years) ineach of the PGA levels stratified by the patients’ PGA level atbaseline in 1986.

To study the effect of aging and disease duration on the4-point PGA, we compared the distribution of these variablesacross the PGA levels at each clinical examination. Also, a mul-tivariate ordered logistic regression model, which is a type oflogistic regression model with an ordinal-depending variablethat has more than 2 categories, was used to calculate ad-justed odds ratios (ORs) and 95% CIs. For the 4 analyses, col-linearity was absent and the proportional odds assumption wastested using a likelihood ratio test and a Brant test, the resultsof which were nonsignificant (P� .05).

RESULTS

THE COHORT OVER TIME

Figure 2 demonstrates that of the patients who wereactively followed up in each year of the PUVA study, theoverwhelming majority were interviewed, and that thenumber of participants who were unavailable for fol-low-up or withdrew from the study was very small.Death of the cohort members was the primary reason

for the decreasing number of participants between 1975and 2005.

CHARACTERISTICS OFSTUDY POPULATION

About three-quarters of the cohort members who werefollowed up participated in the clinical examinations(Table 1). Of the 979 patients, 815 (82.3%) under-went 2 or more clinical examinations between 1985 and2005. The majority of 815 study participants were middle-aged male patients who had psoriasis for more than 25years (Table 1). The age at onset of psoriasis ranged from0 to 68 years (mean [SD] age, 19.0 [17.8] years). Graphi-cally, no bimodal distribution was detected. A little fewerthan 50% of the patients had PGA scores of moderate tosevere psoriasis at 1 or more of the examinations. Thedistribution of PGA levels remained stable over a periodof 20 years, except that a larger percentage of patientswere reported clear of psoriasis at the 2005 examinationcompared with the prior examinations (9.6% vs 4.5%,3.9%, and 5.1%, respectively; P� .03). However, the per-centage of patients with none to mild and moderate tosevere psoriasis was comparable at the different exami-nations (56.8% vs 52.1%, 52.6%, and 55.0%, respec-tively; P� .13).

DISTRIBUTION

In 1986, 1991, 2001, and 2005, the total number of clini-cally examined patients was 889, 979, 580, and 424, re-spectively. Of the patients who were followed up after1991, 92.2% of the 884 patients who underwent 1 ex-amination between 1985 and 2005 underwent a secondexamination. Of the study patients who were followedup in 2005, 8.8% underwent 2, 30.5% underwent 3, and60.8% underwent 4 examinations. The likelihood that a

1400

600

1200

Interviewed Died Lost Withdrew

1000

800

400

200

01975 1978 1981 1984 1987 1990 1993 1996 1999 2002

Study Year

No. o

f Pat

ient

s

Figure 2. Of 1380 cohort members, the number of patients who were interviewed and the cumulative number of patients who died, who were unavailable forfollow-up (Lost), and who withdrew from the study each year.



second examination was reported increased signifi-cantly for each year a patient was followed up after thefirst examinations between 1985 and 2005 (OR, 1.31; 95%CI, 1.27-1.36). A total of 2378 examinations were in-cluded in this study (mean, 3 examinations per pa-tient). The mode of PGA level was mild (48.0%). Of the815 patients, 290 (35.6%) had the exact same PGA score,and 50.3% had a maximum difference of 1 level; 13.3%,of 2 levels; and 0.9%, of 3 levels (ie, had an examinationreporting clear psoriasis and an examination reportingsevere psoriasis) at examinations in which they partici-pated. The difference between the 2 broad severity cat-egories (none or mild vs moderate or severe psoriasis)changed in more than half of the participants (52.3%).Although the variance increased significantly (P� .001)with the maximum number of observations (Figure 3),the majority of patients remained within 1 level of theirbaseline assessment in 1986 (Figure 4). Of the 94 pa-tients who had severe disease in 1986, almost half still

had severe disease in 1991, about 25% remained at thislevel in 2001 and 2005, and between 65% to 85% had atleast moderate disease at subsequent examinations. About90% of 602 patients with mild and moderate psoriasisin 1986 remained within 1 PGA level at the subsequentexaminations. About 70% of patients with no or mild dis-ease in 1986 had these same levels of psoriasis severityat subsequent examinations.

PAIRED COMPARISONS

The median change in PGA score between the 4 exami-nation periods was 0 (25th percentiles of 0.0 and −1.0and 75th percentiles of 0.0 and 1.0). Paired analyses be-tween the 4 assessments showed no significant differ-ence in PGA levels of patients (P� .06). After categoriz-ing the PGA levels of patients with none to mild andmoderate to severe psoriasis, paired analyses showed nostatistical significant differences between periods(P� .33).

INTRAINDIVIDUAL VARIABILITY

Patients’ probability of having the same global psoriasisseverity 1 year later is about 80% but decreases at 5- and10-year intervals (Table2). Of patients with mild to mod-erate disease, about 85% were likely to still have the samelevel of disease after 10 years. It was estimated that al-most half the patients with moderate disease will havemild disease after 10 years. About a third of patients withsevere psoriasis are likely to have mild or moderate dis-ease 10 years later, and almost a quarter of those withno psoriasis will have moderate disease.

Table 3 provides the results of our statistical modelthat predicts the expected duration in years of the total20-year study period for individual time in each PGA levelaccording to their baseline level (the sum of the years inthe rows equals 20 years). Patients with no psoriasis in1986 were likely to have about 5 years with no diseaseand 5 years with moderate to severe disease over a sub-sequent 20-year period. Patients who had mild psoriasisin 1985 would remain in that PGA level for an averageof 12 of 20 subsequent years and be clear for 1 year. Pa-

Table 1. Characteristics of Patients With Psoriasis Who Underwent 2 or More Clinical Examinations

Characteristics

Examination (Year)

1 (1985-1986) 2 (1990-1991) 3 (2000-2001) 4 (2004-2005)

Participants, No. (%)a 865 (73.7)b 743 (74.1)b 571 (78.3)b 400 (76.0)b

Women, No. (%) 333 (38.1) 280 (37.4) 221 (38.1) 151 (37.8)Age, mean (SD), y 54.0 (13.9) 56.1 (13.0) 59.9 (12.6) 63.7 (13.5)Duration of disease, mean (SD), y 32.6 (13.6) 34.9 (12.8) 41.6 (12.2) 45.0 (13.0)Physician global assessment of psoriasis, No. (%)

Clear 40 (4.5) 29 (3.8) 29 (4.8) 38 (9.5)c

Mild 404 (45.4) 340 (44.7) 289 (47.8) 188 (47.0)Moderate 308 (34.6) 270 (35.5) 193 (31.9) 127 (31.8)Severe 121 (13.6) 109 (14.3) 63 (10.4) 47 (11.8)

aOf the patients who underwent at least 2 examinations (n = 815).bPercentage of patients who were followed up during that period.cStatistically significant (P � .02) on �² test.

100

None 1 Level 2 Level 3 Level

60

80

40

20

02

(n = 308)3

(n = 266)4

(n = 241)

No. of Clinical Examinations

Patie

nts,

%

Figure 3. Maximum level of change in physician global assessment score forthe maximum number of clinical examinations.



tients with moderate psoriasis in 1986 were expected tohave moderate or severe disease in 11 of 20 subsequentyears. Those with severe disease at the initial examina-tion would average 6 years of mild, moderate, and se-vere psoriasis and should be clear of psoriasis for onlyhalf a year over 2 decades.

EFFECT OF AGING

A comparison of patients’ age and disease duration at thetime of examination between the 4 PGA levels showedno statistical significant differences in any of the exami-nation periods (P� .18). After adjustment for sex in a mul-tivariate proportional logistic regression model, age anddisease duration were not significantly associated withlevel of PGA in 1986, 1991, 2001, or 2005 (eg, age in 2005,adjusted OR, 0.98 [95% CI, 0.96-1.02]; and duration of

disease in 2005, adjusted OR, 1.00 [95% CI, 0.98-1.03]).

SENSITIVITY ANALYSES

To study whether patients with active or persistent dis-ease were more likely to accept an invitation for a study-sponsored clinical examination, several sensitivity analy-ses were performed. The demographic and diseasecharacteristics of the patients who underwent all exami-nations during follow-up (72.2% of the study popula-tion) were comparable to those of the patients who missedintermediate examinations (data not shown). Restrict-ing the analysis to the 241 participants who underwentall 4 examinations did not change the results substan-tially (data not shown). We compared the PGA levels ofpatients who were examined only in 1986 with those of

100

60

80

40

20

01

19912

20013

2005

Clinical Examination, y

11991

22001

32005


Patie

nts,

%

100

60

80

40

20

0

11991

22001

32005


11991

22001

32005


0

Patie

nts,

%

No change Improved 1 levelImproved 2 levelsImproved 3 levels

No change Worse 1 level Worse 2 levels Worse 3 levels

100

60

80

40

20

0

Patie

nts,

%

100

60

80

40

20

0

Patie

nts,

%

A B

C D

Figure 4. The distribution of change of 4-point physician global assessment (PGA) for subsequent clinical examinations stratified by PGA level at 1986examination. A, No psoriasis (n=33). B, Mild psoriasis (n=347). C, Moderate psoriasis (n=255). D, Severe psoriasis (n=94).



patients who were assessed in both 1986 and 1991. Thesame comparison was performed for examinations in 1991and 2001. There were no significant differences in thedistribution of the PGA levels of the patients with 1 vs 2observations, except that patients who underwent only1 examination were significantly more likely to have se-vere psoriasis than were those who underwent 2 exami-nations (11.9% vs 3.1%, P� .001). When we fitted aMarkov model with all transitions possible (ie, also fromPGA levels 1 to 4), the likelihood ratio test comparingthis model with the presented model showed no signifi-cant difference (df=6, P=.41).

COMMENT

Nearly all data concerning the variability in clinical pso-riasis severity over time come from a few cross-sectionalstudies using self-reported psoriasis severity in time or fromclinical trials with follow-up measured in weeks rather thanyears or decades. The PUVA Follow-up Study cohort isunique. Patients have been followed up for nearly 30 yearsfrom enrollment in a clinical trial irrespective of their useof PUVA or any other therapy. In the first 5 years of the

study, many patients still relied on PUVA and the studyprotocol linked interviews and study-sponsored dermato-logic examinations. By the mid-1980s, use of PUVA haddecreased significantly.9 Also beginning in 1985, the tim-ing of interviews and study-sponsored dermatologic ex-aminations were no longer closely linked to each other orto patients seeking care for their psoriasis. As a result, thestructured dermatologic examinations performed from 1985to 2005 provide a unique assessment of the distribution ofpsoriasis over time in a group of patients who joined thiscohort nearly a decade earlier and of the variability in in-dividuals’ physician-assessed psoriasis severity over a 20-year period. To quantify the variation of the patients’ dis-ease severity over these 20 years, we used new statisticaltechniques to estimate transition rates and the expectednumber of years that patients would be at the different PGAlevels depending on their baseline assessment.

PSORIASIS SEVERITY IN TIME

Based on nearly 2400 structured dermatologic examina-tions, the distribution of psoriasis severity within thisgroup assessed by dermatologists changed little for thePUVA cohort as a whole over a 20-year period. Only about25% of individual study participants had more than 1 PGAlevel change (with up to 3 levels possible) in multipleexaminations. Paired comparisons demonstrated no sig-nificant differences between patients’ PGA levels over time.Moreover, transition rate analysis estimated that pa-tients had about an 80% chance to have the same levelof psoriasis severity after 1 year, but these rates de-creased after 5 and 10 years, especially for the 2 extremePGA levels. At least 50% of patients with moderate to se-vere disease are likely to still be at this level 10 years later,but patients with severe psoriasis will have mild to mod-erate disease in at least two-thirds of the follow-up years.Therefore, at least in this heterogeneous population caredfor by many dermatologists, including many with a spe-cial interest in psoriasis, the likelihood that the extentof disease will change more than 1 PGA level over 1 yearand over 10 years is relatively small, and fewer than 5%will have no psoriasis. As expected, patients who had nopsoriasis at baseline, which may reflect the natural courseof their disease or may be attributable to therapy, havethe best chance of having no or mild disease in the fu-ture. Although treatment exposure and PGA examina-tion results were not linked in this study, overall treat-ment use has been substantial in this cohort. For example,from study entry to 1984, cohort members had used twiceas many PUVA as UV-B treatments, but after 1984, UV-Bwas used 3 times as often (between 1985 and 2005, thecohort used more than 200 000 UV-B treatments). De-spite substantial therapy use over time, only 9.0% and2.3% of the 815 study participants were clear of diseaseat 1 or 2 examinations and only 5 (1.0%) of the 507 pa-tients who underwent 3 or more clinical examinationswere clear at 3 examinations, which may be a spontane-ous outcome or attributable to therapy. These low pro-portions of clearance contrast with those observed in anolder prospective study of an isolated population fol-lowed up for a median interval of 1 year as well as thoseof a study that noted that 75% of patients reported a com-

Table 2. The Transit Rate Between Levels of PhysicianGlobal Assessment of Psoriasis at ConsecutiveExaminations After 1, 5, and 10 Yearsa

Psoriasis, y None Mild Moderate Severe

None1 0.82 0.17 0.01 0.005 0.40 0.47 0.12 0.0210 0.19 0.53 0.23 0.05

Mild1 0.02 0.88 0.10 0.005 0.06 0.63 0.26 0.0510 0.07 0.53 0.30 0.09

Moderate1 0.00 0.15 0.78 0.075 0.02 0.39 0.44 0.1510 0.05 0.46 0.36 0.14

Severe1 0.00 0.02 0.19 0.795 0.01 0.15 0.78 0.1510 0.03 0.37 0.39 0.22

aMultistate Markov modeling (see “Methods” section).

Table 3. Of the Total of 20 Years, the Expected Numberof Years That Patients Will Spend in Each of the Levelsof the Physician Global Assessment (PGA) Stratifiedby PGA of Psoriasis in 1986a

PGA of Psoriasisin 1986 None Mild Moderate Severe

None 5.79 9.28 3.90 1.03Mild 1.23 11.80 5.42 1.55Moderate 0.78 8.15 8.44 2.63Severe 0.55 6.22 7.04 6.18

aMultistate Markov modeling (see “Methods” section).



plete remission of their disease and that 2% of men and8% of women had no signs of the disease for at least 5years.2-4 Both of these studies relied on patient reportsrather than on examination and included a much higherpercentage of persons with mild psoriasis than did thePUVA Follow-up Study.

EFFICACY VS EFFECTIVENESS

The study findings confirm the discrepancy betweenthe efficacy of treatments observed in clinical trials andour success in controlling psoriasis over the long term.In the original PUVA trial, 88% of the patients achievedclearance at completion of the study,12 but 1 years ormore thereafter, including at time points close to the in-troduction of cyclosporine and the biologics, 55% orfewer of those patients were described as having no ormild psoriasis. This difference of 30% may suggest thatclearing psoriasis over the short term is quite feasiblebut that it is much more difficult to substantially reducepsoriasis over the long term, which may be attributableto other issues such as care setting (clinical trials vs reallife), access and availability, safety, convenience, satis-faction, and costs.13-15 Although our study was unable todifferentiate which of these issues or what other factorsmay be the root cause, our findings indicate that thosepatients who sought a cure for their psoriasis 30 yearsago and temporarily achieved good control of their dis-ease have suffered from substantial disease over thelong term.

Psoriasis severity reported at the first 5 examinationsafter the introduction of PUVA (1976-1982), whichwere excluded from the analyses, was significantly lessthan that at later examinations. In the years in whichthe first 5 examinations were performed, the use ofPUVA was significantly greater than in the years beforethe examinations that were performed between 1985and 2005 (Figure 1). However, this level of control di-minished in association with the decline of PUVA useafter it became widely accepted to be carcinogenic inthe 1980s.16 At least in our cohort, the introduction ofnew treatments from 1985 to 1998 (ie, vitamin D de-rivatives, retinoids, and cyclosporine) did not seem tohave substantially affected the severity of the patients’psoriasis. However, the distribution of PGA levels of400 patients with psoriasis who were examined in 2005showed that a significantly larger percentage (9.6% vs�5.1%) of patients were rated clear compared with thepatients at prior examinations. Fewer than one-fourthof those who were clear in 2005 had ever used biolog-ics, but the availability of biologics and the publicityabout psoriasis treatments may have stimulated morepatients to seek care again. Therefore, these recent data,as well as our cohorts’ experience from 1977 to 1982,are compatible with the hypothesis that more intensivetherapy, both new and old, is able to reduce psoriasis, atleast for short periods.

EFFECT OF AGING

A medical claims–based study that used prevalence es-timates that were based on general physicians’ record-

ing diagnoses and assessing the distribution of severityamong different groups suggested that psoriasis may gointo remission in the elderly.6 Our logistic regressionmodel, which was based on a prospective study of a de-fined cohort, indicates that psoriasis severity does not de-crease substantially with age. Also, our study could notconfirm that the extent of psoriasis depends on the du-ration of disease (ie, age at onset) but indicates that itmay depend on family history.17-19

STRENGTHS AND LIMITATIONS

To our knowledge, this is the first prospective study thathas used standardized structured dermatologic exami-nation to assess psoriasis severity longitudinally over 2decades at 4 time points that were not substantially re-lated to treatment use. This is a long-term ecologic studyof a cohort with psoriasis. As such, it has several advan-tages and disadvantages.

Although we documented patient treatments throughnearly annual telephone interviews with members of thecohort, the timing of dermatologic examination was notcoordinated with therapy or therapeutic decisions butrather was arranged to suit the patient’s convenience.Therefore, we cannot link the use of specific treatmentswith specific disease states. Instead, we present a uniqueperspective of physician-assessed severity in a cohort ofpatients who were periodically examined over a 20-yearperiod. A true “natural history” study of a chronic dis-ease such as psoriasis would need to withhold treat-ment from the affected individuals. Such a study wouldbe both impractical and unethical, particularly in a co-hort such as ours. Enrollment in a clinical trial may beassociated with worsening of a disease, with subsequentobserved decreased severity a reflection of “regression tothe mean” rather than a true change in the “usual” dis-ease severity. Since we used examination 1, which wasadministered nearly a decade after enrollment, as our base-line, this potential ascertainment bias is unlikely to havesubstantially influenced our results. However, this phe-nomenon may in part explain the relatively large pro-portion of patients with mild disease. Although cohortmembers with persistent or active psoriasis may have beenmore likely than those with less or inactive disease to ac-cept the invitation for a clinical examination, the num-ber of patients who were unavailable for follow-up or whowithdrew from the cohort study was small. Although sev-eral sensitivity analyses also suggest that this potentialselection bias has not substantially affected our results,study patients had more severe psoriasis in 1977 than co-hort members who were not included in this study. De-spite the fact that patients were invited randomly in a fol-low-up cycle, the probability may have been greater thatthe participating patients who attended the examina-tion were more likely than nonparticipating patients tobe seeking medical care, including therapeutic and non-therapeutic interventions. However, this association isnot likely to be different among the 4 assessments, sug-gesting a nondifferential effect. The study findings arenot generalizable to all patients with psoriasis, becausethis cohort included patients who sought experimentaltherapy 30 years ago at tertiary referral centers. At the



time of patients’ enrollment in the PUVA study, nearly adecade before the first examinations included in theseanalyses were performed, the patients had sufficiently se-vere and/or treatment-resistant psoriasis that both pa-tients and their treating dermatologists at 16 centers,which were selected for their expertise in the treatmentof psoriasis, deemed them appropriate candidates for athen-experimental therapy, which included the sys-temic administration of drug. We believe that our re-sults are likely to be representative of the course of pso-riasis severity among patients with moderate to severepsoriasis who at least initially had access to care at a lead-ing dermatologic center. Although mortality may be as-sociated with clinical psoriasis severity (owing to in-creased incidence of comorbidities),20 it is not likely tobe associated with the degree of disease fluctuations, whichis the objective of this study. The Markov model was usedto create the transition tables. As a statistical model, it isbased on several assumptions. The results of these kindsof analyses should be interpreted with some caution andapplied cautiously in advising individual patients. How-ever, a sensitivity analysis with altered assumptions con-firmed the findings of the presented model. Because ofsoftware limitations, we were unable to calculate 95% CIsfor the transition rates. However, the 95% CIs were es-timated for the transition intensities (hazard matrix),which is in an intermediate step in the process of the tran-sition rate calculation, and were reasonably narrow (datanot shown). We believe that a 4-point PGA scale, whichwas assessed by dermatologists who were trained at in-vestigators’ meetings and through instructions, as partof a structured dermatologic examination, is likely to bea reliable and valid measurement in the assessment of pso-riasis. However, the PGA was not strictly defined and didnot include parameters such as percentage of body sur-face area affected and/or plaque characteristics but iswidely used and has “face” validity.21-23 Although a globalassessment may be relatively insensitive for measuringsmall changes over time, we believe it should be sensi-tive to clinically meaningful changes, which are the fo-cus of this study. In addition to the intraobserver vari-ability of the PGA, interobserver variability may also haveaffected our results, because different investigators mayhave assessed some patients over time. The extent to whichobserver bias affects the PGA in psoriasis is not well docu-mented.22

In conclusion, even with substantial use of psoriasistherapies after 1984, including more than an average of170 UV-B treatments, 50 PUVA treatments, and a yearof methotrexate therapy, our cohort was assessed bydermatologists as having moderate or severe psoriasis atnearly half of almost 2400 examinations performedover a 20-year period. Although at some times the dis-ease may improve or worsen, most individuals in ourcohort had similar extents of psoriasis at subsequent ex-aminations. Three decades after this large diverse groupof patients sought a cure for their psoriasis, they havenot achieved consistent control of their disase. Addi-tional prospective longitudinal studies that assess the(natural) course of psoriasis and its determinants arewarranted.

Accepted for Publication: February 25, 2007.Correspondence: Robert S. Stern, MD, Department of Der-matology, Beth Israel Deaconess Medical Center, 330Brookline Ave, Boston, MA 02215 ([email protected]).Author Contributions: Dr Stern is the principal inves-tigator of the PUVA Follow-up Study. Study concept anddesign: Nijsten and Stern. Acquisition of data: Stern. Analy-sis and interpretation of data: Nijsten, Looman, and Stern.Drafting of the manuscript: Nijsten. Critical revision of themanuscript for important intellectual content: Nijsten, Loo-man, and Stern. Statistical analysis: Nijsten and Loo-man. Obtained funding: Stern. Administrative, technical,and material support: Nijsten and Stern. Study supervi-sion: Stern.Financial Disclosure: None reported.Funding/Support:This study was funded in part by Pub-lic Health Service contract N01-AR-0-2246 from the Na-tional Health Institute of Arthritis and Musculoskeletaland Skin Diseases, National Institutes of Health, Depart-ment of Health and Human Services.Additional Contributions: We thank the patients andprincipal investigators of the multicenter PUVA Fol-low-up Study for their participation; M. Cosmos and S.Paquette for collecting and managing data for this project;D. Spiegelman for input on data analysis; and Chris-topher Jackson, PhD, Imperial College, London, En-gland, for providing access to Markov modeling soft-ware. The following institutions and investigators haveparticipated in the recruitment and follow-up of pa-tients in this study: Stanford University School of Medi-cine, Stanford, California (E. Bauer and E. Abel); Uni-versity of California at San Francisco (J. H. Epstein andJ. Koo); Baylor College of Medicine, Houston, Texas(J. Wolf); Washington Hospital Center, Washington, DC(T. P. Nigra); University of Michigan Medical School, AnnArbor (T. F. Anderson); Columbia University College ofPhysicians and Surgeons, New York, New York (J. Prys-towsky and M. J. Stiller); Mayo Graduate School of Medi-cine, Rochester, Minnesota (M. McEvoy); University ofMiami Medical School, Miami, Florida (J. R. Taylor); MtSinai Medical Center, Miami Beach, Florida (N. Zaias);Temple University School of Medicine, Philadelphia,Pennsylvania (F. Urbach); Beth Israel Deaconess Medi-cal Center, Boston, Massachusetts (R. S. Stern); Dart-mouth Medical School, Hanover, New Hampshire (R. D.Baughman); Yale University of Medicine, New Haven,Connecticut (I. M. Braverman); Duke University Medi-cal Center, Durham, North Carolina (J. Murray); Uni-versity of Pennsylvania Hospitals, Phildelphia (V. P.Werth); and Massachusetts General Hospital, Boston(T. B. Fitzpatrick [deceased], J. Parrish, C. Taylor, andA. J. Sober).

REFERENCES

1. Stern RS, Nijsten T, Feldman SR, Margolis DJ, Rolstad T. Psoriasis is common,carries a substantial burden even when not extensive, and is associated with wide-spread treatment dissatisfaction. J Investig Dermatol Symp Proc. 2004;9(2):136-139.

2. Farber EM, Bright RD, Nall ML. Psoriasis: a questionnaire survey of 2,144 patients.Arch Dermatol. 1968;98(3):248-259.



3. Farber EM, Nall ML. The natural history of psoriasis in 5,600 patients. Dermatologica.1974;148(1):1-18.

4. Molin L. Psoriasis: a study of the course and degree of severity, joint involve-ment, socio-medical conditions, general morbidity and influences of selectionfactors among previously hospitalized psoriatics. Acta Derm Venereol Suppl(Stockh). 1973;53:1-125.

5. Spuls PI, Witkamp L, Bossuyt PM, Bos JD. The course of chronic plaque-typepsoriasis in placebo groups of randomized controlled studies. Arch Dermatol.2004;140(3):338-344.

6. Gelfand JM, Weinstein R, Porter SB, Neimann AL, Berlin JA, Margolis DJ. Preva-lence and treatment of psoriasis in the United Kingdom: a population-based study.Arch Dermatol. 2005;141(12):1537-1541.

7. Stern RS, Laird N, Melski J, Parrish JA, Fitzpatrick TB, Bleich HL. Cutaneous squa-mous-cell carcinoma in patients treated with PUVA. N Engl J Med. 1984;310(18):1156-1161.

8. Stern RS, Thibodeau LA, Kleinerman RA, Parrish JA, Fitzpatrick TB. Risk of cu-taneous carcinoma in patients treated with oral methoxsalen photochemo-therapy for psoriasis. N Engl J Med. 1979;300(15):809-8138.


10. Gladman DD, Farewell VT. Progression in psoriatic arthritis: role of time-varyingclinical indicators. J Rheumatol. 1999;26(11):2409-2413.

11. The R project for statistical computing. http://www.r-project.org/. Accessed Sep-tember 10, 2006.

12. Melski JW, Tanenbaum L, Parrish JA, Fitzpatrick TB, Bleich HL. Oral methox-salen photochemotherapy for the treatment of psoriasis: a cooperative clinicaltrial. J Invest Dermatol. 1977;68(6):328-335.

13. Krueger G, Koo J, Lebwohl M, Menter A, Stern RS, Rolstad T. The impact of pso-riasis on quality of life: results of a 1998 National Psoriasis Foundation patient-membership survey. Arch Dermatol. 2001;137(3):280-284.

14. Zaghloul SS, Goodfield MJ. Objective assessment of compliance with psoriasistreatment. Arch Dermatol. 2004;140(4):408-414.

15. Nijsten T, Margolis DJ, Feldman SR, Rolstad T, Stern RS. Traditional systemictreatments have not fully met the needs of psoriasis patients: results from a na-tional survey. J Am Acad Dermatol. 2005;52(3, pt 1):434-444.

16. Housman TS, Rohrback JM, Fleischer AB Jr, Feldman SR. Phototherapy utiliza-tion for psoriasis is declining in the United States. J Am Acad Dermatol. 2002;46(4):557-559.

17. Melski JW, Stern RS. The separation of susceptibility to psoriasis from age atonset. J Invest Dermatol. 1981;77(6):474-477.

18. Gudjonsson JE, Karason A, Runarsdottir EH, et al. Distinct clinical differencesbetween HLA-Cw*0602 positive and negative psoriasis patients—an analysis of1019 HLA-C– and HLA-B–typed patients. J Invest Dermatol. 2006;126(4):740-745.

19. Henseler T, Christophers E. Psoriasis of early and late onset: characterization oftwo types of psoriasis vulgaris. J Am Acad Dermatol. 1985;13(3):450-456.

20. Gelfand JM, Neimann AL, Shin DB, Wang X, Margolis DJ, Troxel AB. Risk of myo-cardial infarction in patients with psoriasis. JAMA. 2006;296(14):1735-1741.

21. Naldi L, Svensson A, Diepgen T, et al. Randomized clinical trials for psoriasis1977-2000: the EDEN survey. J Invest Dermatol. 2003;120(5):738-741.

22. Ashcroft DM, Wan Po AL, Williams HC, Griffiths CE. Clinical measures of dis-ease severity and outcome in psoriasis: a critical appraisal of their quality. Br JDermatol. 1999;141(2):185-191.

23. Feldman SR, Krueger GG. Psoriasis assessment tools in clinical trials. Ann RheumDis. 2005;64:ii65-ii68.

ARCHIVES Feature

Free color publication if color illustrations enhance thedidactic value of the article.



STUDY

Clinical, Biochemical, and Genetic Studyof 11 Patients With Erythropoietic ProtoporphyriaIncluding One With Homozygous DiseaseCarmen Herrero, MD, PhD; Jordi To-Figueras, PhD; Celia Badenas, PhD; Manuel Mendez, PhD;Patricia Serrano, MD; Rafael Enrıquez-Salamanca, MD, PhD; Mario Lecha, MD, PhD

Objective: To study the mutations in the ferroche-latase gene (FECH) and the phenotypic expression oferythropoietic protoporphyria (EPP) in a group of Span-ish patients.

Design: Case series.

Setting: University-based hospital.

Patients: Eleven unrelated patients with EPP and 19asymptomatic relatives from 10 families.

Main Outcomes Measures: Measurement of proto-porphyrin concentration in red blood cells and feces byfluorometry and chromatography. Analysis of the mu-tations of the FECH gene by single-strand conformationanalysis. Expression of mutations in Escherichia coli.

Results: FECH gene mutations were found in all 11 pa-tients. Ten were heterozygous and carried the IVS3-48Clow-expression allele. Three novel mutations were found:IVS4�1delG, 347-351delC, and 130_147dupl 18. One

patient did not present the IVS3-48C polymorphism andwas found to harbor a novel A185T missense mutationin both alleles. The familial study confirmed a recessivemode of inheritance of the disease. The A185T muta-tion showed a residual activity 4% of normal when ex-pressed in E coli. This patient presented cutaneous pho-tosensitivity similar to the heterozygous cases, but a higherprotoporphyrin accumulation in erythrocytes, micro-cytic anemia, and early signs of liver engagement. FECHmutations were found in 10 healthy relatives, none ofwhom carried the low-expression allele. The frequencyof the IVS3-48C allele among 180 nonporphyric Span-ish individuals was 5.2%.

Conclusions: These findings confirm, among a groupof Spanish patients, that most cases of EPP result fromthe coinheritance of IVS3-48C and a mutation in the FECHgene, and also document the existence of patients withmutations in homozygosity that may present a more se-vere form of the disease.

Arch Dermatol. 2007;143(9):1125-1129

E RYTHROPOIETIC PROTOPOR-phyria (EPP) (Mendelian In-heritance in Man 177000) isan inherited disorder of por-phyrin metabolism caused by

a partial deficiency of ferrochelatase (FECH)(European Community 4.99.1.1), whichcatalyzes the insertion of iron into proto-porphyrin IX to form heme. Patients withEPP accumulate free protoporphyrin mainlyin erythropoietic tissues, but the metabo-lite may also reach the skin, absorb light atcertain wavelengths, and generate free radi-cals, resulting in phototoxic cell injury.Most patients with EPP develop solely pho-tosensitivity commencing in childhood, butin some cases liver complications also ap-pear. The strong hydrophobic nature of themolecule does not allow renal clearance butonly bile secretion. Therefore, massive he-patic accumulation of protoporphyrin in theliver may cause cholestasis and liver dam-

age in a minority of patients with EPP. He-patic involvement may vary in severity,ranging from abnormalities in 20% of thecases to cirrhosis and liver failure in lessthan 5%.1-3 Blood disorders such as mildanemia with hypochromia and microcyto-sis may also be detected in 20% to 50% ofthese patients.2,4,5

The FECH gene is assigned to chromo-some 18q21.3 (GenBank No. D00726),with more than 110 different mutationshaving been described in EPP families todate.6 Individuals who are heterozygousfor these mutations remain asymptom-atic, despite having FECH activity re-duced by about 50%. Clinical expressionof the disease in most cases results from

See also pages 1153, 1194,and 1209

Author Affiliations:Departments of Dermatology(Drs Herrero, Serrano, andLecha) and Biochemistry andMolecular Genetics(Drs To-Figueras and Badenas),Hospital Clinic of Barcelona,University of Barcelona,IDIBAPS (Institutd’Investigacions BiomèdiquesAugust Pi i Sunyer), Barcelona,Spain; and Centro deInvestigacion, Hospital 12 deOctubre, Madrid, Spain(Drs Mendez andEnrıquez-Salamanca).



the coinheritance of a null FECH allele plus a wild-typelow-expressed allele, which further reduces FECHactivity to below a critical threshold of about 35%.7 In ahaplotype segregation analysis,8 it was found that the low-expression allele is the C variant of an intronic single-nucleotide polymorphism (SNP) IVS3-48T/C in intron3 of the FECH gene. Because the IVS3-48C allele is rela-tively common among European populations,6 overtEPP is normally inherited in such a way that it re-sembles an autosomal dominant disease with low clini-cal penetrance.

Erythropoietic protoporphyria is rarely inherited asan autosomal recessive disease with true mutations in bothalleles.9-12

In some areas of southern Europe, the prevalence ofEPP and the genetic defects underlying the disease arenot sufficiently elucidated. Therefore, we investigated themutations in the FECH gene and the phenotypic expres-sion of the disease in a series of Spanish patients withEPP. The molecular, biochemical, and clinical study in-cluded a rare case of recessive EPP.

METHODS

PATIENTS

We studied 11 unrelated patients (6 female and 5 male, with amean age of 12.8 years [range, 4-26 years]) who attended theDepartment of Dermatology of the Hospital Clinic of Barce-lona (Barcelona, Spain) for photosensitivity and suspicion ofporphyria. The diagnosis of EPP was established initially by clini-cal and biochemical criteria and was later confirmed by DNAanalysis of mutations in the FECH gene. A complete hemato-logical and biochemical profile and specific liver function testswere performed in every case. In addition, 19 asymptomaticrelatives from 8 families were also included in the genetic study.The study was conducted in accordance with the Declarationof Helsinki principles and was approved by the hospital ethicscommittee of the Hospital Clinic of Barcelona. Written con-sent was obtained from all patients.

PORPHYRIN DETERMINATION

Protoporphyrin concentration in erythrocytes was deter-mined fluorometrically after solvent partition according to themethod described elsewhere.13-15 Protoporphyrin in feces wasextracted according to Lockwood et al16 and was analyzed byhigh-performance liquid chromatography according to Lim andPeters.17 Porphyrins in urine were analyzed by standard meth-ods of high-performance liquid chromatography.17

LYMPHOCYTE FECH ENZYME ASSAY

FECH activity in lymphocytes was determined by incubating ahomogenate with zinc and mesoporphyrin and measuring zinc-mesoporphyrin formation by fluorometry according essentiallyto the method of Rossi et al,18 as modified by Gouya el al.6 FECHactivity was expressed as nanomoles of zinc-mesoporphyrinformed per hour per milligram of protein at 37°C.

DNA ANALYSIS

We analyzed exons 1 to 11 of the FECH gene and the associ-ated splice donor and acceptor sites. The primers used for exon

amplification are available from the authors. All polymerase chainreaction (PCR) analyses were carried out using the PCR Mas-ter Mix (Promega Corporation, Madison, Wisconsin) follow-ing the manufacturer’s instructions. The following PCR con-ditions were applied: initial denaturizing step (5 minutes at95°C), 35 cycles (1 minute at 95°C, 1 minute at 57°C, and 1minute at 72°C), followed by final extension (10 minutes at72°C) and maintenance at 4°C until single-strand conforma-tional polymorphism or sequencing studies were carried out.Part of intron 3 was sequenced in all samples to test for the IVS3-48T/C polymorphism.

Single-strand conformation analysis was used to detect mu-tations. Three microliters of denatured PCR product was com-bined with loading buffer and loaded into GeneGel Excel 12.5acrylamide gels (Amersham Biosciences, Uppsala, Sweden) andrun following the manufacturer’s instructions. The different mi-grations were silver stained.19 DNA samples with abnormal mi-grations were sequenced. The PCR products were purified usingthe GFX PCR DNA and gel band purification kit (AmershamBiosciences) and automatically sequenced using the BigDye Ter-minator v3.1 Cycle Sequencing Kit (Applied Biosystems, Fos-ter City, California) and an ABI3100 automatic sequencer (Ap-plied Biosystems).

PROKARYOTIC EXPRESSIONOF MUTANT FECH

Mature ferrochelatase was expressed in Escherichia coli strainJM109 (Promega Corporation) using the expression vectorpKK223-3 (Pharmacia Biotech Inc, Piscataway, New Jersey).The wild-type human ferrochelatase complementary DNA(cDNA) was synthesized by PCR excluding the region encod-ing the mitochondrial leader sequence and including a new ATGstart site of translation, and cloned into the EcoRI-HindIII sitesof pKK223-3. This construct was designated pKK-FECH-wt.To generate the mutant construct pKK-FECH-A185T, a frag-ment of FECH cDNA, containing the A185T mutation wasgenerated by PCR site-directed mutagenes20 using the pKK-FECH-wt plasmid as templated. Then, the PCR product wasdigested with the restriction endonucleases MfeI and CveI (NewEngland Biolabs, Beverly, Massachusetts) and the MfeI-CveI frag-ment was ligated as a cassette into the corresponding sites in pKK-FECH-wt, and transformed in E coli strain JM109. The integrityof each expression construct was checked by automated sequenc-ing. The primers are available from the authors.

Bacterial clones, each of them containing one of the plasmidspKK223-3, pKK-FECH-wt, or pKK-FECH-A185T, were grownto log phase and were induced with 5mM isopropylthiogalacto-side for 3 hours. Then, cells were harvested by centrifugation andwashed twice in phosphate-buffered saline. The pellets were re-suspended in 250 µL of lysis buffer (100mM Tris hydrochloride,pH 7.6; 0.1% Triton-X 100) and sonicated on ice. After centrifu-gation, ferrochelatase activity was measured in the supernatantusing protoporphyrin IX and zinc sulfate as substrates, accord-ing to the method described by Camadro and Labbe.21

RESULTS

CLINICAL CHARACTERISTICS

The main clinical and biochemical outcomes of the pa-tients with EPP are given in the Table. All presented acutecutaneous photosensitivity, with the symptoms begin-ning in early childhood in most cases. Only 3 cases pre-sented with chronic skin thickening over the knuckles.Microcytic anemia was present in 4 cases. Liver func-



tion test results were persistently normal in 10 cases, mostwith a long clinical follow-up (mean, 8.6 years; range 1-26years). Only 1 case showed alterations in the liver func-tion test results. The main characteristics of this patientare described in the “Report of a Case” subsection.

MUTATIONS IN THE FECH GENE

Ten different mutations in the FECH gene were found,thereby confirming the molecular heterogeneity of EPP(Table).Fournovelmutationswere found:553G�Ainexon5 (A185T); IVS4�1delG and 347-351delC, both in exon4; and 130_147dupl 18 in exon 1. The A185T mutationwas homozygous, while the other mutations were hetero-zygous. Only the already described 575del5 mutation waspresent in 2 patients. Patients 7, 8, 9, and 10 (I186T, H386P,843delC, and R115X, respectively) were previously stud-ied by Gouya et al,22 but the status of the SNP of the alleletrans to mutant allele was not determined in that study. Allpatients, except patient 1 who was homozygous for IVS3-48T/T, were heterozygous for IVS3-48C/T.

PROKARYOTIC EXPRESSIONOF GENE MUTATION

The A185T mutation was expressed in E coli. FECH spe-cific activity was determined in 4 independent experi-ments. The residual activity was calculated from the valueof the normal FECH construct (pKK-FECH-wt) and was4% of normal activity.

FAMILIAL AND POPULATION ANALYSES

Molecular analysis of 19 asymptomatic relatives from pa-tients 1 through 6 and 8 through 10 were performed. Thisallowed the identification of 10 relatives who were car-riers of heterozygous mutations. Inspection of the in-tron 3 SNP showed that none harbored the IVS3-48C al-lele. All of these relatives had normal values ofprotoporphyrin in blood. Analysis of 180 nonporphyric

unrelated Spanish individuals showed that the IVS3-48C allele was present in 10.5% (allele frequency, 5.2 %).

REPORT OF A CASE

Patient 1 was a 15-year-old girl with microcytic anemia sinceshe was 4 months old and photosensitive dermatitis sinceshe was 4 years old. When she was 10 years old, alter-ations in liver function had been detected, with the per-sistence of microcytic anemia. She presented an itchy ery-thema and edema on the hands and face after sun exposure.On examination, pitted and linear scars on the nose, cheeks,and forehead and mild skin thickening over the knuckleswere observed. There was no family history of photosen-sitivity and no apparent familial consanguinity.

Results from porphyrin analyses included the follow-ing: erythrocyte protoporphyrin level, 32 µmol/L (refer-ence value, �1.5 µmol/L; �90% free protoporphyrin);and fecal protoporphyrin level, 1066 nmol/g of dried fe-ces (reference value, �100 nmol/g), which allowed forthe diagnosis of EPP. The excretion of coproporphyrinsin urine was normal. Results from biochemical analysisincluded the following: aspartate transaminase level, 52U/L (reference value, �40 U/L), alanine aminotransfer-ase level, 109 U/L (reference value, �40 U/L), and �-glu-tamyltransferase level, 57 U/L (reference value, �40 U/L)(to convert to microkatals per liter, multiply by 0.0167);hemoglobin level, 11.5 g/dL (reference range, 12.0-17.0g/dL) (to convert to grams per liter, multiply by 10); me-dian corpuscular volume, 73 fL (reference range, 80-100 fL); median corpuscular hemoglobin concentra-tion, 22.7 pg (26.7-33.3 pg); serum iron level, 59 µg/dL(reference range, 50-150 µg/dL) (to convert to micro-moles per liter, multiply by 0.179); ferritin level, 28 ng/mL(reference range, 15-200 ng/mL) (to convert to pico-moles per liter, multiply by 2.247); and soluble trans-ferrin receptor level, −0.089 mg/dL (reference range,0.083-0.176 mg/dL) (to convert to micromoles per liter,multiply by 0.0123). The appearance of the liver on echog-raphy was normal. A hepatic biopsy revealed normal liver

Table. Major Clinical and Biochemical Features and FECH Genotypes in 11 Patients With EPP

Patient No./Sex/Age atFirst Symptoms, y

Hemoglobin,12-17 g/dL

MCV,80-100 fL

LiverDysfunction

Erythrocyte PP�1.5 µmol/L

Fecal PP�100 nmol/gof Dried Feces

FECHMutation

IVS3-48C/TSNP

1/F/4 11.5 73 Yesa 32 1066 A185T/A185T T/T2/F/2 11.1 75 No 12 ND 757del5 C/T3/M/1 13.7 80 No 11 90 IVS4 � 1delG C/T4/M/1 13.4 79 No 10 715 347-351delC C/T5/F/6 12.0 87 No 13 114 757del5 C/T6/M/3 12.6 75 No 18 327 130_147dupl18 C/T7/F/8 11.8 79 No 17 115 I186T C/C8/M/2 12.3 83 No 11 ND H386P C/T9/M/2 12.5 81 No 5 ND 843delC C/T

10/F/1 11.7 76 No 10 ND R115X C/T11/F/6 12.4 79 No 7 186 P334L C/T

Abbreviations: EPP, erythropoietic protoporphyria; MCV, medium corpuscular volume; ND, not done; PP, protoporphyrin; SNP, single-nucleotide polymorphism.SI conversion factor: To convert the hemoglobin to grams per liter, multiply by 10.aAspartate aminotransferase, alanine aminotransferase, and �-glutamyltransferase levels greater than 40 U/L (to convert to microkatals per liter, multiply by

0.0167).



architecture and the presence of brown material with-out birefringence at polarized light. Hemoglobin elec-trophoresis demonstrated normal percentages of hemo-globin A2 and F, and findings from DNA analysis excludedthe thalassemia trait. Bone marrow examination showednormal percentages of the hematopoietic cell lineages withdysplasia signs in the erythroid cells. The bone marrowkaryotype was 46,XX[20].

Findings from a molecular study of the DNA showedthat this patient was apparently homozygous for a novelA185T missense mutation and homozygous for the T vari-ant of the IVS3-48C/T polymorphism. The residual activ-ity of the mutant FECH in E coli was 4% of normal activ-ity. The FECH activity in peripheral lymphocytes was 0.4nmol of zinc-mesoporphyrin per hour per milligram of pro-tein at 37°C (reference value, mean±SD, 4.5±1). Two yearslater, the erythrocyte protoporphyrin level had increasedto 77 µmol/L (fecal protoporphyrin level, 1221 nmol/g ofdried feces).

Familial investigation showed that both parents anda sister were heterozygous for the A185T mutation, andnone harbored the low-expression IVS3-48C allele, thusconfirming recessive EPP. All were asymptomatic and hadan erythrocyte protoporphyrin concentration of less than1.5 µmol/L.

COMMENT

We present the clinical, biochemical, and molecular char-acteristics of 11 cases of EPP and 19 asymptomatic rela-tives from 10 Spanish families. Among the series of pa-tients with EPP, 1 had homozygous disease. Recessiveinheritance of EPP, with patients presenting mutationsin both alleles is rare. Only 13 cases (12 families) havebeen reported to date,6,9-12 most being compound het-erozygotes. To our knowledge, the patient who was ho-mozygous for an A185T mutation is the second case oftrue homozygous EPP, with the same mutation in bothalleles confirmed by molecular analysis.12 The A185T mu-tation is located within the �3-strand of the enzyme andmay produce a significant structural alteration owing tothe replacement of a hydrophobic amino acid by a polaramino acid. Although the alanine residue at this posi-tion is not invariant among different species, it is con-served in mammals, yeast, and some prokaryotic spe-cies.23,24 Expression of the A185T allele revealed a lowbut still detectable residual activity of 4% in a crude bac-terial extract. These data indicate that the mutation pro-duces important structural and/or functional conse-quences in the FECH enzyme but is not completelydeleterious.

It has been suggested that recessive EPP forms mayhave a higher risk of liver failure than the more usualdominant forms.10 According to published reports, se-vere liver disease has been observed in 4 of 13 patientswith recessive EPP detected to date.6,10,12 This is a highproportion, taking into account that liver failure ap-pears in only in less than 5% of patients with EPP domi-nant forms.6,25

The determinants of liver injury in EPP are not wellunderstood because genetic factors may modulate the ac-

cumulation of protoporphyrin in the liver.26 It is likely,however, that the amount of residual FECH activity incells12,25 may play a key role in determining the courseof the disease. Most of the missense mutations found inheterozygous EPP cases show a drastic reduction or evenundetectable activity when expressed in prokary-otes.27-29 In contrast, in a report of recessive EPP, all themutations retained a FECH residual activity ranging from12% to 50% when expressed in E coli.12 This is in accor-dance with the general view that homozygous cases ofporphyria must retain a substantial residual enzyme ac-tivity to be compatible with life. The A185T mutation,however, showed only 4% of the normal activity in E coli,thus suggesting a severe impairment in functionality. Con-sequently, even if this homozygous patient is free of cho-lestasis signs and the liver markers remain stable to date,the progression of the disease should be considered ofrisk and advocates close clinical follow-up.

Except for the recessive case, all the other patients inour study showed mutations in heterozygosity within theFECH gene. Some of the new mutations can also causesevere impairment in enzyme activity. The c.351delC mu-tation is a single C deletion from a stretch of 5 C thatcauses a frameshift and introduces a premature termi-nation codon of translation in exon 4. As a conse-quence, we can predict the rapid degradation of the mu-tant transcript due to nonsense-mediated messenger RNAdecay.30,31 The c.130_147dupl 18 mutation does not re-sult in a frameshift, but its location within the mitochon-drial leader sequence could result in a defective target-ing to mitochondria or a deficient processing of thepolypeptide. The novel splicing defect IVS4�1delG af-fects the consensus donor splice site of intron 4.32 Un-fortunately, RNA from this patient was not available foranalysis. Nonetheless, this mutation probably results inexon 4 skipping, as was observed in a previously re-ported mutation (IVS4�1G�C) in this donor site.27

All heterozygous cases presented the low-expressionIVS3-48C allele, thereby confirming the common viewthat clinically overt EPP is associated with inheritanceof this allele in trans to a severe FECH mutation.6,12,33,34

The study of the relatives yielded additional confirma-tion of the importance of this polymorphism in EPP, sinceall asymptomatic carriers had the IVS3-48T/T geno-type. Familial investigation of patient 1 showed that sev-eral relatives harbored the A185T mutation, but none hadthe IVS3-48C allele. Therefore, the phenotype of a hy-pothetical carrier of the A185T mutation with the low-expression allele in trans remains unknown.

In conclusion, among our series of Spanish patients withEPP, we have found 10 cases with coinheritance of 1 FECHgene mutation with IVS3-48C and 1 rare case with a se-vere homozygous mutation with very low residual activ-ity. This patient presented a cutaneous photosensitivitysimilar to the heterozygous cases but had a higher proto-porphyrin accumulation, pronounced microcytic ane-mia, and early signs of liver involvement. These findingsemphasize the need for further research on the genotypic-phenotypic relationships in porphyria.

Accepted for Publication: January 26, 2007.Correspondence: Jordi To-Figueras, PhD, Department



of Biochemistry and Molecular Genetics, Villarroel 170,Hospital Clinic Barcelona, 08036 Barcelona, Spain ([email protected]).Author Contributions: Drs To-Figueras and Carmen Her-rero had full access to all the data and take full respon-sibility for the integrity of the data and the accuracy ofthe data analysis. Study concept and design: Herrero, To-Figueras, Mendez, Enrıquez-Salamanca, and Lecha. Ac-quisition of data: Herrero, To-Figueras, Badenas, Men-dez, and Serrano. Analysis and interpretation of data:Herrero, To-Figueras, and Mendez. Drafting of the manu-script: Herrero, To-Figueras, Mendez, and Serrano. Criti-cal revision of the manuscript for important intellectual con-tent: Badenas, Enrıquez-Salamanca, and Lecha. Obtainedfunding: Herrero and To-Figueras. Administrative, tech-nical, and material support: Badenas. Study supervision:Mendez, Enrıquez-Salamanca, and Lecha.Financial Disclosure: None reported.Funding/Support: This study was supported in part bygrants from “Fondo de Investigacion Sanitaria” (03/0489) and “Fundacion Mutua Madrilena.” Both are non-profit organizations that supported the University-based research as a contribution to health progress.Role of the Sponsors: The sponsors did not participatein the conduction of the study; collection, management,analysis, interpretation of the data; or preparation, re-view, or approval of the manuscript.Additional Contributions: We appreciate the participa-tion of the family members in this study. M. Bruguera,MD, provided help in the control of hepatic disease, J.M, Mascaro, MD, collaborated in the recruitment of pa-tients, and M. Sala provided technical assistance.

REFERENCES

1. Nordmann Y. Erythropoietic protoporphyria and hepatic complications. J Hepatol.1992;16(1-2):4-6.

2. Todd DJ. Erythropoietic protoporphyria. Br J Dermatol. 1994;131(6):751-766.3. Meerman L. Erythropoietic protoporphyria: an overview with emphasis on the

liver. Scand J Gastroenterol Suppl. 2000;(232):79-85.4. DeLeo VA, Poh-Fitzpatrick M, Mathews-Roth M, Harber LC. Erythropoietic pro-

toporphyria: 10 years experience. Am J Med. 1976;60(1):8-22.5. Thunell S, Harper P, Brun A. Porphyrins, porphyrin metabolism and porphyrias,

IV: pathophysiology of erythyropoietic protoporphyria—diagnosis, care and moni-toring of the patient. Scand J Clin Lab Invest. 2000;60(7):581-604.

6. Gouya L, Martin-Schmitt C, Robreau AM, et al. Contribution of a common single-nucleotide polymorphism to the genetic predisposition for erythropoieticprotoporphyria. Am J Hum Genet. 2006;78(1):2-14.

7. Gouya L, Puy H, Lamoril J, et al. Inheritance in erythropoietic protoporphyria: acommon wild-type ferrochelatase allelic variant with low expresion accounts forclinical manifestation. Blood. 1999;93(6):2105-2110.

8. Gouya L, Puy H, Robreau AM, et al. The penetrance of dominant erythropoieticprotoporphyria is modulated by expression of wildtype FECH. Nat Genet. 2002;30(1):27-28.

9. Lamoril J, Boulechfar S, Verneuil H, Grandchamp B, Nordmann Y, Deybach J-Ch.Human erythropoietic protoporphyria: two point mutations in the ferrochelatasegene. Biochem Biophys Res Commun. 1991;181(2):594-599.

10. Sarkany RP, Alexander GJ, Cox TM. Recessive inheritance of erythropoietic pro-toporphyria with liver failure. Lancet. 1994;343(8910):1394-1396.

11. Poh-Fitzpatrick MB, Wang X, Anderson KE, Bloomer JR, Bolwell B, Lichtin AE.Erythropoietic protoporphyria: altered phenotype after bone marrow transplan-tation for myelogenous leukemia in a patient heteroallelic for ferrochelatase genemutations. J Am Acad Dermatol. 2002;46(6):861-866.

12. Whatley SD, Mason NG, Khan M, et al. Autosomal recessive erythropoietic pro-toporphyria in the United Kingdom: prevalence and relationship to liver disease.J Med Genet. 2004;41(8):e105.

13. Piomelli S. Free erythrocyte porphyrin in the detection of undue absorption oflead and iron deficiency. Clin Chem. 1977;23(2, pt 1):264-269.

14. Blake D, Poulos V, Rossi R. Diagnosis of porphyria—recommended methodsfor peripheral laboratories. Clin Biochem Rev. 1992;13(suppl 1):S1-S24.

15. Deacon AC, Elder GH. ACP Best Practice No 165: front line tests for the inves-tigation of suspected porphyria. J Clin Pathol. 2001;54(7):500-507.

16. Lockwood WH, Poulos V, Rossi E, Curnow DH. Rapid procedure for fecal por-phyrin assay. Clin Chem. 1985;31(7):1163-1167.

17. Lim CK, Peters TJ. Urine and faecal porphyrin profiles by reversed-phase high-performance liquid chromatography in the porphyrias Clin Chim Acta. 1984;139(1):55-63.

18. Rossi E, Costin KA, Garcia-Webb P. Ferrochelatase activity in human lympho-cytes, as quantified by a new high-performance liquid-chromatographic method.Clin Chem. 1988;34(12):2481-2485.

19. Badenas C, Torra R, Lucero L, Milà M, Estivill X, Darnell A. Mutational analysiswithin the 3� region of the PKD1 gene. Kidney Int. 1999;55(4):1225-1233.

20. Cormack B. Mutagenesis by the polymerase chain reaction. In: Ausubel FM, BrentR, Kingston RE, et al, eds. Current Protocols in Molecular Biology. New York,NY: Wiley & Sons, Inc; 1991:8.5.1-8.5.9.

21. Camadro JM, Labbe P. Purification and properties of ferrochelatase from yeastSaccharomyces cerevisiae: evidence for a precursor from of the protein. J BiolChem. 1988;263(24):11675-11682.

22. Gouya L, Schneider-Yin X, Rufenacht U, et al. Mutations in the ferrochelatasegene of four Spanish patients with erythropoietic protoporphyria. J Invest Dermatol.1998;111(3):406-409.

23. Dailey HA, Sellers VM, Dailey TA. Mammalian ferrochelatase: expression and char-acterization of normal and two human protoporphyric ferrochelatases. J Biol Chem.1994;269(1):390-395.

24. Dailey TA, Dailey HA. Identification of [2Fe-2S] clusters in microbial ferrochelatases.J Bacteriol. 2002;184(9):2460-2464.

25. Chen FP, Risheg H, Liu Y, Bloomer J. Ferrochelatase gene mutations in eryth-ropoietic protoporphyria: focus on liver disease. Cell Mol Biol (Noisy-le-grand).2002;48(1):83-89.

26. Abitbol M, Bernex F, Puy H, et al. A mouse model provides evidence that geneticbackground modulates anemia and liver injury in erythropoietic protoporphyria.Am J Physiol Gastrointest Liver Physiol. 2005;288(6):G1208-G1216.

27. Rufenacht UB, Gouya L, Scheneider-Yin X, et al. Systematic analysis of molecu-lar defects in the ferrochelatase gene from patients with erythropoieticprotoporphyria. Am J Hum Genet. 1998;62(6):1341-1352.

28. Schneider-Yin X, Gouya L, Dorsey M, Rufenacht U, Deybach J-C, Ferreira GC.Mutations in the iron-sulfur cluster ligands of the human ferrochelatase genelead to erythropoietic protoporphyria. Blood. 2000;96(4):1545-1549.

29. Rufenacht UB, Gregor A, Gouya L, Tarcznska-Nosal S, Schneider-Yin X, Dey-bach J-C. New missense mutation in the human ferrochelatase gene in a familywith erythropoietic protoporphyria: functional studies and correlation of geno-type and phenotype. Clin Chem. 2001;47(6):1112-1113.

30. Maquat LE. Defects in RNA splicing and the consequence of shortened transla-tional reading frames. Am J Hum Genet. 1996;59(2):279-286.

31. Hentze MW, Kulozik AE. A perfect message: RNA surveillance and nonsense-mediated decay. Cell. 1999;96(3):307-310.

32. Shapiro MB, Senepathy P. RNA splice junctions of different classes of eukary-otes: sequence statistics and functional implications in gene expression. NucleicAcids Res. 1987;15(17):7155-7174.

33. Risheg H, Fu-Ping Ch, Bloomer JR. Genotypic determinants of phenotype in NorthAmerican patients with erythropoietic protoporphyria. Mol Genet Metab. 2003;80(1-2):196-206.

34. Wiman A, Flodeurs Y, Harper P. Novel mutations and phenotypic effect of thesplice site modulator IVS3-48C in nine Swedish families with erythropoieticprotoporphyria. J Hum Genet. 2003;48(2):70-76.



STUDY

Five-Year Follow-up of a Randomized, ProspectiveTrial of Topical Methyl Aminolevulinate PhotodynamicTherapy vs Surgery for Nodular Basal Cell CarcinomaLesley E. Rhodes, MD, FRCP; Menno A. de Rie, MD; Ragna Leifsdottir, MD; Raymond C. Yu, MD, FRCP;Ingeborg Bachmann, MD; Victoria Goulden, MD, FRCP; Gavin A. E. Wong, MRCP;Marie-Aleth Richard, MD; Alex Anstey, MD, FRCP; Peter Wolf, MD

Objective: To compare 5-year lesion recurrence ratesin primary nodular basal cell carcinoma treated with topi-cal methyl aminolevulinate photodynamic therapy (PDT)or simple excision surgery.

Design: Prospective, randomized, multicenter study.

Setting: University hospital dermatology departments.

Patients: A total of 97 patients, 50 with 53 lesions treatedwith methyl aminolevulinate PDT and 47 with 52 lesionstreated by excision surgery, were included in the per pro-tocolanalysis.Ofthelesionstreatedwithmethylaminolevu-linate PDT and surgery, 49 and 52, respectively, showedcomplete clinical responseat3monthsafter treatmentandwere observed for long-term outcome evaluation.

Interventions: Topical methyl aminolevulinate cream,160 mg/g, applied for 3 hours before illumination (75J/cm2 of red light at 570 to 670 nm) on 2 or 4 occasions(12 [23%] of 53 lesions); or excision surgery.

Main Outcome Measures: Histologically confirmedlesion recurrence, sustained lesion complete response rate

(time-to-event analysis), and investigator assessment ofcosmetic outcome, 5 years after the last treatment.

Results: At 5 years, recurrence was documented in 7(14%) of 49 lesions (95% confidence interval [CI], 6%-27%) treated with methyl aminolevulinate PDT vs 2 (4%)of 52 lesions (95% CI, 1%-13%) treated with excision sur-gery (P=.09). Estimated sustained lesion complete re-sponse rates were 76% (95% CI, 59%-87%) and 96% (95%CI, 84%-99%), respectively (P=.01). More patients treatedwith methyl aminolevulinate PDT than surgery had anexcellent or good cosmetic outcome: 27 (87%) of 31 pa-tients (95% CI, 70%-96%) vs 19 (54%) of 35 patients (95%CI, 37%-71%) (P=.007).

Conclusions: Long-term follow-up indicates superiorefficacy of surgery to methyl aminolevulinate PDT innodular basal cell carcinoma. However, methyl ami-nolevulinate PDT is also an effective treatment for thisindication and exhibits a more favorable cosmeticoutcome.

Arch Dermatol. 2007;143(9):1131-1136

T OPICAL PHOTODYNAMIC

therapy (PDT) is an attrac-tive modality for the treat-ment of basal cell carci-noma (BCC), based on its

generally favorable efficacy and adverseeffect profile and its excellent cosmetic out-come. Moreover, because PDT is a non-invasive procedure causing minimaldamage to surrounding tissue, it offers ad-vantages over excision surgery, particu-larly for the treatment of lesions in cos-metically sensitive areas, such as the face.1-3

However, long-term follow-up data of con-trolled clinical trials of topical PDT vs ex-cision surgery (regarded as the treatmentof choice for nodular BCC)4,5 are lacking.

We performed a prospective, random-ized, multicenter study of simple excisionsurgery vs topical PDT, using methyl ami-

nolevulinate, a drug showing highly selec-tive uptake by malignant cells,6 in the treat-ment of primary nodular BCC. Initial datacovering a 2-year follow-up demonstratedthat methyl aminolevulinate PDT was aneffective treatment for nodular BCC. Clear-ance rates of 93% (vs 100% for excision sur-gery) were seen after 3 months and 85% (vs98% for excision surgery) after 12 months,with no significant difference between treat-ment groups. Methyl aminolevulinate PDTalso gave a significant superior cosmeticoutcome compared with excision surgeryat 3, 12, and 24 months.7 At 24 months, 5lesions that initially cleared with methylaminolevulinate PDT had recurred, com-pared with 1 after surgery.

The potential for lesion recurrence fol-lowing PDT is an issue of clinical concern,particularly in the treatment of deeper

Author Affiliations are listed atthe end of this article.



(nodular) BCC. Nearly two-thirds of all recurrent BCC le-sions appear in the first 3 years after treatment, with 18%appearing between 5 and 10 years posttreatment.8 We there-fore performed a 5-year follow-up to examine the sus-tained complete lesion response and recurrence rate overtime for methyl aminolevulinate PDT vs excision surgeryin patients with nodular BCC.

METHODS

European university hospital dermatology departments par-ticipated in a prospective, randomized study of methyl ami-nolevulinate PDT vs simple excision surgery in patients withnodular BCC. The study design and procedures have been de-tailed in a prior report.7 Patients aged at least 18 years with pre-viously untreated primary nodular BCC, suitable for simple ex-cision surgery, were enrolled from the clinics of participatingcenters. The diagnosis of nodular BCC was confirmed by his-tological examination. Patients with more than 10 eligible le-sions; lesions in the midface region, orbital areas, and ears; le-sions with a longest diameter of less than 6 mm or more than15 mm (face or scalp), more than 20 mm (extremities or neck),or more than 30 mm (trunk); and pigmented or morpheaformBCCs were excluded. Patients were enrolled from October 26,1999, to August 29, 2000, and were followed up until October12, 2005. All patients gave written informed consent before en-rollment. The study was performed in accordance with the Dec-laration of Helsinki, with ethical approval obtained from thelocal ethics committee of each of the participating centers.

Eligible patients were randomized to treatment with PDT usingmethyl aminolevulinate cream,160mg/g (Metvix;PhotoCureASA,Oslo, Norway, and Galderma, Sophia Antipolis, France), or ex-cision surgery.7 In the PDT group, lesion surface preparation wasperformed in a standardized manner between centers, before ap-plication of methyl aminolevulinate cream. Surface scale or crustwas gently removed with a curette or scalpel blade, without an-esthesia, insufficient to cause pain or bleeding. An occlusive andlight-shielding dressing was applied, and after 3 hours, the areawas illuminated with noncoherent red light (wavelength, 570-670 nm; total fluence, 75 J/cm2; and fluence rate, 50-200 mW/

cm2). Patients in the PDT group were treated with either 1 or 2PDT cycles, each composed of 2 PDT sessions, with an intervalof 1 week between sessions. Simple elliptical excision surgery withat least 5-mm margins was performed once under local anesthe-sia in accordance with the usual practice of the center.

Patients with a complete response at 3 months after the lasttreatment (ie, complete disappearance of the lesion as evalu-ated on clinical inspection by the same investigator [M.A.R., V.G.,G.A.E.W., or P.W.]) were followed up at yearly intervals for 5years after the last treatment. Any clinically suspected recur-rence was confirmed by histological examination. Investigator-assessed cosmetic outcome was judged yearly for up to 5 yearsin all patients with a complete response in all lesions, using a4-point scale7 as follows: (1) excellent: no scarring, atrophy, orinduration and slight or no redness or change in pigmentationcompared with adjacent skin; (2) good: no scarring, atrophy, orinduration and moderate redness or increase in pigmentation com-pared with adjacent skin; (3) fair: slight to moderate occur-rence of scarring, atrophy, or induration; and (4) poor: exten-sive occurrence of scarring, atrophy, or induration.

Details of adverse events, including local phototoxicity re-actions reported up to 3 months after last treatment, have beenprovided.7 During the 5-year follow-up, any adverse events thatled to discontinuation were documented.

The primary statistical analysis was based on the per pro-tocol (PP) population, including all eligible patients who com-pleted surgery or the first PDT treatment cycle and had a re-sponse assessment at 3 months or who completed the secondPDT treatment cycle and received treatment in accordance withthe study procedures. In addition, an intention-to-treat analy-sis was performed. The analysis populations (intention-to-treat and PP groups) were almost identical and gave similar re-sults in all aspects. Two patients in each treatment group (4 of101) were excluded from the PP population. The present studytherefore focuses on the population of primary interest (ie, thePP population). Analyses were performed independently(PAREXEL International GmbH, Berlin, Germany) using SASstatistical software (SAS Institute Inc, Cary, North Carolina).

Of primary interest during long-term follow-up was the his-tologically confirmed lesion recurrence rate. Lesion recurrencerates were the proportion of recurrent lesions, of all lesions that

101 Patients randomized and treated

52 In the methyl aminolevulinate PDT group (ITT population) 49 In the surgery group (ITT population)

50 Patients in the PP population 47 Patients in the PP population

31 Patients completed the 5-y follow-up 35 Patients completed the 5-y follow-up

2 Not included1 Protocol violation1 Discontinued because

of an adverse event

2 Not included1 Discontinued because

of an adverse event1 Withdrew consent

19 Not included10 Experienced treatment

failure (all lesions)6 Discontinued because

of an adverse event1 Withdrew consent2 Were lost to follow-up

12 Not included2 Experienced treatment

failure (all lesions)4 Discontinued because

of an adverse event4 Withdrew consent2 Were lost to follow-up

Figure 1. Disposition of patients 5 years after the last treatment. The protocol violation in the methyl aminolevulinate group was also a treatment failure (alllesions). ITT indicates intention-to-treat; PDT, photodynamic therapy; and PP, per protocol.



had a clinical complete response 3 months after treatment. Theywere presented with 95% confidence intervals (CIs), and the treat-ment difference was tested with the Fisher exact test.

A time-to-event approach was used to estimate lesion com-plete response rates over time, using all available informationand keeping the original sample size and randomization scheme.The event of interest was failure, defined as nonresponse at 3months after the last treatment or recurrence at the follow-upintervals of 1, 2, 3, 4, or 5 years.

Censoring occurred when a patient with a lesion in completeresponse discontinued for reasons other than treatment failure.Discontinuations or recurrences were only recorded at the yearlyfollow-up. Thus, the exact time when the event occurred was un-known. Furthermore, some patients had more than 1 lesion. Thelesions formed clusters within these patients and therefore couldno longer be regarded as independent observations. The intervalcensoring and the lesion clusters made the data unsuitable for stan-dard procedures, such as the Kaplan-Meier or the life-table ap-proach. Instead, a complementary log-log model suggested by Guoand Lin9 was used. In a first step, a logistic regression using thecomplementary log-log link function was run with standard soft-ware (PROC LOGISTIC; SAS Institute Inc). In a second step, thestandard errors of the estimates were adjusted for the dependen-cies in the data caused by the lesion clusters, as described usingcomputer software (IML; SAS Institute Inc).

Recurrence rates could not be derived from the time-to-event model because the failure rates were a combination ofnonresponse rates at 3 months after the last treatment and re-currence rates during later follow-ups.

The proportions of patients categorized with excellent or goodoverall cosmetic outcome were summarized with 2-sided 95%CIs. The treatment difference was tested with the Fisher exacttest using the categories of excellent or good vs fair or poor.

RESULTS

Of 103 randomized patients, 101 were treated in the study,52 with methyl aminolevulinate PDT and 49 with sur-gery. The PP population comprised 97 patients, 50 with53 lesions treated with methyl aminolevulinate PDT and47 with 52 lesions treated with excision surgery(Figure 1). Four patients were excluded from the PPpopulation, 1 in the methyl aminolevulinate PDT groupwith a major protocol violation (received a light dose less

than that stipulated) and 3 for whom response at 3 monthscould not be assessed (in the surgery group, 1 died and 1withdrew consent; and in the methyl aminolevulinate PDTgroup, 1 discontinued the study prematurely because oftreatment-related pain). Forty-nine lesions in 46 patientstreated with methyl aminolevulinate PDT and 52 lesionsin 47 patients treated with excision surgery showed com-plete lesion response at 3 months and were included inthe long-term follow-up.

Sixty-six patients, 31 treated with methyl aminolevu-linate PDT and 35 treated with surgery, completed the5-year follow-up evaluation with at least 1 lesion in com-plete response. Twelve patients, 10 treated with methylaminolevulinate PDT and 2 treated with surgery, dis-continued the study before the 5-year follow-up evalu-ation, with treatment failure in all lesions (Figure 1). Only4 patients (2 in each treatment group) were truly lost tofollow-up during the 5-year period. Eight patients (4 ineach group) died during the follow-up, and 2 discontin-ued because of intercurrent disease.

The baseline characteristics of the 2 groups were simi-lar.7 The mean (range) age was 69 (40-95) years in themethyl aminolevulinate PDT group and 67 (38-82) yearsin the surgery group; there were 32 men and 20 womenin the methyl aminolevulinate PDT group and 29 menand 20 women in the surgery group. Most patients in eachgroup had 1 lesion (94% [47 of 50 patients] in the methylaminolevulinate PDT group and 89% [42 of 47 pa-tients] in the surgery group). In each group, most le-sions were less than 20 mm in diameter and were lo-cated on the face or scalp or trunk or neck (Table 1).However, the distribution of these lesions differed be-tween the 2 groups. In the surgery group, there were morelesions on the face or scalp than on the trunk or neck,whereas in the methyl aminolevulinate PDT group, thedistribution of lesions to these areas was similar (Table 1).Most lesions in the methyl aminolevulinate PDT groupwere treated with 1 PDT cycle (41 of 53 lesions [77%]).

LESION RESPONSE AND RECURRENCE

A time-to-event analysis of lesion response over timeshowed that excision surgery was more favorable thanmethyl aminolevulinate PDT (Figure 2). At 5 years af-

Table 1. Baseline Characteristics of Lesionsin the PP Populationa

Characteristic

MethylAminolevulinate

PDT Group(n = 53)

SurgeryGroup

(n = 52)

Lesion locationFace or scalp 21 (40) 32 (62)Trunk or neck 27 (51) 15 (29)Extremities 5 (9) 5 (10)

Longest lesion diameter, mmb

�10 29 (55) 34 (65)�10 and �20 19 (36) 14 (27)�20 2 (4) 3 (6)

Abbreviations: PDT, photodynamic therapy; PP, per protocol.aData are given as number (percentage) of lesions for each group.

Percentages may not total 100 because of rounding or because of missing data.bData are missing for 3 patients in the methyl aminolevulinate PDT group

and 1 in the surgery group.

100

50

75

25

0 12 24 36 48 60Time, mo

Lesi

on C

R Ra

te(C

ompl

emen

tary

Log

-Log

Mod

el)

Methyl aminolevulinate PDT group

Surgery group

Figure 2. Lesion complete response (CR) over time (time-to-event analysis)for the per protocol population. PDT indicates photodynamic therapy.



ter last treatment, the sustained lesion complete re-sponse rate, estimated by the complementary log-logmodel, was 76% (95% CI, 59%-87%) for methyl ami-nolevulinate PDT compared with 96% (95% CI, 84%-99%) for excision surgery in the PP population (P=.01).Lesion recurrence was documented in 14% of lesions withcomplete response after 3 months treated with methyl

aminolevulinate PDT, compared with 4% of lesions treatedwith excision surgery (P=.09) (Table 2). The methylaminolevulinate PDT group recurrences included 5 of 40lesions given 1 treatment cycle and 2 of 9 lesions given2 treatment cycles. No recurrences in the methyl ami-nolevulinate PDT group were observed from 3 to 5 years,and only 1 lesion treated with excision surgery recurred

Table 2. Lesion Complete Response and Recurrence Rates in the Per Protocol Population

Variable

Interval After Last Treatment, y

1 2 3 4 5

Methyl Aminolevulinate PDT Groupa,b

Missing, No. (%) 2 (4) 9 (18) 9 (18) 10 (20) 10 (20)Lesion recurrence (n = 49)

No. of lesions 2 5 7 7 7Proportion (95% CI) of lesions, % 4 (1-14) 10 (3-22) 14 (6-27) 14 (6-27) 14 (6-27)c

Lesion size, mmd

�10 1 4 5 5 5�10 and �20 1 1 2 2 2

Surgery Groupb,e

Missing, No. (%) 1 (2) 3 (6) 7 (13) 10 (19) 11 (21)Lesion recurrence (n = 52)

No. of lesions 0 0 1 2 2Proportion (95% CI) of lesions, % 0 (0-100) 0 (0-100) 2 (0-10) 4 (1-13) 4 (1-13)c

Lesion size, mmd

�10 0 0 0 0 0�10 and �20 0 0 1 2 2

Abbreviations: CI, confidence interval; PDT, photodynamic therapy.aComplete response at 3 months occurred in 49 of 53 lesions (92%) (95% confidence interval, 82%-98%).bP = .12 for the treatment difference at 3 months in lesion response rate.cP = .09 for the treatment difference at 5 years in lesion recurrence rate.dThere were no 20-mm or larger lesions in either group.eComplete response at 3 months occurred in 52 of 52 lesions (100%) (95% confidence interval, 93%-100%).

BA

Figure 3. The appearance of the nodular basal cell carcinoma in a study participant in the photodynamic therapy group (a 72-year-old woman) at baseline (A) andat 5 years, showing an excellent cosmetic outcome and complete lesion response (B). In A and B, the inset shows a close-up.



within this period. In the methyl aminolevulinate PDTgroup, there was no evidence that lesion recurrence washigher in larger lesions because recurrence was docu-mented in 5 lesions with a maximum diameter of up to10 mm and in 2 lesions with a diameter between 11 and20 mm (Table 2). Recurrence rates at 5 years were iden-tical in the intention-to-treat population.

COSMETIC OUTCOME

In the PP population, outcome at 3 months rated as ex-cellent or good was found in 36 of 44 patients (82%) (95%CI, 67%-92%) in the methyl aminolevulinate PDT groupvs 15 of 45 patients (33%) (95% CI, 20%-49%) in the sur-gery group (P� .001). The overall cosmetic outcome 5years after last treatment was also superior with methylaminolevulinate PDT compared with excision surgery. Atfinal assessment, outcome as rated by the investigators[L.E.R., M.A.R., R.L., R.C.Y., I.B., V.G., M.-A.R., A.A., andP.W.]was good or excellent in 27 of 31 patients (87%) (95%CI, 70%-96%) treated with methyl aminolevulinate PDTcompared with 19 of 35 patients (54%) (95% CI, 37%-71%) treated with excision surgery (P=.007) (Figure 3).

COMMENT

Findings from this long-term follow-up study add to evi-dence supporting a role for methyl aminolevulinate PDTin the treatment of primary nodular BCC. The cumula-tive 5-year lesion recurrence rate with methyl aminolevu-linate PDT was 14% compared with 4% for excision sur-gery, the latter being consistent with the published lowrate for this modality.8,10,11 We anticipate that we havecaptured all relevant events and that a longer than 5-yearfollow-up would be unlikely to yield further recur-rences because only 2 recurrent lesions were observedin the methyl aminolevulinate PDT group and 1 in thesurgical group between 2 and 3 years’ follow-up. Therewere no recurrent lesions observed in the methyl ami-nolevulinate PDT group and only 1 in the surgical groupfrom 3 to 5 years’ follow-up.

While the sustained lesion complete response rate at5 years was statistically superior with excision surgerycompared with methyl aminolevulinate PDT (96% vs 76%;P=.01), lesions treated with methyl aminolevulinate PDTshowed a substantially better cosmetic outcome at 5 years,as assessed by the investigator, than those treated withsurgery (P=.007). Thus, the results of our study indi-cate that, whereas simple excision surgery will gener-ally remain the treatment of choice for nodular BCC ame-nable to this intervention, methyl aminolevulinate PDTis also an effective treatment. There are a number of in-stances (eg, when there are multiple lesions; when thereare lesions in sites such as the lower leg, where there ispoor healing; and when there are patients undergoing an-ticoagulation) in which excision can be challenging. More-over, methyl aminolevulinate PDT may be more suit-able for the treatment of lesions in cosmetically sensitiveareas, such as the face (involving approximately 50% oflesions in the present study), where optimal cosmetic out-come is an important clinical consideration.

Systematic reviews4,5,8,11 of treatment modalities forBCC highlight the lack of published long-term fol-low-up data from controlled clinical trials and advocateevaluation of 5-year lesion recurrence rates as a primaryoutcome in clinical studies. The lack of long-term datafrom randomized PDT studies has been a particular con-cern among dermatologists.12 To our knowledge, we pre-sent the first randomized study to report cumulative 5-yearrecurrence rates in nodular BCC comparing topical PDTwith the standard treatment of excision surgery. We con-clude that the moderately low 5-year lesion recurrencerate with methyl aminolevulinate PDT documented inthis study, together with the favorable long-term cos-metic outcome, supports a clinical role for this modalityin the treatment of nodular BCC, particularly when avoid-ance of scarring is a priority.

Accepted for Publication: April 4, 2007.Author Affiliations: Photobiology Unit, Departments ofDermatology, University of Manchester and Salford RoyalFoundation Hospital, Manchester (Dr Rhodes), Univer-sity College Hospital, London (Dr Yu), Leeds General In-firmary, Leeds (Dr Goulden), and Royal Liverpool Uni-versity Hospital, Liverpool (Dr Wong), England; AcademicMedical Centre, Amsterdam, the Netherlands (Dr de Rie);Norra Alvsborgs Lanssjukhus, Trollhattan, Sweden(Dr Leifsdottir); Haukeland Hospital, Bergen, Norway(Dr Bachmann); Hopital Sainte-Marguerite, Marseille,France (Dr Richard); University Hospital of Wales, Car-diff (Dr Anstey); and Medical University Graz, Graz, Aus-tria (Dr Wolf).Correspondence: Lesley E. Rhodes, MD, FRCP, Photo-biology Unit, Department of Dermatology, University ofManchester, Salford Royal Foundation Hospital, Manches-ter M6 8HD, England ([email protected]).Author Contributions: Dr Rhodes had full access to allthe data in the study and takes responsibility for the in-tegrity of the data and the accuracy of the data analysis.Study concept and design: Rhodes, de Rie, Goulden, andWong. Acquisition of data: Rhodes, de Rie, Leifsdottir, Yu,Bachmann, Goulden, Wong, Richard, Anstey, and Wolf.Analysis and interpretation of data: Rhodes, de Rie, Yu,Anstey, and Wolf. Drafting of the manuscript: Rhodes, Yu,and Richard. Critical revision of the manuscript for impor-tant intellectual content: Rhodes, de Rie, Leifsdottir, Yu,Bachmann, Goulden, Wong, Richard, Anstey, and Wolf.Statistical analysis: Rhodes. Obtained funding: Rhodes. Ad-ministrative, technical, and material support: Rhodes, deRie, Yu, Richard, Anstey, and Wolf. Study supervision:Rhodes, Yu, and Wolf.Financial Disclosure: None reported.Funding/Support: This study received financial sup-port from Photocure ASA.Role of the Sponsor: Photocure ASA assisted in thedesign and conduct of the study, performed the datacollection, and assisted in the preparation of the manu-script.Additional Contributions: Parexel GmbH, Berlin, Ger-many, performed the data analysis.



REFERENCES

1. Horn M, Wolf P, Wulf HC, et al. Topical methyl aminolaevulinate photodynamictherapy in patients with basal cell carcinoma prone to complications and poorcosmetic outcome with conventional treatment. Br J Dermatol. 2003;149(6):1242-1249.

2. Vinciullo C, Elliott T, Francis D, et al. Photodynamic therapy with topical methylaminolaevulinate for “difficult-to-treat” basal cell carcinoma. Br J Dermatol. 2005;152(4):765-772.

3. Szeimies RM, Morton CA, Sidoroff A, Braathen LR. Photodynamic therapy fornon-melanoma skin cancer. Acta Derm Venereol. 2005;85(6):483-490.

4. Thissen MRTM, Neumann MHA, Schouten LJ. A systematic review of treatmentmodalities for primary basal cell carcinomas. Arch Dermatol. 1999;135(10):1177-1183.

5. Telfer NR, Colver GB, Bowers PW. Guidelines for the management of basal cellcarcinoma. Br J Dermatol. 1999;141(3):415-423.

6. Angell-Petersen E, Sørensen R, Warloe T, et al. Porphyrin formation in actinic

keratosis and basal cell carcinoma after topical application of methyl5-aminolevulinate. J Invest Dermatol. 2006;126(2):265-271.

7. Rhodes LE, de Rie M, Enstrom Y, et al. Photodynamic therapy using topical methylaminolevulinate vs surgery for nodular basal cell carcinoma: results of multi-center randomized prospective trial. Arch Dermatol. 2004;140(1):17-23.

8. Rowe DE, Carroll RJ, Day CL. Long-term recurrence rates in previously un-treated (primary) basal cell carcinoma: implications for patient follow-up. J Der-matol Surg Oncol. 1989;15(3):315-328.

9. Guo SW, Lin DY. Regression analysis of multivariate grouped survival data.Biometrics. 1994;50(3):632-639.

10. Silverman MK, Kopf AW, Bart RS, Grin CM, Levenstein MS. Recurrence rates oftreated basal cell carcinomas, part 3: surgical excision. J Dermatol Surg Oncol.1992;18(6):471-476.

11. Bath-Hextall F, Bong J, Perkins W, Williams H. Interventions for basal cell car-cinoma of the skin: systematic review. BMJ. 2004;329(7468):705-709.

12. Morton CA, Brown SB, Collins S, et al. Guidelines for topical photodynamic therapy:report of a workshop of the British Photodermatology Group. Br J Dermatol. 2002;146(4):552-567.

Call for Papers

Genetics and Genomic Manuscripts

In March 2008, the JAMA and Archives family of journalswill publish manuscripts in a theme issue devoted to prac-tical applications of genetics and genomics that are cur-rently clinically important or may become clinically rel-evant in the near future. We invite authors to submitmanuscripts reporting the results of original research, es-pecially clinical trials; systematic reviews including meta-analyses; special communications; and commentaries. Evi-dence-based articles will be given priority.

Topics of particular interest include genetic diagnosis,including prenatal tests, and genetic testing, especially forillnesses for which presymptomatic intervention such asprevention of disease or increased surveillance for diseasecan be expected to change the patient’s quality of life. Otherpresymptomatic interventions may include pharmacoge-nomics, gene therapy, evolutionary medicine such as geno-types with proven adaptive responses to emerging infec-tions, genetic counseling, and ethical issues thatdermatologists will need to consider as the role of genet-ics and presymptomatic interventions expands.

ManuscriptsreceivedbyOctober15,2007,willhavethebestchanceforconsiderationforthis themeissue.Allmanu-scriptswillundergoourusual rigorouseditorial reviewpro-cess.AuthorsareencouragedtoconsulttheArchivesInstruc-tions for Authors at http://www.archdermatol.com forguidelines on preparing and submitting manuscripts.

June K. Robinson, MDEditorJeffery P. Callen, MDAssociate Editor



STUDY

Effect of Cold Air Cooling on the Incidenceof Postinflammatory Hyperpigmentation AfterQ-Switched Nd:YAG Laser Treatment of AcquiredBilateral Nevus of Ota–like MaculesWoraphong Manuskiatti, MD; Sasima Eimpunth, MD; Rungsima Wanitphakdeedecha, MD

Objective: To investigate the effect of cold air coolingon the incidence of postinflammatory hyperpigmenta-tion (PIH) after laser treatment in Asian patients.

Design: Randomized, controlled, split-face study.

Setting: Skin laser center of a university hospital.

Patients: Twenty-three Thai women with acquired bi-lateral nevus of Ota–like macules.

Interventions: Patients were treated using a 1064-nmQ-switched Nd:YAG laser at an average fluence of 7.0J/cm2 using a 3-mm spot size. The same laser fluence wasused on both sides of the face in individual patients. Onerandomly selected face side of each patient was cooledusing a cold air cooling device during and 30 secondsbefore and after laser irradiation, and the other side wasirradiated without cooling.

Main Outcome Measures: Occurrence of PIH was ob-jectively evaluated by measuring the melanin index using

a spectrometer, and it was subjectively assessed by 2 non-treating physicians before treatment and once weekly for4 weeks.

Results: Of the 21 patients who completed the study,13 (62%) and 5 (24%) developed PIH on the cooled anduncooled sides, respectively. One patient (5%) had PIHon both the cooled and uncooled sides, and 2 (10%) didnot experience PIH. The cooled sides were significantlymore likely to become hyperpigmented after laser irra-diation than the uncooled sides (relative risk, 2.6; 95%confidence interval, 1.13-6.00; P=.03). The clinical evalu-ation corresponded to the spectrometer reading.

Conclusion: Epidermal cooling with cold air is associ-ated with an increased risk of PIH after Q-switched Nd:YAG laser treatment.

Trial Registration: clinicaltrials.gov Identifier:NCT00287001

Arch Dermatol. 2007;143(9):1139-1143

P OSTINFLAMMATORY HYPERPIG-mentation (PIH) is a fre-quently encountered prob-lem and represents thesequelae of various cutane-

ous disorders and therapeutic interven-tions. The pathogenesis of PIH includes anincrease in melanin production and an ab-normal distribution of this pigment. Af-ter trauma or cutaneous inflammation, me-lanocytes can react with increased ordecreased production of melanin, re-flected clinically as hyperpigmentation orhypopigmentation.1 Postinflammatory hy-perpigmentation is probably the most com-mon adverse effect of laser treatments indark-skinned individuals.2,3 Treatment ofPIH is difficult because there are few, if any,therapeutic options that are consistentlysuccessful. Little is known about whetherPIH can be prevented or minimized.4,5

Acquired bilateral nevus of Ota–like mac-ules (ABNOMs), or Hori nevus, is clini-

cally characterized by blue-brown mac-ules occurring bilaterally on the forehead,temples, eyelids, malar areas, nasal alae, andnasal root. Unlike the nevus of Ota, thesepigmented lesions usually appear in thefourth or fifth decade of life in women (onlyrarely in men) and are not observed in theconjunctiva or mucous membranes of themouth or nose.6,7 Histologically, active mela-nin-synthesizing dermal melanocytes aredispersed in the papillary and middle por-tions of the dermis.8 Q-switched lasers havebeen used successfully as a treatment mo-dality.9-11 However, the incidence of tran-sient PIH after laser treatment is as muchas 50% to 73%.2

Skin cooling has been used to protect theepidermis in a variety of laser dermato-logic procedures, including leg vein treat-ment, hair removal, and port-wine stain re-moval. The use of epidermal cooling duringlaser treatment decreases procedure-associated pain, allows safer treatment of

Author Affiliations:Department of Dermatology,Faculty of Medicine SirirajHospital, Mahidol University,Bangkok, Thailand.



darker-skinned individuals, and allows the use of higherfluences.12-14 Studies to determine the benefit of epider-mal cooling on the prevention of PIH after laser treatmentare limited. Epidermal cooling is thought to reduce the non-specific thermal injury caused by the laser pulses and con-sequently to minimize the incidence of PIH.

The present study was performed to address the ad-vantage of epidermal cooling on reducing the occur-rence of PIH after laser treatment of ABNOMs. How-ever, we found that epidermal cooling with cold air wasassociated with an increased risk of PIH after Q-switched Nd:YAG laser treatment. An unexpected find-ing was a clinically and statistically significant darken-ing of the cooled side compared with the uncooled side.

METHODS

Twenty-three Thai women with ABNOMs (Hori nevus) wereenrolled in the study. The patients were randomized as to whichside of the face would be treated with Q-switched Nd:YAG la-ser and a cold air cooling device vs Q-switched Nd:YAG laseralone by using a table of uniform random digits. The study pro-tocol and informed consent documents were submitted and ap-proved by the Ethics Committee on Research Involving Hu-man Subjects, Faculty of Medicine Siriraj Hospital, MahidolUniversity. Written informed consent was obtained from eachpatient before enrollment.

PATIENT CHARACTERISTICS

The mean age of the patients was 43 years (range, 27-72 years).Their skin types were III (17/23) and IV (6/23). Patients whowere pregnant or lactating, who were taking birth control pillsor undergoing hormone therapy, or who participated in majoroutdoor activities were excluded. Patients who used bleach-ing agents, such as hydroquinone preparation, and those whoreceived chemical peels, laser treatment, or intense pulsed lighttreatment before enrollment were also excluded.

TREATMENT

Lidocaine, 2.5%, and prilocaine, 2.5%, cream (a eutectic mix-ture of local anesthetic) (AstraZeneca LP, Wilmington, Dela-ware) was applied to the lesions 1 hour before treatment, and

a dressing (Tegaderm; 3M, St Paul, Minnesota) was used to oc-clude the area. Each patient received 1 Q-switched Nd:YAG la-ser treatment. The uncooled side of the patient’s face was al-ways treated first, followed by the side with cold air cooling.The clinical end point was defined as the point of immediatewhitening without bleeding and tissue splatter. The same flu-ence setting was used on both sides.

COOLING TECHNIQUE

A commercially available cold air cooling device (CRIOjet AIRMini; CRIO Medizintechnik GmbH, Birkenfeld, Germany) wasused at a cooling level of 4 for all treatments. This instrumentworks with a compressor system similar to those in refrigera-tors and uses ambient air to generate a permanent stream ofcold air with a flow of 500 to 1000 L/min and temperatures aslow as −30°C, depending on the cooling delivery system andthe desired cooling level (range, 1-9). The nozzle-to–cooled sur-face distance was 5 cm (as suggested by the manufacturer). Rela-tive humidity was 60%, and room temperature was approxi-mately 25°C. The nozzle was held with a rotating motion duringthe treatment. The cooled side was always cooled during and30 seconds before and after laser irradiation. The skin surfacetemperature, measured using an infrared thermometer (Mini-Temp MT4; Raytek Corp, Santa Cruz, California) during thetreatment, was 4°C to 5°C.

Postoperatively, antibacterial mupirocin ointment (Bactro-ban Ointment; SmithKline Beecham Pharmaceuticals, Philadel-phia, Pennsylvania) was applied to the treated area. Patients wereinstructed to cleanse the treated sites gently with tap water andto reapply the mupirocin ointment 4 times a day until all crust-ing had subsided. After crusting completely healed, all the pa-tients were instructed to wear a broad spectrum sunscreen witha sun protection factor of 40 and were also asked to avoid sunexposure, which could cause hyperpigmentation.

CLINICAL EVALUATION

Masked assessment of the occurrence of PIH in all the patientswas made by 2 expert dermatologists from the digital photo-graphs taken before treatment (baseline) and 1, 2, 3, 4, and 12weeks after laser treatment. Photographs were taken using a digi-tal camera (D70s; Nikon Corporation, Tokyo, Japan) equippedwith a 60-mm lens (Nikkor; Nikon Corporation) and CanfieldTwinFlash (Canfield Scientific Inc, Fairfield, New Jersey). All thephotographs were taken in raw format with identical conditions

A B

Figure 1. The left side of the face before laser treatment with cold air cooling (A) and the right side of the face before laser treatment without cooling (B).



and camera settings. Standardized views (face front and 45° ob-lique) were used. The liquid crystal display monitor was cali-brated for precise on-screen displays using a colorimeter (Spyder2;Colorvision Inc, Lawrenceville, New Jersey). Clinical observa-tions for cutaneous changes (specifically, erythema, urticaria, blis-tering, dyspigmentation, and scarring) were conducted at 1-hourand 1-week intervals after cold air exposure.

Observable skin changes were graded on the following scale:0, absent; 1, minimal (light pink; barely perceptible skin el-evation; whitening of skin; no evident blister); 2, mild (lightred/pink; slight skin elevation; skin wrinkling; no well-formed blister); 3, moderate (red; clearly visible urticaria; flac-cid blisters); and 4, severe (very red; pronounced urticaria; tenseblisters). Clinical evaluation for the improvement in lesionalpigmentation was subjectively graded at baseline and week 12using a quartile grading scale as follows: poor, less than 25%;fair, 25% to 50%; good, 51% to 75%; and excellent, more than75% improvement.

COLOR READING

The designated lesions of pigmented macules on both sides ofthe face were marked on every patient and mapped with a trans-lucent sheet at the first visit to ensure consistency of location.The degree of lesion lightening was defined as the percentageof darkness reduction compared with baseline. The degree ofpigmentation was assessed using a handheld reflectance spec-trometer (DermaSpectrometer; Cortex Technology, Hadsund,Denmark). By emitting red light at 655 nm and measuring theabsorption coefficients in the skin, the instrument provides areadout of a melanin index (M-index) based on the absorp-tion characteristics of the skin pigmentation. During measure-ments, the device was placed perpendicularly on the skin;blanching was minimized by applying light pressure. The valueused was the average of 3 measurements at each site at eachassessment. A higher M-index value indicates increased satu-ration toward darkness.


Two-tailed paired t test analyses using statistical software (SPSSversion 14; SPSS Inc, Chicago, Illinois) were performed to com-pare mean M-index values for each treatment technique andat baseline vs the 12-week follow-up visit. Relative risk was cal-culated using statistical software (Epi Info version 3.3.2; Cen-ters for Disease Control and Prevention, Atlanta, Georgia).P� .05 was considered statistically significant.

RESULTS

Of the 23 patients, 21 completed all the follow-up visits.Two patients were withdrawn from the study because theydid not attend scheduled visits. Acute erythema was notedin 18 of 21 patients (86%), 1� erythema was observedin 16 of 21 patients (76%), and 2� erythema was pres-ent in 2 of 21 patients (10%). One hour after treatment,the redness disappeared in 12 of 18 patients (67%) whodeveloped erythema. In all the patients, the erythema re-solved in 24 hours. One patient each with skin types IIIand IV developed 1� urticaria. All urticaria resolved bythe day 1 follow-up visit. No blistering was observed inany patients.

The crusting healed completely within 1 and 2 weeksof treatment in 17 of 21 patients (81%) and 4 of 21 pa-tients (19%), respectively. No difference in crusting reso-lution rates was observed between the cooled and un-cooled sides. No patients developed hypopigmentationor scarring. All the patients noted significant pain reduc-tion on the cooled side during laser treatment. No post-operative analgesic treatment was required beyond theapplication of mupirocin ointment and ice compresses.

As judged by the 2 expert dermatologists, of the 21patients who completed the study, 13 (62%) developedPIH on the cooled side, and 5 (24%) had PIH on the un-cooled side (Figures 1, 2, 3, and 4). One patient (5%)had PIH on both the cooled and uncooled sides, and 2(10%) did not experience PIH. The cooled sides were 3times more likely than the uncooled sides to develop PIHafter laser treatment (relative risk, 2.6; 95% confidenceinterval, 1.13-6.00; P=.03). The onset of PIH was at week1, 2, 3, and 4 in 5% (1/21), 62% (13/21), 14% (3/21), and10% (2/21) of the patients, respectively. The PIH was com-pletely resolved 12 weeks after laser treatment in all thepatients except 1 with skin type IV, who developed PIHon the cooled side.

All the patients showed less than 25% overall light-ening of the pigmented macules at week 12. No differ-ence in clinical improvement was observed regarding thecooling used on 1 side during treatment. At baseline, therewas no significant difference in the M-index values of the

A B

Figure 2. The left (cooled) (A) and right (uncooled) (B) sides of the face 1 week after treatment.



cooled and uncooled sides (P=.17). At week 12, therewas also no significant difference in the M-index valuesof the lesions on the cooled (P=.07) and uncooled (P=.55)sides compared with baseline. The M-index value of allareas considered to have PIH was significantly in-creased compared with the baseline value (P=.03).

COMMENT

Postinflammatory hyperpigmentation is one of the mostcommon adverse effects in dark-skinned patients afterlaser treatment. It is not life threatening, but PIH maycause substantial psychological problems. The inci-dence of transient PIH after carbon dioxide laser resur-facing is approximately 37% in patients with all skintypes15 and nearly 70% in those with skin type IV orhigher.16

Hori nevus or ABNOMs is a common pigmentary dis-order in Asian individuals. Q-switched lasers provide fa-vorable treatment responses, but transient PIH after la-ser irradiation affects as much as 50% to 73% of patientsand can last for many months.11,17 Postinflammatory hy-

perpigmentation takes a long time to dissipate. The treat-ment of PIH is difficult and time-consuming, often last-ing many months to achieve the desired results, whichcauses frustration in patients and physicians.1,2 Thus, vari-ous attempts to reduce the occurrence of PIH after skinlaser surgery have included sun avoidance, use of pre-operative and postoperative treatment regimens, and tech-niques for epidermal protection.

Current commercial cooling methods and devices in-clude passive cooling with aqueous gel, active cooling withwater or refrigerated air, and dynamic active cooling withcryogen spray. Lowering the temperature of the skin’s sur-face is a method of selectively controlling the depth atwhich heat is produced in the skin by lasers or pulsedlight sources. Cooling of the epidermis can prevent itstemperature elevation from exceeding the threshold forthermal injury.12 Minimizing the nonspecific thermal in-jury to the skin surface caused by the laser pulses maytherefore lower the incidence of PIH.

A previous study18 of a freeze injury in normal hu-man skin showed that a brief (5-second) freezing of liq-uid nitrogen caused hypopigmentation, with a periph-

A B

Figure 4. The left (cooled) (A) and right (uncooled) (B) sides of the face 12 weeks after treatment. Note the disappearance of the postinflammatoryhyperpigmentation on the left (cooled) side of the face.

A B

Figure 3. The left (cooled) (A) and right (uncooled) (B) sides of the face 2 weeks after treatment. Note the postinflammatory hyperpigmentation of the lesions onthe left (cooled) side of the face.



eral rim of hyperpigmentation in all frozen lesions.Similarly, a recent study by Datrice et al19 also found thattransient hyperpigmentation occurred after cryogen spraycooling exposure in individuals with skin type III orhigher, and all hyperpigmentation was resolved by 8 weekswithout medical intervention.

In trying to address the possible causes of this occur-rence, we tested the effect of the cold air cooling in 2 pa-tients with untreated ABNOMs (not in the present study).The same cooling techniques were performed without la-ser irradiation on 1 side of the face of each patient. Thepatients were followed up every week for 4 weeks. How-ever, none of the patients developed PIH on the cold air–exposed side. Thus, it is unlikely that cold air exposurealone is the mechanism causing PIH. However, some in-flammatory triggers from the laser irradiation may un-derlie the increased incidence of PIH in the present study.

The underlying mechanisms and the variability indi-viduals show for developing hyperpigmentation or hypo-pigmentation are not well understood. An inherited indi-vidual chromatic tendency based on “strong” or “weak”melanocytes and their tendency to respond to trauma orinflammation with either hyperpigmentation or hypopig-mentation has been postulated.1 In addition, keratino-cytes probably play an active role in the regulation of me-lanogenesis and can stimulate melanocytes to becomedendritic and transfer melanosomes.20 We therefore hy-pothesize that the laser-irradiated melanocyte or keratino-cyte may be more hyperactive to stimuli such as cold tem-perature.

In conclusion, the mechanism by which cold air cool-ing is associated with an increased risk of PIH after lasertreatment is unknown. It is possible that the mecha-nism was a melanocyte’s or keratinocyte’s responses tothe laser pulses together with cold air exposure. Futurestudies should address the question of whether the othermethods of epidermal cooling are associated with an in-crease risk of PIH. Of particular interest is the mecha-nism that underlies the cause of hyperpigmentation af-ter cold air cooling.

Accepted for Publication: March 6, 2007.Correspondence: Woraphong Manuskiatti, MD, Depart-ment of Dermatology, Faculty of Medicine, Siriraj Hos-pital, Mahidol University, 2 Pran-nok Rd, Bangkok 10700,Thailand ([email protected]).Authors Contributions: Study concept and design:Manuskiatti and Wanitphakdeedecha. Acquisition of data:Manuskiatti and Eimpunth. Analysis and interpretationof data: Manuskiatti. Drafting of the manuscript: Manuski-atti and Eimpunth. Critical revision of the manuscript for

important intellectual content: Wanitphakdeedecha. Sta-tistical analysis: Manuskiatti and Wanitphakdeedecha. Ad-ministrative, technical, and material support: Eimpunth.Study supervision: Manuskiatti.Financial Disclosure: None reported.Previous Presentation: This study was presented at theAnnual Meeting of the American Society for Dermato-logic Surgery; October 29, 2006; Palm Desert, California.

REFERENCES

1. Ruiz-Maldonado R, Orozco-Covarrubias ML. Postinflammatory hypopigmenta-tion and hyperpigmentation. Semin Cutan Med Surg. 1997;16(1):36-43.

2. Chan HH, Alam M, Kono T, Dover JS. Clinical application of lasers in Asians. Der-matol Surg. 2002;28(7):556-563.

3. Manuskiatti W, Goldman MP. Use of lasers on Asian skin. In: Goldman MP, ed.Cutaneous and Cosmetic Laser Surgery. Philadelphia, PA: Mosby Elsevier; 2006:349-374.

4. Grimes P, Nordlund JJ, Pandya AG, Taylor S, Rendon M, Ortonne JP. Increas-ing our understanding of pigmentary disorders. J Am Acad Dermatol. 2006;54(5)(suppl 2):S255-S261.

5. Pandya AG, Guevara IL. Disorders of hyperpigmentation. Dermatol Clin. 2000;18(1):91-98, ix.

6. Ee HL, Wong HC, Goh CL, Ang P. Characteristics of Hori naevus: a prospectiveanalysis. Br J Dermatol. 2006;154(1):50-53.

7. Hori Y, Kawashima M, Oohara K, Kukita A. Acquired, bilateral nevus of Ota–likemacules. J Am Acad Dermatol. 1984;10(6):961-964.

8. Hori Y, Takayama O. Circumscribed dermal melanoses: classification and his-tologic features. Dermatol Clin. 1988;6(2):315-326.

9. Lam AY, Wong DS, Lam LK, Ho WS, Chan HH. A retrospective study on the effi-cacy and complications of Q-switched alexandrite laser in the treatment of ac-quired bilateral nevus of Ota–like macules. Dermatol Surg. 2001;27(11):937-942.

10. Manuskiatti W, Sivayathorn A, Leelaudomlipi P, Fitzpatrick RE. Treatment of ac-quired bilateral nevus of Ota–like macules (Hori’s nevus) using a combination ofscanned carbon dioxide laser followed by Q-switched ruby laser. J Am AcadDermatol. 2003;48(4):584-591.

11. Polnikorn N, Tanrattanakorn S, Goldberg DJ. Treatment of Hori’s nevus with theQ-switched Nd:YAG laser. Dermatol Surg. 2000;26(5):477-480.

12. Nelson JS, Majaron B, Kelly KM. Active skin cooling in conjunction with laserdermatologic surgery. Semin Cutan Med Surg. 2000;19(4):253-266.

13. Raulin C, Greve B, Hammes S. Cold air in laser therapy: first experiences with anew cooling system. Lasers Surg Med. 2000;27(5):404-410.

14. Zenzie HH, Altshuler GB, Smirnov MZ, Anderson RR. Evaluation of cooling meth-ods for laser dermatology. Lasers Surg Med. 2000;26(2):130-144.

15. Nanni CA, Alster TS. Complications of carbon dioxide laser resurfacing: an evalu-ation of 500 patients. Dermatol Surg. 1998;24(3):315-320.

16. Sriprachya-anunt S, Marchell NL, Fitzpatrick RE, Goldman MP, Rostan EF. Fa-cial resurfacing in patients with Fitzpatrick skin type IV. Lasers Surg Med. 2002;30(2):86-92.

17. Kunachak S, Leelaudomlipi P, Sirikulchayanonta V. Q-switched ruby laser therapyof acquired bilateral nevus of Ota–like macules. Dermatol Surg. 1999;25(12):938-941.

18. Burge SM, Bristol M, Millard PR, Dawber RP. Pigment changes in human skinafter cryotherapy. Cryobiology. 1986;23(5):422-432.

19. Datrice N, Ramirez-San-Juan J, Zhang R, et al. Cutaneous effects of cryogen spraycooling on in vivo human skin. Dermatol Surg. 2006;32(8):1007-1012.

20. Minwalla L, Zhao Y, Le Poole IC, Wickett RR, Boissy RE. Keratinocytes play arole in regulating distribution patterns of recipient melanosomes in vitro. J In-vest Dermatol. 2001;117(2):341-347.



STUDY

Primary Cutaneous Diffuse Large B-CellLymphoma, Leg Type

Clinicopathologic Features and Prognostic Analysis in 60 Cases

Florent Grange, MD, PhD; Marie Beylot-Barry, MD, PhD; Phillipe Courville, MD; Eve Maubec, MD;Martine Bagot, MD, PhD; Beatrice Vergier, MD, PhD; Pierre Souteyrand, MD; Laurent Machet, MD, PhD;Sophie Dalac, MD; Eric Esteve, MD; Isabelle Templier, MD; Emmanuel Delaporte, MD; Marie-Francoise Avril, MD;Caroline Robert, MD, PhD; Stephane Dalle, MD; Liliane Laroche, MD, PhD; Michele Delaunay, MD;Pascal Joly, MD, PhD; Janine Wechsler, MD; Tony Petrella, MD

Objectives: To describe clinicopathologic features andto identify prognostic factors in a large series of primarycutaneous diffuse large B-cell lymphoma, leg type (PCLBCLLT), as defined in the recent World Health Organization–European Organization for Research and Treatment of Can-cer classification of cutaneous lymphomas.

Design: Retrospective multicenter study from the FrenchStudy Group on Cutaneous Lymphomas.

Setting: Nineteen departments of dermatology in 10 re-gions of France.

Patients: Sixty patients with a PCLBCL LT included inthe registry of the French Study Group on CutaneousLymphomas.

Main Outcome Measures: Age, sex, outcome, therapy,B symptoms, cutaneous extent, number of lesions, loca-tion (leg vs nonleg), serum lactate dehydrogenase level,and MUM-1 and Bcl-2 expression were recorded. Disease-specific survival was used as the main end point. Prog-nostic factors were identified using a Cox proportionalhazards model.

Results: Primary cutaneous diffuse large B-cell lym-phoma, leg type is characterized by a predilection forthe leg (72%), a high proportion of Bcl-2 expression(85%), an advanced age at onset (mean age, 76 years),and frequent relapses and extracutaneous dissemina-tion. The overall 5-year disease-specific survival ratewas 41%. Location on the leg and multiple skin lesionswere predictive of death in multivariate analysis.Although no variable related to therapy was signifi-cantly associated with survival, patients recently treatedwith combinations of anthracycline-containing chemo-therapies and rituximab had a more favorable short-term outcome.

Conclusions: Primary cutaneous diffuse large B-celllymphoma, leg type is a distinct entity with a poorprognosis, particularly in patients with multiple tumorson the legs. Despite the advanced age of many patients,the prognosis could be improved with combinationsof anthracycline-containing chemotherapies andrituximab.

Arch Dermatol. 2007;143(9):1144-1150

I N 2005, A NEW WORLD HEALTH

Organization–European Organi-zation for Research and Treat-ment of Cancer (WHO-EORTC)classification of cutaneous lym-

phomas was proposed1 and was widelyaccepted2,3 as an important consensual ad-vance for the characterization and manage-ment of cutaneous lymphomas. In this clas-sification, primary cutaneous B-celllymphomas (PCBCLs) were divided into thefollowing 3 main groups: primary cutane-ous marginal zone B-cell lymphoma, pri-mary cutaneous follicle center cell lym-phoma, and primary cutaneous diffuse largeB-cell lymphoma, leg type (PCLBCL LT).Whereas the first 2 groups had been iden-tified long before and studied in numerous

large reports,4-10 the third group was newlydefined and had not yet been character-ized by large multicenter studies.

Since the report of the WHO-EORTCclassification, 2 studies11,12 have included in-formation about 40 and 51 cases of PCLBCLLT as subgroup analyses of larger series ofPCBCLs of different subtypes. However, nolarge studies specifically dedicated toPCLBCL LT have been published to date,to our knowledge. In addition, althoughPCLBCL LT is known to have a more ag-gressive clinical behavior than other groupsof PCBCLs,1,11,12 we are aware of only 1study11 that has attempted to identify prog-nostic factors in these patients. In this study,we analyzed the outcome, therapy, prog-nostic factors, clinical characteristics, and

Author Affiliations are listed atthe end of this article.



immunohistologic features in a large series of patients withPCLBCL LT.

METHODS

PATIENT SELECTION

A retrospective review of PCBCLs included in the registry of theFrench Study Group on Cutaneous Lymphomas between Janu-ary 1, 1988, and March 31, 2006, was carried out. Patients withPCLBCL LT according to the WHO-EORTC classification1 wereselected for analysis after clinicopathologic review.

Primary cutaneous B-cell lymphomas were defined as B-cell lymphomas manifesting on the skin without evidence ofextracutaneous disease. Among PCBCLs, PCLBCL LT was de-fined histologically as cases with a predominance or confluentsheets of centroblasts and immunoblasts, irrespective of the lo-cation of skin lesions and Bcl-2 protein expression. There wereno recorded patients with associated chronic lymphocytic leu-kemia or Richter syndrome.

To avoid selection biases, events occurring after diagnosiswere not used as exclusion criteria. In particular, patients un-able to receive any treatment because of their poor general con-dition, those who developed extracutaneous disseminationwithin 6 months after diagnosis, and those who died soon there-after were included in the study.

HISTOLOGIC ANDIMMUNOHISTOCHEMICAL REVIEW

All cases were reviewed by a panel of 3 of us (B.V., J.W., and T.P.)who were blinded to the clinical data to confirm the inclusioncriteria based on the predominance (�80%) of centroblasts (largecells with round or oval nuclei and several small nucleoli gener-ally sticking to the nucleus membrane) and immunoblasts (largecells with round nuclei and �1 large central nucleoli). Cases show-ing a large proportion of large centrocytic cells (cleaved cells) wereexcluded. For each case, hematoxylin-eosin stains and CD20- andCD3-immunostained slides from formalin-fixed and paraffin-embedded biopsy specimens were required. In all cases, tumorcells expressed CD20 and were negative for CD3. The cases thatdid not correspond to the morphological and phenotypical cri-teria of centroblastic or immunoblastic B-cell lymphomas wereexcluded. Unstained sections or paraffin blocks were subse-quently collected from each center to perform Bcl-2 and MUM-1immunostainings. The analysis was performed in the same labo-ratory (that of T.P.) at the same time to avoid technique-dependent variability using an automated system (BenchMark;Ventana Medical System, Illkich, France). We used reactive lymphnode specimens as positive and negative controls. The followingmonoclonal antibodies were used for the study: clone 124 for Bcl-2protein and clone mum1p for MUM-1 protein (Dako Company,Globstrup, Denmark). After consensus among reviewers, an es-timated quantification of the proportion of neoplastic large cellsthat showed unequivocal Bcl-2 and MUM-1 positivity was given,ranging from 0% to 100%. As in previous studies, Bcl-2 expres-sion was considered positive when this proportion exceeded 50%.MUM-1 expression was categorized as positive (�50% positivetumorcells), intermediate (30%-49%positive tumorcells), ornega-tive (�30% positive tumor cells).

CLINICAL AND FOLLOW-UP DATA

All medical records were reviewed. The following clinical char-acteristics were recorded at diagnosis and were evaluated forprognostic value: age, sex, B symptoms, number of skin le-

sions, duration of skin lesions before diagnosis, serum lactatedehydrogenase level, anatomical site (arm, leg, head and neck,anterior aspect of the trunk, or posterior aspect of the trunk),and cutaneous extent (namely, “localized” when 1 or multipleskin lesions were restricted to 1 anatomical site and “dissemi-nated” when several anatomical sites were involved). Fol-low-up data were recorded until April 1, 2006, including ini-tial and subsequent therapies, achievement of a completeresponse, cutaneous relapse, extracutaneous progression of thedisease, final status, and date and cause of death. Causes of deathwere ascertained in most cases by physician members of theFrench Study Group on Cutaneous Lymphomas who fol-lowed up patients and in other cases by questioning their gen-eral practitioners. Follow-up ranged from 0.3 to 155 months(mean follow-up, 32 months). Fifty-two patients (86.7%) werefollowed up until death, until the end point, or for longer than5 years, whereas 8 patients (13.3%) were lost to follow-up af-ter less than 5 years (range, 17-49 months).


Disease-specific survival duration was calculated from diagnosisto the date of disease-related death or censoring. Patients whosedeaths were unrelated to lymphoma were considered censored.

Prognostic factors were identified by disease-specific sur-vival univariate and multivariate analyses using a Cox propor-tional hazards model.13 Factors significant at the 0.2 level inunivariate analysis were included in stepwise regression mul-tivariate analyses. Comparisons between subgroups of pa-tients according to factors of prognostic value were performedusing �2 test or Fisher exact test for categorical variables and ttest or Mann-Whitney test for continuous variables. The Kaplan-Meier method was used to estimate lymphoma-specific sur-vival and to construct corresponding survival curves.14 Sur-vival rates were compared separately for each category ofprognostic variable identified by the Cox proportional haz-ards model using the Mantel-Cox test.

RESULTS

BASELINE CLINICAL AND HISTOLOGICCHARACTERISTICS AND FOLLOW-UP DATA

Sixty patients met the inclusion criteria. Thirty-two hadbeen included in a previous study.15 All patients had nega-tive staging investigations at diagnosis, including physi-cal examination (100% of cases), routine laboratory testresults (100%), bone marrow cytologic or histologic fea-tures (90%), chest radiography or thoracic computed to-mography (100%), abdominal ultrasonographic tomog-raphy or abdominal computed tomography (100%), andlymph node histologic features in patients with clini-cally enlarged lymph node (13%).

The main characteristics of patients at diagnosis and thefollow-up data are summarized in Table 1. The female-male sex ratio was 1.6. Patient age ranged from 44 to 96years (mean age, 76 years; median age, 77 years). The per-formance status was 0 (fully active), 1 (ambulatory), 2 (bed-ridden �50% of the time), 3 (bedridden �50% of the time),and 4 (completely bedridden) in 12, 13, 4, 4, and 2 pa-tients, respectively, and was unavailable for 25 patients.Fifty-four patients had cutaneous nodules or tumors, 3 pa-tients had deeply infiltrated plaques, 2 patients had largesubcutaneous tumors, and 1 patient had a leg ulcer. Twenty



patients (33.3%) had 1 lesion, 19 patients (31.7%) had 2to 5 lesions, and 21 (35.0%) had more than 5 lesions. Eightpatients (13.3%)had lesionson the trunk,9patients (15.0%)had lesions on the arm, 11 patients (18.3%) had lesions onthe head, and 43 patients (71.7%) had lesions on the leg(in 5 of these 43 patients, lesions were located both on theleg and at other anatomical sites). Using the previousEORTC classification,16 all 43 of these cases would havebeen classified by most authors as PCLBCL of the leg,whereas 17 cases without leg lesions (28.3%) would havebeen classified as primary cutaneous follicle center cell lym-phoma. In the present study, these 17 cases are referred toas nonleg PCLBCL LT, as defined in the WHO-EORTCclassification.1

Seven patients (11.7%) had a high serum lactate de-hydrogenase level, 6 patients (10.0%) had B symptoms,and 2 patients (3.3%) had both conditions. Bcl-2 stain-ing was positive in 51 patients (85.0%). MUM-1 stain-ing was scored as positive, intermediate, and negative in

23 (56.1%), 5 (12.2%), and 13 (31.7%) of 41 patients,respectively, and was unavailable in 19 patients.

Table 2 summarizes first-line and subsequent thera-pies administered in our 60 patients. First-line therapywas most often radiation therapy (38.3%) or variable com-binations of systemic chemotherapies (49.9%) with orwithout rituximab. In addition, 17 patients (28.3%) re-ceived second-line or third-line therapies. Overall, 39 pa-tients (65.0%) were treated with polychemotherapies asfirst-line or subsequent therapies during the course of thedisease (data not shown). These included 9 patients(15.0%) who received polychemotherapies only with-out anthracycline, 18 patients (30.0%) who received an-thracycline-containing chemotherapies without ritux-imab, and 12 patients (20.0%) who received variouscombinations of anthracycline-containing chemothera-pies and rituximab. No statistically significant differ-ence with respect to therapies administered was ob-served between patients with leg vs nonleg PCLBCL LT.

Table 1. Main Findings at Diagnosis and Follow-up Data and According to Location of Skin Tumorsa

Variable Total Leg Nonleg P Valueb

No. of cases 60 43 17Age, y

Mean (range) 76 (44-96) 78 (44-92) 72 (49-96) .03�75 35 (58.3) 30 (69.8) 5 (29.4) .004

Sex .4Male 23 (38.3) 15 (34.9) 8 (47.1)Female 37 (61.7) 28 (65.1) 9 (52.9)

No. of lesions �.0011 20 (33.3) 8 (18.6) 12 (70.6)�1 40 (66.7) 35 (81.4) 5 (29.4)

Extent .15Localized 48 (80.0) 32 (74.4) 16 (94.1)Disseminated 12 (20.0) 11 (25.6) 1 (5.9)

Serum lactate dehydrogenase level .6Normal 54 (90.0) 38 (88.4) 16 (94.1)High 6 (10.0) 5 (11.6) 1 (5.9)

Immunohistochemistryc

Bcl-2 positive, No. of patients/total No. of patients 51/60 (85.0) 39/43 (90.7) 12/17 (70.6) .1MUM-1 positive, No. of patients/total No. of patients 23/41 (56.1) 18/29 (62.1) 5/12 (41.7) .3

Complete response .2Yes 41 (68.3) 27 (62.8) 14 (82.4)No 19 (31.7) 16 (37.2) 3 (17.6)

Relapsed .13No 15 (36.6) 8 (29.6) 7 (50.0)Yes 26 (63.4) 19 (70.4) 7 (50.0)

Extracutaneous progression .3No 34 (56.7) 23 (53.5) 11 (64.7)Yes, nodal only 9 (15.0) 6 (14.0) 3 (17.6)Yes, visceral with or without nodal 17 (28.3) 14 (32.6) 3 (17.6)

Status .001Alive, disease free 16 (26.7) 5 (11.6) 11 (64.7)Alive with disease 4 (6.7) 4 (9.3) 0Died of lymphoma 31 (51.7) 27 (62.8) 4 (23.5)Died, other cause 9 (15.0) 7 (16.3) 2 (11.8)

Disease-specific survival rate, % .0023 y 53 43 775 y 41 26 77

aData are given as number (percentage) unless otherwise indicated.bLeg vs nonleg primary cutaneous diffuse large B-cell lymphoma, leg type. For 5-year survival rates, P values were calculated from a comparison of survival

curves.cNot available in all patients.dConsidered only in 41 patients who achieved a complete response.



Among 41 patients (68.3%) who achieved at any timea complete response, 26 (63.4%) experienced 1 or sev-eral relapses. Of 60 patients, 26 (43.3%) developed ex-tracutaneous disease. The mean time until extracutane-ous dissemination was 20 months. The dissemination wasrestricted to the lymph nodes in 9 patients. The remain-ing 17 patients had visceral progression associated withlymph node involvement (6 patients) or without (11 pa-tients). The central nervous system was the most fre-quent site of visceral dissemination (7 patients). Othersites included the bones (2 patients), kidney (2 pa-tients), liver (1 patient), spleen (1 patient), testis (1 pa-tient), pancreas (1 patient), breast (1 patient), pelvis (1patient), and brachial plexus (1 patient).

Thirty-one patients (51.7%) died of lymphoma, and9 patients (15.0%) died of unrelated disease. Of 31 disease-specific deaths, 24 (77.4%) followed extracutaneous pro-gression of the disease. Eight other deaths were consid-ered disease-related despite the absence of obviousextracutaneous involvement. These included patients whodeveloped fatal sepsis after chemotherapy and those whodied of secondary infection, major cutaneous tumor bulkand ulceration, or major worsening of their general con-dition. The 3-year and 5-year disease-specific survival rateswere 53% and 41%, respectively.

PROGNOSTIC FACTORS

In univariate analysis, disease-related death was statis-tically significantly associated with location on the leg(P=.003), disseminated distribution (P=.04), and the pres-ence of multiple skin lesions at diagnosis (P=.004). Pa-

tients older than 75 years had a slightly poorer progno-sis (P=.06). Sex, B symptoms, Bcl-2 expression, MUM-1expression, performance status, serum lactate dehydroge-nase level, duration of lesions before diagnosis, and vari-ables related to therapy had no effect on survival. Multi-variate analysis of lymphoma-specific survival using allcandidate variables identified location on the leg (P=.02)and multiple skin lesions (P=.05) as independent factorsassociated with a poorer prognosis (Table 3). Similar re-sults were observed when age was forced in the model.

CHARACTERISTICS OF PATIENTSAND OUTCOMES IN DIFFERENT

PROGNOSTIC GROUPS

Because the topographic subtype (leg vs nonleg) was thestrongest prognostic factor, the main features at diagno-sis and the follow-up data were subsequently analyzedaccording to the location of skin lesions (Table 1). Pa-tients with the leg subtype were older and had more nu-merous skin lesions at diagnosis than patients with non-leg PCLBCL LT; they showed a tendency for less frequentresponse to therapy and more frequent relapses and vis-ceral progressions. The 3-year disease-specific survivalrates were 43% in the leg subtype group and 77% in thenonleg subtype group (P=.002). Kaplan-Meier lymphoma-specific survival curves according to the location of skinlesions are shown in the Figure.

The number of skin lesions at diagnosis also was a ma-jor distinctive factor for predicting survival in the entire

Table 2. Summary of Therapiesa

TherapyFirst-line Therapies

(n=60)Subsequent Therapies

(n=17)

Radiation therapy 23 (38.3)b 3Single-agent chemotherapy 1 (1.7) 2Polychemotherapies without anthracycline 8 (13.3)c 5Anthracycline-containing chemotherapies without rituximab 11 (18.3)d 7Various combinations of anthracycline-containing chemotherapies and rituximab 11 (18.3) 1Other 6 (10.0)e

aData are given as number (percentage).bAssociated with single-agent chemotherapy in 1 patient.cAssociated with radiation therapy in 1 patient.dAssociated with radiation therapy in 3 patients.eConsisted of surgery in 3 patients, oral corticosteroid use in 1 patient, and no therapy in 2 patients.

Table 3. Results From the Multivariate Analysis

Result

Relative Risk(95% Confidence

Interval)P

Value

Location of skin lesions .02Nonleg 1 [Reference]Leg 3.3 (1.1-10.0)

No. of lesions .05Single 1 [Reference]Multiple 2.3 (1.0-5.4)

1.00

0.50

0.75

0.25

00 12 24 36 48 60 72 84

Months

Surv

ival

Nonleg (n = 17)

P = .003

Leg (n = 43)

Figure. Kaplan-Meier lymphoma-specific survival curves according tolocation of skin lesions.



series, with 3-year disease-specific survival rates of 39%in patients with multiple skin lesions vs 77% in those witha single lesion (P=.004). The number of lesions was re-lated to the topographic subtype (Table 1). Only 8 of 43patients (18.6%) in the leg subtype group had a singlelesion at diagnosis. In this subgroup, the 3-year survivalrate was 78% compared with 32% in 35 patients with mul-tiple lesions involving 1 or both legs. No difference insurvival was observed between Bcl-2–positive cases (3-year survival, 54%) and Bcl-2–negative cases (3-year sur-vival, 43%) in the entire series (P=.8).

ROLE OF THERAPY

Although no variable related to therapy was statisticallysignificantly associated with survival in this retrospec-tive study, we further analyzed patients who received atany time various combinations of anthracycline-containing chemotherapies and rituximab (12 patients)and compared them with those who received other thera-pies only (48 patients). These 2 groups did not differ byage, location, or number of skin lesions. They stronglydiffered by period of inclusion because rituximab wasnever used before 2002 and was used only rarely before2004. Therefore, no long-term (�3 year) survival com-parison was available at the end point. The 2-year sur-vival rate differed between the 2 groups (group receiv-ing various combinations of anthracycline-containingchemotherapies and rituximab, 81%; group receivingother therapies, 59%) without reaching a statisticallysignificant difference (P=.3). However, the short-termoutcome was more favorable in patients treated with vari-ous combinations of anthracycline-containing chemo-therapies and rituximab. Among these 12 patients, all but1 (91.6%) achieved a complete response (compared with62% in the group receiving other therapies) (P=.05), and10 of 11 patients had no relapse. At the end point, 9 ofthese patients were alive with a mean follow-up of 19months (8 patients without disease and 1 patient withdisease), 2 patients had died of lymphoma (1 patient ofsepticemia after chemotherapy and 1 patient of brain in-volvement), and 1 patient had died of unrelated disease.A similar comparative analysis between patients treatedwith anthracycline-containing chemotherapies withoutrituximab and other patients failed to disclose any dif-ference in outcomes.

COMMENT

We report herein the largest study of PCLBCL LT, to ourknowledge, and provide new data on its clinicopatho-logic features and prognostic factors. In addition, we pro-vide preliminary results comparing outcomes in pa-tients treated with classic vs new therapeutic regimens.

Most PCBCLs, including almost all small-cell lym-phomas and a large proportion of large-cell lympho-mas, have an indolent clinical course.4-6,9,16-18 However,a subset of PCBCLs with a predominance of large cellscomprises aggressive lymphomas. This finding led re-searchers to look for discriminating prognostic factorswithin this heterogeneous group of primary cutaneous

large B-cell lymphomas. Small case series first identifiedthe location on the leg as a criterion of aggressive-ness.4,19 In further multicenter studies, round-cell mor-phological features20 (ie, the predominance of large cellswith round nuclei over large cells with cleaved nuclei)and Bcl-2 protein expression15 were identified as addi-tional adverse prognostic factors.

In the WHO-EORTC classification,1 the term primarycutaneous diffuse large B-cell lymphoma, leg type was in-troduced beside primary cutaneous follicle center cell lym-phoma and primary cutaneous marginal zone B-cell lym-phoma to designate PCBCLs with a predominance of largecells and a less favorable prognosis. Primary cutaneous dif-fuse large B-cell lymphoma, leg type was primarily de-fined on the basis of morphological features by the pres-ence of confluent sheets of large cells with round nuclei(ie, centroblasts and immunoblasts). Primary cutaneousdiffuse large B-cell lymphoma with a predominance of largecleaved cells and fewer centroblasts was classified withinthe group of primary cutaneous follicle center cell lym-phoma. In addition, it was specified that PCLBCL LT ariseson the leg in most (but not all) cases and demonstratesstrong Bcl-2 expression.1 The role of Bcl-2 expression forclassifying these lymphomas remained unclear. In fur-ther studies, some authors excluded rare Bcl-2–negativecases from the group of PCLBCL LT and included themin the heterogeneous group termed large B-cell lym-phoma, other,11 whereas other authors (as we do in the pres-ent study) included all cases with characteristic round-cell morphological features within the group of PCLBCLLT, irrespective of Bcl-2 expression.12

Using these criteria, we found that PCLBCL LT is char-acterized by a poor prognosis, a high proportion of Bcl-2expression (85%), an advanced age at onset (median age,77 years), and a frequent location on the leg (72% in 43cases, including 5 cases with lesions on the leg and at othersites). The 3-year and 5-year disease-specific survival ratesin the entire group were 53% and 41%, respectively.

Survival rates in the present study (Table 1) were lowerthan those in previous reports. The Italian study Groupfor Cutaneous Lymphomas12 recently reported a 5-year sur-vival rate of 73% in 51 patients with PCLBCL LT, includ-ing 6 patients without leg lesions. Kodama et al11 re-ported a 5-year survival rate of 61.7% in 40 Austrian patientswith Bcl-2–positive PCLBCL LT, 32 of whom had diseaseon the leg. Such differences in survival may result fromvariations in management and therapy or from differ-ences in baseline characteristics of patients resulting fromvarious selection biases. Although patients included in theAustrian study did not differ in age from our patients, thoseincluded in the Italian series were younger (median age,70 years). In addition, only patients who had 6 months offollow-up without extracutaneous dissemination were in-cluded in the Italian series. In our study, patients were in-cluded on the basis of negative initial staging, irrespec-tive of further outcome. Some older patients had multipleand bulky tumors on the legs, could receive only pallia-tive care, and died early after diagnosis. These patients werenot excluded from the study and affected the survival ratein the entire series.

We identified location on the leg as the main nega-tive prognostic factor. Patients with leg tumors had a



3-year disease-specific survival rate of 43%, compared with77% in patients with nonleg PCLBCL LT. This result con-firms and extends previous reports of the aggressive-ness of primary cutaneous large B-cell lymphoma of theleg, as defined in the previous EORTC classifica-tion.16,19,20 However, the prognosis of nonleg PCLBCL LT(Figure) seems poorer than that of primary cutaneousfollicle center lymphoma or primary cutaneous mar-ginal zone B-cell lymphoma, which are characterized by5-year survival rates of 95% to 100%.1,8 Therefore, it seemspreferable to classify these lymphomas on a morphologi-cal basis within the group of PCLBCL LT, as proposedin the recent WHO-EORTC classification,1 rather thanin the group of primary cutaneous follicle center cell lym-phoma, as in the previous EORTC classification.16

The second negative prognostic factor in our study wasthe presence of multiple skin lesions. Although the num-ber of skin lesions had no prognostic value in the Aus-trian series,11 2 previous studies15,20 on aggressive PCBCLsidentified multiple skin lesions as an important adverseprognostic factor. This is in accord with earlier stud-ies9,21,22 of different types of cutaneous or noncutaneouslymphomas that underlined the prognostic value of vari-ables related to tumor burden. In the present study, the3-year disease-specific survival rate was 39% in patientswith multiple skin lesions vs 77% in patients with only 1lesion (P=.004). Most patients with leg involvement hadmultiple tumors. However, those with only 1 tumor on aleg had a favorable prognosis. Whether these patients maybe first treated with less aggressive procedures such as ra-diation therapy alone remains questionable.

In the present study, PCLBCL LT showed intermedi-ate or positive staining for MUM-1 protein in most casesand consistently expressed Bcl-2 protein. This finding con-firms that the expression of MUM-1 and Bcl-2 character-izes not only primary cutaneous large B-cell lymphomasof the leg, as previously demonstrated,23-25 but also PCLBCLLT defined on a morphological basis. Bcl-2 expression wasalmost universal in cases located on the leg and was ob-served in most nonleg cases (Table 1). Overall, only 9 of60 cases had negative Bcl-2 staining. Bcl-2–negativePCLBCL LT is rare and has been poorly characterized todate. Kodama et al11 described 9 cases using the term largeB-cell lymphoma, other. Four cases were located outsideof the leg. No difference in survival was observed be-tween these 9 patients (5-year survival, 50%) and 40 pa-tients with typical Bcl-2–positive PCLBCL LT (5-year sur-vival, 61.7%). We found similar results, with 3-year survivalrates of 43% in Bcl-2–negative patients vs 54% in Bcl-2–positive patients (P=.8). Although these results need tobe confirmed in larger series, it seems suitable to includeall of these cases within the group of PCLBCL LT as de-fined morphologically by confluent sheets of large roundB-cells, irrespective of Bcl-2 staining.

Primary cutaneous diffuse large B-cell lymphoma, legtype is an aggressive lymphoma that requires effectivetherapies. This seems to be a difficult challenge in viewof the advanced age of many patients. Historically, someof these patients received only palliative care, radiationtherapy, or nonaggressive chemotherapies. However, thereport of the effectiveness of various combinations of an-thracycline-containing chemotherapies and rituximab in

older patients with noncutaneous diffuse large B-cell lym-phomas22 led to changes in the French practices regard-ing the cutaneous counterparts of these lymphomas. Therationale for this attitude was enhanced by the demon-stration that rituximab was able to overcome Bcl-2–associated resistance to chemotherapy,26,27 which coin-cided with findings that most PCLBCL LT stronglyexpressed Bcl-2.15,23 In the present study, 12 patients whoreceived various combinations of anthracycline-containing chemotherapies and rituximab were retro-spectively compared with 48 patients who received othertreatments. Although the follow-up was insufficient toobjectively determine a statistically significant differ-ence in survival, more favorable short-term outcomes wereobserved with various combinations of anthracycline-containing chemotherapies and rituximab. While await-ing appropriate controlled trials to confirm these obser-vations, we recommend that patients with PCLBCL LTshould be treated as often as possible with age-adaptedvarious combinations of anthracycline-containing che-motherapies and rituximab, particularly in patients withmultiple skin tumors.

Accepted for Publication: February 25, 2007.Author Affiliations: Department of Dermatology, Hopi-tal Robert Debre, Reims (Dr Grange); Departments of Der-matology (Dr Beylot-Barry) and Pathology (Dr Vergier),Hopital du Haut Leveque, Pessac; Departments of Der-matology (Dr Courville) and Pathology (Dr Joly), Hopi-tal Charles Nicolle, Rouen; Department of Dermatol-ogy, Hopital Bichat (Dr Maubec), and Department ofDermatology, Hopital Tarnier (Dr Avril), Paris; Depart-ments of Dermatology (Dr Bagot) and Pathology (DrWechsler), Hopital Henri-Mondor, Creteil; Departmentof Dermatology, Hotel Dieu, Clermont-Ferrand (DrSouteyrand); Department of Dermatology, Hopital Trous-seau, Tours (Dr Machet); Department of Dermatology (DrDalac) and Centre de Pathologie and Department of Pa-thology (Dr Petrella), Hopital du Bocage, Dijon; Depart-ment of Dermatology, Hopital Porte Madeleine, Orleans(Dr Esteve); Department of Dermatology, Hopital Mi-challon, Grenoble (Dr Templier); Department of Der-matology, Hopital Claude Huriez, Lille (Dr Delaporte);Department of Dermatology, Institut Gustave Roussy,Villejuif (Drs Avril and Robert); Department of Derma-tology, Hopital de l’Hotel-Dieu, Lyon (Dr Dalle); Depart-ment of Dermatology, Hopital Avicenne, Bobigny (DrLaroche); and Department of Dermatology and Cancer-ology, Hopital Pellegrin, Bordeaux (Dr Delaunay); France.Correspondence: Florent Grange, MD, PhD, Departmentof Dermatology, Hopital Robert Debre, Avenue du Gen-eral Koenig, Reims 51100, France ([email protected]).Author Contributions: Study concept and design: Grange.Acquisition of data: Grange, Beylot-Barry, Courville,Maubec, Bagot, Vergier, Souteyrand, Machet, Dalac,Esteve, Templier, Delaporte, Avril, Robert, Dalle, Laroche,Delaunay, Joly, Wechsler, and Petrella. Analysis and in-terpretation of data: Grange and Petrella. Drafting of themanuscript: Grange and Petrella. Critical revision of themanuscript for important intellectual content: Beylot-Barry, Courville, Maubec, Bagot, Vergier, Souteyrand,Machet, Dalac, Esteve, Templier, Delaporte, Avril, Robert,



Dalle, Laroche, Delaunay, Joly, and Wechsler. Statisti-cal analysis: Grange. Study supervision: Grange andPetrella.Financial Disclosure: None reported.

REFERENCES

1. Willemze R, Jaffe ES, Burg G, et al. WHO-EORTC classification for cutaneouslymphomas [published online ahead of print February 3, 2005]. Blood. 2005;105(10):3768-3785.

2. Burg G, Kempf W, Cozzio A, et al. WHO/EORTC classification of cutaneous lym-phomas 2005: histological and molecular aspects. J Cutan Pathol. 2005;32(10):647-674.

3. Slater DN. The new World Health Organization–European Organization for Re-search and Treatment of Cancer classification for cutaneous lymphomas: a prac-tical marriage of two giants. Br J Dermatol. 2005;153(5):874-880.

4. Willemze R, Meijer CJLM, Scheffer E, et al. Diffuse large cell lymphomas of fol-licle center cell origin presenting in the skin: a clinicopathologic and immuno-logic study of 16 patients. Am J Pathol. 1987;126(2):325-333.

5. Santucci M, Pimpinelli N, Arganini L. Primary cutaneous B-cell lymphoma: a uniquetype of low-grade lymphoma: clinicopathologic and immunologic study of 83cases. Cancer. 1991;67(9):2311-2326.

6. Pimpinelli N, Santucci M, Mori M, Vallecchi C, Giannotti B. Primary cutaneousB-cell lymphoma: a clinically homogeneous entity? J Am Acad Dermatol. 1997;37(6):1012-1016.

7. Rijlaarsdam JU, van der Putte SC, Berti E, et al. Cutaneous immunocytomas: aclinicopathologic study of 26 cases. Histopathology. 1993;23(2):117-125.

8. Hoefnagel JJ, Vermeer MH, Jansen PM, et al. Primary cutaneous marginal zoneB-cell lymphoma: clinical and therapeutic features in 50 cases. Arch Dermatol.2005;141(9):1139-1145.

9. Grange F, Hedelin G, Joly P, et al; The French Study Group on Cutaneous Lym-phomas. Prognostic factors in primary cutaneous lymphomas other than my-cosis fungoides and the Sezary syndrome. Blood. 1999;93(11):3637-3642.

10. Fink-Puches R, Zenahlik P, Back B, et al. Primary cutaneous lymphomas: appli-cability of current classification schemes (European Organization for Researchand Treatment of Cancer, World Health Organization) based on clinicopatho-logic features observed in a large group of patients. Blood. 2002;99(3):800-805.

11. Kodama K, Massone C, Chott A, Smolle J, Kerl H, Cerroni L. Primary cutaneouslarge B-cell lymphomas: clinicopathologic features, classification, and prognos-tic factors in a large series of patients. Blood. 2005;106(7):2491-2497.

12. Zinzani PL, Quaglino P, Pimpinelli N, et al; Italian Study Group for CutaneousLymphomas. Prognostic factors in primary cutaneous B-cell lymphoma: the Ital-

ian Study Group for Cutaneous Lymphomas [published online ahead of print Feb-ruary 21, 2006]. J Clin Oncol. 2006;24(9):1376-1382.

13. Cox DR. Regression models and life tables. J R Stat Soc B. 1972;34:187-202.14. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations.

J Am Stat Assoc. 1958;53:157-181.15. Grange F, Petrella T, Beylot-Barry M, et al. Bcl-2 protein expression is the stron-

gest independent prognostic factor of survival in primary cutaneous large B-celllymphomas. Blood. 2004;103(10):3662-3668.

16. Willemze R, Kerl H, Sterry W, et al. EORTC classification for primary cutaneouslymphomas: a proposal from the Cutaneous Lymphoma Study Group of the Eu-ropean Organization for Research and Treatment of Cancer. Blood. 1997;90(1):354-371.

17. Grange F, Bagot M. Prognosis of primary cutaneous lymphomas [in French]. AnnDermatol Venereol. 2002;129(1, pt 1):30-40.

18. Hembury TA, Lee B, Gascoyne RD, et al. Primary cutaneous diffuse large B-celllymphoma: a clinicopathologic study of 15 cases. Am J Clin Pathol. 2002;117(4):574-580.

19. Vermeer MH, Geelen FA, van Haselen CW, et al. Primary cutaneous large B-celllymphomas of the legs: a distinct type of cutaneous B-cell lymphoma with anintermediate prognosis. Arch Dermatol. 1996;132(11):1304-1308.

20. Grange F, Bekkenk MW, Wechsler J, et al. Prognostic factors in primary cuta-neous large B-cell lymphomas: a European Multicenter Study. J Clin Oncol. 2001;19(16):3602-3610.

21. Joly P, Vasseur E, Esteve E, et al; French Study Group for Cutaneous Lympho-mas. Primary cutaneous medium and large cell lymphomas other than mycosisfungoides: an immunohistological and follow-up study of 54 cases. Br J Dermatol.1995;132(4):506-512.

22. Coiffier B, Lepage E. Prognosis of aggressive lymphomas: a study of five prog-nostic models with patients included in the LNH-84 regimen. Blood. 1989;74(2):558-564.

23. Geelen FA, Vermeer MH, Meijer CJ, et al. Bcl-2 protein expression in primarycutaneous large B-cell lymphoma is site-related. J Clin Oncol. 1998;16(6):2080-2085.

24. Paulli M, Viglio A, Vivenza D, et al. Primary cutaneous large B-cell lymphomasof the leg: histogenetic analysis of a controversial clinicopathologic entity. HumPathol. 2002;33(9):937-943.

25. Hoefnagel JJ, Dijkman R, Basso K, et al. Distinct types of primary cutaneous largeB-cell lymphoma identified by gene expression profiling. Blood. 2005;105(9):3671-3678.

26. Coiffier B, Lepage E, Brière J, et al. CHOP chemotherapy plus rituximab com-pared with CHOP alone in elderly patients with diffuse large B-cell lymphoma.N Engl J Med. 2002;346(4):235-242.

27. Mounier N, Brière J, Gisselbrecht C, et al. Rituximab plus CHOP (R-CHOP) over-comes bcl-2–associated resistance to chemotherapy in elderly patients with dif-fuse large B-cell lymphoma. Blood. 2003;101(11):4279-4284.



OBSERVATION

Familial Acanthosis NigricansDue to K650T FGFR3 MutationDavid R. Berk, MD; Elaine B. Spector, PhD; Susan J. Bayliss, MD

Background: Acanthosis nigricans is a feature of sev-eral syndromes caused by activating mutations of the fi-broblast growth factor receptor 3 gene (FGFR3), includ-ing Crouzon syndrome with acanthosis nigricans,thanatophoric dysplasia, and severe achondroplasia withdevelopmental delay and acanthosis nigricans (SADDANsyndrome).

Observations: We describe a healthy 4-year-old Afri-can American girl with generalized acanthosis nigricanssince infancy. Her father had a history of acanthosis ni-gricans since childhood, in addition to Crohn disease,obesity, and adult-onset diabetes mellitus. A pedigree with

numerous affected family members was constructed.Other than slightly short stature, no associated anoma-lies were found, including dysmorphic features or skel-etal or neurologic defects. Genetic testing revealed a pre-viously undescribed, heterozygous lysine to threoninemutation at codon 650 of the FGFR3 gene in the 4 af-fected family members who were tested.

Conclusion: Extensive acanthosis nigricans in early child-hood, especially with a family history of acanthosis ni-gricans, may warrant testing for FGFR3 mutations.

Arch Dermatol. 2007;143(9):1153-1156

N UMEROUS GENETIC SYN-dromes are associatedwith acanthosis nigri-cans (AN), particularlysyndromes character-

ized by obesity, hyperinsulinemia, and/or craniosynostosis.1 In a recent review,Torley et al1 divided these genetic syn-dromes into insulin resistance syndromesand fibroblast growth factor (FGF) de-fects. Insulin resistance syndromes arecaused by mutations in the insulin recep-tor (leprechaunism and Rabson-Menden-hall syndrome), peroxisome proliferator-activated receptor-� (insulin-resistant

diabetes mellitus with AN and hyperten-sion), 1-acylglycerol-3-phophate O-acyl-transferase-2 (Berardinelli-Seip syn-drome),seipin(Berardinelli-Seipsyndrome),lamin A/C (Dunnigan syndrome), and Al-strom syndrome genes. Fibroblast growthfactor defects associated with AN includespecific activating mutations of FGFR2(Beare-Stevenson syndrome) and FGFR3(Crouzon syndrome with AN and thanato-phoric dysplasia [TD], and severe achon-droplasiawithdevelopmentaldelayandAN[SADDAN syndrome]).

We describe herein a family with AN as-sociated with a previously undescribed ly-sine to threonine mutation at codon 650 ofthe FGFR3 gene. All 4 affected family mem-bers who were tested demonstrated the mu-tation. The inheritance pattern was auto-somal dominant, and there were no obviousassociated skeletal or neurological abnor-malities other than short stature.

REPORT OF CASES

A 4-year-old African American girl was re-ferred for evaluation of generalized, mini-mally pruritic skin thickening since in-fancy. She was diagnosed as having atopicdermatitis by her pediatrician years earlierbut had not responded to treatment withtopical corticosteroids. She was otherwisehealthy without diabetes mellitus and wastaking no medications. Neurological devel-opment was normal. Her father (age, 25years) described having a similar skin con-dition since early childhood. He was slightlyobeseandhadCrohndiseaseandmildadult-onset diabetes mellitus (diet controlled).There was no history of consanguinity.

On physical examination, the patienthad generalized, thick, velvety, hyperpig-mented plaques that were most promi-nent on her neck, back, and axillae(Figure 1). There was no erythema or li-

See also pages 1125,1194, and 1209

Author Affiliations:Department of InternalMedicine, Division ofDermatology, and Departmentof Pediatrics, WashingtonUniversity School of Medicine,St Louis, Missouri (Drs Berkand Bayliss); and DNADiagnostic Laboratory andDepartment of Pediatrics,University of Colorado atDenver and Health and SciencesCenter, Aurora (Dr Spector).



chenification. Her oral mucosa, palms, and external geni-talia were normal. Height and weight at age 5.5 years were105.5 cm (eighth percentile) and 17.3 kg (20th percen-tile), respectively. Results from laboratory evaluations (com-plete blood cell count, urinalysis, and random glucose andhemoglobin A1c measurements) were normal. Examina-tion of the patient’s father revealed similarly thick, vel-vety, hyperpigmented plaques localized to the neck, back,and axillae. His height was 156 cm (� fifth percentile). Overthe past few years, his weight had fluctuated from 48 to106 kg due to his Crohn disease. Neither the patient norher father had facial dysmorphism (eg, frontal bossing) ortibial bowing. Familial AN was diagnosed.

Because activating FGFR3 mutations have been iden-tified in several forms of syndromic AN, this gene wasevaluated for mutations. Genetic testing was performedon DNA isolated from saliva (Oragene; DNA Genotek Inc,Ottawa, Ontario, Canada) followed by sequencing of ex-ons 10, 13, and 15 of the FGFR3 gene. A single nucleo-tide change was observed in exon 15 (Figure 2). Thepatient and her father both demonstrated a previouslyundescribed lysine to threonine mutation at codon 650(K650T) in one FGFR3 gene, suggesting autosomal domi-nant inheritance. No mutation was identified in the pa-tient’s mother or the patient’s 3-year-old sister, both ofwhom were clinically unaffected. Genetic counseling wasperformed. There was no change with topical pimecro-limus and steroids.

The father’s 29-year-old cousin (Figure3) and her 11-year-old daughter (Figure 4) were subsequently evalu-ated for the same condition and diagnosed as having fa-milial AN, resulting in a pedigree with multiple affectedrelatives (Figure 5). The daughter’s height and weightat age 11 years were 139 cm (26th percentile) and 40.9kg (68th percentile), respectively. Her mother’s height andweight were 145 cm (� fifth percentile) and 64.2 kg (25th-50th percentile), respectively. No further associated anoma-lies were identified in these 2 patients or reported in theirrelatives, including dysmorphic features, diabetes melli-tus, Crohn disease, or skeletal or neurologic defects. Boththis child and her mother demonstrated heterozygousK650T mutations on genetic testing, identical to their 2relatives who had been previously tested.

COMMENT

Acanthosis nigricans is characterized by velvety, hyper-pigmented plaques often accentuated in the flexures. Nu-merous forms of AN have been described including be-

Figure 1. Patient at age 4 years with thick, velvety, hyperpigmented plaqueinvolving the neck and chest.

A/CG GA A A AC C

G GAA A A A AC C

B

A

Figure 2. Sequencing of exon 15 of the FGFR3 gene was performed usingthe following primers: forward, 5� GTA AAA CGA CGG CCA GT AGG TGT GGGTGG AGT AGG; and reverse, 5� CAG GAA ACA GCT ATG ACC TC AGG CGCCAT CCA CTT 5� CTG TCA CCG TAG CCG TGA AG. Sequencing wasperformed using an ABI 3130 sequencer (Applied Biosystems, Foster City,California). A, The arrow indicates K650T mutation; A→C change atnucleotide 1950, amino acid 650. B, Normal sequence.



nign (nonsyndromic, insulin resistance-associated AN),malignant (paraneoplastic), obesity-associated (pseudo-AN), acral, medication-induced (corticosteroids, estro-gens, oral contraceptives, niacin, triazinate, somatotro-phin, and diethylstilbestrol), and syndromic AN(including isolated, “pure” familial AN).2,3

Generalized AN is rare and often paraneoplastic,4 evenin very early childhood. However, generalized AN in earlychildhood may also suggest isolated (nonsyndromic, non-insulin resistance–associated) familial AN.4-8 Further-more, there are reports of benign, generalized AN in earlychildhood without a positive family history, malig-nancy, associated syndrome, causative medication, or co-morbid condition.9-15

Only a few families with isolated (nonsyndromic, non-insulin resistance–associated) familial AN have been re-ported in the literature.4-8 Inheritance tends to be auto-somal dominant with variable penetrance. This “pure”familial AN usually appears in infancy, stabilizes at pu-berty, and is not associated with obesity or diabetes melli-tus. The term familial AN is confusing because AN as-sociated with insulin resistance also tends to be hereditaryand the term familial AN has been used in the literaturein such cases.16,17 Comorbid conditions have rarely beenreported with isolated familial AN. Chuang et al18 re-ported a family with AN and ectodermal defects includ-ing madarosis. In addition, 2 cases of benign AN (no fam-ily history) in early childhood have been associated withpyramidal tract degeneration.11,15 Familial AN may be con-fused with ichthyosis hystrix,5 generalized epidermal nevi,or lichenification in atopic dermatitis.

The pathogenesis of AN is poorly understood. In pa-tients with hyperinsulinemia, excess insulin may di-rectly or indirectly stimulate epidermal proliferationthrough the insulin-like growth factor 1 receptor.19

Growth factors (particularly transforming growth fac-tor-�) produced by tumors may cause malignant ANthrough the epidermal growth factor receptor.20

Fibroblast growth factors are involved in angiogen-esis, embryogenesis, mitogenesis, and wound healing.There are more than 20 human FGF ligands and 4 hu-man FGF receptors. The FGFR3 gene is located at 4p16.3and encodes a transmembrane receptor tyrosine kinase,which down-regulates long-bone growth. FGFR3 muta-tions have been reported in several dermatologic condi-tions including seborrheic keratoses,21 epidermal nevi,22

and syndromic AN. Several autosomal dominant syn-dromes are caused by specific FGFR3 mutations, includ-ing achondroplasia, Muenke syndrome, hypochondropla-sia, TD I and II, Crouzon syndrome with AN, and SADDANsyndrome. The latter 3 disorders demonstrate AN.

Codon 650 of FGFR3 is located in its tyrosine kinasedomain II. Mutations of this codon have been reportedin skeletal disorders including hypochondroplasia(K650N and K650Q), SADDAN syndrome (K650M), TD I(K650M), and TD II (K650E), as well as malignant con-ditions including bladder cancer and multiple myeloma(K650M, K650E, and K650T). Germline K650T muta-tions have never been reported. Bellus et al23 studied theeffects of various amino acid changes on the activationof FGFR3 in vitro. Constructs were made that containedall of the possible FGFR3 mutations at amino acid 650.

Figure 3. Cousin (age, 29 years) demonstrating thick, velvety,hyperpigmented plaques on the neck and posterior auricular region.

Figure 4. Cousin (age, 11 years) demonstrating thick, velvety,hyperpigmented plaques on the neck and forehead.



In a [32P]-ATP autophosphorylation assay, the relativeactivity of the mutations compared with that of a wild-type construct was as follows: K650R, 1.5x; K650N, 3.7x;K650Q, 4.9x; K650T, 3.1x; K650E, 9.6x; and K650M,18.1x. These data suggest that since the K650T muta-tion was a relatively weak activator of FGFR3, it wouldhave a less severe effect on phenotype than the muta-tions found in hypochondroplasia, SADDAN syn-drome, TD I, and TD II, as was observed in the patientsdescribed herein.

Activating FGFR3 mutations and/or ectopic overex-pression have been reported in a number of tumors, suchas bladder,24 cervical,25 multiple myeloma,26 and colo-rectal cancers.27 Tyrosine kinase inhibitors of FGFR3 arebeing tested in several hematologic malignancies withpositive results.26 Topical preparations of these inhibi-tors may prove beneficial for seborrheic keratoses, epi-dermal nevi, and/or AN.22

In summary, we describe a family with AN caused byan autosomal dominant K650T mutation in the FGFR3gene. Extensive AN in early childhood, especially witha family history of AN, may warrant testing for FGFR3mutations.

Accepted for Publication: March 27, 2007.Correspondence: David R. Berk, MD, Division of Der-matology, Washington University School of Medicine, 660S Euclid, Campus Box 8123, St Louis, MO 63110 ([email protected]).Author Contributions: Dr Berk had full access to all ofthe data in the study and take responsibility for the in-tegrity of the data and the accuracy of the data analysis.Study concept and design: Berk and Bayliss. Acquisition ofdata: Berk and Spector. Analysis and interpretation of data:Berk and Spector. Drafting of the manuscript: Berk andSpector. Critical revision of the manuscript for importantintellectual content: Berk, Spector, and Bayliss. Statisticalanalysis: Berk. Administrative, technical, and material sup-

port: Berk, Spector, and Bayliss. Study supervision: Berkand Spector.Financial Disclosure: None reported.

REFERENCES

1. Torley D, Bellus GA, Munro CS. Genes, growth factors and acanthosis nigricans.Br J Dermatol. 2002;147(6):1096-1101.

2. Curth HO. Classification of acanthosis nigricans. Int J Dermatol. 1976;15(8):592-593.

3. Curth HO. Acanthosis nigricans. Birth Defects Orig Artic Ser. 1971;7(8):31-39.4. Inamadar AC, Palit A. Generalized acanthosis nigricans in childhood. Pediatr

Dermatol. 2004;21(3):277-279.5. Curth HO, Aschner BM. Genetic studies on acanthosis nigricans. AMA Arch Derm.

1959;79(1):55-66.6. Curth HO. Benign acanthosis nigricans. Arch Dermatol. 1967;96(5):603-604.7. Tasjian D, Jarratt M. Familial acanthosis nigricans. Arch Dermatol. 1984;120(10):

1351-1354.8. Dhar S, Dawn G, Kanwar AJ, Nada R. Familial acanthosis nigricans. Int J Dermatol.

1996;35(2):126-127.9. Curth HO. Benign type of acanthosis nigricans. Arch Derm Syphilol. 1936;34(3):

353-366.10. Pindborg JJ, Gorlin RJ. Oral changes in acanthosis nigricans (juvenile type): sur-

vey of the literature and report of a case. Acta Derm Venereol. 1962;42:63-71.11. Maranon G, Alvarez Cascos M. Benign juvenile acanthosis nigricans; its relation

to congenital abnormalities. Acta Derm Venereol. 1957;37(3):249-258.12. Skiljevic DS, Nikolic MM, Jakovljevic A, Dobrosavljevic DD. Generalized acan-

thosis nigricans in early childhood. Pediatr Dermatol. 2001;18(3):213-216.13. Uyttendaele H, Koss T, Bagheri B, et al. Generalized acanthosis nigricans in an

otherwise healthy young child. Pediatr Dermatol. 2003;20(3):254-256.14. Rai VM, Balachandran C. Generalized acanthosis nigricans in childhood. Der-

matol Online J. 2006;12(6):14.15. Curth HO. Significance of acanthosis nigricans. AMA Arch Derm Syphilol. 1952;

66(1):80-100.16. Taylor G, James MP, Simpson H. Familial acanthosis nigricans. J R Soc Med.

1994;87(3):169.17. Friedman CI, Richards S, Kim MH. Familial acanthosis nigricans: a longitudinal

study. J Reprod Med. 1987;32(7):531-536.18. Chuang SD, Jee SH, Chiu HC, Chen JS, Lin JT. Familial acanthosis nigricans with

madarosis. Br J Dermatol. 1995;133(1):104-108.19. Cruz PD Jr, Hud JA Jr. Excess insulin binding to insulin-like growth factor re-

ceptors: proposed mechanism for acanthosis nigricans. J Invest Dermatol. 1992;98(6)(suppl):82S-85S.

20. Haase I, Hunzelmann N. Activation of epidermal growth factor receptor/ERK sig-naling correlates with suppressed differentiation in malignant acanthosis nigricans.J Invest Dermatol. 2002;118(5):891-893.

21. Logie A, Dunois-Larde C, Rosty C, et al. Activating mutations of the tyrosine ki-nase receptor FGFR3 are associated with benign skin tumors in mice and humans.Hum Mol Genet. 2005;14(9):1153-1160.

22. Hafner C, van Oers JM, Vogt T, et al. Mosaicism of activating FGFR3 mutationsin human skin causes epidermal nevi. J Clin Invest. 2006;116(8):2201-2207.

23. Bellus GA, Spector EB, Speiser PW, et al. Distinct missense mutations of the FGFR3lys650 codon modulate receptor kinase activation and the severity of the skel-etal dysplasia phenotype. Am J Hum Genet. 2000;67(6):1411-1421.

24. van Rhijn BW, van Tilborg AA, Lurkin I, et al. Novel fibroblast growth factor re-ceptor 3 (FGFR3) mutations in bladder cancer previously identified in non-lethalskeletal disorders. Eur J Hum Genet. 2002;10(12):819-824.

25. Cappellen D, De Oliveira C, Ricol D, et al. Frequent activating mutations of FGFR3in human bladder and cervix carcinomas. Nat Genet. 1999;23(1):18-20.

26. Grand EK, Chase AJ, Heath C, Rahemtulla A, Cross NC. Targeting FGFR3 in mul-tiple myeloma: inhibition of t(4;14)-positive cells by SU5402 and PD173074.Leukemia. 2004;18(5):962-966.

27. Jang JH, Shin KH, Park JG. Mutations in fibroblast growth factor receptor 2 andfibroblast growth factor receptor 3 genes associated with human gastric and co-lorectal cancers. Cancer Res. 2001;61(9):3541-3543.

∗∗ ∗

∗∗ ∗

Of an additional 11 siblings, 4 were affected

Figure 5. Pedigree of family. Shaded boxes indicate affected family members.The asterisks indicate family members tested for FGFR3 mutations.



OBSERVATION

Eczematoid Graft-vs-Host Disease

A Novel Form of Chronic Cutaneous Graft-vs-Host Diseaseand Its Response to Psoralen–UV-A Therapy

Daniel Creamer, MD; Claire L. Martyn-Simmons, MRCP(England); Genevieve Osborne, MRCP(England);Michelle Kenyon, MSc; Jon R. Salisbury, MD; Stephen Devereux, MD; Antonio Pagliuca, MD;Aloysius Y. Ho, MD; Ghulam J. Mufti, MD; Anthony W. P. du Vivier, MD

Background: Chronic cutaneous graft-vs-host disease(GVHD) is generally classified by whether lesions havea lichenoid or sclerodermatous morphology. Other un-usual clinical forms have been reported that exhibit thefeatures of dermatomyositis and lupus erythematosus.Within a large population of individuals who under-went allogeneic stem cell transplantation because of he-matologic malignancy, a group of patients was identi-fied in whom severe and persistent eczema developed.

Observations: We prospectively evaluated 10 adult pa-tients with unexplained eczematous dermatosis after al-logeneic hematopoietic stem cell transplantation. The der-matosis developed between 2 and 18 months (mean, 7.5months) after receipt of the transplant, exhibited the typi-cal clinical features of dermatitis, and became erythro-dermic in each case. The patient group had strong riskfactors for chronic cutaneous GVHD: 8 had received atransplant from an unrelated donor, 7 had evidence ofextracutaneous GVHD, and 7 had a history of acute cu-taneous GVHD. Sampling of lesional skin revealed the

histologic features of GVHD coexisting with the changesof dermatitis. The patients were treated with topical cor-ticosteroid and systemic immunosuppressive agents. Sixpatients also received psoralen–UV-A. Four patientsachieved prolonged remission. Six patients died, 5 of in-fective complications and 1 of relapsed leukemia.

Conclusions: The eczematous dermatosis observed rep-resents a novel form of chronic cutaneous GVHD thatwe named eczematoid GVHD. Eczematoid GVHD is anaggressive, chronic dermatosis that requires substantialimmunosuppression therapy to achieve control. It is as-sociated with a poor prognosis. Although atopy can betransmitted to an individual from a hematopoietic stemcell transplant, none of the donors in this series gave ahistory of an atopic disorder. Therefore, other factors mustbe implicated in provoking the expression of an eczema-tous phenotype in individuals with underlying chronicgraft-vs-host activity.

Arch Dermatol. 2007;143(9):1157-1162

G RAFT-VS-HOST DISEASE

(GVHD) is a multisys-tem disease initiated by al-logeneic T lymphocytesthat recognize foreign tis-

sue antigens in the host.1 The disease usu-ally develops after hematopoietic stem celltransplantation (HSCT), a therapeuticmethod used primarily to treat hemato-logic malignancy. However, it may also oc-cur after transfusion of nonirradiated bloodproducts, transplantation of solid organs,and maternal blood transfer in an immu-nodeficient fetus. Graft-vs-host disease re-sults in complications in 40% to 80% of al-logeneic HSCTs and is a major cause ofmorbidity and mortality.2

Graft-vs-host disease is divided intoacute and chronic forms that have distinctdisease patterns and are conventionally dif-ferentiated by whether onset is before or af-ter 100 days following transplantation.Acute GVHD follows a graft-vs-host (GVH)

reaction targeted against epithelia of skin,gastrointestinal tract, and liver and is mani-fested with rash and diarrhea and with ab-normal liver function test results. The erup-tion of acute GVHD is characterized bymaculopapular exanthem with acral accen-tuation or, rarely, is manifested as wide-spread epidermal necrolysis.3 Clinically,chronic cutaneous GVHD has been classi-fied by whether lesions have a scleroder-moid or lichenoid appearance.4 Scleroder-moid GVHD has many clinicopathologicpatterns including cases resembling mor-phea, lichen sclerosus, and eosinophilicfasciitis.5 The diffuse form of scleroder-moid GVHD is associated with deep-seated fibrosis and joint contractures. Li-chenoid GVHD is characterized by thepresence of violaceous, indurated papulesand plaques that resemble lichen planus.Other papulosquamous GVHD subtypeshave been reported including psoriasi-form, keratosis pilaris–like, and asteatotic

Author Affiliations:Departments of Dermatology(Drs Creamer, Martyn-Simmons, Osborne, and duVivier) and Histopathology(Dr Salisbury), King’s CollegeHospital; and Department ofHaematological Medicine,King’s College Hospital andKing’s College London(Ms Kenyon and Drs Devereux,Pagliuca, Ho, and Mufti),London, England.



forms.5 Rarely, GVHD variants occur that exhibit the fea-tures of autoimmune connective tissue diseases such asdermatomyositis and lupus erythematosus.6,7 Two re-ports cite eczemalike GVHD.8,9

We studied patients who developed cutaneous GVHDafter HSCT performed in the Department of Haemato-logical Medicine at King’s College Hospital, London, En-gland. Patterns of GVHD manifestation, natural history,response to treatment, and outcome were recorded forall patients. Within this population, a distinct group hasbeen identified characterized by persistent, widespread,chronic eczematous dermatosis occurring several monthsafter allogeneic HSCT. The eruption demonstrates the his-topathologic changes of both dermatitis and a GVH re-action, and we, therefore, suggest that it represents a mani-festation of chronic cutaneous GVHD that warrantsreporting and further investigation.

METHODS

PATIENTS

Between January 1, 2001, and December 31, 2004, 254 patientsunderwent allogeneic HSCT in the Department of Haematologi-cal Medicine at King’s College Hospital. Of these patients, 10 wereidentified who developed chronic eczematous dermatosis afterHSCT (eTable; available at: http://www.archdermatol.com). The

group consisted of 8 men and 2 women, with a mean age at on-set of dermatosis at 54.5 years (age range, 18-65 years). All pa-tients had undergone allogeneic HSCT for the treatment of a he-matologic malignancy (acute myeloid leukemia, 3 patients;chronic myeloid leukemia, 1; myelodyplastic syndrome, 2; chronicmyelomonocytic leukemia, 2; myelofibrosis, 1; and acute lym-phoblastic leukemia, 1). Dermatosis developed between 2 and18 months (mean, 7.5 months) after HSCT. One patient had ahistory of atopic dermatitis, 2 had a history of asthma and hay-fever, and 1 had a history of psoriasis. Drugs being taken by pa-tients at the onset of dermatosis included prednisolone cyclo-sporine, tacrolimus, phenoxymethyl-penicillin, fluconazole,amphotericin B, and acyclovir.

HEMATOPOIETIC STEMCELL TRANSPLANTATION

All 10 patients received an allogeneic stem cell transplant. Stemcells were derived from the peripheral blood in 7 patients andbone marrow in 3. In 8 patients, the transplant was from a vol-unteer unrelated donor; in 2, the donor was a sibling (eTable).There was a donor-recipient sex mismatch in 3 of 10 trans-plantations and a donor-recipient cytomegalovirus mismatchin another 3 transplantations (eTable). None of the donors re-ported a history of atopy (atopic dermatitis, asthma, or ec-zema). One donor reported a history of mild psoriasis (the re-cipient was patient 5). The conditioning regimen, describedpreviously,10 included alemtuzumab (Campath; Genzyme Corp,Cambridge, Massachusetts) in all patients but was otherwise

A B

Figure 1. Widespread, intensely pruritic, eczematous eruption in patient 7 (A) and patient 2 (B). During the course of the illness, all patients developederythroderma with thermoregulatory dysfunction and dependent edema (B).



of reduced intensity in 9 patients (fludarabine phosphate, 150mg/m2; busulfan, 8 mg/kg; and alemtuzumab, 100 mg) and ofstandard intensity in patient 7 (total body irradiation; cyclo-phosphamide; busulfan, 8 mg/kg; and alemtuzumab, 100 mg).

RISK FACTORS FOR GVHD

All 10 patients had received GVHD prophylaxis with cyclospor-ine and all transplants were T cell depleted in vivo using alem-tuzumab in the conditioning regimen. Matching for donor rela-tion, sex, and cytomegalovirus status is described above in“Hematopoietic Stem Cell Transplantation” subsection of the“Methods” section (eTable). Four patients had evidence of chronichepatic, gastrointestinal GVHD, or both, occurring concur-rently with the eczematous dermatosis. Seven patients were knownto have had preceding acute cutaneous GVHD (eTable), and, inthese patients, the eczematous dermatosis occurred 2 to 16 months(mean, 7.7 months) after acute GVHD. In 6 patients, the erup-tion appeared some months after acute GVHD (quiescent pat-tern), and in 1 patient (patient 3), it developed immediately af-ter the acute episode (progressive pattern). Three patients (patients4, 7, and 9) received donor lymphocyte infusions after HSCT(eTable). In these patients, the donor lymphocyte infusion wasadministered 13 weeks (patient 7) or 14 weeks (patients 4 and9) after HSCT and 2 to 6 weeks before the onset of eczematousdermatosis. All patients received standard GVHD prophylaxis withcyclosporine and prednisolone in the first few months after HSCT.In 5 patients, the eczematous dermatosis developed as prophy-lactic immunosuppression therapy was being tapered in the 2 to4 months after HSCT.

CLINICAL FEATURES

The eruption was similar in each patient and consisted of dif-fuse erythema and fine scaling, suggestive of eczema (Figure1).

Signs of impetiginization (colonization by Staphylococcus au-reus) were frequently observed. Weeping in involved skin wascommon. Vesicles or pustules were not seen. Pruritus was usu-ally severe and accompanied by excoriation. One patient hadmarked ichthyotic scaling, especially on the face and scalp(Figure 2). Palmoplantar hyperkeratosis was prominent in 7of 10 patients (Figure 2). In all patients, the eczema rapidly be-came widespread, leading to erythroderma (exfoliative derma-titis). Erythrodermic involvement was complicated by ther-moregulatory dysfunction, thirst, and dependent edema(Figure 1).

HISTOPATHOLOGIC FEATURES

Biopsy specimens were obtained from the affected skin in allpatients. In each case, the histologic features of GVHD coex-isted with the changes of dermatitis (Figure 3). The GVH re-action was indicated by satellite cell necrosis, and the pres-ence of parakeratosis, lymphocyte exocytosis, and epidermalspongiosis reflected the clinical appearance of eczema (Figure 3).The dermal changes were less marked, often showing a sparseperivascular lymphocytic infiltrate. Eosinophils were ad-mixed in the infiltrate in 4 patients. The combined features werenot always present in the first biopsy specimen, and, in somepatients, as many as 3 samples were necessary to indicate theeczematoid pattern of GVHD.

TREATMENT

First-line therapy for all patients was a regimen of an emol-lient, a potent topical corticosteroid ointment, and a sedatingantihistamine taken at bedtime. Oral antibiotics were given whenthe dermatosis was complicated by secondary bacterial infec-tion. Nine of 10 patients required hospital admission at leastonce during the course of their illness. Inpatient management

A B

Figure 2. Palmoplantar hyperkeratosis was present in 7 of 10 patients. A, Patient 4. B, One patient (patient 3) had marked ichthyotic scaling overlying the eczema.



was considered necessary when systemic symptoms arising fromthe erythroderma (eg, thirst, shivering, or malaise) were promi-nent or if first-line therapy was ineffective in controlling thedermatosis. As well as delivering maximal topical therapy, eryth-roderma was treated with systemic immunosuppression withone or more of the following: psoralen–UV-A (PUVA), pred-nisolone, methylprednisolone, azathioprine, cyclosporine, ta-crolimus, or mycophenolate mofetil (Table).

In an attempt to limit systemic immunosuppression, 6 pa-tients were treated with PUVA using oral psoralen (5-methoxypsoralen) and a UV-A dosimetry regimen similar tothat used to treat atopic dermatitis (Table). Psoralen–UV-A wasdelivered twice weekly starting at a dose of 0.5 J/cm2 and in-creasing the dose by 0.5 J/cm2 increments at each treatment toa maximum dose of 5 J/cm2. The duration of PUVA therapy was12 to 20 weeks.

RESULTS

The course of dermatosis was prolonged in all 10 pa-tients and did not tend to remit spontaneously. The short-est period of active involvement was 4 months, in 2 pa-tients; both patients died with active eczematoid GVHD4 months after the onset of dermatosis. The longest pe-riod of active involvement was 20 months (mean, 9.2months; Table). The shortest follow-up was 4 months,

and the longest was 42 months (mean, 15.5 months).Seven of the 10 patients also developed features of ex-tracutaneous chronic GVHD, with hepatic involvementin 5 patients, bronchiolitis obliterans in 2, and involve-ment of the gastrointestinal tract in 2 (Table).

In 2 patients, the dermatosis was controlled with topi-cal agents and systemic immunosuppressant agents (pred-nisolone and mycophenolate mofetil in patient 8 and pred-nisolone, tacrolimus, and mycophenolate mofetil inpatient 10). Four of the 6 patients treated with PUVA re-sponded successfully, achieving complete cutaneous clear-ance, and remission was maintained with only low-dosesystemic immunosuppression therapy (eg, predniso-lone, 1-5 mg/d) or no systemic immunosuppressiontherapy. In 2 of the patients who responded to PUVA (pa-tients 2 and 6), treatment was hampered by the pres-ence of secondary infection with methicillin-resistantS aureus, and the dermatosis resolved only after staphy-lococcal eradication.

Six patients died. The cause of death was consideredto be bacterial sepsis in 2 patients, disseminated Asper-gillus species infection in 2, and cryptococcosis in 1(Table); 1 patient had a relapse of acute myelocytic leu-kemia. Two patients (patients 4 and 7) who died of sep-tic complications had required prolonged exposure to highdoses of systemic immunosuppressant agents to controlthe cutaneous GVHD. Secondary skin infection in pa-tient 4 was probably the source of fatal overwhelmingsepsis.

Four patients1,5,8,10 have achieved long-term remis-sion (12-42 months; mean, 24 months) of the eczema-toid GVHD; 2 (patients 1 and 5) received PUVA and 2(patients 8 and 10) did not (Table). In these patients,the hematologic malignancy has also remained in re-mission.

COMMENT

We describe 10 patients in whom a widespread, chronicdermatitic eruption developed after allogeneic HSCT.The striking feature in all of our patients was the coex-istence of eczematous histopathologic features withchanges of GVHD. All of the patients had risk factorsfor the development of chronic cutaneous GVHD: 8 hadreceived a transplant from an unrelated donor, 7 hadevidence of extracutaneous GVHD, and at least 5 had ahistory of acute cutaneous GVHD. We believe there iscompelling evidence to suggest that the eczematousdermatosis observed in our patients represents a novelform of chronic cutaneous GVHD that we have namedeczematoid GVHD.

The differential diagnosis of a dermatitic eruptionoccurring after allogeneic HSCT should include the fol-lowing: an eczematous drug reaction, the precipitationof preexisting dermatitis, or the acquisition of atopicdiathesis from donor stem cells. Each potential cause ofeczema was considered in our patients. All of the pa-tients were receiving multiple drugs, but none of thesewere known to be associated with the induction of aneczematous drug eruption. Three patients had allergicdisorders and 1 had mild eczema during childhood;

A

B

Figure 3. A, Note parakeratosis, mild acanthosis, and an upper dermal andintraepidermal lymphocytic infiltrate (hematoxylin-eosin, originalmagnification �40). B, The dermatopathologic condition exhibits thecombined features of both eczema and a graft-vs-host reaction. There isspongiosis (short arrow) and satellite cell necrosis (long arrows)(hematoxylin-eosin, original magnification �100).



however, the remaining 7 patients had no history sug-gestive of an atopic susceptibility. None of the patientswere known to have another eczematous dermatosis(eg, seborrhoeic dermatitis or allergic contact dermati-tis) before HSCT. Although it is recognized that atopycan be transmitted to an individual from HSCT,11 noneof the donors in this series gave a history of an atopicdisorder. We, therefore, suggest that in these patients,eczema, observed both clinically and histopathologi-cally, was expressed as a manifestation of chronicGVHD. Although the pathophysiology of chronicGVHD remains poorly understood, it is recognized thatboth alloreactive and autoreactive T cells have a role inmediating tissue damage.12 Thymic injury from acuteGVHD may prevent the deletion of autoreactive clones,which can, therefore, promote a range of inflammatoryresponses.13 In murine chronic cutaneous GVHD, un-controlled collagen synthesis is driven by transforminggrowth factor–� generated by an inflammatory processwith a Th2-predominant cytokine profile.14 In many ec-zematous diseases, a Th2 response is also typical.15 Amodel to explain the development of an eczematousGVH response in our patients will, therefore, need toincorporate the complex interaction between antigen-presenting cells (including epidermal dendritic cells ofdonor origin) and autoreactive and allogenic T cellsstimulated in the presence of Th2 cytokines. Externalstimuli, most notably, bacterial antigens, will influencethe severity of eczematoid GVHD, as with other formsof eczema. Alemtuzumab, a monoclonal antibody thattargets CD52 on mature T lymphocytes, was adminis-tered to all patients as a part of the conditioning regi-men before HSCT. Alemtuzumab-induced T-cell deple-tion may be central to the development of eczematoidGVHD; however, larger studies are needed to explorethe relation between conditioning regimens and GVHreactions.

The development of GVHD after allogeneic HSCT haspotential clinical benefits because concomitant graft-vs-leukemia effects can control residual malignancy. In ourseries, only 1 patient had a relapse of the original hema-tologic malignancy. Nevertheless, the severity of this formof GVH reaction resulted in substantial morbidity andmortality. All of the patients with eczematoid GVHD de-

veloped erythroderma, which was poorly tolerated. Theseverity of the skin involvement necessitated, in mostcases, admission to the hospital for intensive topicaltherapy and treatment with systemic immunosuppres-sant agents. Six patients died; in 5, the cause of death wasconsidered to be an infective complication. Three pa-tients had systemic fungal infections, and 1 died of over-whelming bacterial sepsis. In one of these patients, sep-ticemia was probably a complication of intractable skininfection. The need to control the GVHD with systemicimmunosuppression therapy may have had a permis-sive effect in the progression to fatal sepsis. Problems en-countered with systemic immunosuppression therapyprompted the use of more effective skin-directed treat-ment. Six patients were treated with PUVA, and 4 of theseresponded well, achieving complete cutaneous clear-ance. Psoralen–UV-A is, therefore, recommended as a use-ful adjunctive and immunosuppression-sparing treat-ment in this form of GVHD.

Eczematoid GVHD is a severe, often erythrodermiceruption associated with considerable morbidity and mor-tality. While this dermatosis represents a complex man-agement problem for dermatologists treating patients withhematologic malignancies, the eczematous phenotypepoints to immunopathologic pathways previously unex-plored in the GVH reaction.

Accepted for Publication: May 3, 2007.Correspondence: Daniel Creamer, MD, Department ofDermatology, King’s College Hospital, Denmark Hill, Lon-don SE5 9RS, England ([email protected]).Author Contributions: Dr Creamer had full access to allthe data in the study and takes responsibility for the in-tegrity of the data and the accuracy of the data analysis.Study concept and design: Creamer, Salisbury, Mufti, anddu Vivier. Acquisition of data: Creamer, Martyn-Simmons, Osborne, Kenyon, Salisbury, Devereux, Pa-gliuca, Ho, Mufti, and du Vivier. Analysis and interpre-tation of data: Creamer, Martyn-Simmons, Osborne,Salisbury, Pagliuca, and Mufti. Drafting of the manu-script: Creamer, Salisbury, Mufti, and du Vivier. Criticalrevision of the manuscript for important intellectual con-tent: Creamer, Martyn-Simmons, Osborne, Kenyon, Sal-isbury, Devereux, Pagliuca, Ho, Mufti, and du Vivier. Sta-

Table. Treatment and Follow-up of Patients With Chronic Eczematoid GVHD

PatientNo. Treatment Response

Duration ofDisease, mo

Duration ofFollow-up, mo Outcome; Extracutaneous GVHD

1 PUVA, prednisolone, tacrolimus Good 20 24 Remission (skin and blood); no extracutaneous GVHD2 PUVA, prednisolone, tacrolimus Good 17 20 Died (relapse of AML); hepatic GVHD3 Prednisolone, cyclosporine Good 4 4 Died (cryptococcosis); hepatic GVHD4 PUVA, methylprednisolone, cyclosporine Poor 12 12 Died (bacterial sepsis); hepatic and GI tract GVHD5 PUVA Good 8 12 Remission (skin and blood); no extracutaneous GVHD6 PUVA, prednisolone, cyclosporine Good 6 7 Died (aspergillosis); hepatic GVHD7 PUVA, prednisolone, cyclosporine Poor 9 9 Died (aspergillosis); hepatic GVHD8 Prednisolone, MMF Good 6 42 Remission (skin and blood); bronchiolitis obliterans9 Prednisolone Good 4 4 Died (bacterial sepsis); GI tract GVHD

10 Prednisolone, tacrolimus, MMF Good 6 20 Remission (skin and blood); bronchiolitis obliterans

Abbreviations: AML, acute myeloid leukemia; GI, gastrointestinal; GVHD, graft-vs-host disease; MMF, mycophenolate mofetil; PUVA, psoralen–UV-A.



tistical analysis: Mufti. Administrative, technical, or materialsupport: Creamer, Martyn-Simmons, Osborne, Salis-bury, Devereux, and Mufti. Study supervision: Creamer,Salisbury, Ho, Mufti, and du Vivier.Financial Disclosure: None reported.Additional Information: The eTable is available at http://www.archdermatol.com.

REFERENCES

1. Billingham RE. The biology of graft-versus-host reactions. Harvey Lect. 1966-1967;62:21-78.

2. Ferrara JL, Deeg HJ. Graft-versus-host disease. N Engl J Med. 1991;324(10):667-674.

3. Johnson ML, Farmer ER. Graft-versus-host reactions in dermatology. J Am AcadDermatol. 1998;38(3):369-396.

4. Hood AF, Soter NA, Rappeport J, Gigli I. Graft-versus-host reaction: cutaneousmanifestations following bone marrow transplantation. Arch Dermatol. 1977;113(8):1087-1091.

5. Hymes SR, Turner ML, Champlin RE, Couriel DR. Cutaneous manifestations ofchronic graft-versus-host disease. Biol Blood Marrow Transplant. 2006;12(11):1101-1113.

6. Ollivier I, Wolkenstein P, Gherardi R, et al. Dermatomyositis-like graft-versus-host disease. Br J Dermatol. 1998;138(3):558-559.

7. Girardi M, Herreid P, Tigelaar R. Specific suppression of lupus-like graft-versus-

host disease using extracorporeal photochemical attenuation of effectorlymphocytes. J Invest Dermatol. 1995;104(2):177-182.

8. Tanasescu S, Balguerie X, Thomine E, et al. Eczema-like cutaneous graft versushost disease treated by UV-B therapy in a 2-year-old child [in French]. AnnDermatol Venereol. 1999;126(1):51-53.

9. Sloane JP, Thomas JA, Imrie SF, Easton DF, Powles RL. Morphological and im-munohistological changes in the skin in allogeneic bone marrow recipients.J Clin Pathol. 1984;37(8):919-930.

10. Ho AY, Pagliuca A, Kenyon M, et al. Reduced-intensity allogeneic hematopoieticstem cell transplantation for myelodysplastic syndrome and acute myeloid leu-kemia with multilineage dysplasia using fludarabine, busulphan, and alemtu-zumab (FBC) conditioning [published online ahead of print April 1, 2004]. Blood.2004;104(6):1616-1623. doi:10.1182/blood-2003-12-4207.

11. Bellou A, Kanny G, Fremont S, Moneret-Vautrin DA. Transfer of atopy followingbone marrow transplantation. Ann Allergy Asthma Immunol. 1997;78(5):513-516.

12. Sullivan KM, Parkman R. The pathophysiology and treatment of graft-versus-host disease. Clin Haematol. 1983;12(3):775-789.

13. Weinberg K, Blazar BR, Wagner JE, et al. Factors affecting thymic function afterallogeneic hematopoietic stem cell transplantation. Blood. 2001;97(5):1458-1466.

14. McCormick LL, Zhang Y, Tootell E, Gilliam AC. Anti-TGF-beta treatment pre-vents skin and lung fibrosis in murine sclerodermatous graft-versus-host dis-ease: a model for human scleroderma. J Immunol. 1999;163(10):5693-5699.

15. van Reijsen FC, Bruijnzeel-Koomen CA, Kalthoff FS, et al. Skin-derived aeroallergen-specific T-cell clones of Th2 phenotype in patients with atopic dermatitis. J Al-lergy Clin Immunol. 1992;90(2):184-193.

Announcement

Manuscript Submission

v Before preparing a manuscript authors shouldreview the Instructions for Authors available at http://www.archdermatol.com.

v Manuscripts are submitted to all sections ofthe Archives by Web access at http://manuscripts.archdermatol.com.

v Authors may check the status of their manuscriptas it proceeds through the review and decision processby logging into http://manuscripts.archdermatol.com.

v It is important for authors to update their contactinformation, especially their e-mail addresses, by log-ging into http://manuscripts.archdermatol.com. Publi-cation of accepted manuscripts may be delayed by ourinability to locate authors, who need to approve copy-edited manuscript proofs.



WEB-ONLY CONTENT

eTable. Chronic Eczematoid GVHD: Patient and Donor Characteristicsa

Patient No./Sex/Age, y Diagnosis Donor Type

Donor Sex/Age, y

CMVStatus R/D

AcuteGVHD DLI

Onset of Eczematoid GVHDAfter Transplantion, mo

1/M/54 MDS/RAEB Unrelated M/26 −/− Yes No 182/M/65 MDS/AML Unrelated M/44 −/− No No 23/M/46 Myelofibrosis Unrelated M/25 �/� Yes No 34/M/62 CML Sibling M/70 �/− Yes Yes 55/F/56 AML Unrelated M/37 −/� Yes No 106/F/64 MDS Unrelated F/34 −/− No No 47/M/18 ALL Unrelated M/40 �/− No Yes 48/M/54 CMML Unrelated M/43 �/� Yes No 69/M/63 CML/AML Unrelated F/26 −/− No Yes 4

10/M/62 CMML Sibling F/63 −/− Yes No 18

Abbreviations: ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; CML, chronic myeloid leukemia; CMML, chronic myelomonocytic leukemia;CMV, cytomegalovirus; DLI, donor lymphocyte infusion; GVHD, graft-vs-host disease; MDS, myelodysplastic syndrome; RAEB, refractory anemia with excessblasts; R/D, recipient/donor; −, negative; �, positive.

aAll patients underwent a conditioning regimen with alemtuzumab (Campath; Genzyme Corp, Cambridge, Massachusetts) and received GVHD prophylaxis.

(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), SEP 2007 WWW.ARCHDERMATOL.COME1


OBSERVATION

Epidermolysis Bullosa Nevus

An Exception to the Clinical and Dermoscopic Criteria for Melanoma

Sarah H. Cash, MD; Tara T. Dever, MD; Patrice Hyde, MD; Jason B. Lee, MD

Background: Large acquired melanocytic nevi that oc-cur in patients with epidermolysis bullosa (EB), re-ferred to as EB nevi, may pose a diagnostic challenge be-cause of their clinical and dermoscopic resemblance tomelanoma. These unconventional melanocytic nevi havebeen encountered in all categories of hereditary EB, mostof them in childhood. Although some of the reported caseshave an alarming clinical appearance that is indistin-guishable from melanoma, long-term follow-up has con-firmed the benign nature of these rarely encountered me-lanocytic lesions. The histopathologic patterns of thesenevi range from a banal congenital pattern to the prob-lematic persistent pseudomelanoma pattern.

Observation: We describe the clinical, dermoscopic,and histopathologic features of a large EB nevus in a tod-

dler. Clinically, the lesion was markedly asymmetricaland irregularly pigmented with foci of stippled pigmen-tation and scarring, which easily fulfilled the ABCD cri-teria for melanoma. Accordingly, a false-positive scoreresulted when dermoscopy was performed. Histopatho-logically, a pattern of persistent melanocytic neoplasmwas observed. In the following 18 months, dynamicchanges of the lesion included near-complete disappear-ance of the pigment, which was replaced by scar, milia,and areas of healing ulcers.

Conclusion: Epidermolysis bullosa nevi are dynamic me-lanocytic lesions that may simulate melanoma.

Arch Dermatol. 2007;143(9):1164-1167

R ECENTLY, BAUER ET AL1 PRO-posed the term epidermoly-sis bullosa (EB) nevi forlarge acquired melano-cytic nevi that are encoun-

tered in all forms of hereditary EB. Theirclinical importance lies in the fact that theymay simulate melanoma clinically and der-moscopically. In that 20-year prospec-tive study of patients with EB nevi, how-ever, no melanoma arose in associationwith these melanocytic nevi. We herein de-scribe the clinical, dermoscopic, and his-topathologic features of a large EB nevusin a toddler. After 18 months of follow-up, the pigment almost completely disap-peared, with replacement by scar, milia,and areas of healing ulcers.

REPORT OF A CASE

A 3-year-old Korean boy with a knownmutation for recessive dystrophic EB wasnoted on a routine follow-up visit to havea large pigmented lesion on the right thigh.The mother reported a rapid develop-ment of the pigmented lesion during a pe-

riod of several months that coincided withrepeated blistering at the site. Physical ex-amination revealed a large (9�10-cm),markedly asymmetric, irregularly pig-mented patch on the right lateral thigh(Figure 1). In addition, areas of scar-ring with stippled pigmentation and re-gressive changes (ie, hypopigmentationand scarring) were found within the le-sion (Figure 1). Results of the examina-tion revealed no palpable lymph nodes.Dermoscopic findings included an atypi-cal network and irregular dots and glob-ules with colors that ranged from blue tovarying shades of brown and black(Figure 2). Examination of a shave bi-opsy specimen from an area with mul-tiple colors revealed a cleft at the dermo-epidermal junction consistent with thepatient’s known recessive dystrophic EBand increased pigment within the epider-mis and dermis (Figure 3). Within theepidermis, there was an uneven distribu-tion of solitary melanocytes and a mild de-gree of scatter (ie, pagetoid spread) ac-companied by the presence of melanin atall levels of the epidermis, including the

Author Affiliations:Department of Dermatologyand Cutaneous Biology, ThomasJefferson University Hospital,Philadelphia, Pennsylvania(Drs Cash, Hyde, and Lee); andNational Naval Medical Center,Bethesda, Maryland (Dr Dever).Dr Cash is now in privatepractice in Dermatology Groupof the Carolinas, Concord,North Carolina.



cornified layer. Within the dermis, there were discretelynested, heavily melanized monomorphous melanocytesenveloped in fibrosis accompanied by numerous mela-nophages (Figure 4). No mitotic figures were ob-served. Immunohistochemical staining for Ki-67 showedno increase in the proliferative index. The histopatho-logic pattern was consistent with a persistent melano-cytic neoplasm and, hence, the differential diagnosis wasbetween a persistent melanocytic nevus and persistentmelanoma. The monomorphous melanocytes, lack of any

mitotic figures, normal proliferative index, and knowl-edge of the EB nevus phenomenon eventually led to thediagnosis of EB nevus. Because of the reported benignclinical course of these nevi, no therapeutic interven-tion was offered, only close clinical follow-up. To our sur-prise, during 18 months of follow-up, there was com-plete regression of epidermal pigment. What remainedwere areas of healing ulcers, hypopigmented scars, anda subtle hue of dermal pigmentation in some foci(Figure 5). Whether or not the melanocytic lesion ac-tually regressed completely was not confirmed by histo-pathologic evaluation. At 24 months, the patient was well,and development of additional melanocytic lesions wasnot observed.

COMMENT

Although rarely encountered, EB nevi may not be so rareamong patients with EB. Bauer et al1 reported a preva-lence rate of 14% (12 of 86) among the patients in theirEB registry. The lesions are usually encountered in chil-dren with all major categories of hereditary EB. The meanages of patients in reported cases of recessive dystro-phic EB and EB simplex are 7 and 11 years, respec-tively.1-5 No predilection for sex or site appears to exist,except that the lesions are encountered at the sites of re-peated blisters. They are often eruptive in nature, ie, theyappear rapidly and enlarge within a few weeks to a few

Figure 1. A large pigmented lesion with an irregular border, colorvariegation, and areas of regression, scarring, and stippled pigmentation(arrow).

Figure 2. Dermoscopic image (original magnification �10) shows irregulardots and globules, blue-gray pigment, and hypopigmented areas.

Figure 3. A shave biopsy specimen shows a cleft between thedermoepidermal junction, effaced rete ridges, and pigmented nestedmelanocytes in the superficial dermis (hematoxylin-eosin, originalmagnification �40).

100 µm

Figure 4. Higher-power magnification of the specimen from Figure 3 showsfibrosis of the dermis, the presence of numerous melanophages, and nestedmonomorphous melanocytes (hematoxylin-eosin, original magnification�200).



months. The reported lesions have ranged from 3 to 15cm in diameter with an appearance that ranged from con-genital melanocytic nevus to melanoma. The alarmingclinical features included not only their large size but alsothe presence of marked asymmetry, irregular and/orstippled pigmentation, variegation in color, scarring, andfoci of regression. On a morphologic basis alone, with-out the benefit of knowing the clinical context, the EBnevus may be indistinguishable from melanoma at times.

Although Gallardo et al5 observed no melanoma-associated dermoscopic features in their series of EB neviin a single patient, Lanschuetzer et al6 observed melanoma-associated dermoscopic features that included a multi-component pattern, an atypical pigment network, and ir-regular dots and globules in many of the patients in theirEB registry. Even the milky red areas highly character-istic of melanoma were present in several of their nevi.When dermoscopic algorithms were performed, a false-positive score resulted in most of the lesions in their se-ries. According to the authors,6 these apparently alarm-ing clinical and dermoscopic findings can be explainedby the known pathogenesis of EB nevi, ie, the repeateddisruption of the dermoepidermal junction, fibrosing in-flammation, scar formation, and neovascularizationaccount for their clinical and dermoscopic features.Marghoob and Kopf7 have previously reported that a per-sistent melanocytic nevus frequently yields a false-positive dermoscopic score. It should be no surprise thenthat asymmetrical and irregularly pigmented EB nevi witha histopathologic pattern of persistent melanocytic neo-plasm yield a false-positive dermoscopic score.

The histopathologic pattern of EB nevus ranges fromthe readily recognizable congenital pattern to a prob-lematic persistent/pseudomelanoma pattern.1-5 Most per-sistent patterns occur in patients with recessive dystro-phic EB, whereas the congenital pattern occurs morefrequently in patients with EB simplex. Obviously, thefibrosing inflammation that is more frequently encoun-tered in recessive dystrophic EB accounts for this obser-vation. Some of the large lesions, even those with an omi-nous clinical appearance, display a banal compoundcongenital pattern, ie, monomorphous melanocytes in

nests and fascicles spanning the papillary and reticulardermis with angiotropism and adnexotropism.1,3,8 Otherpatients, like ours, have shown the persistent pattern, ie,uneven distribution of melanocytes within the epider-mis and dermal melanocytes enveloped in fibrosis. Thepersistent pattern raises the differential diagnosis of per-sistent melanoma vs persistent melanocytic nevus. Di-agnosing persistent nevus becomes problematic if sec-tions from the previous procedure are not available and/orif only the portion of the persistent melanocytic neo-plasm is biopsied, which may lack the readily recogniz-able conventional benign melanocytic pattern. In casesof large EB nevi, in which the entire lesion often cannotbe sampled, the problem of the persistent pattern maynot be resolvable on the basis of histologic sections alonebecause there are no previous sections to review and theentire lesion may not be available for review. Our pa-tient represented such a dilemma. The clues that the le-sion on our patient was benign included melanocytes inthe dermis that were monomorphous, a lack of mitosis,and, accordingly, a normal proliferative index. In addi-tion, the lesion in our patient had a striking clinical re-semblance to a large superficial spreading melanoma,but this clinical and histopathologic pattern of mela-noma, although common in adults, is exceedingly rarein children. Melanomas in children are usually papularor nodular and thus have a prominent vertical orienta-tion clinically and histopathologically rather than thehorizontal orientation that is typical of the superficialspreading type of melanoma.9,10 The knowledge of theEB nevus phenomenon and the histopathologic featureseventually convinced us of the benign nature of the le-sion on our patient.

Although it is obvious that the recurrent blisters andthe subsequent inflammation and/or fibrosis alter the mor-phologic features of melanocytic nevi in patients with EB,the exact mechanism by which the recurrent blistersachieve the alteration is unknown. Speculations include(1) induction by the Koebner phenomenon, meaning thatdisruption of melanocytes at the site of blisters some-how induces proliferation during reepitheliazation,2 and(2) seeding of a blister cavity by melanocytes from a pre-existing nevus, which then proliferates owing to cyto-kines that are released by the inflammatory process causedby the blisters.8 Irrespective of the mechanism, the re-peated blisters at the site of the nevi result in a dynamicgrowth pattern that may lead to clinical lesions, whichmay be asymmetrical and multicolored and have regres-sive changes due to chronic fibrosing inflammation ratherthan to specific immune-mediated process mountedagainst melanocytes.

An EB nevus represents a collision between an inflam-matory disease and a melanocytic nevus that results in un-conventional phenotypic expression of the nevus. Colli-sion between lichen sclerosus et atrophicus and melanocyticnevus represents an analogous phenomenon.11,12 These col-lision phenomena have led to diagnostic difficulties for theclinicians and histopathologists because the expected clini-cal and histopathologic findings of the melanocytic le-sions are significantly altered. Similar to the EB nevus, thepersistent melanocytic pattern is observed histopathologi-cally in cases of collision with lichen sclerosus.11 The well-

Figure 5. The lesion from Figure 1 after 18 months of follow-up showsregression of epidermal pigment.



circumscribed nature of these lesions allows for correctinterpretation in most instances. In cases of a large EB ne-vus, however, circumscription often cannot be assessedbecause only a portion of the lesion is available for inter-pretation. This adds to the diagnostic difficulty com-pared with the cases of lichen sclerosus. Previously re-ported cases of eruptive melanocytic nevi that appearedshortly after episodes of Stevens-Johnson syndrome andbullous erythema multiforme have been suggested as analo-gous phenomena.13-16 In these instances, there is no col-lision, per se; rather, the 2 phenomena occur metachro-nously rather than synchronously. Accordingly, diagnosticdifficulties are not reported because there are no signifi-cant clinical and histopathologic alterations of the nevi’smorphologic features. Curiously, analogous phenomenahave not been reported with any frequency in other blis-tering diseases known to occur in childhood, such as lin-ear IgA dermatosis. In 1 reported case of an 8-year-old girlwith bullous pemphigoid of the vulva, a melanocytic le-sion at the site of the blisters eventually regressed com-pletely. As in our patient, the lesion resembled mela-noma clinically and dermoscopically.17

Although the association between squamous cell car-cinoma and chronic scarring inflammatory dermatosisis well known, it is unclear whether such an associationexists for melanoma in general. Data suggest that a higherincidence of melanoma may exist in patients with EB,18-20

especially recessive dystrophic EB, but there are no de-tailed, well-documented cases in the literature, in con-trast to the well-documented and now well-known higherincidence of squamous cell carcinoma in patients withEB. Furthermore, the higher incidence has not been cor-roborated by experience in other countries. Because EBnevi may simulate melanoma, they have the potential tofalsely increase the incidence and prevalence of mela-noma among patients with EB if they are misconstruedas such. Clinicians and histopathologists should be awareof this phenomenon to avoid misdiagnosis and unnec-essary therapeutic intervention. An EB nevus should beconsidered a distinct diagnostic possibility and shouldbe excluded if melanoma is suspected in a child with EB.Once it has been established clinically, dermoscopi-cally, and histopathologically that a lesion is an EB ne-vus, close clinical follow-up of these unconventional me-lanocytic nevi is an acceptable option, especially for nevithat are not amenable to simple excision owing to theirlarge size.

Accepted for Publication: February 20, 2007.Correspondence: Jason B. Lee, MD, Department of Der-matology and Cutaneous Biology, Thomas Jefferson Uni-versity Hospital, 833 Chestnut St, Ste 704, Philadelphia,PA 19107 ([email protected]).Author Contributions: Dr Lee had full access to all thedata in the study and takes responsibility for the integ-

rity of the data. Study concept and design: Lee. Acquisi-tion of data: Cash and Hyde. Analysis and interpretationof data: Cash, Dever, and Lee. Drafting of the manuscript:Cash, Dever, Hyde, and Lee. Critical revision of the manu-script for important intellectual content: Dever, Hyde, andLee. Administrative, technical, and material support: Cashand Hyde. Study supervision: Cash, Hyde, and Lee.Financial Disclosure: None reported.

REFERENCES

1. Bauer JW, Schaeppi H, Kaserer C, Hantich B, Hintner H. Large melanocytic neviin hereditary epidermolysis bullosa. J Am Acad Dermatol. 2001;44(4):577-584.

2. Soltani K, Pepper MC, Simjee S, Apatoff BR. Large acquired nevocytic nevus in-duced by the Koebner phenomenon. J Cutan Pathol. 1984;11(4):296-299.

3. Hoss DM, McNutt NS, Carter DM, Rothaus KO, Kenet BJ, Lin AN. Atypical me-lanocytic lesions in epidermolysis bullosa. J Cutan Pathol. 1994;21(2):164-169.

4. Stavrianeas NG, Katoulis AC, Moussatou V, et al. Eruptive large melanocytic ne-vus in a patient with hereditary epidermolysis bullosa simplex. Dermatology. 2003;207(4):402-404.

5. Gallardo F, Toll A, Malvehy J, et al. Large atypical melanocytic nevi in recessivedystrophic epidermolysis bullosa: clinicopathological, ultrastructural, and der-moscopic study. Pediatr Dermatol. 2005;22(4):338-343.

6. Lanschuetzer CM, Emberger M, Laimer M, et al. Epidermolysis bullosa naevi re-veal distinctive dermoscopic pattern. Br J Dermatol. 2005;153(1):97-102.

7. Marghoob AA, Kopf AW. Persistent nevus: an exception to the ABCD rule ofdermoscopy. J Am Acad Dermatol. 1997;36(3, pt 1):474-475.

8. Lanschuetzer CM, Emberger M, Hametner R, et al. Pathogenic mechanisms inepidermolysis bullosa naevi. Acta Derm Venereol. 2003;83(5):332-337.

9. Ferrari A, Bono A, Baldi M, et al. Does melanoma behave differently in youngerchildren than in adults? a retrospective study of 33 cases of childhood mela-noma from a single institution. Pediatrics. 2005;115(3):649-654.

10. Mones JM, Ackerman AB. Melanomas in prepubescent children: review com-prehensively, critique historically, criteria diagnostically, and course biologically.Am J Dermatopathol. 2003;25(3):223-238.

11. Carlson JA, Mu XC, Slominski A, et al. Melanocytic proliferations associated withlichen sclerosus. Arch Dermatol. 2002;138(1):77-87.

12. El Shabrawi-Caelen L, Soyer HP, Schaeppi H, et al. Genital lentigines and mela-nocytic nevi with superimposed lichen sclerosus: a diagnostic challenge. J AmAcad Dermatol. 2004;50(5):690-694.

13. Shoji T, Cockerell CJ, Koff AB, Bhawan J. Eruptive melanocytic nevi after Stevens-Johnson syndrome. J Am Acad Dermatol. 1997;37(2, pt 2):337-339.

14. Kirby JD, Darley CR. Eruptive melanocytic naevi following severe bullous disease.Br J Dermatol. 1978;99(5):575-580.

15. Burns DA, Sarkany I. Junctional naevi following toxic epidermal necrolysis. ClinExp Dermatol. 1978;3(3):323-326.

16. Kopf AW, Grupper C, Baer RL, Mitchell JC. Eruptive nevocytic nevi after severebullous disease. Arch Dermatol. 1977;113(8):1080-1084.

17. Lebeau S, Braun RP, Masouye I, Perrinaud A, Harms M, Borradori L. Acquiredmelanocytic naevus in childhood vulval pemphigoid. Dermatology. 2006;213(2):159-162.

18. Chorny JA, Shroyer KR, Golitz LE. Malignant melanoma and a squamous cellcarcinoma in recessive dystrophic epidermolysis bullosa. Arch Dermatol. 1993;129(9):1212.

19. Fine JD, Johnson LB, Tien H, et al. Skin cancer and inherited epidermolysis bul-losa (EB): analysis of the National EB Registry cohort by disease type and sub-type [abstract 12]. J Invest Dermatol. 1994;103(6):846.

20. Fine JD, Johnson LB, Tien H, et al. Risk of skin cancers and inherited epider-molysis bullosa (EB): determination of differences across major EB subtypes,as assessed by lifetable analysis techniques [abstract 21]. J Invest Dermatol.1994;103(6):848.



OBSERVATION

Herpes Gestationis in a Mother and Newborn

Immunoclinical Perspectives Based on a Weekly Follow-upof the Enzyme-Linked Immunosorbent Assay Indexof a Bullous Pemphigoid Antigen Noncollagenous Domain

Yumi Aoyama, MD; Kanako Asai, MD; Kana Hioki, MD; Michinori Funato, MD;Naomi Kondo, MD; Yasuo Kitajima, MD

Background: Herpes gestationis (HG) is a rare, autoim-mune, bullous disease that occurs during the second orthird trimester and usually resolves over weeks or monthsafter delivery. Neonates with HG are rare (estimated at 1per 100 000 cases). Although anti–180-kDa bullous pem-phigoid (BP180) autoantibody and transfer of this auto-antibody are known as the cause, to our knowledge, nocoordinated analysis of clinical symptoms and anti-BP180 antibody enzyme-linked immunosorbent assay ti-ters has been reported in a mother and neonate with HG.

Observations: We describe a 33-year-old woman withHG and her neonate with vesicular erythematous le-sions and the weekly follow-up results of the BP180 non-collagenous domain (NC16a) enzyme-linked immuno-sorbent assay.

Conclusions: Almost the same titer of pathogenic anti-body as that in the mother is transferred to the neonate.The plasma elimination half-life of anti-BP180 antibody isapproximately 15 days in mother and neonate. An abrupttwin peak increase in the BP180 enzyme-linked immuno-sorbent assay index from maternal serum was observed justbefore and after delivery, possibly explaining why HG usu-ally occurs in the last trimester of pregnancy and exacer-bates postpartum. Lesions in the neonate resolve withouttreatment far before pathogenic antibody disappears, sug-gesting that factors other than anti-BP180 antibodies maybe involved in the generation of eruptions. Frequent test-ing of the BP180 enzyme-linked immunosorbent assaygreatly facilitates therapeutic planning.

Arch Dermatol. 2007;143(9):1168-1172

H ERPES GESTATIONIS (HG),also known as pemphi-goid gestationis, is a rare,autoimmune, bullous dis-ease that occurs during

the second or third trimester, but it hasbeen reported in the first trimester.1 It flaresat delivery and usually resolves spontane-ously over weeks or months after deliv-ery. Herpes gestationis is clinically char-acterized by pruritic urticarial papules orplaques, polymorphous eruptions, and an-nular or figured erythematous-edema-tous lesions evolving into vesicles and tenseblisters. These skin eruptions usually starton the abdomen and spread over the wholebody, including the extremities; the face,scalp, and mucous membranes are muchless involved.2,3 Immunologically, HG ischaracterized by linear deposition of C3with or without associated IgG at the base-ment membrane zone on direct immuno-fluorescence4,5 and by the presence on se-rum of antibodies to the 180-kDa bullouspemphigoid (BP) antigen (BP180), whichis type 17 collagen contained in thehemidesmosomal components.6,7 Epi-tope mapping has revealed that HG and

BP autoantibodies primarily bind at a com-mon antigenic site within the noncollag-enous domain (NC16a) of BP180.8 MostHG serum samples are positive for theNC16a domain on immunoblot analysis(93%), with enzyme-linked immunosor-bent assay (ELISA) (88%) using the BP180NC16a domain as the antigen.9

We describe herein a mother and neo-nate with HG, with precise weekly fol-low-up studies of the BP180 NC16a ELISA(BP180 ELISA) index during treatment.The present case displayed BP180 ELISAtiters in the umbilical artery and vein of1224.5 and 1021.6, respectively, com-pared with 1521.8 in maternal venousblood; the plasma elimination half-life ofBP180 antibody was approximately 15 daysduring the first 33 days after birth in themother and neonate. The BP180 ELISA ac-tivity had been eliminated by day 112 af-ter delivery in the neonate, but the ma-ternal level remained higher than normal(at 44.3) as of 19 months after delivery.We present an immunoclinical perspec-tive of HG based on BP180 ELISA analy-sis of the present case.

Author Affiliations:Departments of Dermatology(Drs Aoyama, Asai, Hioki, andKitajima) and Pediatrics(Drs Funato and Kondo), GifuUniversity School of Medicine,Gifu City, Japan.



REPORT OF A CASE

A 33-year-old woman with a 1-week history of itchyexudative or urticarial erythema in a figurate and annu-lar pattern, with tense vesicles on the abdomen, pre-sented in gestational week 18 of her third pregnancy(Figure 1A). Examination of the skin revealed palmarplantar pruritic dyshidrosislike vesicles (Figure 1B). Hermedical history included a presumptive diagnosis of pru-ritic urticarial papules and plaques of pregnancy in thefirst pregnancy, with delivery of a healthy neonate, anda spontaneous abortion with pruritic urticarial papulesand plaques of pregnancy in her second pregnancy. Noother history of diseases was elicited. Because severe itch-ing and extension of lesions to the entire body were iden-tified, the patient was admitted to Gifu University Hos-pital, Gifu City, for corticosteroid treatment. The otherwisehealthy newborn boy, weighing 2264 g, had developedannular erythema, with vesicles ranging from coin to wal-nut size and appearing when he was 3 days old, involv-ing the face, trunk, and extremities (Figure 2A and B).The neonate also developed dyshidrosislike vesicles onthe soles, as did the mother (Figure 2C). These lesionsresolved in the neonate by the time he was 10 days old,without treatment. A previous study10 of HG revealed thatHLA-DR3 and HLA-DR4 are 2 predominant histocom-patibility complex II molecules that are common to pa-tients with HG. In our case, the mother and neonate shareda common HLA-DR antigen, because the mother ex-pressed DR4/15 and the neonate expressed DR4.

Skin biopsy specimens were obtained from the edgeof erythema with vesicles on the arm of the mother. Ahistopathological examination using hematoxylin-eosin staining demonstrated marked subepidermal edemaand blisters, with an inflammatory infiltrate mainly com-prising eosinophils in blisters and the upper dermis

(Figure 3A). Direct immunofluorescence of the skin bi-opsy specimens showed linear C3 deposits in the base-ment membrane zone (Figure 3B), whereas IgG wasweakly positive and IgA and IgM were not detected.

The BP180 ELISA titers were determined using a BP180NC16a ELISA kit (MBL, Nagoya, Japan). Briefly, each well

BA

Figure 1. Exudative and erythematous lesions with clusters of vesicles on the abdomen (A) and vesicles on the palm (B) of the mother.

BA

C

Figure 2. Annular erythematous-edematous lesions on the face (A) andabdomen (B) and vesicles on the soles (C) of the neonate.



of standard 96-well microtiter plates was coated with re-combinant glutathione S-transferase NC16a or the sameamount of recombinant glutathione S-transferase. Se-rum samples, diluted 101-fold, 16 times 101-fold, or 64times 101-fold, were incubated in duplicate for reac-tion, and ELISA indexes were determined in accordancewith the instructions of the manufacturer.

The mother was initially treated with 30 mg/d (0.6mg/kg per day) of prednisolone for the first 2 weeks,when the BP180 ELISA index was 2500. By approxi-mately 3 weeks after the initiation of treatment withoral prednisolone at this dosage, eruptions had resolveddespite 1 occasion of exacerbation and the ELISA indexhad also increased once to 5950 before decreasing to2360 (Figure 4). Because eruptions had mostly re-solved, the prednisolone dosage was gradually taperedto 20 mg/d over the following 4 weeks, although theELISA index remained somewhat higher (approxi-mately �2360) than normal, and this dosage was main-tained until delivery, because any further decrease inprednisolone dosage led to flares in disease activity. At1 week after delivery, prednisolone was decreased by 5mg/d, resulting in severe clinical recurrence with amarked increase in the BP180 ELISA index. Given thissevere recurrence of clinical symptoms and the in-creased ELISA index, the prednisolone dosage was in-creased to 35 mg/d for 1 week and gradually decreasedto 15 mg/d over 5 months. The mother has been clini-cally in remission with treatment of 15 mg/d of pred-nisolone and 100 mg/d of azathioprine. The time courseof the BP180 ELISA index in the mother revealed thatanti-BP180 activity increased dramatically 1 month be-fore and 2 weeks after delivery to maximum indexes of2955 and 4300, respectively, and gradually decreasedwith prednisolone treatment to approximately 50 by 9months after delivery.

The BP180 ELISA indexes in the umbilical arteryand vein were 1225 and 1022, respectively, comparedwith 1522 in maternal venous blood. Furthermore, theplasma elimination half-life of the BP180 antibody inthe neonate was approximately 15 days during the ini-tial 33 days after delivery (Figure 5). Anti-BP180 anti-body activity in the neonate, as determined by theBP180 ELISA, was completely eliminated by the timethe neonate was 112 days old. Eruptions in the neonateresolved even when the titer of pathogenic antibodywas still high (BP180 ELISA index, approximately 770),far before the pathogenic antibody titer disappeared.This may suggest that anti-BP180 antibody is insuffi-cient to generate skin lesions.

6000

5000

4000

5500

4500

3500

3000

2500

2000

1500

1000

500

0Oct12

Dec12

Nov12

Apr12

Jan12

May12

Feb12

Jun12

Mar12

Jul12

Aug12

2004 2005

BP18

0 EL

ISA

Inde

x

Prednisolone, mg/d

30 25 20 35 15

50 100

Azathioprine, mg/dMinomycin, mg/d

Delivery

a

b

Erythema

Vesicles

c

Figure 4. Time course of clinical symptoms, treatments, and the 180-kDabullous pemphigoid (BP180) enzyme-linked immunosorbent assay (ELISA)index of the mother: an abrupt twin peak increase in the BP180 ELISA indexis seen in maternal serum just before (a) and after (b) delivery; oralprednisolone, 20 to 30 mg/d, sustains remission, whereas markedexacerbation is seen at the second peak of the BP180 ELISA index just afterdelivery and mild exacerbation is present 4 months after delivery (c).

BA

Figure 3. Hematoxylin-eosin staining of a vesicular erythematous lesion from the mother (A) and a clear linear deposit of C3 shown on direct immunofluorescenceof biopsy specimens from perilesional skin of the mother (B). In A, papillary dermis shows marked edema associated with eosinophils and blisters in whicheosinophils and fibrin strands are contained. The epidermis appears intact.



COMMENT

The present case study of a mother and neonate with HGfocused on the follow-up of BP180 ELISA levels, provid-ing several novel insights that may lead to a better under-standing of HG. This was an extreme rare opportunity tobe able to follow the BP180 ELISA index in a mother andneonate, because most neonates of mothers with HG areborn with normal skin and neonates with vesicular le-sions are rare (estimated at 1 per 100 000 cases).11,12

Our case clearly proves that vesicular erythematouslesions in the neonate are caused by transplacental pas-sage of pathogenic antibodies against BP180 NC16a, fromthe mother to the neonate, as shown by the fact that BP180ELISA indexes in the umbilical artery and vein were 1225and 1022, respectively, compared with 1522 in mater-nal venous blood. Interestingly, indexes from the neo-nate at 1 and 4 days after birth were 1154 and 784, re-spectively. Also, eruptions in the neonate disappeared bythe age of 8 days, much faster than the disappearance ofanti-BP180 activity, suggesting that some unknown fac-tors may be involved in the generation of blisters.

The plasma elimination half-life of IgG is thought to bearound 14 days in the average human adult. The presentneonatal case revealed for the first time, to our knowl-edge, a plasma elimination half-life for anti-BP180 anti-body of approximately 15 days during the first 33 days af-ter birth. The pathway of anti-BP180 antibody eliminationfrom plasma may involve consumption by binding to an-tigen in patients with BP in addition to physiological deg-radation of other IgG levels. Antibody bound to BP180,which is expressed over the entire basal cell surface andon hemidesmosomes in the basement membrane zone, isdegraded by internalization from the membrane surface intobasal cells13 and by several proteolytic enzymes generatedduring the inflammation process because of complimentactivation at the basement membrane zone. However, thepresent study suggests that the consumption of patho-genic IgG in the basement membrane zone is not substan-tial, because the 15-day elimination half-life of patho-genic IgG in the first 33 days is approximately the same asthe physiological elimination half-life of IgG.

Herpes gestationis occurs during the second or thirdtrimester (mean onset, 21 weeks) or, in some cases, dur-ing the first trimester.1 It flares at delivery in 75% of cases.14

In most cases, it regresses spontaneously over weeks ormonths after delivery. This seems to be associated withanti-BP180 antibody activity, as shown in an abrupt twinpeak increase in the BP180 ELISA index from maternalserum just before and after delivery (Figure 4). Exactlywhy pathogenic antibody increases before and after de-livery remains unclear, but the chorionic membrane,which is rich in BP180 antigen, may become labile at thisstage of pregnancy and delivery, resulting in the releaseof antigen into the maternal bloodstream, in turn acti-vating the production of pathogenic antibody.

The elimination rate of pathogenic antibody from thesecond peak on day 40 (shown as “b” in Figure 4) to thetrough level on day 64 after delivery in maternal serumclosely parallels that in the neonate (Figure 5), for thebaseline BP180 ELISA indexes of the mother (2955) and

neonate (1154) normalized to 100%. After postdeliveryday 60, pathogenic antibody disappeared from the neo-nate, whereas antibody in the mother continued to showa mild up-and-down slope (peak “c” in Figure 4) andstayed in the abnormal range, with a BP180 ELISA in-dex of 43.3, suggesting that the disease activity to pro-duce pathogenic antibody remains present, although muchless extensive. This suggests a diagnosis of chronic per-sistent HG or progression to bullous pemphigoid.

Monitoring the ELISA index is useful when cortico-steroid tapering is planned after all skin lesions are gone.Even if all the lesions are gone, we may have to think ofkeeping a minimum dosage of corticosteroids to sup-press antibody production to prevent recurrence, whenELISA indexes remain in high titer, as shown in the moth-er’s case (Figure 4).

In summary, we have presented herein an interestingcoordinate analysis of clinical severity and time course ofthe BP180 ELISA index, demonstrating the following: (1)almost the same titer of pathogenic antibody as that in themother is transferred to the neonate; (2) the plasma elimi-nation half-life of anti-BP180 antibody is approximately15 days during the first 33 days after delivery in motherand neonate; (3) an abrupt twin peak increase in the BP180ELISA index from maternal serum was observed just be-fore and after delivery, possibly explaining why HG usu-ally occurs in the last trimester of pregnancy and exacer-bates postpartum; (4) lesions in the neonate resolve farbefore pathogenic antibody disappears, and without treat-ment, suggesting that factors other than anti-BP180 anti-bodies may be involved in the generation of the eruption;and (5) frequently testing the BP180 ELISA index greatlyfacilitates treatment planning.

Accepted for Publication: February 20, 2007.Correspondence: Yumi Aoyama, MD, Department of Der-matology,GifuUniversitySchoolofMedicine,Yanagido1-1,Gifu City 501-1194, Japan ([email protected]).Author Contributions: Dr Aoyama had full access to allthe data in this study and takes responsibility for the in-tegrity of the data and the accuracy of the data analysis.Study concept and design: Aoyama. Acquisition of data:Aoyama, Asai, Hioki, Funato, and Kondo. Analysis and

100

80

6050

402520

00 8~11 5033 100 150

Time After Delivery, d

BP18

0 EL

ISA

Inde

x, %

MotherNeonate

Figure 5. Elimination curve of the 180-kDa bullous pemphigoid (BP180)enzyme-linked immunosorbent assay (ELISA) index for mother and neonate.Starting points for the ELISA index of the mother (2955) and neonate (1154)were normalized as 100%. The elimination half-life of anti-BP180 antibody isinitially approximately 15 days in mother and neonate in the first 33 daysafter delivery.



interpretation of data: Aoyama and Kitajima. Drafting ofthe manuscript: Aoyama. Critical revision of the manu-script for important intellectual content: Aoyama, Asai,Hioki, Funato, Kondo, and Kitajima. Obtained funding:Aoyama. Administrative, technical, and material support:Kondo. Study supervision: Kitajima.Financial Disclosure: None reported.Funding/Support: This study was supported by grants-in-aid for scientific research from the Ministry of Edu-cation, Culture, Sports, Science, and Technology of Ja-pan; Health and Labor Sciences Research Grants forResearch on Measures for Intractable Disease; and theMinistry of Health, Labor, and Welfare of Japan.

REFERENCES

1. Jenkins RE, Hern S, Black MM. Clinical features and management of 87 patientswith pemphigoid gestationis. Clin Exp Dermatol. 1999;24(4):255-259.

2. Shornick JK, Bangert JL, Freeman RG, Gilliam JN. Herpes gestationis: clinicaland histologic features of twenty-eight cases. J Am Acad Dermatol. 1983;8(2):214-224.

3. Lin MS, Arteaga LA, Diaz LA. Herpes gestationis. Clin Dermatol. 2001;19(6):697-702.

4. Provost TT, Tomasi TB Jr. Evidence for complement activation via the alternatepathway in skin diseases, I: herpes gestationis, systemic lupus erythematosus,and bullous pemphigoid. J Clin Invest. 1973;52(7):1779-1787.

5. Bushkell LL, Jordon RE, Goltz RW. Herpes gestationis: new immunologic findings.Arch Dermatol. 1974;110(1):65-69.

6. Giudice GJ, Wilske KC, Anhalt GJ, et al. Development of an ELISA to detect anti-BP180 autoantibodies in bullous pemphigoid and herpes gestationis. J InvestDermatol. 1994;102(6):878-881.

7. Matsumura K, Amagai M, Nishikawa T, Hashimoto T. The majority of bullous pem-phigoid and herpes gestationis serum samples react with the NC16a domain ofthe 180-kDa bullous pemphigoid antigen. Arch Dermatol Res. 1996;288(9):507-509.

8. Lin MS, Gharia M, Fu CL, et al. Molecular mapping of the major epitopes of BP180recognized by herpes gestationis autoantibodies. Clin Immunol. 1999;92(3):285-292.

9. Sitaru C, Powell J, Messer G, Brocker EB, Wojnarowska F, Zillikens D. Immu-noblotting and enzyme-linked immunosorbent assay for the diagnosis of pem-phigoid gestationis. Obstet Gynecol. 2004;103(4):757-763.

10. Shornick JK, Stastny P, Gilliam JN. High frequency of histocompatibility anti-gens HLA-DR3 and DR4 in herpes gestations. J Clin Invest. 1981;68(2):553-555.

11. Chorzelski TP, Jablonska S, Beutner EH, Maciejowska E, Jarzabek-ChorzelskaM. Herpes gestations with identical lesions in the newborn: passive transfer ofthe disease? Arch Dermatol. 1976;112(8):1129-1131.

12. Katz A, Minto JO, Toole JW, Medwidsky W. Immunopathologic study of herpesgestationis in mother and infant. Arch Dermatol. 1977;113(8):1069-1072.

13. Kitajima Y, Nojiri M, Yamada T, Hirako Y, Owaribe K. Internalization of the 180kDa bullous pemphigoid antigen as immune complexes in basal keratinocytes:an important early event in blister formation in bullous pemphigoid. Br J Dermatol.1998;138(1):71-76.

14. Baxi LV, Kovilam OP, Collins MH, Walther RR. Recurrent herpes gestationis withpostpartum flare: a case report. Am J Obstet Gynecol. 1991;164(3):778-780.



EVIDENCE-BASED DERMATOLOGY: STUDY

SECTION EDITOR: MICHAEL BIGBY, MD; ASSISTANT SECTION EDITORS: DAMIANO ABENI, MD, MPH; ROSAMARIA CORONA, DSc, MD;URBÀ GONZALEZ, MD, PhD; ABRAR A. QURESHI, MD, MPH; HYWEL WILLIAMS, MSc, PhD, FRCP;

CRITICALLY APPRAISED TOPIC FEATURE EDITOR: MICHAEL J. SLADDEN, MAE, MRCP(UK)

Balancing the Benefits and Risks of Drug Treatment

A Stated-Preference, Discrete Choice Experiment With Patients With Psoriasis

Elizabeth M. Seston, PhD; Darren M. Ashcroft, PhD; Christopher E. M. Griffiths, MD, FRCP

Objective: To examine the extent to which theattributes of a treatment affect patients’ choice of treat-ment for psoriasis and the magnitude and nature oftrade-offs between risks and benefits of treatment.

Design: A questionnaire, including a stated-preference,discrete choice experiment, was used to elicit patients’ pref-erences for the treatment of psoriasis.

Setting: Dermatology clinics in England.

Patients: A total of 126 patients with psoriasis.

Main Outcome Measures: Preferences of patients for,and trade-offs between, the 6 attributes of time to mod-erate (50%) improvement, relapse, and risks of experi-encing skin irritation, high blood pressure, liver dam-age, and skin cancer.

Results: The mean age of respondents was 47.6 years, and

the mean duration of psoriasis was 23 years. All 6 at-tributes were important factors affecting choice of treat-ment. The results indicated that patients with psoriasis pri-oritized low risk of skin cancer (�=−0.054; P� .01) andliver damage (�=−0.054; P� .01) and preferred treat-ment that resulted in a shorter time to achieve a moderateimprovement (�=−0.034; P� .01) over a longer time to re-lapse (�=0.028; P�.01). Patients were most willing to waitlonger for a treatment to work if the likelihood of skin can-cer or liver damage was reduced.

Conclusions: This study shows that treatment at-tributes influence patients with psoriasis in their choiceof treatment. The results of the discrete choice experi-ment presented herein indicate that most respondentswould be willing to trade between different aspects of treat-ment to achieve improvements in their psoriasis and mini-mize the risks of adverse events.

Arch Dermatol. 2007;143(9):1175-1179

P SORIASIS IS A CHRONIC IN-flammatory, hyperprolifera-tive skin disease that affectsapproximately 1% to 2% ofthe general population in the

United Kingdom.1 The morbidity associ-ated with psoriasis is well documented interms of its impact on patients’ quality oflife.2-4 Many patients experience prob-lems with body image, self-esteem and self-concept, poor psychological adaptation,and feelings of stigmatization, shame, andembarrassment concerning their appear-ance.2,5,6 The disease also places an eco-nomic burden on patients, particularlythose with severe psoriasis, because theyincur out-of-pocket expenses.7

There is currently no cure for psoriasis,although a wide range of therapies are avail-able, varying considerably in terms of effi-cacy and toxic effects.8 Levels of adher-ence to treatment among patients withpsoriasis are low, with nonadherence ratesreported in 40% of patients.9,10 Factors suchas efficacy9,11 and duration of treatment10 caninfluence the decision whether to con-

tinue with treatment. Previous studies12,13

have shown that many patients with pso-riasis often feel frustrated with the ineffec-tiveness of current therapies and want treat-ment to be more aggressive. It has also beensuggested that dermatologists (and otherphysicians) may not appreciate fully the im-pact that skin diseases have on individualpatients.14,15

Little is known about how patients withpsoriasis evaluate the risks and benefits ofthe treatments they are offered. Given thelow levels of adherence among these pa-tients, information about the decision-making process could be useful for physi-cians considering which treatment to offer.This study used a stated-preference, dis-crete choice experiment (DCE)16 to ex-plore the attitudes of patients toward therisks and benefits of drug treatment for pso-riasis. The objectives were to (1) identify theextent to which the attributes of a treat-ment (eg, degree of improvement, dura-tion of remission, and adverse effects) af-fected patients’ choice of treatment for theirpsoriasis, (2) determine the hierarchical im-

Author Affiliations: School ofPharmacy and PharmaceuticalSciences (Drs Seston andAshcroft) and DermatologyCentre, Hope Hospital(Dr Griffiths), University ofManchester, Manchester,England.



portance of these attributes, and (3) explore the extent towhich patients make trade-offs between risks and benefitsof drug treatment for psoriasis.

METHODS

TREATMENT ATTRIBUTES

Using a DCE, we explored patients’ preferences.16 In a DCE,key attributes of a good or service are identified. Realistic andtradable levels are assigned to these attributes, and scenarios

containing the attributes are created. Scenarios are paired to-gether, and respondents are asked to choose between the sce-narios on the basis of the levels presented. The exercise is thenrepeated with the values of the attributes within the scenariosbeing altered systematically.

The treatment attributes used in the DCE were derived froma review of the published literature, and their relevance wasconfirmed through discussions with 3 consultant dermatolo-gists (one of whom was C.E.M.G.).17-21 In all, 6 attributes wereincluded, namely: (1) time to achieve a moderate (50%) im-provement, (2) time to relapse, (3) risk of skin irritation, (4)risk of high blood pressure, (5) 20-year risk of liver damage,and (6) 20-year risk of skin cancer. Our survey was adminis-tered to patients currently being treated for psoriasis in rou-tine practice and not in a trial setting. We therefore chose a 50%improvement as one of the attributes because this has previ-ously been acknowledged to be a clinically important end pointin the assessment of psoriasis.22,23 The levels used for each ofthese attributes are shown in Table 1.

Thelevelswereorganizedintotreatmentscenariosusingafrac-tional factorial design. This type of design was used because a fullfactorialdesignwouldhaveproduced512(4�4�4�2�2�2)treatment scenarios, which was felt to be too burdensome for re-spondents.Usingexperimentaldesigntheory,wewereabletoiden-tify 16 pair-wise treatment choices in which the levels were var-ied independently, thus avoiding multicollinearity24 (a situationin which 2 variables within the model are highly correlated).

These 16 choice sets were split randomly into 2 separate ver-sions of the questionnaire, each containing 8 choice sets, whichwere distributed to patients. For each choice set, the respon-dents were asked to select which treatment scenario they wouldprefer for the management of their psoriasis. Each version ofthe questionnaire also included a single choice set in which 1of the treatment options would always be expected to be cho-sen as a test for consistency. Each patient, therefore, com-pleted 8 pair-wise choices and 1 test of consistency. Table 2shows an example of a choice set presented to respondents.

Demographic data were also collected from each patient re-lating to age, sex, duration of psoriasis, and age at onset. TheSelf-Assessment Psoriasis Area and Severity Index, which hasa 0 to 72-point scale, was used to determine clinical severity ofa patient’s psoriasis, with a higher score indicating greater se-

Table 1. Treatment Attributes, Definitions, and LevelsUsed in the Discrete Choice Experiment

Treatment Attribute DefinitionAttributeLevelsa

Time to achievemoderate (50%)improvement inpsoriasis

Time taken for psoriasisto improve by 50%following treatment

4, 8, 12, or24 wk

Time to relapse Time taken to lose theimprovement inpsoriasis obtainedfrom the treatmentonce the treatment isstopped

2, 6, 12, or24 wk

Risk of experiencing skinirritation duringtreatment

Percentage risk ofexperiencing skinirritation while usingtreatment

0%, 10%, 20%,or 50%

Risk of high bloodpressure duringtreatment

Percentage risk ofexperiencing highblood pressure whileusing treatment

0% or 10%

20-y risk of experiencingliver damage

Percentage risk of liverdamage fromtreatment over a 20-yperiod

0% or 10%

20-y risk of experiencingskin cancer

Percentage risk ofdeveloping squamouscell or basal cell skincancer as a result oftreatment over a 20-yperiod

0% or 20%

aSee the “Methods” section.

Table 2. Example of Choice Set Usedin the Discrete Choice Experiment

Treatment Attribute Treatment A Treatment B

Time to achieve moderate (50%)improvement in psoriasis, wk

4 12

Time to relapse, wk 2 24Risk of experiencing skin irritation during

treatment, %0a 10b

Risk of high blood pressure duringtreatment, %

10b 0a

20-y risk of experiencing liver damage,% 0a 10b

20-y risk of experiencing skin cancer, % 20c 0a

Which treatment would you prefer?(Please choose 1 only.)

A B

aNo risk.bOne in 10 people.cTwo in 10 people.

Section Editor’s Note

Does determining patients’ preferences have to be so com-plicated? Probably. Ryan et al1 reviewed the state of the artof eliciting public preferences for health care. Quantitativetechniques include ranking exercises, rating scales, andchoice-based techniques (as used in this article by Sestonetal).Qualitative techniques include interviews,casestudyanalyses, theDelphi technique, complaintsprocedures, fo-cus groups, concept mapping, citizens’ juries, consensuspanels, public meetings, and nominal group techniques.All techniqueshavestrengthsandlimitationsthatwerewellreviewed by Ryan et al.1 A thorough description of discretechoice experiments and an example of their use in rheu-matology can be found in a second article by Ryan et al.2

Michael Bigby, MD

1. Ryan M, Scott DA, Reeves C, et al. Eliciting public preferences forhealthcare: a systematic review of techniques. Health Technol Assess.2001;5(5):1-186..

2. Ryan M, Bate A, Eastmond CJ, Ludbrook A. Use of discrete choiceexperiments to elicit preferences. Qual Health Care. 2001;(10 suppl1):i55-i60.



verity.25 In addition, respondents were asked to rank the 6 treat-ment attributes in order of importance ranging from 1 (mostimportant) to 6 (least important). The study was approved bythe local National Health Service research ethics committeesin the 3 areas where the study was conducted and also by theUniversity of Manchester Ethics Committee.

STUDY SAMPLEAND DATA ANALYSIS

Patients were recruited from the dermatology departments of3 Acute National Health Service Hospital Trusts located in north-west England. Questionnaires were distributed to patients whowere attending a dermatology outpatient clinic, and postage-paid envelopes were provided for the returns.

Because each respondent provided up to 8 responses (hav-ing completed 8 choice sets), a random effects probit modelwas considered appropriate, as this accounts for correlationsbetween responses given by the same individual. A probitmodel is a popular specification of a generalized linear model,using the probit link function. Because the response is a seriesof binomial results, the likelihood is often assumed to followthe binomial distribution. Let Y be a binary outcome variable,and let X be a vector of regressors. The probit model assumesthat

Pr (Y=1|X=x)=�(x��),

where � is the cumulative distribution function of the stan-dard normal distribution. Parameters for � are typically esti-mated by maximum likelihood.26

All calculations were conducted using Intercooled Stata sta-tistical software (version 8.0; StataCorp LP, College Station,Texas). In the model, the choice of treatment option was thebinary-dependent variable, and the differences in levels for eachof the 6 attributes were the independent variables. Differenceswere calculated by subtracting the levels. A main-effects lin-ear additive model was estimated in which �� is the depen-dent variable; �o is the constant, representing unobserved in-fluences on choices, �j is the coefficient of the attributes includedin the design, e represents the error term due to differencesamong observations, and u represents the error terms due todifferences among respondents26:

��=�o��1IMPROVE��2RELAPSE��3SKIN��4BLOOD��5LIVER��6CANCER�e�u.

The theoretical validity of the model was tested by exam-ining the signs and significance of coefficient estimates in re-lation to a priori hypotheses. The longer the time to improve-ment and increased risks of adverse effects of treatment were

all expected to have negative coefficients, whereas the longerthe time to relapse was expected to have a positive coefficient.

The trade-offs that patients were willing to make were de-termined by calculating the marginal rate of substitution, thatis, the ratio of any 2 attributes.27 Trade-offs were calculated bydividing the coefficients of the attributes by the coefficient fortime to improvement (�1), giving the estimated trade-offs intime to improvement for a 1-unit change in each of the otherattributes. Confidence intervals for coefficients and the mar-ginal rate of substitution were derived using nonparametric boot-strapping with Stata software (StataCorp LP).28 Bootstrappingallows the confidence intervals to be determined through re-peated sampling with replacement, using parameter point es-timates and their estimated variance-covariance matrix.

The respondents’ treatment choices were explored to see ifthey showed evidence of nontrading behavior among the dif-ferent treatment attributes. Respondents were classified as hav-ing a dominant preference if they always chose the “best” levelof a particular attribute and if they also ranked that attributeas the most important in the ranking exercise.29 To determinewhether including respondents with dominant preferences af-fected the results of the random effects probit model, the datawere analyzed in 2 separate models by including and exclud-ing these respondents.

RESULTS

INFLUENCE OF ATTRIBUTESON TREATMENT CHOICES

In all, 126 patients with psoriasis completed the question-naire. A total of 64 respondents (51%) were female, and62 (49%) were male. Their mean age was 47.6 years (range,21-82 years), the mean duration of their psoriasis was 22.8years (range, 1-63 years), their mean age at onset of pso-riasis was 24.9 years (range, birth to 67 years), and the meanSelf-Assessment Psoriasis Area and Severity Index score was11.1 (range, 0.4-55.2). All 126 respondents passed the con-sistency test, suggesting that they had understood the DCE.Twenty-nine respondents (23%) showed evidence of domi-nant preferences for specific treatment attributes. Most ofthese (19/29) always chose the option for which there wasno risk of experiencing skin cancer. A few respondents alsoshowed evidence of dominant preferences for time to re-lapse (5 subjects), no risk of liver damage (3 subjects), andno risk of high blood pressure (2 subjects).

Table 3. Regression Results From Random Effects Probit Model and Marginal Rates of Substitution in 126 Patients With Psoriasis

Treatment AttributeCoefficient(95% CI)a

Trade-offs Relativeto Time to Improvement, wk

PreferenceInterpretation

Time to achieve moderate (50%) improvement −0.034 (−0.042 to −0.026) NA NATime to relapse 0.028 (0.020 to 0.035) 0.8 (0.6 to 1.0) 1-wk increase in time to relapseRisk of experiencing skin irritation during treatment −0.010 (−0.017 to −0.007) 0.3 (0.2 to 4.0) 1% Reduction in risk of skin irritationRisk of high blood pressure during treatment −0.038 (−0.051 to −0.026) 1.1 (0.8 to 1.5) 1% Reduction in risk of high blood pressure20-y risk of experiencing liver damage −0.054 (−0.068 to −0.042) 1.6 (1.2 to 1.9) 1% Reduction in 20-y risk of liver damage20-y risk of experiencing skin cancer −0.054 (−0.0629 to −0.049) 1.6 (1.4 to 1.8) 1% Reduction in 20-y risk of skin cancerConstant termb 1.331 (1.135 to 1.555) NA NA

Abbreviations: CI, confidence interval; NA, not applicable.aData are presented as � values (range).bThe constant term is automatically included in the Stata software model (StataCorp LP, College Station, Texas) and indicates that unobserved attributes (ie,

attributes not included in our model) are also influencing treatment choices.



Two random effects probit models were estimated; 1with all respondents and 1 with “dominant” respon-dents excluded. A comparison of the models indicatedno significant differences in coefficient estimates, rela-tive size of the coefficients, or the significance of the fitof the models. Therefore, only the model that includedall respondents is presented in Table 3. The coeffi-cients for all 6 treatment attributes were statistically sig-nificantly different from zero (P� .01), indicating thatall the attributes were important factors affecting pa-tients’ choice of treatment. The constant term was alsosignificant (P� .05), indicating that other unobservedattributes were also likely to influence treatment prefer-ences.

The size and direction of the coefficients were in ac-cordance with our a priori hypotheses. The negative signson the coefficient for the adverse effects indicated thatthe higher the risk of experiencing these events, the lesslikely patients were to choose this scenario. Similarly, thenegative sign for the time to improvement attribute in-dicated that the longer the time to achieve a moderateimprovement, the less likely the patients were to choosethat scenario. Positive values for the time to relapse in-dicated that the greater the duration of time to relapse,the more likely the patients were to choose that sce-nario.

RELATIVE IMPORTANCEOF THE TREATMENT ATTRIBUTES

Comparison of the magnitude of the coefficients indi-cated that respondents considered both the long-term risksof experiencing skin cancer and liver damage to be the mostimportant adverse events influencing their choice of treat-ment, followed by the risk of experiencing high blood pres-sure. The risk of skin irritation was considered to be leastimportant. The patients also prioritized time to moderateimprovement over time to relapse. This corresponds withthe results of the ranking exercise in which 30 of 103 re-spondents ranked the time to achieving a moderate im-provement in psoriasis as the most important attribute(29.1%), whereas 17 of 103 (16.5%) ranked time to re-lapse as the most important, as shown inTable4. (It shouldbe noted that 23 respondents [18.3%] did not complete theranking exercise.)

TRADE-OFFS BETWEEN THE BENEFITSAND RISKS OF TREATMENT

Table 3 also shows the marginal rates of substitution be-tween time to achieve a moderate improvement in theseverity of psoriasis and the other attributes; that is, howmuch more time to improvement the respondents werewilling to trade off to achieve an improvement in 1 levelof the other treatment attributes. For example, the re-sults show that respondents were willing to wait 0.8 weekslonger to achieve a moderate improvement if the treat-ment prolonged the time to relapse following such im-provement by 1 week. The least important marginal ratesof substitution were for the risk of skin irritation duringtreatment, and the most important were for the 20-yearrisks of liver damage and skin cancer.

COMMENT

This study has shown that treatment attributes, such asadverse effects, time to improvement, and time to re-lapse, influence the treatment preferences of patients withpsoriasis. Most respondents would be willing to tradeamong different aspects of treatment to achieve an im-provement in their psoriasis with minimal adverse ef-fects. In other words, patients indicated that they wouldwait longer for a treatment to work if the chance of a se-vere adverse effect, such as skin cancer or liver damage,was considerably reduced. Respondents considered thelong-term risks of skin cancer and liver damage to be themost important adverse effects and were prioritized aboveshort-term risks of drug-induced hypertension or skinirritation.

Almost a quarter of the sample were unwilling to tradeamong the attributes and always chose the treatment op-tion with a particular level of their chosen attribute, mostoften no risk of skin cancer. The level of nontrading in thisstudy was low compared with other DCE studies16,29 inwhich nontrading respondents have accounted for 30% to71% of all respondents. It is important to take these intoaccount because they reveal strong preferences on the partof some respondents. These studies suggest that some pa-tients may not accept a treatment offered to them, no mat-ter how efficacious, if they have serious concerns about ad-verse effects. These findings may provide some insight intowhypatientsdemonstrate intentionalnonadherence to treat-ments for psoriasis and provide a useful starting point formeaningful dialogue between patients and physicians aboutchoice of treatment.

Previous studies have explored attitudes to risks andbenefits of treatments in patients with rheumatoid ar-thritis and other musculoskeletal conditions. Fraenkelet al30 found that as many as 60% of respondents wouldnot accept an arthritis treatment if there was a risk, how-ever small, of developing cancer. Another study31 showedthat patients would be willing to accept the risk of treat-ment if the benefits were seen to be high. More recently,it was reported that patients with arthritis prioritized elimi-nating the risk of adverse effects, whether short- or long-term, much more highly than the chance of the treat-ment providing benefit.32

Table 4. Ranking of Attributes by 103 Respondentsa

Treatment Attribute

RespondentsRanking It as

“Most Important,”No. (%)

Time to achieve moderate (50%) improvement 30 (29.1)Time to relapse 17 (16.5)Risk of experiencing skin irritation during

treatment2 (1.9)

Risk of high blood pressure during treatment 9 (8.7)20-y risk of experiencing liver damage 17 (16.6)20-y risk of experiencing skin cancer 28 (27.2)

aThere are some missing values.



To our knowledge, this is the first study to examinethe nature and extent of some of the specific trade-offsthat patients routinely face when choosing among the risksand benefits of drug treatments for psoriasis. However,there are several limitations that should be taken into ac-count in interpreting the findings. First, the patients werepresented with scenarios in which we systematically ma-nipulated specific treatment attributes. Respondents weremaking decisions based on the state of their psoriasis atthe time of the study; however, their attitudes may havevaried according to the severity of their psoriasis. In thewords of one patient, who provided written commentsto support her answers, “If my psoriasis was severe, whichindeed it has been at various times of my life, if I lookback . . . I would probably, in desperation, have agreedto any treatment that would improve my skin condi-tion.” The results were averaged across the sample, andas a result there is inevitable variation between patients.

It is likely that our sample does not represent fully thegeneral population of patients with psoriasis; rather, itis more representative of those being treated in derma-tology clinics in the hospital setting. Future work shouldexplore whether these findings are mirrored by those de-rived from patients who are being cared for in ambula-tory care alone. Likewise, it would also be of interest toexplore patients’ preferences toward other treatment at-tributes, such as the route or mode of administration andthe associated treatment costs.

Nonetheless, this study has provided novel insight intothe attitudes of patients with psoriasis toward the risksand benefits of the treatments they receive. Given theheavy psychological burden associated with psoriasis, thehigh levels of dissatisfaction with the effectiveness of cur-rent treatment, and poor levels of adherence, a greaterunderstanding of how these patients perceive their treat-ment may enhance the dialogue between them and theirphysicians.

Accepted for Publication: March 16, 2007.Correspondence: Elizabeth M. Seston, PhD, School ofPharmacy and Pharmaceutical Sciences, Stopford Build-ing, Oxford Road, University of Manchester, Manches-ter M13 9PT, England ([email protected]).Author Contributions: Dr Ashcroft had full access to allthe data in the study and takes responsibility for the in-tegrity of the data and the accuracy of the data analysis.Study concept and design: Seston, Ashcroft, and Grif-fiths. Acquisition of data: Seston and Ashcroft. Analysisand interpretation of data: Seston, Ashcroft, and Grif-fiths. Drafting of the manuscript: Seston and Ashcroft. Criti-cal revision of the manuscript for important intellectual con-tent: Seston, Ashcroft, and Griffiths. Statistical analysis:Seston and Ashcroft. Obtained funding: Ashcroft and Grif-fiths. Administrative, technical, and material support: Ash-croft. Study supervision: Ashcroft and Griffiths.Financial Disclosure: None reported.Funding/Support: Galderma International Ltd pro-vided financial support for the study via an unrestrictededucational grant.Disclaimer: The views expressed in this article are thoseof the authors alone.Additional Contributions: The staff at the hospitals as-

sisted with the distribution of questionnaires. We thankall of the patients who took part in the study.

REFERENCES

1. Nevitt GJ, Hutchinson PE. Psoriasis in the community: prevalence, severity andpatients’ beliefs and attitudes towards the disease. Br J Dermatol. 1996;135(4):533-537.

2. Ramsay B, O’Reagan M. A survey of the social and psychological effects of psoriasis.Br J Dermatol. 1988;118(2):195-201.

3. Fried RG, Friedman S, Paradis C, et al. Trivial or terrible? the psychosocial im-pact of psoriasis. Int J Dermatol. 1995;34(2):101-105.

4. Rapp SR, Feldman SR, Exum ML, Fleischer AB Jr, Reboussin DM. Psoriasis causesas much disability as other major medical diseases. J Am Acad Dermatol. 1999;41(3, pt 1):401-407.

5. Jowett S, Ryan T. Skin disease and handicap: an analysis of the impact of skinconditions. Soc Sci Med. 1985;20(4):425-429.

6. Ginsburg IH, Link BG. Feelings of stigmatization in patients with psoriasis. J AmAcad Dermatol. 1989;20(1):53-63.

7. Feldman SR, Fleischer AB, Reboussin DM, Rapp SR, Bradhaam DD, Exum ML.The economic impact of psoriasis increases with psoriasis severity. J Am AcadDermatol. 1997;37(4):564-569.

8. Fairhurst DA, Ashcroft DM, Griffiths CEM. Optimal management of severe plaqueform of psoriasis. Am J Clin Dermatol. 2005;6(5):283-294.

9. Richards HL, Fortune DG, Griffiths CEM. Adherence to treatment in patients withpsoriasis. J Eur Acad Dermatol Venereol. 2006;20(4):370-379.

10. van de Kerkhof PC, De Hoop D, de Korte J, Cobolens SA, Kuipers MV. Patientcompliance and disease management in the treatment of psoriasis in theNetherlands. Dermatology. 2000;200(4):292-298.

11. Zaghoul SS, Goodfield MJ. Objective assessment of compliance with psoriasistreatment. Arch Dermatol. 2004;140(4):408-414.

12. Krueger G, Koo J, Lebwohl M, Menter A, Stern RS, Rolstad T. The impact of pso-riasis on quality of life. Arch Dermatol. 2001;137(3):280-284.

13. Nijsten T, Rolstad T, Feldman SR, Stern RS. Members of the National PsoriasisFoundation: more extensive disease and better informed about treatment options.Arch Dermatol. 2005;141(1):19-26.

14. Jemec GB, Wulf HC. Patient-physician consensus on quality of life in dermatology.Clin Exp Dermatol. 1996;21(3):177-179.

15. Hermansen SE, Helland CA, Finlay AY. Patients’ and doctors’ assessment of skindisease handicap. Clin Exp Dermatol. 2002;27(3):249-250.

16. Ryan M, Gerard K. Using discrete choice experiments to value health care pro-grammes. Appl Health Econ Health Policy. 2003;2(1):55-64.

17. Griffiths CEM, Clark CM, Chalmers RJ, Li Wan Po A, Williams HC. A systematicreview of treatments for severe psoriasis. Health Technol Assess. 2000;4(40):1-125.

18. Ashcroft DM, Li Wan Po AL, Williams HC, Griffiths CEM. Systematic review ofcomparative efficacy and tolerability of calcipotriol in treating chronic plaquepsoriasis. BMJ. 2000;320(7240):963-970.

19. Ho VC, Griffiths CEM, Albrecht G, et al; PISCES Study Group. Intermittent shortcourses of cyclosporin (Neoral�) for psoriasis unresponsive to topical therapy:a 1-year multicentre, randomized study. Br J Dermatol. 1999;141(2):283-291.

20. Van Dooren-Greebe RJ, Kuijpers Al, Mulder J, De Boo T, Van de Kerkhof PC.Methotrexate revisited. Br J Dermatol. 1994;130(2):204-210.

21. Marcil I, Stern RS. Squamous-cell cancer of the skin in patients given PUVA andciclosporin: nested cohort crossover study. Lancet. 2001;358(9287):1042-1045.

22. Gordon KB, Feldman SR, Koo J, Menter A, Rolstad T, Krueger G. Definitions ofmeasures of effect duration for psoriasis treatments. Arch Dermatol. 2005;141(1):82-84.

23. Carlin CS, Feldman SR, Krueger JG, Menter A, Krueger G. A 50% reduction in thePsoriasis Area and Severity Index (PASI 50) is a clinically significant endpoint inthe assessment of psoriasis. J Am Acad Dermatol. 2004;50(6):859-866.

24. Street DJ, Burgess L, Louviere JJ. Quick and easy choice sets: constructing optimaland nearly optimal stated choice experiments. Int J Res Mark. 2005;22:459-470.

25. Feldman SR, Fleischer AB, Reboussin DM, et al. The Self-Administered Psoria-sis Area and Severity Index is valid and reliable. J Invest Dermatol. 1996;106(1):183-186.

26. Greene WH. Models with discrete dependent variables. In: Greene WH, ed. Econo-metric Analysis. 4th ed. Upper Saddle River, NJ: Prentice Hall; 2000:811-896.

27. Small KA, Rosen HS. Applied welfare economics with discrete choice models.Econometrica. 1981;49:105-130.

28. Efron B, Tibshirani RJ. An Introduction to the Bootstrap. New York, NY: Chap-man & Hall; 1994:168-176.

29. Scott A. Identifying and analysing dominant preferences in discrete choice ex-periments: an application in health care. J Econ Psychol. 2002;23:383-398.

30. Fraenkel L, Bogardus ST, Concato J, Felson DT. Unwillingness of rheumatoid arthri-tis patients to risk adverse effects. Rheumatology (Oxford). 2002;41(3):253-261.

31. Berry DC, Bradlow A, Bersellini E. Perceptions of the risks and benefits of medi-cines in patients with rheumatoid arthritis and other painful musculoskeletalconditions. Rheumatology (Oxford). 2004;43(7):901-905.

32. Fraenkel L, Bogardus ST, Concato J, Felson DT, Wittink DR. Patient preferences forthe treatment of rheumatoid arthritis. Ann Rheum Dis. 2004;63(11):1372-1378.



EVIDENCE-BASED DERMATOLOGY: RESEARCH COMMENTARY

Use of Antibiotic OintmentAfter Clean Cutaneous SurgeryMichael Bigby, MD; Department of Dermatology, Harvard Medical School, Boston, Massachusetts, andBeth Israel Deaconess Medical Center, Boston

Commentary on: Randomized clinical trial ofthe effect of applying ointment to surgicalwounds before occlusive dressingDixon AJ, Dixon MP, Dixon JB

Br J Surg. 2006;93(8):937-943

Ointments are often applied to surgical wounds after cleancutaneous surgery in the belief that they speed and im-prove wound healing. The effectiveness of this practicewas tested in a large randomized control trial.Question: Does application of no ointment, paraffin oint-ment, or mupirocin ointment lower complication ratesor improve wound healing after clean cutaneous sur-gery when occlusive dressings are applied?Design: A blinded randomized clinical trial was under-taken to evaluate the effect of applying mupirocin oint-ment to a wound before occlusive dressing, in compari-son with no ointment or sterile paraffin.Setting: Skincanceronly Clinic, Geelong, Alfred Hospi-tal, Melbourne, Australia.Patients: All patients were newly referred to the Skin-canceronly Clinic from July 1, 2002, to December 31,2003. One surgeon performed all procedures, which in-cluded punch biopsy, elliptical excision, incisional bi-opsy, skin grafts, and skin flaps.Interventions: No ointment, paraffin ointment, or mupiro-cin ointment was placed on sutured wounds of patients priorto application of moist occlusive dressings. All patients weregiven a detailed instruction sheet regarding wound treat-ment, warning signs, and details of return appointments.At 6 to 9 months after surgery, patients were surveyed toassess the wounds, with a response rate of 74.0%.Main Outcome Measures: The primary outcome was de-velopment of a wound infection defined as purulent su-ture site, suture abscess, cellulitis, infective necrosis, largesubcuticular abscess, regional lymphadenitis, or septi-cemia assessed at suture removal. When there was ab-scess formation or evidence of involvement beyond thelocal site, a wound swab was taken for culture; other-wise, the infection was assessed clinically. In the ab-sence of suppuration, a wound was considered infectedif 3 of the following 4 signs were present: discharge, pain,erythema, or induration. Secondary outcomes includedscarring, suture markings, suture reaction, hypertro-phy, keloid formation, discoloration, wound depres-

sion, wound elevation, dog ears, and contracture. Re-corded adverse reactions included postoperative bleeding,allergy to dressing, allergy to skin preparation, contactdermatitis, local recurrence, subcutaneous fibrosis, granu-loma, dehiscence, pruritus, persistent pain, nerve dam-age, ectropion, nodal involvement, and distant metasta-ses. All key outcome incidences were analyzed with the�2 test.

Results: Some 778 patients with 1801 surgical woundsfollowing excision of skin lesions were enrolled in thetrial. No ointment was placed on 510 sutured woundsof 247 patients, paraffin ointment was placed on 729wounds of 269 patients, and mupirocin ointment on 562wounds of 262 patients. There were no significant dif-ferences in outcome for all primary end points evalu-ated (Table1). The infection rate was 1.4% with no oint-ment, 1.6% for paraffin, and 2.3% for mupirocin (P= .49).Total complication rates were 3.5%, 4.7%, and 4.8% forno ointment, paraffin, and mupirocin, respectively(P= .59). Skin necrosis occurred in 0, 1, and 6 patients,respectively. There was a clinical and statistical increasein wound infections in lower limb lesions.

At 6 to 9 months after surgery, patients were sur-veyed to assess the wounds, with a response rate of 74.0%.Some 10.9%, 10.3%, and 8.2% of patients had a neutralor negative perception of their wounds at 6 to 9 monthsafter surgery with no ointment, paraffin ointment, andmupirocin ointment, respectively (P= .65). There was nodifference in postoperative pain, degree of inconve-nience, or overall level of satisfaction with treatment.

Authors’ Conclusions: Putting ointment on a surgicalwound before occlusive dressing does not benefit the pa-tient. In view of the risk of antibiotic resistance, mupiro-cin ointment is not indicated for clean surgical wounds.

Comment

The trial was well reported. A sample size calculation wasperformed, and adequate numbers of patients were re-cruited. Patients (not wounds) were randomized pro-spectively. The randomization process (an independentperson drew 1 of 150 disks [50 each of 3 different colorsfrom a barrel] without replacement) was well de-scribed. Drawing disks without replacement results in“block randomization,” designed to produce groups of



equal size. The authors state on page 938 that, “The sur-geon and the patients were not aware of the allocation”but the means of concealment of allocation were not de-scribed. Multiple wounds were evaluated in random-ized patients. Because 1 group received no ointment,masking was not entirely possible. Intention-to-treatanalysis (ITT) was performed for wound healing and de-velopment of infection because all wounds were fol-lowed up until healing was complete (to time of sutureremoval or longer if healing was slow or complicationswere experienced). Intention-to-treat analysis was not per-formed but was calculable for long-term results with thedata provided from survey results. Equal numbers of pa-tients failed to return surveys of long-term results fromthe groups receiving no ointment, paraffin ointment, andmupirocin ointment.

All patients had a “moist occlusive dressing,” al-though the specifics of the dressing and its aftercare werenot described. The corresponding author supplied the in-formation on my request. The dressings were variable butwere all of a moist occlusive type as described by Eagle-stein et al.1 In the main, an initial cover like Melolin orTelfa was used. Then the dressing was sealed with tapecircumferentially. Patients were asked to leave the dress-ing on until review. If they wanted to take it off, they wereencouraged to reinforce the dressing and leave it on.

This trial was well conducted and reported. Two flawsdeserve comment. The first is the lack of detailed de-scription of the occlusive dressing used in the pub-lished report. The second relates to the unit of analysis.Although individual patients were randomized into thestudy, analysis was performed using the wound as theunit of analysis—an approach that is problematic if re-sponses to treatment are clustered within a few individu-als with multiple wounds who might be overrepre-sented in one of the treatment groups. The number of

wounds was quite different in the 3 groups (510, 729,and 562 wounds compared with 247, 269, and 262 in-dividual patients, respectively), yet no adjustment for clus-tering effects was made in the analysis. The authors rec-ognized this flaw and commented that the differences ininfection rates for just the first wound were similar to thefindings when all wounds were considered.

The results (Table 1) fully justify the authors’ con-clusions. This article should lead to a paradigm shift inthe management of postoperative wound care.

In a similar randomized, double-blind study pub-lished more than 10 years ago, Smack et al2 demon-strated that, compared with bacitracin ointment, whitepetrolatum is a safe, effective wound care ointment forpostoperative wound care using nonocclusive dressingsthat are changed daily. Procedures included shave andpunch biopsy, electrodessication and curettage, exci-sion, Mohs surgery, and dermabrasion. Compared withbacitracin ointment, white petrolatum possesses a nearlyequally low infection rate and minimal risk for induc-tion of allergy (Table 2). Their data2 indicated that usingwhite petrolatum would produce an insignificantly higherinfection rate (number needed to harm [NNH] 91, evalu-able; 127 ITT) and a lower rate of contact allergy (NNH−112, evaluable; −150 ITT).

This study was also well reported. A power calcula-tions was performed; however, adequate numbers of sub-jects were not recruited because the actual infection ratewas lower than the 3% that was assumed in the calcula-tion. Because of the less-than-adequate number of sub-jects recruited, the authors could not exclude the possi-bility that using white petrolatum would produce a higherinfection rate with an NNH as high as 37, evaluable or52 ITT. The method of randomization and concealmentof allocation were adequate. Patient and evaluator blind-ing was successful. An ITT analysis was not performed

Table 1. Postoperative Occlusive Dressinga

ComplicationNo Ointment

(n = 510 Wounds) 95% CIParaffin

(n = 729 Wounds) 95% CIMupirocin

(n = 562 Wounds) 95% CI P Valueb

Woundinfection

7 (1.4) (0.4-2.4) 12 (1.6) (0.7-2.6) 13 (2.3) (1.1-3.6) .49

Skinnecrosis

0 (0) (0-0.6) 1 (0.1) (0-0.4) 6 (1.1) (0.2-1.9) .007,

Abbreviation: CI, confidence interval.aData are given as the number of events (percentage).bFor the �2 test.

Table 2. Postoperative Nonocclusive Dressing

ComplicationWhite Petrolatum a

(n = 610 Wounds)Bacitracin a

(n = 597 Wounds) Difference (95% CI) NNT (95% CI)

Infection (evaluable) 9 (2.0) 4 (0.9) 0.011 (−0.004 to 0.027) 91 (−250 to 37)Infection (ITT) 9 (1.5) 4 (0.7) 0.008 (−0.004 to 0.019) 127 (−280 to 52)Contact dermatitis (evaluable) 0 (0) 4 (0.9) −0.009 (NP) −112 (NP)Contact dermatitis (ITT) 0 (0) 4 (0.7) −0.007 (−0.015 to 0.002) −150 (−68 to 667)

Abbreviations: CI, confidence interval; ITT, intention to treat; NNT, number needed to treat; NP, not provided.aData are given as number (percentage).



but was calculable. The results (Table 2) justify the au-thors’ conclusions.

Bottom Line: No ointment is necessary when postop-erative wounds are covered with occlusive dressings.Petrolatum is an acceptable ointment for use when non-occlusive dressings that are changed daily are used.Adoption of these changes instead of using antibioticointments on postoperative wounds will result in con-siderable cost savings.

Accepted for Publication: March 14, 2007.Correspondence: Michael Bigby, MD, Department of Der-matology, Beth Israel Hospital, 330 Brookline Ave, Bos-ton, MA 02215 ([email protected]).Financial Disclosure: None reported.

1. Eaglstein WH, Davis SC, Mehle AL, Mertz PM. Optimal use of an occlusive dress-ing to enhance healing: effect of delayed application and early removal on woundhealing. Arch Dermatol. 1988;124(3):392-395.

2. Smack DP, Harrington AC, Dunn C, et al. Infection and allergy incidence in am-bulatory surgery patients using white petrolatum vs. bacitracin ointment: a ran-domized controlled trial. JAMA. 1996;276(12):972-977.

Archives Web Quiz Winner

C ongratulations to the winner of our June quiz, Far-zam Gorouhi, MD, Center for Research and Train-

ing in Skin Diseases and Leprosy, Medical Sciences, Uni-versity of Tehran, Tehran, Iran. The correct answer toour May challenge was factitial dermatitis. For a com-plete discussion of this case, see the Off-Center Fold sec-tion in the July Archives (Heydendael VMR, HoekzemaR. Acute blue patch on the forearm. Arch Dermatol. 2007;143[7]:937-942).

Be sure to visit the Archives of Dermatology Web site(http://www.archdermatol.com) to try your hand at theinteractive quiz. We invite visitors to make a diagnosisbased on selected information from a case report or otherfeature scheduled to be published in the following month’sprint edition of the Archives. The first visitor to e-mailour Web editors with the correct answer will be recog-nized in the print journal and on our Web site and willalso receive a free copy of The Art of JAMA II.



EDITORIAL

Evidence-Based Dermatology Section Welcomesa New Feature: Critically Appraised Topic

S INCE 1999, THE EVIDENCE-BASED DERMATOL-ogy (EBD) section has fulfilled its purposeof alerting clinicians to important advancesin dermatology to help them use the best andmost up-to date evidence when making de-

cisions about patient care. My assistant editors and I se-lect original articles from the biomedical literature to sum-marize in structured, value-added abstracts, and clinicalexperts comment on them. The procedures we follow aswe attempt to achieve this purpose are (1) detecting, usingprestated criteria, the best original and systematic re-view articles, which may or may not include meta-analysis, on the cause, course, diagnosis, prevention, andtreatment of disorders and quality of care or economicsin dermatology; (2) introducing these articles with de-clarative titles and summarizing them accurately in struc-tured abstracts that describe their objectives, methods,results, and evidence-based conclusions; (3) adding brief,

highly expert commentaries to place each of these sum-maries in its proper clinical and health care context; and(4) disseminating the summaries to clinicians in a timelyfashion.

In this issue of the Archives, we introduce criticallyappraised topics (CATs) and the CATs’ feature editor,Michael J. Sladden, MAE, MRCP(UK), to the EBD sec-tion. CATs are written summaries of the application ofthe practice of evidence-based medicine to specific prob-lems that are commonly encountered in clinical prac-tice (ie, translating the problem into an answerable, 4-part,well-built question; finding the best evidence to answerthe question; critically appraising the evidence for its va-lidity, magnitude, and precision; and applying the evi-dence to specific patients). In short, CATs are mini-

Michael Bigby, MD, Section Editor Damiano Abeni, MD, MPH Rosamaria Corona, DSc, MD

Abrar A. Qureshi, MD, MPH Michael J. Sladden, MAE, MRCP(UK) Hywel Williams, MSc, PhD, FRCP

Urbà González, MD, PhD

See also page 1187



systematic reviews of narrow, patient-focused questionsthat are not easy to find in a textbook. For example, after aclinical encounter, a patient-oriented question might bedeveloped, which would then be followed by a search forrelevant high-quality information to answer that ques-tion. The studies that are found will be briefly and criti-cally appraised and then applied back to the patient alongwith the commentary.

If you are interested in contributing to the EBD sec-tion and wish to discuss the topic with the section edi-tor, please e-mail your note of interest to [email protected]. Please specify which kind of contributionyou are interested in providing:

v Critical Appraisal of published articles, whose au-thors are invited to reply.

v Systematic Reviews, which may or may not in-clude meta-analysis.

v Original Contributions.

v Evidence-Based Reports, which are original manu-scripts that are often the work of a consensus confer-ence using the evidence-based approach to evaluate thestatus of a disease and its therapy or to formulate a pub-lic health policy.

v CATs.

Once the manuscript is complete, it is submitted via Web-based process at http://manuscripts.archdermatol.com.The author designates 1 of the 5 categories of submis-sions to the EBD section. The assistant section editorsand I look forward to reviewing your submissions.

Correspondence: Dr Bigby, Beth Israel Deaconess Medi-cal Center, TCC-2, 330 Brookline Ave, Boston, MA 02215([email protected]).

Correction

Error in Byline. In the article titled “Breaking Strengthof Barbed Polypropylene Sutures: Rater-Blinded, Con-trolled Comparison With Nonbarbed Sutures of Vari-ous Calibers,” by Rashid et al, published in the July is-sue of the Archives (2007;143[7]:869-872), the firstauthor’s name should have read as follows: Rashid M.Rashid, MD, PhD.

Michael Bigby, MDSection Editor



EVIDENCE-BASED DERMATOLOGY: CRITICALLY APPRAISED TOPIC

What Is the Chance of a Normal Pregnancyin a Woman Whose Fetus Has Been Exposedto Isotretinoin?Michael J. Sladden, MAE, MRCP(UK); Karen E. Harman, MD, MRCP; Department of Medicine, University of Tasmania,Launceston, Australia (Dr Sladden), and Department of Dermatology, Leicester Royal Infirmary, Leicester, England(Dr Harman)

Clinical Question: Recently, one of our patients who wastaking isotretinoin became pregnant. This occurred de-spite appropriate counseling, a negative pregnancy testresult before commencing treatment, and use of the com-bined oral contraceptive pill Microgynon 30 (ScheringHealth Care, West Sussex, England) (levonorgestrel, 150µg, ethinylestradiol, 30 µg). The patient did not want toterminate her pregnancy. The aim of this Critically Ap-praised Topic is to explore the literature to determine thechance of delivering a healthy child after fetal exposureto isotretinoin; the types of fetal malformations associ-ated with it; and what monitoring should be performed.

Background

Since introduction of the drug in 1982, over 2000 preg-nancies in the United States have been affected by fetalexposure to isotretinoin,1 most resulting in spontane-ous or elective abortions.2-6

Literature Search

We searched the Medline and EmBase databases from1966 to March 2007 using the terms isotretinoin or Ac-cutane or Roaccutane and pregnancy or birth defect.

Appraisal of the Evidence

We found 469 articles in the literature search and chose2 case series3,4 that prospectively identified and fol-lowed up pregnancies in which the fetus was exposed toisotretinoin for which abortion was not elected.3,4 We ex-tracted data from these 2 prospective studies to developour clinical bottom line because these studies are likelyto be less biased than the retrospective studies. These 2studies collected data from a combined total of 151 preg-nant women in the United States aged 14 years to olderthan 35 years whose fetuses were exposed to isotreti-noin.

v In one study3 of 115 pregnancies, there were 21 spon-taneous abortions (18%). Of the 94 live births, 61 werehealthy infants (65% of births, 53% of pregnancies), 26had congenital malformations consistent with isotreti-noin embryopathy (28% of births, 23% of pregnancies),and 7 had other problems (7% of births). Therefore, 28%of live births had congenital malformations consistent with

isotretinoin embryopathy (95% confidence interval, 19%-37%).3

v In the second study4 of 36 pregnancies, there were8 spontaneous abortions (22%). Of the 28 live births, 23were healthy infants (82% of births, 64% of pregnan-cies), and 5 had congenital malformations (18% of births,14% of pregnancies).

The main abnormalities found in isotretinoin embry-opathy are craniofacial, central nervous system, cardio-vascular, and thymic.1,3,4,7,8

v Craniofacial: ear defects, dysmorphism, cleft pal-ate, depressed nasal bridge, hypertelorism;

v Central nervous system: hydrocephalus, micro-cephaly, facial nerve palsy, cortical and cerebellar de-fects;

v Cardiovascular: Fallot’s tetralogy, transposition ofgreat vessels, septal defects, aortic arch hypoplasia;

v Thymic: ectopia, hypoplasia, aplasia; andv Miscellaneous: spina bifida, limb reduction.

In addition, fetal exposure to isotretinoin is associ-ated with high risk of adverse outcome with respect tomental functioning.9 The United Kingdom National Tera-tology Information Service10 estimates that in fetal ex-posure to isotretinoin, 30% of infants with no gross mal-formations have mental retardation, and up to 60% haveimpaired neuropsychological function.

The National Teratology Information Service recom-mends that women who wish to continue their preg-nancy after fetal exposure to isotretinoin should have al-pha-fetoprotein testing at 16 to 19 weeks’ gestation andundergo a targeted ultrasound scan and echocardiogra-phy at 20 to 21 weeks’ gestation.10 These investigationswould give some indication of the risks of structural mal-formations so that parents can plan support services and,in rare instances, in utero intervention could be per-formed, if appropriate.

The high rate of fetal exposure to isotretinoin and itsserious teratogenicity are clearly illustrated. The US Foodand Drug Administration has recently approved the“iPLEDGE” risk management program,11 which is de-signed to reduce the risk of fetal exposure to isotreti-

For editorial commentsee page 1185



noin. However, it is also important that dermatologistsprevent pregnant women from taking the medication(document proof of no pregnancy) and prevent womenwho are taking it from getting pregnant (use of 2 formsof birth control).

Limitations

There is no dose of oral isotretinoin that is safe for use inpregnant women3,12 and, consequently, there are no pub-lished studies of women who took isotretinoin through-out pregnancy. Therefore, information about safety mustbe obtained from studies in which isotretinoin was takenfor acne during some portion of pregnancy.13

Reported outcomes of retrospectively and prospec-tively ascertained cases differ considerably.3 Therefore,we have based our conclusions on data from prospec-tive studies because of the strong likelihood of bias (es-pecially reporting bias) associated with retrospective stud-ies. Published prospective outcome data are available foronly a small proportion of pregnancies in which the fe-tus was exposed to isotretinoin because most of these preg-nancies are not reported in the literature.

The level of fetal exposure to isotretinoin varies frompregnancy to pregnancy so it is possible that isotretinoin-related problems may be higher for women who con-tinue taking isotretinoin for a longer duration before dis-covering that they are pregnant. However, there isinsufficient data to address this issue. There is little in-formation about the timing of spontaneous abortions,either in weeks or trimesters. There is little follow-up dataon infants with no gross malformation to determine therisk of developmental disabilities later in life.

Clinical Bottom Line

In pregnancies in which the fetus is exposed to isotreti-noin,

v The risk of spontaneous abortion is approximately20%;

v In pregnancies that progress, 65% to 82% of neo-nates appear normal at birth, but there is insufficient datato determine how many will later develop isotretinoin-related problems;

v There is an 18% to 28% risk of isotretinoin embry-opathy;

v There is no safe level of exposure: any exposure cancause malformation;

v The main abnormalities are craniofacial, cardiac, cen-tral nervous system, and thymic; and

v Women who choose to continue their pregnancy re-quire careful support and monitoring.

Finally, it is important that dermatologists preventpregnant women from taking isotretinoin and preventwomen who are taking it from getting pregnant.

What Happened to Our Patient?

Our patient ceased taking isotretinoin as soon as shediscovered that she was pregnant, at approximately 6 to7 weeks’ gestation. We discussed with her the evidenceregarding isotretinoin and birth defects. She elected tocontinue with her pregnancy and underwent regularultrasound scans, performed by her obstetrician. Shedelivered a healthy baby girl, with no apparent birthdefect. At age 18 months, her daughter was developingnormally.

REFERENCES

1. Lindstrom J. Isotretinoin: background and regulatory history. FDA Joint Derma-tologic and Ophthalmic Drugs & Drug Safety and Risk Management AdvisoryCommittee Meeting; February 26–27, 2004; Gaithersburg, MD. http://www.fda.gov/ohrms/dockets/ac/04/slides/4017s1.htm. Accessed March 12, 2007.

2. Abroms L, Maibach E, Lyon-Daniel K, Feldman SR. What is the best approach toreducing birth defects associated with isotretinoin? PLoS Med. 2006;3(11):e483.

3. Dai WS, LaBraico JM, Stern RS. Epidemiology of isotretinoin exposure duringpregnancy. J Am Acad Dermatol. 1992;26(4):599-606.

4. Lammer EJ, Chen DT, Hoar RM, et al. Retinoic acid embryopathy. N Engl J Med.1985;313(14):837-841.

5. Mitchell AA, Van Bennekom CM, Louik C. A pregnancy-prevention program inwomen of childbearing age receiving Isotretinoin. N Engl J Med. 1995;333(2):101-106.

6. Berard A, Azoulay L, Koren G, Blais L, Perreault S, Oraichi D. Isotretinoin, preg-nancies, abortions and birth defects: a population-based perspective. Br J ClinPharmacol. 2007;63(2):196-205.

7. Lynberg MC, Khoury MJ, Lammer EJ, Waller KO, Cordero JF, Erickson JD.Sensitivity, specificity, and positive predictive value of multiple malformationsin isotretinoin embryopathy surveillance. Teratology. 1990;42(5):513-519.

8. Stern RS, Rosa F, Baum C. Isotretinoin and pregnancy. J Am Acad Dermatol.1984;10(5, pt 1):851-854.

9. Adams J, Lammer EJ. Neurobehavioral teratology of isotretinoin. Reprod Toxicol.1993;7(2):175-177.

10. National Teratology Information Service. Exposure to Isotretinoin During Pregnancy.Newcastle upon Tyne, England: National Teratology Information Service, Re-gional Drug and Therapeutics Centre; 2001.

11. Food and Drug Administration. iPLEDGE update. 2006. http://www.fda.gov/cder/drug/infopage/accutane/iPLEDGEupdate200603.htm. Accessed May 28, 2007.

12. Goldsmith LA, Bolognia JL, Callen JP, et al. American Academy of DermatologyConsensus Conference on the safe and optimal use of isotretinoin: summary andrecommendations [published correction for a dosage error appears in J Am AcadDermatol. 2004;51(3):348]. J Am Acad Dermatol. 2004;50(6):900-906.

13. Rothman KF, Pochi PE. Use of oral and topical agents for acne in pregnancy.J Am Acad Dermatol. 1988;19(3):431-442.



ON THE HORIZON

SECTION EDITOR: GARY S. WOOD, MD; ASSISTANT SECTION EDITORS: CRAIG A. ELMETS, MD; MOLLY A. HINSHAW, MD;JAY C. KLEMME, MD, MPH; MARK R. PITTELKOW, MD; MARTIN A. WEINSTOCK, MD, PhD; DAVID T. WOODLEY, MD

Accessible Evidence-Based Medicine

Critically Appraised Topics

David A. Barzilai, MD, PhD; Martin A. Weinstock, MD, PhD; Department of Dermatology (Drs Barzilai and Weinstock),Brown University, and Department of Veterans Affairs Medical Center (Dr Weinstock), Providence, Rhode Island

Learning Evidence-Based Veterinary MedicineThrough Development of a CriticallyAppraised TopicHardin LE, Robertson S

J Vet Med Educ. 2006;33(3):474-478

Evidence-based veterinary medicine is a relatively newfield of study. Increased knowledge of medicine coupledwith the increased ability of computers and other elec-tronic devices present overwhelming information. Thecritically appraised topic (CAT) is one method to gatherand evaluate information related to a clinical question.CATs in informatics are short summaries of evidence, usu-ally found through literature searches, in response to aspecifically stated, clinically oriented problem or ques-tion. This article describes a study in which each first-year veterinary student developed a CAT as a class project.The results of this project indicate that students were ableto successfully develop CATs and that this exercise helpedthem understand evidence-based veterinary medicine con-cepts. Though some modification in this project will bemade in the future, overall it was a worthwhile effort andwill remain as an activity in the course.

COMMENT

The cumulative knowledge base of pertinent data for pa-tient management is tremendous and rapidly growing indermatology. To facilitate efficient and accurate acqui-sition of the best current knowledge, evidence-based medi-cine (EBM) has emerged as the science of integrating “thebest research evidence with our clinical expertise and ourpatient’s unique values and circumstances.”1(p1) CATs andCAT banks (collections of CATs) will become an impor-tant part of this effort in the future. In the evidence-based hierarchy, CATs are between narrative reviews andsystematic reviews in quality and come in handy whensystematic reviews are not available.

While a full discussion of EBM is beyond the scopeof this brief discussion,1-5 using EBM to answer a clini-cal query involves the following steps: (1) formalizingthe question into a precisely worded, answerable ques-

tion; (2) systematically searching out the best evidenceavailable; (3) critically appraising the evidence; (4) in-tegrating the data with clinical expertise and patient val-ues; and (5) archiving the results and learning from steps1 through 4. In clinical practice, steps 2 and 3 (search-ing out and critically appraising the best evidence) areoften the most challenging. Step 5 is essential to allowothers to benefit from an EBM review, and this is wherethe CATs and a CAT bank are useful.

The quality of clinical evidence can be categorized in acontinuum ranging from lower-level evidence such as acolleague’s subjective opinion or a solitary case report tosystematic reviews based on double-blind, randomized,controlled trials. Systematic reviews are regarded as thehighest level of data because they involve comprehensivesearches of the literature and use explicit criteria for evalu-ating and incorporating data into the conclusions. A fre-quent challenge in clinical practice, however, is to find asystematic review that addresses a particular query. Such

Editor’s Note

Critically Appraised Topics

How wonderfully appropriate that the benefits of CATsare recognized in evidence-based veterinary medicine!The medical care providers for our feline friends wrapthe methods of informatics around clinical questions justas those felines that we share living space with wrap them-selves around our legs. In this issue, the Archives intro-duces the CATs feature of the Evidence-Based Derma-tology section. Our planning for this feature began inJanuary 2007. We are delighted to be able to implementthis clinically relevant feature for our readers and inviteyou to send your suggestions for topics to be addressedto the CATs feature editor, Michael Sladden, MAE,MRCP(UK) at [email protected].

Now our canine pals are begging for equal represen-tation, and who can resist those wet noses and waggingtails. We must recognize DOGs, disease-oriented groups.In recent years, the Archives published the work of sev-eral multicenter groups who focus on a single disease toperform randomized clinical trials to provide the high-est level of evidence for therapeutic interventions. Thechallenging work of these visionary seeing eye DOGs isleading the way to improve therapy for our patients.

June K. Robinson, MD

See also page 1187

(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), 2007 WWW.ARCHDERMATOL.COM1189


reviews are often not available, and to create one couldtake a year or more to reach a conclusion (by whichtime the question might no longer be relevant to thecase that prompted the query). Narrative reviews aremore likely to have some answer, but these are a poorsolution because they do not reliably, comprehensively,or critically appraise the evidence and hence ultimatelyresemble expert opinion. In practice, dermatologistsmay use medical search engines like PubMed or evenWeb search engines like Google to find their answers,but the first answer is often not the most current, repre-sentative, or accurate. Finally, none of these methodspermits an easy system for storing the answer for futurereference or sharing results with colleagues.

What if a tool existed to help rapidly identify valu-able data, critically appraise its quality, and then store itas a 1-page summary for future reference? There is sucha tool, the CAT. Literally speaking, CATs are 1-page sum-maries of evidence, but what makes them so powerful isthe standardized approach to conducting evidence-based searches and appraising the evidence. The otherpremise behind CATs is that they are educational. Thebest way to learn EBM is through answering real-life clini-cal queries developed in practice.

The Centre for Evidence-Based Medicine, Heading-ton, England, has developed publicly accessible and freesoftware for generating CATs, CATMaker. This softwarecan be downloaded to any personal computerwith access to the Web (www.cebm.net/catmaker.asp).CATMaker software facilitates data entry, helps carry outcalculations to determine the clinical importance of dif-ferences detected in the study, and generates an output filethat can be edited by a word processor. In addition, the soft-ware is equipped with an online help system designed fornovices to EBM searches. The output of this process is asimple 1-page CAT summary that can readily be shared withcolleagues. With this software, the output can be con-verted to HTML format for Web-based searching, printedout locally for filing away the results, and saved as a “Kit-ten,” which serves as the basis of future updates to the CAT.The clinical bottom-line section of the CAT summarizesthe results.

Multiple online CAT banks are currently available(Table). By August 2007, ebDerm.org will be releasinga user community–generated international database ofdermatology CATs. The ebDerm.org CATs will be con-structed using CATMaker software.

Several important limitations of CATs should be noted.First, the study or studies under inquiry, in contrast withsystematic reviews, may not represent the best or onlydata on a subject.1 Second, because CATs are user-driven and generally not subject to the highest level ofpeer review, they are vulnerable to bias. Third, CATs, likeother data sources, may no longer represent the most re-cent best data if they are not periodically updated. Forthese reasons, if a systematic review can be found to an-swer a clinical question (such as through the CochraneSkin Group), this, rather than a CAT, would be the bestsource of data. Finally, the generalizability of a CAT toany individual practice setting must be considered. CATsummaries of severe psoriasis cases seen in the derma-tology department of an urban tertiary medical centermight have limited application to moderate psoriasis casesfound in a rural middle class setting.

Despite these limitations, the CAT is a powerful toolfor improving the quality of evidence-based searches andfor practicing EBM skills. CATs can also save tremen-dous time when a previously generated CAT resemblesyour query and can be quickly updated to reflect the bestcurrent data. In the future, CATs may also be incorpo-rated as a method for practicing evidence-based searchand critical appraisal skills in dermatology residency oras continuing medical education for dermatologists.

Correspondence: David A. Barzilai, MD, PhD, Depart-ment of Dermatology, Brown University, 593 Eddy St,APC 10th Floor, Providence, RI 02903 ([email protected]).Financial Disclosure: Dr Barzilai is the founder ofebDerm.org. Dr Weinstock is on the advisory board ofebDerm.org.

REFERENCES

1. Straus SE, Richardson WS, Glasziou P, Haynes RB. Evidence-Based Medicine: Howto Practice and Teach EBM. 3rd ed. New York, NY: Churchill Livingstone; 2005.

2. Hardin LE, Robertson S. Learning evidence-based veterinary medicine through de-velopment of a critically appraised topic. J Vet Med Educ. 2006;33(3):474-478.

3. Barzilai DA, Freiman A, Dellavalle RP, Weinstock MA, Mostow EN. Dermatoepi-demiology. J Am Acad Dermatol. 2005;52(4):559-578.

4. Barzilai DA. Evidence-based dermatology: a brief introduction. http://www.vgrd.org/archive/cases/2007/ebd/ebd.htm. Accessed May 2, 2007.

5. Bigby M. Evidence-based medicine in a nutshell: a guide to finding and using thebest evidence in caring for patients. Arch Dermatol. 1998;134(12):1609-1618.

Table. Critically Appraised Topic (CAT) Web Sites

Name Description URL

BestBets CATs in a variety of clinical areas www.bestbets.orgCAT bank CATs in a variety of clinical areas www.minervation.com/cebm2/cats/allcats.htmlDermatoscopia Spanish language dermatology CATs www.dermatoscopio.clebDerm.org (evidence-based dermatology) Will host English language CATs in August 2008 ebDerm.org

(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), 2007 WWW.ARCHDERMATOL.COM1190


EDITORIAL

Long-term Treatment for Severe Psoriasis

We’re Halfway There, With a Long Way to Go

O VER THE LAST 30 YEARS, THE LANDMARK

PUVA [psoralen–UV-A] Follow-upStudy has demonstrated the impor-tance of clinical epidemiology researchin making informed treatment deci-

sions for patients with psoriasis. When PUVA was firstintroduced, psoriasis was widely believed to be an epi-dermal cell proliferation disorder, and there were few sys-temic treatment options available at that time.1 Thirty yearslater, psoriasis is believed to be an immunologic disor-der, and more new systemic therapies have been ap-proved to treat it in the last 4 years than in the previous30 years combined.2-8 Our objective criterion regardingwhich patients have severe psoriasis and therefore are can-didates for systemic therapy has also evolved during thisperiod, declining from 20% to 30% body surface area(BSA) in the 1970s to 1990s to 5% more recently.9,10 Withthe increasing recognition of the impact of psoriasis onhealth-related quality of life and the advent of novel thera-pies targeting its immunopathogenesis, the treatment ofpsoriasis is undergoing a revolution. As patients with pso-riasis are increasingly being treated with systemic agentson a long-term basis, the PUVA study provides an im-portant reminder of the challenge involved in makingclinical decisions based on a scientific understanding ofthe disease’s natural history and the robust long-termsafety and efficacy data of its treatments.

First, the PUVA experience exemplifies the critical im-portance of long-term studies to fully define the risks andbenefits of a novel treatment. In 1974, years before theimportance of PUVA in targeting T cells and dendriticcells was recognized, this treatment was first shown tobe effective for severe psoriasis in 21 patients who hadat least 50% BSA involvement.1,11 The use of PUVA fortreating very severe psoriasis was a major development,as all patients in the trial achieved complete clearanceand reported adverse effects were minimal. In 1977, a largemulticenter clinical trial involving more than 1300 pa-tients who received more than 45 000 treatments dem-onstrated that only 3% of patients failed to achieve clear-ance with this regimen and that adverse effects due toPUVA therapy were uncommon, temporary, and gener-ally mild.12 Although PUVA therapy was considered anexperimental technique, with limited long-term safetydata, it was widely used in the United States, with an es-timated 35 000 patients being treated in 1978 alone.13 In

1979, the first observations of cutaneous carcinoma werereported in the cohort; however, it was not clear if PUVAtherapy was responsible for the excess number of skincancers, as affected patients had a history of treatmentwith ionizing radiation or a history of skin cancer.13 In1984, about 2 years after being approved by the Food andDrug Administration, PUVA was definitively linked toan excess risk of squamous cell carcinoma when the co-hort had an average follow-up of 5.7 years.14 In 1997, whenthe median follow-up of patients reached 19 years, PUVAuse was associated with an increased risk of melanoma,a finding that remains controversial.15

Although it took 10 years to clearly demonstrate therisk of squamous cell carcinoma from PUVA therapy andeven longer to detect a potential association with mela-noma, the PUVA study represents a success in definingthe long-term safety of a systemic psoriasis treatment. Cur-rently, the most robust safety data for psoriasis treat-ments are derived predominantly from randomized con-trolled clinical trials. However, these trials are generallyof short duration, measured in weeks to months, whereaspsoriasis is a lifelong disease that requires several de-cades of treatment for control.16 Although clinical trialsare well suited to define the efficacy of an agent, they areparticularly prone to miss the effects of drugs that aredelayed and/or uncommon (eg, cancer, cardiovascular dis-ease, and serious infections).17 The current drug ap-proval process leaves us with wide gaps in our knowl-edge of treatment safety, which is particularly problematicwhen the therapy is to be used in large populations ofpatients on a long-term basis.18 In particular, existing safetydata of systemic therapies for severe psoriasis are lim-ited in the duration of drug exposure and in the numberof patients who are receiving follow-up. Therefore, thepotential risks associated with truly long-term treat-ment of psoriasis remain to be further defined for seri-ous end points such as malignancy.19

The current PUVA Follow-up Study, as analyzed byNijsten et al,20 also provides unique data regarding theeffectiveness of long-term treatment of severe psoriasis.During the follow-up of 815 patients who underwent 2378skin examinations from 1985 to 2005, approximately 50%of such examinations demonstrated that patients had mildto no skin disease. Moreover, Nijsten and coauthors notethat the likelihood that the extent of psoriasis will changemore than 1 physician global assessment level over 1 yearand over 10 years is relatively small. These observationssuggest that our treatment approach is about 50% effec-tive over the long term, given that the average BSA of pa-tients who initially entered the multicenter PUVA trial

See also page 1113



was 33%. Based on the ecological design used for thePUVA study, in which treatment use and psoriasis se-verity were not linked at the individual level, it is un-clear if these patients had mild or no psoriasis becauseof the natural history of psoriasis or because of the useof psoriasis treatments. The natural history and deter-minants of psoriasis remission and flare remain poorlyunderstood as reviewed by Nijsten and colleagues. Thisstudy therefore underscores the need for further pro-spective epidemiological investigations to determine thetrue rate and determinants of spontaneous psoriasis im-provement and flare, as well as how psoriasis may leadto other disease states that are associated with chronicTH1 inflammation, such as psoriatic arthritis, metabolicdisease, atherosclerosis, and myocardial infarction.21-24

Such information would better inform patients about thenatural history of their psoriasis, would improve clini-cal trial designs and the interpretation of safety data, andwould allow clinicians and patients to make more in-formed decisions about long-term vs intermittent therapy.

Remarkably, Nijsten et al20 report that despite havingaccess to a wide array of dermatologic therapeutic mo-dalities, including extensive use of phototherapy andmethotrexate in the study population, about 50% of pa-tients continued to have moderate to severe psoriasis ontheir follow-up examinations. This observation demon-strates that we still have a long way to go when it comesto consistently controlling severe psoriasis. For ex-ample, a recent study of 1657 contacts of the NationalPsoriasis Foundation demonstrated that 37% of pa-tients with moderate psoriasis (defined as 3%-10% BSA)and 39% of patients with severe disease (�10% BSA) werenot receiving any therapy.25 Furthermore, more than 80%of patients with severe psoriasis were not receiving sys-temic therapy or phototherapy. These data suggest thatsevere psoriasis remains uncontrolled in the majority ofpatients affected. Further studies are necessary to deter-mine the degree to which this is attributable to patientfactors (such as changing treatment preference, adher-ence, or access to therapies) or to factors related to treat-ment failure over time. Although short-term clinical trialsin highly selected patient populations have shown theefficacy of systemic therapies for psoriasis, we have largegaps in our knowledge of the long-term effectiveness ofthese treatments when they are used on a long-term ba-sis in a broader population of patients with psoriasis. Forexample, the limited data we have on methotrexate sug-gest that the drug is highly efficacious over the short term(16 weeks) but that over 1 year of follow-up about 25%of patients discontinue therapy because of adverse ef-fects, indicating that long-term effectiveness is well be-low short-term efficacy.26 Similarly, long-term studies ofbiologic antibody therapy (chimeric and humanized) forinflammatory diseases show high short-term efficacy butlower long-term effectiveness as a result of drug discon-tinuation associated with adverse effects or loss of treat-ment response during extended periods of treat-ment.7,27,28

Dermatologists and patients have scant data to makehead-to-head comparisons of systemic therapies, and ourgeneral assumptions about which treatments are most safe,efficacious, and cost-effective based on short-term trial

data may not be valid when the end point is more clini-cally relevant (eg, long-term effectiveness). The burdenof treatment success and failure over the patient’s life-time is eloquently described by John Updike,29 whowrote,“ . . . when I am at last too ill for all of these de-manding and perilous palliatives, the psoriasis like a smol-dering fire in damp peat will break out and spread tri-umphantly; in my dying I will become hideous, I willbecome what I am.”

Over the last 30 years, tremendous progress has beenmade in our knowledge of the pathogenesis of psoriasisand in the development of systemic therapies to treat thisdisease. This progress has been well documented by themore than 20 000 medical publications related to pso-riasis since PUVA was introduced. Despite this progress,we still have major gaps in our basic knowledge of thenatural history of psoriasis, and a large percentage of pa-tients with extensive psoriasis will continue to suffer fromthe burden of this disease for decades.

Correspondence: Dr Gelfand, Department of Dermatol-ogy and Center for Clinical Epidemiology and Biostatis-tics, University of Pennsylvania, 3600 Spruce St, 2 Malo-ney Bldg, Philadelphia, PA 19104 ([email protected]).Financial Disclosure: Dr Gelfand has been a consultantfor Genentech, Centocor, Amgen Inc, Wyeth, Warner-Chilcott, Novartis, Signum, Follica, and Therakos andhas received honoraria from CME Consultants and Emeri-tus Educational Sciences as well as grants from Biogen-idec, AMGEN, Astellas, Centocor, and Longport Inc. Heis also on the Medical Advisory Board for the NationalPsoriasis Foundation and on the Clinical Guidelines Com-mittee for the American Academy of Dermatology.Funding/Support: Dr Gelfand is supported by grantK23AR051125 from the National Institute of Arthritis andMusculoskeletal and Skin Diseases.Role of the Sponsor: The sponsor had no role in this edi-torial.Additional Contributions: John R. Stanley, MD, andDaniel Roling, MD, provided helpful feedback on earlydrafts of this manuscript.

REFERENCES

1. Parrish JA, Fitzpatrick TB, Tanenbaum L, Pathak MA. Photochemotherapy of pso-riasis with oral methoxsalen and longwave ultraviolet light. N Engl J Med. 1974;291(23):1207-1211.

2. Gottlieb AB. Immunologic mechanisms in psoriasis. J Am Acad Dermatol. 1988;18(6):1376-1380.

3. Krueger JG, Bowcock A. Psoriasis pathophysiology: current concepts ofpathogenesis. Ann Rheum Dis. 2005;64(suppl 2):ii30-ii36.

4. Leonardi CL, Powers JL, Matheson RT, et al. Etanercept as monotherapy in pa-tients with psoriasis. N Engl J Med. 2003;349(21):2014-2022.

5. Gordon KB, Papp KA, Hamilton TK, et al. Efalizumab for patients with moderateto severe plaque psoriasis: a randomized controlled trial. JAMA. 2003;290(23):3073-3080.

6. Ellis CN, Krueger GG. Treatment of chronic plaque psoriasis by selective target-ing of memory effector T lymphocytes. N Engl J Med. 2001;345(4):248-255.

7. Menter A, Feldman SR, Weinstein GD, et al. A randomized comparison of con-tinuous vs. intermittent infliximab maintenance regimens over 1 year in the treat-ment of moderate-to-severe plaque psoriasis. J Am Acad Dermatol. 2007;56(1):31.e1-e15.

Joel M. Gelfand, MD, MSCE



8. Gordon KB, Langley RG, Leonardi C, et al. Clinical response to adalimumab treat-ment in patients with moderate to severe psoriasis: double-blind, randomizedcontrolled trial and open-label extension study. J Am Acad Dermatol. 2006;55(4):598-606.

9. Roenigk HH Jr, Auerbach R, Maibach H, Weinstein G, Lebwohl M. Methotrexatein psoriasis: consensus conference. J Am Acad Dermatol. 1998;38(3):478-485.

10. Pariser DM, Bagel J, Gelfand JM, et al. National Psoriasis Foundation clinical con-sensus on disease severity. Arch Dermatol. 2007;143(2):239-242.

11. Baker BS, Swain AF, Griffiths CE, Leonard JN, Fry L, Valdimarsson H. EpidermalT lymphocytes and dendritic cells in chronic plaque psoriasis: the effects of PUVAtreatment. Clin Exp Immunol. 1985;61(3):526-534.

12. Melski JW, Tanenbaum L, Parrish JA, Fitzpatrick TB, Bleich HL. Oral methox-salen photochemotherapy for the treatment of psoriasis: a cooperative clinicaltrial. J Invest Dermatol. 1977;68(6):328-335.

13. Stern RS, Thibodeau LA, Kleinerman RA, Parrish JA, Fitzpatrick TB. Risk of cu-taneous carcinoma in patients treated with oral methoxsalen photochemo-therapy for psoriasis. N Engl J Med. 1979;300(15):809-813.


15. Stern RS, Nichols KT, Vakeva LH. Malignant melanoma in patients treated forpsoriasis with methoxsalen (psoralen) and ultraviolet A radiation (PUVA): thePUVA Follow-up Study. N Engl J Med. 1997;336(15):1041-1045.

16. Naldi L, Svensson A, Diepgen T, et al. Randomized clinical trials for psoriasis1977-2000: the EDEN survey. J Invest Dermatol. 2003;120(5):738-741.

17. Gelfand JM. Pharmacovigilance: verifying that drugs remain safe. In: WolvertonS, ed. Comprehensive Dermatologic Drug Therapy. 2nd ed. Philadelphia, PA:Elsevier; 2007:1017-1026.

18. Okie S. Safety in numbers—monitoring risk in approved drugs. N Engl J Med.2005;352(12):1173-1176.

19. Gelfand JM, Shin DB, Neimann AL, Wang X, Margolis DJ, Troxel AB. The risk oflymphoma in patients with psoriasis. J Invest Dermatol. 2006;126(10):2194-2201.

20. Nijsten T, Looman CWN, Stern RS. Clinical severity of psoriasis in last 20 yearsof PUVA Study. Arch Dermatol. 2007;143(9):1113-1121.

21. Gelfand JM, Neimann AL, Shin DB, Wang X, Margolis DJ, Troxel AB. Risk of myo-cardial infarction in patients with psoriasis. JAMA. 2006;296(14):1735-1741.

22. Neimann AL, Shin DB, Wang X, Margolis DJ, Troxel AB, Gelfand JM. Prevalenceof cardiovascular risk factors in patients with psoriasis. J Am Acad Dermatol.2006;55(5):829-835.

23. Gelfand JM, Gladman DD, Mease PJ, et al. Epidemiology of psoriatic arthritis inthe population of the United States. J Am Acad Dermatol. 2005;53(4):573.

24. Ludwig RJ, Herzog C, Rostock A, et al. Psoriasis: a possible risk factor for de-velopment of coronary artery calcification. Br J Dermatol. 2007;156(2):271-276.

25. Horn EJ, Patel KM, Fox C, Chiou F, Dann F, Lebwohl M. Treatment patterns andutilization of systemic therapy and phototherapy among psoriasis patients[abstract]. J Am Acad Dermatol. 2007;56:2725.

26. Heydendael VM, Spuls PI, Opmeer BC, et al. Methotrexate versus cyclosporinein moderate-to-severe chronic plaque psoriasis. N Engl J Med. 2003;349(7):658-665.

27. Bartelds GM, Wijbrandts CA, Nurmohamed MT, et al. Clinical response to adali-mumab: the relationship with anti-adalimumab antibodies and serum adali-mumab concentrations in rheumatoid arthritis [published online ahead of printMarch 9, 2007]. Ann Rheum Dis.

28. Hyrich KL, Lunt M, Watson KD, Symmons DP, Silman AJ. Outcomes after switch-ing from one anti–tumor necrosis factor alpha agent to a second anti–tumor ne-crosis factor alpha agent in patients with rheumatoid arthritis: results from a largeUK national cohort study. Arthritis Rheum. 2007;56(1):13-20.

29. Updike J. Self-Consciousness. New York, NY: Alfred A Knopf Inc; 1989.



EDITORIAL

Dermatology and the Human Genome

An Epidemiologic Approach

A S THE BODY’S MOST ACCESSIBLE ORGAN, SKIN

provides a unique window on physi-ologic processes governed ultimately bythe genome. Cutaneous signs have longbeen used to diagnose systemic disease,

from hypothyroidism to syphilis,1 and many genetic syn-dromes include pathognomonic skin findings. The fieldof medical genetics extends the search for patterns be-yond individual patients to inheritance within families;statistical genetics and genetic epidemiology use quan-titative methods to analyze these patterns in families andpopulations. Now the tools of molecular genetics are beingused to identify thousands of mutations associated withhundreds of mostly rare genetic diseases. In the pro-cess, they are revealing unexpected genetic complexity,blurring the distinction between Mendelian and com-plex disorders.2 New approaches are needed to find mean-ingful patterns in the outpouring of genomic data; we be-lieve that epidemiology will provide a crucial link betweendermatology and the human genome.

Three articles in this issue of the Archives describe novelmutations in genes associated with inherited dermato-ses: FGFR3 with acanthosis nigricans (AN),3 CYLD with

multiple familial trichoepithelioma,4 and FECH with eryth-ropoietic protoporphyria (EPP).5 Each of these findingsadds incrementally to the growing web of knowledge thatconnects variations in genetic sequences with biologicprocesses, clinical findings, and epidemiologic observa-tions. For example, some FGFR3 mutations that pro-duce amino acid substitutions at position 650 are knownto cause severe skeletal dysplasias; association of theK650R mutation with AN, a relatively mild condition, isconsistent with its small effect on function in a system-atic study of all possible FGFR3 missense mutations atthat locus.3 Variation in CYLD affects its negative regu-lation of nuclear factor–kappa B, a key signaling proteinin a pathway that has been implicated in cancer and manyother diseases.6 Confirmation that FECH mutations causeEPP only when coinherited with another low-expression allele is consistent with the autosomal domi-nant inheritance pattern observed in European popula-tions, where one such allele is common.7

How will larger patterns emerge from the accumula-tion of thousands of small studies? The first systematiccatalog of inherited diseases and their genetic causes waspublished in 1966 by Victor A. McKusick,8 a founder of

modern medical genetics. Now in its 12th edition, Men-delian Inheritance in Man8 has a continuously updated,online counterpart known as OMIM (Online MendelianInheritance in Man).9 The OMIM editors regularly ap-pend newly reported findings, creating essentially chrono-logical summaries. Currently, OMIM includes more than2000 summaries that refer to genes or syndromes asso-ciated with skin changes. In addition, various disease-or gene-focused databases have been created to serveparticular research communities, such as the HumanIntermediate Filament Database, which catalogs muta-tions in epithelial keratin and related genes.10 Several jour-nals, including Human Mutation and Human Genetics, pub-lish brief descriptions of novel mutations along with namesof authors and digital object identifiers (DOIs) that per-mit their retrieval by PubMed. The Human Genome Varia-tion Society (HGVS), which promotes the “collection, docu-mentation and free distribution of genomic variationinformation and associated clinical variations,”11 recentlyannounced a new initiative, the Human Variome Project,which aims to catalog all human mutations worldwide andmake the information freely available online.12

Most mutation databases are essentially lists based oninvestigations of the genetic characteristics of indi-vidual persons and families—the information equiva-lent of a case series. In contrast, epidemiologic studiesthat link a genotype with phenotypic and environmen-tal variables produce multidimensional data. The epide-miologic approach, which is based on precise defini-tions, careful sampling, and systematic analysis, allowsinvestigators to interpret their findings in a context largerthan that of individual persons or families. Molecular ge-netic analysis became practical for use in epidemiologicstudies only in the last 10 years or so; however, pub-lished reports already number in the thousands. The mostpopular design for studies of genetic associations com-pares the distributions of genotypes between affected andunaffected groups, which is equivalent to an epidemio-logic case-control study. Because the inherited geno-type is invariant, studies of gene-disease associations areable to avoid bias related to concurrent ascertainment ofcause and effect, which is an important limitation of tra-ditional case-control studies. When based on sound epi-demiologic research methods, such studies can providea strong basis for inference.13

The term human genome epidemiology (HuGE) was pro-posed to distinguish population-based analyses of geno-type frequencies, gene-disease associations, and gene-gene and gene-environment interactions from thetraditional studies of genetic epidemiology, which fo-




cus on heritability analysis and gene discovery.14 TheHuGE Network (HuGENet) is an international, volun-tary collaboration for promoting this population per-spective and encouraging the synthesis of published andunpublished data by consortiums of investigators.15 Inaddition to devising a “road map” for this endeavor,16

HuGENet has focused on developing tools to support re-searchers, including a curated, searchable, online data-base of PubMed abstracts reporting studies in human ge-nome epidemiology.17,18 As of April 16, 2007, the databasecontained abstracts for 26 635 articles, including 247 re-lated to dermatology. These articles, which were pub-lished in more than 50 journals in the United States, Eu-rope, and Asia, examined 145 genes and more than 100conditions, of which the most frequently studied werepsoriasis (61 articles), melanoma (34 articles), and atopicdermatitis (22 articles).

The systematic review and synthesis of informationon human genetic variation in relation to health and dis-ease is increasingly recognized as crucial to realizing thepromised medical and public health benefits of ge-nomic research.16 This challenge is too large and com-plex to undertake individually or to base on a single dis-cipline, methodology, database, or other infrastructure;a networked approach may have a better chance of suc-cess.19 Advocacy groups like PXE International can playan important role by creating registries and organizingresearch consortiums.20 By promoting partnerships amongclinicians, basic scientists, and epidemiologists, the fieldof dermatology can help promote the productive inte-gration and translation of research findings into betterpreventive medicine and clinical care. The editors of theArchives of Dermatology would like to encourage the sub-mission of articles that synthesize information on ge-netic variants in relation to diseases, both common andrare, in the form of HuGE reviews and meta-analyses thatwill integrate these findings into a larger context. An on-line handbook published by HuGENet and available on-line provides methodologic guidance in developing, con-ducting, and reporting individual associations and theirintegration into systematic reviews.21

Correspondence: Dr Gwinn, National Office of PublicHealth Genomics, Centers for Disease Control and Pre-

vention, 4770 Buford Hwy, MS K-89, Atlanta, GA 30341([email protected]).Financial Disclosure: None reported.

REFERENCES

1. Templeton HJ Skin lesions as diagnostic aids in general medicine. http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=1655038. Accessed April 16, 2007.

2. Sidransky E. Heterozygosity for a Mendelian disorder as a risk factor for com-plex disease. Clin Genet. 2006;70(4):275-282.

3. Berk DR, Spector EB, Bayliss SJ. Familial acanthosis nigricans due to K650T FGFR3mutation. Arch Dermatol. 2007;143(9):1153-1156.

4. Espana A, Garcıa-Amigot F, Aguado L, Garcıa-Foncillas J. A novel missense mu-tation in the CYLD gene in a Spanish family with multiple familial trichoepithelioma.Arch Dermatol. 2007;143(9):1209-1210.

5. Herrero C, To-Figueras J, Badenas C, et al. Clinical, biochemical, and genetic studyof 11 patients with erythropoietic protoporphyria including one with homozy-gous disease. Arch Dermatol. 2007;143(9):1125-1129.

6. Courtois G, Gilmore TD. Mutations in the NF-kappaB signaling pathway: impli-cations for human disease. Oncogene. 2006;25(51):6831-6843.

7. Gouya L, Puy H, Robreau AM, et al. The penetrance of dominant erythropoieticprotoporphyria is modulated by expression of wildtype FECH. Nat Genet. 2002;30(1):27-28.

8. McKusick VA. Mendelian Inheritance in Man: A Catalog of Human Genes and Ge-netic Disorders. Baltimore, MD: Johns Hopkins University Press; 1998.

9. OMIM—Online Mendelian Inheritance in Man. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=OMIM. Accessed April 16, 2007.

10. Human Intermediate Filament Database. http://www.interfil.org. Accessed April16, 2007.

11. HGVS—Human Genome Variation Society. http://www.hgvs.org. Accessed April16, 2007.

12. Cotton RG; Participants in the 2006 Human Variome Project meeting. Recom-mendations of the 2006 Human Variome Project meeting. Nat Genet. 2007;39(4):433-436.

13. Hattersley AT, McCarthy MI. What makes a good genetic association study? Lancet.2005;366(9493):1315-1323.

14. Khoury MJ, Little J, Burke W, eds. Human Genome Epidemiology: A ScientificFoundation for Using Genetic Information to Improve Health and Prevent Disease.New York, NY: Oxford University Press; 2004.

15. National Office of Public Health Genomics. Human Genome Epidemiology Network.http://www.cdc.gov/genomics/hugenet/. Accessed April 16, 2007.

16. Ioannidis JP, Gwinn M, Little J, et al. A road map for efficient and reliable humangenome epidemiology. Nat Genet. 2006;38(1):3-5.

17. GDPInfo. Genomics and Disease Prevention Information System. http://apps.nccd.cdc.gov/genomics/GDPQueryTool/default.asp. Accessed April 16, 2007.

18. Lin BK, Clyne M, Walsh M, et al. Tracking the epidemiology of human genes inthe literature: the HuGE Published Literature database. Am J Epidemiol. 2006;164(1):1-4.

19. Ioannidis JP, Bernstein J, Boffetta P, et al. A network of investigator networks inhuman genome epidemiology. Am J Epidemiol. 2005;162(4):302-304.

20. Terry SF, Terry PF, Rauen KA, Uitto J, Bercovitch LG. Advocacy groups as re-search organizations: the PXE International example. Nat Rev Genet. 2007;8(2):157-164.

21. Little J, Higgins J, eds. The HuGENet HuGE Review Handbook, Version 1.0. http://www.genesens.net/_intranet/doc_nouvelles/HuGE%20Review%20Handbook%20v11.pdf. Accessed April 16, 2006.

Marta Gwinn, MD, MPHMuin J. Khoury, MD, PhD



Editor’s Note

The Next Frontier of Dermatologic Research

Recently, dermatology research has identified new genetic alterations associated with diseases. In 2005, the Archives wasprivileged to publish the first of these reports, which can be summarized as “name that tune” by identifying the single er-roneous note causing the disease.1 Clinicians longed for progress beyond the stupefying identification of variants of geneticdisorders, dared to hope for gene therapy for dermatologic diseases, and questioned the relevance of ever increasing reportsof 2 or 3 cases with a few codons different from other cases of the same disease. The “tune” can be named in 1 note. Now, wehear the sweet melody of therapy in at least 1 case of junctional epidermolysis bullosa!2

Our next frontier is prevention of dermatologic diseases by the identification and modification of environmental risk factorsamong persons susceptible to disease due to genotype.3 To realize this goal, the principles of epidemiologic research will be ap-plied to molecular genetic analysis of populations.4 The Archives invites HuGE Reviews and meta-analyses submitted under thearticle category “Reviews” as described in our Instructions for Authors at http://archderm.ama-assn.org/misc/ifora.dtl.

June K. Robinson, MDEditor

1. Gao M, Wang PG, Yang S, et al. Two frameshift mutations in the RNA-specific adenosine deaminase gene associated with dyschromatosis symmetrica hereditaria.Arch Dermatol. 2005;141(2):193-196.

2. Mavilio F, Pellegrini G, Ferrari S, et al. Correction of junctional epidermolysis bullosa by transplantation of genetically modified epidermal stem cells.Nat Med. 2006;12(12):1397-1402.

3. Coughlin SS. The intersection of genetics, public health, and preventive medicine. Am J Prev Med. 1999;16(2):89-90.4. Gwinn M, Khoury MJ. Dermatology and the human genome: an epidemiologic approach. Arch Dermatol. 2007;143(9):1194-1195.



OFF-CENTER FOLD

SECTION EDITORS: MICHAEL E. MING, MD, MSCE; ASSISTANT SECTION EDITORS: CARRIE ANN R. CUSACK, MD; SENAIT W. DYSON, MD;JACQUELINE M. JUNKINS-HOPKINS, MD; VINCENT LIU, MD; KARLA S. ROSENMAN, MD

Hyperpigmented Keratotic NodulesSer Ling Chua, MRCP; Kusum Kulkarni, MRCPath; Eric Saihan, FRCP; Queen’s Medical Centre, Nottingham, England

REPORT OF A CASE

A 59-year-old woman of Indian descent presented witha 6-month history of an increasing number of itchy hy-perpigmented nodules on her trunk and limbs. The le-sions were preceded by pruritus. Treatment with topi-cal 0.05% clobetasone butyrate cream and oral cetirizine

reduced the pruritus but did not affect the appearanceof the lesions. The patient had a history of type 2 diabe-tes mellitus and hypertension. Ultrasonography showedthat her liver was enlarged, with a fatty echo texture.

Physical examination revealed multiple 3- to15-mm-diameter hyperpigmented lesions with a centralkeratotic plug (Figure 1). The Koebner phenomenon

was seen (Figure 2). There was sparing of her backin “difficult-to-reach” areas. A 4-mm punch biopsyspecimen obtained from a lesion on the left upperback area was sent for histopathologic examination(Figure 3).

What is your diagnosis?

Verrucous Papules and Plaques in a Pediatric PatientMercedes E. Gonzalez, MD; Fiona P. Blanco, MD; Maria C. Garzon, MD; Columbia University, New York, New York

REPORT OF A CASE

A 12-year-old boy presented with a several-month his-tory of enlarging verrucous lesions on his face and leftforearm. Similar lesions had first appeared on the cen-tral area of his face 3 years earlier and had been excised.His medical history included failure to thrive, which re-quired gastrostomy tube feeding; Arnold-Chiari malfor-mation, which had been corrected when he was 1 yearold; idiopathic hypertrophic subaortic stenosis, which re-quired 2 corrective surgical procedures; global develop-mental delay; and hypertension, which required pro-pranolol therapy. His family history was unremarkable.

Physical examination revealed coarse facial features,large lips, a short neck, and 4 facial verrucous papules(Figure 1). Follicular papules were present over thelateral surface of both arms and on the face. The palmswere hyperkeratotic, with accentuation of the creases.There was a large 2�2-cm verrucous plaque on the

extensor surface of the left forearm (Figure 2), fromwhich a shave biopsy specimen was obtained.


Diffuse Nodules in a Woman With Renal FailureErin B. Lesesky, MD; Michelle T. Pelle, MD; Terence C. O’Grady, MD; University of California, San Diego

REPORT OF A CASE

A 47-year-old woman presented with a 1-year history ofan erythematous plaque on her left leg. The plaque wasnot painful when it first developed, but it became ten-der when it ulcerated. She denied trauma and systemicsymptoms. There was no previous treatment. Her medi-cal history was significant for type 2 diabetes, chronicrenal failure, cardiomyopathy, and gouty arthritis.

Physical examination revealed an 11�21-cm ery-thematous plaque on the lower area of the left leg, withscattered firm, white papules and areas of ulceration(Figure 1), and a 7-cm, rubbery, nontender, reddish yel-low nodule on the dorsal aspect of the right hand(Figure 2). Also, pink to white firm nodules were seenin a linear distribution on the lateral aspect of the fin-gers, the dorsal aspect of both hands, and the extensoraspect of the forearms. A punch biopsy specimen was

obtained from the plaque on the left lower leg area(Figure 3).


Tender Nodules on the Palms and SolesLisa Esler-Brauer, MD; Ilene Rothman, MD; University at Buffalo School of Medicine and The Women and Children’s Hospital of Buffalo, Buffalo, New York

REPORT OF A CASE

A 10-year-old white boy presented with a 3-day historyof painful, red, swollen hands. Physical examinationrevealed multiple, erythematous, 1- to 3-cm, tender nod-ules on his palms (Figure 1). A few similar smaller

nodules were present on the plantar aspect of the middleof both heels. The patient had an associated low-gradefever but was otherwise healthy and active. He hadrecently spent a lot of time in ice hockey practice. Hisprimary care physician had initially evaluated him, andthe results of rapid streptococcal screening were

negative, no murmur was appreciated, and no treatmenthad been initiated. A 4-mm punch biopsy specimen wasobtained from a palmar lesion for microscopic evalua-tion (Figure 2 and Figure 3).


Figure 1. Figure 2. Figure 3.

Figure 1. Figure 2.



(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), SEP 2007 WWW.ARCHDERMATOL.COM1201-1206


Hyperpigmented Keratotic NodulesDiagnosis: Acquired perforating collagenosis.

MICROSCOPIC FINDINGS

Microscopic examination showed a cup-shaped defor-mity of the epidermis that was filled with a plug consist-ing of collagen, parakeratosis, and inflammatory exu-date. The underlying epidermis was atrophic, but theadjacent epidermis was acanthotic. Dermal collagen atthe base of the plug was almost perpendicular to the over-lying epidermis. There was a sparse mixed inflamma-tory exudate around the upper dermal vessels.

DISCUSSION

The main histologic abnormality in perforating derma-toses is the transepithelial elimination of material fromthe dermis. There are 4 main types of primary perforat-ing dermatoses: Kyrle disease, elastosis perforans ser-piginosa, reactive perforating collagenosis, and perforat-ing folliculitis. Perforating dermatoses have also beenreported in association with chronic renal failure and dia-betes mellitus. The term acquired perforating dermatoseshas been used to describe them. The lesions are typi-cally dome-shaped papules and nodules with a centralkeratotic plug. They have been considered a distinct

entity from primary perforating dermatoses owing to theirdifferences in clinical and histologic presentation fromthe original descriptions of the primary perforating der-matoses.1-3

Acquired perforating dermatosis has been reported tooccur in 4.5% to 10% of patients receiving hemodialysisin North America. In Dundee, Scotland, the reportedprevalence was 11% in patients who were undergoing di-alysis, 78% of whom were diabetic.2 In most patients, pru-ritus precedes or develops at the same time as the skinlesions. The Koebner phenomenon is frequently ob-served. Acquired perforating dermatosis is also associ-ated with lymphoma, nephropathy of various causes (in-cluding obstructive nephropathy, glomerulonephritis,hypertensive nephrosclerosis, heroin use, and human im-munodeficiency virus infection), hypothyroidism, hy-perparathyroidism, neurodermatitis, liver dysfunction,and hypertension.4

The pathogenesis of acquired perforating dermatosisis unknown. Trauma due to scratching has been postu-lated as a precipitating factor. This theory is supportedby the linear distribution and reproducibility of the le-sions by scratching of the skin.4 In diabetic patients, ithas been suggested that scratching results in localized der-mal necrosis due to poor blood supply as a result of dia-betic vasculopathy. The necrotic tissue is then extruded

through the epidermis.5 In patients with renal impair-ment, the dermal microdeposition of substances such asuric acid has been suggested as a cause of acquired per-forating dermatosis.6

To our knowledge, there is no well-established therapyfor acquired perforating dermatosis. In some patients, theskin lesions resolve spontaneously. There have been re-ports of successful treatment of both primary and ac-quired perforating dermatoses with topical, intrader-mal, and systemic corticosteroids; topical and systemicretinoids; UV-B phototherapy; methotrexate; and allo-purinol.1,2,4

REFERENCES

1. Patterson JW, Richmond VA. The perforating disorders. J Am Acad Dermatol. 1984;10(4):561-581.

2. Morton CA, Henderson IS, Jones MC, et al. Acquired perforating dermatosis in aBritish dialysis population. Br J Dermatol. 1996;135(5):671-677.

3. Rapini RP, Hebert AA, Drucker CR. Acquired perforating dermatosis. Arch Dermatol.1989;125(8):1074-1078.

4. Faver IR, Daoud MS, Daniel Su WP. Acquired reactive perforating collagenosis.J Am Acad Dermatol. 1994;30(4):575-580.

5. Cochran RJ, Tucker SB, Wilkin JK. Reactive perforating collagenosis of diabetesmellitus and renal failure. Cutis. 1983;31(1):55-58.

6. Haftek M, Euvrard S, Kanitakis J, et al. Acquired perforating dermatosis of diabe-tes mellitus and renal failure: further ultrastructural clues to its pathogenosis.J Cutan Pathol. 1993;20(4):350-355.

Verrucous Papules and Plaques in a Pediatric PatientDiagnosis: Cutaneous papillomas associated withCostello syndrome.

MICROSCOPIC FINDINGSAND CLINICAL COURSE

Histopathologic examination of the left arm lesion showeda hyperplastic epidermis with vascular ectasia and a mono-nuclear cell infiltrate in the dermis. The findings wereconsistent with verrucous acanthoma (papilloma). Thepatient was referred to the plastic surgery department todiscuss treatment options for the facial papillomas.

DISCUSSION

Approximately 150 cases of Costello syndrome havebeen identified since the condition was described in 1977by Costello1 and later named by Der Kaloustian et al2 in1991. The major features include growth and mental re-tardation, with a sociable personality; coarse facies(macrocephaly, thick lips, big tongue, depressed nasalbridge, and low-set ears with thick lobes); short neck;musculoskeletal abnormalities (tight Achilles tendon, hy-perextensible fingers, and limited joint mobility), car-diac defects (hypertrophic cardiomyopathy, cardiacmalformations, and supraventricular arrhythmias); in-creased risk for malignant tumors (rhabdomyosarcoma,ganglioneuroblastoma, neuroblastoma, and bladdercarcinoma); and distinctive cutaneous findings.3 Cuta-neous papillomas are a unique skin manifestation of Cos-tello syndrome, although they are present in only about

50% of patients.4 Papillomas occur on the face, anus, ax-illae, elbows, knees, vocal cords, and abdomen and usu-ally develop between the 2nd and 15th years of life. Thepathogenesis is unknown, but mechanical stress and hu-man papillomavirus have been implicated.5 Clinically, thepapillomas resemble verrucae, but histologically, verru-cae can be distinguished by the presence of koilocytes,vertical columns of parakeratosis, and foci of clumpedkeratohyaline granules. In addition to verrucae, the dif-ferential diagnosis includes epidermal nevus, verrucouscarcinoma, keratoacanthoma, actinic keratosis, squa-mous cell carcinoma, lipoid proteinosis, deep fungal in-fections, and epidermodysplasia verruciformis. Treat-ment options are limited to destruction and excision ofthe lesions, both of which may result in profound scar-ring. Patients with Costello syndrome also invariably haveloose, redundant skin, especially of the neck, hands, andfeet, which results in the appearance of deep creases onthe palms and soles, as seen in our patient. Acanthosisnigricans, pigmented acral nevi, vascular birthmarks, hy-perkeratosis, hyperpigmentation of the skin, thin deep-set nails, thick eyebrows, and sparse curly scalp hair alsofrequently occur.4

Historically, Costello syndrome was considered avariant of Noonan syndrome and was synonymous withcardiofaciocutaneous syndrome, but these 3 conditionswere found to have separate genetic loci.6 The derange-ment in Costello syndrome is a de novo mutation of theHRAS gene (v-Ha-ras Harvey rat sarcoma viral onco-gene homologue), located on chromosome 11, resultingin permanent activation of the HRAS protein anduncontrolled cell division.7 Our patient was diagnosed

as having both Noonan syndrome and cardiofaciocuta-neous syndrome before receiving the diagnosis of Cos-tello syndrome at the age of 6 years. The average age atdiagnosis in 1 review was 14 years, which further indi-cates that establishing the correct diagnosis may be achallenge.8 Genetic testing is now possible, but thediagnosis of Costello syndrome still relies heavily onclinical recognition of the pattern of malformations.Characteristic skin features provide early clues to thediagnosis, thereby facilitating the prevention and man-agement of the more serious systemic effects of Costellosyndrome.

REFERENCES

1. Costello JM. A new syndrome: mental subnormality and nasal papillomata. AustPaediatr J. 1977;13(2):114-118.

2. Der Kaloustian VM, Moroz B, McIntosh N, Watters AK, Blaichman S. Costellosyndrome. Am J Med Genet. 1991;41(1):69-73.

3. Costello syndrome. http://www.ncbi.nlm.nih.gov/entrez/dispomim.cgi?id=218040. Accessed June 11, 2007.

4. Hennekam RC. Costello syndrome: an overview. Am J Med Genet C Semin MedGenet. 2003;117(1):42-48.

5. Torrelo A, Lopez-Avila A, Medeiro IG, Zambrano A. Costello syndrome. J Am AcadDermatol. 1995;32(5, pt 2):904-907.

6. Wieczorek D, Majewski F, Gillessen-Kaesbach G. Cardio-facio-cutaneous (CFC)syndrome—a distinct entity? report of three patients demonstrating the diagnos-tic difficulties in delineation of CFC syndrome. Clin Genet. 1997;52(1):37-46.

7. Gripp KW, Lin AE, Stabley DL, et al. HRAS mutation analysis in Costello syn-drome: genotype and phenotype correlation. Am J Med Genet A. 2006;140(1):1-7.

8. White SM, Graham JM Jr, Kerr B, et al. The adult phenotype in Costello syndrome.Am J Med Genet A. 2005;136(2):128-135.

Diffuse Nodules in a Woman With Renal FailureDiagnosis: Chronic tophaceous gout.

MICROSCOPIC FINDINGSAND CLINICAL COURSE

The biopsy specimen revealed amorphous pink to graymaterial in the dermis, surrounded by lymphocytes andmultinucleated foreign body–type giant cells, with no evi-dence of vasculitis or panniculitis. A von Kossa stain wasnegative for calcium. These microscopic findings wereconsistent with a diagnosis of a gouty tophus. Labora-tory data revealed normal levels of serum calcium andphosphorus and an elevated level of uric acid (12.2 mg/dL[to convert to micromoles per liter, multiply by 59.485];reference range, 2.6-7.5 mg/dL).

Oral allopurinol therapy (100 mg every other day; al-ternate-day dosing because of renal failure) was initi-ated, and hemodialysis was started 3 months later. Thepatient’s uric acid level decreased markedly to 5.6 mg/dLover a 5-month period. No new tophi have developed sincethe allopurinol therapy was begun. In this case, transepi-dermal elimination of the tophi caused multiple largeulcers and subsequent pyodermas that yielded methicillin-resistant Staphylococcus aureus and vancomycin-resistant Enterococcus faecalis on culture. Meticulouswound care, frequent surveillance for infection, allopu-rinol, and antibiotics as needed are the mainstays of ourpatient’s treatment.

DISCUSSION

Gout, one of the cutaneous deposition disorders, is a meta-bolic disease in which the underlying abnormality is hy-peruricemia. The clinical manifestations include acutegouty arthritis, cutaneous tophi, and uric acid nephro-lithiasis and nephropathy.1 Hyperuricemia can result fromoverproduction or underexcretion of uric acid, which isthe end product of the catabolism of purines and is pri-marily excreted by the kidneys.1 Predisposing factors forthe development of gout include dietary overindul-gence, obesity, excessive alcohol intake, renal insuffi-ciency, and ingestion of certain medications, includingdiuretics and cyclosporine.2,3 The 3 classic stages of gout,which can occur after years of asymptomatic hyperuri-cemia, are acute gouty arthritis; intercritical gout, whichincludes the intervals between attacks; and chronic to-phaceous gout.2

Chronic tophaceous gout has a prevalence of less than5% among patients with gout because of the availabilityof effective antihyperuricemic medications.2 The aver-age interval from the first gouty attack to the onset of to-phi is 12 years.4 Tophi present as firm, smooth or mul-tilobulated, white to red dermal or subcutaneous nodules.They are most commonly located in the skin overlyingjoints or the helix of the ear and are often nontender.1

Diagnosis of gout during an acute attack is made bydemonstration of urate crystals in synovial fluid.

However, urate crystals can also be obtained from atophus or visualized histologically in excised tissue.Ideally, specimens are placed in alcohol, which pre-serves the crystals, but formalin-fixed tissue also exhib-its key features, including deposits of amorphous mate-rial containing needlelike clefts within the dermis orsubcutis and a surrounding infiltrate of lymphocytesand foreign body–type giant cells.1 Treatment of acuteattacks includes colchicine and nonsteroidal anti-inflammatory drugs. Chronic tophaceous gout is besttreated by allopurinol and reduction of risk factors,with a goal of lowering the serum uric acid level to lessthan 6 mg/dL, which will prevent further attacks andcan lead to resorption of tophi.5 Our patient improvedmarkedly with allopurinol therapy and the discontinua-tion of diuretic treatment once hemodialysis wasstarted.

REFERENCES

1. Franke I, Gollnick H. Deposition diseases. In: Bolognia J, Jorizzo J, Rapini R, eds.Dermatology. New York, NY: Mosby Elsevier Ltd; 2003:669-672.

2. O’Duffy JD, Hunder GG, Kelly PJ. Decreasing prevalence of tophaceous gout. MayoClin Proc. 1975;50(5):227-228.

3. Fam AG, Assaad D. Intradermal urate tophi. J Rheumatol. 1997;24(6):1126-1131.4. Gutman AB. The past four decades of progress in the knowledge of gout, with an

assessment of the present status. Arthritis Rheum. 1973;16(4):431-445.5. Yamanaka H, Togashi R, Hakoda M, et al. Optimal range of serum urate concen-

trations to minimize risk of gouty attacks during anti-hyperuricemic treatment.Adv Exp Med Biol. 1998;431:13-18.

Tender Nodules on the Palms and SolesDiagnosis: Palmoplantar eccrine hidradenitis(PEH).

MICROSCOPIC FINDINGS

Examination of the hematoxylin-eosin–stained skin bi-opsy specimen from the erythematous tender nodule re-vealed a dense neutrophilic inflammatory infiltrate aroundthe eccrine sweat glands.

DISCUSSION

Palmoplantar eccrine hidradenitis is a distinct entity withcharacteristic clinical findings. First reported in 1988, itwas named traumatic plantar urticaria.1 Since then, 47 caseshave been reported under 5 different names, including id-iopathic (palmo)plantar hidradenitis and recurrent pal-moplantar hidradenitis.2 It primarily affects healthy chil-dren and is characterized by an abrupt appearance oferythematous, tender nodules on the palms and soles thatmay last 1 to 4 weeks and cause considerable difficulty inwalking.2 Most eruptions are limited to the plantar sur-faces, although palmar lesions have been documented.3

Typical histiologic findings include a neutrophilic periec-crine infiltrate in the setting of varying degrees of super-ficial and deep perivascular infiltrates of neutrophils, lym-phocytes, and histiocytes as well as septal panniculitis.3

The etiology of PEH remains unknown. Local me-chanical or thermal trauma, moisture, recent streptococ-cal infection, and hypersensitivity reaction have all beenconsidered.3 Naimer et al4 postulate that it is exposure

to wet footwear that triggers PEH, as a correlation was notedbetween the eruption and strenuous activity, such as bal-let dancing. Interestingly, the majority of the lesions in ourcase were on the palmar surface and may be attributed tothe effect of hockey glove use that created a moist envi-ronment similar to a shoe. Some speculate that excessivesweating, whether intrinsic (hyperhidrosis) or extrinisic(exertion induced), may cause obstruction of the eccrineduct, with subsequent rupture leading to the inflamma-tory cascade.5 This could explain eruptions related to hottub use. The differential diagnosis of PEH includes ery-thema nodosum or other atypical panniculitis, erythemamultiforme, insect bites, vasculitis, chilblains, migratoryangioedema, embolic disease, and cellulitis.

Most patients respond to bed rest, which presump-tively decreases sweat secretion to allow resolution of in-flammation. Gradual resumption of activity after resolu-tion of lesions is recommended. Relapses occur in 50% ofpatients.6 Topical and systemic steroids as well as sys-temic antibiotics and nonsteroidal anti-inflammatory drugshave been used to treat PEH, without clear benefit.2,3 Pal-moplantar eccrine hidradenitis is a characteristic clinicalentity that is benign and self-limiting. It is likely more com-mon than has been realized in the pediatric population andshould be recognizable by dermatologists based on its dis-tinctive presentation and clinical findings.

REFERENCES

1. Metzker A, Brodsky F. Traumatic plantar urticaria—an unrecognized entity? J AmAcad Dermatol. 1988;18(1, pt 1):144-146.

2. Rubinson R, Larralde M, Santos-Munoz A, Parr V, de Parra NP. Palmoplantar ec-crine hidradenitis: seven new cases. Pediatr Dermatol. 2004;21(4):466-468.

3. Landau M, Metzker A, Gat A, Nen-Amitai D, Brenner S. Palmoplantar eccrine hi-dradenitis: three new cases and review. Pediatr Dermatol. 1998;15(2):97-102.

4. Naimer SA, Zvulunov A, Ben-Amitai D, Landau M. Plantar hidradenitis in childreninduced by exposure to wet footwear. Pediatr Emerg Care. 2000;16(3):182-183.

5. Ben-Amitai D, Hodak E, Landau M, Metzker A, Feinmesser M, David M. Idiopathicpalmoplantar eccrine hidradenitis in children. Eur J Pediatr. 2001;160(3):189-191.

6. Simon M Jr, Cremer H, von den Driesch P. Idiopathic recurrent palmoplantar hi-dradenitis in children: report of 22 cases. Arch Dermatol. 1998;134(1):76-79.

Submissions

Clinicians, local and regional societies, and residents andfellows in dermatology are invited to submit quiz casesto this section. Cases should follow the established pat-tern of a Report of a Case section of less than 150 words,followed by the Diagnosis, Microscopic Findings, and Dis-cussion. The discussion should be between 285 and 350words. References are limited to 9. The text should be sub-mitted double-spaced, with the right margin ragged. Pho-tomicrographs and illustrations must be wider than theyare tall (horizontal orientation), sharply focused with goodcolor balance, and submitted as separate JPG files with eachfile numbered with the figure number. Material must beaccompanied by the required copyright transfer state-ment (see authorship form in Instructions for Authors[http://archderm.ama-assn.org/misc/auinst_crit.pdf]).Manuscripts should be submitted via our online manu-script submission and review system (http://manuscripts.archdermatol.com). Please indicate in your cover letterthat the manuscript is a submission to Off-Center Fold.

(REPRINTED) ARCH DERMATOL/ VOL 143 (NO. 9), SEP 2007 WWW.ARCHDERMATOL.COM1201-1206


CORRESPONDENCE

RESEARCH LETTERS

A Novel Missense Mutationin the CYLD Gene in a Spanish FamilyWith Multiple Familial Trichoepithelioma

B rooke-Spiegler syndrome (BSS) includes the com-bination of spiradenomas, cylindromas, andtrichoepitheliomas. It has been postulated that

BSS results from defects in the regulation of putative stemcells of the folliculosebaceous-apocrine unit.1 This fol-licular dysregulation may give rise to 3 different geno-dermatoses: familial cylindromatosis (FC), multiple fa-milial trichoepithelioma (MFT), or the classic triad of BSS.

The gene for FC was mapped to chromosome16q12-q13.2 At present, 32 different germline muta-tions in the CYLD gene have been described,3 20 infamilies with FC, 8 in families with MFT, and 4 in fami-lies with BSS. Taken together, these observations sug-gest that these inherited syndromes associated with skinappendage tumors not only share a common genetic ba-sis but also may represent phenotypic variation of the samedisease.4

Report of a Case. An 8-year-old girl came to our depart-ment with her parents. Her mother, grandmother, and 2aunts were diagnosed as having trichoepithelioma sincechildhood (Figure 1). The parents of our patient wantedto know if their daughter might harbor any genetic sus-ceptibility for these cutaneous lesions.

Blood samples were obtained from available familymembers and 110 unrelated controls. Genomic DNA wasextracted; all coding exons were amplified by polymer-

ase chain reaction; and further sequencing analysis wasperformed. We identified 1 mutation not previously re-ported. The mutation was found in all patients but notin the healthy members of this family. The change was apoint mutation in exon 20 (G2687C) that resulted in sub-stitution of glycine at 896 by alanine (Figure 2). Themutation was not detected in 110 unrelated controls.

Comment. Herein, we report a novel CYLD genemutation at nucleotide 2687 that carries out 1 aminoacid change at glycine 896 in the 4 affected membersof this family but not in the proband. The fact that wehave not detected this change in 110 unaffected controlsmakes a contribution to the genotype-phenotype corre-lation in MFT.

The CYLD gene is considered a negative regulator ofnuclear factor kappa B(NF-�B).5 Thereby, inhibition or

1G/A

1G/A

2G/A

3G/A

1G/G

I.

II.

III.

A B

Figure 1. Clinical picture (A) and family pedigree (B) for the present case. A,Multiple trichoepitheliomas are present on the upper eyelids, nose, nasolabialfolds, and the upper lip in the mother of the proband. B, Pedigree of thefamily: for individuals whose DNA samples have been analyzed, the allelesequences at codon 896 have been indicated as G (glycine) or A (alanine).The proband (arrow) encodes for a G on both alleles (G/G), whereas theaffected family members are heterozygous for the mutation (black circles)encode an A on the mutant allele.

A B

C C C C C C C CCC A A A A A A A A A AG G G G G G G G G G S G G G GGT T T T T T T T T T T T TT

Wild-type sequence Mutant sequenceG896AExon 20Intron 19

Figure 2. Genetic analysis of the family with multiple familial trichoepithelioma. Shown are wild-type DNA (A) and mutant sequences (B) of exon 20 of CYLD fromcontrol and affected members, respectively.




inactivationofCYLDenhancestheactionofNF-kBandleadsto increased resistance to apoptosis and carcinogenesis.6

To our knowledge, this mutation has not been previ-ously described. The known mutations of the CYLD geneare mostly located in the C-terminal portion.

Germline mutations display tissue-specific functionloss. Another possibility would be that the germline mu-tation determines the tissues where the preferred sec-ond hit occurs. When the second hit occurs in eccrine-apocrine cells, the patients become susceptible to multiplecylindromas; in hair follicle cells, the patients exhibit sus-ceptibility to multiple cylindromas or MFT.6

Our study shows the importance of mutation screen-ing of the CYLD gene in patients affected with FC and MFTas well as their relatives to identify early clinical manifes-tations. Analysis of control volunteers in this report con-firms the role of this missense mutation as the cause of thissyndrome. Further studies evaluating the effect of this mu-tation in animal models must be considered.

Correspondence: Dr Espana, Department of Dermatol-ogy, University Clinic of Navarra, University of Na-varra, PO Box 31080, Pamplona, Navarra, Spain ([email protected]).Financial Disclosure: None reported.Funding/Support: This study was supported in part bythe Spanish Academy of Dermatology.Role of the Sponsor: The sponsor had no role in the de-sign and conduct of the study, in the collection, analysisand interpretation of data, or in the preparation review,or approval of the manuscript.

1. Clarke J, Ioffreda M, Helm KF. Multiple trichoepithelioma: a folliculosebaceous-apocrine genodermatosis. Am J Dermatopathol. 2002;24(5):402-405.

2. Fenske C, Banerjee P, Holden C, Carter N. Brooke-Spiegler syndrome locusassigned to 16q12-q13. J Invest Dermatol. 2000;114(5):1057-1058.

3. The Human Gene Mutation Database at the Institute of Medical Genetics inCardiff [registration required]. http://www.hgmd.cf.ac.uk/ac/all/php?gene=CYLD. Accessed February 9, 2007.

4. Bowen S, Gill M, Lee DA, et al. Mutations in the CYLD gene in Brooke-Spiegler syndrome, familial cylindromatosis, and multiple trichoepithe-lioma: lack of genotype-phenotype correlation. J Invest Dermatol. 2005;124(5):919-920.

5. Regamey A, Hohl D, Liu JW, et al. The tumor suppressor CYLD interacts withTRIP and regulates negatively nuclear factor kB activation by tumor necrosisfactor. J Exp Med. 2003;198(12):1959-1964.

6. Liang YH, Gao M, Sun LD, et al. Two novel CYLD gene mutations in Chinesefamilies with trichoepithelioma and a literature review of 16 families with tricho-epithelioma reported in China. Br J Dermatol. 2005;153(6):1213-1215.

A Novel PTPN11 Gene Mutationin a Patient With LEOPARD Syndrome

I n 1969, Gorlin et al1 described an autosomal domi-nant syndrome encompassing multiple lentigines,electrocardiographic abnormalities, ocular hyper-

telorism, pulmonary stenosis, abnormal genitalia, retar-dation of growth, and sensorineural deafness, currentlyknown as LEOPARD syndrome (LS). Recently, it has beenreported that most cases of LS are probably related to het-erozygous mutations of PTPN11 (protein-tyrosine phos-

phatase, nonreceptor type 11), a gene encoding a tyrosine-phosphatase protein named SHP-2, with 2 particular “hotspots” in exons 7 and 12.2-5 Despite overlapping clinicalmanifestations, LS is distinct from Noonan syndrome, an-other PTPN11 gene mutation–related disorder but witha different mutation spectrum. Herein we report the firstcase to our knowledge of typical LS featuring a newPTPN11 gene mutation.

Report of a Case. A 39-year-old woman with a medicalhistory of deafness and a familial background of Downsyndrome in a sister was referred for evaluation of pig-mentary changes that first appeared during infancy as-sociated with mild facial dysmorphism. Clinical exami-nation disclosed multiple light or dark brown maculesof varied sizes scattered throughout her whole body sur-face including her face, palmoplantar areas, lips, and con-junctiva (Figure1). There were no lentigines on the othermucous membranes. Facial examination revealed hyper-telorism. Electrocardiography showed a first-degree atri-oventricular block, whereas heart ultrasound evalua-tion did not find any valve abnormality. Hearinginvestigations confirmed sensorineural deafness. Therewere no urogenital abnormalities, endocrinopathy, orgrowth retardation. A diagnosis of LS was establishedbased on the presence of 4 criteria.

After obtaining the patient’s consent, we undertookdirect sequencing of the PTPN11 coding region and dis-covered a previously undescribed (to our knowledge) het-erozygous missense mutation in exon 13, namely aG1493T transversion leading to an R498L change in aminoacid sequence (Figure2). No genetic analysis of her rela-tives could be carried out to establish a diagnosis of denovo or inherited mutation.

Comment. The SHP-2 phosphatase plays several impor-tant roles in cellular physiologic function, mainly in cellproliferation, differentiation, migration, and adhe-sion.6-8 This protein contains 2 main domains: a C-terminal protein-tyrosine phosphatase (PTP) domain in-volved in catalytic activity and 2 N-terminal Src homology2 (SH2) domains interacting with the PTP domain, keep-ing it folded and inactive (Figure 2).3 To our knowl-edge, only 7 PTPN11 mutations have been reported inpatients with LS, all of them in the PTP domain in exons7 (Y279C and Y279S), 12 (T468M and A461T), and 13

Figure 1. Numerous lentiginous elements scattered throughout the trunk.

Agustın Espana, MDFermin Garcıa-Amigot, PhDLeyre Aguado, MDJesus Garcıa-Foncillas, MD



(Q506P, Q510E, and Q510G).2,4,9-11 Two of them (Y279Cand T468M) represent more than 90% of the identifiedchanges. The mutational site in our patient is close to thePTP-SH2 domains interacting tract and to the previ-ously reported mutations affecting exon 13. Accord-ingly, this change is probably relevant as to LS patho-mechanisms. PTPN11 mutations in Noonan syndromeare likely to result in a gain of function with “perma-nent” catalytic activity of the protein, whereas recent re-sults unexpectedly suggest that LS-related PTPN11 mu-tations result in a decrease in the phosphatase functionof SHP-2.12-14 Understanding why topographically closelyrelated mutations on the same gene may have oppositefunctional consequences but still result in close pheno-types will be a rewarding challenge.

Correspondence: Dr Dereure, Department of Dermatol-ogy, University of Montpellier I, Hopital Saint-Eloi, 80Ave A. Fliche, 34295 Montpellier CEDEX 5, France([email protected]).Financial Disclosure: None reported.

1. Gorlin RJ, Anderson RC, Blaw M. Multiple lentigenes syndrome. Am J DisChild. 1969;117(6):652-662.

2. Digilio MC, Conti E, Sarkozy A, et al. Grouping of multiple-lentigines/LEOPARD and Noonan syndromes on the PTPN11 gene. Am J Hum Genet.2002;71(2):389-394.

3. Legius E, Schrander-Stumpel C, Schollen E, Pulles-Heintzberger C, Gewil-lig M, Fryns JP. PTPN11 mutations in LEOPARD syndrome. J Med Genet.2002;39(8):571-574.

4. Keren B, Hadchouel A, Saba S, et al. PTPN11 mutations in patients withLEOPARD syndrome: a French multicentric experience. J Med Genet. 2004;41(11):e117.

5. Ogata T, Yoshida R. PTPN11 mutations and genotype-phenotype correla-tions in Noonan and LEOPARD syndromes. Pediatr Endocrinol Rev. 2005;2(4):669-674.

6. Kroll J, Waltenberger J. The vascular endothelial growth factor receptorKDR activates multiple signal transduction pathways in porcine aortic en-dothelial cells. J Biol Chem. 1997;272(51):32521-32527.

7. Chang Y, Ceacareanu B, Dixit M, Sreejayan N, Hassid A. Nitric oxide-inducedmotility inaorticsmoothmusclecells: roleofproteintyrosinephosphataseSHP-2and GTP-binding protein Rho. Circ Res. 2002;91(5):390-397.

8. Maheshwari M, Belmont J, Fernbach S, et al. PTPN11 mutations in Noonansyndrome type I: detection of recurrent mutations in exon 3 and 13. HumMutat. 2002;20(4):298-304.

9. Conti E, Dottorino T, Sarkozy A, et al. A novel PTPN11 mutation in LEOPARDsyndrome. Hum Mutat. 2003;21(6):654.

10. Sarkozy A, Obregon MG, Conti E, et al. A novelPTPN11 gene mutation bridgesNoonan syndrome, multiple lentigines/LEOPARD syndrome and Noonan-like/multiple giant cell lesion syndrome. Eur J Hum Genet. 2004;12(12):1069-1072.

11. Digilio MC, Sarkozy A, Pacileo G, Limongelli G, Marino B, Dallapiccola B.PTPN11 gene mutations: linking the Gln510Glu mutation to the “LEOPARDsyndrome phenotype.” Eur J Pediatr. 2006;165(11):803-805.

12. Hanna N, Montagner A, Lee WH, et al. Reduced phosphatase activity of SHP-2in LEOPARD syndrome: consequences for PI3K binding on Gab1. FEBS Lett.2006;580(10):2477-2482.

13. Tartaglia M, Martinelli S, Stella L, et al. Diversity and functional conse-quences of germline and somatic PTPN11 mutations in human disease. AmJ Hum Genet. 2006;78(2):279-290.

14. Kontaridis MI, Swanson KD, David FS, Barford D, Neel BG. PTPN11 (SHP2)mutations in LEOPARD syndrome have dominant negative, not activating,effects. J Biol Chem. 2006;281(10):6785-6792.

COMMENTS AND OPINIONS

Sponsorship of Graduate MedicalEducation: One Successful Model

I n the March 2007 issue of the Archives, Loo et al1 dis-cuss the recruitment and retention of academic der-matologicclinician-educators.Theycall fornewideas

to help provide needed support for teaching and mentor-ship, suggesting that this will lead to better job satisfac-tion and retention rates. It is my strong belief that robustdepartmental mentoring programs supported by active,dedicated senior faculty members working in concert withjunior faculty members and residents is the key to increas-ing recruitment and retention of academic dermatologists.

Implementing effective role modeling and mentor-ship involves a wide array of necessary elements that Iwill not expand on other than to say that the educa-tional environment requires the philosophical, physi-cal, and monetary support of the leadership. From wheredoes the last of these, money, become available?

Medical schools and institutions concerned with theadvancement of medicine and the development of the next

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

3′5′

Y279 T468M R498L

N N-SH2 C-SH2 Catalytic region (PTP) C

3 104 112 216 221 524

Figure 2. Genomic structure of the PTPN11 gene and corresponding functional organization of the PTPN11-encoded SHP-2 phosphatase. Exons are representedas shaded boxes. Also shown are the main LEOPARD syndrome–associated mutations (hot spots Y279C and T468M) and our patient’s R498L mutation affectingexon 13. N and C are terminal indicators; PTP indicates protein-tyrosine phosphatase; SH2, Src homology 2.

Aurelie Du-Thanh, MDHelène Cave, PharmD, PhDDidier Bessis, MDCarine Puso, MDJean-Jacques Guilhou, MDOlivier Dereure, MD, PhD



generation of physicians cite their commitment to the triadof patient care, research, and education. The relative valuean institution places on each of these areas dictates whichwill be most scrutinized in evaluation, promotion, andretention decisions. If research is the driving value, Na-tional Institutes of Health grants and other funding sourcesprovide the currency to support these endeavors. Insti-tutions that claim superior patient care as their drivingforce have an obvious source of funds—those generatedby the activity itself. Education alone among the trium-virate is an unfunded priority.

How has graduate medical education survived to thispoint without a sponsor, a source of dollars seeminglynecessary to support any endeavor? Excess funds fromclinical revenue helped support education until the 1990s,but certainly such funds have disappeared with the ad-vent of managed care. There has been a move among asmall number of institutions to dedicate dollars to thismission and provide graded stipends based on teachingtime. However, this is not a widespread phenomenon,and the compensation is sparse relative to the effort ex-pended. Generally, then, graduate medical education isan unfunded priority that has been considered a callingand kept alive by the volunteer efforts of dedicated, sup-portive practitioners from the community as well as ide-alistic, committed faculty educators.

In Philadelphia, Pennsylvania, Paul Gross, MD(Figure), is an example of the former and an innovatorof a welcome new kind of support for graduate medicaleducation. Dr Gross graduated from the University of

Pennsylvania School of Medicine in 1962 and com-pleted his residency in dermatology at the same institu-tion in 1968. From then until 2006, he headed the der-matology service at one of our affiliated teaching hospitals,Pennsylvania Hospital, in the Center City neighbor-hood of Philadelphia. Without compensation, Paul an-chored the dermatology residency teaching rotation, pro-vided in-patient consultation services, and educated scoresof internists, family practitioners, and interns in his pri-vate practice office. He is a stalwart attendee at our weekly“Duhring Conference” grand rounds, sharing his clini-cal expertise to help diagnose and treat patients with un-common clinical diseases. He and his wife, Anita, are regu-lar attendees at all of our departmental get-togethers,graduation dinners, named lectures, and alumni re-unions. In short, he is the epitome of the dedicated, sup-portive practitioner who daily contributes voluntarily ofhis time and expertise to help guarantee the future of ourspecialty.

Dr Gross has, however, taken his support to a higherlevel, one that might also serve as a model to promote much-needed funding for full-time dermatologic clinician-educators and departmental education initiatives. He as-tutely recognizes not only that we need to educate youngacademic clinician-educators but also that to do so re-quires financial support. Acting on this insight, he has beenable to secure philanthropic contributions from his grate-ful patients to support an innovative Clinical-Educator Fel-lowship at the University of Pennsylvania. This postresi-dency experience helps interested dermatologists obtainneeded skills to excel in academic medicine.

In addition, Dr Gross understands that the teachingof dermatology residents requires the program directorto spend a growing amount of time to effectively fulfillthis role. A grateful patient has donated the funds to en-dow a chair for dermatologic education, with the con-dition that it be named in Dr Gross’ honor. Endowmentof professorial chairs to support basic science research-ers and/or chairs of departments is the most commonamong these types of generous philanthropic ventures.Chairs for directors of education are unusual, and it is atestament of Dr Gross’ vision that he was able to estab-lish one at the University of Pennsylvania. Such profes-sorships help the department to ensure that the third pil-lar of excellence, education, will be supported inperpetuity. While education is a shared responsibilityamong faculty members, as are all areas of scholarship,these gifts guarantee the permanent presence of a dedi-cated expert whose main mission is to bolster the train-ing program, mentor the young, fulfill the educationaladministrative requirements, and initiate innovative learn-ing experiences and techniques.

Dr Gross has provided a role model to our residentsof how important private practice volunteer educatorsare to academic departments. We are fortunate to havemore than 50 volunteers from the community who giveat least 50 hours of teaching a year to the University ofPennsylvania Department of Dermatology. I see a directconnection between these exceptional men and womenand the fact that the vast majority of our graduates overthe past 10 years are involved in full-time or voluntaryteaching activities.

Paul Gross, MD



The overlay of philanthropic monetary support di-rected to fellowships and professorships will help sustainour next generation of learners and educators in a differ-ent way. As has been documented, approximately 10% ofour clinician-educators leave academic medicine yearly.1

Residency program directors have a 15% to 20% turnover(unpublished data, W.D.J., 2004). Chairs for mid-level andsenior educators validate the importance of this methodof scholarship to the department and the university. Theyprovide a tangible goal to aspire to, help sustain careers,and stabilize tenure in important positions such as resi-dency program directorships.

I hope that Dr Gross’ vision, and the resulting posi-tive experience at the University of Pennsylvania, stimu-lates additional discussion of support for resident and stu-dent education and invigorates funding of their endeavorsin innovative ways.

Correspondence: Dr James, Department of Dermatol-ogy, Hospital of the University of Pennsylvania, 3600Spruce St, 2 Rhoads Pavilion, Philadelphia, PA 19104([email protected]).Financial Disclosure: None reported.

1. Loo DS, Liu CL, Geller AC, Gilchrest BA. Academic dermatology manpower:issues of recruitment and retention. Arch Dermatol. 2007;143(3):341-347.

Narrowband UV-B Phototherapyfor Extragenital Lichen Sclerosus

I n the January issue of the Archives, Colbert and col-leagues1 report on the successful use of narrow-band (NB) UV-B phototherapy in a patient with

widespread extragenital lichen sclerosus (LS). We fullyagree that NB UV-B is an attractive alternative for pa-tients with extragenital LS if UV-A1 devices are not avail-able. After having treated several patients with NB UV-B,we believe that at least 30 irradiation sessions (similarto NB UV-B treatment for localized scleroderma) are nec-essary to produce significant changes in the clinical sta-tus.2 Importantly, patients must be advised that their skindisease might appear “deteriorated” during photo-therapy because the healthy surrounding skin tans morestrongly than LS lesions.

Colbert and colleagues1 have speculated that the mainmechanism of action of NB UV-B in LS might be the in-crease of matrix metalloproteinase (MMP) levels, fi-nally resulting in a reduction of skin sclerosis. This in-spired us to perform MMP-1 immunostaining in a patientwith LS treated with NB UV-B. The Figure shows strongMMP-1 immunopositivity in the cytoplasm of several fi-broblasts after NB UV-B phototherapy. Our researchgroup3 has recently demonstrated that interleukin 6 (IL-6), a multifunctional cytokine that has been shown toincrease the production of collagen and glycosaminogly-cans, is significantly down-regulated during UV-A1 pho-totherapy in morphea.3 Since IL-6 concentration has alsobeen reported to be increased in LS,4 the efficacy of NBUV-B might be explained by a depletion of proinflam-matory cytokines such as IL-6 in addition to the induc-

tion of MMPs. Data recently reported by Choi et al5 in-dicate that NB UV-B irradiation reduces type 1 collagenconcentrations in human skin fibroblasts by inhibitingtransforming growth factor �1 expression and stimulat-ing the release of MMP-1.

The good tolerability and low carcinogenicity linked toNB UV-B should also give rise to future studies on photo-therapy for genital LS,6 which is much more common andmore afflicting for the patient than extragenital disease.

Correspondence: Dr Kreuter, Department of Dermatol-ogy and Allergology, Ruhr University Bochum, Gudrun-strasse 56, D-44791 Bochum, Germany ([email protected]).Financial Disclosure: None reported.

1. Colbert RL, Chiang MP, Carlin CS, Fleming M. Progressive extragenital li-chen sclerosus successfully treated with narrowband UV-B phototherapy. ArchDermatol. 2007;143(1):19-20.

2. Kreuter A, Hyun J, Stucker M, et al. A randomized controlled study of low-doseUVA1,medium-doseUVA1,andnarrowbandUVBphototherapy inthetreatmentof localized scleroderma. J Am Acad Dermatol. 2006;54(3):440-447.

3. Kreuter A, Hyun J, Skrygan M, et al. Ultraviolet A1-induced downregulationof human beta-defensins and interleukin-6 and interleukin-8 correlates withclinical improvement in localized scleroderma. Br J Dermatol. 2006;155(3):600-607.

4. Romero LI, Pincus SH. In situ localization of interleukin-6 in normal skin andatrophic cutaneous disease. Int Arch Allergy Immunol. 1992;99(1):44-49.

5. Choi CP, Kim YI, Lee JW, Lee MH. The effect of narrowband ultraviolet B onthe expression of matrix metalloproteinase-1, transforming growth factor-beta1 and type I collagen in human skin fibroblasts. Clin Exp Dermatol. 2007;32(2):180-185.

6. Gambichler T, Breuckmann F, Boms S, Altmeyer P, Kreuter A. NarrowbandUVB phototherapy in skin conditions beyond psoriasis. J Am Acad Dermatol.2005;52(4):660-670.

Exclusively Benign Dermoscopic Patternin a Patient With Acral Melanoma

T he article by Saida et al1 on the significance ofdermoscopic patterns in detecting malignantmelanoma is of great importance to us. Based on

the results of this large multicenter study conducted in

William D. James, MD

A B

Figure. Immunohistochemical staining for matrix metalloproteinase (MMP)1 in a patient with extragenital lichen sclerosus before (A) and after (B)narrowband UV-B phototherapy (hematoxylin-eosin, original magnification�40 for both panels). B, The cytoplasm of lesional fibroblasts (arrows)shows strong immunopositivity for MMP-1 after phototherapy.

Alexander Kreuter, MDThilo Gambichler, MD



Japan, the authors reported that 2 dermoscopic pat-terns, the parallel ridge pattern and irregular diffuse pig-mentation pattern, were highly specific for the diagno-sis of acral melanoma.

They found that 7 of 103 melanomas revealed a be-nign dermoscopic pattern consisting of either a parallelfurrow pattern or a latticelike pattern.2 However, the au-thors1 also mention that in all melanomas displaying be-nign dermoscopy patterns, this benign pattern was ob-served only focally and that the predominant pattern wasthe malignant pattern.

We completely agree that both benign and malignantpatterns can coexist within the same lesion. Thus, to avoidmisdiagnosing acral melanoma, we recommend that ac-ral lesions be scanned in their entirety, under dermos-copy, to determine the presence or absence of malig-nant patterns and to ensure that no focal area within thelesion with a malignant pattern is overlooked. Biopsy isrecommended for any lesion that reveals a malignant pat-tern, no matter how small the focus.3 If after scanningthe entire lesion the dermoscopist is reassured that nomalignant patterns are present within any part of the le-

sion, then, and only then, should he or she look for thepresence of benign dermoscopy patterns.

Report of a Case. Unfortunately, exceptions occur: onrare occasions, melanomas can present with an exclu-sively benign dermoscopy pattern. We describe a 28-year-old woman of Italian descent who presented with apigmented lesion on the heel, which first appeared 2 yearsearlier and progressively enlarged. Clinical examina-tion revealed a 13-mm, brown, homogeneous, and sym-metric macule (Figure 1). Under dermoscopy, no ma-lignant dermoscopy patterns were evident anywherewithin the lesion (Figure 2). The lesion displayed a ho-mogeneous, parallel furrow pattern with dots and glob-ules, which best fit the double-line parallel furrow pat-tern associated with dots and globules, a variant of thebenign parallel furrow pattern.4,5

However, owing to the disconcerting history of its re-cent onset and growth we were prompted to override ourdermoscopic diagnosis and take biopsy specimens of thelesion. The lesion was surgically removed and histo-pathologic analysis revealed the presence of an atypicalmelanocytic proliferation at the dermoepidermal junc-tion with confluent nests of atypical melanocytes and pag-etoid spread—findings consistent with a diagnosis of mela-noma in situ (Figure 3 and Figure 4).

Comment. What can we learn from this case? First, it isimportant to remember not to disregard a patient’s medi-cal history. Our patient reported that her heel lesion wasrelatively new and was enlarging. This fact alone shouldraise suspicion, since a new or changing lesion is a sen-sitive indicator of melanoma. Despite the fact that the le-sion had a benign dermoscopy pattern, the patient’s medi-cal history was the main impetus for us to take biopsyspecimens and correctly identify her melanoma.

Second, the size of an acral lesion appears to be im-portant. Several studies have shown that acquired acrallesions larger than 7 mm in diameter have a higher prob-ability of being melanoma, regardless of other morpho-logic criteria.6-8

Third, the presence of large globules is an unusual fea-ture for benign lesions on palms and soles. It is our sin-

Figure 1. Clinical image of an acquired pigmented lesion on the right heel ofa 28-year-old woman.

Figure 2. Dermoscopy image of the same lesion shown in Figure 1 showinga benign dermoscopy pattern throughout the lesion. The pigmentationfollows the furrows, and the ridges are relatively hypopigmented (double-lineparallel furrow pattern associated with dots and globules). No malignantdermoscopy pattern can be observed.

Figure 3. Histologic image of the lesion showing a very atypical melanocyticpattern at the dermoepidermal junction with numerous confluent nests ofatypical melanocytes (Fontana-Masson stain, original magnification �5).



cere hope that some of our observations will result in thetimely detection of acral melanomas.

Correspondence: Dr Braun, Department of Dermatol-ogy, University Hospital Zurich, Gloriastr 31, Zurich 8091,Switzerland ([email protected]).Financial Disclosure: None reported.

1. Saida T, Miyazaki A, Oguchi S, et al. Significance of dermoscopic patterns indetecting malignant melanoma on acral volar skin: results of a multicenterstudy in Japan. Arch Dermatol. 2004;140(10):1233-1238.

2. Menzies SW, Zalaudek I. Why perform dermoscopy? the evidence for its rolein the routine management of pigmented skin lesions. Arch Dermatol. 2006;142(9):1211-1212.

3. Altamura D, Altobelli E, Micantonio T, Piccolo D, Fargnoli MC, Peris K.Dermoscopic patterns of acral melanocytic nevi and melanomas in a whitepopulation in central Italy. Arch Dermatol. 2006;142(9):1123-1128.

4. Saida T, Oguchi S, Miyazaki A. Dermoscopy for acral pigmented skin lesions.Clin Dermatol. 2002;20(3):279-285.

5. Marghoob AA, Braun RP, Kopf AW. Atlas of Dermoscopy. New York, NY: Tay-lor Francis; 2005.

6. Saida T, Yoshida N, Ikegawa S, Ishihara K, Nakajima T. Clinical guidelinesfor the early detection of plantar malignant melanoma. J Am Acad Dermatol.1990;23(1):37-40.

7. Saida T, Ishihara Y, Tokuda Y. Effective detection of plantar malignantmelanoma. Int J Dermatol. 1993;32(10):722-725.

8. Saida T. Malignant melanoma on the sole: how to detect the early lesionsefficiently. Pigment Cell Res. 2000;13(suppl 8):135-139.

In reply

I enjoyed reading the comments from Braun et al on the sig-nificance of the dermoscopic patterns seen in acral mela-nocytic lesions. Acral volar skin is the most prevalent siteof malignant melanoma in nonwhite populations. In Japa-nese patients, about half of all cutaneous melanomas affectthis anatomic site. Therefore, early accurate detection of ma-lignant melanoma of this site is essential to improve the prog-nosis, particularly for nonwhites.

The introduction of dermoscopy has immensely contrib-uted to the early detection of acral melanoma. Early acral

melanoma shows a unique dermoscopic pattern of bandlikepigmentation on the ridges of the skin markings, which myresearch groups have called the parallel ridge pattern.1,2 Aprevious study performed by my group revealed that, in earlyacral melanoma, the sensitivity and the specificity of thispattern is 86% and 99%, respectively.3 Irregular diffuse pig-mentation is another important pattern, also often de-tected in more or less advanced acral melanoma. In con-trast, the major dermoscopic patterns of melanocytic nevion acral volar skin are the parallel furrow pattern, the lat-ticelike pattern, and the fibrillar pattern.4

I completely agree with Braun et al in emphasizing theimportance of careful examination of an entire lesion be-cause the malignant patterns of acral melanoma are onlypartly detected in some cases. I also agree with them thatthe benign dermoscopic patterns can be detected also in ac-ral melanoma. However, in melanoma, the benign patternsare found only focally or irregularly distributed within alesion. In my experience, if a melanocytic lesion on volarskin exhibits 1 of these benign dermoscopic patterns in a mo-notonous fashion, the lesion is histopathologically diag-nosed as a benign nevus without exception.

Thus, the case presented by Braun et al is very interest-ing. Clinically, the lesion was oval in shape and regular inshades of brown color. However, the rapid growth and thelarger size, 13 mm in diameter, are exceptional in a benignnevus. The dermoscopic finding of this lesion was the typi-cal feature of the parallel furrow pattern associated with thebrown globules regularly distributed along both sides of thesulci. No malignant dermoscopic patterns were detectedwithin the lesions. Except for the larger size and the rapidgrowth, all of these clinical and dermoscopic findings stronglysuggested that this was a benign nevus. Unexpectedly, how-ever, histopathologic examination revealed confluent nestsof atypical melanocytes as well as pagetoid spread in thelower epidermis. Braun et al finally diagnosed this case asmelanoma in situ.

Histopathologic analysis is still the gold standard to de-termine the diagnosis of melanocytic neoplasm. Thus, thiscase could really be melanoma in situ. Under the principlethat every rule has its exceptions, acral melanoma could ex-ceptionally show the typical parallel furrow pattern. But inthis particular case, I wish to be prudent and agree with thediagnosis of Braun et al.

Conflict between dermoscopic and histopathologic find-ings is emerging as an important subject. Soyer et al5 havenoted that dermoscopy could challenge the monopoly of der-matopathology in determining the diagnosis of malignantmelanoma. In my opinion, the lesion described by Braun etal shows completely regular, symmetric, homogeneous, andorderly dermoscopic features, thereby strongly suggestinga benign nature.

Furthermore, regarding the histopathologic findings ofthis case, I want to add 1 comment. My group has recentlynoticed that despite banal clinical features, some melano-cytic nevi located at transition portions between glabrousand nonglabrous skin and also in interdigital areas oftenhistopathologically exhibit random, irregular distributionof melanocytes as solitary units as well as in nests withinthe epidermis, mimicking the findings of melanoma in situ(unpublished data, 2007). Owing to these histopathologicfindings, melanocytic nevi on these areas could be overdi-

Figure 4. Detail ofFigure 3 showing confluent nests of atypical melanocytesas well as pagetoid spread of suprabasal cells. There is some fibrosis as wellas pigment incontinence in the dermis. Histopathologic diagnosis wasmelanoma in situ (hematoxylin-eosin, original magnification �20).

Ralph P. Braun, MDOlivier Gaide, MDAndreas M. Skaria, MDAlfred W. Kopf, MDJean-Hilaire Saurat, MDAshfaq A. Marghoob, MD



agnosed as melanoma (“pseudomelanoma”), as is the caseof melanocytic nevus of a young woman’s genitalia. The me-lanocytic lesion described by Braun et al seems to be lo-cated at this transition zone, and thus, in my view, the his-topathologic findings must be carefully interpreted. Of course,validity of this concept of pseudomelanoma at the transi-tion areas must be critically evaluated in further studies.

Correspondence: Dr Saida, Department of Dermatol-ogy, Shinshu University School of Medicine, 3-1-1 Asahi,Matsumoto 390-8621, Japan ([email protected]).

1. Oguchi S, Saida T, Koganehira Y, Ohkubo S, Ishihara Y, Kawachi S. Charac-teristic epiluminescent microscopic features of early malignant melanoma onglabrous skin. Arch Dermatol. 1998;134(5):563-568.

2. Saida T, Oguchi S, Miyazaki A. Dermoscopy for acral pigmented skin lesions.Clin Dermatol. 2002;20(3):279-285.

3. Saida T, Miyazaki A, Oguchi S, et al. Significance of dermoscopic patterns indetecting malignant melanoma on acral volar skin. Arch Dermatol. 2004;140(10):1233-1238.

4. Saida T, Oguchi S, Ishihara Y. In vivo observation of magnified features ofpigmented lesions on volar skin using video macroscope. Arch Dermatol. 1995;131(3):298-304.

5. Soyer HP, Massone C, Ferrara G, Argenziano G. Limitations of histopatho-logic analysis in the recognition of melanoma: a plea for a combined diag-nostic approach of histopathologic and dermoscopic evaluation. Arch Dermatol.2005;141(2):209-211.

Evidence Insufficient to RecommendMelanoma Surveillance FollowingPhototherapy for Jaundice

M atichard et al1 reported an association be-tween receipt of blue-light phototherapy forneonatal jaundice and numbers of melano-

cytic nevi at least 2 mm in diameter at age 8 or 9 years.Although they stated that their results “should be inter-preted with caution,”1(p1603) they concluded that ex-posed children “should undergo dermatologic preven-tive measures and surveillance for the development ofmelanoma.”1(p1599)

We believe that this recommendation is premature. Theauthors studied only 18 children who had been exposedto phototherapy, and the selection of these exposed chil-dren was via an entirely different mechanism from the se-lection of the control group. This could easily have intro-duced confounding factors that were not measured. Itappears that the original hypothesis that the authors soughtto test was that phototherapy might increase the total ne-vus count, which it did not: mean counts, 28.5 vs 28.7(P=.88). The association was present only when atten-tion was restricted to nevi 2 to 5 mm in diameter. The au-thors do not provide a biological rationale for restrictionof the association to nevi of that size. Finally, a larger studyof this issue failed to find any association between photo-therapy and nevi,2,3 and a study of whether phototherapywas a risk factor for melanoma4 found that 0 of 30 mela-noma cases had included exposure to phototherapy, com-pared with 11 of 120 controls (odds ratio, 0; upper limitof 95% confidence interval, 1.55).5

Counseling families of infants exposed to photo-therapy that their child needs to be watched for mela-

noma is not a trivial matter. Much more evidence thanwas provided by Matichard et al1 is needed before it canbe recommended.

Correspondence: Dr Newman, Department of Epidemi-ology and Biostatistics, 185 Berry St, Ste 5700, San Fran-cisco, CA 94107-1762 ([email protected]).Financial Disclosure: None reported.

1. Matichard E, Le Henanff A, Sanders A, Leguyadec J, Crickx B, Descamps V.Effect of neonatal phototherapy on melanocytic nevus count in children. ArchDermatol. 2006;142(12):1599-1604.

2. Wiecker TS, Luther H, Buettner P, Bauer J, Garbe C. Moderate sun exposureand nevus counts in parents are associated with development of melanocyticnevi in childhood: a risk factor study in 1,812 kindergarten children. Cancer.2003;97(3):628-638.

3. Bauer J, Buttner P, Luther H, Wiecker TS, Mohrle M, Garbe C. Blue-light pho-totherapy of neonatal jaundice does not increase the risk for melanocytic ne-vus development. Arch Dermatol. 2004;140(4):493-494.

4. Berg P, Lindelof B. Is phototherapy in neonates a risk factor for malignantmelanoma development? Arch Pediatr Adolesc Med. 1997;151(12):1185-1187.

5. Dean AG, Dean JA, Burton AH, Dicker RC. Epi Info: a general-purpose mi-crocomputer program for public health information systems. Am J Prev Med.1991;7(3):178-182.

In reply

My group’s study1 demonstrates that the number of mela-nocytic nevi larger than 2 mm in diameter is higher in chil-dren aged between 8 and 9 years who were exposed to neo-natal phototherapy than in those who were not exposed. Ina multivariate analysis, we have studied the respective rolesof both sun exposure and neonatal phototherapy (Table).

As summarized in the Table, sun exposure is the majorfactor for nevi smaller than 2 mm. In addition, the data inthe Table support the fact that neonatal phototherapy is themajor risk factor for nevi larger than 2 mm in diameter.

It can be postulated that sun exposure is the major riskfactor for the most recent melanocytic nevi (<2 mm). Nevilarger than 5 mm might represent congenital nevi forwhich the genetic predisposition could be the most impor-tant risk factor.

This classification of melanocytic nevi into different di-ameters (larger than 2 mm, 2-5 mm, and larger than 5 mm)is usual in studies evaluating the prevalence of common me-lanocytic nevi in children and its relationship with pigmen-tary characteristics and sun exposure.2-5 In most studies, smallnevi (<2 mm) and small dark brown pigmented frecklescalled lentigo simplex are excluded. But until now, the is-sue of the size of the nevi has not been totally resolved. How-ever, some authors5 include these small nevi in the total num-ber of nevi, as my group did.1

Retrospective documentation of neonatal phototherapy isoften very difficult to find and often lacks important details(eg, bilirubin level throughout therapy, duration of expo-sure, and dose of phototherapy). Hence the entry criteria formy study group was the presence of complete phototherapydocumentation and the ability to find the child. The controlgroup was selected to match for age, geographic area, andFitzpatrick skin type, thus decreasing both the likelihood ofbias and the number and influence of confounding factors. Ourmethod of case selection avoided the problems of recall biasthat are inherent in studies based on interviews, as con-ducted by Bauer and colleagues.6 The different conclusions

Toshiaki Saida, MD, PhD

Thomas B. Newman, MD, MPHM. Jeffrey Maisels, MB BCh



drawn by my group’s study and by Bauer et al may be theresult of the different methods used to obtain the data.

Although the relatively small sample size of my group’sstudy1 suggests caution in interpreting these data (particu-larly our multivariate analysis) and may limit generaliza-tion of the results, the study poses an important question:can it now be simply stated that phototherapy dose not in-duce a modification of nevus count?

Nonetheless, we strongly recommend that all neonatalphototherapy sessions be meticulously documented for usein further research and risk stratification in patients’ rec-ords. We believe that individuals who receive intensive pho-totherapy should undergo periodic dermatologic follow-upif they develop nevi.

Correspondence: Dr Descamps, Bichat-Claude BernardHospital, Department of Dermatology, AssistancePublique des Hopitaux de Paris, 46 rue Henri Huchard,75018 Paris, France ([email protected]).

1. Matichard E, Le Henanff A, Sanders A, Leguyadec J, Crickx B, Descamps V.Effect of neonatal phototherapy on melanocytic nevus count in children. ArchDermatol. 2006;142(12):1599-1604.

2. Siskind V, Darlington S, Green L, Green A. Evolution of melanocytic nevi onthe faces and necks of adolescents: a 4 y longitudinal study. J Invest Dermatol.2002;118(3):500-504.

3. Autier P, Severi G, Pedeux R, et al; European Organisation for Research andTreatment of Cancer Melanoma Group. Number and size of nevi are influ-enced by different sun exposure components: implications for the etiology ofcutaneous melanoma (Belgium, Germany, France, Italy). Cancer Causes Control.2003;14(5):453-459.

4. MacLennan R, Kelly JW, Rivers JK, Harrison SL. The Eastern Australian Child-hood Nevus Study: site differences in density and size of melanocytic nevi inrelation to latitude and phenotype. J Am Acad Dermatol. 2003;48(3):367-375.

5. Valiukeviciene S, Miseviciene I, Gollnick H. The prevalence of common ac-quired melanocytic nevi and the relationship with skin type characteristicsand sun exposure among children in Lithuania. Arch Dermatol. 2005;141(5):579-586.

6. Bauer J, Buttner P, Luther H, Wiecker TS, Mohrle M, Garbe C. Blue light pho-totherapy of neonatal jaundice does not increase the risk for melanocytic ne-vus development. Arch Dermatol. 2004;140(4):493-494.

Allergic Contact Dermatitis:Another Adverse Effect of Over-the-counterTopical Hydrocortisone

W e read with interest the commentary in theMarch 2007 issue of the Archives of Dermatol-ogy titled “Topical Hydrocortisone From Pre-

scription to Over-the-Counter Sale.”1 While this article aptlydescribes the historical debate regarding approval for over-the-counter drug (OTC) use, it neglects to include allergiccontact dermatitis from corticosteroids as 1 of the adverseeffects of topical hydrocortisone.2-9

Corticosteroid allergy was first described in 1959 byBurckhardt10 but has become more widely recognizedsince the introduction of tixocortol-21-pivalate as a patch-test marker in the late 1980s. Prevalence of allergy to tixo-cortol, a marker for allergy to hydrocortisone and othergroup A corticosteroids, has ranged from 2.3% to 2.9%in large patch-tested populations in North America. Clini-cal relevance to the presenting dermatitis is docu-mented in more than 80% of cases.

The ready availability of OTC hydrocortisone prod-ucts likely contributes to this high prevalence and clini-cal relevance. Clues to the diagnosis of steroid contactallergy include worsening of dermatitis with use or achange in the clinical pattern of the disease. Impor-tantly, this allergen is not available in commercial patch-test kits such as the T.R.U.E. test (Allerderm, Phoenix,Arizona) in the United States. Clinicians should be awareof the risk of allergic contact dermatitis (ACD) from cor-ticosteroids and be prepared to refer patients for com-prehensive patch testing.

Correspondence: Dr Watsky, 330 Orchard St, Ste 103,New Haven, CT 06511 ([email protected]).Financial Disclosure: None reported.

1. Ravis SM, Eaglstein WH. Topical hydrocortisone from prescription to over-the-counter sale. Arch Dermatol. 2007;143(3):413-415.

2. Lutz ME, el-Azhary RA, Gibson LE, Fransway AF. Contact hypersensitivityto tixocortol pivalate. J Am Acad Dermatol. 1998;38(5, pt 1):691-695.

3. Scheuer E, Warshaw E. Allergy to corticosteroids: update and review of epi-demiology, clinical characteristics, and structural cross-reactivity. Am J Con-tact Dermat. 2003;14(4):179-187.

4. Marks JG, Belsito DV, DeLeo VA, et al. North American Contact DermatitisGroup standard tray patch test results (1992 to 1994). Am J Contact Dermat.1995;6:160-165.

5. Marks JG, Belsito DV, DeLeo VA, et al. North American Contact DermatitisGroup patch test results for the detection of delayed-type hypersensitivityto topical allergens. J Am Acad Dermatol. 1998;38(6, pt 1):911-918.

6. Marks JG Jr, Belsito DV, DeLeo VA, et al. North American Contact Derma-titis Group patch test results, 1996-1998. Arch Dermatol. 2000;136(2):272-273.

7. Marks JG Jr, Belsito DV, DeLeo VA, et al. North American Contact Derma-titis Group patch test results, 1998 to 2000. Am J Contact Dermat. 2003;14(2):59-62.

8. Pratt MD, Belsito DV, DeLeo VA, et al. North American Contact DermatitisGroup patch test results, 2001-2002 study period. Dermatitis. 2004;15(4):176-183.

9. Warshaw EM, Rietschel RL, Sasseville D, et al. North American Contact Der-matitis Group patch test results 2003-2004 study period. Dermatitis. In press.

10. Burckhardt W. Contact eczema caused by hydrocortisone. Hautarzt. 1959;10:42.

In reply

We thank Watsky and Warshaw for their interest and in-sightful comments. As they point out, ACD was not listedas an adverse effect of treatment with topical hydrocorti-

Table. Statistical Multivariate Analysis of Nevus Risk Factor

Exposure Category

P Value in the Variance Analysis Model by Lesion Size

Any Size �2 mm 2-5 mm �2 mm �5 mm

Intensive phototherapy .50 .18 �.001 �.001 .29Solar exposure during holidays .02 .02 .40 .46 .80Severe sunburns .94 .84 .51 .47 .45

Kalman L. Watsky, MDErin M. Warshaw, MD, MS

Vincent Descamps, MD, PhD



sone in our commentary.1 We did not mention the potentialadverse effect of ACD to topical hydrocortisone because itdid not play a role in the debate regarding the appropriate-ness of permitting OTC sale of topical hydrocortisone prepa-rations.

As pointed out by Watsky and Warshaw, the readyavailability of OTC hydrocortisone products (especially instructural class A) likely contributes to a high prevalenceof clinically significant positive patch test results to corti-costeroid markers in selected patch-tested populations.How this translates to the prevalence of corticosteroid-induced ACD in the population at large is apparently notknown. At what level corticosteroid-induced ACD might bea public health concern requiring consideration of remov-ing products from the market is not only an intriguingquestion but one that could, depending on the answer, re-sult—following the law of unintended consequences—inproving correct, after all, the initial reaction of the derma-tology community.

Correspondence: Dr Eaglstein, Research and Develop-ment, Stiefel Laboratories Inc, 255 Alhambra Cir, CoralGables, FL 33134 ([email protected]).

1. Ravis SM, Eaglstein WH. Topical hydrocortisone from prescription to over-the-counter sale. Arch Dermatol. 2007;143(3):413-415.

VIGNETTES

Successful Treatment of Pityriasis VersicolorWith 5-Aminolevulinic AcidPhotodynamic Therapy

P ityriasis versicolor is a common chronic super-ficial fungal infection caused by the organism Mal-assezia furfur. Although many conventional treat-

ments have been shown to be successful in treatingpityriasis versicolor, therapy of long-term duration maybe needed in many cases, and recurrence is not uncom-mon. In vitro experimental investigations have demon-strated that several fungal strains can be effectively in-activated by irradiation with visible light wavelengths inthe presence of photosensitizer.1-3 However, to our knowl-edge, there have been no clinical trials of photodynamictherapy (PDT) against pityriasis versicolor.

Report of a Case. A 37-year-old man presented with a2-year history of asymptomatic, light brown, scaly patcheson both axillae (Figure 1A). Microscopic examinationof samples stained with 10% potassium hydroxide re-vealed short, thick, fungal hyphae and spores of varioussizes (Figure 2A). We tried to treat the lesions with

A B

Figure 1. A 37-year-old man with a 2-year history of scaly patches on both axillae. A, Sharply marginated, light brown hyperpigmentation with fine scales on theaxilla. B, Clearance of pityriasis versicolor without reinfection after 16 weeks of follow-up after treatment with 5-aminolevulinic acid in combination withphotodynamic therapy.

William H. Eaglstein, MDScott M. Ravis



5-aminolevulinic acid (ALA) in combination with PDT.Topical 20% ALA (Medac, Hamburg, Germany) in pet-rolatum vehicle was applied to the lesions and coveredwith an occlusive polyurethane film (Tegaderm; 3M,St Paul, Minnesota). After 4 hours, excess ALA was re-moved, and the lesions were irradiated with light fromlight-emitting diodes (mean±SD wavelength, 630±50nm). The light intensity was 100 mW/cm2, and the lightdoses were 70 J/cm2 for the left axilla and 80 J/cm2 forthe right axilla. The procedure was repeated 2 weeks laterwith increased light dose increments of 10 J/cm2 on eachside. The area cleared within 4 weeks, and the patientwas observed for an additional 3 months without rein-fection (Figure 1B). One week after the first treatmentand 10 weeks after the last treatment, samples stainedwith 10% potassium hydroxide were again taken fromboth axillae and these smears showed no fungalhyphae or spores (Figure 2B). Of note, the clearancewas maintained throughout the warm and humidsummer of Korea.

Comment. To our knowledge, only 1 clinical trial of PDTtreatment for mycotic infection has been reported.4 In thatstudy, interdigital mycosis of the feet was irradiated with75 J/cm2. The clearance was seen in 6 of 9 patients after1 (n=4) or 4 (n=2) treatments. However, 4 patients re-ported recurrences 4 weeks after the last treatment. Theauthors suggested that in vivo environmental condi-tions such as temperature, humidity, and pH of the in-terdigital skin could induce a poor cell uptake of ALAand a deficient biosynthesis of protoporphyrin IX. In ad-dition, the irregular tridimensional shape of this pecu-liar anatomic area could lead to a nonuniform delivery

of light and/or ALA cream. The better outcome in ourcase with a similar light dose may be due to the differentcausative fungal strain and relatively uniform delivery oflight and ALA cream to the affected sites.

In the treatment of pityriasis versicolor, standard drugtreatments are prolonged, and the appearance of drug-resistant strains is becoming more frequent. The multi-plicity of cellular targets in fungi should reduce the riskof selection of photomutant, resistant strains after PDT,and this risk should be further minimized by the lack ofmutagenic effects of PDT.5 Our results of prompt clear-ance and no recurrence of pityriasis versicolor after PDTindicate that this approach may be a promising treat-ment for pityriasis versicolor.

Correspondence: Dr Y. C. Kim, Department of Derma-tology, Ajou University School of Medicine, 5 Wonchon-Dong, Yeongtong-Gu, Suwon 443-721, South Korea([email protected]).Financial Disclosure: None reported.

1. Wainwright M. Photodynamic antimicrobial chemotherapy (PACT). J Anti-microb Chemother. 1998;42(1):13-28.

2. Strakhovskaya MG, Shumarina AO, Fraikin GY, Rubin AB. Synthesis of pro-toporphyrin IX induced by 5-aminolevulinic acid in yeast cells in the pres-ence of 2,2;-dipyridyl. Biochemistry (Mosc). 1998;63(6):725-728.

3. Smijs TG, van der Haas RN, Lugtenburg J, Liu Y, de Jong RL, SchuitmakerHJ. Photodynamic treatment of the dermatophyte Trichophyton rubrum andits microconidia with porphyrin photosensitizers. Photochem Photobiol. 2004;80(2):197-202.

4. Calzavara-Pinton PG, Venturini M, Capezzera R, Sala R, Zane C. Photody-namic therapy of interdigital mycoses of the feet with topical application of5-aminolevulinic acid. Photodermatol Photoimmunol Photomed. 2004;20(3):144-147.

A B

Figure 2. Microscopic examination of samples stained with 10% potassium hydroxide (original magnification �400 for both panels). A, Before treatment,numerous short, thick fungal hyphae and spores of various sizes are seen. B, No fungal hyphae or spores are present 10 weeks after the last treatment.

Young Jin Kim, MDYou Chan Kim, MD, PhD



5. Calzavara-Pinton PG, Venturini M, Sala R. A comprehensive overview of pho-todynamic therapy in the treatment of superficial fungal infections of the skin.J Photochem Photobiol B. 2005;78(1):1-6.

Leukemia Cutis: A Presenting Signin Acute Promyelocytic Leukemia

Report of a Case. A 63-year-old woman developed 3 ery-thematous, pruritic plaques on her left lower extremity.She called her dermatologist and was prescribed pred-nisone for likely contact dermatitis from poison ivy. Twoweeks later, the patient developed fatigue and bruising.Findings from a complete blood cell count were abnor-mal, demonstrating blasts (promyelocytes predomi-nant), anemia, and thrombocytopenia. The patient wasadmitted for workup.

On admission, her white blood cell count was 4500cells/µL (normal range, 4500-11 000 cells/µL); hemato-crit level was 31.4% (normal range, 35%-45%); and plate-let count was 33�103/µL (normal range, 153�103/µLto 367�103/µL). A manual differential count demon-strated 62% promyelocytes. Analysis of bone marrow bi-opsy specimens showed markedly hypercellular mar-row with 90% promyelocytes and a positive translocationof 15:17 on fluorescence in situ hybridization. Flow cy-tometry revealed cells positive for CD13, CD33, CD38,and CD117 and negative for CD14, CD64, CD2, CD11,CD3, CD20, CD7, CD34, and HLA-DR, which was con-sistent with acute promyelocytic anemia. Polymerase chainreaction revealed a 0.436 ratio of the PML-RAR-α onco-gene in her bone marrow.

A dermatology consultant was called to evaluate 3 ery-thematous, nontender, indurated plaques with necroticcenters on her left calf, ranging in size from 5 to 9 mm(Figure 1). Scattered petechiae and ecchymoses werepresent. The differential diagnosis included vasculitis, leu-kemia cutis, arthropod assault, panniculitis, deep fun-gal infection, and atypical mycobacteria. Tissue cul-tures were negative for bacteria, fungal elements, andmycobacteria. Histopathologic analysis revealed, pan-dermally and in the subcutaneous fat, dense, perivascu-lar, and interstitial infiltrate of malignant, immature,granulocytic cells with extensive tissue hemorrhage. Thesetumor cells contained polymorphic nuclei with abun-dant granular cytoplasm. Occasional mitotic figures werepresent. The immature hematopoietic cells stained dif-fusely positive for chloracetic esterase by Leder stain andwere CD34 negative. A few immature cells contained eo-sinophilic, cytoplasmic granules. No epidermotropismor vascular or neural invasion was noted (Figure 2). Thefindings confirmed the diagnosis of leukemia cutis, andthe morphologic features were consistent with promy-elocytic leukemia.

The patient underwent a chemotherapeutic regimenof daunorubicin and all-trans retinoic acid (ATRA) fol-lowed by cytarabine, daunorubicin, and arsenic triox-ide. Four months after her initial presentation, a bonemarrow biopsy specimen showed no abnormalities, andher disease is currently in remission. The leukemia cu-tis lesions on her calf are well healed.

Comment. Leukemia cutis is the dissemination of sys-temic leukemia and a poor prognostic sign. It presentsas papules, plaques, or nodules, with a characteristic vio-laceous color.

Diffuse infiltration of leukemic cells into the dermis,subcutis, blood vessels, and skin adnexa are seen on his-tologic examination.1 Immunohistochemical study char-acterizes the immunophenotypes of leukemic cells.

Acute promyelocytic leukemia (APL), classified asAML-M3 (acute myeloid leukemia, subtype M3) in theFrench-American-British classification system, ac-counts for approximately 10% to 15% of acute myeloidleukemia in adults.2 Patients may have abundant abnor-mal progranulocytes, severe coagulopathy, and a bal-anced reciprocal translocation, t(15;17)(q22;q12), in leu-kemic cells.3 Fusion of the promyelocytic leukemia gene(PML) on chromosome 15 and the retinoic acid recep-tor � (RAR-�) on chromosome 17 to form the oncogenePML-RAR-� occurs.2 At 5 years, 70% to 80% of patientswith APL are disease free with treatment, which in-cludes ATRA and anthracycline-based chemotherapy.2

Figure 1. Three erythematous, nontender, indurated plaques with necroticcenters ranging in size from 5 to 9 mm were present on the patient’s left calf.

Figure 2. In a biopsy specimen taken from a lesion on the patient’s calf,immature hematopoietic cells stained diffusely positive for chloraceticesterase by Leder stain (original magnification �40).



From 2% to 20% of patients with lymphocytic andgranulocytic leukemia and 10% to 50% with monocyticleukemia develop leukemia cutis.4 To our knowledge,fewer than 25 cases of skin infiltration by APL cells havebeen previously described.5 Interestingly, extramedul-lary involvement tends to occur after the first remissionin APL, perhaps as a complication of ATRA therapy. Incontrast, our patient presented with lesions of leukemiacutis prior to diagnosis and treatment.

To our knowledge, this represents the first case of leu-kemia cutis as a presenting sign of APL. Clinicians shouldhave a high suspicion for cutaneous findings in a pa-tient with weakness and fatigue because it may heraldthe diagnosis of malignancy.

Correspondence: Dr Martin, Department of Derma-tology, University of Maryland School of Medicine,405 W Redwood St, Sixth Floor, Baltimore, MD 21201([email protected]).Financial Disclosure: None reported.

1. Gambichler T, Herde M, Hoffmann K, Altmeyer P, Jansen T. Poor prognosisof acute myeloid leukemia associated with leukemia cutis. J Eur Acad Derma-tol Venereol. 2002;16(2):177-188.

2. Lowenberg B, Griffin JD, Tallman MS. Acute myeloid leukemia and acute pro-myelocytic leukemia. Hematology Am Soc Hematol Educ Program. 2003:82-101.

3. Giralt S, O’Brian S, Weeks E, et al. Leukemia cutis in acute promyelocytic leu-kemia: reports of three cases after treatment with all-trans retinoic acid. LeukLymphoma. 1994;14(5-6):453-456.

4. Ratnam KV, Khor CJL, Su WPD. Leukemia cutis. Dermatol Clin. 1994;12(2):419-431.

5. Ueda K, Kume A, Furukawa Y, Higashi N. Cutaneous infiltration in acute pro-myelocytic leukemia. J Am Acad Dermatol. 1997;36(1):104-106.

Necrolytic Migratory Erythema:The Outermost Markerfor Glucagonoma Syndrome

Report of a Case. A 56-year-old woman presented withsignificant loss of weight, a painful oral inflammation,and a skin eruption on the trunk, lower extremities, andthe intertriginous and genital areas. The physical exami-nation showed a patient with cachexia and striking erup-tions on the skin and mucous membranes. The lesionsconsisted of annular, confluent, superficially eroded ery-thematous plaques with elevated, well-demarcated edgesand bullas at the sites of friction (Figure 1). Faint areasof lacy erythema were noted on the buttocks and arms.Severe stomatitis, painful glossitis, and bilateral angularcheilitis were present (Figure 2). Annular erythema-tous lesions with slightly elevated borders were seen onthe genital area.

An elevated erythrocyte sedimentation rate (65 mm/h;normal, �12 mm/h), high fasting serum glucose level(155 mg/dL; normal, 63-108 mg/dL) (to convert to mil-limoles per liter, multiply by 0.0555), and low hemoglo-

bin level (100 g/L; normal, 120-140 g/L) were detected.A peripheral blood smear showed microcytic, hypochro-mic red blood cells with anisocytosis.

Histopathologic findings revealed severe parakerato-sis, mild acanthosis, and persistent loss of granular layer.Necrosis of keratinocytes in the squamous layer and sub-corneal blistering were present. Mild vascular dilatationand perivascular lymphocytic infiltrate were seen in thedermis.

An abdominal computed tomographic scan sug-gested a pancreatic lesion and showed multiple hypoecho-genic lesions with diameters of 2 to 3 cm. Magnetic reso-nance imaging of the pancreas revealed a diffuse massin the tail. The plasma glucagon level was elevated at 127ng/L (normal, �60 ng/L).

The patient underwent pancreatic debulking andpartial metastasectomy of the liver. The histologic andimmunohistochemical investigations revealed a differ-entiated endocrine tumor with vascular invasion andsomatostatin receptor immunopositivity. Treatmentwith intramuscular octreotide (30 mg every 28 days)and interferon alfa (6 MU subcutaneously 3 times perweek) was introduced together with 6 cycles of hepaticchemoembolization. No remission was achieved, andthe patient died 14 months after the first physical ex-amination.

Tania R. Markowski, BADonna Bilu Martin, MDGrace F. Kao, MDLinda Lutz, MDApril Deng, MD, PhDAnthony A. Gaspari, MD

Figure 1. Bulla on the fourth toe due to mechanical trauma.

Figure 2. The tongue is red and atrophic; bilateral angular cheilitis is seen.



Comment. Necrolytic migratory erythema (NME) patho-genesis is quite controversial. In vitro studies have dem-onstrated that glucagon stimulates epidermal arachi-donic acid synthesis1 thus directly affecting the skinstructures.

The cellular protein p62 (which is an adapter in signal-transducing pathways and a component of the inclu-sion bodies of various chronic degenerative disorders)is either absent from, or only focally expressed in, glu-cagonomas in contrast to its abundance in a well-functioning pancreas. Since p62 is involved in proapop-totic signal transduction, its loss probably renders tumorcells less sensitive to apoptosis.2

At the same time, p16 and p27 levels are decreasedconsistently in pancreatic endocrine neoplasms, andbecause these proteins act as inhibitors of cyclin-dependent kinases, their insufficiency may stimulate theirtransformation and migration.3

Very often NME precedes the onset of systemic symp-toms and identifies the main pathologic process.4 Ow-ing to its distinct clinical picture and well-defined his-topathologic findings, NME seems to be a dependablemorphologic clue to a subtle diagnosis.5 Remarkably, clini-cal presumption and histologic findings were the basesin the present case for the detection of this localized neo-plastic proliferation and to confirm the sophisticated clini-cal entity.

Correspondence: Dr Gantcheva, Department of Derma-tology and Venereology, Faculty of Medicine, Sofia Medi-cal University, 1 St Georgi Sofiisky St, 1431 Sofia, Bul-garia ([email protected]).Financial Disclosure: None reported.

1. Peterson LL, Shaw JC, Acott KM, Mueggler PA, Parker F. Glucagonoma syn-drome increases epidermal arachidonic acid. J Am Acad Dermatol. 1984;11(3):468-473.

2. Lackner C, Dlaska D, Fuchsbichler A, et al. p62 Protein is expressed in pan-creatic beta cells J Pathol. 2005;206(4):402-408.

3. Tomita T. Cyclin-dependent kinase (cdk6) and p16 in pancreatic endocrineneoplasms. Pathology. 2004;36(6):566-570.

4. Chastain MA. The glucagonoma syndrome: a review of its features and dis-cussion of new perspectives. Am J Med Sci. 2001;321(5):306-320.

5. Kheir SM, Omura EF, Grizzle WE, Herrera GA, Lee I. Histologic variation inthe skin lesions of the glucagonoma syndrome. Am J Surg Pathol. 1986;10(7):445-453.

Lymphomatoid Granulomatosis Inducedby Imatinib-Treatment

Report of a Case. An 89-year-old woman with liver me-tastases from a gastrointestinal stromal tumor (GIST) wastreated with imatinib mesylate, initial dose of 400 mg/d.Three months later, the liver tumors decreased. Again 3months later, a computed tomographic (CT) scan re-vealed 3 intrapulmonary lesions that radiologically re-sembled an infectious complication or metastases of theGIST. Owing to the age of the patient and continuousdecline of her physical status, a lung biopsy was ex-cluded. The progressing pulmonary tumors were clini-

cally interpreted as metastases of the GIST, and there-fore imatinib mesylate treatment was continued.

Nine months later, the patient presented with mul-tiple subcutaneous nodules as large as 25�30 mm onboth legs and the lower back. Histologic findings re-vealed a dense lymphocytic infiltrate, mainly of the sub-cutis, extending through the dermis into the epidermis,with central necrosis and scattered giant cells. The infil-trate was both angiocentric and angiodestructive. Im-munohistochemically, the lymphocytes were mainly CD20positive (Figure 1). Epstein-Barr virus (EBV) in situ hy-bridization revealed EBV RNA in a distinct populationof infiltrating lymphocytes.

Clonality analysis of the CDR3 region of the immu-noglobulin heavy-chain gene revealed the identical clonein 2 distant biopsy specimens. A large B-cell lymphomaresembling lymphomatoid granulomatosis (LYG) wasdiagnosed.

Whole-body CT scan revealed, in addition to hepaticGIST tumor metastasis, progressive pulmonary tumors thatresembled lymphomas or granulomas (Figure 2). Bloodcell count findings showed leukopenia (2600 white bloodcells/µL) and lymphopenia (10.8% lymphocytes). Immu-nophenotyping of the peripheral blood showed an in-creased CD4/CD8 ratio of 9:1 and 1.9% B cells.

Cutaneous and most likely pulmonary LYG was di-agnosed. As the pulmonic lymphomas lead to progres-sive dyspnea, imatinib mesylate therapy was discontin-ued and replaced by a symptomatic dexamethasonetreatment, which was stopped by the patient after 4 weeks.Two months after discontinuation, the subcutaneous nod-ules had vanished completely, and the pulmonary infil-trates could be detected only as small residual nodules(Figure 2). The patient’s physical condition continu-ously improved, and the leukocyte and lymphocyte countsreturned to normal range. The liver metastases of GISTdid not show any change in size.

Comment. Lymphomatoid granulomatosis is a rare, an-giocentric, angioinvasive, EBV-positive, B-cell lympho-proliferative disease that primarily involves the lung andskin,1 histologically characterized by angiocentric and an-

Mary L. Gantcheva, MD, PhDValentina K. Broshtilova, MDAdriana I. Lalova, MD, PhD

A

B

Figure 1. Histologic images demonstrating large blastic B cells surroundingthe necrosis under hematoxylin-eosin (A) (original magnification � 20) andCD20 stain (B) (original magnification � 200).



gioinvasive lymphoid infiltrates. The presence of EBV-positive B cells is critical for the diagnosis.

Imatinib mesylate, a tyrosine kinase inhibitor target-ing the Bcr-Abl fusion tyrosine kinase,2 is currently usedin the management of metastatic and inoperable GIST.Common adverse effects include eosinophilia, fasciitis,and mycosis fungoides–like reactions.3 Bekkenk et al4 ob-served the occurrence of EBV-positive, primary, cutane-ous, B-cell lymphoma during imatinib mesylate treat-ment that resolved after the therapy was discontinued.

Epstein-Barr virus–associated lymphomas occur mostcommonly in severely immunosuppressed patients andoften spontaneously resolve after reconstitution of theimmune system. Imatinib mesylate may act throughT-cell depletion; it can inhibit T-cell proliferation andT-cell activation in vitro and reduce the expansion of cy-totoxic T lymphocytes in response to EBV.5 Since, in thepresent case, cutaneous and pulmonary LYG occurredduring imatinib mesylate therapy and resolved entirelyafter the treatment was discontinued, modulation ofT-cell function was most likely critically involved in thedevelopment of LYG.

Correspondence: Dr Yazdi, Department of Dermatology,University of Tubingen, Liebermeisterstr 25, 72076 Tub-ingen, Germany ([email protected]).Financial Disclosure: None reported.

1. Fauci AS, Haynes BF, Costa J, et al. Lymphomatoid granulomatosis: prospec-tive clinical and therapeutic experience over 10 years. N Engl J Med. 1982;306(2):68-74.

2. Druker BJ, Sawyers CL, Kantarjian H, et al. Activity of a specific inhibitor ofthe BCR-ABL tyrosine kinase in the blast crisis of chronic myeloid leukemiaand acute lymphoblastic leukemia with the Philadelphia chromosome. N EnglJ Med. 2001;344(14):1038-1042.

3. Jardin F, Courville P, Lenain P, et al. Concomitant eosinophilia, fasciitis, andmycosis fungoides-like reaction with antinuclear autoantibodies in chronicmyeloid leukaemia: role of a T-cell clone induced by imatinib. Lancet Oncol.2005;6(9):728-729.

4. Bekkenk MW, Vermeer MH, Meijer CJ, et al. EBV-positive cutaneous B-celllymphoproliferative disease after imatinib mesylate. Blood. 2003;102(12):4243.

5. Seggewiss R, Lore K, Greiner E, et al. Imatinib inhibits T-cell receptor-mediated T-cell proliferation and activation in a dose-dependent manner. Blood.2005;105(6):2473-2479.

Focal Acne During Topical TacrolimusTherapy for Vitiligo

T acrolimus is a macrolide antibiotic produced byStreptomyces tsukubaensis and with strong T-specific immunosuppressant activity. Primarily

used for treatment of atopic dermatitis, it has been usedto treat other inflammatory and immunologic skin dis-orders, including vitiligo, with encouraging results.1 Fo-cal hypertrichosis has been reported during topical ta-crolimus therapy for childhood vitiligo,2 but acne doesnot seem to be a frequently reported adverse effect.

Report of a Case. An 18-year-old woman presented withvitiligo on the left side of her chin and neck and on bothsides of the dorsum. The lesions had been present for 4months and had been enlarging progressively. Findingsof clinical evaluation and laboratory tests were all nor-mal. As a first treatment, she had undergone a 3-monthcourse of topical 0.05% desonide cream without any clini-cal response. She was put on a regimen of topical 0.1%tacrolimus, applied at bedtime, for 3 months, resultingin repigmentation of nearly 90% of the chin and cervicallesions and 50% of the dorsal spots.

However, at the end of 3 months, the cervical and chinspots showed an extensive eruption of inflammatory andnoninflammatory lesions of acne, with papules, pus-tules, and closed comedones (Figure) not present pre-viously. The macules of the dorsum did not present anyacne lesions. Tacrolimus treatment was discontinued, andthe patient was treated with oral doxicycline and topicaladapalene, with complete clearance of the eruption in 2months.

Comment. Treatment with topical calcineurin inhibi-tors is considered a safe and useful therapeutic optionfor several dermatoses. Rosaceiform reactions have been

A B C

Figure 2. Radiologic images. A, Computed tomographic scan showing multiple large pulmonary tumors. B and C, Chest radiographs taken during imatinib therapy(B) and after discontinuation of imatinib treatment (C).

Amir S. Yazdi, MDGisela Metzler, MDSusanne Weyrauch, MDMark Berneburg, MDMichael Bitzer, MDHans-Konrad Muller-Hermelink, MDMartin Rocken, MD



described after topical tacrolimus and even pimecroli-mus treatment for inflammatory disorders of the face.3,4

Used in a vehicle-controlled study for atopic dermatitisin adults, 0.1% topical tacrolimus therapy resulted in acneas an adverse event at a mean±SE rate of 7.1%±2.02%(P� .05).5

As mechanisms and incidence of these adverse find-ings are not well established, a more accurate follow-upof patients undergoing long-term topical therapy withthese drugs would perhaps be advisable, to search for pos-sible, although infrequent, unwanted effects on the pi-losebaceous units such as acne, hypertrichosis, and ro-saceiform reactions.

Correspondence: Dr L. Bakos, Department of Derma-tology, Hospital de Cl ınicas de Porto Alegre–Universidade Federal do Rio Grande do Sul, Ramiro Bar-cellos 2350, Porto Alegre RS 90035003, Brazil ([email protected]).Financial Disclosure: None reported.

1. Travis LB, Weinberg JM, Silverberg NB. Successful treatment of vitiligo with0.1% tacrolimus ointment. Arch Dermatol. 2003;139(5):571-574.

2. Prats Caelles I, Pinto PH, Casado ELA, Laguna RL. Focal hypertrichosis dur-ing topical tacrolimus therapy for childhood vitiligo. Pediatr Dermatol. 2005;22(1):86-87.

3. Antille C, Saurat JH, Lubbe J. Induction of rosaceiform dermatitis during treat-ment of facial inflammatory dermatoses with tacrolimus ointment. ArchDermatol. 2004;140(4):457-460.

4. El Sayed F, Ammoury A, Dhaybi R, et al. Rosaceiform eruption to pimecrolimus.J Am Acad Dermatol. 2006;54(3):548-549.

5. Soter NA, Fleiscer AB, Webster GF, et al. Tacrolimus ointment for the treat-

ment of atopic dermatitis in adults, II: safety. J Am Acad Dermatol. 2001;44(1)(suppl):S39-S46.

Pneumocystis carinii Pneumoniain Infant Treated With Oral Steroidsfor Hemangioma

I nfantile hemangiomas are common, benign, and self-limited tumors of infancy. A recent multi-institutionstudy found that 30% of infants experience compli-

cations related to their hemangioma.1 Of those seen in ter-tiary medical centers, 12% require treatment with oral cor-ticosteroids.1 Ulceration, ocular compromise, and risk ofpermanent disfigurement are the most frequent reasons forintervention. There are no agents approved by the US Foodand Drug Administration for the treatment of hemangio-mas, but current practice includes the use of high-dose cor-ticosteroids. The beneficial effects of corticosteroids werefirst described in the 1960s, and despite decades of treat-ment, the literature consists primarily of case reports andsmall case series. The effects of steroid administration tosuch a young patient population have not been systemati-cally evaluated. High-dose corticosteroids are an indepen-dent risk factor for the development of Pneumocystis cari-nii pneumonia (PCP).2 However, most physicians treatinginfants with oral steroids for hemangiomas do not use an-timicrobial prophylaxis. We report the second known caseof PCP developing in an infant treated with oral cortico-steroids for a hemangioma.

Report of a Case. A girl with a large facial infantile he-mangioma not affecting her airway was treated by an out-side institution with 15 mg of oral prednisolone daily be-ginning at age 2 months (Figure). Attempts to wean herfrom steroids had failed owing to rapid growth of thehemangioma.

At age 7 months, she developed severe respiratory dis-tress with fever, increased respiratory rate, intercostal re-tractions, grunting, and bilateral lung infiltrates on chestradiograph. She was transferred to a pediatric intensivecare unit and required mechanical ventilation. Analysisof endotracheal tube aspirates revealed the presence ofPCP. She responded well to treatment with intravenoustrimethoprim sulfamethoxazole but still required intu-bation for 10 days.

By day 15, she had made a full recovery and was dis-charged home on a gradual oral corticosteroid taper andoral trimethoprim sulfamethoxazole twice weekly for PCPprophylaxis. An extensive laboratory workup revealed nounderlying primary immunodeficiency, and her infec-tion was attributed to immunosuppression secondary tooral corticosteroid therapy.

Comment. Pneumocystis carinii pneumonia is a poten-tially devastating opportunistic infection that is com-mon in patients with AIDS. Less frequently, PCP is seenin patients with hematologic malignancies, organ trans-plantation, inflammatory disorders, and solid tumors. Inpatients without AIDS who develop PCP, greater than 90%of them were undergoing treatment with systemic cor-ticosteroids when diagnosed.2

Figure. Inflammatory papules, pustules, and closed comedones overpartially repigmented lesions of vitiligo on the cervical region and chin.

Lucio Bakos, MD, PhDRenato Marchiori Bakos, MD, MSc



To our knowledge only 1 similar case has been re-ported.3 A 2-month-old infant developed severe respira-tory distress from diffuse airway hemangiomas and wastreated with 3.5 mg/kg per day of prednisone. After 5weeks of therapy, she was readmitted to the hospital withlabored breathing and required mechanical ventilation.A bronchoalveolar lavage revealed PCP. Intravenous tri-methoprim sulfamethoxazole was administered, and me-chanical ventilation was discontinued after 3 days.3

Trimethoprim sulfamethoxazole is effective prophy-laxis against the development of PCP. The most ac-cepted regimen is 150 mg of trimethoprim/m2 per dayplus 750 mg of sulfamethoxazole/m2 per day divided twicedaily on 3 consecutive days weekly.4 Currently, there areno PCP prophylaxis guidelines for infants undergoing oralcorticosteroid treatment for hemangiomas. Prophylaxisfor PCP should be a consideration in this setting, giventhe 2 case reports now in the literature.

Correspondence: Dr Maronn, Department of Dermatol-ogy, Medical College of Wisconsin, 9200 W WisconsinAve, Milwaukee, WI 53226 ([email protected]).Financial Disclosure: None reported.

1. Haggstrom A, Drolet BA, Chamlin SL, et al. Prospective study of infantile hem-angiomas: demographic, prenatal and perinatal characteristics. J Pediatr. 2007;150(3):291-294.

2. Yale SH, Limper AH. Pneumocystis carinii pneumonia in patients without ac-quired immunodeficiency syndrome: associated illnesses and prior cortico-steroid therapy. Mayo Clin Proc. 1996;71(1):5-13.

3. Aviles R, Boyce TG, Thompson DM. Pneumocystis carinii pneumonia in a3-month-old infant receiving high-dose corticosteroid therapy for airwayhemangiomas. Mayo Clin Proc. 2004;79(2):243-245.

4. Hughes WT, Rivera GK, Schell MJ, et al. Successful intermittent chemopro-

phylaxis for Pneumocystis carinii pneumonitis. N Engl J Med. 1987;316(26):1627-1632.

Clinical and Dermoscopic Featuresof Agminated Blue Nevus

B lue nevi are usually solitary but may rarely ap-pear grouped in an agminated pattern. Clini-cally, they present as a group of multiple blue nevi

in a circumscribed skin area smaller than 10 cm in di-ameter.1 The skin between the papules may not be dis-colored and may show a speckled or uniform blue-black or brown pigmentation.1 We report a case ofagminated blue nevus.

Report of a Case. A 59-year-old woman presented withan asymptomatic dermatosis localized on the anteriorlower extremity of her left leg characterized by brown-blue macules and papules. It initially developed duringher late teens, and although its size had increased slightlyover the earlier years, during the last 20 years, its size,color, and texture had not changed. No family history

Figure 1. Clinical image revealing numerous grouped pigmented lesions onthe left leg of a 59-year-old woman. A cluster of about 20 blue-brown to slategray macules and papules of various sizes, some of which were confluent,can be observed (original magnification �10).

∗∗

Figure 2. Dermoscopic image of part of the lesion exhibiting a cluster of bluenevi on the anterior lower extremity on the left leg of a 59-year-old woman.Multiple, grouped, homogeneous, confluent, steel blue to brown-bluepigmented areas fading into the surrounding skin can be seen. On the surface ofsome steel blue areas, linear pigmented structures appear as darker sulci(diamond). In addition, blue areas, some globules and dots (black arrow), abrown veil (black arrowhead), and brown-blue and out-of-focus pseudopods(white arrows) can be recognized. A small rim of tan pigmentation (asterisk)can also be seen around some blue nevi (original magnification �10).

Figure. Girl with large facial infantile hemangioma.

Mandi L. Maronn, MDTimothy Corden, MDBeth A. Drolet, MD



of melanoma or similar nevi was reported. Clinical ex-amination revealed a group of approximately 20 blue-brown to slate gray macules and papules of various sizes,some confluent, and within a perimeter of 4�2 cm; theskin between the macules and papules had either the colorof the surrounding normal skin or a brown-gray pigmen-tation (Figure 1).

In the dermoscopic image (Figure 2), a cluster oflesions was recognizable, all with a homogeneous pat-tern, the morphologic hallmark of blue nevus, charac-terized by multiple, grouped, homogeneous, confluent,steel blue to brown-blue pigmented areas. On the sur-face of some of the steel blue areas, typical linear pig-mented structures2 were present (Figure 2). A brown veil,a small rim of tan pigmentation, gray-blue globules anddots, and out-of-focus pseudopods were also seen withinthe diffuse pigmentation (Figure 2).

Two biopsy specimens were taken with a 5-mm skinpunch, one of a steel blue macule, the other of a brown-blue papule. The histopathologic examination revealeda superficial type of blue nevus characterized by a band-like proliferation of dendritic melanocytes in the super-ficial dermis (Figure 3). Both biopsy specimens showedidentical features of a superficial blue nevus, but the sec-ond one showed a slightly more pronounced melanin pig-mentation of the epidermal basal layer (Figure 3A). Usingboth the clinical and dermoscopic images, we diag-nosed the lesion as a speckled or agminated blue nevus.

Comment. The few cases of agminated nevi described inthe literature have been reported as plaque type,3 erup-tive,4 and patch type,5 corresponding histopathologi-cally to common or cellular blue nevus.1 Our case wasclinically similar to one described by Hendricks4 as aneruptive blue nevus.

Dermoscopically, in addition to structureless areas, ourcase presented the local features associated with blue nevialready reported in the literature.2

Agminated blue nevi may present a diagnostic prob-lem in that the differential diagnosis includes agmi-nated blue nevus or agminated intradermal Spitz nevuscombined with speckled lentiginous nevus.5 Since dif-ferential diagnosis of these nevi must also include mela-noma and malignant blue nevus, a biopsy and histo-pathologic examination are necessary for the diagnosis.4

Our case is of special interest because for the first timeto our knowledge, the dermoscopic features of this rareentity are described.

Correspondence: Dr Pizzichetta, Division of Medical On-cology C–Preventive Oncology, Centro di Riferimento On-cologico, Via Pedemontana Occidentale 12, I-33081 Avi-ano, Italy ([email protected]).Financial Disclosure: None reported.Additional Contributions: Anna Maria Colussi, RN, per-formed editing for this article.

1. Velez A, del-Rio E, Martin-de-Hijas C, Furio V, Sanchez Yus E. Agminatedblue nevi: case report and review of the literature. Dermatology. 1993;186(2):144-148.

2. Stolz W, Braun-Falco O, Bilek P, Landthaler M, Burgdorf WC, Cognetta AB.Color Atlas of Dermoscopy. 2nd ed. Berlin, Germany: Blackwell Publishing;2002.

3. Pittman JL, Fisher BK. Plaque-type blue nevus. Arch Dermatol. 1976;112(8):1127-1128.

4. Hendricks WM. Eruptive blue nevi. J Am Acad Dermatol. 1981;4(1):50-53.5. Misago N, Narisawa Y, Kohda H. A combination of speckled lentiginous ne-

vus with patch-type blue nevus. J Dermatol. 1993;20(10):643-647.

A

B C

Figure 3. Histopathologic examination revealed a superficial type of blue nevus characterized by a bandlike proliferation of dendritic melanocytes in the superficialdermis. A, Dense, diffuse, bandlike proliferation of dendritic melanocytes in the superficial dermis. Note that there is no junctional involvement at all. An increasedmelanin pigmentation of the epidermal basal layer can also be seen (hematoxylin-eosin, original magnification � 40). B, Higher magnification reveals morphologicdetails of numerous dendritic melanocytes and also a few melanophages (hematoxylin-eosin, original magnification �400). C, Scanning magnification nicelyshows that all the dendritic melanocytes are clearly S-100 positive (S-100 staining, original magnification � 40).

Maria A. Pizzichetta, MDH. Peter Soyer, MDCesare Massone, MDLorenzo Cerroni, MD



skINsight

SECTION EDITOR: JAMES M. GRICHNIK, MD, PhD

Subungual ExostosisMarıa Elena Sanchez-Castellanos, MD; Cecilia Sandoval-Tress, MD; Patricia Ramırez-Barcena, MD;Instituto Dermatologico de Jalisco “Dr. Jose Barba Rubio,” Jalisco, Mexico (Drs Sanchez-Castellanosand Ramırez-Barcena). Dr Sandoval-Tress is in private practice in Jalisco.

S UBUNGUAL EXOSTOSIS IS AN OUTGROWTH OF

normal bone or calcified cartilaginous re-mains characterized by the triad of pain, naildystrophy, and characteristic radiographicfeatures.

Case 1 involves a 13-year-old boy who presentedwith a 12-month history of an enlarging and painfullesion localized on the dorsum of the right great toe thatappeared after a trauma. Physical examination revealedthe presence of a flesh-colored, well-circumscribedelevation that measured 2.0�1.5 cm that caused ony-cholysis of the affected nail (Figure 1A).

In case 2, a 17-year-old boy presented with a 4-yearhistory of an asymptomatic growth localized on

the third right toe. Physical examination revealed anerythematous, well-defined growth that measured1.5�1.0 cm that caused onycholysis of the affected nail(Figure 2A).

In both cases, clinical diagnosis of subungual exos-tosis was established and confirmed by radiographic ex-amination that showed a well-defined osseous growth(Figure 1B and Figure 2B). Histologic analysis showed aproliferative fibrocartilaginous cap that merged into hya-line cartilage forming mature trabecular bone at its baseby enchondral ossification. Both lesions were removedsurgically, resulting in complete relief of symptoms withno signs of recurrence.

BA

Figure 1.

BA

Figure 2.



archives of dermatologylib.ajaums.ac.ir/booklist/593498.pdfsarah h. cash, md; tara t. dever, md;...

Documents