Epidermolysis Bullosa Acquisita With Extensive Esophageal Erosions

An Asian male in his 40s with a 7-year history of pemphigoid disease presented to our department. The patient exhibited persistent blisters on the extremities and oral erosions, with minor scarring (Figure 1a,b,f) (BPDAI: 12), despite oral prednisolone (PSL) at 3 mg/day. The initial diagnosis of pemphigoid disease was based on clinical features, histopathological subepidermal blister (Figure 1c), and linear deposition of IgG, IgM, and C3, but not IgA, at the dermal-epidermal junction on direct immunofluorescence. As the specific autoantigen remained unidentified, serological evaluation was repeated, revealing elevated anti-collagen VII (COL7) IgG (42.6 U/mL; normal < 6 U/mL) by enzyme-linked immunosorbent assay (ELISA) (Figure 1f), establishing the diagnosis of epidermolysis bullosa acquisita (EBA). The lesions partially improved after increasing PSL to 6 mg/day with dapsone (25 mg/day) (BPDAI: 6). Eleven months later, the patient reported minor dysphagia. Since laryngoscopy detected no pharyngeal or laryngeal involvement and no cutaneous relapse, treatment was not escalated. One month later, dysphagia progressed to difficulty even drinking water. Esophagogastroscopy identified extensive esophageal erosions (Figure 1d). Histopathology showed separation between the mucosal epithelium and lamina propria with neutrophilic infiltration (Figure 1e), and anti-COL7 IgG titer was elevated (83.2 U/mL) (Figure 1f), supporting an EBA flare, although direct immunofluorescence was not technically feasible. Intensive treatment with PSL (0.5 mg/kg/day), plasmapheresis, and intravenous immunoglobulin resulted in clinical improvement within 3 weeks, with decreased anti-COL7 IgG (5.2 U/mL) (Figure 1f). The patient has since remained stable for 1 year under treatment with 10 mg/day PSL, 25 mg/day dapsone, and 1500 mg/day mycophenolate mofetil. EBA is a pemphigoid disease characterized by autoantibodies against COL7, predominantly of the IgG isotype [1]. Clinically, it is classified into classical mechanobullous and inflammatory types. This case, manifesting minimal erythema and friction-induced blisters, is compatible with the mechanobullous subtype. While both subtypes can involve mucosal lesions, esophageal involvement is rare, reported in 2.6% of EBA [2]. In this case, esophageal severity did not parallel cutaneous severity, suggesting partially distinct mechanisms between mucosal and skin lesions. Based on former literature suggesting IgA-type autoantibody involvement in mucosal lesions [3, 4], we examined anti-COL7 IgA levels by our in-house ELISA, following published protocols [5]. The optical density (OD) was slightly above the control cutoff at baseline (0.134; cutoff < 0.092, defined as the mean + 3 standard deviations of healthy controls (N = 4)), increased during the esophageal flare (0.218), and decreased after treatment (0.141), paralleling mucosal activity (Figure 1f). Nonetheless, involvement of IgG-type autoantibodies cannot be excluded, as both isotypes increased at the mucosal flare. We further speculate that behavioral minimization of cutaneous friction decreased skin activity, whereas unavoidable esophageal mechanical stress permitted mucosal exacerbation. These factors, IgA-associated autoimmunity and unavoidable esophageal friction, may underlie the skin–esophageal severity mismatch. Alternatively, this case may be interpreted as a transition from mechanobullous-type EBA to mucous membrane pemphigoid associated with IgA-type COL7 antibodies, which would better describe the skin–mucosal severity discrepancy. This case highlights the critical need for thorough evaluation, including esophagogastroscopy with monitoring of both IgG- and IgA-type anti-COL7 antibody levels, in EBA patients with dysphagia. The authors have nothing to report. Prof. Daisuke Tsuruta and Prof. Naoki Oiso are Editorial Board members of the Journal of Dermatology and co-authors of this article. To minimize bias, they were excluded from all editorial decision-making related to the acceptance of this article for publication. The other authors declare no conflicts of interest. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.