Are You Confident of the Diagnosis?

What you should be alert for in the history

Patients with dermatitis herpetiformis (DH) usually complain of weeks to months of intense pruritus that has a stinging or burning quality. Only about 30% to 40% of patients admit to gastrointestinal symptoms of celiac disease. When patients do complain of abdominal symptoms, the most common symptoms are bloating and crampy abdominal pain. Diarrhea is a less common symptom and patients may even have constipation. Approximately 10% to 20% of DH patients will have a family history of celiac disease or DH; 10% to 20% of patients with DH will have a history of a thyroid disorder including hypothyroidism or hyperthyroidism.

A history of other autoimmune diseases including diabetes mellitus, Addison’s disease, and vitiligo are less common associations.

Characteristic findings on physical examination

The physical findings of DH vary in severity. Milder cases will have only grouped excoriated papules on the knees or extensor forearms (Figure 1). Other common areas of involvement include the buttocks, scalp, beltline and eyelids. In more severe cases papulovesicles will be seen in a similar distribution and may diffusely involve the torso and extremities. Visible blisters are the exception rather than the rule. Such vesicles and bullae are especially severe when suppressive therapy with dapsone is abruptly discontinued (Figure 2). Rare patients will have bloated abdomen and diffuse abdominal tenderness.

Continue Reading

Figure 1.

DH grouped papulovesicles on the knees.

Figure 2.

Vesiculo-bullous lesions of DH after stopping Dapsone

Expected results of diagnostic studies

The diagnosis is best made on biopsy. Histopathology reveals a neutrophilic infiltrate in the papillary dermis with an associated lymphohistiocytic perivascular infiltrate. More advanced lesions will demonstrate infiltration of the dermal papillae with neutrophils, dermal papillary edema and basement membrane zone separation (Figure 3).

Figure 3.

Histology of a DH lesion demonstrating neutrophil infiltration in dermal papillae and dermal papillary edema

Frequently the intense pruritus results in excoriation and subsequent destruction of the characteristic histopathologic findings. In such cases the biopsy will reveal only excoriations or what appear to be eczematous findings with minimal spongiosis. Direct immunofluorescence is essential for definitive diagnosis. The characteristic deposition of granular IgA in dermal papillae is seen in virtually all cases when clinically normal-appearing skin immediately adjacent to a lesion is biopsied and processed for direct immunofluorescence (Figure 4); 60% to 70% of patients will demonstrate IgA tissue transglutaminase antibodies in their serum. This is a reflection of the underlying celiac disease.

Figure 4.

Even patients who have negative IgA tissue transglutaminase antibody levels may still have significant histologic celiac disease associated with their DH. Virtually all patients with DH have some degree of histologic celiac disease and will respond to dietary gluten restriction as noted below. Recently it has been shown that the IgA antibodies deposited in dermal papillae are the result of circulating IgA antibodies to epidermal transglutaminase. IgA epidermal transglutaminase antibodies can be measured in serum and are present in greater than 90% of DH patients on a gluten-containing diet; 99% of DH and celiac disease patients will have either the DQ2 or DQ8 haplotype compared with 20% of controls.

HLA typing alone is of little diagnostic value since only 5% of patients carrying the DQ2 or DQ8 haplotype have either celiac disease or DH. Perhaps the main use of HLA typing in celiac disease or DH is to rule out these disorders. If patients do not have the DQ2 or DQ8 haplotype, it is extremely unlikely that they have either DH or celiac disease.

Diagnosis confirmation

The histologic findings of DH may be mimicked by other neutrophilic dermatoses including linear IgA bullous dermatosis and bullous lupus erythematosus. Linear IgA bullous dermatosis may be characterized by diffuse vesicles and bullae, which clinically closely resemble bullous pemphigoid. However, at least one third of the patients with linear IgA bullous dermatosis have clinical findings similar to that of DH.

Bullous lupus erythematosus is usually characterized by scattered rather than grouped bullae and lesions are not distributed on extensor surfaces as is the case in DH. Bullous lupus erythematosus patients also usually demonstrate associated findings of systemic lupus erythematosus. Patients with linear IgA bullous dermatosis and bullous lupus erythematosus do not respond to dietary gluten restriction and do not have associated celiac disease. Therefore it is essential that confirmation of the diagnosis of DH be made by direct immunofluorescence.

Who is at Risk for Developing this Disease?

There is a male predominance in DH patients of 1.5:1. This is in contrast to celiac disease alone, which is more common in females. DH may have onset from early childhood to late adult life, but the median age of onset is in the fourth decade. There is a hereditary predisposition characterized by specific HLA genotypes, DQ2 and DQ8. These genotypes occur in only about 20% of Caucasian controls. They are much less frequent in Asians and those of African descent. For that reason DH is much less common in people of Asian and African background.

The prevalence of DH in the Caucasian population has been reported to be as high as 75 per 100,000 in Finland to 11 per 100,000 in the state of Utah in the United States. The annual incidence is approximately 1 to 4 per 100,000/year.

What is the Cause of the Disease?

DH is a cutaneous manifestations of celiac disease. Virtually all patients are sensitive to gluten and will improve with long-term adherence to a strict gluten-free diet. Intestinal biopsies in DH patients span the entire spectrum of histopathologic abnormality in celiac disease. Some cases demonstrate only minimal intraepithelial lymphocytes, while others demonstrate severe villus atrophy. Approximately 10% to 15% of the cases demonstrate no significant histopathologic abnormality on small intestinal biopsy, but still respond to gluten restriction. It seems likely that these patients have extremely low-grade celiac disease that is not readily detected on random intestinal biopsies.

The pathogenesis of DH is closely related to that of celiac disease. The antigen-presenting cells in the intestinal mucosa present alpha gliadin antigen to T cells in the context of the characteristic HLA genotype. This presentation is accentuated by the deamidation of gliadin by the enzyme tissue transglutaminase (TG2). Activation of T cells and a cellular immune response to gliadin ensues. This cellular immune response then releases cytokines that inflame the intestinal epithelium and are responsible for the villus atrophy.

For uncertain reasons, celiac disease patients also develop IgA antibodies to tissue transglutaminase (TG2). These antibodies have come to represent the most reliable serologic test used for the diagnosis of celiac disease. Levels of IgA TG2 antibodies correlate roughly with the severity of the intestinal inflammatory process. DH patients have IgA tissue transglutaminase antibodies in 70% to 80% of the cases. Patients with less severe intestinal inflammation are likely not to have these antibodies. That does not mean that these patients do not have celiac disease. Some of those patients may have significant villus atrophy, while others may have no detectable intestinal inflammation, but will nonetheless respond to dietary gluten restriction.

The granular IgA antibodies in the skin of DH patients represent IgA antibodies to epidermal transglutaminase (TG3). These antibodies occur a longer time after the IgA tissue transglutaminase immune response. This likely is the reason that DH has a later onset in life than celiac disease, which frequently has onset in childhood. It seems likely that epitope spreading from tissue transglutaminase to epidermal transglutaminase dictates the eventual development of DH in approximately 10% to 20% of celiac patients.

Once the IgA epidermal transglutaminase antibodies reach the skin, they bind to soluble epidermal transglutaminase antigen that is produced by keratinocytes. These immune complexes are then cross-linked (likely involving the cross-linking function of epidermal transglutaminase) and firmly bound to connective tissue structures in the papillary dermis. It then seems likely that trauma stimulates the inflammatory release of epidermal cytokines and enzymes that trigger the inflammatory process.

In addition, cytokines produced by the inflammatory process in the intestine, such as interleukin (IL)-8, provide a systemic inflammatory millieu with neutrophil activation that promotes the local inflammatory process and infiltration of dermal papillae with neutrophils. These neutrophils then release proteases that produce the cutaneous inflammation and the eventual separation of the basement membrane zone secondary to enzymatic degradation of the lamina lucida with resultant blister formation.

Systemic Implications and Complications

Since virtually all DH patients have some degree of celiac disease, they are subject to the complications and associations that one would expect with celiac disease. Only a minority of DH patients have intestinal symptoms in spite of the significant number who have fairly severe intestinal inflammation. As serologic testing has become more available and accurate for celiac disease, we have found that many patients with histopathologic celiac disease have relatively minor intestinal symptoms and systemic manifestations. Intestinal symptoms may not be severe and may be limited to crampy abdominal pain and constipation.

Although diarrhea and thin stature are the traditional hallmarks of celiac disease, they are seldom seen in DH patients, and it is now recognized that the vast majority of celiac patients do not present with this extreme manifestation of malnutrition. IgA tissue transglutaminase antibody should be measured in all DH patients. If positive, these antibody levels can be followed as an index of response to gluten restriction and adherence to a gluten-free diet. Virtually all DH patients have some degree of celiac disease and respond to dietary gluten restriction; it is not necessary to perform small-intestinal biopsies on DH patients as a routine. Patients who need to be convinced that they have celiac disease may be helped by the intestinal biopsy process.

It may also be helpful to perform small-intestinal biopsy if the patient is refractory to dietary gluten restriction. When patients do not respond to dietary gluten restriction, it is likely that they have not fully eliminated gluten from their diet. Small-intestinal biopsy may help to confirm this. Both celiac disease and DH are associated with autoimmune thyroiditis and subsequent hypothyroidism. This is more common in female patients but can occur in both genders. Routine evaluation for thyroid dysfunction including a TSH and thyroxine levels is recommended, especially if patients have symptoms suggestive of thyroid insufficiency.

Pernicious anemia, Addison’s disease, and vitiligo occur much less frequently and only need to be evaluated if clinical symptoms suggest those disorders or cutaneous manifestations of vitiligo become obvious.

All celiac disease patients, including those who present with DH, have an increased risk of iron deficiency anemia secondary to malabsorption of iron. Evaluation for iron deficiency with serum iron, iron-binding capacity and ferritin levels is recommended, especially if red blood cell indices are noted to be microcytic on complete blood count.

Celiac disease is associated with an increased risk of lymphoma and osteoporosis (presumably related to malabsorption of calcium and vitamin D). Because DH patients have celiac disease, one would also expect an increased risk of lymphoma and osteoporosis. Initial studies in DH patients suggested a prevalence of lymphoma increased at 1%.

Recently another study has suggested the prevalence of lymphoma is not increased in DH and that there is no increased risk of fracture as would be suspected secondary to an increased prevalence of osteoporosis. At this time it seems reasonable to assume that the risk of those complications is quite low, but may exist if larger cohorts were studied. It is agreed that any possible increase in the incidence of lymphoma would almost certainly decrease to the baseline incidence of a not-at-risk population, if the patient adheres to a strict gluten-free diet.

Treatment Options

Gluten-free diet. Elimination of gluten from the diet will eventually produce clearing of the skin disease in DH. This is likely to take several months and requires strict adherence to the gluten-free diet. It has been the authors experience that patients who have had symptoms for short periods of time (weeks to months) respond quite rapidly to dietary gluten restriction. However, patients who have been maintained on suppressive therapy with dapsone for long periods of time respond very slowly to gluten restriction (possibly years to response).

Long-term adherence to a gluten-free diet treats the underlying cause of the skin disease and will eventually allow elimination of suppressive therapy with dapsone, sulfapyridine, or other suppressive drugs. Adherence to gluten-free diet will also reduce the possibility of long-term complications from the underlying celiac disease, such as lymphoma and osteoporosis. Consultation with a dietitian familiar with gluten restriction is recommended. There are also several patient groups listed on the Internet with extensive information on gluten-free diet therapy.

Deviation from the diet will frequently produce flares of DH. If the diet is abandoned, most cases will eventually recur within weeks to months. Some cases will have a long-term remission of DH in spite of resuming a normal diet. The reason for these remissions is unclear.

Dapsone. Dapsone will produce relief of the intense pruritus within 48 to 72 hours. However, symptoms and signs recur within 48 to 72 hours if dapsone is discontinued. Initial therapy with 25mg daily is recommended. The dose can then be gradually increased 25mg every week until adequate control is obtained. After complete control is obtained and the patient is adhering to a gluten-free diet, the patient can then attempt to decrease the dose gradually on a weekly basis as disease severity is mitigated by dietary gluten restriction. This should eventually allow complete discontinuance of dapsone therapy.

Before beginning dapsone therapy, glucose-6-phosphate dehydrogenase levels should be checked in blacks and those of southern Mediterranean origin, since deficiency of this enzyme may be associated with catastrophic hemolysis in patients treated with dapsone.

Sulfapyridine. Sulfapyridine can be used in place of dapsone in some patients. It can be used in the same way as suppressive therapy while awaiting the effect of dietary gluten restriction. The usual starting dose is 500mg 3 times a day with increase to 1000mg 3 times a day if needed. At the time of response to gluten restriction, the dose can be decreased on a weekly basis to control the cutaneous symptoms. Sulfapyridine is not commercially available in the United States at this time but can be formulated by compounding pharmacies. For uncertain reasons some patients do not respond to sulfapyridine in any dose.

Colchicine. Colchicine has been helpful in some cases when patients are unable to take dapsone or sulfapyridine. Initial dosing of 0.3mg twice a day is recommended with increase to 0.6mg twice a day if necessary. Limiting factors in treatment include the onset of diarrhea or leukopenia.

Tetracycline and niacinamide. Tetracycline 500mg 3 times a day and niacinamide 500mg 3 times a day has been described as helpful in some cases in which dapsone or sulfapyridine was not tolerated.

Elemental diet. DH improves dramatically in 2 to 3 weeks if the patient restricts the diet to amino acids alone or small chain polypeptides. The mechanism for this is unclear but this method can be used for severe and refractory disease. The patient can then begin a gluten-free diet for long-term control of the disease.

Topical therapy. Occasional new lesions are not an indication for increase in suppressive therapy. Such lesions can be treated effectively with high potency topical corticosteroid gel. Extensive cutaneous disease does not respond well to topical therapy of any type. Although it makes some sense from a theoretical point of view, no successful topical therapy with dapsone has been reported.

Heparin. Patients with severe disease who are unable to tolerate dapsone therapy will respond to systemic anti-coagulation therapy with heparin. This will relieve the symptoms but will require long-term therapy or substitution of suppressive therapy described above. The coagulation system has been implicated in the pathogenesis of local lesions of DH. Serum devoid of coagulation factors will produce lesions when injected locally in DH patients. Anticoagulated blood does not have the same effect. Consequently treatment of refractory cases with heparin has been described and may be of use in extreme situations.

Optimal Therapeutic Approach for this Disease

The optimal therapeutic approach is to first make absolutely certain that the diagnosis is correct by confirming the diagnosis by direct immunofluorescence. The most common cause of poor response to therapy as noted above is incorrect diagnosis. A strict gluten-free diet should be recommended. If symptoms are bothersome, dapsone 25mg daily can be started. The dose is then increased by 25mg every week until complete control is obtained.

When DH is completely controlled for several months and the patient is adhering to a gluten-free diet, they may then attempt to decrease the dapsone dose by 12 1/2 to 25mg each week until an optimal dose is achieved. Continuing attempts to decrease the dose will eventually result in complete discontinuance of dapsone therapy. Topical steroid therapy in the form of high potency steroid gel can then be used to treat individual lesions. Occasional (2 to 3/week) development of new lesions is not an indication for increasing dapsone dose.

Eventual reintroduction of gluten into the diet Is a questionable strategy that has been recommended by some. Dapsone Is far and away the best medication at suppressing cutaneous lesions. Although it can not be stated definitively, most authorities believe that dapsone can be used if the patient is “allergic” to sulfa. If dapsone is not tolerated, sulfapyridine can be tried. Patients who are allergic to sulfa will react to sulfapyridine.

Some patients may not respond to sulfapyridine in any dose. In that situation, other therapies noted above can be tried. In desperate situations where the patient is extremely uncomfortable and unable to tolerate dapsone or sulfapyridine, treatment with an elemental diet or heparin can be undertaken to control symptoms. Elemental diet has the advantage of healing the underlying celiac disease and therefore allowing remission.

Although tetracycline, niacinamide and colchicine have been described as effective, the author has not found them to be effective.

Side effects of dapsone may limit its use. If leukopenia occurs, it is best to stop dapsone and try alternative therapy. All patients will develop some degree of hemolytic anemia and so this is not to be considered a reason for stopping therapy. Compensated hemolytic anemia usually results and methemoglobinemia also occurs in all patients to some degree. It is not an indication for changing dapsone dose or treating with methylene blue unless cardio-respiratory compromise has become a limiting factor.

Peripheral neuropathy from dapsone is usually dose-dependent and may not be reversible. For that reason, strict adherence to a gluten-free diet in such cases is strongly recommended to avoid dapsone therapy and prevent progressive peripheral neuropathy. Some patients will claim that they do not respond to a gluten-free diet. In virtually all of those cases a response to elemental diet occurs. This makes it likely that the patient is not adhering well to gluten restriction.

Patient Management

Patients with DH should be started on a gluten-free diet. While awaiting a response to gluten restriction, dapsone can be initiated to allow control of the intense itching. While on dapsone patients should have a complete blood count every 1 to 2 weeks for the first 3 months and then every 6 months thereafter. Leukopenia, when it occurs, usually comes in the first 3 months. Hemolytic anemia occurs in virtually all patients to some degree and should be followed to make certain that it is not excessive when chronic dapsone therapy is used. Compensated hemolytic anemia will produce reticulocytosis, which will be seen in the CBC as macrocytosis (elevated mean corpuscular volume), since reticulocytes are larger than mature red cells.

Liver function tests should be monitored every 6 to 12 months. Minimal abnormalities of hepatic enzymes may occur at any dose but more severe elevations of hepatic enzymes occur at high doses (2 to 3mg per kilogram body weight). Monitoring symptoms of peripheral neuropathy is also recommended in patients on chronic high-dose therapy. It is beneficial to try to minimize the dapsone dose using concomitant gluten restriction to minimize the risk of peripheral neuropathy that may not be reversible even with subsequent discontinuance of dapsone therapy.

DH patients who choose to be on suppressive therapy alone are likely to have a lifelong commitment to either dapsone or sulfapyridine. Patients who choose a strict gluten-free diet will eventually be able to discontinue suppressive therapy. Reinstitution of a regular diet will eventually result in recurrence of the DH symptoms, although this may take months to years. A small percentage (12% in one study) of patients will be able to discontinue all therapy and maintain a long-term remission. The reason for this is totally unknown. Whether those patients will have an increased risk of lymphoma or osteoporosis is unclear.

Unusual Clinical Scenarios to Consider in Patient Management

The most common clinical issue in DH patients is whether or not to adhere to a gluten-free diet. Patients quickly recognize that dapsone will completely control their skin disease. Many patients are able to control their disease completely on 25 to 50 mg of dapsone daily with minimal side effects. In addition, patients may have no intestinal symptoms of celiac disease and therefore not see the benefit of gluten restriction. This choice not only subjects the patients to the long-term side effects of dapsone but also the complications of untreated subclinical celiac disease, which may include an increased risk of lymphoma, osteoporosis, and associated immunologic disorders such as diabetes and hypothyroidism.

Many dermatologists in the United States have recommended to their patients to simply take dapsone. They have claimed that the gluten-free diet therapy is too difficult. This decision should be left to the patient. There is now much greater availability of gluten-free foods and many patients prefer to treat their underlying disorder rather than treating only the symptoms as is being done with dapsone. It is the responsibility of the physician to explain the therapeutic options to the patient and then allow the patient to participate in the decision-making process.

What is the Evidence?

Alonso-Llamazares, J, Gibson, LE, Rogers, RS. “Clinical, pathologic,and immunopathologic features of dermatitis herpetiformis: review of the Mayo Clinic experience”. Int J Dermatol. vol. 46. 2007. pp. 910-19. (This review of 264 patients diagnosed over 26 years shows a 12% incidence of symptomatic celiac disease, attesting to the fact that most patients are without symptoms. Lymphoproliferative disorders were found in seven patients (2.6%). Interestingly, ulcerative colitis was present in 6/264. This is the first report to suggest an association of DH with ulcerative colitis. This study was done at a referral center and may not reflect the true incidence of associated disorders in a general population of DH patients.)

Collin, p, Huhtala, H, Virta, L, Kekkonen, L, Reunala, T. “Diagnosis of celiac disease in clinical practice: physician’s alertness to the condition is essential”. J Clin Gastroenterol. vol. 41. 2007. pp. 152-6. (This documents the incidence and prevalence of both dermatitis herpetiformis and celiac disease in a predominantly Caucasian northern European population.)

Hall, RP, Benbenisty, KM, Mickle, C, Takeuchi, F, Streilein, RD. “Serum IL-8 in patients with dermatitis herpetiformis is produced in response to dietary gluten”. J invest Dermatol. vol. 127. 2007. pp. 2158-65. (This work indicates that the inflammatory response produced by gluten in the intestine likely plays an important role in the eventual pathogenesis of the skin disease.)

Lewis, NR, Logan, RF, Hubbard, RB, West, J. “No increase risk of fracture, malignancy or mortality in dermatitis herpetiformis: a cohort study”. Aliment Pharmacol Ther. vol. 27. 2008. pp. 1140-7. (Previous studies have suggested that patients with DH have an increased risk of lymphoma as do patients who have celiac disease alone. This study failed to confirm this important association and also indicates that there is not an increased risk of fracture in DH patients as might be expected if DH patients had osteoporosis at the same rate as their celiac disease counterparts.)

Paek, SY, Steinberg, SM, Katz, SI. “Remission in dermatitis herpetiformis. A cohort study”. Arch Dermatol. vol. 147. 2011. pp. 302-5. (This confirms that a small percentage (12% in this study) of DH patients go into spontaneous remission over time. It brings up the question of whether or not such patients should no longer adhere to a gluten-free diet after suppressive therapy with dapsone or sulfapyridine has been stopped.)

Rose, C, Rocker, EB, Zillikens, D. “Clinical histological and immunopathological findings in 32 patients with dermatitis herpetiformis Duhring”. J Dtsch Dermatol Ges. vol. 8. 2010. pp. 265-70. (This documents the presence of intestinal abnormalities in patients with DH and confirms the presence of serologic markers for both IgA tissue and epidermal transglutaminase in a consecutively diagnosed population.)

Zone, JJ, Meyer, LJ, Petersen, MJ. “Deposition of granular IgA relative to clinical lesions in dermatitis herpetiformis”. Arch Dermatol. vol. 132. 1996. pp. 912-8. (This provides definitive evidence on the best location for biopsy for direct immunofluorescence in diagnosing dermatitis herpetiformis. It also demonstrates that virtually all DH patients can be diagnosed by direct immunofluorescence if the proper site for biopsy is chosen.)

Zone, JJ. “Skin manifestations of celiac disease”. Gastroenterology. vol. 128. 2005. pp. s87-s91. (This reviews the association of celiac disease and DH. It reviews the spectrum of celiac disease that may be associated with skin disease.)

Zone, JJ, Schmidt, LA, Taylor, TB, Hull, CM, Sotiriou, MC, Jaskowski, TD. “Dermatitis herpetiformis sera or goat anti-transglutaminase 3 transferred to human skin-grafted mice mimics dermatitis herpetiformis immunopathology”. J Immunol. vol. 186. 2011. pp. 4474-80. (This work is the definitive study on pathogenesis to show that the IgA deposition in DH is caused by IgA epidermal transglutaminase antibodies.)