Dermatitis herpetiformis

[2] [4]

A patient has scratched every vesicle away, but the elbows, knees, and buttocks still carry paired clusters of excoriated papules. There is little bowel history. That combination is the clue: the itch and distribution are often more visible than the blisters, and silent gut disease does not argue against DH. [1][2]

DH is the specific cutaneous manifestation of gluten-sensitive coeliac autoimmunity, not a herpes infection. It is a chronic autoimmune subepidermal blistering disease in which IgA complexes containing epidermal transglutaminase 3 (TG3) deposit in the papillary dermis and recruit neutrophils. [2][8]

[2] [3]

About three-quarters of patients have small-bowel villous atrophy and the remainder usually have coeliac-type inflammatory change. Overt gastrointestinal symptoms are uncommon, so the absence of diarrhoea does not exclude the disease. [2]

DH is not a separate entity; it is coeliac disease presenting in the skin. The same gluten-driven immune response, HLA susceptibility, and small-bowel pathology that produce coeliac disease produce DH. The principal difference is the anatomical target: the gut is injured by T-cell-mediated inflammation driven by deamidated gliadin, while the skin is injured by IgA immune complexes that deposit at the dermal papillae.[1][2]

There is no accepted drug-induced or serology-based subtype classification of DH. For exams, describe the phenotype rather than inventing subtypes: classic excoriated extensor DH; atypical acral purpura or urticarial/prurigo-like disease; and DH with negative conventional coeliac serology. The diagnosis still turns on the clinical pattern plus correctly performed DIF. [3][4]

[2] [3]

The hallmark is symmetric, grouped, small papulovesicles on extensor surfaces, and the itch is brutal. The elbows (involved in over 90 percent, often the first site), knees, buttocks and sacrum, scalp, and shoulders bear the brunt; forearms, ankles, and trunk less often. The vesicles are tense, clear or haemorrhagic, a few millimetres across, in clusters.[1][3]

DH is uncommon, usually adult-onset, and varies markedly by population. Reported prevalence in Europe and the United States is 11.2–75.3 per 100,000. Contemporary reviews describe only a slight male excess, with mean onset around 50 years in one large clinical review; fixed “2:1” and “age 20–40” rules are unsafe. [2][3]

[2] [3]

Nearly all affected patients carry HLA-DQ2 or HLA-DQ8; about 90% carry DQ2 and most of the remainder DQ8. These alleles are common in the general population, so a positive result is not diagnostic. Their main value is the high negative predictive value in selected uncertain cases. [2][3][6]

Associated autoimmune disease, particularly autoimmune thyroid disease and type 1 diabetes, is more frequent than in the general population. Screen according to symptoms, history, and the coeliac pathway rather than ordering an indiscriminate autoimmune panel. Adult first-degree relatives can be offered IgA anti-TG2 with total IgA according to current coeliac guidance. [3][7]

The disease begins in the gut and manifests in the skin, and each step is the target of one limb of treatment.[3][4][7]

The safe chain is gluten → gut TG2 response → TG3-directed skin autoimmunity → granular papillary IgA–TG3 complexes → neutrophilic blister. Do not state that anti-TG2 simply “cross-reacts” with TG3 as an established mechanism; how the cutaneous response matures and why complexes localise to papillary dermis are not fully resolved. [2][3]

[2] [3]

The diagnostic test — and the biopsy site that decides it

Think “itch first, vesicles second.” DH causes severe, usually symmetrical pruritus on extensor elbows, knees, buttocks, sacral area, scalp, neck, or upper back. Grouped papules, urticarial plaques, small vesicles, erosions, crusts, and excoriations may coexist; scratching often removes the primary vesicles. Mucosal disease is unusual. [1][3]

The bowel history may be silent or limited to bloating, diarrhoea, constipation, anaemia, weight loss, or nutritional/bone symptoms. Malabsorption complications occur, but are less prominent in DH than in classical symptomatic coeliac disease. [2][3]

[1] [4]

The discriminator line: granular IgA at the dermal papillae plus extensor papulovesicles plus coeliac serology is DH; linear IgA at the BMZ is linear IgA bullous dermatosis; linear IgG/C3 at the BMZ with tense bullae in an older patient is bullous pemphigoid.[7]

DIF pattern separates the closest autoimmune mimics. Clinical morphology narrows the field, but it cannot replace immunofluorescence. [8]

[3] [8]

Dapsone — the rapid bridge, but check G6PD first

Map morphology and distribution, then look for consequences of coeliac disease and scratching. Examine elbows, knees, buttocks, sacrum, scalp, neck, and upper back; document symmetry, intact vesicles, crusting, excoriation, pigment change, and secondary infection. Ask about sleep loss, diet, bowel symptoms, weight change, anaemia, fracture risk, and autoimmune symptoms. [1][3]

There is no validated bedside sign or severity score required for diagnosis. A changing phenotype with large bullae, mucosal involvement, or a non-extensor distribution should reopen the differential rather than be labelled “refractory DH.” [3][8]

• Dapsone given without a G6PD check precipitating severe haemolysis — the textbook preventable error.[8]
• Dapsone agranulocytosis missed because the patient was not warned to report fever or sore throat in the first 12 weeks.[8]
• Biopsying the blister for DIF and getting a false negative, delaying diagnosis for years.[7]
• Missing the coeliac disease — every DH patient needs a duodenal biopsy or validated serology, bone density assessment, and micronutrient screening.[2]
• Forgetting total IgA — selective IgA deficiency falsely lowers IgA coeliac serology.[2]
• Treating with topical steroids alone — symptomatic relief only, no effect on the disease or the coeliac risk.[1]
• A refractory rash blamed on treatment failure when the real cause is ongoing gluten ingestion.[3]

Take the diagnostic sample before a gluten-free diet whenever feasible. A prolonged GFD can reduce skin IgA deposits; pharmacological treatment is not the usual cause of a negative DIF. [3]

Skin biopsy: two sites, two containers

1. DIF: take normal-appearing perilesional skin adjacent to an active lesion. Send it in the laboratory-approved immunofluorescence transport system, such as Michel medium when transport is delayed; do not place it in formalin. Granular IgA in papillary dermis and/or the dermoepidermal junction is the characteristic finding. Reported sensitivity is 90–95%; lesional or inflamed samples have more false negatives. [3][8]
2. Routine histology: if required, take a separate fresh intact vesicle or early erythematous lesion and send it in formalin. Papillary neutrophilic microabscesses and a neutrophil-rich subepidermal blister support DH, but routine histology may be nonspecific and is not independently diagnostic. [3]

If DIF is negative but suspicion remains high, first check the biopsy site, transport, and whether a GFD preceded testing. Repeat DIF from a new normal-appearing perilesional site and reconsider linear IgA disease, bullous pemphigoid, scabies, or another mimic; do not use an empirical dapsone response as proof. [3][8]

Coeliac and DH serology

• IgA anti-TG2 plus total serum IgA is the practical coeliac screen while the patient is eating gluten. If IgA is deficient, use an IgG-based pathway such as IgG anti-TG2 and/or IgG deamidated gliadin peptide according to the laboratory and guideline. [6]
• IgA anti-endomysial antibody is highly specific but operator-dependent; current adult ESsCD guidance does not require it routinely to confirm every positive anti-TG2 result. It remains useful in selected unclear cases. [6]
• IgA anti-TG3 is more specific to DH and may help when available, but assays, availability, and cut-offs are not uniform; DIF remains the diagnostic anchor. [3][4]
• HLA-DQ2/DQ8 is reserved for diagnostic clarification. A negative result strongly argues against DH/coeliac disease; a positive result is common and does not confirm either disease. [3][6]

Is duodenal biopsy mandatory? No.

A correctly diagnosed DH case does not require routine duodenal biopsy solely to prove coeliac disease. Consider gastroenterology-led endoscopy when the skin diagnosis is uncertain, gastrointestinal symptoms are severe or incongruent, serology/histology conflicts, or a complication such as refractory coeliac disease or malignancy is suspected. The degree of villous atrophy does not decide whether a confirmed DH patient needs a lifelong GFD. [1][3][5]

Baseline consequence assessment

Assess nutritional status with history, examination, anthropometry, diet review, and targeted laboratory testing. FBC and iron status are commonly appropriate; folate, B12, calcium, albumin, alkaline phosphatase, vitamin D, liver tests, thyroid testing, and glucose are selected by symptoms, deficits, comorbidity, and the local coeliac pathway. Correct documented deficiencies. [7]

A DXA scan is not compulsory for every low-risk patient at diagnosis. Current adult ESsCD guidance recommends risk-stratified bone assessment: DXA after about one year of GFD for adults with added risks such as delayed diagnosis, severe malabsorption or weight loss, fragility fracture, other osteoporosis risks, or Down syndrome; it may be considered around midlife even without those factors. [7]

Management — Resuscitation & Immediate Control

DH itself rarely needs resuscitation; dapsone toxicity can. Before treatment, secure diagnostic biopsies, start dietetic referral, and assess whether rapid drug control is needed. Urgently assess fever or sore throat (agranulocytosis), jaundice/hepatitis, cyanosis or low pulse oximetry despite oxygen, or a saturation gap (methaemoglobinaemia), and symptomatic anaemia after dapsone. [3][4]

[3]

Management — Definitive & Stepwise

Use the diet to switch off disease and dapsone to bridge the delay. Dapsone relieves itch within hours to a few days and usually stops new blisters within about 24–36 hours; it does not heal enteropathy or reduce lymphoma risk. [2][3]

[2] [3] [7]

Lifelong gluten-free diet

A strict lifelong GFD is first-line for every confirmed DH patient. It treats skin and enteropathy, reduces medication need, and is associated with lower lymphoma risk. Skin improvement begins over weeks to months, but full control and dapsone withdrawal average about two years and can take longer; the old “clear by 6–12 months” promise is too precise. Dietitian support is essential for cross-contamination, nutritional adequacy, cost, and culturally appropriate substitutions. [1][2][3][7]

Dapsone: dose, contraindication screen, and monitoring

For an adult who needs rapid control, one conservative regimen is oral dapsone 25–50 mg once daily, increased gradually toward 100 mg daily only if needed, using the lowest effective dose. Higher doses are specialist decisions because toxicity is dose-related. Avoid routine initiation or obtain specialist oversight in G6PD deficiency and major haematological, cardiac, or pulmonary disease; review interacting medicines and pregnancy/lactation before prescribing. [2][3][4]

Before the first dose, obtain CBC with differential, liver function, renal function, and G6PD activity. A published dermatology protocol checks CBC weekly for month 1, every 2 weeks for the next 8 weeks, then every 3–4 months; it checks liver tests every 2 weeks in month 1 and then every 3–4 months with renal function. Local protocols may differ. Reticulocytes and methaemoglobin are not universal serial tests: request them when haemolysis or methaemoglobinaemia is suspected. [3]

Warn about dose-related haemolysis and methaemoglobinaemia, and rare agranulocytosis, hepatitis, neuropathy, and dapsone hypersensitivity syndrome. Fever, sore throat, rash with systemic illness, jaundice, cyanosis, dyspnoea, or unusual fatigue warrants prompt drug review and testing. [3]

When dapsone is unsuitable

Sulfapyridine 1–2 g/day or sulfasalazine 1–2 g/day can be considered by a specialist, with drug-specific blood and urine monitoring. Potent topical corticosteroids are adjuncts for local itch, not substitutes for GFD or systemic control. Tetracycline/nicotinamide, azathioprine, ciclosporin, methotrexate, colchicine, mycophenolate, and rituximab are supported mainly by case reports or small series; they are not routine “next steps.” [3][4]

Specific Subtypes & Scenarios

Atypical morphology changes the index of suspicion, not the diagnostic standard. Children may show acral purpura or excoriated papules; older adults may present through anaemia or bone disease rather than diarrhoea. Both still need properly sited DIF and age-appropriate coeliac assessment. [3][4]

In pregnancy or lactation, maintain a nutritionally adequate GFD and involve dermatology, obstetric medicine, and pharmacy before systemic treatment. Do not teach blanket dapsone safety: maternal disease, alternatives, dose, maternal haemolysis/metHb risk, fetal or neonatal risk, and feeding plans require individual assessment. [4][7]

Iodide exposure and indomethacin have been reported to exacerbate existing DH. That does not establish a separate drug-induced, non-coeliac DH subtype or justify blaming all NSAIDs. [2]

Complications & Pitfalls

The common preventable harms are a bad biopsy, an unsafe dapsone start, and an unsupported surveillance programme. [3][7]

• A lesional or formalin-fixed DIF sample can miss the diagnostic IgA pattern. [3][8]
• Normal G6PD activity does not exclude dapsone haemolysis or methaemoglobinaemia. [3]
• Routine methaemoglobin/reticulocyte testing for every stable patient is not the same as symptom-triggered testing. [3]
• Confirmed DH does not automatically mandate duodenal biopsy, baseline DXA, repeat biopsy, or annual broad micronutrient panels. Use clinical and risk-based indications. [3][7]
• “Refractory” rash most often demands review of diagnosis, gluten exposure, and adherence before immunosuppression. [3][7]

Lymphoma risk is increased early after DH diagnosis in observational series, but the absolute risk remains low and is not limited to EATL. Long-term GFD lowers risk. Current coeliac guidance does not recommend routine malignancy screening beyond population programmes without clinical suspicion; investigate alarm symptoms such as persistent pain, weight loss, fever/night sweats, obstruction, or lymphadenopathy. [1][3][7]

Prognosis & Disposition

With a strict GFD, prognosis and quality of life are usually excellent. Most patients are managed as outpatients with dermatology, dietetic, and coeliac-care input. Admit or urgently assess only for serious treatment toxicity, severe infection, major nutritional compromise, or concerning gastrointestinal complications. [2][3]

Do not promise a 6–12-month finish. Cutaneous remission and dapsone withdrawal average about two years on a strict GFD, with wide individual variation. Persistent lesions require an adherence and diagnosis review; persistent gastrointestinal symptoms need a separate coeliac work-up. [2][3][7]

Special Populations

[3] [4] [7]

Immunocompromised patients or those already receiving immunosuppression may have altered morphology and competing diagnoses. Do not infer “refractory DH” from treatment failure; obtain specialist review and re-check immunopathology. [8]

Evidence, Guidelines & Regional Differences

The disease is global; the testing and follow-up pathways are local. EADV provides a DH-specific European consensus, the Italian societies cover coeliac disease and DH across ages, and the updated ESsCD documents provide current adult coeliac diagnosis and risk-stratified follow-up. These sources agree on correctly sited DIF, lifelong GFD, short-term symptom control when needed, and multidisciplinary care. [4][5][6][7]

[3] [6] [7] [3] [5]

Exam Pearls

[2] [3] [7] [3] [4]

Self-test: Does confirmed DH require routine duodenal biopsy and baseline DXA?

No. A correctly confirmed DH diagnosis does not require duodenal biopsy solely to prove coeliac disease. Endoscopy is reserved for uncertainty, severe/incongruent symptoms, or suspected complications. DXA is risk-stratified rather than universal in low-risk adults.

[1] [3] [7]

Self-test: What does normal G6PD testing change?

It reduces concern for the extreme oxidant risk of G6PD deficiency; it does not make dapsone harmless. Dose-related haemolysis and methaemoglobinaemia can still occur, so use the lowest effective dose, monitor CBC/liver/renal function, and investigate symptoms with reticulocytes or methaemoglobin as indicated.

[3]

References

1. [1]Salmi TT. Dermatitis herpetiformis Clin Exp Dermatol, 2019.PMID 31093998
2. [2]Reunala T, Hervonen K, Salmi T. Dermatitis Herpetiformis: An Update on Diagnosis and Management Am J Clin Dermatol, 2021.PMID 33432477
3. [3]Nguyen CN, Kim SJ. Dermatitis Herpetiformis: An Update on Diagnosis, Disease Monitoring, and Management Medicina (Kaunas), 2021.PMID 34441049
4. [4]Görög A, Antiga E, Caproni M, et al. S2k guidelines (consensus statement) for diagnosis and therapy of dermatitis herpetiformis initiated by the European Academy of Dermatology and Venereology (EADV) J Eur Acad Dermatol Venereol, 2021.PMID 34004067
5. [5]Zingone F, Maimaris S, Auricchio R, et al. Guidelines of the Italian societies of gastroenterology on the diagnosis and management of coeliac disease and dermatitis herpetiformis Dig Liver Dis, 2022.PMID 35858884
6. [6]Al-Toma A, Zingone F, Branchi F, et al. European Society for the Study of Coeliac Disease 2025 Updated Guidelines on the Diagnosis and Management of Coeliac Disease in Adults. Part 1: Diagnostic Approach United European Gastroenterol J, 2025.PMID 40999951
7. [7]Al-Toma A, Branchi F, Zingone F, et al. European Society for the Study of Coeliac Disease (ESsCD) 2025 Updated Guidelines on the Diagnosis and Management of Coeliac Disease in Adults. Part 2: Management, Follow-Up, and Complex Disease Courses United European Gastroenterol J, 2026.PMID 41831197
8. [8]van Beek N, Holtsche MM, Atefi I, et al. State-of-the-art diagnosis of autoimmune blistering diseases Front Immunol, 2024.PMID 38903493