Linear IgA bullous dermatosis and ulcerative colitis treated by proctocolectomy

ARTICLE

Auteur(s) : Giacomo CALDAROLA¹, Vito ANNESE², Fabrizio BOSSA², Riccardo PELLICANO¹

¹Department of Dermatology, Casa Sollievo della Sofferenza Hospital-IRCCS, Viale Cappuccini, San Giovanni Rotondo, Italy
²Gastrointestinal & Endoscopy Units, Casa Sollievo della Sofferenza IRCCS Hospital, San Giovanni Rotondo, Italy

In 2004 a 24-year-old man presented with a diffuse, non-pruritic, bullous eruption with an annular arrangement that had arisen a week earlier on the trunk, axillae and scalp. Multiple erosions of the oral mucosa made speaking and eating difficult. The patient also had a 1-year history of ulcerative colitis (UC) with extensive colon involvement (pancolitis).

Findings of subepidermal blistering on the histological examination; linear IgA but no IgG, IgM or complement deposits at the BMZ on direct immunofluorescence; and IgA directed against a 120 kD molecule on Western blot analysis led to a diagnosis of Linear IgA dermatosis (LAD). Mild iron deficiency anemia and a slightly increased erythrocyte sedimentation rate were also detected. A malignancy was excluded.

Treatment with dapsone dramatically improved the cutaneous lesions in a few weeks and allowed suboptimal control of the skin disorder over the next 3 years. Treatment of the UC symptoms with oral and rectal mesalazine, several cycles of oral prednisone, and azathioprine (suspended after 1 month because of acute mild pancreatitis) did not achieve an acceptable control of the IBD. He also refused biological therapy with infliximab. In February 2006, he was admitted to a gastroenterology ward for severe UC, not controlled by high doses of intravenous steroids and cyclosporine. Worsening of symptoms and malnutrition led to total colectomy. Six months from surgery the cutaneous disease had not recurred and dapsone was tapered off. Three years after the operation the patient has experienced no new cutaneous eruptions.

Ulcerative colitis is a relapsing, non-transmural inflammatory disease confined to the colon mucosa: it is frequently described in association with autoimmune bullous diseases, particularly LAD. LAD is characterized by subepidermal blisters, characteristically arranged in a “clusters of jewels” configuration, and IgA antibodies against some BMZ antigens. To our knowledge, 15 patients with UC and LAD have been described to date. The association is remarkable, given the extreme rarity of LAD. In a review of 70 British patients with LAD, Paige et al. found concomitant UC in five (7.1%), while the prevalence of ulcerative colitis in the UK is 0.05% [1]. Both diseases present some analogies in their pathogenesis, such as the role of IgA1 and TNF alpha, but are related to a different genetic predisposition, as reflected by the specific HLA haplotypes found in most patients. No recurrent haplotypes were found in the patients affected concomitantly with both diseases (data not shown).

Since in our patient, and in all reported cases, UC predated LAD by months or years, it has been suggested that the bowel injury may be the first step in the LAD pathogenesis for these patients. In fact, since LAD has been described in patients taking sulfasalazine, mesalazine, steroids or even no drugs, its onset in patients with UC is unlikely to be drug-induced. Nonetheless, a close correlation has recently been described for the first time between the appearance of blisters and the activity of the IBD [2]. Finally, the resolution of the skin disorder after removal of the entire colon, appears to support this hypothesis. Colectomy was effective in our patient and in a number of reported cases [3-5]. Paige et al. described two LAD patients who did not respond to colectomy [1], but the information on these patients and their treatments is incomplete; indeed, in two other cases reported as cured [3, 4] the first resection was also ineffective and the residual rectal stump had to be removed (proctocolectomy) to achieve complete cutaneous healing.

In patients with LAD and UC, the bowel disease is probably closely related to the production of LAD autoantibodies directed against the BMZ, probably through a cross-reaction with some exogenous auto-antigens or through the epitope spreading phenomenon. These would be the same mechanisms invoked to explain the presence of autoantibodies against epidermolysis bullosa acquisita antigen in the sera of patients with an IBD [6]. Finally, regression and/or control of the IBD through surgical or medical treatment is likely to benefit the cutaneous disease.

Acknowledgements

References

2 Taniguchi T, Maejima H, Saito N, Katsuoka K, Haruki S. Case of linear IgA bullous dermatosis-involved ulcerative colitis. Inflamm Bowel Dis 2008; [Epub ahead of print].

3 Handley J, Shields M, Dodge J, et al. Chronic bullous disease of childhood and ulcerative colitis. Pediatr Dermatol 1993; 10: 256-8.

4 Egan CA, Meadows KP, Zone JJ. Ulcerative colitis and immunobullous disease cured by colectomy. Arch Dermatol 1999; 135: 214.

5 Walker SL, Banerjee P, Harland CC, et al. Remission of linear IgA disease associated with ulcerative colitis following panproclocolectomy. Br J Dermatol 2000; 143: 1341-2.

6 Chen M, O’Toole EA, Sanghavi J, et al. The Epidermolysis Bullosa Acquisita Antigen (Type VII Collagen) is Present in Human Colon and Patients with Crohn’s Disease have Autoantibodies to Type VII Collagen. J Invest Dermatol 2002; 118: 1059-64.