P.P. Seah

CLEARANCE OF SKIN LESIONS IN DERMATITIS HERPETIFORMIS AFTER GLUTEN WITHDRAWAL

Abstract 24 patients with dermatitis herpeti-formis (D.H.) have been treated with a gluten-free diet (G.F.D.) for periods varying from four months to five years. Of 20 patients who have taken a G.F.D. for a year or longer, 16 (80%) have been able to stop taking dapsone or substantially reduce the dose of the drug. 10 patients have been able to stop taking dapsone completely and are now free of all skin lesions. No patient was able to reduce his dapsone requirements before five months, and 1 patient took twelve months to do this. The length of time taken to be able to stop dapsone completely varied from eight to forty-eight months. In another group of 20 patients with D.H. who did not take a G.F.D. but have been followed up for a similar length of time, there was no significant alteration in dapsone requirements. Reintroduction of gluten in 4 patients who were completely free of skin lesions on a G.F.D. produced irritation and blisters within a week in 3 and within three weeks in the fourth. Dapsone was required for the immediate control of these lesions, which were subsequently controlled again by a G.F.D. alone. Electron-microscopic studies have shown subepidermal membrane-bound vacuoles in clinically normal skin in all 7 patients studied whose eruption was being controlled by dapsone, but these vacuoles were not present in any of the 6 patients whose eruption was controlled by a G.F.D. alone. These results show that the skin lesion in D.H., like the gut lesion, is gluten dependent and that both lesions are part of the same disease process. The length of time for patients to become free of skin lesions after beginning a G.F.D. is stressed, and probably accounts for the previous reports in which it has been stated that the skin lesions are not influenced by a G.F.D.

TISSUE ANTIBODIES IN DERMATITIS HERPETIFORMIS AND ADULT CŒLIAC DISEASE

Abstract A new antibody of IgG class reacting with connective tissue in rat and human organs was found in 17% of 29 patients with dermatitis herpetiformis and in 36% of 31 patients with adult coeliac disease. The antibody appeared to be directed against reticulin rather than basement membrane. In 34% of the dermatitis-herpetiformis patients antinuclear factors of IgG or IgM class were present.

ANTI-RETICULIN ANTIBODIES IN CHILDHOOD CŒLIAC DISEASE

Abstract 19 patients with childhood cœliac disease and 28 controls have been investigated for the presence of anti-reticulin antibody. The antibody was found in 14 of the 19 patients with cœliac disease. 10 of these 19 were on a normal diet, and in 9 of these the antibody was found; whereas in the 9 patients on a gluten-free diet, it was found in 5. The antibody was also found in 1 of 28 control subjects. The anti-reticulin antibody is apparently specific for cœliac disease and may be of help in the diagnosis of the disease.

Dermatitis herpetiformis: an evaluation of diagnostic criteria

Forty‐two patients in whom a clinical diagnosis of dermatitis herpetiformis (DH) had been, made, were studied. All patients had a small intestinal biopsy, a biopsy of uninvolved skin for detection of the presence of IgA deposits by immunofluorescence, serum and red cell folate estimations and examination of the serum for anti‐reticulin antibody. The response of the skin eruption to dapsone was noted. Thirty patients had biopsies of skin lesions for routine histological examination.

IMMUNOGLOBULINS IN THE SKIN IN DERMATITIS HERPETIFORMIS AND CŒLIAC DISEASE

Abstract The unaffected skin of eighteen patients with dermatitis herpetiformis (D.H.), twenty-two patients with cœliac disease (C.D.), and eight controls were examined using direct immunofluorescence and class-specific fluorescein-conjugated anti-human IgA, IgM, and IgG antisera. All eighteen patients with D.H. showed IgA deposits in the skin: in seventeen the deposits were only found in the dermal papillae, whilst in one it was found in a continuous line below the basement membrane, confirmed by immuno-electronmicroscopy. IgM deposits were also found in the dermal papillae in three patients with D.H. and IgG deposits below the basement membrane in one patient. In cœliac disease, however, only one of the twenty-two patients showed papillary IgA deposits and one had continuous IgM deposits. These immunoglobulin deposits in D.H. and C.D. seem to be on the reticulin of the dermal papillae. It is suggested that in D.H. there is a fault of the reticulin in the skin and small intestine, whilst in cœliac disease it is present in the small intestine but not in the skin. The reticulin cross-reacts with gluten complexes to give rise to an immunological reaction. In support of this hypothesis we have demonstrated cross-reactivity between gluten and reticulin.

Immunoglobulins in the skin in dermatitis herpetiformis and their relevance in diagnosis.

Eighty skin biopsies from fifty patients with dermatitis herpetiformis (DH) have been examined for immunoglobulin deposits by direct immunofluorcscence. IgA was found in all fifty patients. However, in two patients no IgA was detected in their first biopsy, and it is stressed that if the clinical suspicion of DH is high and no IgA is found in a single biopsy, then the biopsy should be repeated.

Lymphocytic infiltration of epithelium in diagnosis of gluten-sensitive enteropathy.

The macroscopic appearance of the mucosa of the small intestine and the lymphocytic infiltration of the epithelium were studied in 27 patients with dermatitis herpetiformis and in 11 control subjects. The mucosa was abnormal in appearance in 13 of the patients and normal in 14 patients and in all the controls. In 25 (93%) of the patients the intraepithelial lymphocyte count was significantly raised compared with the controls. The increased lymphocytic infiltration of the epithelium in the patients probably represented an underlying immunological reaction of the small intestine to gluten, since the infiltration lessened in five out of six patients after a year on a gluten-free diet and in all of four patients after three years on a gluten-free diet. Increased lymphocytic infiltration of the epithelium of the small intestine seems a surer sign of gluten sensitivity than the macroscopic appearance of the mucosa, and a diagnosis of gluten-sensitive enteropathy may no longer be excluded when the mucosa appears normal. Further evidence of the significance of increased lymphocytic infiltration is that patients with normal-looking mucosa but with raised intraepithelial lymphocyte counts often had low serum folate levels.

The small intestine in dermatitis herpetiformis

Small intestinal biopsies from 43 patients with dermatitis herpetiformis have been studied. The diagnosis of dermatitis herpetiformis was made on clinical and histological criteria and the presence of IgA deposits in the uninvolved skin. The macroscopic appearance of the intestinal biopsy was flat in 13, convoluted in 10, leaves only in eight, and fingers and leaves in 12. Twenty small intestinal biopsies from patients who did not have dermatitis herpetiformis or gastrointestinal disorder showed leaves only in three and fingers and leaves in 17. The mean total lymphocyte count per 1000 epithelial cells for this control group was 159 ± SE 13; for the dermatitis herpetiformis patients with flat biopsies it was 464 ± SE 27; for the convoluted biopsies 365 ± SE 59; for leaves only 535 ± SE 39; and for the fingers and leaves biopsies 301 ± SE 31. The counts for all four groups are significantly greater than the control group (P < 0.001). Three of the 43 patients with dermatitis herpetiformis had lymphocyte counts below 200 per 1000 epithelial cells, and four of our controls had counts greater than 200 but none above 300. In the control group the mean counts for lymphocytes per 1000 epithelial cells in the basal position of the intestinal epithelium was 79 ± SE 8; in the dermatitis herpetiformis flat biopsies it was 89 ± SE 11; for the convoluted biopsies 93 ± SE 12; for the leaves only biopsies 121 ± SE 18 and for the fingers and leaves 98 ± SE 13. However, the mean lymphocyte count in the perinuclear position was 81 ± SE 7 for the controls, 362 ± SE 23 for the flat dermatitis herpetiformis biopsies; 251 ± SE 46 for the convoluted specimens; 402 ± SE 43 for the leaves only specimens, and 195 ± SE 25 for the fingers and leaves biopsies. The mean lymphocyte count for the supranuclear position was 0·55 ± SE 0·22 for the control group; 13·7 ± SE 2·3 for the flat dermatitis herpetiformis biopsies; 20·0 ± SE 6·9 for the convoluted biopsies; 14·5 ± SE 3·3 for the leaves only biopsies; and 8·3 ± SE 2·6 for the fingers and leaves biopsies. Thus in gluten-sensitive enteropathy the increase in lymphocytes in the intestinal epithelium is in the perinuclear and supranuclear position. The ratio of basal lymphocytes to peri- and supranuclear lymphocytes appears to be 1:1 in normal intestinal epithelium, but approximately 1:4 in the gluten-sensitive enteropathy of dermatitis herpetiformis.

Antireticulin antibody: Incidence and diagnostic significance

Sera from 101 patients with adult coeliac disease, 46 patients with childhood coeliac disease, 50 patients with dermatitis herpetiformis, and 479 patients with various other diseases, including skin, gastrointestinal, haematological, and immunological disorders, have been tested for the presence of the antireticulin antibody. Positive sera were retested at higher dilutions. Antireticulin antibody was only found in a significant proportion of patients with three diseases, ie, coeliac disease, dermatitis herpetiformis, and Crohns disease. Antireticulin antibody was present in 38 out of 101 patients (38%) with adult coeliac disease, 27 out of 46 patients (59%) with childhood coeliac disease, 11 out of 50 patients (22%) with dermatitis herpetiformis, and nine out of 38 patients (24%) with Crohns disease. In the 434 other patients with various disorders the antireticulin antibody was present in only six 1·4%) (two patients were pregnant, one had vitiligo, one had tropical sprue, one had reticulum cell sarcoma, and one had pernicious anaemia). In patients with gluten-sensitive enteropathy, ie, coeliac disease and dermatitis herpetiformis, there was a significantly higher incidence in patients taking a normal diet compared with those on a gluten-free diet. The presence of antireticulin antibody would appear to be particularly helpful in diagnosing childhood coeliac disease as it was found in 22 out of 26 patients (85%) taking a normal diet.

Pemphigus controlled by sulphapyridine

A patient with an II‐year history of a generalized blistering eruption is described. The eruption was completely controlled within I week of taking sulphapyridine and recurred within 4 days of with‐drawal of this drug. On account of these observations a diagnosis of dermatitis herpetiformis was made. Histological, immunological (and small intestinal) studies now show that this patient had pemphigus.