Gluten-dependent Small Bowel Mucosal Transglutaminase 2–specific IgA Deposits in Overt and Mild Enteropathy Coeliac Disease

Abstract

Objectives: 

In coeliac disease, immunoglobulin (Ig)A–class autoantibodies against transglutaminase-2 are produced in the small intestinal mucosa, where they are deposited extracellularly. It remains unclear whether positive intestinal transglutaminase-2-targeted IgA deposits in subjects having normal small bowel mucosal morphology are signs of early-stage coeliac disease. We evaluated the gluten dependency of these deposits in overt and mild enteropathy coeliac disease.

Patients and Methods: 

All together 48 subjects suspected of coeliac disease but having normal small bowel mucosal villi were enrolled; 28 of them had latent coeliac disease. The remaining 20 having positive intestinal IgA deposits adopted a gluten-free diet before villous atrophy had developed. For comparison, 13 patients with overt coeliac disease and 42 noncoeliac controls were studied. Small bowel mucosal transglutaminase-2–specific autoantibodies were compared with villous morphology, intraepithelial lymphocyte densities, and serum coeliac autoantibodies.

Results: 

Intestinal IgA deposits were seen in all but 1 of the patients with latent coeliac disease, when the morphology was still intact; the intensity of these deposits increased as villous atrophy developed and decreased again on a gluten-free diet. In 20 patients with intestinal IgA deposits in normal villi, the intensity of the deposits decreased with the diet similarly to that seen in patients with overt coeliac disease. Mucosal IgA deposits were seen initially only in 5% of noncoeliac controls and in 8% after extended gluten consumption.

Conclusions: 

The response of small bowel mucosal transglutaminase-2–specific IgA deposits for dietary intervention was similar in overt and mild enteropathy coeliac disease. Detection of such IgA deposits thus offers a good diagnostic tool to uncover early-stage coeliac disease.

The current diagnostic criteria for coeliac disease are based on small bowel mucosal villous atrophy with crypt hyperplasia ⁽¹⁾. The gluten-triggered mucosal lesion develops gradually from mucosal inflammation to elongation of crypts and finally to overt villous atrophy ^(2,3), the atrophy thus covering only the end stage of the disease. Evidence suggests that coeliac patients may suffer from gluten-sensitive symptoms and signs even before villous atrophy has developed ^(4–7). Minor small bowel mucosal morphological and inflammatory changes such as increased densities of CD3+ and γδ+ intraepithelial lymphocytes (IELs) are unspecific and subject to false diagnosis ^(8–11), and new reliable tools to detect mild enteropathy coeliac disease are therefore warranted. A specific test also should encompass proof of gluten dependency when patients are placed on gluten-free dietary treatment.

Apart from mucosal changes, a typical feature of untreated coeliac disease is the presence of serum immunoglobulin (Ig)A–class autoantibodies against transglutaminase-2 (TG2) ^(12,13), which offer high sensitivity and specificity ^(14,15). In fact, positive autoantibodies in patients having normal small bowel mucosal villous morphology do not necessarily constitute a false positive finding because they may be a predictive sign of forthcoming mucosal villous atrophy and coeliac disease ^(9,16–18). These autoantibodies are produced in the small intestinal mucosa ^(19,20), where they may be present even when there are no measurable levels in the sera ^(5,21,22). Intestinal TG2-specific IgA deposits may offer a diagnostic tool for detecting early-stage coeliac disease without villous atrophy ⁽⁹⁾. These intestinal antibodies (with absence in serum) also have been found in patients with diabetes mellitus and in first-degree relatives of coeliac patients, without full certainty as to whether these antibodies are gluten induced or related to coeliac disease ^(23–25). The demonstration of gluten dependency in intestinal TG2-specific IgA deposits would further enhance the specificity of the test.

Our aim was to evaluate the gluten dependency and diagnostic value of small bowel mucosal TG2-specific IgA autoantibody deposits in overt and mild enteropathy coeliac disease. Our series included children and adults with latent coeliac disease, who initially showed normal small bowel mucosal villous architecture but subsequently developed mucosal villous atrophy and celiac disease while continuing on a normal gluten-containing diet. In addition, we found individuals evincing small bowel mucosal TG2-specific IgA autoantibody deposits in intact villi who were directly placed on an experimental gluten-free diet to find evidence of gluten sensitivity. The data were compared with those from patients with overt coeliac disease and noncoeliac controls.

PATIENTS AND METHODS

The study group comprised 48 patients suspected of coeliac disease but found to have normal small bowel mucosal villous architecture. Of these, 28 had latent coeliac disease because they developed mucosal villous atrophy during the follow-up when continuing on a gluten-containing diet (Table 1). In these patients, the median follow-up time from normal small bowel mucosal architecture to villous atrophy was 1.7 years (range 0.2–7.4 years). After the mucosal deterioration, the patients were placed on a gluten-free diet, the median duration of the diet being 1.4 years (range 0.8–5.9 years). The remaining 20 patients in the study group were found to have small bowel mucosal IgA deposits in intact villi (defined in this study as potential coeliac disease) (Table 1); to find evidence for their gluten dependency, these individuals were placed on an experimental gluten-free diet even if they did not fullfil the current diagnostic criteria for coeliac disease. After 1 year, the response to the dietary treatment was evaluated.

TABLE 1: Demographic data on study patients

The controls comprised 13 subjects with overt coeliac disease with small bowel mucosal villous atrophy and crypt hyperplasia and 42 patients with suspicion of coeliac disease but no evidence of villous atrophy in 2 successive biopsy samples on a gluten-containing diet with a median interval of 6.0 years (range 0.8–10.0 years) (Table 1). The study protocol was approved by the Ethical Committee of Tampere University Hospital. All of the subjects gave written informed consent.

Small Bowel Mucosal Morphology and Intraepithelial Lymphocytes

Small bowel mucosal biopsies were taken upon upper gastrointestinal endoscopy and in small children with a Watson capsule. For morphometrical analysis, the samples were paraffin embedded, processed, and stained with hematoxylin-eosin. The villous height–crypt depth ratio was counted in well-oriented biopsy samples as previously described ⁽²⁶⁾ and a ratio <2.0 was considered compatible with coeliac disease.

One piece of capsule biopsy or 2 forceps biopsy specimens were freshly embedded in optimal cutting temperature compound (OTC, Tissue-Tec, Miles, Elkhart, IN), frozen in liquid nitrogen, and stored at −70°C. Immunohistochemical studies were carried out on 5-μm-thick frozen sections. CD3+ IELs were stained with monoclonal antibody Leu-4 (Becton Dickinson, San Jose, CA) and γδ+ IELs with T-cell receptor bearing cell γ antibody (T Cell Diagnostics, Woburn, MA). Positive IELs were counted from immunohistochemically stained specimens with a ×100 light microscope objective as described elsewhere ⁽²⁷⁾. The reference value for CD3+ IELs was set at 37 cells per millimeter of epithelium and for γδ+ cells at 4.3 cells per millimeter of epithelium ⁽²⁷⁾.

Small bowel Mucosa TG2-specific IgA Deposits

Small bowel mucosal TG2-specific IgA deposits were investigated in unfixed, 5-μm-thick, frozen sections of small bowel specimens by direct immunofluorescence using fluorescein isothiocyanate–labelled rabbit antibody against human IgA (Dako, Glostrup, Denmark) at a dilution of 1:40 in phosphate-buffered saline, pH 7.4. In coeliac disease a clear subepithelial IgA deposition can be found below the basement membrane along the villous and crypt epithelium and around mucosal vessels; this is in contrast to normal small bowel samples, in which IgA is detected only inside the plasma and epithelial cells ⁽²¹⁾. The IgA deposits were graded semiquantitatively from 0 to 3 according to their intensity along the basement membrane in the villous crypt area, as previously described ⁽⁵⁾. It has been shown that these mucosal IgA deposits are specifically targeted against TG2 ^(21,22). For double labelling, sections were stained for human IgA (green, as above, this paragraph) and for TG2 (red) using monoclonal mouse antibodies against TG2 (CUB7402, NeoMarkers, Fremont, CA), followed by rhodamine-conjugated anti-mouse immunoglobulin antibodies (Dako), both diluted 1:200 in phosphate-buffered saline. In our laboratory, intraobserver and interobserver agreement in the detection of present or absent TG2-specific IgA deposits has been 98% among 5 investigators. All of the evaluations were carried out blindly without knowledge of disease history or laboratory findings.

Serology

Serum IgA-class reticulin (ARA) or endomysial antibodies were determined by an indirect immunofluorescence method as described earlier ⁽¹⁴⁾. A positive staining pattern seen in a serum dilution of 1:5 or more was considered positive in both tests. During the study period, endomysial antibodies replaced ARA in clinical practice. If the endomysial antibodies result was not available, then the ARA result was used instead. These 2 coeliac autoantibody tests have proved in our laboratory to be almost identical ⁽²⁸⁾, and both have been shown to be directed against TG2 ⁽¹³⁾. In this study, these antibodies are indicated as serum coeliac autoantibodies. In retrospect, it was impossible here to determine serum TG2-antibodies by enzyme-linked immunosorbent assay in all of the patients and controls.

HLA Typing

HLA typing was based on polymerase chain reaction with allele-specific primers identifying HLA DQ2 and DQ8, and performed with a Dynal DQ low-resolution SSp kit (Dynal, Oslo, Norway). In coeliac disease, 90% to 95% of patients carry the HLA DQ2-haplotype and most of the rest carry HLA DQ8 ⁽²⁹⁾.

Statistics

Quantitative data are expressed as medians and ranges and qualitative data as percent of abnormal values. Statistical differences were evaluated using the Mann-Whitney U, Wilcoxon, and McNemar tests, as appropriate. All of the testing was 2-sided, and P < 0.05 were considered statistically significant. All of the calculations were performed with the Statistical Package for Social Sciences version 14.0 (SPSS, Chicago, IL).

RESULTS

Small bowel mucosal TG2-specific IgA deposits already were seen in the first biopsy in all but 1 of the patients with latent coeliac disease, when mucosal villous architecture was still intact (Table 2, Fig. 1). When the patients continued on a normal gluten-containing diet, small bowel mucosal villous atrophy and crypt hyperplasia developed; in parallel, the intensity of mucosal IgA deposits increased, and the deposits were present in all of the patients at the time of the diagnosis of coeliac disease (Figs. 1 and 2). The intensity of mucosal IgA deposits again decreased significantly when the patients were placed on a gluten-free diet, albeit minor depositions still were seen in many. A total of 7 patients in the latent coeliac disease group were initially serum autoantibody negative; 6 of them still had positive small bowel mucosa IgA deposits, 2 had family history for coeliac disease, and all 6 cases tested had coeliac-type HLA. Sixteen of the 20 patients having TG2-specific small bowel mucosal IgA deposits in intact villi (potential coeliac disease) agreed to a follow-up biopsy after 1 year on a gluten-free diet. The intensity of the deposits decreased in 15 (94%) of the 16 on a gluten-free diet (Figs. 1 and 2). Parallel to these changes, abdominal symptoms and signs of malabsorption resolved in 13 (65%) and improved in 5 (25%) of the patients whilst on the gluten-free diet (Table 3). TG2-specific small bowel mucosal IgA deposits were present in all of the patients with untreated overt coeliac disease, and the intensity of the deposits decreased with the diet. In noncoeliac controls the deposits were seen initially in 1 of 22 (5%) and in 3 of 38 (8%) patients on a gluten-containing diet in the follow-up biopsies (Fig. 1); these deposit-positive cases did not present with any particular clinical sign and had no family history of coeliac disease. The histological, serological, and HLA DQ findings are set out in Table 2 and Figure 1. At the first biopsy in the noncoeliac control group, densities of CD3+ IELs were increased in 41% and γδ+ IELs in 48%, and during the study the cell densities decreased significantly with time without any dietary intervention (Table 3, Fig. 1). At the outset, 4 (10%) of the noncoeliac controls were coeliac autoantibody positive, but all of them seroconverted negatively during the study (Table 2).

TABLE 2: Abnormal findings and HLA DQ2 or DQ8 haplotypes in different patient groups at the time of first biopsy

FIG. 1: Changes in small bowel mucosal tissue transglutaminase-2 (TG2)–specific immunoglobulin (Ig)A deposits, villous height–crypt depth ratios, and densities of CD3+ intraepithelial lymphocytes in patients having latent, potential, and overt coeliac disease and in noncoeliac controls at different time points. Gluten = normal gluten-containing diet; GFD = gluten-free diet.

FIG. 2: A–F, Small bowel mucosal immunoglobulin (Ig)A deposits in a 48-year-old woman having latent coeliac disease. The first small bowel mucosal biopsy (B) showed normal villous morphology, but (D) 2 years later when she continued on a normal gluten-containing diet villous atrophy with crypt hyperplasia developed. Villous morphology (F) recovered on a gluten-free diet. Serum celiac autoantibodies were negative throughout, but (A) strong subepithelial tissue transglutaminase (TG2)-targeted IgA autoantibody deposits (yellow, arrow) already were present in the small bowel mucosa in the first biopsy, and (C) also when villous atrophy was detected. The deposits disappeared (E) with a gluten-free diet. G-L, Mucosal IgA deposits are shown from a 21-year-old woman with potential coeliac disease having (G) normal small bowel mucosal architecture and strong subepithelial tissue TG2-specific IgA deposits. After 1 year on a gluten-free diet her symptoms recovered and IgA deposits disappeared (L). IgA is stained with green (H, J), TG2 with red (I,K), and colocalization of IgA and TG2 is shown in yellow (A,C,G).

TABLE 3: Findings and subjective symptoms in 20 patients having potential coeliac disease before and after GFD

DISCUSSION

The present study confirms earlier data indicating that the borders of coeliac disease clearly extend beyond small bowel mucosal villous atrophy ⁽³⁰⁾. The current diagnostic criteria have been applied for many years and small intestinal mucosal atrophy has been sine qua non for the diagnosis ⁽¹⁾. To revise the criteria of coeliac disease beyond villous atrophy requires the demonstration of gluten dependency of the symptoms and histology in genetically susceptible individuals. We have shown in this article that by careful examination we are able to find coeliac disease in patients with normal villous structure. Once the criteria have been revised, the inevitable consequence is that these patients should be treated with a gluten-free diet. In this study, mucosal TG2-specific IgA deposits were accurate markers for gluten sensitivity; they were detected early on in the disease process, their intensity increased as enteropathy progressed on a gluten-containing diet, and their intensity decreased along with mucosal recovery on a gluten-free diet. These intestinal coeliac autoantibody deposits proved to be better initial markers for gluten sensitivity than small bowel mucosal IEL densities, and they also were able to detect serum coeliac autoantibody-negative cases having mild enteropathy coeliac disease (Table 2). This was demonstrated in patients with latent coeliac disease for whom the development of villous atrophy had been confirmed by follow-up on a gluten-containing diet. Furthermore, in patients with TG2-specific IgA deposits but no evidence of progression to overt coeliac disease, the deposits were shown to be gluten dependent (Fig. 1). The process was akin to that of overt coeliac disease and latent coeliac disease. In this study, the majority of patients with latent coeliac disease had experienced gluten-dependent symptoms before subsequent diagnostic enteropathy had developed (Table 1).

Furthermore, 18 (90%) of the 20 patients found to have small bowel mucosal TG2-specific IgA deposits, but considered in view of normal small bowel histology not to have celiac disease, benefited from gluten-free dietary treatment (Table 3). Similar cases have been reported in the literature ^(6,31,32); some subjects have even been diagnosed as having osteopenia or osteoporosis ^(4,7,33). Of note, in this study group the nonresponsive symptoms were refractory epilepsy and arthritis, indicating that these extraintestinal manifestations were not gluten dependent. Assuming that the diagnostic criteria for the disease should indeed be widened to cover mild enteropathy coeliac disease, some issues should be stressed to avoid overdiagnosis of the disorder. Marsh 1-type small bowel mucosal lymphocytosis is an unspecific finding ^(8–10), as also seen in the present study; the densities of CD3+ and γδ+ IELs also decreased with time in noncoeliac controls without any dietary intervention (Fig. 1). Some noncoeliac patients evinced negative seroconversion during the follow-up; these cases were ARA positive without HLA DQ2 or DQ8.

Furthermore, 8% of these controls had minor small bowel mucosal IgA deposits without any signs of gluten sensitivity in the follow-up biopsy. Long-term follow-up is needed to show whether these subjects will eventually develop overt coeliac disease ^(7,34,35). To conclude, no single test alone can reliably detect early-stage coeliac disease without villous atrophy, but gluten-dependent TG2-specific IgA deposits offer a good diagnostic tool. Based on the findings in this study, we suggest that the deposits should be investigated when coeliac disease is suspected but the small bowel mucosal villous morphology is equivocal. In symptomatic patients (as was the case in the present series) having signs of minor enteropathy coeliac disease without villous atrophy, gluten-free dietary treatment should be considered. In asymptomatic cases, the benefits of early treatment are more ambiguous and subject to further studies. In the meantime, follow-up with a normal gluten-containing diet is recommended. In coeliac disease, however, gluten-induced symptoms may occur outside the intestine ⁽³⁰⁾, and during the follow-up the wide clinical spectrum of the disease should be kept in mind.