Linear IgA Disease Complicated by Erythema Multiforme · ly developed erythema multiforme, possibly due to dapsone. Linear IgA dermatosis should be consid-ered in patients presenting

27 Dermatol Sinica, March 2006

From National Skin Centre, Singapore and Department of Pathology, Singapore General Hospital*Accepted for publication: September 05, 2005Reprint requests: Dr Ching Yin Neoh, National Skin Centre,1 Mandalay Road, Singapore 308205TEL:65-6253 4455 FAX:65-6253 3225 E-mail:cyneoh@nsc.gov.sg

IgAMalcolm W Greaves M Priyanthi Kumarasinghe*

*

Linear IgA Disease Complicated by Erythema MultiformeChing Yin Neoh Malcolm W Greaves M Priyanthi Kumarasinghe*

A 55-year-old Chinese man presented with acute onset of a vesicobullous rash on his face, axillae,genitalia and trunk. The skin biopsy showed a subepidermal bulla. The diagnosis was made by directimmunofluorescence of the perilesional skin which revealed characteristic linear deposition of IgAalong the dermoepidermal junction. He was successfully treated with dapsone initially but subsequent-ly developed erythema multiforme, possibly due to dapsone. Linear IgA dermatosis should be consid-ered in patients presenting with bullous lesions, even when the distribution of the rash is atypical.Alternative diagnoses should be entertained when patients experience an unexpected relapse whilst ontreatment. We highlight the important points in the clinical presentation, diagnosis and the problemsencountered in the treatment of this uncommon skin disorder.(Dermatol Sinica 24: 27-31, 2006)

Keywords: Bullous disease, Dapsone, Erythema multiforme, Linear IgA dermatosis

55

IgA

IgA dapsone

dapsone IgA

( 24: 27-31, 2006)

Dermatol Sinica, March 2006 28

IgA

INTRODUCTIONLinear IgA dermatosis is an uncommon

autoimmune vesicobullous disease in Singapore.1

It is characterized by linear deposition of IgAautoantibody at the basement membrane zone.We report a patient with linear IgA dermatosiswho responded briefly to dapsone therapy butlater developed a drug reaction, in the form oferythema multiforme.

CASE REPORTA 55-year-old Chinese man presented with

a two week history of an itchy vesicobullousrash. The rash started on the face and lips, andthen spreading to the trunk and axillae. It firstpresented as papules which progressed tobecome vesicles. He denied any form of topicaltreatment on the body. Systemic review did notreveal any other symptoms.

There was no history of previous episodesof skin problems and he did not have any familyhistory of skin diseases. He had not taken anymedications but confessed to drinking herbal teafrom Chinese medical halls intermittently forthe past one year. He had otherwise no past his-tory of note. He worked as a hawker assistant.

Physical examination revealed a vesicob-ullous eruption with symmetrical flexural distri-bution including the axillae and genitalia. Thebullae were tense and the Nikolsky sign wasnegative (Fig. 1). There were no mouth lesionsand the rest of the physical examination wasunremarkable. Because of the atypical distribu-tion, the provisional clinical diagnosis was lin-

ear IgA disease but bullous pemphigoid couldnot be excluded on clinical grounds.

Histological examination of the skin biop-sy showed a subepidermal bulla containing fib-rinous exudate with an admixture of neutrophilsand occasional eosinophils (Fig. 2). The dermiscontained a perivascular lymphocytic infiltrate.Direct immunofluorescence studies showed pre-dominant linear deposits of IgA, with IgG andC3 in the basement membrane zone and in thefloor of the bulla (Fig. 3). Indirect immunofluo-rescence examination of the patient's serumshowed no anti-basement membrane or anti-intercellular substance antibodies.

The diagnosis of linear IgA disease wasmade and he was started on prednisolone 50mgonce daily. The patient was reviewed 2 weekslater. He had run out of prednisolone for 4 days

Fig. 1 Bullae on the thigh of the patient.

Fig. 2 Subepidermal bullous lesion with a fibrinous exudates andadmixture of neutrophils and eosinophils in the floor of thebulla. ( H&E original magnification, x200)

Fig. 3 Linear IgA deposits with direct immunoflourescence studies.

29 Dermatol Sinica, March 2006

and there were active bullous lesions on theaxillae and face. Prednisolone was reintroducedand the dose was increased to 60 mg per day.He was also started on dapsone 50mg a day, hisglucose-6-phosphate dehydrogenase status hav-ing been found to be positive. Other bloodinvestigations including full blood count, renaland liver function tests were normal. One weeklater, his skin lesions were improved with driedvesicles. Over the next 2 weeks, prednisolonewas tailed down to 40mg a day and dapsoneincreased to 100mg a day. 3 weeks later, he pre-sented to clinic with generalised erythema andswollen palms with generalized target-like skinlesions. He was admitted with a diagnosis of anerythema multiforme-like drug eruption possi-bly secondary to dapsone.

Examination at this time revealed a sym-metrical erythematous maculopapular and targe-toid rash on the arms, trunk and legs and vesi-cles on the left arm. There was no mucosalinvolvement. Dapsone was withdrawn and hewas maintained on prednisolone 40mg a day. Abiopsy of a targetoid skin lesion showed inter-face dermatitis compatible with early lesions oferythema multiforme (Fig. 4). Blood investiga-tions were unremarkable except for mildanaemia of 11.8g/dL (14.0-18.0), which couldbe secondary to dapsone therapy. The renal and

liver function tests were normal. The rash sub-sided after five days, but he developed newvesicles in the axillae and trunk, possibly arelapse of the linear IgA disease. He was dis-charged with an increased dose of prednisolone50mg a day. On review a week later, he did nothave new bullae and the dose of prednisolonewas gradually tapered. The haemoglobin levelsnormalized spontaneously 1 month after dis-charge and had since remained stable.

DISCUSSIONLinear IgA dermatosis is one of the rarest

subepidermal blistering diseases in WesternEurope and is more common in China. It is sur-prisingly not common in Singapore, where themajority of the population is Chinese.1

Typically, it occurs in patients over thirty yearsof age, with the average age of onset after sixtyyears old and is more common in females.2 Thedisease is characterized by annular or groupedpapules, vesicles or bullae on a normal or ery-thematous base. The lesions are typically dis-tributed symmetrically along the extensor sur-faces and are often pruritic.

Linear IgA dermatosis may mimic der-matitis herpetiformis or bullous pemphigoid,but it is a separate disorder that is bothimmunopathologically and immunogeneticallydistinct.2, 3 There may also be mucosal involve-ment, clinically indistinguishable from cicatri-cial pemphigoid.3 Chronic bullous disease ofchildhood is an immunopathologically similardisease which differs only in age of onset andthat it typically resolves within 2 to 4 years ofonset. 2, 3

Most cases of linear IgA dermatosis areidiopathic, although it has been associated withinfections, malignancies4 and drugs, includingvancomycin,5 fruosemide6 and captopril.7 Thereare also two reports of ultraviolet light inducedIgA dermatosis.8 Our patient denied taking anydrugs around the time his skin lesions appeared.It is unlikely that the Chinese herbal tea wasresponsible for inducing his disease.

Subepidermal bullae containing predomi-nantly neutrophils seen in linear IgA disease

Fig. 4 Interface dermatitis with basal vacuolar change and lympho-cytic infiltrate obliterating the dermo-epidermal interface,suggestive of early erythema multiforme. (H&E originalmagnification x100)

Dermatol Sinica, March 2006 30

IgA

can be indistinguishable from dermatitis her-petiformis on light microscopy. Those with apredominant infiltrate of eosinophils may mim-ic bullous pemphigoid. The type of inflammato-ry infiltrate present may depend on the age ofthe lesion and the time of the biopsy. In our ownpatient, histological examination revealed a sub-epidermal bulla with predominant neutrophilsand direct immunofluoresence of perilesionalskin showed linear deposits of IgA along thedermoepidermal junction. Indirect immunofluo-rescence on sodium chloride split skin, whichwas not done in our patient, yields variableresults with immunoreactants mapping to theroof or floor of the split or both.9 There is evi-dence of antigenic heterogeneity in idiopathiclinear IgA dermatosis as demonstrated byWestern immunoblotting. Target antigensinvolved include BP230/BPAG1, BP180/BPAG2,LAD antigen, type VII collagen and a 97kDaantigen representing a cleaved ectodomain ofBPAG2 (BP180). IgG autoantibodies may alsobe involved.10 Linear IgA deposits have beenreported in a few other conditions, includingcutaneous varicella-zoster infection.11

A rare group of bullous pemphigoidpatients may also have IgA, IgG/C3 depositsalong the basement membrane zone and distinc-tion between linear IgA disease may be chal-lenging.3 The classification of these patients hasbeen debated upon. Some authorities classifyonly patients with IgA deposits as linear IgAdisease while others categorize patients withboth IgG and IgA on the basis of the predomi-nant immunoglobulin on the directimmunoflourescence. The diagnosis of bullouspemphigoid would be more convincing if therewere circulating anti-basement membraneautoantibodies against BP180 or type VII colla-gen. These tests were not performed in ourpatient. It can be argued that the erythema mul-tiforme skin lesions in our patient may representone of the protean manifestations of linear IgAdisease. However, the time course of theappearance and disappearance of the erythemamultiforme-like rash correlated well with theadministration and withdrawal of dapsone. It is

also unlikely that the patient's linear IgA diseaseshould flare without other exacerbating triggers,after having good initial response to dapsone.The patient also subsequently developed morecharacteristic linear IgA vesicles, reminiscent ofhis initial presentation, after dapsone was with-drawn.

Linear IgA dermatosis has a variablecourse and may last for years with spontaneousresolution in some patients. Patients with linearIgA dermatosis respond favourably to dapsone,but some require low dose prednisolone to sup-press blister formation. Our patient developederythema multiforme whilst on dapsone, whichmay have been due to this drug.12 Adverse reac-tions to dapsone are more likely to occur inpatients with a slow acetylator phenotype.13

Cutaneous reactions to dapsone are well recog-nized and include the dapsone hypersensitivitysyndrome,14 toxic epidermal necrolysis andStevens Johnson syndrome.15, 16, 17 Dapsoneremains one of the most effective drugs fortreatment of linear IgA dermatosis and intoler-ance to dapsone may impede the future manage-ment of his skin condition.

In conclusion, we have described a patientwith linear IgA dermatosis who developed ery-thema multiforme possibly due to dapsone. Thiscomplication should be considered in anypatient with linear IgA bullous dermatosis whodevelops an otherwise unexplained flareup dur-ing dapsone treatment.

REFERENCES1. Wong SN, Chua SH: Spectrum of subepidermal

immunobullous disorders seen at National SkinCentre, Singapore: a 2 year review. Br J Dermatol147: 476-480, 2002.

2. Bickle KM, Roark TR, Hsu S: AutoimmuneBullous Dermatoses: a review. Am Fam Phys 65:1861-1870, 2002.

3. Herron MD, Zone JJ: Dermatitis herpetiformis andlinear IgA bullous dermatosis. In: Bolognia JL,Jorizzo JL, Rapini RP, et al., eds. Dermatology. 1sted. London: Elsevier Limited, 479-489, 2003.

4. McEvoy MT, Conolly SM: Linear IgA dermatosis:association with malignancy. J Am Acad Dermatol22: 59-63, 1990.