Journal of the DERMATOLOGYAmerican Academy ofphdres.caregate.net/curriculum/pdf-files/Derm_refs/PemphigusVariants.pdfautoantibodies of patients with pemphigus vulgaris and pemphigus

649

From the Division of Immunodermatology, Department ofDermatology, Northwestern University Medical School, Chicagoa;the Department of Dermatology, Kurume University School ofMedicineb; the Department of Dermatology, Keio UniversitySchool of Medicine, Tokyoc; and the Medicine Service, Section ofDermatology, Lakeside Division, Veterans Affairs Chicago HealthCare System.d

Supported by a Clinical Investigator Award (K08-AR01961, NationalInstitutes of Health, Bethesda, Md) (to L. S. C.).

Reprint requests: Lawrence S. Chan, MD, Department ofDermatology, Tarry 4-721, Mail Code T225, 300 E Superior St,Chicago, IL 60611-3010. E-mail address: [email protected].

Copyright © 1999 by the American Academy of Dermatology, Inc.0190-9622/99/$8.00 + 016/2/96936

Pemphigus describes a group of chronic bul-lous diseases, originally named by Wichmann in1791.1 All acquired forms of pemphigus are nowcharacterized as autoimmune blistering diseasespresenting clinically with flaccid intraepidermalblisters and erosions of the skin and mucousmembranes, histologically with acantholysis, andimmunopathologically with in vivo–bound andcirculating autoantibodies against keratinocyte

cell surface components.2 Pemphigus is usuallydivided into two distinct categories depending onblister location: pemphigus vulgaris and pemphi-gus foliaceus, each with its own variants (pem-phigus vegetans and pemphigus erythematosus,respectively). Evidence has revealed that the IgGautoantibodies of patients with pemphigusvulgaris and pemphigus foliaceus recognizedesmosomal components, desmoglein 3 anddesmoglein 1, respectively.3-5 A complete reviewof these diseases, including pemphigus vulgaris,foliaceus, vegetans, and erythematosus, has beenpreviously published.2 During the past 3 decades,rare forms of pemphigus have been described,including pemphigus herpetiformis, IgA pemphi-gus, and paraneoplastic pemphigus. Most of theautoantigens targeted by the autoantibodies frompatients affected by these new pemphigus variantshave been recently identified. In light of this newinformation, this article extensively reviews theclinical presentation, histology, immunopatholo-

CONTINUING MEDICAL EDUCATION

The new pemphigus variants

Neha D. Robinson, MD,a Takashi Hashimoto, MD,b Masayuki Amagai, MD,c andLawrence S. Chan, MDa,d Chicago, Illinois, and Kurume and Tokyo, Japan

Pemphigus describes a group of autoimmune diseases characterized by blisters and ero-sions of the skin and mucous membranes, acantholysis by histology, and autoantibodiesdirected against epidermal cell surface components. Since the early 1970s, the followingnew clinical variants of pemphigus have been reported: pemphigus herpetiformis, IgApemphigus, and paraneoplastic pemphigus. In recent years, significant data have beenobtained from laboratory investigation on these rare and atypical variants, especiallyregarding their specific target antigens. We review these variants, their clinical presenta-tions, histologic findings, immunopathology, target antigens, theories of pathogenesis,treatment modalities, and clinical courses. (J Am Acad Dermatol 1999;40:649-71.)

Learning objective:After completing this article, readers should be familiar with pemphi-gus herpetiformis, IgA pemphigus, and paraneoplastic pemphigus. On the basis of theirunique clinical, histologic, and immunofluorescence findings, readers should also be ableto differentiate these new pemphigus variants from the common forms of pemphigus.

Journal of theAmerican Academy of

DERMATOLOGYVOLUME 40 NUMBER 5 PART 1 MAY 1999

Journal of the American Academy of DermatologyMay 1999

5 cases of pemphigus herpetiformis (6%). In thisstudy, the mean age at onset was 65 years (range,45-83 years of age). In other studies, the age rangeat onset was 31 to 81 years without a male orfemale predilection.8,9,20

Pemphigus herpetiformis combines the clinicalfeatures of dermatitis herpetiformis with theimmunologic and histologic features of pemphigus(Fig 1). In the study of 15 patients by Maciejow-ska, Jablonska, and Chorzelski,8 all patientspresented with erythematous, vesicular, bullous, orpapular lesions, often in a herpetiform pattern.Oral mucous membranes were involved in 4 cases,and severe pruritus accompanied the eruption in 6patients. In cases of other studies, the skin was theonly affected site.19 Most patients described bySanti et al9 experienced pruritus and an eruptioncharacterized by superficial vesicles and inflam-matory plaques. Unfortunately, the clinical presen-tation of pemphigus herpetiformis is atypical, andthe lesions are often initially recognized as variousother bullous diseases, such as dermatitis herpeti-formis, bullous pemphigoid, linear IgA bullous

gy, target antigens, and treatment of these newforms of rare and atypical pemphigus.

EPIDEMIOLOGY AND CLINICALPRESENTATION (Table I)Pemphigus herpetiformis

Pemphigus herpetiformis is recognized as a dis-tinct variant of pemphigus by its pruritus, theinvariable presence of eosinophils and/or neu-trophils, and its tendency to respond to sulfones.The disease was originally named and the diag-nostic criteria were described by Jablonska et al6

in 1975, although patients with similar clinicalphenotypes were described as early as 1955 byFloden and Gentale7 under the heading “dermatitisherpetiformis with acantholysis.” Since then atleast 40 additional cases have been presented inthe literature.8-18 The study by Maciejowska,Jablonska, and Chorzelski8 found 15 cases of pem-phigus herpetiformis out of 205 cases of pemphi-gus (7.3%). In 1998, Micali, Musumeci, andNasca19 performed an epidemiologic analysis of84 cases of pemphigus in eastern Sicily and found

650 Robinson et al

Table I. Clinical presentation, histology, and immunofluorescence findings of the new pemphigusvariants

Disease Clinical presentation Histopathology DIF IIF

Pemphigus Pruritic, erythematous, Eosinophilic spongiosis, IgG deposits in Circulating IgGherpetiformis vesicular, bullous, or w/ or w/o acantholysis; upper or entire autoantibodies

papular lesions, often intraepidermal pustules epidermal cell to epithelialin herpetiform pattern; filled w/ eosinophils surfaces cell surfacesoccasional mucous or neutrophilsmembrane involvement

IgA pemphigus Pruritic, flaccid vesicles IEN type: Suprabasal IEN type: IgA Circulating IgAand/or pustules in pustules, neutrophil deposits in lower autoantibodiesannular pattern with infiltration, scanty or entire epidermal to epithelial cellcentral crusting, acantholysis cell surfaces surfaces (in sometimes hypopyon; SPD type: Subcorneal SPD type: IgA 50% of cases)rare mucous pustules, neutrophil deposits in uppermembrane infiltration, scanty epidermal cellinvolvement acantholysis surfaces

Paraneoplastic Polymorphous skin Suprabasilar intraepithelial IgG deposits in entire Circulating IgG pemphigus eruption, consisting acantholysis, epidermal cell autoantibodies

of blisters, erosions, Keratinocyte necrosis, surfaces +/– to epithelialand targetoid lesions; and vacuolar-interface granular-linear cell surfaces;severe mucous change complement circulating IgGmembrane deposition along autoantibodiesinvolvement basement membrane to rat bladder

zone and epidermal epithelium in cell surfaces 75% of cases

(sensitivity)

+/–, With or without; DIF, direct immunofluorescence; IIF, indirect immunofluorescence; w/, with; w/o,without.

Journal of the American Academy of DermatologyVolume 40, Number 5, Part 1

dermatosis, or pemphigus foliaceus8 (Table II).Nevertheless, severe pruritus seems to be a com-monly presented symptom.9,11

IgA pemphigus

IgA pemphigus is a newly characterized groupof autoimmune intraepidermal blistering diseasespresenting with a vesiculopustular eruption, neu-trophil infiltration, acantholysis, and in vivo–bound and circulating IgA autoantibodies thattarget cell surface components of the epidermis.IgA pemphigus as a disease entity was firstdescribed by Wallach, Foldes, and Cottenot21 in1982 under the name subcorneal pustular der-matosis and monoclonal IgA,although IgA depo-sition in the epidermis in patients with the clinicalphenotype of subcorneal pustular dermatosis wasreported in 1979 by Sneddon and Wilkinson andothers.22,23 Subsequently, multiple cases havebeen reported under several different names,including intraepidermal neutrophilic IgA der-matosis, intercellular IgA dermatosis, IgA pemphi-

gus foliaceus, IgA herpetiform pemphigus,intraepidermal IgA pustulosis,and intercellularIgA vesiculopustular dermatosis.21,24-30

There are two distinct types of IgA pemphigus:the subcorneal pustular dermatosis (SPD) type andthe intraepidermal neutrophilic (IEN) type, alsoreferred to as “IgA pemphigus foliaceus” and “IgApemphigus vulgaris,” respectively.24,31However, aconsensus has not been reached regarding thisnomenclature. Patients with both types of IgApemphigus clinically present with flaccid vesiclesor pustules or both on erythematous or normal skin(Figs 2 and 3). The pustules tend to coalesce toform an annular or circinate pattern with crusts inthe central area. The sites of predilection are theaxillary and groin areas, but the trunk, proximalextremities, and lower aspect of the abdomen are

Robinson et al 651

Fig 1. Clinical presentation of pemphigus herpeti-formis. Annular, erythematous plaques with peripheralvesicles.

Table II. Differential diagnoses of the new pem-phigus variants

Pemphigus variants Differential diagnoses

Pemphigus 1. Dermatitis herpetiformisherpetiformis 2. Pemphigus foliaceus

3. Bullous pemphigoid4. Linear IgA bullous dermatosis5. Pemphigus vulgaris

IgA pemphigus 1. Subcorneal pustular dermatosis2. Dermatitis herpetiformis3. Pemphigus foliaceus4. Pemphigus herpetiformis5. Linear IgA bullous dermatosis

Paraneoplastic 1. Erythema multiforme majorpemphigus 2. Erythema multiforme with

pemphigus vulgaris3. Toxic epidermal necrolysis4. Epidermolysis bullosa acquisita5. Pemphigus vulgaris6. Drug-induced pemphigus7. Bullous pemphigoid8. Cicatricial pemphigoid9. Aphthous stomatitis

10. Behçet’s disease11. Lichen planus with pemphigus

vulgaris12. Persistent herpes simplex virus

infection

Fig 2. Clinical presentation of the SPD type of IgApemphigus. Cluster of vesicles with central crusting onerythematous skin.


unusual paraneoplastic bullous disease.39 The 5diagnostic criteria initially proposed for paraneo-plastic pemphigus include the following38:

1. Painful mucosal ulcerations and blisters and apolymorphous skin eruption, with papular lesionsprogressing to blisters and erosive lesions affectingthe trunk, extremities, palms, and soles, in the con-text of an occult or confirmed underlying neoplasm

2. Histologic findings of vacuolar interfacechange, keratinocyte necrosis, and intraepidermalacantholysis

3. Deposition of IgG and C3 in the epidermalintercellular spaces, and granular-linear comple-ment deposition along the epidermal basementmembrane zone on direct immunofluorescence

4. Serum autoantibodies that bind not only tothe cell surfaces of skin and mucosa in a patterntypical of pemphigus, but also to simple, colum-nar, and transitional epithelia

5. Serum autoantibodies that recognize epider-mal antigens of 250, 230, 210, and 190 kd byimmunochemical techniques such as immunopre-cipitation

A new term,neoplasia-induced pemphigus,aswell as a modification of the original set of diag-nostic criteria was later proposed by Camisa andHelm40 in 1993. This term was proposed after anobservation that the mucocutaneous changes maypersist after the tumor was in complete remis-sion.40 However, for the remainder of this reviewarticle, we will use the term paraneoplastic pem-phigusuntil a general consensus has been reached.The modified criteria proposed by Camisa andHelm are as follows40:

Major criteria:Polymorphous mucocutaneous eruptionConcurrent internal neoplasiaCharacteristic serum immunoprecipitation findings

Minor criteria:Positive cytoplasmic staining of rat bladder epithelium by indirect immunofluorescence

Intercellular and basement membrane zone immunoreactants on direct immunofluores-cence of perilesional tissue

Acantholysis in biopsy specimen from at least one anatomic site of involvement

According to the proposal by Camisa andHelm,40 patients should be considered to have neo-plasia-induced pemphigus if all 3 major or 2 majorand 2 or more minor criteria are met.

commonly involved. Mucous membrane involve-ment is rare.22,24,32Pruritus is also a significantsymptom that may interfere with the patient’sdaily activities. An unusual presentation of a wide-spread vesicular eruption with herpetiform distrib-ution, similar to pemphigus herpetiformis, wasdescribed in a patient with the SPD type of IgApemphigus as well.33

In the literature to date, there are more than 60cases of IgA pemphigus, with a slight predomi-nance of the SPD type. IgA pemphigus usuallyoccurs in middle-aged or elderly persons, althoughoccasional reports of the disease in childhood havebeen described. The disease has an average age atonset of approximately 48 years (range, 5-92years). There is also a slight female predominance(56%).22,30-32,34-36There seems to be no particulargeographic distribution; patients have beendescribed throughout Japan, Europe, and theUnited States. A pemphigus vegetans variant of theIEN type of IgA pemphigus occurring duringimmunosuppressive drug therapy was described ina 7-year-old boy.36 Another report described a 7-year-old girl with lesions involving not only theskin, but also the oral mucosa.37

Paraneoplastic pemphigus

Before the formal characterization by Anhalt etal38 in 1990 as an autoimmune disease with a poly-morphous blistering eruption, mucocutaneousulcerations, and an underlying neoplasm, severalcases of paraneoplastic pemphigus were describedin the literature, often reported as unusual cases ofpemphigus vulgaris, erythema multiforme, or an

652 Robinson et al

Fig 3. Clinical presentation of the IEN type of IgApemphigus. Pustules with central crusting on ery-thematous skin.


The cutaneous lesions of paraneoplastic pemphi-gus are quite variable, consisting of a mixture ofblisters, erosions, and target lesions (Fig 4). In theinitial study by Anhalt et al,38 patients experiencedpruritic blisters that ruptured easily, affecting theupper half of the trunk, head and neck, and proximalextremities. Occasionally, lesions on the extremitiesconsisted of tense blisters or target lesions with cen-tral blister formation, resembling those seen in bul-lous pemphigoid or erythema multiforme.38,41-43

Confluent erythema in the V area of the upper chestand back was also common.38 It has been observedthat some eruptions resembling lichen planus orlichen planus pemphigoides occur as well. Theselichenoid eruptions may be the sole cutaneous man-ifestations of the disease, or they may develop inlesions that had previous blistering.44-47The occur-rence of blisters and lichenoid lesions on the palmsand soles can often be used to differentiate paraneo-plastic pemphigus from pemphigus vulgaris, inwhich lesions of the palms and soles are unusual.Patients with chronic lichenoid skin lesions mayoccasionally also have ulcerative paronychiallesions, which can be painful and debilitating.39

The most observed clinical feature of paraneo-plastic pemphigus is an intractable stomatitis, whichis usually the earliest presenting sign, and isextremely resistant to therapy.39 The stomatitis con-sists of painful erosions and ulcerations of theoropharynx and vermilion borders of the lips.38,39

Most patients also have a severe pseudomembra-nous conjunctivitis.38 Lam et al48 describe a patientwho initially had an erosive conjunctivitis, whichprogressed to scarring and bilateral obliteration ofthe conjunctival fornices. Labial, gingival, buccal,lingual, esophageal, laryngeal, tracheobronchial,nasopharyngeal, oropharyngeal, vaginal, and penilemucosal lesions have also been observed.38,43,48,49

Fullerton et al43 describe a patient with dyspnea outof proportion to findings by arterial blood gas analy-sis and chest radiography. Bronchoscopy revealed adenuded respiratory epithelium, and direct immuno-fluorescence of a biopsy specimen of the bronchialmucosa was consistent with paraneoplastic pemphi-gus. Postmortem examination revealed pemphigus-like lesions of the esophageal and trachealmucosa.43 Involvement of the gastrointestinal andrespiratory epithelia was also suggested by Lam etal48 by bronchoscopy and upper gastrointestinalradiologic studies, but the pathology from either sitewas not documented. Only 1 case has been

described with the total absence of mucosal involve-ment, and this patient had a favorable outcome withcorticosteroid treatment alone.50

As stated in the diagnostic criteria, paraneoplas-tic pemphigus is associated with an underlyingneoplasm, either malignant or benign. In approxi-mately two thirds of cases, the skin disease occursin patients with an existing neoplasm. In theremaining one third of cases, neoplastic lesions aredetected after the mucocutaneous disease occurs.39

Thus examination for an occult neoplasm isimportant in suspected cases of paraneoplasticpemphigus, especially with computed tomograph-ic scanning of the chest, abdomen, and pelvis.39

The various neoplasms associated with para-neoplastic pemphigus in decreasing order of fre-quency are as follows: non-Hodgkin’s lymphoma,chronic lymphocytic leukemia, Castleman’stumor, thymoma (malignant and benign), poorlydifferentiated sarcoma, Waldenström’s macroglob-ulinemia, inflammatory fibrosarcoma, bron-chogenic squamous cell carcinoma, round-cellliposarcoma, Hodgkin’s disease, and T-cell lym-phoma.38,40,43,44,46,47,50-63One patient has beendescribed by Fullerton et al43 in whom paraneo-plastic pemphigus developed while his non-Hodgkin’s lymphoma was in apparent completeremission after autologous bone marrow trans-plantation. Autopsy revealed good engraftment ofthe autologous bone marrow, without gross ormicroscopic evidence of recurrent lymphoma.43

Another case report describes a patient with para-neoplastic pemphigus in the absence of a knownneoplasm. Unfortunately, an autopsy was notperformed.41

Robinson et al 653

Fig 4. Clinical presentation of paraneoplastic pemphi-gus. Erosions and ulcerations of the palate.


Jablonska, and Chorzelski,8 biopsy specimensrevealed various features: eosinophilic spongiosis,with or without isolated acantholytic cells in theupper layers of the epidermis; abortive bullae withslight acantholysis; or intraepidermal pustules filledwith neutrophils and eosinophils. On the contrary,biopsy specimens in the study by Santi et al9 invari-ably revealed eosinophilic spongiosis and acanthol-ysis. Recent reports have indicated a wide variationof inflammatory cell infiltration with approximate-ly 20% eosinophil-dominant, 20% neutrophil-dominant, and 60% mixed neutrophils andeosinophils.10 Because the histologic findings varyaccording to the evolution of the skin lesions andtypical findings of pemphigus emerge only later inthe disease process, several biopsies may be neededto reach the correct diagnosis (Fig 5).

IgA pemphigus

Histopathologic examination of the two IgApemphigus types documents the level of intraepi-dermal pustule or blister formation. In the SPDtype of IgA pemphigus, pustules are located sub-corneally in the upper epidermis (Fig 6), whereasin the IEN type, suprabasilar pustules in the loweror entire epidermis are present (Fig 7). There is anaccompanying superficial acantholysis that is fre-quently sparse, as well as a significant neutrophilinfiltration.22,24,31


The histologic findings of lesions in paraneo-plastic pemphigus show considerable variability,

What is significant in this list of neoplasms isthat Castleman’s tumor is disproportionately asso-ciated with paraneoplastic pemphigus.39 Castle-man’s tumor is also referred to as angiofollicularlymph node hyperplasia, giant lymph node hyper-plasia, and Castleman’s pseudolymphoma.39

Affected nodes appear predominantly in the thoraxand abdominal and pelvic retroperitoneal spaces;extranodal forms of the disease also occur.39

Castleman’s tumors have also been associated withother autoimmune disorders, such as myastheniagravis and autoimmune cytopenias.64-66

Since the original description of paraneoplasticpemphigus by Anhalt et al38 in 1990, more than 70cases of paraneoplastic pemphigus have beendescribed in the literature to date.39-44,46-48,50-63,67

There is a wide geographic distribution, includingthe following ethnicities: Polish, Japanese,Dutch, Iranian, Hispanic, Bosnian, andAmerican.38,46,53,54,58Age at onset ranges from 7to 77 years (mean, 51 years).* Paraneoplasticpemphigus is rare in children, although two casereports have described paraneoplastic pemphigusin children 7 and 13 years of age.53,58There doesnot seem to be a gender predominance.40

HISTOPATHOLOGYPemphigus herpetiformis

The histologic findings in pemphigus herpeti-formis vary greatly and often are noncontributory.In the studies by Jablonska et al6 and Maciejowska,

654 Robinson et al

Fig 5. Histologic findings in pemphigus herpetiformis.Early acantholysis with a mixed inflammatory cellinfiltrate in the epidermis. (Hematoxylin-eosin stain;original magnification ×100.)

Fig 6. Histologic findings in the SPD type of IgA pem-phigus. Subcorneal blister with neutrophil infiltration.(Hematoxylin-eosin stain; original magnification×100.)

*References 38-44, 46-48, 50-53, 55, 56, 58, 59, 61-63, 67.


reflecting the polymorphism of the clinicallesions.38,39,67The stomatitis is severe, and manybiopsy specimens from ulcerative lesions will onlydemonstrate nonspecific inflammation.39Anhalt etal38 describe suprabasilar intraepithelial acanthol-ysis, necrosis of individual keratinocytes, andvacuolar-interface change as the predominant his-tologic findings in their series of patients (Fig 8).According to Horn and Anhalt,67 the major histo-logic features include epidermal acantholysis,suprabasilar cleft formation, dyskeratotic kera-tinocytes, vacuolar change of the basilar epider-mis, and epidermal exocytosis of inflammatorycells. These diagnostic features are expressed inclinical lesions demonstrating vesiculation. Theperivascular inflammatory cell infiltrate was main-ly composed of mononuclear cells, probably lym-phocytes.67 A unique combination of suprabasalacantholysis and dyskeratotic keratinocytes wasvisualized throughout the epidermis in 44% ofspecimens. A “tombstone appearance” of the basalkeratinocytes lining the floor of the blister ofsuprabasal clefts was also noted in some cases.Unlike many eruptions with interface changes, nodyskeratotic cells were noted in the upper dermis.There was no evidence of cellular atypia or vas-culitis present in any of the 16 specimens.67

The suprabasilar acantholysis and blister forma-tion of paraneoplastic pemphigus are similarly seenin pemphigus vulgaris, but the characteristic pem-phigus vulgaris findings of neutrophils andeosinophils in the papillary dermis and eosinophilic

epidermal spongiosis are not seen in paraneoplasticpemphigus.67 The suprabasal cleft does notprogress with overt formation of bullae, distin-guishing paraneoplastic pemphigus histologicallyfrom herpes gestationis and incontinentia pigmenti,which are characterized by both bullae and dysker-atotic keratinocytes.67 The lichenoid skin lesionsdemonstrate areas of dense lymphocytic infiltrate inthe papillary dermis and individual lymphocytesinfiltrating in the epithelium with occasional indi-vidual cell necrosis.39 Specimens that display thefeatures of interface dermatitis (dyskeratotickeratinocytes, basal vacuolization, and epidermalexocytosis) cannot always be distinguished histo-logically from other disorders with interfacedermatitis, including erythema multiforme, toxicepidermal necrolysis, fixed drug eruption, lupuserythematosus, graft-versus-host disease, and sec-ondary syphilis. The patterns of inflammation,absence of acantholysis, and immunofluorescencefindings, however, distinguish these diseases fromparaneoplastic pemphigus.67

IMMUNOPATHOLOGYPemphigus herpetiformis

Because the clinical presentation and histologicfindings are often atypical, the more reliable basisfor diagnosis of pemphigus herpetiformis isimmunopathology. Direct immunofluorescenceperformed on perilesional skin biopsy specimenshas invariably revealed in vivo–bound IgG deposi-tion at the keratinocyte cell surface primarily in the

Robinson et al 655

Fig 7. Histologic findings in the IEN type of IgA pem-phigus. Intraepidermal blister with acantholysis and apredominantly neutrophil infiltration. (Hematoxylin-eosin stain; original magnification ×50.)

Fig 8. Histologic findings of paraneoplastic pemphi-gus. Suprabasilar acantholysis and cleft forma-tion, dyskeratotic keratinocytes, and an infiltrationof mononuclear cells in a lichenoid pattern. (Hema-toxylin-eosin stain; original magnification ×75.)


sional skin specimens, usually in a pattern similarto pemphigus IgG deposition.22,24-27,32,34In theSPD type of IgA pemphigus, IgA deposition islimited to the upper epidermal cell surfaces (Fig10), whereas in the IEN type of IgA pemphigus,there is intercellular IgA deposition restricted tothe lower epidermis or throughout the entire epi-dermis (Fig 11).24,30,32 The subclass of invivo–bound and circulating IgA autoantibodieshas also been determined and is exclusivelyIgA1.22,26,31,37

Despite the consistent in vivo deposition of IgAin patients’ skin, only approximately 50% ofpatients have detectable circulating autoantibodiesby indirect immunofluorescence.22,24,28 In thestudy by Hashimoto, Ebihara, and Nishikawa32 in1996, nearly all patients’ sera (16 of 17) demon-strated circulating autoantibodies using indirectimmunofluorescence on normal human skin.Low serum titers were also noted in the majorityof these cases, usually between 1:10 and1:320.22,24,31,32

The study by Supapannachart and Mutasim69 in1993 also demonstrated that a patient’s serumspecimen that did not exhibit circulating IgAautoantibodies by standard indirect immunofluo-rescence techniques on standard substrates, suchas monkey esophagus or normal human skin, didcontain antibodies on prolonged incubation underexplant culture conditions. This study proposedthat skin explant culture may be more sensitive indetecting circulating autoantibodies, and this tech-

upper epidermis (Fig 9).6,8,9,11Indirect immunoflu-orescence has also demonstrated circulating IgGanti-epithelial cell surface autoantibodies in mostpemphigus herpetiformis patients described.6,8,9,11

The IgG4 subclass was found to be the predominantIgG subclass of in vivo–bound and circulatingautoantibodies in the studies by Santi et al,9 Kuboet al,20 and O’Toole et al.68

IgA pemphigus

IgA deposition in the intercellular substance ofthe epidermis is present in all cases of IgA pem-phigus by direct immunofluorescence of perile-

656 Robinson et al

Fig 9. Immunofluorescence findings of pemphigus her-petiformis. Direct immunofluorescence demonstratesIgG deposition at the epidermal cell surfaces, sparingthe basal layers. (Original magnification ×100.)

Fig 11. Immunofluorescence findings of the IEN typeof IgA pemphigus. Indirect immunofluorescencedemonstrates circulating IgA autoantibodies binding tothe cell surfaces of the entire epidermis of normalhuman skin. (Original magnification ×100.)

Fig 10. Immunofluorescence findings of the SPD typeof IgA pemphigus. Indirect immunofluorescencedemonstrates circulating IgA autoantibodies binding tothe upper epidermal cell surfaces of normal humanskin. (Original magnification ×100.)


nique may be utilized in detecting IgA pemphigusantibodies in serum specimens that are negative bystandard indirect immunofluorescence.


Direct immunofluorescence of lesional or peri-lesional skin or mucosa specimens has invariablyrevealed IgG deposition in the epidermal cell sur-faces, in a pattern typical to that seen in pemphigusvulgaris, but often in a focal or faint depositionpattern (Fig 12).38,39,67A distinctive finding wasthe combined presence of a granular-linear com-plement deposition pattern along the basementmembrane zone as well as the epidermal cell sur-faces.38 False-negative direct immunofluorescenceresults are more common in paraneoplastic pem-phigus than in pemphigus vulgaris; thereforerepeated biopsies may be necessary to establishthe diagnosis.39

The diagnosis of paraneoplastic pemphigus canalso be established by the presence of circulatingIgG autoantibodies by indirect immunofluores-cence examination of monkey esophagus, humanskin, and mouse skin. Titers usually range from1:320 to 1:5120. The autoantibodies have beenfound to be polyclonal, with IgG subclasses in thefollowing descending order: IgG1, IgG2, IgG4,IgG3. Lambda light chains predominate over κlight chains.38,60

The epidermal cell surface binding pattern inparaneoplastic pemphigus sera is identical to thatseen in pemphigus vulgaris or foliaceus.38,49,57,67

To distinguish the circulating antibodies fromthose seen in pemphigus vulgaris, nonstratifiedand stratified squamous epithelia are used as sub-strates for indirect immunofluorescence. Whenindirect immunofluorescence is performed onmonkey esophagus (stratified epithelium) androdent urinary bladder (nonstratified epithelium),autoantibodies from sera of patients with paraneo-plastic pemphigus will react with both substrates;sera from patients with pemphigus vulgaris or foli-aceus will only react with monkey esophagus.39

Urinary bladder was chosen as the preferred tissuebecause it is a complex transitional epitheliumwith a high density of desmosomes.57 In the initialstudy by Anhalt et al,38 the autoantibodies boundconsistently and strongly to the epithelium of uri-nary bladder, respiratory epithelium, small bowelepithelium, and colon epithelium, and with vari-able intensity to intercalated disks of myocardium,

skeletal muscle, and thyroid epithelium. The mostintense binding was observed in the urinary blad-der epithelium.

Initially, indirect immunofluorescence onrat bladder epithelium seemed highly reliable,but reports involved only a limited number ofpatients.70 Subsequently, reports by Helou,Allbritton, and Anhalt,57 in a study of 28 patientswith paraneoplastic pemphigus, demonstrated aspecificity of 83% and a sensitivity of 75% of indi-rect immunofluorescence studies on rat bladderepithelium. With amplification of the immunoflu-orescence signal by indirect complement fixation,the sensitivity of the technique was increased to89%.57 Studies were also performed on nonepithe-lial tissues in which desmosomes are present, suchas myocardium and liver. Indirect immunofluores-cence on liver was specific (96.5%) but insensitive(43%). False-positive results have also beenreported.71 Therefore indirect immunofluores-cence on rat bladder epithelium is an adequatescreening test for paraneoplastic pemphigus, but isfalse negative in as many as 25% of patients.Immunoprecipitation must therefore be performedfor confirmation of the diagnosis of paraneoplasticpemphigus by identification of the characteristicantigen complex.57

TARGET ANTIGENS (Table III, Fig 13)Pemphigus herpetiformis

The recognition of target antigens has alwaysbeen a focus of extensive study in pemphigus

Robinson et al 657

Fig 12. Immunofluorescence findings of paraneoplas-tic pemphigus. Indirect immunofluorescence demon-strates characteristic cytoplasmic and cell surfacestaining of rat bladder epithelium by circulating IgGautoantibodies. (Original magnification ×100.)


adsorption studies further confirm the specificityof the IgG autoantibodies.68

In 1997, Kubo et al20 described a patient withpemphigus herpetiformis with reactivity only withdesmoglein 3 by immunoblot analysis. Direct andindirect immunofluorescence also demonstratedkeratinocyte cell surface staining in the lower lay-ers of the epidermis, where desmoglein 3 is locat-

research. Accumulated evidence from immuno-blotting and immunoprecipitation has revealedthat the IgG autoantibodies of most patients withpemphigus herpetiformis target the desmosomalcomponent, desmoglein 1 exclusively.9,68Becausedesmoglein 1 is predominantly located in theupper epidermis, this may explain the preferentialbinding of IgG to the upper epidermis. Immuno-

658 Robinson et al

Fig 13. Schematic diagram of the target autoantigens in the new pemphigus variants.72-76

(Although desmoglein 1 is depicted here in the desmosome of a basal keratinocyte, it ispredominantly located in the upper epidermis.)

Table III. Delineated autoantigens for the new pemphigus variants

Molecular ReferenceDisease Autoantigen Location weight (kd) No. (s)

Pemphigus herpetiformis Desmoglein 1 (majority) Upper epidermis 160 9, 68, 77Desmoglein 3 Lower epidermis 130 20, 77

IgA pemphigus Desmoglein 3 Lower epidermis 130 31IEN type (one case only)

IgA pemphigus Desmocollin 1 Upper epidermis 105-115 30, 32, 35SPD type

Paraneoplastic pemphigus Desmoplakin I Desmosome 250 38Desmoplakin II Desmosome ~210 60Bullous pemphigoid Hemidesmosome 230 38antigen 1 (BP230)

Envoplakin Desmosome 210 81, 83Periplakin Desmosome 190 55, 82, 83Undetermined 170 85Desmoglein 3 Lower epidermis 130 84Desmoglein 1 Upper epidermis 160 84


ed.20 Thus both desmoglein 1 and desmoglein 3are recognized by pemphigus herpetiformisautoantibodies.77

IgA pemphigus

There are 2 types of desmosomal cadherins:desmoglein and desmocollin, both of which occuras 3 isoforms, desmoglein 1, 2, and 3 and desmo-collin 1, 2, and 3, encoded by different genes.78,79

In 1991, Ebihara et al30 demonstrated by immuno-blotting that serum from a patient with the IENtype of IgA pemphigus reacted exclusively with a120-kd protein in normal human skin extract and adesmosome-enriched bovine snout epidermis sam-ple. In this study, sera from 3 patients with theSPD type of IgA pemphigus also reacted with adoublet of 115-kd and 105-kd proteins, whichappeared to be identical to desmocollins 1 and 2 inthe bovine desmosome sample. Years later,Hashimoto, Ebihara, and Nishikawa32 demonstrat-ed by immunoblotting that IgA antibodies in 10 of17 IgA pemphigus sera (7 SPD type and 3 IENtype) reacted variously with either bovinedesmoglein 1 or desmocollin, not with humandesmocollin. The significance of the results of thisstudy was unclear at that time, since no IgA pem-phigus sera had been shown to react with humandesmocollin by immunoblotting. The reason forthis may be that IgA pemphigus sera react withconformation-dependent epitopes on desmocollin,because most pathogenic autoantibodies recognizeconformation-sensitive epitopes of desmogleins inpemphigus foliaceus and pemphigus vulgaris.3,80

Immunoprecipitation is a useful method to detectconformation-dependent epitopes by IgG antibod-ies, but a robust immunoprecipitation method withIgA antibodies has not yet been established.

More recently, Hashimoto et al35 have revealedthat human desmocollin 1 is an autoantigen for theSPD type of IgA pemphigus. In this study, mam-malian expression vectors containing the entire cod-ing sequence of human desmocollin 1, 2, and 3 wereconstructed and transfected into COS7 cells.Immunofluorescence of the COS7 cells demonstrat-ed that IgA autoantibodies of 6 cases of the SPDtype of IgA pemphigus reacted with the surface ofcells expressing desmocollin 1 (Fig 14), but not withcells expressing desmocollin 2 or desmocollin 3.None of the 7 cases of the IEN type of IgA pemphi-gus reacted with cells transfected with any desmo-collins. This study also demonstrated that immuno-

fluorescence using cells transfected with cDNAencoding antigenic proteins is a useful method fordefining antigens whose epitopes are conformation-dependent and therefore cannot be detected withstandard immunoblotting techniques.35

The target antigen for the IEN type of IgA pem-phigus has also recently been identified in 1patient. Wang et al31 demonstrated in immunoblot-ting studies that the IgA autoantibodies of a patientwith the IEN type of IgA pemphigus reacted torecombinant desmoglein 3, which is also recog-nized by IgG autoantibodies from classic pemphi-gus vulgaris.4 In the immunoblotting studies, apartially purified baculovirus-expressed recombi-nant desmoglein 3 protein was used. The specifici-ty of the IgA autoantibodies against desmoglein 3was further confirmed by immunoadsorption.31

Others, however, have found no desmoglein 3 ordesmoglein 1 reactivities by enzyme-linkedimmunosorbent assay (ELISA) in 5 patients withthe IEN type of IgA pemphigus (Amagai andHashimoto, unpublished observation). Furtherstudies in large numbers of patients are needed toconfirm the target antigen of the IEN type of IgApemphigus.


The paraneoplastic pemphigus autoantigensidentified to date include cytoplasmic proteins ofthe plakin gene family (desmoplakin I and II, bul-lous pemphigoid antigen 1 [BP230], envoplakin

Robinson et al 659

Fig 14. Immunofluorescence findings of the SPD typeof IgA pemphigus. Indirect immunofluorescencedemonstrates that sera from the SPD type of IgA pem-phigus clearly stains the cell surface of COS7 cellstransfected with desmocollin 1 in a granular pattern.(Original magnification ×300.)


plakin.55,74,82,83The 170-kd antigen is a trans-membrane protein, which may be a critical antigenin the pathogenesis of tissue injury in paraneoplas-tic pemphigus. The 170-kd antigen is now recog-nized by autoantibodies in almost all cases ofparaneoplastic pemphigus, but its true identity hasnot yet been determined.39,85 Reports have alsosuggested that a 130-kd protein recognized by para-neoplastic pemphigus sera may be desmoglein 3(the target antigen of pemphigus vulgaris).52,54,88,89

A recent study by Amagai et al84 revealed thatdesmoglein 3 is a key cell surface target antigen inparaneoplastic pemphigus. ELISA using bac-ulovirus-expressed recombinant desmogleins 1and 3 revealed that 100% of paraneoplasticpemphigus sera tested were positive againstdesmoglein 3, whereas 64% of sera were also pos-itive against desmoglein 1. All of 12 sera testedimmunoprecipitated desmoglein 3.84 Becauseautoantibodies against desmogleins have beenproven to be pathogenic in passive transfer exper-iments,38,39 and desmogleins are extracellularlylocated and therefore more accessible to immunetargeting, they may be more important as the path-ogenic target antigens.

PATHOGENESISPemphigus herpetiformis

Anti-desmoglein antibodies in pemphigus her-petiformis induce spongiosis with eosinophil infil-tration, but rarely produce acantholysis, in contrastto classical pemphigus.20 A theory has beenproposed to explain why the anti-desmoglein anti-bodies give rise to different clinical and histologicpresentations. It is thought that the autoantibodiesin classical pemphigus and in pemphigus herpeti-formis recognize different epitopes on the antigenmolecules.20 Autoantibodies in classical pemphi-gus may inhibit the adhesive function ofdesmoglein or activate a signaling pathway toinduce proteinases, leading to acantholysis.90,91

Autoantibodies in pemphigus herpetiformis maybind different epitopes and may not inhibit theadhesive function or induce the signaling pathwayof proteinases.20

In a study by O’Toole et al,68 biopsy specimensfrom two patients with pemphigus herpetiformisdemonstrate a predominantly neutrophilic infiltra-tion with a co-localization of in vivo–bound IgGand interleukin-8 (IL-8), a potent neutrophilchemoattractant, at the upper epidermis, where

and periplakin), desmosomal antigens (desmoglein3 and desmoglein 1), and an undetermined 170-kdtransmembranous antigen.38,60,74,75,81-85Plakinsare molecules that are thought to link the keratinintermediate filaments to the cell surface, specifi-cally to desmosomes and hemidesmosomes.76,86Inthe desmosomes, the known plakins are desmo-plakin, envoplakin, and periplakin,75,76 whereasbullous pemphigoid antigen 1 is the major plaqueprotein of the epidermal hemidesmosome.86

In the initial report by Anhalt et al,38 the serumof all 5 patients immunoprecipitated an identicalcomplex of 4 polypeptides from human kera-tinocyte extracts, with estimated molecularweights of 250, 230, 210, and 190 kd. The210-kd and 190-kd bands were most consistentlyvisualized. A diffuse band at 170 kd was also occa-sionally visualized, which was suspected to be adegradation product of one of the higher-molecu-lar-weight antigens.38 The identities of the 250-kdand 230-kd antigens were subsequently deter-mined to be desmoplakin I and bullous pem-phigoid antigen 1 (BP230), respectively.38

Later, studies identified the 210-kd antigen asdesmoplakin II, another major cytoplasmic plaqueprotein of desmosomes.60 In the study byHashimoto et al,54 immunoblotting of epidermalextracts and bovine desmosome preparationsrevealed a difference of mobility on the gelbetween the 210-kd protein detected by paraneo-plastic pemphigus sera and desmoplakin II, detect-ed by the anti-desmoplakin monoclonal antibody.These results suggested that the 210-kd antigenwas not desmoplakin II, as previously thought.54

More confirmatory evidence of the differencebetween desmoplakin II and the 210-kd proteinstems from the observation that the 210-kd antigenand desmoplakin I are not always detected simul-taneously by the same sera.54,58 Desmoplakins Iand II are known to be produced from the samegene by alternative splicing, and desmoplakin II isa part of desmoplakin I.87 Therefore, if the 210-kdantigen is desmoplakin II, desmoplakin I and the210-kd antigen should always be detected simulta-neously. The 210-kd antigen, the faster migratingband, was later determined by cDNA cloning andbiochemical studies to be envoplakin, which ishomologous to desmoplakin and a novel precursorprotein of the cornified envelope.75,81,83 Morerecently, the 190-kd antigen was identified asperiplakin, which is also homologous to desmo-

660 Robinson et al


neutrophil infiltration and acantholysis are appar-ent. This study also demonstrated the in vitro acti-vation of IL-8 cytoplasmic expression and secre-tion in cultured keratinocytes by the purified IgGof the patients with pemphigus herpetiformis. Thekeratinocytes treated with this IgG secreted a 5- to7-fold increase of IL-8 (determined by ELISA)compared with controls. These results suggestedthat the IgG autoantibodies in pemphigus herpeti-formis are responsible for recruiting neutrophils tothe upper epidermis by inducing keratinocyte IL-8secretion (Fig 15).68

IgA pemphigus

In 1993, Supapannachart and Mutasim69 wereable to induce intraepidermal acantholysis withtwo IgA pemphigus (SPD type) serum specimensusing skin explant cultures. Immunofluorescencestaining showed that the antibody was fixed to thecell surface antigens (3 hours) before the onset ofacantholysis (24 hours), suggesting a possible rolefor the circulating antibody in the induction ofacantholysis.69

Because desmoglein 3 is a common target anti-gen for pemphigus vulgaris and a subset of theIEN type of IgA pemphigus, it is unclear why dif-ferent isotypes of autoantibodies develop towardthe same antigen. The generation of antibodies is acomplex immunologic process involving T and Blymphocytes. T cells direct the production of anti-bodies indirectly through effects on the B cells’transformation into plasma cells.90 Studies have

shown that helper T-cell subset 2 (Th2) is respon-sible for directing IgA production in mice.92 TheseTh2 cells in turn secrete IL-5, which then stimu-lates B cells to produce IgA.93 A similar processmay be involved with IgA pemphigus (Fig 16).

There are also theories for the prominent neu-trophil infiltration present in IgA pemphigus.Recently, the specific binding site for the mono-cyte/granulocyte IgA-Fc receptor (CD89) has beenlocalized to the constant domain of human IgA1distal to the hinge region.94 This arrangement ofthe binding site may possibly provide resistance toprotease degradation, allowing for efficient bind-ing of neutrophils.95

It has also been demonstrated that there is anabsence of complement deposition in most casesof IgA pemphigus.22,26,28,69 In the study byHashimoto et al26 in 1987, by means of comple-ment-enhancing indirect immunofluorescencetesting using anti-C1q, C4, C3, and C5 antiserum,no evidence of complement activation was seen inthese IgA antibodies. Junctional deposits of C3were noted in only 4 of the 29 cases reviewed byWallach.22 This may imply that the acantholysis inIgA pemphigus may be complement-independent.


The actual cause of paraneoplastic pemphigushas always been a matter of speculation.39 Onetheory describes an antitumor immune responsethat cross-reacts with normal epithelial proteins.39

This theory is based on similar tumor-associated

Robinson et al 661

Fig 15. Proposed pathogenesis of neutrophil-dominant pemphigus herpetiformis.


have been known to be anomalously producedby tumors that do not normally possess desmo-somes. These tumors include sarcomas, lym-phomas, neuroblastomas, and meningiomas.100-102

Desmoplakins have also been detected byimmunostaining in reactive lymph nodes, tonsils,non-Hodgkin’s B-cell lymphomas, and synovialsarcomas.103 These findings can support thehypothesis that these tumors may anomalouslyproduce these antigens and that autoantibodiesdirected against these tumor antigens becomecross-reactive with epithelial antigens, causing themucocutaneous disease.60

Other hypotheses also exist: treatment withcytokines may have induced paraneoplastic pem-phigus in a patient with underlying lymphoma orWaldenström’s macroglobulinemia.104 The induc-tion of autoimmunity may be due to a dysregulat-ed cytokine production by tumor cells.39 In certaincases of non-Hodgkin’s lymphoma, chronic lym-phocytic leukemia, and Castleman’s disease, thetumors secrete massive amounts of IL-6 invitro.105-107 Because IL-6 is known to promote

autoimmune phenomena occurring in other organsystems. For example, in certain ovarian carcino-mas, the neoplasm may produce a neuroectoder-mal protein. These patients experience an autoan-tibody response against the neuroectodermal pro-tein normally present in Purkinje cells in thecerebellum, resulting in ataxia.96,97Cancer-associ-ated retinopathy is another similar phenomenon.39

Certain patients with small cell carcinoma of thelung anomalously express proteins normally foundexclusively in the choroid and retina of the eye.These patients developed autoantibodies againstthis tumor-associated protein and becameblind.98,99 Applying this theory to paraneoplasticpemphigus, it is known that desmoplakins areexpressed in thymomas and Castleman’s tumors,and it is possible that these patients’ autoantibod-ies were initiated against desmoplakins in thesetumors. However, the majority of affected patientshave lymphomas or chronic leukemias of B-cellorigin, which do not naturally produce desmo-somes or express desmoplakins.60

Desmosomes and “desmosome-like junctions”

662 Robinson et al

Fig 16. Proposed pathogenesis of IgA pemphigus.


B-cell differentiation and drive immunoglobulinproduction, and dysregulated IL-6 production hasbeen demonstrated in certain autoimmune dis-eases, a link between tumor-associated cytokineover-expression and autoimmunity could exist.39

The pathophysiologic relevance of the paraneo-plastic pemphigus autoantibodies in the inductionof acantholysis and consequent blister formation isnow clear. It has been demonstrated that whenautoantibodies from paraneoplastic pemphigussera are injected into neonatal mice, acantholysisand cutaneous blisters are induced, demonstratingthat these autoantibodies are pathogenic in para-neoplastic pemphigus, and not merely an epiphe-nomenon.38,39The acantholytic changes are indis-tinguishable from those observed with passivetransfer of pemphigus vulgaris autoantibodies intoanimals.39 Ultrastructural examination of the skinof the mice revealed a loss of cell-cell adhesion,but normal basal cell–dermal cell attachment. Theepidermal cells also showed mitochondrialswelling and cytoplasmic vacuolization.38

Because most of the autoantibodies recognizespecific cytoplasmic plaque proteins by immuno-precipitation, questions have been raised as to howthese anti-cytoplasmic protein autoantibodies canbe directly involved in the induction of acantholy-

sis, since the cytoplasmic location of the autoanti-gens does not allow circulating antibodies toaccess them.84 Thus the pathophysiologic rele-vance of these anti-plakin autoantibodies is notclear. Amagai et al84 have recently demonstratedthat affinity-purified anti-desmoglein 3 antibodiesare pathogenic and cause acantholysis and blisterformation in neonatal mice by passive transfer.Removal of the anti-desmoglein 3 antibodies byimmunoadsorption eliminated this pathogenicability of paraneoplastic pemphigus sera. There-fore it is more likely that anti-desmoglein 3autoantibodies present in these patients initiate theacantholytic process and cause damage to the cellmembranes. Once the membrane is damaged,autoantibodies against the plakin family areinduced, which then enter the cell and bind the tar-get autoantigens by an epitope spreading phenom-enon (Fig 17).84,108The possible pathogenic roleof autoantibodies against the undetermined 170-kdtransmembrane cell surface antigen has been sug-gested but not been clearly defined.39,85

TREATMENT

Because pemphigus herpetiformis, IgA pem-phigus, and paraneoplastic pemphigus are rare andnewly described variants of pemphigus, compre-

Robinson et al 663

Fig 17. Proposed pathogenesis of paraneoplastic pemphigus.


of pemphigus. Plasmapheresis should be used inconjunction with other agents to suppress theotherwise inevitable rebound increase in anti-body synthesis afterward.109 In the study byMaciejowska, Jablonska, and Chorzelski,8 onepatient was treated with plasmapheresis in addi-tion to prednisone and cyclophosphamide, result-ing in rapid remission. A relapse after 9 monthswas treated with prednisone and cyclophos-phamide, resulting in complete remission.

IgA pemphigus

Dapsone has proved to be the drug of choice inthe treatment of IgA pemphigus.22-25,36,37Withtherapeutic doses, dapsone interferes with themyeloperoxidase-hydrogen peroxide, halide-medi-ated cytotoxic system of neutrophils. This actionresults in reduced cytotoxicity to adjacent cells.110

Not surprisingly, dapsone is effective in disorderswith prominent neutrophil infiltration. Dapsonewas effective in 16 of 23 patients, usually at a doseof 100 mg daily.22 If dapsone is not well tolerated,sulfapyridine in combination with prednisone orphotochemotherapy with PUVA with etretinate areuseful alternatives.27,111

In some cases the effects of dapsone are eitherinsufficient or only transient26,112; therefore etreti-nate is an excellent second-line treatment. Theretinoid derivative etretinate was effective in4 patients with IgA pemphigus, according toWallach.22 Etretinate is thought to act as an anti-inflammatory agent by its inhibition of the migra-tion of monocytes and neutrophils in vivo.113

Some patients have responded well to a combina-tion of dapsone 50 to 100 mg daily and etretinate20 to 30 mg daily.29,112Other patients may requireetretinate to treat subsequent flares.21,27 Becauseetretinate is no longer available in the UnitedStates, we assume that acitretin may be as effectivefor these indications.

In the treatment of IgA pemphigus, good resultshave been achieved with moderate doses of sys-temic corticosteroids alone in 5 patients and incombination with immunosuppressive agents in 2patients.22 The suggested corticosteroid dose is 0.5to 1 mg/kg daily. Topical corticosteroid therapymay also be adequate in some cases of IgA pem-phigus. It is also an extremely useful adjuvant tosystemic therapy when disease is more severe, per-mitting a reduction in requirements for cortico-steroid therapy.109 The patient described by Wang

hensive studies of large numbers of patients thusfar have not taken place. The following summaryof treatment options is taken from personal experi-ences of physician-investigators on small numbersof cases.

Pemphigus herpetiformis

A common problem is the choice of propertreatment. Good results have been demonstratedby combined treatment with dapsone and lowdoses of systemic corticosteroids, but in severalcases, a regimen of corticosteroids in combinationwith immunosuppressants, such as azathioprineand cyclophosphamide, is necessary.109

Dapsone is the drug of first choice for the treat-ment of pemphigus herpetiformis with doses rang-ing from 100 to 300 mg daily.8,11Dapsone may begiven as monotherapy or in combination with sys-temic corticosteroids and immunosuppressants.109

Cases of pemphigus with low or negative antibodytiters or those demonstrating eosinophilic spongio-sis appear more likely to respond to dapsone ther-apy as well.109Sulfapyridine 2 g daily may also beeffective.11

Systemic corticosteroids, usually in combina-tion with dapsone, sulfapyridine, or immunosup-pressive agents, were required to obtain remissionin 14 of 15 patients with pemphigus herpetiformisin the study by Maciejowska, Jablonska, andChorzelski.8 In the study by Ingber andFeuerman,11 remission was achieved with low-dose prednisone or dapsone, or both. Prednisonedoses used by various physicians ranged from 10to 120 mg daily for initial control.8,11

Immunosuppressive drugs, such as azathioprineand cyclophosphamide, can be added to reducecorticosteroid doses and resulting complications.Azathioprine is the most commonly used immuno-suppressive agent, usually as an adjunct to corti-costeroids, but occasionally as monotherapy.109

Azathioprine 100 to 200 mg has been used in com-bination with prednisone to treat pemphigus her-petiformis, remission obtained in one of twocases.8 Cyclophosphamide 50 to 100 mg daily, incombination with prednisone, was effective inobtaining remission in 3 patients after failure ofother therapy.8

Plasmapheresis has been used in an attempt toreduce the titers of pathogenic autoantibodies inpatients with pemphigus. It often results in clinicalimprovement in otherwise therapy-resistant cases

664 Robinson et al


et al31 was treated with mid-potency topicalsteroids for several weeks, with resolution oflesions and no recurrence of disease 3 months afterdiscontinuation of therapy.

Azathioprine is a commonly used immunosup-pressant in pemphigus, but does not seem to beeffective in treating IgA pemphigus. In combina-tion with fluocortolone 80 mg daily, azathioprine150 mg daily was used in the treatment of onepatient without benefit.29 PUVA, alone or in com-bination with etretinate, was effective in 3 patientswith IgA pemphigus.22,27 Colchicine was alsoeffective in one of two patients with IgA pemphi-gus.22 Aggressive therapy with plasmapheresis,prednisone, and cyclophosphamide has also beenused for a recurrence after initial treatment withdapsone and prednisone.33


Patients with paraneoplastic pemphigus andbenign tumors, such as thymomas or Castleman’stumors, should have the tumor surgically excised.The majority of these patients will either improvesubstantially or clear completely.38,39,52,53,59,67

Complete resolution of lesions usually occurs 6 to18 months after excision of a benign neo-plasm.39,59In the one patient presented by Jansen,Plewig, and Anhalt,59 progressive mucocutaneouslesions proved refractory to high-dose cortico-steroids, aromatic retinoids, and immunosuppres-sion. After surgical removal of a Castleman’stumor, there was complete resolution of all symp-toms with no recurrences for more than 3 years.

Regarding malignant neoplasms, there is noconsensus on a standard effective therapeutic regi-men.39 Cutaneous lesions respond more quickly totherapy, whereas stomatitis is generally refractoryto most forms of therapy.39 Induction ofchemotherapy has occasionally resulted incomplete resolution of the malignancy (a diffuselarge-cell lymphoma and a T-cell lymphoblasticlymphoma) and a slow resolution of the skinlesions.56,58 In general, corticosteroids produceonly partial resolution of lesions.38,39,114Dega etal50 presented a patient with Hodgkin’s lymphomain remission and an atypical presentation ofparaneoplastic pemphigus, absence of mucosalinvolvement. After treatment with oral prednisone1 mg/kg daily, the skin lesions cleared completely,with the patient still free of disease 18 monthsafter diagnosis. This favorable outcome with corti-

costeroid use alone in this case is unusual and maybe due to the lack of adverse prognostic factors,such as severe mucosal involvement and a pro-gressive underlying cancer.50

Other attempted treatments that are generallyunsuccessful include immunosuppression withcyclophosphamide or azathioprine, cyclosporine,gold, dapsone, plasmapheresis, and photophere-sis.38-41,44,47,51,63Only a few successful outcomeshave been described. Krunic et al46 described apatient with an encapsulated round-cell liposarco-ma who, after tumor resection and administrationof prednisone (0.75-1.0 mg/kg) and azathioprine(100 mg/day), experienced a significant reductionof skin lesions in 2 weeks and a gradual healing oforal lesions. Another patient treated with pred-nisone 40 mg and azathioprine 100 mg daily expe-rienced complete resolution of skin lesions andmarked alleviation of the stomatitis with onlyresidual ulceration of the tongue 2 years after diag-nosis.44 The combination of oral cyclosporine 5mg/kg daily and prednisone 35 mg daily has alsobeen reported to heal erosions of the mucous mem-branes substantially within 2 months in a patientwith chronic lymphocytic leukemia.63

Izaki et al61 reported a potential therapeuticeffect of plasmapheresis in a patient with paraneo-plastic pemphigus and chronic lymphocyticleukemia. The treatment resulted in a decrease inthe antibody titer from 1:640-1280 to 1:20-40, aswell as some clearing of the skin and mucosallesions. Two other patients described in the litera-ture also received plasmapheresis therapy, in com-bination with systemic prednisone, experiencingsignificant clearing of skin and mucosallesions.43,115 Schoen et al62 recently described apatient with paraneoplastic pemphigus and aninflammatory myofibroblastic tumor who showedonly minimal improvement after excision of theneoplasm and administration of corticosteroids. Animmunoapheresis regimen was then initiated usingsheep anti-human-IgG bead-formed agarose gel,leading to disappearance of circulating autoantibod-ies and the patient’s recovery. Postapheresis ofhigh-dose immunoglobulin was administered toprevent autoantibody resynthesis.62

CLINICAL COURSE AND PROGNOSISPemphigus herpetiformis

The most significant observation in the follow-up of patients with pemphigus herpetiformis is the

Robinson et al 665


generally clears substantially or goes into com-plete remission once the neoplasm is surgicallyexcised.39 Follow-up serology has shown a reduc-tion in levels of circulating antibodies against theparaneoplastic pemphigus antigens by more than90%.39 Reported cases of paraneoplastic pemphi-gus with malignant neoplasms have generally fol-lowed a rapidly progressive and invariably fatalcourse.38,39,67 Anhalt39 reported that 30 of 33patients with malignancies died of complicationsof the syndrome or of treatment 1 month to 2 yearsafter diagnosis. Two rare cases of prolonged sur-vival, both associated with chronic lymphocyticleukemia, have been described: one patient wasstill alive 3 years after the onset of disease44; theother patient died nearly 8 years after the onset ofparaneoplastic pemphigus.63

The disease course does not parallel that of theunderlying malignancy and has even occurred inpatients apparently cured of their malignancy.39,43

In 1 case, paraneoplastic pemphigus developedafter successful autologous bone marrow trans-plantation for non-Hodgkin’s lymphoma. Thepatient was free of detectable tumor by autopsy,but died of respiratory complications of paraneo-plastic pemphigus.43 The mucosal disease isparticularly refractory and often progresses toinvolve the respiratory tract, resulting in respira-tory failure.114

Of the patients who have not survived, death isusually secondary to sepsis, gastrointestinalbleeding, multiple organ failure, or respiratoryfailure.39 Patients with autoimmune disease asso-ciated with B-cell neoplasms tend to have a highincidence of autoimmune cytopenias, often pre-disposing them to subsequent sepsis.39 Anhalt39

reported that 6 of 33 patients died of pulmonaryfailure and had severe hypoxia with a relativelyclear chest roentgenogram before death. Thepatient described by Fullerton et al43 with pul-monary involvement received a bronchial biopsybefore death, which revealed IgG deposition inthe intercellular spaces of the bronchial epitheli-um in vivo. Considering that the autoantibodiesof paraneoplastic pemphigus react with desmo-plakins and desmoplakins are present in respira-tory epithelium, it is possible that the respiratoryfailure may have been due to autoantibody-mediated injury to bronchial epithelium, resultingin plugging of terminal alveoli and ventilation/perfusion abnormalities.39

benign course of the disease. The disease respondswell to treatment with a tendency to achieve com-plete remission and is controlled with a low doseof steroids.11 Some physicians observed that theclassic features of pemphigus foliaceus, or occa-sionally pemphigus vulgaris, will eventuallydevelop in many patients with pemphigus herpeti-formis. In some cases, after treatment has beendiscontinued, the cutaneous lesions transform intoa pemphigus foliaceus phenotype.8,9 Once thistransformation occurs, the treatment regimenshould include corticosteroids and immunosup-pressive drugs. In the study by Maciejowska,Jablonska, and Chorzelski,8 the presence ofdetectable circulating autoantibodies in pemphigusherpetiformis also appears to be associated withgreater resistance to therapy.

IgA pemphigus

A follow-up study up to a maximum of 22 yearshas determined that IgA pemphigus carries achronic, benign course and responds well to appro-priate therapy.22 IgA pemphigus, as a superficialblistering disease, usually heals without scarring,if appropriate treatment is provided.31 The milderclinical presentation and benign course may alsobe explained by the lower IgA autoantibody titerand the inability of IgA to activate complement bythe classical pathway.31 Recurrences of lesionshave been noted after termination of treatment orreduction in drug dosage.26

IgA pemphigus is also associated with othermedical disorders. In the cases reviewed byWallach22 in 1992, 6 of the 29 patients had anassociated monoclonal gammopathy. This gam-mopathy was of the IgA class with κ light chainsin 5 of the 6 patients. Two gammopathies werebenign; one patient had a B-cell lymphoma, andtwo patients had myeloma. In two patients, themonoclonal gammopathy developed only yearsafter the onset of the dermatosis.22 In those caseswith an associated malignant IgA gammopathy,the prognosis was related to the malignancy.29

Gastrointestinal diseases may also be associatedwith IgA pemphigus. One case each of Crohn’sdisease and gluten-sensitive enteropathy have beenreported in the literature.22


As stated earlier, in patients with paraneoplasticpemphigus and benign neoplasms, the disease

666 Robinson et al


COMPLICATIONS OF TREATMENTSystemic corticosteroids

Systemic corticosteroids are the most usefuldrugs in treating the common forms of pemphigus,rapidly inducing remission in the majority ofpatients. After their introduction in the 1950s, therewas a dramatic decline in mortality; however, thecomplications of corticosteroids caused by the pro-longed high-maintenance doses resulted in signifi-cant morbidity.116-118The most common complica-tions of prolonged corticosteroid use are quiteextensive:hypertension, diabetes mellitus, pancre-atitis, peptic ulcers, opportunistic infections, moodalteration, psychosis, hirsutism, osteoporosis,aseptic necrosis of the femoral and humoral heads,abdominal striae, menstrual irregularity, truncaland facial obesity, hypothalamus-pituitary-adrenalaxis suppression, cataracts, glaucoma, purpura,hypopigmentation, cutaneous atrophy, impairedwound healing, and proximal myopathy.Routinebiochemistry, urinalysis, and determinations ofblood pressure and weight should be performed atregular intervals. Bone densitometry, especially inpostmenopausal women, should be performed at 1-to 2-year intervals. Ophthalmologic examinationsshould be performed regularly to screen forcataracts and glaucoma.119

Azathioprine

The side effects of azathioprine include terato-genicity, idiopathic hepatitis, increased susceptibil-ity to infection secondary to a dose-dependent bonemarrow depression, and an increased risk of inter-nal and cutaneous malignancy. Patients shouldundergo urinalysis, complete blood cell count, renaland liver function tests, and determination of elec-trolyte levels before starting therapy and at regularintervals as long as therapy is continued.120

Cyclophosphamide

Complications of cyclophosphamide includenausea, vomiting, alopecia, sterility, teratogenicity,hemorrhagic cystitis, bone marrow depression, andan increased risk of internal malignancy. Patientsshould be monitored as for azathioprine, but withthe addition of urinalysis weekly for the first 8 to12 weeks, then every 2 weeks thereafter.120

Dapsone and sulfonamides

Side effects of dapsone include anorexia, nau-sea, vomiting, hepatic damage, mood changes,

exfoliative dermatitis, cutaneous hypersensitivityreactions, and dose-related methemoglobinemiaand hemolysis.

Cimetidine may be given to improve gastroin-testinal tolerance to dapsone. Baseline and initial-ly weekly laboratory tests should include a com-plete blood cell count, renal and liver functiontests, and determination of electrolyte levels.Glucose-6-phosphate dehydrogenase deficiencymust be excluded before therapy. Evaluation ofmethemoglobin levels can be obtained as clinical-ly indicated.121-123The sulfonamides may causegastrointestinal symptoms, drug-induced rashes,blood dyscrasias, and renal failure.124

Plasmapheresis

Potential complications with plasmapheresisinclude maintaining venous access, a tendency tobleed, electrolyte imbalances, pulmonary edema,allergic reactions, fever, chills, hypotension, andsepticemia. Frequent monitoring of vital signs andcardiac monitoring is required during the proce-dure. Weekly coagulation studies, liver functiontests, hepatitis serology, and determination of elec-trolyte levels should also be performed.125,126

We thank Dr Kathleen J. Green (Departments ofPathology and Dermatology, Northwestern UniversityMedical School) and Dr Jonathan C. R. Jones(Department of Cell and Molecular Biology, North-western University Medical School) for their help inpreparing the schematic diagram of the target autoanti-gens in the new pemphigus variants.

REFERENCES

1. Thivolet J. Pemphigus: past, present and future.Dermatology 1994;189(suppl 2):26-9.

2. Dugan EM, Anhalt G, Diaz LA. Pemphigus. In: JordanRE, editor. Immunologic diseases of the skin. Norwalk(CT): Appleton & Lange; 1991. p. 279-91.

3. Amagai M, Hashimoto T, Green KJ, Shimizu N,Nishikawa T. Antigen-specific immunoadsorption ofpathogenic autoantibodies in pemphigus foliaceus. JInvest Dermatol 1995;104:895-901.

4. Amagai M, Klaus-Kovtum V, Stanley JR. Autoantibodiesagainst a novel epithelial cadherin in pemphigus vul-garis, a disease of cell adhesion. Cell 1991;67:869-77.

5. Amagai M, Koch PJ, Nishikawa T, Stanley JR.Pemphigus vulgaris antigen (desmoglein 3) is localizedin the lower epidermis, the site of blister formation inpatients. J Invest Dermatol 1996;106:351-5.

6. Jablonska S, Chorzelski T, Beutner E, Chorzelska J.Herpetiform pemphigus, a variable pattern of pemphi-gus. Int J Dermatol 1975;14:353-9.

7. Floden CH, Gentale H. A case of clinically typical der-matitis herpetiformis (M. Duhring) presenting acanthol-ysis. Acta Derm Venereol (Stockh) 1955;35:128.

Robinson et al 667


29. Zillikens D, Miller K, Hartmann AA, Burg G. IgA pem-phigus foliaceus: a case report. Dermatologica 1990;181:304-7.

30. Ebihara T, Hashimoto T, Iwatsuki K, Takigawa M, AndoM, Ohkawara A, et al. Autoantigens for IgA anti-inter-cellular antibodies of intercellular IgA vesiculopustulardermatosis. J Invest Dermatol 1991;97:742-5.

31. Wang J, Kwon J, Ding X, Fairley JA, Woodley DT, ChanLS. Nonsecretory IgA1 autoantibodies targeting desmo-somal component desmoglein 3 in intraepidermal neu-trophilic IgA dermatosis. Am J Pathol 1997;150:1901-7.

32. Hashimoto T, Ebihara T, Nishikawa T. Studies ofautoantigens recognized by IgA anti-keratinocyte cellsurface antibodies. J Dermatol Sci 1996;12:10-7.

33. Chorzelski TP, Beutner EH, Kowalewski C, OlszewskaM, Maciejowska E, Seferowicz E, et al. IgA pemphigusfoliaceus with a clinical presentation of pemphigus her-petiformis. J Am Acad Dermatol 1991;24:839-44.

34. Lutz ME, Daoud MS, McEvoy MT, Gibson LE. Sub-corneal pustular dermatosis: a clinical study of tenpatients. Cutis 1998;61:203-8.

35. Hashimoto T, Kiyokawa C, Mori O, Miyasato M,Chidgey MAJ, Garrod DR, et al. Human desmocollin 1(Dsc 1) is an autoantigen for the subcorneal pustular der-matosis type of IgA pemphigus. J Invest Dermatol1997;109:127-31.

36. Weston WL, Friednash M, Hashimoto T, Seline P, HuffJC, Morelli JG. A novel childhood pemphigus vegetansvariant of intraepidermal neutrophilic IgA dermatosis. JAm Acad Dermatol 1998;38:635-8.

37. Teraki Y, Amagai N, Hashimoto T, Kusunoki T,Nishikawa T. Intercellular IgA dermatosis of childhood:selective deposition of monomer IgA1 in the intercellu-lar space of the epidermis. Arch Dermatol 1991;127:221-4.

38. Anhalt GJ, Kim SC, Stanley JR, Korman NJ, Jabs DA,Kory M. Paraneoplastic pemphigus: an autoimmunemucocutaneous disease associated with neoplasia. NEngl J Med 1990;323:1729-35.

39. Anhalt GJ. Paraneoplastic pemphigus. Adv Dermatol1997;12:77-96.

40. Camisa C, Helm TN. Paraneoplastic pemphigus is a dis-tinct neoplasia-induced autoimmune disease. ArchDermatol 1993;129:883-6.

41. Ostezan LB, Fabré VC, Caughman SW, Swerlick RA,Korman NJ, Callen JP. Paraneoplastic pemphigus in theabsence of a known neoplasm. J Am Acad Dermatol1995;33:312-5.

42. Fried R, Lynfield Y, Vitale P, Anhalt G. Paraneoplasticpemphigus appearing as bullous pemphigoid-like erup-tion after palliative radiation therapy. J Am AcadDermatol 1993;29:815-7.

43. Fullerton SH, Woodley DT, Smoller BR, Anhalt GJ.Paraneoplastic pemphigus with autoantibody depositionin bronchial epithelium after autologous bone marrowtransplantation. JAMA 1992;267:1500-2.

44. Camisa C, Helm TN, Liu YC, Valenzuela R, Allen C,Bona S, et al. Paraneoplastic pemphigus: a report of threecases including one long-term survivor. J Am AcadDermatol 1992;27:547-53.

45. Plewig G, Jansen T, Jungblut RM, Roher HD. Castlemantumor, lichen planus, and pemphigus vulgaris:Paraneoplastic association of immunologic disease?Hautarzt 1990;41:662-70.

46. Krunic ALJ, Kokai D, Bacetic B, Kesic V, Nikolic MM,Petkovic S, et al. Retroperitoneal round-cell liposarcoma

8. Maciejowska E, Jablonska S, Chorzelski T. Is pemphigusherpetiformis an entity? Int J Dermatol 1987;26:571-7.

9. Santi CG, Maruta CW, Aoki V, Sotto MN, Rivitti EA,Diaz LA, et al. Pemphigus herpetiformis is a rare clinicalexpression of nonendemic pemphigus foliaceus, fogoselvagem, and pemphigus vulgaris. J Am Acad Dermatol1996;34:40-6.

10. Huhn KM, Tron VA, Nguyen N, Trotter MJ. Neutrophilicspongiosis in pemphigus herpetiformis. J Cutan Pathol1996;23:264-9.

11. Ingber A, Feuerman EJ. Pemphigus with characteristicsof dermatitis herpetiformis: a long-term follow-up of fivepatients. Int J Dermatol 1986;25:575-9.

12. Beissert S, Metze D, Luger TA, Schwarz T. Pemphigusherpetiformis. Hautarzt 1993;44:594-7.

13. Passarini B, Palmerio B, Costa AM, Neri I. Pemfigoerpetiforme: descrizione di un caso. G Ital DermatolVenereol 1989;124:411-3.

14. Mutevelic-Arslanagic N, Bokonjic M, Sehic C, Ganic F.Pemphigus herpetiformis (gerpetiformnyi pemfigus).Vestn Dermatol Venerol 1989;10:72-4.

15. Weidner GO, Kleinhans D. Pemphigus herpetiformis. ZHautkr 1984;59:1663-8.

16. Skripkin IuK, Butov IuS, Khamaganova AV.Gerpetiformnyi pemfigus u dvukh bol’nykh. VestnDermatol Venerol 1984;5:43-5.

17. Bauer R, Stadler R, Immel C, Orfanos CE. Akantholyseund eosinophile Spongiose: Pemphigus herpetiformis.Erfolgreiche Retinoidtherapie. Hautarzt 1983;34:13-7.

18. Bauer R, Orfanos CE. Pemphigus herpetiformis: eineklinische Variante des Pemphigus vulgaris. Hautarzt1980;31:540-4.

19. Micali G, Musumeci ML, Nasca MR. Epidemiologicanalysis and clinical course of 84 consecutive cases ofpemphigus in eastern Sicily. Int J Dermatol 1998;37:197-200.

20. Kubo A, Amagai M, Hashimoto T, Doi T, HigashiyamaM, Hashimoto K, et al. Herpetiform pemphigus showingreactivity with pemphigus vulgaris antigen (desmoglein3). Br J Dermatol 1997;137:109-13.

21. Wallach D, Foldes C, Cottenot F. Pustulose sous-cornee,acantholyse superficielle et IgA monoclonale. AnnDermatol Venereol 1982;109:959-63.

22. Wallach D. Intraepidermal IgA pustulosis. J Am AcadDermatol 1992;27:993-1000.

23. Sneddon IB, Wilkinson DS. Subcorneal pustular der-matosis. Br J Dermatol 1979;100:61-8.

24. Beutner EH, Chorzelski TP, Wilson RM, Kumar V,Michel B, Helm F, et al. IgA pemphigus foliaceus: reportof two cases and a review of the literature. J Am AcadDermatol 1989;20:89-97.

25. Huff JC, Golitz LE, Kunke KS. Intraepidermal neu-trophilic IgA dermatosis. N Engl J Med 1985;313:1643-5.

26. Hashimoto T, Inamoto N, Nakamura K, Nishikawa T.Intercellular IgA dermatosis with clinical features ofsubcorneal pustular dermatosis. Arch Dermatol1987;123:1062-5.

27. Tagami H, Iwatsuki K, Iwase Y, Yamada M. Subcornealpustular dermatosis with vesiculo-bullous eruption:demonstration of subcorneal IgA deposits and a leuko-cyte chemotactic factor. Br J Dermatol 1983;109:581-7.

28. Hodak E, David M, Ingber A, Rotem A, Hazaz B,Shamai-Lubovitz O, et al. The clinical and histopatho-logical spectrum of IgA-pemphigus: report of two cases.Clin Exp Dermatol 1990;15:433-7.

668 Robinson et al


associated with paraneoplastic pemphigus presenting aslichen planus pemphigoides-like eruption. Int J Dermatol1997;36:526-9.

47. Stevens SR, Griffiths CEM, Anhalt GJ, Cooper KD.Paraneoplastic pemphigus presenting as a lichen planuspemphigoides-like eruption. Arch Dermatol 1993;129:866-9.

48. Lam S, Stone MS, Goeken JA, Massicotte SJ, Smith AC,Folberg R, et al. Paraneoplastic pemphigus, cicatricialconjunctivitis, and acanthosis nigricans with patchy der-matoglyphy in a patient with bronchogenic squamouscell carcinoma. Ophthalmology 1992;99:108-13.

49. Mutasim DF, Pelc NJ, Anhalt GJ. Paraneoplastic pem-phigus. Dermatol Clin 1993;11:473-81.

50. Dega H, Laporte JL, Joly P, Gabarre J, André C.Paraneoplastic pemphigus associated with Hodgkin’sdisease. Br J Dermatol 1998;138:196-7.

51. Schlesinger T, McCarron K, Camisa C, Anhalt G.Paraneoplastic pemphigus occurring in a patient with B-cell non-Hodgkin’s lymphoma. Cutis 1998;61:94-6.

52. Chiewchanvit S, Hashimoto T, Chaiwun B, Nishikawa T.A pemphigus case with long term survival, implicatingthe spectrum between paraneoplastic pemphigus andpemphigus vulgaris. Int J Dermatol 1997;36:957-8.

53. Lemon MA, Weston WL, Huff JC. Childhood paraneo-plastic pemphigus associated with Castleman’s tumour.Br J Dermatol 1997;136:115-7.

54. Hashimoto T, Amagai M, Watanabe K, Chorzelski TP,Bhogal BS, Black MM, et al. Characterization of para-neoplastic pemphigus autoantigens by immunoblotanalysis. J Invest Dermatol 1995;104:829-34.

55. Borradori L, Trueb RM, Jaunin F, Limat A, Favre B,Saurat JH. Autoantibodies from a patient with paraneo-plastic pemphigus bind periplakin, a novel member ofthe plakin family. J Invest Dermatol 1998;111:338-40.

56. Herrada J, Phillips DK, Hebert AA, Cabanillas F, JordonRE. A progressive blistering eruption in a patient withlymphoma. Arch Dermatol 1997;133:97, 100-1.

57. Helou J, Allbritton J, Anhalt GJ. Accuracy of indirectimmunofluorescence testing in the diagnosis of paraneo-plastic pemphigus. J Am Acad Dermatol 1995;32:441-7.

58. Rybojad M, Leblanc T, Flageul B, Bernard PH, Morel P,Schaison G, et al. Paraneoplastic pemphigus in a childwith a T-cell lymphoblastic lymphoma. Br J Dermatol1993;128:418-22.

59. Jansen T, Plewig G, Anhalt GJ. Paraneoplastic pemphi-gus with clinical features of erosive lichen planus associ-ated with Castleman’s tumor. Dermatology 1995;190:245-50.

60. Oursler JR, Labib RS, Ariss-Abdo L, Burke T, O’KeefeEJ, Anhalt GJ. Human autoantibodies against desmo-plakins in paraneoplastic pemphigus. J Clin Invest 1992;89:1775-82.

61. Izaki S, Yoshizawa Y, Kitamura K, Kato H, HashimotoH, Korman NJ, et al. Paraneoplastic pemphigus: poten-tial therapeutic effect of plasmapheresis. Br J Dermatol1996;134:987-9.

62. Schoen H, Foedinger D, Derfler K, Amann G,Rappersberger K, Stingl G, et al. Immunoapheresis inparaneoplastic pemphigus. Arch Dermatol 1998;134:706-10.

63. Perniciaro C, Kuechle MK, Colon-Otero G, RaymondMG, Spear KL, Pittelkow MR. Paraneoplastic pemphi-gus: a case of prolonged survival. Mayo Clin Proc1994;69:851-5.

64. Castleman B, Iversen L, Menendez V. Localized medi-

astinal lymph node hyperplasia resembling thymoma.Cancer 1956;9:822-30.

65. Keller AR, Hochholzer L, Castleman B. Hyaline vascu-lar and plasma cell types of giant lymph node hyperpla-sia of the mediastinum and other locations. Cancer1972;29:670-83.

66. Carrington PA, Anderson H, Harris M, Walsh SE,Houghton JB, Morgenstern GR. Autoimmune cytopeniasin Castleman’s disease. Am J Clin Pathol 1990;94:101-4.

67. Horn TD, Anhalt GJ. Histologic features of paraneoplas-tic pemphigus. Arch Dermatol 1992;128:1091-5.

68. O’Toole EA, Mak LL, Guitart J, Hashimoto T, AmagaiM, Woodley DT, et al. Activation of interleukin-8 secre-tion in pemphigus herpetiformis epidermis by anti-desmoglein 1 IgG autoantibodies [abstract]. J InvestDermatol 1998;110:509.

69. Supapannachart N, Mutasim DF. The distribution of IgApemphigus antigen in human skin and the role of IgAanti-cell surface antibodies in the induction of intraepi-dermal acantholysis. Arch Dermatol 1993;129:605-8.

70. Liu AY, Valenzuela R, Helm TN, Camisa C, Melton AL,Bergfeld WF. Indirect immunofluorescence on rat blad-der transitional epithelium: a test with high specificityfor paraneoplastic pemphigus. J Am Acad Dermatol1993;28:696-9.

71. Fodinger D, Bocskor B, Wolff K, Rappersberger K.Erythema multiforme: demonstration of autoantibodiesagainst desmoplakin I [abstract]. Clin Res 1993;41:419A.

72. Kowalczyk AP, Bornslaeger EA, Norvell SM, Palka HL,Green KJ. Desmosomes: intercellular adhesive junctionsspecialized for attachment of intermediate filaments. IntRev Cytol 1999;185:237-302.

73. Moll R, Moll I. Epidermal adhesion molecules and base-ment membrane components as target structures ofautoimmunity. Virchows Arch 1998;432:487-504.

74. Ruhrberg C, Hajibagheri MA, Parry DA, Watt FM.Periplakin, a novel component of cornified envelopesand desmosomes that belongs to the plakin family andforms complexes with envoplakin. J Cell Biol 1997;139:1835-49.

75. Ruhrberg C, Hajibagheri MA, Simon M, Dooley TP,Watt FM. Envoplakin, a novel precursor of the cornifiedenvelope that has homology to desmoplakin. J Cell Biol1996;134:715-29.

76. Ruhrberg C, Watt FM. The plakin family: versatile orga-nizers of cytoskeletal architecture. Curr Opin Genet Dev1997;7:392-7.

77. Ishii K, Amagai M, Komai A, Ebihara T, Chorzelski TP,Nishikawa T, et al. Desmoglein 1 and desmoglein 3 asautoimmune target of herpetiform pemphigus [abstract].J Invest Dermatol 1998;110:510A.

78. Amagai M. Adhesion molecules I: Keratinocyte-kera-tinocyte interactions; cadherins and pemphigus. J InvestDermatol 1995;104:146-52.

79. Buxton RS, Cowin P, Franke WW, Garrod DR, GreenKH, King IA, et al. Nomenclature of the desmosomalcadherins. J Cell Biol 1993;121:481-3.

80. Amagai M, Hashimoto T, Shimizu N, Nishikawa T.Absorption of pathogenic autoantibodies by the extracel-lular domain of pemphigus vulgaris antigen (Dsg 3) pro-duced by baculovirus. J Clin Invest 1994;94:59-67.

81. Kim SC, Kwon YD, Lee IJ, Lee IJ, Chang SN, Lee TG.cDNA cloning of the 210-kDa paraneoplastic pemphigusantigen reveals that envoplakin is a component of theantigen complex. J Invest Dermatol 1997;109:365-9.

Robinson et al 669


paraneoplastic cerebellar degeneration. N Engl J Med1990;322:1844-51.

98. Thirkill CE, Fitzgerald P, Sergott RC, Roth AM, TylerNK, Keltner JL. Cancer-associated retinopathy (CARsyndrome) with antibodies reacting with retinal, optic-nerve, and cancer cells. N Engl J Med 1989;321:1567-71.

99. Grunwald GB, Kornguth SE, Towfighi J, Sassani J,Simmonds MA, Housman CM, et al. Autoimmune basisfor visual paraneoplastic syndrome in patients with smallcell lung carcinoma: retinal immune deposits and abla-tion of retinal ganglion cells. Cancer 1987;60:780-6.

100. Erlandson RA, editor. Diagnostic transmission electronmicroscopy of human tumors. New York: MassonPublishing; 1981.

101.Ghadially FN, editor. Ultrastructural pathology of thecell matrix. London: Butterworth; 1982.

102. Geiger B, Schmid E, Franke WW. Spatial distribution ofproteins specific for desmosomes and adherens junctionsin epithelial cells demonstrated by double immunofluo-rescence microscopy. Differentiation 1983;23:189-205.

103. Miettinen M. Keratin subsets in spindle cell sarcomas:keratins are widespread but synovial sarcoma contains adistinctive keratin polypeptide pattern and desmoplakin.Am J Pathol 1991;138:505-13.

104. Kirsner RS, Anhalt GJ, Kerdel FA. Treatment with alphainterferon associated with the development of paraneo-plastic pemphigus. Br J Dermatol 1995;132:474-8.

105. Yee C,Biondi A, Wang XH, Iscove NN, de Sousa J,Aarden LA, et al. A possible autocrine role for inter-leukin-6 in two lymphoma cell lines. Blood 1989;74:798-804.

106.Emilie D, Devergne O, Raphael M, Coumbaras LJ,Galanaud P. Production of interleukin-6 in high grade Blymphomas. Curr Top Microbiol Immunol 1992;182:349-55.

107. Biondi A, Rossi V, Bassan R, Barbui T, Bettoni S, SironiM, et al. Constitutive expression of the interleukin-6gene in chronic lymphocytic leukemia. Blood 1989;73:1279-84.

108.Chan LS, Vanderlugt CJ, Hashimoto T, Nishikawa T,Zone JJ, Black MM, et al. Epitope spreading: lessonsfrom autoimmune skin diseases. J Invest Dermatol 1998;110:103-9.

109. Huilgol SC, Black MM. Management of the immuno-bullous disorders. II. Pemphigus. Clin Exp Dermatol1995;20:283-93.

110.Stendahl O, Molin L, Dahlgren C. The inhibition ofpolymorphonuclear leukocyte cytotoxicity by dapsone: apossible mechanism in the treatment of dermatitis her-petiformis. J Clin Invest 1978;62:214-22.

111.Piette W, Burken RR, Ray TL. Intraepidermal neu-trophilic IgA dermatosis: presence of circulating pem-phigus-like IgA antibody specific for monkey epithelium[abstract]. J Invest Dermatol 1987;88:512.

112. Burrows D, Bingham EA. Subcorneal pustular dermato-sis and IgA gammopathy. Br J Dermatol 1984;111:91-3.

113.Norris DA, Osborn R, Robinson W, Tonnesen MG.Isotretinoin produces significant inhibition of monocyteand neutrophil chemotaxis in vivo in patients with cysticacne. J Invest Dermatol 1987;89:42-3.

114. Tankel M, Tannenbaum S, Parekh S. Paraneoplasticpemphigus presenting as an unusual bullous eruption. JAm Acad Dermatol 1993;29:825-8.

115. Su WPD, Oursler JR, Muller SA. Paraneoplastic pem-phigus: a case with high titer of circulating anti–base-

82. Mahoney MG, Aho S, Uitto J, Stanley JR. The membersof the plakin family of proteins recognized by paraneo-plastic pemphigus antibodies include periplakin. J InvestDermatol 1998;111:308-13.

83. Kiyokawa C, Ruhrberg C, Nie Z, Karashima T, Mori O,Nishikawa T, et al. Envoplakin and periplakin are com-ponents of the paraneoplastic pemphigus antigen com-plex. J Invest Dermatol 1998;111:1236-8.

84. Amagai M, Nishikawa T, Nousari HC, Anhalt GJ,Hashimoto T. Antibodies against desmoglein 3 (pemphi-gus vulgaris antigen) are present in sera from patientswith paraneoplastic pemphigus and cause acantholysis invivo in neonatal mice. J Clin Invest 1998;102:775-82.

85. Anhalt GJ, Hilu J, Takehara K, Lee A, Casciola-Rosen L.Identification of a putative transmembrane autoantigenin paraneoplastic pemphigus [abstract]. Clin Res1993;41:407A.

86. Stanley JR. Cell adhesion molecules as targets of autoan-tibodies in pemphigus and pemphigoid, bullous diseasesdue to defective epidermal cell adhesion. Adv Immunol1993;53:291-325.

87. Green KJ, Parry DA, Steinert PM, Virata ML, WagnerRM, Angst BD, et al. Structure of the human desmo-plakins: implications for function in the desmosomalplaque. J Biol Chem 1990;265:2603-12.

88. Joly P, Thomine E, Gilbert D, Verdier S, Delpech A,Prost C, et al. Overlapping distribution of autoantibodyspecificities in paraneoplastic pemphigus and pemphigusvulgaris. J Invest Dermatol 1994;103:65-72.

89. Hashimoto T, Amagai M, Wang N, Nishikawa T,Karashima T, Moria O, et al. Novel non-radioisotopeimmunoprecipitation studies indicate involvement ofpemphigus vulgaris antigen in paraneoplastic pemphi-gus. J Dermatol Sci 1998;17:132-9.

90. Morioka S, Naito K, Ogawa H. The pathogenic role ofpemphigus antibodies and proteinase in epidermal acan-tholysis. J Invest Dermatol 1981;76:337-41.

91. Hashimoto K, Shafran KM, Webber PS, Lazarus GS,Singer KH. Anti-cell surface pemphigus autoantibodystimulates plasminogen activator activity of human epi-dermal cells: a mechanism for the loss of epidermalcohesion and blister formation. J Exp Med 1983;157:259-72.

92. Coffman RL, Seymour BWP, Lebman DA. The role ofhelper T cell products in mouse B cell differentiation andisotype regulation. Immunol Rev 1988;102:5-28.

93. Coffman RL, Shrader B, Carty T, Mosmann TR, BondMW. A mouse T cell product that preferentially enhancesIgA production. I. Biologic characterization. J Immunol1987;139:3685-90.

94. Carayannopoulos L, Hexham JM, Capra JD.Localization of the binding site for monocyte IgA-Fcreceptor (CD89) to the domains boundary between Cγ2and Cγ3 in human IgA1. J Exp Med 1996;183:1579-86.

95. Weisbart RH, Kacena A, Schuh A, Golde DW. GM-CSFinduces human neutrophil IgA-mediated phagocytosisby an IgA Fc receptor activation mechanism. Nature1988;332:647-8.

96. Brashear HR, Greenlee JE, Jaeckle KA, Rose JW.Anticerebellar antibodies in neurologically normalpatients with ovarian neoplasms. Neurology 1989;39:1605-9.

97. Furneaux HM, Rosenblum MK, Dalmau J, Wong E,Woodruff P, Graus F, et al. Selective expression ofPurkinje-cell antigens in tumor tissue from patients with

670 Robinson et al


ment membrane zone autoantibodies. J Am AcadDermatol 1994;30:841-4.

116. Ryan JG. Pemphigus: a 20-year survey of experiencewith 70 cases. Arch Dermatol 1971;104:14-20.

117. Rosenberg FR, Sander S, Nelson CT. Pemphigus: a 20-year review of 107 patients treated with corticosteroids.Arch Dermatol 1976;112:962-70.

118. Lever WF, White H. Treatment of pemphigus with corti-costeroids: results obtained in 46 patients over a periodof 11 years. Arch Dermatol 1963;87:12-26.

119. Truhan AP, Ahmed AR. Corticosteroids: a review withemphasis on complications of prolonged systemic thera-py. Ann Allergy 1989;62:375-91.

120. Ho VC, Zloty DM. Immunosuppressive agents in derma-tology. Dermatol Clin 1993;11:73-85.

121. Rogers RS. Dapsone and sulfapyridine therapy of pem-phigoid diseases. Australas J Dermatol 1986;27:58-63.

122. Stern RS. Systemic dapsone. Arch Dermatol 1993;129:301-3.

123. Coleman MD. Dapsone: modes of action, toxicity andpossible strategies for increasing patient tolerance. Br JDermatol 1993;129:507-13.

124. Huilgol SC, Black MM. Management of the immuno-bullous disorders. I. Pemphigoid. Clin Exp Dermatol1995;20:189-201.

125. Guillot B, Donadio D, Guilhou JJ, Meynadier J. Longterm plasma exchange therapy in bullous pemphigoid.Acta Derm Venereol (Stockh) 1986;66:73-5.

126. Bystryn JC. Plasmapheresis therapy of pemphigus. ArchDermatol 1988;124:1702-4.

Robinson et al 671

Documents

Journal of the DERMATOLOGYAmerican Academy ofphdres.caregate.net/curriculum/pdf-files/Derm_refs/PemphigusVariants.pdfautoantibodies of patients with pemphigus vulgaris and pemphigus