Disappearance of epidermal transglutaminase and IgA deposits from the papillary dermis of patients with dermatitis herpetiformis after a long-term gluten-free diet

Dear Editor, Dermatitis herpetiformis (DH) is an itchy, blistering skin disease characterized by the deposition of granular IgA in the papillary dermis. It is regarded as the cutaneous manifestation of coeliac disease, an autoimmune‐mediated condition affecting the small intestine. In addition to skin symptoms, patients with DH have mostly subclinical small‐intestinal villous atrophy and crypt hyperplasia or at least coeliac‐type inflammatory changes. Moreover, patients with DH have specific antibodies targeting epidermal transglutaminase 3 (TG3), the dominant autoantigen in DH.1 These IgA class TG3 antibodies can be found in the serum of the majority of untreated patients.2 Moreover, the pathognomonic IgA at dermoepidermal junctions also targets TG3 that is localized at this site only in DH.1 The treatment of choice for DH is a lifelong strict gluten‐free diet (GFD), during which the dermal IgA disappears slowly. However, after several years on a GFD, when the rash has healed, IgA is still present in the majority of patients.3,4,5 Whether the disappearance of TG3 from the papillary dermis follows a similar or longer time span is not known. With this in mind, we investigated 33 patients with DH: seven in which the DH was untreated with active rash at the time of diagnosis (untreated group); 21 with DH in clinical remission adhering to a strict GFD (mean duration 22 years, range 5–38) (remission group); and five with active disease following either a strict (n = 3) or partial (n = 2) GFD (mean duration 28 years, range 9–40) (active disease group) for the presence of dermal TG3 and IgA (Table 1). Characteristics, duration of gluten‐free diet (GFD), and cutaneous and serological findings in 33 patients with dermatitis herpetiformis at the time of the study and small‐bowel mucosal histology at diagnosis and at the time of the studya TG3, transglutaminase; F, female; M, male; na, not available; SVA, subtotal villous atrophy; PVA, partial villous atrophy. aSeven patients were untreated (untreated group), 21 were in clinical remission and had no rash on a GFD (remission group) and five were nonresponsive, i.e. had active rash and needed dapsone during long‐lasting GFD treatment (active disease group). Presence of IgA and TG3 in the papillary dermis was examined with single and double stainings and the results interpreted as strong (+++), moderate (++) and weak (+) positive or negative (−). The small‐bowel mucosal findings were graded as normal villous architecture, PVA or SVA. bCut‐off value for serum TG3 antibody positivity > 22 AU mL−1; cdapsone (7·5–25 mg daily) stopped at the time of the skin biopsy; ddapsone (12·5–100 mg daily) needed to control the rash; eGFD not strict. Characteristics, duration of gluten‐free diet (GFD), and cutaneous and serological findings in 33 patients with dermatitis herpetiformis at the time of the study and small‐bowel mucosal histology at diagnosis and at the time of the studya TG3, transglutaminase; F, female; M, male; na, not available; SVA, subtotal villous atrophy; PVA, partial villous atrophy. aSeven patients were untreated (untreated group), 21 were in clinical remission and had no rash on a GFD (remission group) and five were nonresponsive, i.e. had active rash and needed dapsone during long‐lasting GFD treatment (active disease group). Presence of IgA and TG3 in the papillary dermis was examined with single and double stainings and the results interpreted as strong (+++), moderate (++) and weak (+) positive or negative (−). The small‐bowel mucosal findings were graded as normal villous architecture, PVA or SVA. bCut‐off value for serum TG3 antibody positivity > 22 AU mL−1; cdapsone (7·5–25 mg daily) stopped at the time of the skin biopsy; ddapsone (12·5–100 mg daily) needed to control the rash; eGFD not strict. Skin biopsies were taken from perilesional skin or, when there was no rash, routinely from normal‐appearing forearm skin. Sections cut from the biopsies were stained using FITC‐conjugated rabbit polyclonal TG3 antibody (1 : 100) (A030, Zedira, Darmstadt, Germany) or, alternatively, a goat polyclonal TG3 antibody (1 : 25)6 as the primary antibody, followed by Alexa Fluor 488‐conjugated antigoat antibody (1 : 500) (705–546–147, Jackson ImmunoResearch Laboratories, West Grove, PA, U.S.A.) as the secondary antibody. For double stainings, the sections were further incubated with TRITC‐conjugated goat antihuman IgA (1 : 50) (A18786, Thermo Fisher Scientific, Waltham, MA, U.S.A.). IgA class TG3 antibody levels in patient serum were determined using an antihuman‐epidermal‐transglutaminase IgA enzyme‐linked immunosorbent assay kit (Immundiagnostik, Bensheim, Germany). Moreover, data on small‐bowel mucosal morphology at diagnosis and at the time of the study was available for a subset of patients. The study protocol was approved by the ethical committee of Tampere University Hospital, Tampere, Finland. All patients in the untreated and active disease groups with active rash showed strong or moderate IgA deposits in the papillary dermis, which colocalized with TG3 (Table 1, Fig. 1). In contrast, of the 21 patients in the remission group, only eight had IgA and TG3 in the papillary dermis whereas the rest were negative for both. Apart from three individuals, the staining intensities of IgA and TG3 were of the same magnitude (Table 1). Corresponding results were obtained with a noncommercial TG3 antibody6 used previously in addressing TG3 deposition in the skin of patients with DH (data not shown).7 Immunofluorescence staining for transglutaminase 3 (TG3) and IgA in skin sections of patients with dermatitis herpetiformis (DH). Representative pictures of sections from (a) a patient with untreated DH showing strong (+++) staining; (b) a gluten‐free diet (GFD)‐treated patient showing moderate (++) staining and (c) a GFD‐treated patient showing negative (–) staining. TG3 is shown in green, IgA in red and their colocalization in yellow. Scale bar = 100 μm in all panels. The eight patients in remission but with IgA and TG3 deposits had been on the strict GFD diet for a mean of 19 years (range 5–33) and the 13 without any deposits a mean of 22 (range 8–38) years (Table 1). The presence or absence of IgA and TG3 deposits did not thus show any correlation with the duration of GFD treatment (P = 0·078). Serum TG3 antibodies were positive in only two individuals in the active disease group and in two in the remission group (Table 1), and there was no association with the antibody levels and the presence of cutaneous IgA and TG3. Moreover, the absence or presence of IgA and TG3 deposits was not associated with small‐bowel mucosal damage at diagnosis or at the time of the study or dapsone medication. To conclude, we have demonstrated that IgA and TG3 disappear in parallel from the papillary dermis of patients with DH, as seen in a previous study showing that TG3 is always found in conjunction with IgA in active DH.7 The disappearance is particularly slow and occurs only in a subset of patients despite a long and strict GFD. In addition, it appears that the response to GFD treatment in terms of rash and small‐bowel mucosal damage was equally good regardless of the presence of dermal IgA deposits. Altogether, the present findings provide further evidence for the prevailing view that IgA and TG3 exist as immune complexes in the skin of patients with DH.1,8,9 Funding sources: this study was supported by the Academy of Finland, the Finnish Medical Foundation, the Sigrid Jusélius Foundation, the Päivikki and Sakari Sohlberg Foundation, the Foundation for Pediatric Research and the Competitive State Research Financing of the Expert Area of Tampere University Hospital. Conflicts of interest: none declared.