Abstract
Objective. To evaluate the presence of celiac disease in patients with systemic sclerosis (SSc). The association of autoimmune diseases with celiac disease has been reported, but few publications deal with the combination of SSc and celiac disease.
Methods. We investigated the presence of anti-tissue transglutaminase (anti-tTG) antibodies and serum antiendomysial antibodies (anti-EMA) in 50 patients with SSc. All subjects were on a gluten-containing diet. Duodenal mucosa histology and biopsy culture were performed in anti-tTG-positive patients; anti-EMA and IgA, IgG1 anti-tTG were detected in culture supernatants.
Results. The incidence of celiac disease in patients with SSc was found to be 8%. Serum anti-tTG antibody-positive results were detectable in 5 out of 50 patients with SSc, but only in 4 of them was the diagnosis confirmed by histological results (Marsh classification).
Conclusion. Our data show an increased prevalence of celiac disease in patients with SSc.
Systemic sclerosis (SSc) is a clinically heterogeneous disorder of the connective tissue characterized by immune activation, microvascular injury, and fibrosis. Major organ-based complications involving the lungs, heart, kidneys, and gastrointestinal (GI) tract determine mortality and morbidity. Clinical manifestations depend both on organ injury and on subsets of SSc: diffuse cutaneous SSc (dcSSc) is characterized by short-term Raynaud’s phenomenon (RP), truncal and acral skin involvement, interstitial lung disease, renal failure, diffuse GI disease, myocardial involvement, erectile dysfunction; limited cutaneous SSc (lcSSc) is characterized by RP and limited skin involvement for many years, but also significant incidence of pulmonary hypertension, trigeminal neuralgia, skin calcification, and telangiectasia1,2. Irrespective of disease classification, SSc is typically associated with RP; up to 95% of patients with SSc have secondary RP. Scleroderma RP is characterized by microvascular damage and high plasma adrenomedullin and endothelin-l concentrations3. After the skin, the GI tract is the second most common target of SSc4. Digestive system involvement occurs in 50% of patients with SSc. Dyspepsia, restricted distension of the gastric antrum, and diffuse GI dysmotility are frequent features in SSc. These defects are independent from the occurrence of autonomic neuropathy5. Small bowel disease can be associated with a variety of symptoms such as intermittent bloating, abdominal cramps, chronic diarrhea, and malabsorption frequently requiring prokinetic drugs, laxatives, or proton-pump inhibitors, but the absence of symptoms occurs in about one-third of patients6–8.
Celiac disease (CD), an autoimmune enteropathy induced by wheat gliadin in genetically susceptible individuals, is one of the most common immunologically mediated GI diseases in Europe9. This illness shows a multitude of symptoms and signs involving several organs, although the diagnosis is based on histological observation in the intestinal mucosa of gluten-dependent villous atrophy10. In its common form, the presence of autoantibodies against endomysium and tissue transglutaminase (tTG) antigens, together with their disappearance with a gluten-free diet, confirms the diagnosis11,12. Antiendomysial antibodies (EMA) and anti-tTG antibodies provide useful evidence for screening and followup of patients with CD13–15. Although definitive diagnosis of CD is still based on histological findings of villous atrophy in duodenal specimens, detection of EMA and anti-tTG antibodies may be useful if histological examination is not specific for CD16. CD may coexist with connective tissue diseases, but identification of this association is difficult because CD may be atypical or may mimic a rheumatologic condition; in addition, autoimmune diseases may show typical symptoms of CD. All these findings often create confusion and delay correct diagnosis17. Recent reports showed links between CD and other autoimmune disorders such as type I diabetes, autoimmune thyroiditis, Sjögren’s syndrome, ankylosing spondylitis, polymyositis, rheumatoid arthritis, and psoriatic arthritis as well as autoimmune liver disease and Addison’s disease18–24.
Few publications have reported CD in patients with SSc25–29. The pathogenetic mechanism of the association between gluten enteropathy and other autoimmune diseases is still unclear: genetic predisposition, immunological mechanisms, and environmental factors have to be considered. Our aim was to investigate the incidence of CD in patients with SSc.
MATERIALS AND METHODS
Subjects
The study group consisted of 50 consecutive patients (43 female, 7 male; mean age 51 ± 14.5 yrs) with SSc diagnosed according to the American Rheumatism Association (American College of Rheumatology; ACR) criteria admitted to the Clinical Immunology and Allergy Unit of Sapienza University. All fulfilled the ACR criteria for SSc. They were subsequently divided into lcSSc and dcSSc subgroups according to LeRoy, et al1. Disease activity and disease severity in SSc was measured using the Valentini Scleroderma Disease Activity Score30 and Medsger Scleroderma Disease Severity Scale31. Table 1 lists the main baseline epidemiological and clinical features of the patients.
All study procedures were in accord with the ethical standards of the institutional committee responsible for human experimentation.
Circulating antibodies
Serum was collected from all patients. IgA and IgG anti-tTG antibodies were detected using an ELISA in which microtiter plate wells were coated with recombinant human tTG (Eu-tTG kit; Eurospital, Trieste, Italy)32,33. According to the manufacturer’s instructions, serum was diluted 1:100 and the absorbance was measured at 405 nm. The lower limit of positivity of IgG and IgA antibodies was, respectively, 29 AU/ml and 7 AU/ml. IgA and IgG1 EMA were screened by the direct immunofluorescent method on cryostat sections of monkey esophagus (Antiendomysium kit; Eurospital). IgA and IgG1 isotype EMA were measured in sera diluted 1:5 from all subjects showing anti-tTG-positive results34. EMA-positive results were identified by the typical reticulin-like staining of smooth-muscle bundles.
Duodenal biopsy examination
Four duodenal mucosa specimens were obtained at esophagogastroduodenoscopy. Two specimens were submitted for routine histological examination according to the Marsh/Oberhuber classification35. The results compatible with class III (a, b, and/or c) were considered pathognomonic for CD35. Two specimens were submitted to an organ culture system for 48 h at 37°C: one in the presence and one in the absence of a peptic-tryptic digest of gliadin (1 g/l) as described36,37. Culture supernatants were collected and stored at −70°C until they were used for both IgA and IgG1 EMA detection36,37.
Statistical analysis
All results are expressed as mean ± standard deviation. Analysis was done for all the 50 patients. Differences in continuous variables were evaluated by Student t test. Chi-square test and Fisher’s exact test were used for categorical variables. P value < 0.05 was considered significant.
RESULTS
Five of 50 (10%) SSc patient samples were positive for IgA anti-tTG antibodies in sera. Table 2 lists the main baseline epidemiological and clinical features of the patients presenting anti-tTG antibodies in sera. One of these 5 patients also was positive for IgG anti-tTG. IgA EMA-positive results were present in only 2 of 5 patients (Table 3). Four of 5 patients presenting IgA anti-tTG antibodies and/or EMA-positive results underwent endoscopic examination, and one refused this examination. These 4 patients showed subtotal or total atrophy of intestinal mucosa on histological examination (types IIIa to IIIc of the Marsh classification modified by Oberhuber, et al35). Organ culture investigations showed positive EMA and anti-tTG antibodies in 2 (50%) of the 4 SSc patients showing total atrophy of intestinal mucosa; specimens of the other 2 were excluded because of tissue necrosis (Table 3).
Esophagogastroduodenoscopy was performed for routine checks of gastroesophageal reflux disease in 20 patients with serum IgA anti-tTG antibody or EMA-negative results. In all patients, standard endoscopy showed no mucosal features (loss of duodenal folds, scalloping, nodularity, mosaic) suspicious for CD and alerting the endoscopist to take small intestinal biopsy specimens. However, biopsy specimens were taken from the duodenal bulb in 6 patients. The histological examination in these patients was normal.
In the study population of SSc patients the prevalence of CD was 8% according to detection of serum anti-tTG antibodies and histological classification. Results from the 50 SSc patients showed no correlation between the presence of CD and epidemiological or clinical findings (sex, age, disease duration, onset of RP, SSc subset, antinuclear antibody pattern, and SSc-specific autoantibody).
DISCUSSION
The increased prevalence of CD in patients with autoimmune pathologies has been reported in several studies, although the role of CD in the clinical course of different autoimmune diseases is unclear.
We found increased IgA anti-tTG antibodies in serum samples from 5 (10%) patients with SSc, but only 2 of them presented IgA EMA-positive results. Histological findings, according to the Marsh classification35, were positive for all of the 4 patients examined. Further, we found that EMA and anti-tTG antibodies were detectable in 2 culture supernatants of 4 patients studied; specimens of 2 of them were not adequate for evaluation because of tissue necrosis. Although screening for CD routinely evaluates the presence of specific serological markers (anti-tTG and EMA antibodies) in patients’ sera, confirmation of CD is still based on the histological observation of villous atrophy in duodenal specimens. In the presence of CD symptomatology, culture supernatants may be useful if histological examination is negative or uncertain. EMA assay from a cultured intestinal biopsy can detect gluten-sensitive enteropathy, characterized by an infiltrative/hyperplastic histological pattern, that is often associated with negative serum EMA38. However, culture supernatants may give a negative result from specimens of small size or with tissue necrosis.
Our previous data suggested that if anti-tTG antibodies were present in EMA-negative patients it could be indicative only of tissue lesion and not expression of CD39–41. Indeed it is possible to hypothesize that tTG is not the only antigen of the EMA. One report described that although tTG immunoabsorption is enough to abrogate reactivity of the anti-tTG ELISA, it fails to completely rule out EMA binding activity42. Moreover, another study showed the existence of EMA antigenic epitopes not related to tTG43. Tissue transglutaminase, the autoantigen well recognized for its role in the pathogenesis of CD, could perhaps not be the only autoantigen involved in gluten-dependent autoimmune reactions. Other normally “cryptic” autoantigens could be unmasked and cause a self-aggressive immunological response following the gliadin-initiated inflammatory process44. Immune response to cryptic antigens is not the only pathogenetic mechanism of autoimmune disease: genetic predisposition and molecular mimicry (triggering antigens) may also play a role.
In addition to genetic predisposition and exposure to triggering antigens, development of autoimmune diseases is related to the loss of mucosal barriers connected to environmental factors45.
Finally, the duration of gluten exposure following the gliadin-initiated inflammatory process has been reported to be a cause of the high prevalence of autoimmune disorders in celiac patients46. It appears that organ-specific autoantibodies are gluten-dependent and disappear with a correct gluten-free diet11.
In our study, CD was confirmed by the presence of villous atrophy in duodenal specimens in only 4 patients (8%). In SSc patients with GI symptoms and iron deficiency anemia, it is important to investigate for coexistence of CD. It is not known if the association of these 2 diseases is linked to the presence of a common pathogenic mechanism (an immune response against self-antigens) and/or a common genetic background.
Our study demonstrated a high prevalence of celiac disease among patients with SSc. Routine screening for celiac disease should be considered in patients with SSc.
Footnotes
- Accepted for publication December 22, 2008.