Dermatitis herpetiformis Duhring – evaluation of disease severity and tissue transglutaminase levels in 122 patients on dapsone therapy

ABSTRACT

The aim of our study was to characterize the group of patients suffering from dermatitis herpetiformis Duhring (DH) and to make the statistical analysis of monitored parameters. We retrospectively analysed the data of 122 patients with DH (age, gender, the severity of DH, tissue transglutaminase antibodies IgA (IgA-TGA) level and dapsone dose). The mild form of DH was recorded in 69 patients (56.6%), the moderate form in 48 patients (39.3%) and the severe form in 5 patients (4.1%); 118 patients (96.7%) are on dapsone therapy, 3.2% of patients are only on gluten-free diet. The median of average IgA-TGA in the mild form is 6.25 U/ml, in the moderate form 14.45 U/ml and in the severe form 10.33 U/ml. Although the significant relation of the IgA-TGA to the disease severity was not confirmed, there is some trend in the increase of the IgA-TGA level from mild to moderate and severe form of DH.

KEYWORDS:

• dermatitis herpetiforms
• dapsone
• tissue transglutaminase IgA antibodies (IgA-TGA)
• disease severity
• celiac disease
• gluten-free diet

Introduction

Dermatitis herpetiformis Duhring (DH) is an inflammatory disease of the skin, considered as the specific cutaneous manifestation of celiac disease (CD). Both DH and CD occur in gluten-sensitive individuals, share the same Human Leucocyte Antigen (HLA) haplotypes (DQ2 and DQ8), and improve following the administration of a gluten-free diet. Moreover, almost all DH patients show typical CD alterations at the small bowel biopsy, ranging from villous atrophy to augmented presence of intraepithelial lymphocytes, as well as the generation of circulating autoantibodies against tissue transglutaminase. Detecting granular IgA deposits at the dermal–epidermal junction by direct immunofluorescence from perilesional skin represents the most specific diagnostic tool. Furthermore, assessing serum titres of autoantibodies against epidermal transglutaminase, the supposed autoantigen of DH may also serve as a clue for the diagnosis (Antiga et al., 2019). Due to the underlying latent celiac disease in the vast majority of the patients, some gastroenterologists and dermatologists call it the celiac disease of the skin (Antiga et al., 2019; Kárpáti, 2004). Wheat is one of the major cereals consumed throughout the world and there has been a radical increase in the population suffering from many wheat-related disorders. On the other hand, DH is a rare disease that occurs prevalently in Caucasian individuals (Alonso-Llamazares & Gibson, 2007; Graziano & Rossi, 2018; Zhang et al., 2012). In the Europe and the USA, the prevalence of DH ranges from 11.2 to 75.3 per 100,000, with the highest reported in Finland, whereas the incidence ranges from 0.4 to 2.6 per 100,000 people per year (Graziano & Rossi, 2018). Etiopathogenesis has an immunological cause but is not fully understood. All recent data related to the pathophysiology and to the clinical features of DH indicate that DH is a gastrointestinal gluten-induced IgA-TG3 immune complex-mediated disease (Alonso-Llamazares & Gibson, 2007; Kárpáti, 2004). The genetic predisposition to the development of gluten sensitivity underlies the disease. It is known that there is a higher incidence of genotypes HLA DR3, HLA DQw2 in 80–90% of patients, HLA B8 and HLA DQ8 in 10–20% of cases, as well as adult celiac disease (Marietta et al., 2004; Zhang et al., 2012). Cutaneous lesions begin with itching or a burning sensation with development in erythematous papules and urticarial plaques. There are grouped vesicles and tense blisters with centrifugal growth, whose contents may be serous or haemorrhagic, with symmetrical distribution (Alonso-Llamazares & Gibson, 2007; Korman, 2001; Marietta et al., 2004; Salavec & Boštíková, 2017). The treatment is successful in patients on a gluten-free diet and in patients who tolerate dapsone (Korman, 2001; Salavec & Boštíková, 2017). Dapsone (4,4'-diaminodiphenylsulfone) was synthesized a century ago and continues to be a powerful therapeutic tool in many skin diseases (Korman, 2001; Salavec & Boštíková, 2017). The course of DH is characteristic with periods of remission and exacerbation. An emotional event or infection may trigger a disease worsening (Kárpáti, 2004; Marietta et al., 2004). Patients with DH usually show the specific antibodies that can be found in patients with celiac disease (Alonso-Llamazares & Gibson, 2007; Marietta et al., 2004; Sárdy et al., 2002; Zhang et al., 2012). Among them, tissue transglutaminase antibodies IgA (IgA-TGA) are considered the most sensitive and specific ones and should be tested as the first-line serologic investigation in patients with a suspected DH (Alonso-Llamazares & Gibson, 2007; Marietta et al., 2004; Sárdy et al., 2002; Zhang et al., 2012). The IgA-TGA level measuring is suitable for the evaluation of the status of jejunum mucous membrane and patient adherence (Alonso-Llamazares & Gibson, 2007; Marietta et al., 2004; Sárdy et al., 2002; Zhang et al., 2012). Contrary to the other bullous disorders, DH patients have no circulating autoantibodies binding to the cutaneous basement membrane components or to other adherent structures of the skin, but they have gluten-induced IgA autoantibodies against transglutaminase TG 2 and TG3 also called tissue-TG (t-TG) and epidermal-TG (e-TG), respectively (Aeschlimann & Thomazy, 2000). Over time, different authors demonstrated that the enzymatic activity of TG2 is implicated in several diseases as Huntington disease, Alzheimer disease and celiac disease (Nemes et al., 2001). Levels of anti-TG2 correlate with bowel damage and gluten-free diet adherence in DH/CD patients and are measured using an enzyme-linked immunosorbent assay (ELISA) (Hornbeck, 2015). Sárdy et al. (Marietta et al., 2004) demonstrated, the target autoantigen of DH is represented by TG3 that shares a 64% of homology with TG2 and overlapping sensitivity and specificity (Jaskowski et al., 2009). In addition to anti-TG2 antibodies, also those antiendomysium antibodies (EMA) have become relatively sensitive and specific tools for initial detection of gluten-sensitive disease (Kagnoff, 2006). Despite the lack of randomized controlled trials, dapsone has been considered the first-line treatment in patients with DH for over 70 years. According to the literature, the side effects of drugs are dose-dependent (Wozel & Ch, 2014). Methemoglobinaemia is generally less than 5% and does not exceed 12% in DH. The signs and symptoms occur at methemoglobin 3%, characterized by cyanosis, a greyish colour, weakness, headaches, tachycardia, nausea and abdominal pain (Wozel & Ch, 2014). Other adverse effects are peripheral neuropathy (reversible with dose reduction); morbiliform eruption, erythema nodosum, erythema multiforme, exfoliative dermatitis, toxic epidermal necrolysis, severe hypoalbuminaemia with anasarca, leucopoenia and agranulocytosis, which may be fatal during the first three months, cholestasis and hepatitis. Due to both the side effects and gravity of some cases, patients taking dapsone must be closely and continuously monitored (Wozel & Ch, 2014).

Aim of the study

The aim of our study was to characterize retrospectively the group of patients suffering from DH and to analyse some parameters such as dapsone dose (= Disulone dose), immunoglobulin (Ig)A antibodies against tissue transglutaminase (IgA-TGA) and the severity of DH. We evaluate the patient's age in confirmation of the diagnosis and the prevalence of disease.

Statistical evaluation is used for the assessment (1) the level of IgA-TGA to the dapsone dose, (2) the severity of DH to the dapsone dose, (3) the severity of DH to the level of IgA-TGA, (4) the age and the gender to monitored parameters.

Patients and methods

Patients suffering from DH were included in the study. All these patients were examined at the Department of Dermatology, Faculty Hospital Hradec Králové, Charles University, Czech Republic and at the Department of Dermatology, Faculty Hospital Pilsen, Charles University, Czech Republic during 1980–2019. It means that in this study the patients suffering from DH were included from two regions in the Czech Republic. In the region of Hradec Králové, there are 551,000 inhabitants and the area of this region is 4759 square km. In the region of Pilsen, there are 584,000 inhabitants and the area of this region is 7649 square km.

The diagnosis of DH was based on the confirmation of histology from the skin lesion and on the laboratory examination-specific antibodies such as antiendomysium, antireticulin, antigliadin, tissue transglutaminase antibodies (IgA-TGA). Before the initiation of dapsone therapy, patients had undergone a careful clinical evaluation that had included a complete history and physical examination. Routine laboratory checks were recommended.

In all included patients we evaluated these data (recorded items characterization):

1. The severity of DH – mild, moderate, severe. The clinical examination of the patients with DH was performed every 3–6 months. The severity of DH was evaluated during the last 3 years for this study.

   • mild form – sporadic, slightly pruritic erythematous papules

   • moderate form – some grouped lesions with moderate itching

   • severe form – widespread lesions with intensive pruritus.

2. Immunoglobulin (Ig)A antibodies against tissue transglutaminase (IgA-TGA) – It was measured with the method of an enzyme-linked immunosorbent assay (ELISA) at the Department of Allergy and Clinical Immunology, Faculty Hospital Hradec Králové and in the Department of Allergy and Immunology, Faculty Hospital Pilsen, Czech Republic. The immunoglobulin (Ig)A antibodies against tissue transglutaminase ≤10 U/ml were assessed as negative, the level > 10 U/ml as positive. The measurement of IgA-TGA was performed in all included patients during the treatment with dapsone every six months. ELISA (enzyme-linked immunosorbent assay) is a plate-based assay technique designed for detecting and quantifying peptides, proteins, antibodies and hormones. In ELISA, an antigen must be immobilized to a solid surface and then complexed with an antibody that is linked to an enzyme (Hornbeck, 2015). Detection is accomplished by assessing the conjugated enzyme activity via incubation with a substrate to produce a measurable product. The most crucial element of the detection strategy is a highly specific antibody–antigen interaction. The ELISA test kit provides a semi-quantitative in vitro assay for human antibodies of the IgG class against tissue transglutaminase in serum or plasma. The test kit contains microtitre strips each with 8 break-off reagent wells coated with human tissue transglutaminase. In the first reaction step, diluted patient samples are incubated in the wells. In the case of positive samples, specific IgG antibodies (also IgA and IgM) will bind to the antigens. To detect the bound antibodies, a second incubation is carried out using an enzyme-labelled anti-human IgG (enzyme conjugate), which is capable of promoting a colour reaction (Hornbeck, 2015).

3. The dapsone dose – the daily dose of dapsone was monitored and the average daily dapsone dose was calculated in every patient in the last 3 years.

Statistical analysis

We evaluated in statistics:

1. The relation of the IgA-TGA level to the dapsone dose – Fisheŕs exact test, Spearman Rank Correlation Coefficient

2. Severity of DH to dapsone dose – Fisheŕs exact test

3. Severity of DH to the IgA-TGA level - unpaired F - test Kruskal–Wallis One-Way ANOVA with the level of significance set to 5%.

4. Evaluation of age and gender to the followed parameters (severity of DH, IgA-TGA level, dapsone dose)

   • the IgA-TGA levels, dapsone dose to age – Spearman test

   • the severity of DH to age – the test Kruskal–Wallis One-Way ANOVA

   • the severity of DH to gender – the Chi-square test and with Fisheŕs exact test

   • the IgA-TGA level and dapsone dose to gender – Mann–Whitney U and Wilcoxon Rank-Sum Test for Difference.

Patients were divided according to IgA-TGA level in the following categories 0–10, 10–50, 50–100, more than 100 U/ml. Based on the individual dapsone dose, patients were stratified to the following categories – dapsone doses in mg 25, 25–50, 50, 50–100, 100 and more. Pairs of these categories were entered in the contingency tables and Fisher’s Exact test of these variables was performed with the level of significance set to 5%. The average dapsone dose and the average IgA-TGA level were calculated in every patient during the last 3 years too. Due to the violation of normality, we also used for statistical analysis Spearman Rank Correlation Coefficient.

The prevalence of DH was calculated per 100,000 for East Bohemia and West Bohemia. The age of patients, when the diagnosis of DH was confirmed, was also calculated.

The results were processed in the Department of Medical Biophysics of Medical Faculty of Charles University in Hradec Králové.

Results

Altogether 122 patients with the diagnosis of DH were included in the study. The characteristics of patients are shown in Table 1 (92 men, 30 women, average age 56 years, min. 25 years, max. 86 years, average age in men 55.1 years, in women 56.9 years). The mild form of DH was recorded in 69 patients (56.6%) with the average age of diagnosis 33.5 years, the moderate form in 48 patients (39.3%) with the average age of diagnosis 34.1 years and the severe form in 5 patients (4.1%) with the average age of diagnosis 29.8 years. The prevalence of DH in East Bohemia is 13.79 a and in West Bohemia 7.74 per 100,000.

Table 1. Characteristics of patients. 122 patients (= 100%) suffering from DH included in the study, 118 patients (96.7%) on dapsone therapy, 4 patients (3.2%) only on gluten-free diet, 92 men, 30 women, average age 56 years, min. 25 years, max. 86 years

Disease severity	No of patients	Average age of included patients (years)	Average age of diagnosis (years)
Mild form	69 (56.6%) (49 men, 20 women)	Men 55.2 Women 57.2	33.5
Moderate form	48 (39.3%) (39 men, 9 women)	Men 54.9 Women 56.3	34.1
Severe form	5 (4.1%) (4 men, 1 women)	Men 54.9 Women 56.3	29.8

Out of 122 patients (= 100%), 118 patients are on dapsone therapy (96.7%), 4 patients are on the gluten-free diet (3.2%). All these patients were regularly followed and examined (every 3–6 months) at the Department of Dermatology, Faculty Hospital Hradec Králové and Charles University in Hradec Králové and at the Department of Dermatology, Faculty Hospital Pilsen during 1980–2019. Below described data were evaluated:

1) The relation of the IgA-TGA level to dapsone dose

The level of IgA-TGA ≤ 10 U/ml was recorded in 75 patients (61.5%), IgA-TGA 10 −50 U/ml in 18 patients (14.8%), IgA-TGA 50–100 U/ml in 5 patients (4.1%) and > 100 U/ml in 8 patients (6.6%). The daily average dose 25 mg during the last years was recorded in 15 patients (12.3%), the dose 25–50 mg in 13 patients (10.7%), the dose 50 mg in 39 patients (31.9%), the dose 50–100 mg in 17 patients (13.9%) and the dose more than 100 mg in 22 patients (17.5%).

We evaluated the relation between the IgA-TGA level and dapsone dose.

1. The average dapsone dose (the daily average dose during the last 3 years) and the average IgA-TGA level were calculated in every patient. Due to the violation of normality, we used for statistical analysis Spearman Rank Correlation Coefficient (= 0.13), which means a very weak degree of relation Table 2, Graph to Table 2.

2. Patients were divided according to IgA-TGA into the following categories 0–10 U/ml, 10–50 U/ml, 50–100 U/ml, more than 100 U/ml and according to the dapsone dose (the daily average dose during the last 3 years), patients were stratified to the following categories: dapsone dose 25 mg, 25–50mg, 50 mg, 50–100 mg, over 100 mg. The relation between IgA-TGA level and the dapsone dose is shown in Table 3. In this kind of statistical evaluation (Fisher,s exact test), the relation between the IgA-TGA level and the dapsone dose was not confirmed (p-value = 0.223. We calculated the relative frequency also – Complement to Table 3 (given IgA-TGA and given dapsone dose).

Table 2. The relation between the mean IgA tissue transglutaminase antibodies (IgA-TGA) level and the average dapsone dose. Due to the violation of normality, we used statistical analysis Spearman Rank Correlation Coefficient (= 0.1330*), which means a very weak degree of relation.

	Age	Dapsone dose	Level IgA-TGA
Pearson correlation	0.0573	1.0000	0.2485
Spearman Correlation	0.0442	1.0000	0.1330*
Pearson P-value	0.5307		0.0366
Spearman P-value	0.5307		0.0366
Count	122	122	71

Graph to Table 2

The relation between the mean IgA tissue transglutaminase antibodies (IgA-TGA) level (axis y) and the average dapsone dose (mg), (axis x).

Table 3. The number of patients according to the level IgA-TGA (in categories 0–10, 10–50, 50–100, more than 100 U/ml, the number of patients 122 = 100%) and according to the dapsone dose (the categories 25, 25–50, 50, 50–100, 100 mg), (Fisher’s exact test).

	The tissue transglutaminase antibodies level - IgA-TGA (U/ml) – No. of patients
	0 to 10 U/ml	10 to 50 U/ml	50 to 100 U/ml	Over 100 U/ml	Total patients (%)
Dose of Dapsone – No. of patients
25 mg	11 (9.0%)	4(3.3%)	0(0%)	0(0%)	15(12.3%)
25–50 mg	10 (8.2%)	2(1.6%)	0(0%)	1(0.9%)	13(10.7%)
50 mg	27(22.1%)	8(6.6%)	2(1.6%)	2(1.6%)	39(31.9%)
50–100 mg	11(9.1%)	4(3.3%)	0(0%)	2(1.6%)	17(13.9%)
100 mg	16(13.1%)	0(0%)	3(2.5%)	3(2.5%)	22(18.0%)
Total patients	75 (61.5%	18 (14.8%)	5 (4.1%)	8 (6.6%)	106 (86.9%)
p-value	p-value = 0.223
Complement to Table 3 The relative frequency according to the dapsone dose. The number of patients (relative frequency in %) according to the IgA-TGA (in categories 0–10, 10–50, 50–100, more than 100 U/ml) and according to the dapsone dose (the categories 25, 25–50, 50, 50–100, 100 mg, the number of patients in the category = 100%).

	The tissue transglutaminase antibodies level - IgA-TGA (U/ml)
Dose of dapsone	0 to 10 U/ml	10 to 50 U/ml	50 to 100 U/ml	Over 100 U/ml	Total patients %
25 mg	11(73.3%)	4(26.7%)	0.0%	0.0%	15 = 100.0%
25–50 mg	10(76.9%)	2(15.4%)	0.0%	1(7.7%)	13 = 100.0%
50 mg	27(69.2%)	8(20.5%)	2(5.1%)	2(5.1%)	39 = 100.0%
50–100 mg	11(64.7%)	4(23.5%)	(0.0%)	2(11.8%)	17 = 100.0%
100 mg	16(72.7%)	0.0%	3(13.6%)	3(13.6%)	22 = 100.0%
p-value = 0.223
Complement to Table 3 – the relative frequency according to the level IgA-TGA The number of patients (relative frequency in %) according to the IgA-TGA (in categories 0–10, 10–50, 50–100, more than 100 U/ml, the number of patients in the category = 100%) and according to the dapsone dose (the categories 25, 25–50, 50, 50–100, 100 mg).

	The tissue transglutaminase antibodies level – IgA-TGA (U/ml)
Dose of dapsone	0 to 10 U/ml	10 to 50 U/ml	50 to 100 U/ml	over 100 U/ml
Total patients	75 = 100%	18 = 100%	5 = 100%	8 = 100%
25 mg	11 (14.7%)	4(38.9%)	0.0%	0.0%
25–50 mg	10(13.3%)	2(11.1%)	0.0%	1(12.5%)
50 mg	27(36.0%)	8 (44.4%)	2 (40.0%)	2(25.0%)
50–100 mg	11(14.6%)	4(38.9%)	(0.0%)	2(25.0%)
100 mg	16 (21.3%)	0.0%	3 (60.0%)	3(37.5%)
p-value	p-value = 0.223

2) The relation of DH severity and the dapsone dose

The relation between the dapsone dose (the daily average dose during the last 3 years) and the severity of DH in patients with mild, moderate, and severe form is shown in Table 4. The relation was confirmed (Fisheŕs exact test, p – value = 0 .000619). We calculated the relative frequency also – Complement to Table 4 (given severity and given dapsone dose).

Table 4. Dapsone dose (the average daily dose during the last 3 years in the categories of doses of dapsone 25, 25–50, 50, 50–100, 100 mg) and number of patients with mild, moderate, severe form of DH, 122 patients included in the study = 100%. The relation was confirmed, Fisheŕs exact test.

	Dapsone dose (mg)
	25 mg	25–50 mg	50 mg	50–100 mg	100 mg	total patients%
No. of patients mild form	17 (13.9%)	7 (5.7%)	26 (21.3%)	7 (5.7%)	9 (7.4%)	66 (54.1%)
No. of patients moderate form	0 (0%)	7(5.7%)	16 (13.1%)	12 (9.8%)	12 (9.8%)	47 (38.2%)
No. of patients severe form	0 (0%)	0 (0%)	2 (1.6%)	1 (0.8%)	2 (1.6%)	5 (4.1%)
Total no. of patients	17(13.9%)	14 (11.5%)	44 (36.1%)	20 (16.4%)	23 (18.9%)	118 (96.7%)
p-value	p-value = 0.000619
Complement to Table 4. – the relative frequency according to the severity of disease The number of patients (relative frequency in %) according to the severity of disease (mild, moderate and severe form and to the dapsone dose (the categories 25, 25–50, 50, 50–100, 100 mg)
	Dapsone dose (mg)
	25 mg	25–50 mg	50 mg	50–100 mg	100 mg	Total patients %
No. of patients mild form	17 (25.8%)	7 (10.6%)	26 (39.4%)	7 (10.6%)	9 (13.6%)	66 = 100.0%
No. of patients moderate form	0.0%	7 (14.9%)	16 (34.0%)	12 (25.5%)	12 (25.5%)	47 = 100.0%
No. of patients severe form	0.0%	0.0%	2 (40.0%)	1 (20.0%)	2 (40.0%)	5 = 100.0%
p-value	p-value = 0.000619
Complement to Table 4 – the relative frequency according to the dapsone dose. The number of patients (relative frequency in %) according to the severity of disease (mild, moderate and severe form), and to the dapsone dose (the categories 25, 25–50, 50, 50–100, 100 mg).
	Dapsone dose (mg)
	25 mg	25–50 mg	50 mg	50–100 mg	100 mg	Total patients%
Total number of patients	17 = 100%	14 = 100%	44 = 100%	20 = 100%	23 = 100%	118 = 100%
No. of patients mild form	17 (100%)	7 (50%)	26 (59.1%)	7 (35%)	9 (39.1%)	66 (55.9%)
No. of patients moderate form	0	7 (50%)	16 (36.4%)	12 (60%)	12 (52.2%)	47 (39.8%)
No. of patients severe form	0	0	2 (4.5%)	1 (4.5%)	2 (8.7%)	5 (4.2%)
p - value	p-value = 0.000619

As only a small number of reported patients with severe form of DH, we made a collapse of number of patients with moderate and severe form together. The relation between the dapsone dose and the severity of DH in patients with mild form and moderate + severe form was confirmed too (Fisheŕs exact test, p-value = 0.000099).

3) The relation of DH severity to the IgA-TGA level

The comparison of levels IgA – TGA according to the severity of DH (mild, moderate, severe form) is demonstrated in Table 5, (test Kruskal–Wallis One-Way ANOVA). The median of average IgA-TGA in the mild form is 6.25 U/ml, in the moderate form 14.45 U/ml and in the severe form 10.33 U/ml. We show the 1st quartile and 3rd quartile, the relation was not confirmed, p- value = 0.166. We show the graph (Box plot) with the relation between the severity of DH and the IgA-TGA level (Graph to Table 5).

Table 5. The comparison of levels IgA - TGA according to the severity of DH (mild, moderate, severe form), test Kruskal–Wallis One-Way ANOVA. The relation was not confirmed, p-value = 0.166.

N	Form	Median of average TGA (U/ml)	1st quartile	3rd quartile	Kruskal–Wallis One-Way ANOVA
69	Mild	6.25	1.8	44.94	0.166
48	Moderate	14.45	4.3	102.06
5	Severe	10.33	0.75	169.37

The median of average IgA-TGA in mild form is 6.25 U/ml, in moderate form 14.45 U/ml and in severe form 10.33 U/ml. We show the 1st quartile and 3rd quartile, p- value = 0.166.

The graph to Table 5. This graph (Box plot) shows the relation between the severity of DH and the IgA-TGA level.

4) The Evaluation of the relation between Dapsone dose, severity of disease and IgA-TGA to the age of patients and to the gender.

The evaluation of the relation between dapsone dose and the age of patients was done; the relation is not confirmed, Spearman test is 0.044, Table 6, Graph 6(a). The evaluation of the relation between the IgA-TGA and age of patients was done; the relation is not confirmed, Spearman test is 0.022, Table 6, Graph 6(b).

Table 6. The evaluation of the relation between dose of dapsone and age of patients was calculated, the relation is not confirmed, Spearman test is 0.044* Table 6, Graph 6(a). The evaluation of the relation between the IgA-TGA and the age of patients was calculated, the relation is not confirmed, Spearman test is 0.022**Table 6, Graph 6(b).

	Age	Dapsone dose	Level IgA-TGA
Pearson Correlation	1.0000	0.0573	0.1094
Spearman Correlation	1.0000	0.0442*	0.0221**
Pearson P-value		0.5307	0.3637
Spearman P-value		0.5307	0.3637
Count		122	71

Graph 6 a). The relation dapsone dose (Disulone dose ) and age. The relation is not confirmed. Spearman test is 0.044.

Graph 6 b). The level IgA-TGA and age. The relation is not confirmed. Spearman test is 0.022.

The evaluation of the severity of DH to age – evaluated with the test Kruskal–Wallis One-Way ANOVA, p-value = 0.739, the relation was not confirmed.

The evaluation of the severity of DH to gender, the Chi-square test and Fisheŕs exact test,

p- value = 0.43636, the relation was not confirmed.

The evaluation of dapsone dose (Disulone dose) to gender, Mann–Whitney U and Wilcoxon Rank-Sum Test for Difference, p- value = 0.28702, the relation was not confirmed.

The evaluation IgA-TGA to gender, Mann–Whitney U and Wilcoxon Rank-Sum Test for Difference,: p-value = 0.52643, the relation was not confirmed.

Discussion

The purpose of our study was to evaluate in the group of patients suffering from DH the important clinical and laboratory parameters and to characterize this group of patients. The majority of patients (56.6%) suffer from the mild form, only 4.1% of patients suffer from the severe form. Gluten-free diet is recommended to all patients with DH. Of 122 patients (= 100%), 118 patients (96.7%) are on dapsone therapy and try to strictly keep on a gluten-free diet, 4 patients (3.2%) are only on a gluten-free diet without dapsone therapy.

We have chosen some parameters such as the severity of the disease, dapsone dose and the level of IgA-TGA to characterize the group of patients. We did not confirm the relation of the followed parameters (severity of DH, dapsone dose and IgA-TGA) to gender and to age.

The IgA-TGA level measuring is suitable for the evaluation of the status of jejunum mucous membrane and patient adherence to gluten-free diet and to the therapy. We've been trying to determine whether a certain dapsone dose is associated with negative IgA-TGA levels. We have not shown this dependence; according to our results, the dapsone dose is individual and we cannot determine the dose of dapsone that leads to the negative level of IgA-TGA. Dapsone has some adverse effects, so we are trying to find the optimum dose of dapsone which will lead to the lower level of IgA-TGA and which has a significant effect also on improving the skin finding. The dapsone dose 100 mg is given to 13.6% of patients with the mild form, to 25.5% of patients with the moderate form and to 40.0% of patients with the severe form. The median of average IgA-TGA in the mild form is 6.25 U/ml, in the moderate form 14.45 U/ml and in the severe form 10.33 U/ml. One study analysed whether the quantification of autoantibodies against tissue transglutaminase could be used to predict mucosal destruction and disease severity in patients with gluten sensitivity. One hundred seventy patients with coeliac disease, comprising 52 children with severe malabsorption (group I), 59 children with mild symptoms (group II), 59 adults (group III), 134 patients with DH (DH) and 131 disease controls, were studied. High levels of IgA-TGA and IgG-TGA antibodies were associated with the grade of mucosal villous atrophy and a more severe clinical presentation. According to the results, the combined measurement of IgA-TGA and IgG-TGA enables a noninvasive prediction of small intestinal villous atrophy with high accuracy, and may reduce the need for a biopsy in patients with suspected celiac disease (Dahlbom et al., 2010). According to the literature, epidermal transglutaminase (TGe) and closely related tissue transglutaminase (tTG) are considered to be autoantigens in DH, because a majority of DH patients have IgA antibodies specific for TGe and for tTG. It is unknown where and how these autoantigen-specific IgAs are generated in DH. The levels of circulating IgA-TGA and anti-TGe IgA in DH correlate with each other and that both appear to correlate with the degree (extent) of enteropathy. Intestinal damage is associated with the production of anti-tTG IgA and anti-TGe IgA in DH patients (Bonciolini et al., 2019; Leonard et al., 2017). Reunala et al. followed up patients with DH during a strict gluten-free diet treatment and found that circulating IgA autoantibodies to TG3 decrease in parallel with TG2 and EmA antibodies. This shows that TG3 antibodies do not offer any additional advantage over TG2 and EmA antibodies when monitoring the gluten-free diet treatment (Reunala et al., 2015). According to Japanese case report, the monitoring of circulating IgA antibodies to transglutaminases can be a marker of disease activity and therapeutic efficacy in DH patients (Kurosaki et al., 2018). In our study, the negative level of tissue transglutaminase antibodies IgA (≤ 10 U/ml) was recorded in 75 patients (61.5%) from 122 patients included in the study. The majority of our patients are on the daily dose 50 mg (31.9% patients), 22 patients (17.5%) are on the dose more than 100 mg. Although no significant relation between the IgA-TGA and the dapsone dose and the disease severity was confirmed, we can observe that the level of IgA-TGA over 100 U/ml is recorded in a higher number of patients according to the increasing dose of dapsone. We can also observe some trend in the levels of IgA-TGA according to the severity of DH, but the significant relation was not confirmed.

All patients included in our observation were examined at gastroenterology department for the evaluation of celiac disease. The prevalence of the celiac disease in Czech Republic is 1: 200–1:250 (50–70,000 patients). All included patients in our study are on the gluten-free diet. Gluten-free diet requires strict monitoring of all ingested foods; it is recommended to eliminate also wheat, oats, barley, rye (contains avenine, which causes cross-reactivity, often oats got contaminated during processing). We have observed the worsening of skin findings and new papulo-vesicular lesions and pruritus after the diet mistake. The majority of these patients eliminated gluten, but they recorded that flares are reproducible after a casual ingestion of gluten. In these patients, we did not recommend the increase of the dapsone dose, because the skin findings improved following the strict diet regimen. Gluten-free diet (GFD) alleviates gastrointestinal symptoms much more rapidly than the rash: it takes an average of 2 years of gluten-free diet for complete elimination of the cutaneous lesions, which invariably recur within 12 weeks after the reintroduction of gluten (Garioch et al., 1994; Hardman et al., 1997; Janatuinen et al., 1995; Lewis et al., 1996; Rottmann, 1992). Observing 133 DH patients on a long-term gluten-free diet, Garioch et al. reported the following advantages: reduced or no need for medication, resolution of enteropathy and the correlated malabsorption of essential nutrients (and therefore prevention of alimentary deficiency of iron, vitamin B12 and folate), a general feeling of well-being, protective effects against the development of intestinal lymphoma (Garioch et al., 1994; Hardman et al., 1997; Janatuinen et al., 1995; Lewis et al., 1996; Rottmann, 1992). Although this diet offers many benefits in the management of DH, it is not easy to realize by many DH patients (Garioch et al., 1994; Hardman et al., 1997; Janatuinen et al., 1995; Lewis et al., 1996; Rottmann, 1992). A GFD requires scrupulous monitoring of all ingested foods; it is time-consuming and socially restricting. Strict adherence to a GFD requires extensive knowledge of foods and diet, thus consultation with a dietician and involvement in DH support groups are strongly encouraged. Rottmann provides a detailed list and description of foods permitted and those to be avoided. More studies are required to determine whether a long-term GFD will decrease the incidence of concurrent autoimmune conditions in patients with DH (Garioch et al., 1994; Hardman et al., 1997; Janatuinen et al., 1995; Lewis et al., 1996; Rottmann, 1992). In our study, of 122 patients (= 100%), 118 patients (96.7%) are on the gluten-free diet together with Dapsone therapy; diet alone was not sufficient to improve the skin finding in these patients. On the other hand, four patients are only on the gluten-free diet (3.2%) and suffer from the mild form of DH.

We did not observe severe adverse effects in our patients on dapsone therapy; in addition, we observed also the improvement of clinical signs of celiac disease. We started usually with the initial dose between 100 and 200 mg per day and the response occurred within some days, with no new lesions appearing. We recommend the patient should take a minimum dose, sufficient to suppress the disease. Some of our patients take 25 mg of dapsone per week, but others need up to 100 mg daily. Regarding the adverse effect of dapsone in our group of patients, we observed leucopoenia and thrombocytopenia only in one patient, with the reduction of dapsone dose to 25 mg daily, the blood count improved. In eight (6.6%) patients, we observed methemoglobinaemia above 5%, in all these cases we recommended the reduction of dapsone dose to 25 mg daily without the worsening of skin finding, 50 mg of dapsone can cause some haemolysis. Doses of 150 mg may decrease 2 grams of haemoglobin, which can be asymptomatic in healthy patients or trigger several symptoms and signs in patients with cardiac or lung disease, or even in elderly people (Wozel & Ch, 2014). Methemoglobinaemia is generally less than 5% and does not exceed 12% in DH. Patients, who cannot tolerate the use of dapsone, may benefit from sulfapyridine (1–1.5 g / day, tetracycline 2 g/day, together with nicotinamide 1.5 g/day or cyclosporine for resistant cases) (Wozel & Ch, 2014). Dapsone must be used with caution in the following conditions: Glucose-6-phosphate dehydrogenase deficiency, met-Hb-reductase deficiency, severe hepatopathy, cardiac insufficiency/heart failure, pulmonary diseases, co-medication with met-Hb-inducing drugs or compounds, respectively (Wozel & Ch, 2014). Rogers et al. selected dapsone as an alternative anti-inflammatory agent in a subset of patients with bullous pemphigoid. According to this study, cicatricial pemphigoid may be added to the list of Dapsone-responsive dermatoses (Rogers et al., 1982).

Sulfasalazine and sulphamethoxypyridazine might provide an effective alternative to dapsone especially when it fails to control the disease or the therapy is complicated by adverse events. The suggested dosages are of 1–2 g/day for sulfasalazine and of 0.25–1.5 g/day for sulphamethoxypyridazine. Both drugs share common adverse effects, including hypersensitivity reactions, haemolytic anaemia, proteinuria and crystalluria; thus, a full blood count with differential and urine microscopy with urinalysis should be carried out prior to starting treatment and monthly for the first 3 months of therapy, and thereafter once every 6 months. However, the most common adverse effects are nausea, anorexia and vomiting, which can be prevented by prescribing the enteric-coated forms of the drugs (Caproni et al., 2009; Willsteed et al., 2005).

According to our long-term follow-up, it is necessary to confirm the diagnosis by an objective response to a gluten-free diet and to detect and manage non-compliance. We also assess a patient's compliance with a gluten-free diet, to reinforce the importance of such compliance and to evaluate the possible development of intestinal malabsorption and/or DH-associated conditions. Beyond this, there are no clear guidelines as to the optimal means to monitor adherence to a gluten-free diet. In general, monitoring adherence to a gluten-free diet with serological investigations (i.e. anti-tTG or EMA) is sensitive for major but not for minor transient dietary indiscretions (Beutner et al., 2000; Beutner & Plunkett, 2006; Kagnoff, 2006). In our study, diet adherence was monitored by serological testing and skin lesion observation. Several authors have suggested that small bowel biopsy is no longer regarded as mandatory for the diagnosis of celiac disease at least in a subgroup of patients, (Hill & Holmes, 2008). In DH, some authors have demonstrated IgA-TGA specificity higher than 90%, and a sensitivity ranging from 47% to 95% (Desai et al., 2005).

Conclusion

The data of 122 patients with DH were retrospectively analysed. The majority of patients (56.6%) suffer from the mild form, only 4.1% of patients suffer from the severe form of DH; 96.7% of patients are on dapsone therapy, 3.2% of patients are only on a gluten-free diet. The majority of patients are on the daily dapsone dose 50 mg (31.9% patients). Dapsone therapy is well tolerated, and the adverse effects are rare. The relation between the dapsone dose and the disease severity was confirmed. The dapsone dose 100 mg is given to 13.6% of patients suffering from the mild form, to 25.5% of patients suffering from the moderate form and to 40% of patients with the severe form. Although we did not confirm the significant relation of the IgA-TGA to the dapsone dose and to the disease severity, we can observe that the IgA-TGA over 100 U/ml is recorded in a higher number of patients according to the increasing dapsone dose and there is also some trend in the increase of the IgA-TGA level from the mild to moderate and severe form of DH. We did not confirm the relation of the followed parameters (severity of DH, dapsone dose and IgA-TGA) to gender and to age.

Acknowledgments

The authors would like to thank Professor Roy S. Rogers III, from Mayo Clinic Arizona, for his valuable comments and help with editing of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).