Eyelid deformities from pyodermatitis pyostomatitis vegetans

Gregory Moloney,P. Dolman

DOI: https://doi.org/10.1111/j.1442-9071.2011.02583.x

2011-12-01

Clinical and Experimental Ophthalmology

Abstract:Pyostomatitis pyodermatitis vegetans (PSPDV) is an inflammation of the skin and mucous membranes associated commonly with inflammatory bowel disease, particularly ulcerative colitis. Ocular adnexal involvement with this condition is rare. We describe a case of PSPDV presenting with conjunctival involvement. A 50-year-old woman was referred to our clinic for repair of upper lid cicatricial entropions and nodules following 4 years of recurrent upper eyelid inflammation (Figs 1,2). On her first episode, she had associated flu-like symptoms with a history of ulcerative colitis controlled by a combination of azathioprine and mesalamine. She denied oral or skin complications. Her first ophthalmologist incised both upper lids for suspected chalazia but this caused increased conjunctival ulceration and scarring. Conjunctival biopsies found a mixed inflammatory cell infiltrate with prominent eosinophilia and neutrophilia, with lymphocytes in deeper layers (Fig. 3). Immunofluorescent stains were negative for IgG, IgA, IgM, C3 and fibrin. Ocular cicatricial pemphigoid was suspected, and dapsone was added at 50 mg/day, increasing gradually to 100 mg/day. This additional medication appeared to provide some control of the inflammation but necrosis and scarring developed nonetheless. Her disease settled after 18 months of therapy with combination azathioprine, mesalamine and dapsone. Several short pulses of oral prednisone were given throughout this period for flare-ups of her colitis. She was referred to our clinic 3 years later for repair of a left upper eyelid nodule and cicatricial entropion. Her unusual history prompted review of her condition. At the time of initial symptoms her full blood count was normal apart from an eosinophilia (0.8 giga/L, normal less than 0.5). She was noted to have snail track scarring in the buccal mucosa and her diagnosis changed to PSPDV (Fig. 4). The left upper lid nodule with associated trichiasis was excised with a full thickness wedge and direct closure, using perioperative oral corticosteroid cover. Pyostomatitis pyodermatitis vegetans is an inflammatory disorder of unknown aetiology affecting the mucosal surfaces of the mouth, the intertriginous skin and rarely the conjunctiva. Characteristically the process begins as small pustules in each location that coalesce into ‘snail tracks’. Vegetating lesions of the skin may erupt. Typical histopathology includes intraepithelial and subepithelial abscesses, a preponderance of eosinophils diminishing with time, underlying lymphocytic and neutrohpilic infiltrate and possible acanthosis and hyperkeratosis of overlying epithelium. An acute peripheral eosinophilia is present in 90% of cases. Ocular adnexal involvement with PSPDV is rare. Leibovitch et al. described a case masquerading as folliculitis and reviewed two previous papers describing ocular involvement. The diagnosis may be challenging as no diagnostic test exists for PSPDV. It is differentiated from pemphigobullous disease based on clinical features together with supportive histology and negative immunofluorescence. Biopsy results may vary based on the site and timing of the process. An early eosinophilia gives way to a mixed chronic inflammatory cell infiltrate and eventually tissue necrosis, fibrosis and scarring. Inflammatory bowel disease is well known to be associated with a variety of dermatites including Sweet’s syndrome, Pyoderma Gangrenosum, erythema elevatum diutinum and granuloma faciale. There has been considerable debate over the classification of this clinical spectrum with recent trend moving towards unification under the term neutrophilic dermatoses. Some authors postulate that differing levels of skin involvement by the same pathogenic process confers the different clinical appearances of this group. All share infiltration of the skin by normal neutrophils with tissue inflammation and necrosis. PSPDV may be distinct from these disorders but has considerable clinical overlap. Our case bears some similarities to a series of pyoderma gangrenosum affecting the eyelids. The authors describe a nodular inflammatory process followed by necrosis with predilection for the central 1/3 of the upper eyelid with sparing of the tarsal margin. However, we favour the diagnosis of PSPDV based on the painless nature of the lesions, the presence of local and peripheral eosinophilia, the characteristic oral findings and the primary involvement of the conjunctiva rather than the lid skin. Atypical features of our case are that our patient denies ever having skin involvement elsewhere and that she had minimal stomatitis, recognized only in retrospect by the presence of scarring. Her tempered presentation may have resulted from baseline immunosuppression with azathioprine, mesalasine, dapsone and, some months later, periodic oral prednisone pulses. All these agents are known to suppress PSPDV and other neutrophilic dermatoses. Systemic steroids are perhaps the most potent agent in the short term. It is unsure whether more aggressive earlier therapy might have spared some of the ultimate deformity of the eyelids. Ophthalmologists should be aware of the cutaneous manifestations of inflammatory bowel disease for several reasons. Although ocular involvement is rare, findings are potentially reversible with immunosuppression. Conversely if allowed to progress, extensive tissue necrosis Clinical and Experimental Ophthalmology 2011; 39: 910–924 doi: 10.1111/j.1442-9071.2011.02583.x