Review Article

Cancer-associated genodermatoses: apersonal history

Introduction

David Bickers and I are of roughly the sameage and our paths have crossed many timesover the years. Our special interest is nourish-ing German–American interactions in derma-tology. As our scientific interests have neveroverlapped, I instead offer, for David’s 65thbirthday, a personal recollection of theprogress which has been made in bringingcancer-associated genodermatoses from earlywarning signs for systemic malignancies toextensively researched genetic disorders, whoseresponsible genes have given importantinsights into the molecular biology of cancer.My two special interests have been genoder-matoses and dermatopathology – not two sub-jects one instinctively links – but I will alsotry to show how in this group of disordersthey fit together nicely.Another link, which David and I have, is

Columbia University. I certainly would not behere today but for the fact that a red-headedAlabamian grade school teacher and a youngGerman law student both lived at InternationalHouse 1 year in the 1930s while pursuingadvanced degrees at Columbia. The two met,got married, and I am their only child. Myother Columbia link is through the man who

stimulated my interest in clinical genetics, BobGorlin. He in turn was introduced to clinicalgenetics by a dermatologist, Helen OllendorffCurth, a long-time clinical faculty member atColumbia, who was an international expert onacanthosis nigricans.

Why bother?

As I have lectured across the USA and Eur-ope about cancer-associated genodermatosesover the past 30 years, the almost invariablequestion is ‘Why bother with such rare dis-eases?’. One of the great attractions of derma-tology for me was the possibility to walk intoa room, glance at a patient and make a pro-nouncement such as ‘You ought to have yourthyroid checked; you have a increased risk ofa medullary thyroid carcinoma.’ In manyinstances, cutaneous findings precede systemictumors by many years and the possibility foreither careful monitoring, prophylactic treat-ment or genetic counselling of relatives at riskis available. In addition, it has been clearsince the days of Warthin (1) a century agothat familial cancer has the potential to offerinsights into sporadic cancer. In no field hasthis played out to be more true than in

Burgdorf WHC. Cancer-associated genodermatoses: a personalhistory.Exp Dermatol 2006: 15: 653–666. � The Author 2006. Journalcompilation � 2006 Blackwell Munksgaard.

Abstract: Cancer-associated genodermatoses are a group of geneticdisorders inherited in an autosomal-dominant fashion in which uniquecutaneous findings are a reliable marker for the risk of developinginternal malignancies. The historical, clinical and dermatopathologicalaspects of basal cell nevus syndrome, Muir–Torre syndrome, Cowdensyndrome, Carney complex and Birt–Hogg–Dube syndrome arereviewed in a personal and informal fashion. The latest advances inthe molecular genetics of the disorders are also summarized.

Walter H. C. Burgdorf

Department of Dermatology, LudwigMaximilian University, Munich, Germany

Key words: basal cell nevus syndrome – Birt–Hogg–Dube syndrome – cancer-associatedgenodermatoses – Carney complex – Cowdensyndrome – Muir–Torre syndrome

Walter Burgdorf, MDTraubinger Str. 45A82327 TutzingGermanyTel.: +49 8158 7159e-mail: wburgdorf@gmx.de

Accepted for publication 9 June 2006

653

Experimental Dermatology 2006: 15: 653–666Blackwell Munksgaard . Printed in Singaporedoi: 10.1111/j.0906-6705.2006.00463.x

Copyright � The Author 2006. Journal compilation � 2006 Blackwell Munksgaard

Experimental Dermatology

ISSN 0906-6705

cancer-associated genodermatoses. As a der-matopathologist, one fascinating aspect is thatmultiple cutaneous tumors, usually adnexalneoplasms, often indicate a genodermatosis,while a solitary but microscopically identicallesion is of little significance (2).

Basal cell nevus syndrome

Abbreviation BCNSOMIM 109400Gene PTCH at 9q22.3Cutaneousfindings

Multiple basal cell carcinomas, palmo-plantar pits

Systemicmalignancies

Medulloblastoma, ovarian fibrosarcoma

Other clinicalfindings

Macrocephaly, odontogenic keratocyst, cardiacfibroma, ovarian fibroma, fetal rhabdomyoma, calci-fication of falx cerebri

Synonyms Gorlin syndrome, Gorlin–Goltz syndrome,nevoid basal cell carcinoma syndrome

My first introduction to basal cell nevus syn-drome (BCNS) happened in 1974 as a first yeardermatology resident at the University of Min-nesota. A delightful lady, M.S. (Fig. 1), was aregular patient of my chairman, Bob Goltz. Shewas the first patient Bob Gorlin and he haddescribed in their classic paper on BCNS in theNew England Journal of Medicine in 1960 (3).M.S. has had hundreds of basal cell carcinomas(BCC) removed, most superficial, and was veryfussy about allowing residents to work on her.She, for unclear reasons, accepted me, and for

the next few years, I had the honour of curet-ting or excising many of her lesions. She taughtme to always listen to the patient, as she couldfind a new BCC better than any physician. Thesecond patient with BCNS was L.P., who pre-sented in an extraordinary way. Working in apet shop, he was bitten by a poisonous snake,which had been given to Owen Wangensteen,the revered Chairman of the Department ofSurgery at Minnesota. A friend in the storetook L.P. to the hospital on his motorcycle andthey were involved in an accident. When BobGorlin first met L.P., the latter not only had afractured femur but also a funny-shaped headand lots of skin cancers.

The next lesson I learned was as a junior fac-ulty member in Oklahoma. The residentsshowed me a 7-year-old boy with a few tiny tel-angiectatic papules on his fingers. They weredisturbingly confident and I was totally con-fused, offering ‘treated warts’ as a diagnosis.The answer was BCNS – tiny papules in a childwith palmar pits and fairly obvious frontalbossing. The residents were so sure because theyhad been treating the child’s mother, grand-mother and great-grandfather for years. In thisfamily, ‘little skin cancers’ were nothing extra-ordinary, and surely no indication for gettingexcited.

The relevant dermatopathology question hereis – can one see anything microscopically toindicate which BCC are taken from patientswith BCNS? The answer was long ‘Yes, lesionsfrom syndrome patients are more likely to cal-cify.’ This surely has not been my experienceand in any case, looking at a patient is a muchquicker way to ascertain association. Once thegene had been identified, then this questionbecame more interesting and we will return toit.

A dermatologist was part of the researchteam that helped identify the PTCH gene, ahomologue of the patched gene in Drosophila.Erwin Epstein Jr from UC-SF, a close personalfriend of David’s, moved from clinical practiceback into the laboratory to associate with aworldwide group of researchers and produce aclassic paper in 1996 (4). Since then, the worldof PTCH has become incredibly complex andlinguistically challenging. In a simplified form,PTCH encodes a transmembrane protein Ptchwhich interacts with Smoh protein (fromSMOH or smoothened gene) to serve as recep-tor for the sonic hedgehog signalling moleculeShh (5). When Shh is missing, Ptch suppressesSmoh; when Shh binds, Smoh is freed from

Figure 1. M.S. Gorlin and Goltz’s first patient with basal cellnevus syndrome with scars where hundreds of basal cell carci-nomas had been removed over the years.

Burgdorf

654

inhibition with increased signal transductionand activation of target genes. Thus, PTCH is,in one sense, a tumor-suppressor gene andSMOH is a proto-oncogene. Mutations in bothare important in the development of sporadicBCC (6); thus, understanding this rare syn-drome has done what I and many others sug-gested would be possible – given us an insightinto the most common human malignancy(7,8).Returning to the microscopic appearance of

the tumors, it became clear in the 1990s thatBCC were derived from hair germ in mostcases, not from ordinary basal layer keratino-cytes. The UC-SF group with Erv Epsteinand Phil LeBoit noted that PTCH +/) micedevelop follicular neoplasms which resembletrichoblastomas, but when they are exposedto UV or ionizing radiation, the tumorsevolve in BCC (9). The human equivalentmay well be the infundibulocystic BCC des-cribed by Tozawa and Ackerman in 1987(10), and expanded upon by Walsh and Ack-erman a few years later (Fig. 2) (11). Theyemphasized that any type of BCC could befound in BCNS patients with nodular beingmost common, but pointed out that when abiopsy was encountered where virtually everyinfundibulum was associated with a smallinfundibulocystic BCC, the diagnosis was BCNS– or perhaps in rare cases multiple hereditaryinfundibulocystic BCC without other features ofBCNS (12). A most important parallel clinicalpoint is that 10% of patients with BCNS haveabsolutely no BCC (13).A picture is still worth 1000 words. One

afternoon a few years ago, I was moonlight-ing in a dermatology practice in Munich whena bus driver came in. He had BCNS, knownin Germany as Gorlin–Goltz-Syndrom and

was very impressed that I actually knewGorlin and Goltz; he soon demanded to seeonly me, the temporary help. I brought in apicture (Fig. 3) of the three of us togetherwith Mark Allen Everett and after that hewas in awe of my knowledge, which had notradically changed. Then, I decided to askboth Bob Goltz and Bob Gorlin how theypreferred to have their syndrome named. BobGorlin in a modest way said he identifiedmore with BCNS than with any other of hissyndromes; in the fourth edition of Syndromesof the Head and Neck (14), he used the titleGorlin (nevoid BCC) syndrome. Bob Goltzsaid he was happy to be associated in anyway.

Finally, in contrast to the other syndromesmentioned below, systemic malignancies arenot a great problem in BCNS. The main riskis childhood medulloblastomas. The exactrisk is hard to determine because medullobla-stoma is the most common childhood braintumor, accounting for about 25% of suchlesions (15). Only 3% of all medulloblastomasoccur in patients with BCNS, while about 5%of BCNS patients develop medulloblastoma.Returning to the laboratory, the activation ofthe Shh pathway predisposes neural progen-itor cells to develop into medulloblastomas byinactivating the retinoblastoma tumor-suppres-sor gene (Rb) and inducing the proto-onco-gene N-myc (16). The main clinical factor isthat radiation therapy is the treatment ofchoice for medulloblastoma; often children areirradiated before the diagnosis of BCNS hasbeen made. As the skin of BCNS patients isexquisitely radio-sensitive, such unfortunateindividuals often develop thousands of BCCin the radiation fields.

Figure 2. Infundibulocystic basal cell carcinoma (photomicro-graph courtesy of Peter Soyer, MD, Graz, Austria).

Figure 3. Mark Allen Everett, Robert J. Gorlin, Walter Burg-dorf, Robert W. Goltz; Albuquerque NM, 1993.

Cancer-associated genodermatoses

655

Muir–Torre syndrome

Abbreviation MTSOMIM 158320Genes Mismatch repair genes

MSH2 at 2p22–p21MLH1 at 3p21.3

Cutaneous findings Multiple sebaceous tumors (especiallycystic sebaceous tumors),multiple keratoacanthomas

Systemic malignancies Carcinomas of colon, endometrium,ovary and many other organs

Synonyms Hereditary non-polyposis colorectal cancer,Lynch syndrome, cancer family syndrome

In 1979, Aly Fahmy, a general pathologist atthe Oklahoma City Veterans AdministrationHospital, asked me to come by and look atsome slides which he and his colleague JanPitha found interesting. They were from apatient whom I already had seen in clinic; R.S.had had lots of funny cutaneous tumors (Fig. 4),but I knew little else about him. Aly and Janwere way ahead of me. In those days, beforecomputers, they had little index cards containingall the previous pathological diagnoses on eachpatient and R.S.’s card was most interesting. Inaddition to keratoacanthomas and sebaceoustumors, he had already multiple primary gastro-intestinal carcinomas at a relatively early age.Dr Elson B. Helwig at the Armed Forces Insti-tute of Pathology had agreed that they were allprimary tumors. R.S. died several years laterfrom radiation-induced pneumonia, but wastumor-free after having nine primary systemicmalignancies over more than a decade.I was still in the dark, but Aly and Jan knew

the answer – Muir–Torre syndrome (MTS).Douglas Torre, one of the pioneers in dermato-logical cryotherapy and a private practitioner inManhattan, had presented a patient at the New

York Dermatologic Society with several gastro-intestinal tumors, prolonged survival and a col-lection of peculiar cutaneous tumors includingan ‘epidermoid cyst with squamous cell carci-noma’ and multiple poorly defined sebaceoustumors (17). In the interval between Torre’spresentation and the publication of the meetingproceedings in the Archives of Dermatology,E. G. Muir, a surgeon, and his colleagues fromLondon described a Maltese patient with mul-tiple keratoacanthomas, sebaceous neoplasmsand gastrointestinal tumors (18). Dr Helwigand his fellow Dave Rulon, a dermatopatholo-gist, had also recently written about the associ-ation of sebaceous neoplasms and visceraltumors but had lacked sufficient information todiagnose our case (19).

About this time, two exogenous factorsforced me to become further interested inMTS. First, my chairman, Mark Allen Everett(also shown in Fig. 3), had agreed to arrangea dermatopathology meeting in New Orleansand simply told me I would be presenting‘something’ there. I asked Aly if he wouldhelp me and he did. We identified two otherveterans with multiple malignancies, and Ithen had a ‘moment of understanding’, recall-ing a patient I had presented to HermannPinkus a few months before. J.M. was ayoung lady with a personal history of thyroidcancer, a family history of gastrointestinalcancer and a single red papule on her thigh.The lesion was a small basaloid tumor with afew sebaceous cells scattered throughout;Pinkus had been stumped but identified thesebocytes. She became our fourth MTSpatient. All fit the pattern – multiple systemictumors, early onset and relatively long survi-val. After our paper on R.S. written by BobSchosser, our dermatopathology fellow, hadbeen accepted by Cancer (20), Wolf-IngoWorret, a young German visiting dermatopa-thologist, helped us write up all four patientsin the German literature. Because of thevagaries of publishing schedules, the paper inHautarzt appeared first with the initial Ger-man-language mention of MTS (21).

Next, Henry Lynch asked me to write achapter in a book, he and Ramon Fusaro wereediting, entitled Cancer-associated Genoderma-toses. My assignment was ‘Dermatopathologicaspects of cancer-associated genodermatoses’(22). This chapter in 1982 convinced me that Iwas not wasting my time trying to combinedermatopathology and clinical genetics. I hadthe opportunity to learn from another master

Figure 4. R.S. My first patient with Muir–Torre syndrome;notice the sebaceous carcinoma of the chin, as well as the typ-ical paranasal lesions and scars.

Burgdorf

656

clinician-geneticist, Henry T. Lynch of Omaha.Henry shared his insights into MTS with meand convinced me that MTS was simply part ofthe cancer family syndrome (CFS), today knownas hereditary non-polyposis colorectal cancer 1and 2 (HNPCC1 and HNPCC2) (Lynch syn-drome 1 and 2) (23–25). He had several largepedigrees of CFS, some of which he had inher-ited from Warthin’s files at the University ofMichigan (26). In this collective, Henry hadalready identified a few patients with sebaceousneoplasms (27,28), whom we would now identifyas either MTS or HNPCC2 with colonic andextracolonic tumors. HNPCC is caused bymutations in a variety of DNA mismatch repair(MMR) genes; MSH2 (29) is almost alwaysresponsible for HNPCC2 (MTS), while MLH1(30) is equally important in HNPCC1. Therevised Bethesda guidelines incorporate much ofthe carefully acquired clinical and genetic datafrom Lynch and his colleagues (31).Our first patient nicely summarizes the entire

HNPCC story; he developed multiple tumors,primarily gastrointestinal, but also involvingother organs, at a relatively early age but sur-vived far longer than age-matched controls withsimilar sporadic tumors (32). Another distin-guishing feature is that the colon carcinomasare far more likely to be on the right side, as inour case, than in familial adenomatous polypo-sis (FAP) with its cutaneous marker, Gardnersyndrome. By then, I had numerous biopsiesfrom our four patients with MTS and hadreceived cases in consultation from around thecountry. So, I had a burgeoning collection. Iwas becoming more and more convinced thatthe cutaneous lesions in MTS frequently defiedclassification, often showed overlaps betweensebaceous tumors and keratoacanthomas andon occasion were most distinctive. All thisinformation I published with Jan and Aly in anarticle in American Journal of Dermatopathologyin 1986 (33).Some 15 years later, Roland Kruse, a young

German dermatologist at that time working inthe genetics department in Bonn, called me todiscuss MTS. His group made many new dis-coveries about MTS, showing they had thesame spectrum of MMR gene mutations asHNPCC (34), although MSH2 was more likelyto be involved (35). They also showed the pres-ence of microsatellite instability in skin tumors(36) and demonstrated that the loss of hetero-zygosity (LOH) was not the preferred mode oftumor development in MTS (37). Rolandplaced me in contact with Arno Rutten, a der-

matopathologist in Friedrichshafen, Germany,who had a long interest in MTS and had provi-ded much of the pathological material forRoland’s studies.

I agreed to bring my collection of biopsies toFriedrichshafen, where Arno and I spent a daylooking at funny sebaceous lesions. We decidedthat the most distinctive tumor was what wethen designated a cystic sebaceous tumor(CST), a deep cystic sebaceous proliferation(Fig. 5) (38). I had identified such lesions inpassing a decade earlier, but now Arno foundat least one in half the patients in his much lar-ger collective. Prospectively, we identified sev-eral patients who had only a CST, butinvestigation revealed a personal or family his-tory that strongly suggested MTS and thengermline DNA repair gene mutations werefound (39).

Mathiak, Rutten and others refined the pro-cess even further. Their group employed com-mercial monoclonal antibody stains with whichthe presence of MSH2 or MLH1 protein hadbeen identified in gastrointestinal biopsy sec-tions (40). In CST and other tumors withmicrosatellite instability, in most instances,there was no staining for the appropriate geneproducts in the cutaneous tumors. Thus, it waspossible to exactly identify the gene harbouringthe germline mutation responsible for a wholespectrum of problems simply by staining theright skin biopsy in the right way. This is trulyan elegant combination of dermatopathologyand genetics.

Figure 5. Cystic sebaceous tumor, deep in dermis with obvi-ous glandular areas (histological material courtesy of ArnoRutten, MD, Friedrichshafen, Germany).

Cancer-associated genodermatoses

657

Cowden syndrome

Abbreviation CSOMIM 158350Gene PTEN at 10q23.31Cutaneous findings Facial, acral and oral mucosal papules;

lipomas, haemangiomas,pigmented macules

Systemic malignancies Carcinomas of breast, thyroid;hamartomatous polyposis withlow risk of colon cancer

Other clinical findings Macrocephaly, cerebellar dysfunctionSynonyms Multiple hamartoma syndrome

I first learned about Cowden syndrome (CS)under the wrong name – multiple hamartomasyndrome. Peyton Weary, Bob Gorlin, BillGentry (my favourite clinical teacher at Minne-sota) and several of Peyton’s colleagues fromVirginia published a paper on this disorder in1972 (41). As it was a Minnesota-Virginiapaper, I accepted it as the whole answer. I sawseveral of Bill Gentry’s patients and many biop-sies but never grasped the essence of the dis-ease. In 1976, Bob Goltz took me with him tothe Michigan Dermatologic Society meetingheld one spring weekend in Dearborn Villagenear Detroit. My most vivid memory of themeeting is a case discussed by Amir Mehreganin which the patient had a baffling array ofcutaneous tumors – lipomas, hair follicle tum-ors, warts and peculiar fibrous lesions – andwas introduced under a complex new namewhich I have forgotten. Bob Goltz was one ofthe guest discussants and in his inimitable waypraised the clinicians, praised Amir, indicatedthis was just about the most puzzling thing hehad ever seen, but asked if he could pose apossible alternative diagnosis – multiple hamar-toma syndrome. The minute he said it, everyonein the audience simultaneously clapped theirforeheads in the ‘Oh, why didn’t I think ofthat?’ routine, and I became a fan of thisdisorder.I went to the library immediately on the fol-

lowing Monday to read about my new interestand as always bumped into Bob Gorlin, themost conspicuous user of Diehl Hall (the med-ical library) at the University of Minnesota. Hereminded me that Ken Lloyd in Youngstown,OH, had first described the syndrome and toldme of his personal first encounter with CS. Hewas visiting the Armed Forces Institute ofPathology and was shown a tray of slides byDr Southern Hooker, a descendent of GeneralFighting Joe Hooker of Civil War fame. I later

asked Bob to write down these events and nowI quote…

It was the case of a woman who had lots ofbumps on her lips, gingiva and skin, who hadhuge breasts and breast cancer. About sixmonths later Lauren Ackerman contacted meand said, ‘Bob I have a patient with this, this,this and this.’ I said, ‘My goodness, Lauren, Ijust (saw slides) of a gal with almost the identi-cal findings.’

The doctors caring for the first case were twoyoung internal medicine residents in Youngs-town, OH, Kenneth Lloyd and Macey Dennis.They published their case in Annals of InternalMedicine in 1963 (42), choosing the name Cow-den syndrome to honour their patient RachelCowden, who later died of metastatic cancer.Ken Lloyd went on to become a dermatologistand still practises today in Youngstown.

Carcinoma of the breast remains the majorrisk for CS patients. About 75% of femalepatients have fibrocystic disease or otherbreast changes and fully a quarter developcarcinoma. Prophylactic mastectomy is oftenrecommended (43). In addition, recent evi-dence shows that male individuals with CSare also at risk for breast cancer (44). Around50% of patients have hamartomatous polypsof the gastrointestinal tract, but the risk ofmalignant degeneration is very small. Simi-larly, about two-thirds have thyroid disease,but once again, malignant change is rare.Theo Starink, a dermatologist in Amsterdam,accumulated considerable data from severallarge Dutch pedigrees (45).

Cowden syndrome evolved into an even moreinteresting dermatopathological issue. The faciallesions were identified as trichilemmomas(46,47), and the dermal lesions occasionallyshowed a peculiar cut-onion layered patternwhich was designated as sclerosing fibroma(48). Although multiple sclerotic fibromas werefirst described, soon solitary sporadic ones wereidentified (49). Two giants of dermatopathology,Marty Brownstein and Bernie Ackerman,argued for years whether the trichilemmomaswere warts or not. Bernie was sure all trichi-lemmomas were warts (50), and Marty was justas sure they were not (51).

While at the University of New Mexico, Isaw a college freshman, who had presented asan infant with an enlarged head but did nothave hydrocephalus and now in college, washaving trouble concentrating and developinglots of warts. The referring physician hadalready thought of CS, so I could only agree,

Burgdorf

658

but I learned that many patients are originallyevaluated by paediatricians for increased headcircumference. I also found out that Lhermitte–Duclos syndrome (52), a form of cerebellarhypertrophy, was closely associated with CS(53).The next major jump in our knowledge of CS

occurred at Columbia University. The group ofRamon Parson in the Institute for Cancer Gen-etics reported, in early 1997 in Science, on thefinding that the PTEN gene, a member of theprotein tyrosine phosphatase family, is mutatedin human brain, breast and prostate cancers(54). Parson’s group had already joined effortswith the group of Monica Peacocke in DavidBicker’s dermatology department. Together,these researchers found out that mutations inthe PTEN gene are also responsible for CS(55). Subsequently, diagnostic criteria wereestablished for CS and everything seemed clear(56). The diagnostic criteria included mucocuta-neous lesions:1. six or more facial papules are present, of

which three or more must be trichilemmomas;2. cutaneous facial papules and oral mucosal

papillomatosus are present;3. oral mucosal papillomatosus and acral

keratoses are present;4. six or more palmoplantar keratoses are

present.Mutations in PTEN are responsible for a

befuddling list of syndromes, which are some-times given the group name Bannayan–Riley–Ruvalcaba (57) and at other times furthersubdivided into Bannayan–Zonana syndrome(58), Riley–Smith syndrome and Ruvalcaba–Myhre–Smith syndrome. Even the names areconfusing – two different Smiths were involved;Harris ‘Pete’ Riley was a long-time Chairman ofPediatrics at Oklahoma and kind to me when Iran paediatric dermatology clinics at OklahomaChildren’s Hospital. Today, we know all thesealphabet soup syndromes are caused by PTENmutations (59) and can be viewed as ‘PTENosis’or paediatric variants of CS, featuring haeman-giomas, lipomas, juvenile polyps, macrocephalyand the classic pigmented macules of the penis(speckled pecker sign). In addition, mutations inPTEN also cause Lhermitte–Duclos syndrome.The association with juvenile polyposis syn-drome has been more confusing; latest informa-tion suggests that PTEN mutations shouldexclude this diagnosis (60), while two other caus-ative genes for JPS have been identified. Thediagnostic criteria are now more comprehensiveand address the PTEN spectrum (61).

I met Jorg Schaller, a dermatopathologistin Duisburg, Germany, about this time. Itturned out that he had been in the Depart-ment of Dermatology at the University ofLeipzig when I lectured there during the deep-est days of the German Democratic Republic(East Germany) but had missed my talk. Atone meeting, Jorg presented the data he andAngela Rohwedder had produced, indicatingthat many solitary trichilemmomas containhuman papillomavirus (HPV) DNA (62). Itseemed to me more and more likely that allthe cutaneous lesions in CS represented differ-ent stages in the same pathological processinvolving follicular proliferation and perifolli-cular fibrosis. Clinically, everything fits withwarts – white mucosal papules; papillomatousfacial and acral tumors, sometimes involvinghair follicles; and palmo-plantar papules. PeterItin, Stephan Lautenschlager and I collectedbiopsies from seven patients with CS whichwe studied histologically with Jorg Schaller,while Angela Rohwedder looked for HPVDNA. We learned that only a small percent-age of facial papules in CS are trichilemmo-mas; our batting average was abysmal, eventhough we chose typical wart-like faciallesions (Fig. 6). We were able to confirm thatthere is a spectrum between wart, seborrheickeratosis-like lesion, trichilemmoma and scler-otic fibroma (Fig. 7), but the HPV searchtaught us another lesson. The incidence ofHPV in facial skin of adults, especially in thehair follicles, is so high that to obtain statisti-cally significant results, we would have had tostudy more CS-related skin lesions than hadever been biopsied, probably by a factor ofthousand or even million. Thus, our frequentfinding of HPV DNA did not help answerthe question (63). Personally, I wonder if

Figure 6. Elderly woman with Cowden syndrome and verru-cous facial papules (photograph courtesy of Peter Itin, MD,Basel, Switzerland).

Cancer-associated genodermatoses

659

PTEN mutations can somehow lead to anepidermal growth pattern which resembles thatinduced by HPV; I have asked lots of PTENexperts and never received an answer. Anysuggestions?

Carney complex

Abbreviation CNCOMIM CNC1 160980

CNC2 605244Genes CNC1: PRKAR1A at 17q23–q24

CNC2: CNC2 at 2p16Cutaneous findings Myxomas, multiple freckles or lentigines,

blue neviSystemic malignancies Large-cell calcifying Sertoli cell tumor,

psammomatous melanotic schwannomaOther clinical findings Cardiac myxoma, pigmented nodular

adrenocortical disease withCushing syndrome, myxoidmammary fibroadenoma

Synonyms NAME syndrome, LAMB syndrome

For many years, I read about Carney complex(CNC) and saw cases at meetings, but neverpersonally diagnosed or followed a patient. Ihad written to J. Aidan Carney, a pathologistat the Mayo Clinic, a number of times request-

ing information for lectures or review papers.The minute I read his classic paper (64), I real-ized that he was correct to lump together bothNAME (65) and LAMB (66) syndromes whichwere clearly describing the same patients withslightly different choice of words and emphasis.I was sure that when I saw a patient, I wouldimmediately recognize the admixture of frecklesand blue nevi and promptly make a brilliantdiagnosis. It turned out differently.

In the midst of a busy clinic, a resident calledme in to look at some molluscum before he cur-etted them. J.H. was a tall teenage boy, accom-panied by his mother, a hospital employeewhom I knew in passing. The resident said,‘J. has a few molluscum. I’m going to curettethem and not get pathology, as there is no needof wasting the family’s money.’ (In defense ofthe resident, we rarely submitted molluscum forpathology). But…J. had two little glassypapules on his eyelids (Fig. 8). I knew they werenot molluscum but was totally stumped. Ibriefly looked at the rest of his skin; he hadnothing else. Then his mom asked ‘Could thishave anything to do with his bilateral large-cellcalcifying Sertoli cell tumor of the testes?’ Isaid, ‘Just that these are myxomas in Carneycomplex.’ Histological examination of the tan-gentially excised papules proved me correct(Fig. 9). I examined J. again; he had absolutelyno pigmented lesions. I ran back to my officeand called Dr Carney. He excitedly told me thathe was doing a paper on myxomas of the eye-lids and external ear as almost dead-certainmarkers for CNC. Soon thereafter, Ferreiroand he published their findings in this area (67).

The extraordinary finding in my first caseswas the lack of pigmented lesions. Seventyper cent of patients have either freckles orlentigines; the matter has never been resolved.

Figure 7. Trichilemmoma with papillomatous or warty chan-ges.

Figure 8. J.H. My first patient with Carney syndrome withconjunctival myxomas.

Burgdorf

660

The ‘E’ in NAME refers to ephelides, whilethe ‘L’ in LAMB indicates lentigines; in myexperience there are both. The most strikingcutaneous findings are the myxomas (68),nodules with lacy epithelial strands and lakesof mucin. They are almost pathognomonic ofCNC. Another unusual feature is epithelioidblue nevi which should suggest CNC (69) butmay be sporadic (70).I stayed in touch with Dr Carney and

learned much from him. First of all, one mustsay Carney complex, because Carney triadrefers to gastric (epithelioid) leiomyosarcoma,functioning extra-adrenal paraganglioma andpulmonary chondroma (71). In CNC, the pre-valence of cardiac myxomas is 70% and thatof Cushing syndrome is 30%. Carney spent agreat deal of time trying to track downHarvey Cushing’s first patient, Minnie G.,with what is now known as Cushingsyndrome; it appears likely that Minnie G.actually had CNC (72). Similarly, WilliamYoung re-evaluated the first patient with pri-mary pigmented nodular adrenocortical disease50 years later and she too had CNC (73).The main problem in CNC is the cardiac

myxomas, which are histologically benign, butpotentially life-threatening. In a familialpatient, the myxomas tend to occur at earlierage than in a sporadic case, more often aremultiple, involve any chamber and tend torecur. They may cause emboli or cardiac fail-ure and the surgery itself is risky, so that per-haps 25% of CNC patients die from theirmyxomas. Psammomatous melanotic schwan-nomas usually involve the gastrointestinaltract (74) but have been reported in the skin

(75); they metastasize in about 10% of cases.The breast tumors are benign (76).

Carney complex was one of the first diseasesto turn out to be caused by mutations in atleast two different genes, without discernibledifferences in clinical appearance. Mutations inPRKAR1A (77), a protein kinase, at 17q23–q24and a not characterized gene at 2p16 (78) causeexactly the same problems. Who would havethought 30 years ago that two different genescould cause exactly the same autosomal-domin-ant disorder? The possibility never crossed mymind. Recently, I was lucky enough to be ableto join Dr Carney in writing the section onCNC in the latest World Health Organizationbook on skin tumors (79).

Birt–Hogg–Dube syndrome

Abbreviation BHDSOMIM 135150Gene FLCN (folliculin) at 17p11.2Cutaneous findings Multiple perifollicular fibromas

and trichodiscomasSystemic malignancies Renal cell carcinomas

(unusual histological types)Other clinical findings Spontaneous pneumothoracesSynonyms Hornstein–Knickenberg syndrome

I go back a very long way with Birt–Hogg–Dubesyndrome (BHDS). Arthur Birt was a dermatol-ogist in Winnipeg, Canada, who frequently vis-ited the University of Minnesota for statedermatology meetings or occasionally just Fridayafternoon grand rounds. He was often accom-panied by his colleague Arthur Anhalt, father ofGrant Anhalt (who needs no further introduct-ion to readers of this journal). This trip aloneshows the devotion of Dr Birt to dermatologicaleducation; the two cities are 450 miles (720 km)apart. Dr Birt was Winnipeg through andthrough. His father and grandfather had beenmayors of the city. He completed his entire med-ical training there and then practised in the sameoffice for over 50 years. The syndrome whichmade Winnipeg famous was initially identifiedby Jim Dube, a endocrinologist originally fromTrinidad who discovered a family with multiplethyroid cancers and funny skin lesions. Hereferred the family to Arthur Birt who biopsiedthe lesions and sent the specimens to GeorginaHogg, a pathologist. She identified them as folli-cular tumors but sought help from HermanPinkus who was able to provide an almost-cor-rect diagnosis, as discussed below. She also sent

Figure 9. Myxoma with lacy epithelial strands originatingfrom central follicle.

Cancer-associated genodermatoses

661

slides to Bob Goltz and I still have some of theoriginal slides which she shared with me. TheCanadians then published their description ofthe syndrome in 1977 (80).My first encounter with a patient with BHDS

was at the University of Oklahoma. I was sittingin my office at Oklahoma Children’s Hospitalreading a paper by Otto Hornstein and MonikaKnickenberg on a new syndrome (81). The phonerang and Dennis Weigand was calling from theVeterans Administration Hospital across thestreet. He said ‘I have a patient who has beenboarded out of the service with the diagnosis ofacne scars, but he has something else. Could youtake a look?’ I walked over, entered the roomand saw a man who looked identical to the pic-tures I had been glancing at a few minutesbefore. I confidently said, ‘He’s got Hornstein–Knickenberg syndrome.’ This tour de force con-vinced Dennis and the residents that perhaps thenew faculty member was not so bad. There arestill divided opinions, but to me it is clear thatHornstein and Knickenberg were describing thesame syndrome as Birt, Hogg and Dube; theyclearly published first – 2 years earlier.C.P. had hundreds of small white dome-

shaped perifollicular papules most prominenton face and chest, clearly not acne scars(Fig. 10). I biopsied 10 lesions; five showedperifollicular fibrosis with lacy epithelial strands(fibrofolliculoma) (Fig. 11) and five showed adisc-shaped subepidermal proliferation of spin-dle cells (trichodiscoma) (Fig. 12). We referredthe patient for internal medicine work-upbecause Hornstein and Knickenberg had men-tioned a risk of gastrointestinal carcinoma,while Birt and colleagues had suggested a poss-ible association with thyroid cancer. Nothingwas found, but he died a few years later of met-astatic colon carcinoma.

I saw several more patients clinically, but themost interesting question was the nature of thelesions under the microscope. The hair disc is awell-established sensory structure, prominent oncat and rabbit lips, and especially well seen ingloves made from the hide of the peccary, a pig-like mammal of the Southwestern Desert of theUSA locally known as javelina. Felix Pinkus firstdescribed the ‘Haarscheibe’ (German for hairdisc) in 1902 (82) and even identified it in peccar-ies (83). It consists of an acanthotic epidermisoverlying a spindle-shaped proliferation; the epi-dermis is rich in Merkel cells and the dermal ele-ments are neural, as the hair disc is a highlyspecialized neuroreceptor. Felix’s son, HermannPinkus, identified cutaneous tumors resemblinghair discs which he named trichodiscomas (84).Initially Burnier and Rejsek (85) and laterZackheim and Pinkus (86) had describedperifolliculomas. With this background, it had

Figure 11. Fibrofolliculoma with branching lacy epithelialstrands arising from follicle and surrounded by perifollicularfibrosis.

Figure 12. The original trichodiscoma described by Birt,Hogg and Dube (histological material courtesy of GeorginaHogg, MD, Winnipeg, Canada).

Figure 10. C.P. My first patient with Birt–Hogg–Dube syn-drome. Numerous follicular papules.

Burgdorf

662

been relatively easy for Hermann Pinkus to sug-gest a diagnosis for the slides sent to him byHogg. Because of the presence of perifollicularfibrosis interlaced with thin epithelial strands orfingers, the Winnipeg group as well as Weintrauband Pinkus (87) proposed the name fibrofollicu-loma.Unfortunately, the issue was more complex

and Pinkus was not 100% correct. I met one ofmy best friends over this issue. Gerd Plewig,when he was Chairman of the Department ofDermatology in Dusseldorf, Germany, becameinterested in the trichodiscomas; he and hisco-workers clearly showed that they were fibrousproliferations, not neural, in a poster presentedat the American Academy of Dermatology in1987 (88). I visited Gerd in Dusseldorf on one ofmy German trips just to see these slides and webecame close friends. Even though clinical pro-fessorships are almost non-existent in Germany,I have one enabling me to work closely withGerd in the Department of Dermatology at theLudwig Maximilian University in Munich since1993, editing and translating several books, aswell as taking numerous hikes and bike trips.The next step in understanding these lesions

was made by Bernie Ackerman who often lec-tured about what I call ‘Ackerman’s rule ofmultiple adnexal neoplasms’. I paraphrase, buthe stated, ‘If a patient has multiple lesionswhich are interpreted with two different histo-pathological diagnoses, but a good cliniciancannot tell them apart, then the lesions areprobably variants on a single process.’ This issurely the case in MTS where the sebaceouslesions and keratoacanthomas overlap and inCS where trichilemmomas and fibromas mergetogether; it is also true in BHDS where the bestname is probably that proposed by CharlieSteffen – mantleoma (89). The sebaceous mantleconsists of strands of epithelial cells that eman-ate from the infundibulum of a hair follicle anddrop down aside the follicle in the form of amantle or skirt; sebocytes appear at the tips ofthese chords and evolve into sebaceous lobulesand glands. Later, the mantle and sebaceousglands involute with fibrosis. Thus, perifollicu-lar fibrosis, lacy strands and finally pure fibrosisoccur. The connection between this normal pro-cess and the microscopic appearance of fibrofol-liculomas and trichodiscomas is obvious. TwoGerman workers, Tilman Schulz and WolfgangHartschuh, took things a step further andshowed that if one sectioned carefully, featuresof both the epithelial proliferation and fibrosiscould be found in the same specimens (90).

My most fascinating contact with BHDS –even though I contributed nothing – was aphone call from Burton Zbar at the NationalInstitutes of Health perhaps 10 years ago. PaulDuray, one of my oldest (better said, longestserving) friends, with whom I worked at Henne-pin County Hospital in the spring of 1975, hadtold Burt I was an expert on BHDS. The storyhad evolved as follows: the NIH had becomeinterested in families with multiple, often bilat-eral, renal cell carcinoma, often with unusualhistological types (usually chromophobe, onco-cytic or overlap, with the common sporadicclear cell carcinoma found in only 10%). MariaTurner and other clinicians at the NIH noticedimmediately that many of these patients hadcutaneous findings – multiple perifollicularpapules which on biopsy were mantleomas, sothey knew that BHDS was the answer (91).

In 2002, Zbar’s group was successful in iden-tifying mutations in the folliculin (FLCN) geneunderlying BHDS (92). Folliculin is a cytoplas-mic protein with a predicted size of 64 kDa.Alternative splicing of the FLCN gene results intwo isoforms, a full-length form and one thatuses an initiation codon in intron 6. Althoughthe precise function of folliculin and its alter-nate splice variant is not clear, it is thought tobe a tumor suppressor with LOH initiatingtumor formation. All germline mutations inpatients suffering from BHDS described so farare splice-site mutations or insertions/deletionsthat result in putative protein truncation andhaplo-insufficiency.

Today, it appears that Birt’s suggestions ofassociated thyroid carcinoma and Hornstein’sof gastrointestinal carcinoma were incorrect;the major tumor risk is that of renal cell carci-noma, as well as of spontaneous pneumothoraxand pulmonary cysts (93). The picture may bemore complex. My last contact with BHDS wasjust a few weeks ago. One of the junior facultiesin Gerd Plewig’s department called to tell meshe and colleagues in Nuremburg were follow-ing twin sisters with BHDS and dysplasticnevus syndrome, whose mother had similar skinlesions but had died of melanoma. One of thepatients has an endometrial carcinoma. Whoknows where this trail will lead?

My big question

One thing fascinates me about these disorders –the unexplained paradox about why they causemalignant systemic tumors and very oftenbenign cutaneous tumors. Of course, in BCNS,

Cancer-associated genodermatoses

663

the skin tumors are malignant, but generationsof dermatologists have argued about the terms‘basal cell carcinoma’ and ‘basal cell epitheli-oma’ because of the almost 0% lifetime risk ofmetastasis. In MTS, only a very rare sebaceoustumor is malignant. Our first patient had asmall-cell sebaceous carcinoma of the soft tissueof the chin, mimicking a Merkel cell tumorunder the microscope. When I presented thecase, Richard Winkelmann, a superb dermato-pathologist among his other considerable tal-ents, doubted it was a carcinoma. Severalmonths later, the patient had jaw metastases. InCS, CNC and BHDS, the cutaneous lesions arecompletely benign. I have no explanation. Onceagain, maybe, a young experimentally orientedreader of this journal can solve this puzzle forme.My purpose in writing this highly anecdotal

account has not been to highlight my ownachievements. They have been minimal and ofinterest to no one, not even to my mother.Instead, I present them just to show how if oneremains alert, communicates with everyonepossible and always tries to learn somethingnew, there are many interesting adventures tobe had in the field of cancer-associated genoder-matoses. Over the 30 some years, David and Ihave been active in dermatology. These fivesyndromes have advanced from being odd dis-eases set aside for the ‘stamp collectors’ to fas-cinating diseases, giving us new insights intocutaneous and systemic tumor development.

Acknowledgements

Many old friends and some new acquaintances helped merefresh my memory and expand my knowledge. This hope-fully complete list includes: Grant Anhalt, J. Aidan Car-ney, Michael Flaig, Jorge Frank, Robert Goltz, RobertGorlin, Wolfgang Hartschuh, Robert Jackson, Peter Itin,Roland Kruse, Peter Lane, Kenneth Lloyd, Ingrid Moll,Eileen Murray, Gerd Plewig, Julia Reifenberger, ArnoRutten, Jorg Schaller, Jack Toole and Peyton Weary.

References

1. Warthin A S. Heredity with reference to carcinoma.Arch Intern Med 1913: 12: 546–555.

2. Burgdorf W H, Koester G. Multiple cutaneous tum-ors: what do they mean? J Cutan Pathol 1992: 19:449–457.

3. Gorlin R J, Goltz R W. Multiple nevoid basal-cell epi-thelioma, jaw cysts and bifid rib. A syndrome. N EnglJ Med 1960: 262: 908–912.

4. Johnson R L, Rothman A L, Xie J et al. Humanhomolog of patched, a candidate gene for the basalcell nevus syndrome. Science 1996: 272: 1668–1671.

5. Stone D M, Hynes M, Armanini M et al. Thetumour-suppressor gene patched encodes a candidatereceptor for sonic hedgehog. Nature 1996: 384: 129–134.

6. Reifenberger J, Wolter M, Knobbe C B et al. Somaticmutations in the PTCH, SMOH, SUFUH and TP53genes in sporadic basal cell carcinomas. Br J Dermatol2005: 152: 43–51.

7. Cohen M M Jr. The hedgehog signaling network. AmJ Med Genet A 2003: 123: 5–28.

8. Taipale J, Beachy P A. The hedgehog and Wnt sig-nalling pathways in cancer. Nature 2001: 411: 349–354.

9. Aszterbaum M, Epstein J, Oro A et al. Ultravioletand ionizing radiation enhance the growth of BCCsand trichoblastomas in patched heterozygous knock-out mice. Nat Med 1999: 5: 1285–1291.

10. Tozawa T, Ackerman A B. Basal cell carcinoma withfollicular differentiation. Am J Dermatopathol 1987:9: 474–482.

11. Walsh N, Ackerman A B. Infundibulocystic basal cellcarcinoma: a newly described variant. Mod Pathol1990: 3: 599–608.

12. Requena L, Farina M C, Robledo M et al. Multiplehereditary infundibulocystic basal cell carcinomas: agenodermatosis different from nevoid basal cell carci-noma syndrome. Arch Dermatol 1999: 135: 1227–1235.

13. Gorlin R J. Nevoid basal-cell carcinoma syndrome.Medicine (Baltimore) 1987: 66: 98–113.

14. Gorlin R J, Cohen M M Jr, Hennekam R C M.Syndromes of the Head and Neck, 4th edn. NewYork: Oxford University Press, 2001.

15. Campbell R M, Mader R D, Dufresne R G Jr. Men-ingiomas after medulloblastoma irradiation treatmentin a patient with basal cell nevus syndrome. J AmAcad Dermatol 2005: 53: S256–S259.

16. Sanai N, Alvarez-Buylla A, Berger M S. Neural stemcells and the origin of gliomas. N Engl J Med 2005:353: 811–822.

17. Torre D. Multiple sebaceous tumors. Arch Dermatol1968: 98: 549–551.

18. Muir E G, Bell A J, Barlow K A. Multiple primarycarcinomata of the colon, duodenum, and larynx asso-ciated with kerato-acanthomata of the face. Br J Surg1967: 54: 191–195.

19. Rulon D B, Helwig E B. Multiple sebaceous neo-plasms of the skin: an association with multiple vis-ceral carcinomas, especially of the colon. Am J ClinPathol 1973: 60: 745–752.

20. Fahmy A, Burgdorf W H, Schosser R H et al. Muir–Torre syndrome: report of a case and reevaluation ofthe dermatopathologic features. Cancer 1982: 49:1898–1903.

21. Worret W I, Burgdorf W H, Fahmy A et al. Torre–Muir syndrome. Sebaceous gland neoplasms, kerato-acanthomas, multiple internal cancers and heredity.Hautarzt 1981: 32: 519–524.

22. Burgdorf W. Dermatopathologic aspects of cancer-associated genodermatoses. In: Lynch H T, Fusaro RM, eds. Cancer-associated Genodermatoses. NewYork: Van Nostrand, 1982: 145–170.

23. Lynch H T, Kimberling W, Albano W A et al. Hered-itary nonpolyposis colorectal cancer (Lynchsyndromes I and II). I. Clinical description ofresource. Cancer 1985: 56: 934–938.

Burgdorf

664

24. Lynch H T, Lanspa S, Smyrk T et al. Hereditary non-polyposis colorectal cancer (Lynch syndromes I & II).Genetics, pathology, natural history, and cancer con-trol, Part I. Cancer Genet Cytogenet 1991: 53: 143–160.

25. Lynch H T, de la Chapelle A. Hereditary colorectalcancer. N Engl J Med 2003: 348: 919–932.

26. Lynch H T, Krush A J. Cancer family ‘‘G’’ revisited:1895–1970. Cancer 1971: 27: 1505–1511.

27. Lynch H T, Lynch P M, Pester J et al. The cancerfamily syndrome. Rare cutaneous phenotypic linkageof Torre’s syndrome. Arch Intern Med 1981: 141:607–611.

28. Lynch H T, Fusaro R M, Roberts L et al. Muir–Torresyndrome in several members of a family with a variantof the cancer family syndrome. Br J Dermatol 1985:113: 295–301.

29. Fishel R, Lescoe M K, Rao M R et al. The humanmutator gene homolog MSH2 and its association withhereditary nonpolyposis colon cancer. Cell 1993: 75:1027–1038.

30. Papadopoulos N, Nicolaides N C, Wei Y F et al.Mutation of a mutL homolog in hereditary colon can-cer. Science 1994: 263: 1625–1629.

31. Umar A, Boland C R, Terdiman J P et al. RevisedBethesda guidelines for hereditary nonpolyposis colo-rectal cancer (Lynch syndrome) and microsatelliteinstability. J Natl Cancer Inst 2004: 96: 261–268.

32. Myrhoj T, Bisgaard M L, Bernstein I et al. Hereditarynon-polyposis colorectal cancer: clinical features andsurvival. Results from the Danish HNPCC register.Scand J Gastroenterol 1997: 32: 572–576.

33. Burgdorf W H, Pitha J, Fahmy A. Muir–Torre syn-drome. Histologic spectrum of sebaceous prolifera-tions. Am J Dermatopathol 1986: 8: 202–208.

34. Kruse R, Rutten A, Lamberti C et al. Muir–Torrephenotype has a frequency of DNA mismatch-repair-gene mutations similar to that in hereditary non-polyposis colorectal cancer families defined by theAmsterdam criteria. Am J Hum Genet 1998: 63: 63–70.

35. Kruse R, Lamberti C, Wang Y et al. Is the mismatchrepair deficient type of Muir–Torre syndrome confinedto mutations in the hMSH2 gene? Hum Genet 1996:98: 747–750.

36. Kruse R, Rutten A, Schweiger N et al. Frequency ofmicrosatellite instability in unselected sebaceous glandneoplasias and hyperplasias. J Invest Dermatol 2003:120: 858–864.

37. Kruse R, Rutten A, Hosseiny-Malayeri H R et al.‘‘Second hit’’ in sebaceous tumors from Muir–Torrepatients with germline mutations in MSH2: allele lossis not the preferred mode of inactivation. J InvestDermatol 2001: 116: 463–465.

38. Rutten A, Burgdorf W, Hugel H et al. Cystic seba-ceous tumors as marker lesions for the Muir–Torresyndrome: a histopathologic and molecular geneticstudy. Am J Dermatopathol 1999: 21: 405–413.

39. Kruse R, Rutten A, Malayeri H R et al. A novelgermline mutation in the hMLH1 DNA mismatchrepair gene in a patient with an isolated cysticsebaceous tumor. J Invest Dermatol 1999: 112: 117–118.

40. Mathiak M, Rutten A, Mangold E et al. Loss ofDNA mismatch repair proteins in skin tumors frompatients with Muir–Torre syndrome and MSH2 orMLH1 germline mutations: establishment of immu-

nohistochemical analysis as a screening test. Am JSurg Pathol 2002: 26: 338–343.

41. Weary P E, Gorlin R J, Gentry W C Jr et al. Multiplehamartoma syndrome (Cowden’s disease). Arch Der-matol 1972: 106: 682–690.

42. Lloyd K M 2nd, Dennis M. Cowden’s disease. Apossible new symptom complex with multiple systeminvolvement. Ann Intern Med 1963: 58: 136–142.

43. Walton B J, Morain W D, Baughman R D et al.Cowden’s disease: a further indication for prophylacticmastectomy. Surgery 1986: 99: 82–86.

44. Fackenthal J D, Marsh D J, Richardson A L et al.Male breast cancer in Cowden syndrome patients withgermline PTEN mutations. J Med Genet 2001: 38:159–164.

45. Starink T M, van der Veen J P, Arwert F et al. TheCowden syndrome: a clinical and genetic study in 21patients. Clin Genet 1986: 29: 222–233.

46. Brownstein M H, Mehregan A H, Bikowski J B.Trichilemmomas in Cowden’s disease. JAMA 1977:238: 26.

47. Brownstein M H, Mehregan A H, Bikowski J B et al.The dermatopathology of Cowden’s syndrome. Br JDermatol 1979: 100: 667–673.

48. Starink T M, Meijer C J, Brownstein M H. The cuta-neous pathology of Cowden’s disease: new findings.J Cutan Pathol 1985: 12: 83–93.

49. Metcalf J S, Maize J C, LeBoit P E. Circumscribedstoriform collagenoma (sclerosing fibroma). Am JDermatopathol 1991: 13: 122–129.

50. Ackerman A B. Trichilemmoma. Arch Dermatol1978: 114: 286.

51. Brownstein M H. Trichilemmoma. Benign folliculartumor or viral wart? Am J Dermatopathol 1980: 2:229–231.

52. Lhermitte J, Duclos P. Sur un ganglioneurome diffusdu cortex du cervelet. Bull Assoc Fr Cancer 1920: 9:99–107.

53. Padberg G W, Schot J D, Vielvoye G J et al. Lher-mitte–Duclos disease and Cowden disease: a singlephakomatosis. Ann Neurol 1991: 29: 517–523.

54. Li J, Yen C, Liaw D et al. PTEN, a putative proteintyrosine phosphatase gene mutated in human brain,breast, and prostate cancer. Science 1997: 275: 1943–1947.

55. Liaw D, Marsh D J, Li J et al. Germline mutations ofthe PTEN gene in Cowden disease, an inherited breastand thyroid cancer syndrome. Nat Genet 1997: 16:64–67.

56. Eng C. Will the real Cowden syndrome please standup: revised diagnostic criteria. J Med Genet 2000: 37:828–830.

57. Arch E M, Goodman B K, Van Wesep R A et al.Deletion of PTEN in a patient with Bannayan–Riley–Ruvalcaba syndrome suggests allelism with Cowdendisease. Am J Med Genet 1997: 71: 489–493.

58. Marsh D J, Dahia P L, Zheng Z et al. Germlinemutations in PTEN are present in Bannayan–Zonanasyndrome. Nat Genet 1997: 16: 333–334.

59. Marsh D J, Kum J B, Lunetta K L et al. PTEN muta-tion spectrum and genotype–phenotype correlations inBannayan–Riley–Ruvalcaba syndrome suggest a singleentity with Cowden syndrome. Hum Mol Genet 1999:8: 1461–1472.

60. Waite K A, Eng C. Protean PTEN: form and func-tion. Am J Hum Genet 2002: 70: 829–844.

Cancer-associated genodermatoses

665

61. Pilarski R, Eng C. Will the real Cowden syndromeplease stand up (again)? Expanding mutational andclinical spectra of the PTEN hamartoma tumour syn-drome. J Med Genet 2004: 41: 323–326.

62. Rohwedder A, Keminer O, Hendricks C et al. Detec-tion of HPV DNA in trichilemmomas by polymerasechain reaction. J Med Virol 1997: 51: 119–125.

63. Schaller J, Rohwedder A, Burgdorf W H et al. Identi-fication of human papillomavirus DNA in cutaneouslesions of Cowden syndrome. Dermatology 2003: 207:134–140.

64. Carney J A, Gordon H, Carpenter P C et al. Thecomplex of myxomas, spotty pigmentation, and endo-crine overactivity. Medicine (Baltimore) 1985: 64: 270–283.

65. Atherton D J, Pitcher D W, Wells R S et al. A syn-drome of various cutaneous pigmented lesions, myx-oid neurofibromata and atrial myxoma: the NAMEsyndrome. Br J Dermatol 1980: 103: 421–429.

66. Rhodes A R, Silverman R A, Harrist T J et al. Muco-cutaneous lentigines, cardiomucocutaneous myxomas,and multiple blue nevi: the ‘‘LAMB’’ syndrome. J AmAcad Dermatol 1984: 10: 72–82.

67. Ferreiro J A, Carney J A. Myxomas of the externalear and their significance. Am J Surg Pathol 1994: 18:274–280.

68. Carney J A, Headington J T, Su W P. Cutaneousmyxomas. A major component of the complex ofmyxomas, spotty pigmentation, and endocrine overac-tivity. Arch Dermatol 1986: 122: 790–798.

69. Carney J A, Ferreiro J A. The epithelioid blue nevus.A multicentric familial tumor with important associa-tions, including cardiac myxoma and psammomatousmelanotic schwannoma. Am J Surg Pathol 1996: 20:259–272.

70. O’Grady T C, Barr R J, Billman G et al. Epithelioidblue nevus occurring in children with no evidence ofCarney complex. Am J Dermatopathol 1999: 21: 483–486.

71. Carney J A, Sheps S G, Go V L et al. The triad ofgastric leiomyosarcoma, functioning extra-adrenalparaganglioma and pulmonary chondroma. N Engl JMed 1977: 296: 1517–1518.

72. Carney J A. The search for Harvey Cushing’s patient,Minnie G., and the cause of her hypercortisolism. AmJ Surg Pathol 1995: 19: 100–108.

73. Young W F Jr, Carney J A, Musa B U et al. FamilialCushing’s syndrome due to primary pigmented nod-ular adrenocortical disease. Reinvestigation 50 yearslater. N Engl J Med 1989: 321: 1659–1664.

74. Carney J A. Psammomatous melanotic schwannoma.A distinctive, heritable tumor with special associa-tions, including cardiac myxoma and the Cushing syn-drome. Am J Surg Pathol 1990: 14: 206–222.

75. Carney J A, Stratakis C A. Epithelioid blue nevus andpsammomatous melanotic schwannoma: the unusualpigmented skin tumors of the Carney complex. SeminDiagn Pathol 1998: 15: 216–224.

76. Carney J A, Toorkey B C. Myxoid fibroadenoma andallied conditions (myxomatosis) of the breast. A herit-able disorder with special associations including

cardiac and cutaneous myxomas. Am J Surg Pathol1991: 15: 713–721.

77. Kirschner L S, Carney J A, Pack S D et al. Mutationsof the gene encoding the protein kinase A type I-alpharegulatory subunit in patients with the Carney com-plex. Nat Genet 2000: 26: 89–92.

78. Matyakhina L, Pack S, Kirschner L S et al. Chromo-some 2 (2p16) abnormalities in Carney complextumours. J Med Genet 2003: 40: 268–277.

79. Burgdorf W, Carney J A. Carney complex. In: LeBoit PE, Burg G, Weedon D et al., eds. Pathology and Genet-ics of Skin Tumors. Lyon: IARC Press, 2006: 291–292.

80. Birt A R, Hogg G R, Dube W J. Hereditary multiplefibrofolliculomas with trichodiscomas and acrochor-dons. Arch Dermatol 1977: 113: 1674–1677.

81. Hornstein O P, Knickenberg M. Perifollicular fibro-matosis cutis with polyps of the colon – a cutaneo-intestinal syndrome sui generis. Arch Dermatol Res1975: 253: 161–175.

82. Pinkus F. Uber einen bisher unbekannten Nebenappa-rat des Menschen: Haarscheiben. Dermatol Z 1902: 9:465–469.

83. Pinkus F. Uber die Haarscheiben des Nabelschweins(Pekari-Dicotyles). Zugleich ein Beitrag zur Kenntnisder Muskulatur der Haare. Arch Dermatol Syphiligr1933: 169: 379–396.

84. Pinkus H, Coskey R, Burgess G H. Trichodiscoma. Abenign tumor related to haarscheibe (hair disk).J Invest Dermatol 1974: 63: 212–218.

85. Burnier M, Rejsek P. Fibromes sous-cutanes peripil-aires multiples du cou. Bull Soc Fr Dermatol Syphiligr1925: 32: 242–243.

86. Zackheim H S, Pinkus H. Perifollicular fibromas.Arch Dermatol 1960: 82: 913–917.

87. Weintraub R, Pinkus H. Multiple fibrofolliculomas(Birt–Hogg–Dube) associated with a large connectivetissue nevus. J Cutan Pathol 1977: 4: 289–299.

88. Rodder O, Schulz-Ehrenburg U, Veh R W, Plewig G.Hair disk (Haarscheibe) and trichodiscoma (Poster).In: 46th Annual Meeting, American Academy of Der-matology. San Antonio, 1987.

89. Steffen C. Mantleoma. A benign neoplasm with man-tle differentiation. Am J Dermatopathol 1993: 15:306–310.

90. Schulz T, Hartschuh W. Birt–Hogg–Dube syndromeand Hornstein–Knickenberg syndrome are the same.Different sectioning technique as the cause of differenthistology. J Cutan Pathol 1999: 26: 55–61.

91. Toro J R, Glenn G, Duray P et al. Birt–Hogg–Dubesyndrome: a novel marker of kidney neoplasia. ArchDermatol 1999: 135: 1195–1202.

92. Nickerson M L, Warren M B, Toro J R et al. Muta-tions in a novel gene lead to kidney tumors, lung walldefects, and benign tumors of the hair follicle inpatients with the Birt–Hogg–Dube syndrome. CancerCell 2002: 2: 157–164.

93. Zbar B, Alvord W G, Glenn G et al. Risk of renaland colonic neoplasms and spontaneous pneumotho-rax in the Birt–Hogg–Dube syndrome. Cancer Epi-demiol Biomarkers Prev 2002: 11: 393–400.

Burgdorf

666