Celiac disease (CD) is an immune-mediated disease, which is triggered by the ingestion of gliadin and other toxic prolamines in genetically susceptible individuals. It is characterized by a dysregulated immune response at gut level (1) that results in enteropathy of variable severity.
Today the analysis of the small intestinal biopsy is still considered the gold standard for a definitive diagnosis of CD; however, serologic tests are largely used to identify in a noninvasive way the patients that warrant a jejunal biopsy. In fact, untreated patients produce several gluten-dependent antibodies; of these, anti-gliadin (AGA) antibodies (IgA class and IgG class) and deamidated gliadin peptide antibodies (D-AGA) are targeted to dietary gliadin, whereas IgA anti-endomysium (anti-EMA) and IgA anti-tissue transglutaminase 2 (anti-TG2) are autoantibodies to closely related epitopes of the tissue transglutaminase 2 (TG2) (2,3).
In different studies performed in children, the sensitivity of IgA-class AGA for the diagnosis of CD ranges widely from 52% to 95% (4,5) and the specificity varies from 68% to 98% (6,7), whereas IgG-class AGA tests have a sensitivity and specificity ranging from 88% to 100% and from 36% to 97%, respectively (7,8). The sensitivity of these tests is higher in children younger than 2 years of age than in older patients, as shown in 1983 by Savilahti et al (9). A new generation of AGA antibody assays has been developed recently to detect antibodies to synthetic deamidated homologous gliadin peptides in the serum of patients with CD (10,11); nevertheless, these tests, although highly sensitive and specific, are not yet widely implemented, and the test results in young children are still lacking. Over the years, in clinical practice the AGA test has given way to the EMA test, because the latter has a sensitivity and specificity approaching 100% (6,7); however data in the literature show that the assay may be less accurate in children younger than 2 years, with a sensitivity of 83% to 87% (5,12,13). As far as anti-TG2 autoantibodies are concerned, they are present in the serum of untreated patients with CD and are now detected by an ELISA test based on the use of human recombinant tissue transglutaminase; the sensitivity of the test in children ranges from 91% to 97%, with a specificity varying from 96% to 100% (10,14,15), these values being similar to those in EMA. Even in this case the sensitivity is lower in children younger than 2 years with values of 67% to 83% (5,13); previous studies (5,13,16) have shown that in this age group, the sensitivity of the AGA test is higher than that of the anti-TG2 antibodies test.
We now know that anti-TG2 autoantibodies are primarily produced in the small bowel mucosa of patients with CD (17) and can be found deposited below the epithelial basement membrane and around the crypts; it seems that such subepithelial IgA anti-TG2 are specific for gluten-induced enteropathy (18). Recent findings suggest that these deposits are already present in the early phases of disease before the development of the mucosal damage (19) and before they can be detected in serum (18). This is why anti-TG2 antibodies can still be localized only at the level of the small intestinal mucosa in some patients with CD without serum EMA and TG2 positivity.
The aim of our retrospective study was to evaluate how many children younger than 2 years of age and diagnosed as having CD were negative for serum anti-TG2 antibodies and to test the hypothesis that in this group of patients, TG2-specific–IgA deposits could instead be present at the mucosal level.
PATIENTS AND METHODS
We enrolled 283 children who underwent small-intestinal biopsy because of gastrointestinal symptoms at the Department of Pediatrics in Naples University Hospital “Federico II” from January 2000 to December 2007. The children presented a lesion of small intestinal mucosa classified as Marsh type 3 (M3) (variable degree of villous atrophy) according to Marsh classification modified by Oberhuber (1999) (20), and were diagnosed as having CD according to the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN) 1990 criteria (villous atrophy with crypts hyperplasia of the small intestinal mucosa while the patients eat gluten and a full clinical remission on gluten-free diet). Our study population was composed of 2 groups of patients: 104 children younger than 2 years of age (range 3–24 months, median age 15 months) and 179 children older than the age of 2 years (range 2 years 6 months–17 years, median age 7 years). Patients deficient in IgA were excluded from our study.
Serum samples from all of the patients were collected at the time of intestinal biopsy. Serum levels of IgA-class AGA and IgG-class AGA, EMA, and anti-TG2 antibodies were evaluated.
Moreover 56 of 104 patients with CD younger than 2 years of age and 40 of 179 patients with CD older than the age of 2 years were also investigated for the presence of TG2-related extracellular IgA deposits. As a control group, we used 18 patients with villous atrophy and a diagnosis that was unrelated to CD: 10 had been diagnosed as having multiple food intolerances and 8 were allergic to cow's milk. All of them had negative serum EMA and anti-TG2 antibodies. We investigated a total of 114 children for the presence of mucosal deposits of anti-TG2 antibodies.
Enzyme immunoassays for the detection of specific IgA and IgG antibodies against α-fraction of gliadin were used. These ELISA tests (α-Gliatest S Chromo IgA and α-Gliatest S Chromo IgG) were provided by Eurospital (Italy). Values are positive when more than or equal to 55 U/mL.
EMA and Anti-TG2 Antibodies
Serum IgA EMA was detected by indirect immunofluorescence on a section of frozen human umbilical cord as a source of antigen (21). IgA anti-TG2 was identified by ELISA technique using a kit based on human recombinant antigen (Kit Eu-tTg IgA, Eurospital, Trieste) (21). For IgA anti-TG2, values are positive when more than or equal to 7 U/mL.
Intestinal Deposits of Anti-TG2
Frozen biopsy fragments were evaluated for the presence of IgA anti-TG2 extracellular deposits, as previously described (22). In brief, 5-μm frozen sections from each patient were fixed in acetone and were examined by double immunofluorescence. After a 15-minute preincubation with normal rabbit serum (1:100, Dako), the sections were covered with a monoclonal mouse antibody against guinea pig TG2 (CUB 7402) (1:200, NeoMarkers) for 1 hour at room temperature in a humid chamber. The sections were washed in phosphate buffered saline (PBS) and incubated for 30 minutes in the dark with a mixture of fluorescein isothiocyanate-labeled rabbit antibody against human IgA (1:100, Dako), to reveal (in green) IgA deposits, and R-phycoerythrin–labeled rabbit anti-mouse antibody (1:40, Dako) to reveal (in red) tissue transglutaminase 2. Finally, the sections were rinsed several times in PBS and mounted by glycerol/PBS (1:10). The colocalization images of IgA mucosal deposits and TG2 that resulted in a yellow band were analyzed using confocal microscopy (LSM510, Zeiss). Five sections from each sample were evaluated by 2 blinded investigators who had no knowledge of the patients' clinical or laboratory findings. Intraobserver and interobserver variations were each 98% regarding the detection of the presence or absence of anti-TG2 mucosal deposits.
Statistical differences between the study groups were evaluated using Pearson χ2 test. Values of P < 0.05 were considered to be significant.
Serum AGA, IgA EMA, and IgA Anti-TG2 Antibodies
As far as serum AGA antibodies are concerned, in the group of patients younger than 2 years of age, 89.4% and 94.2% of patients showed high serum levels of IgA-class and IgG-class AGA, respectively, whereas in the group of patients older than 2 years, 67.0% showed serum levels of IgA-class AGA and 84.3% of IgG-class AGA that were higher than the cutoff (Table 1).
In the group of children younger than 2 years of age, serum IgA EMA were detected in 88.4% of the patients, whereas among older children these antibodies were found in 98.3% of the patients. In addition, 87.5% children younger than 2 years and 96.1% older than 2 years of age showed high serum levels of anti-TG2 autoantibodies. These differences were statistically significant (P < 0.05) (Table 1).
Eleven of 104 (10.6%) children younger than 2 years were negative for both serum EMA and anti-TG2 autoantibodies. The clinical characteristics of those patients are listed in Table 2. Only 1 was negative for all of the evaluated antibodies, including AGA of both isotypes (patient 8 in Table 2). In the group of children older than 2 years, only 1 patient was negative for serum EMA and anti-TG2 antibodies; in this patient, only the serum levels of IgG-class AGA were higher than the cutoff. In this last category, no patients were negative for all of the tested antibodies.
Intestinal Deposits of Anti-TG2 IgA
Among 56 children younger than 2 years of age studied for intestinal deposition of anti-TG2, 46 of 56 (82.1%) were positive for serum EMA and presented serum levels of anti-TG2 higher than the cutoff. The remaining 10 cases were negative for serum EMA and/or anti-TG2 antibodies: 9 of 10 were positive for IgA-class and/or IgG-class AGA whereas only 1 was negative for all tested antibodies (patient 8 in Table 2). In these 2 subgroups of patients, mucosal deposits were positive in 39 of 46 (84.7%) and in 2 of 10 (20.0%) (Fig. 1A, B, C), respectively. In total, mucosal deposits were present in 41 of 56 (73.2%) children aged younger than 2 years. An image from a child younger than 2 years with serum positivity but without mucosal deposits of anti-TG2 antibodies is reported in Figure 1F.
All of the 40 patients with CD older than 2 years of age were positive for serum EMA and anti-TG2 antibodies and also were positive for intestinal deposits of anti-TG2 IgA.
All of the 18 patients with villous atrophy (M3) who were negative for serum EMA and anti-TG2 antibodies and had a diagnosis unrelated to CD, were negative for intestinal deposits. Results are summarized in Figure 2.
The pattern of positivity was characterized by a thick yellow band with a clear localization below the basement membrane, along the villous and crypt as well as around mucosal vessels in patients older than 2 years. Even when the deposits were in the same areas in all of the examined samples from patients younger than 2 years of age, we observed a patchy distribution. Some areas were completely negative, whereas others presented evident subepithelial deposits of anti-TG2 IgA (Fig. 1C).
In our study we confirmed the values of serologic tests such as anti-TG2, EMA, and IgA- and IgG-class AGA for the diagnosis of CD even in the group of children younger than 2 years of age. We showed that the majority of children younger than age 2 years diagnosed as having CD, presented elevated levels of EMA and anti-TG2 antibodies; however, in this age group approximately 10% of the cases lacked serum EMA and anti-TG2 antibodies. In these patients the diagnosis was based on a combination of different factors: suggestive symptoms, compatible human leukocyte antigen (HLA) (when available), villous atrophy, clinical recovery, and negative serology on gluten-free diet, exclusion of other causes of enteropathy (mainly, gut infections and cow's milk allergy), and in 1 case a clinical relapse after a gluten challenge. Unfortunately we were not able to perform HLA typing in all of the patients, but the 2 we have assessed presented HLA alleles associated with CD. We think that diagnosis other than CD is unlikely, but in the future those patients will undergo a formal gluten challenge.
We do not believe that the reduced sensitivity in patients younger than 2 years of age is because of the limits of the assay. For children younger than 2 years, the figures are comparable to those of the other studies, and the high sensitivity in older children would refute this hypothesis. We have thus confirmed that the sensitivity of EMA and anti-TG2 tests in children younger than 2 years old is lower than in children older than 2 years (5,12,13) and in adults; moreover, we have noted that there are patients who are positive for anti-TG2 and/or EMA without positivity for IgA-class and IgG-class AGA. On the basis of these observations, we conclude that in the group of patients younger than 2 years of age, the measurement of serum IgA-class and IgG-class AGA is still an important practice, but these tests should be combined with anti-TG2 and/or EMA to gain better sensitivity. However, this could reduce the specificity of the diagnosis so that in this category of patients, greater attention should be devoted to combine serologic results with clinical and histologic data and to verify the clinical response to gluten-free diet (GFD). Simell et al (23) have recently described the natural history of CD-associated antibodies in the first few years of age in patients genetically at risk for developing CD. They demonstrated that EMA, anti-TG2, and IgA-class AGA may fluctuate during the first 2 years of life, whereas IgG-class AGA usually remains positive. In this study, IgA-class AGA and IgG-class AGA were detectable in serum for months before anti-TG2 antibodies appeared. We thus hypothesize that IgA-class and IgG-class AGA are the first antibodies that appear in the serum of children with CD, and the other specific CD autoantibodies are produced in a later stage. Because of their precocity and their stability in blood, IgG-class AGA may be the best candidates in this age group for the diagnosis of CD; unfortunately, they usually occur in the general population (24), and only a fraction of IgG-class AGA-positive children develop overt CD. Some recent data have demonstrated the value of D-AGA in the diagnosis and follow-up of patients with CD revealing their sensitivity and specificity to be near to those of anti-TG2 antibodies (10,11). Research on D-AGA even in patients younger than 2 years of age could increase diagnostic accuracy in this age group, and in the future it will be important to apply this method to these subjects.
Recent findings (18,19) have shown the value of the detection of mucosal anti-TG2 autoantibody deposits in CD diagnosis. We recently validated this technique in a pediatric population older than 2 years of age (22). Unfortunately, it requires frozen biopsy samples. In the present study we wanted to apply this method to cryostat intestinal sections of the children younger than 2 years of age to verify its sensitivity in this age group. Our aim was primarily to determine whether detection of IgA anti-TG2 antibodies deposits could be useful in clinical practice in the diagnosis of patients having serum-negative EMA and anti-TG2. Our data showed that the detection of mucosal anti-TG2 antibodies is specific for CD even in children younger than 2 years of age; none of the patients without CD presented extracellular deposits of anti-TG2. Nonetheless, in children with CD younger than 2 years of age, we did not find this test to be more sensitive than serum anti-TG2 antibody measurement; in fact, deposits are present in 73% (41/56) compared with 82% (46/56) for anti-TG2 serum antibodies. In particular, 84.7% had increased serum levels of EMA and/or anti-TG2 presented mucosal deposits, even in a patchy distribution, whereas the same deposits were detectable in only 20% of patients without serum EMA and/or anti-TG2 autoantibodies. We think that this last finding may be related to the natural fluctuation of these autoantibodies in serum. The discrepancy of serum anti-TG2 and intestinal deposits could also be related to the affinity of anti-TG2 antibodies. In their early stage of production these autoantibodies show a minor affinity to their antigen and could only be detected in the serum of patients with CD; over time their affinity to extracellular TG2 increases so that they can bind to it. They can be found in the mucosa of the small intestine; eventually their affinity could become so high that they remain only at a mucosal level, disappearing into the serum (19). All of this could explain the 15.3% of children with serum-positive EMA and anti-TG2 antibodies but without mucosal deposits.
The finding of patients with CD, intestinal atrophy in the absence of signs of production of anti-TG2 antibodies (absence in serum, absence of intestinal deposits), supports the idea that those autoantibodies against extracellular matrix component are not necessary for the development of mucosal damage, although once they are produced they may contribute to it (25).
In conclusion, our data show that in the group of patients younger than 2 years of age, the use of the AGA test still has an important role to play in terms of increased diagnostic sensitivity. The detection of mucosal deposits of anti-TG2 IgA could help in the diagnosis of patients negative for serum markers, but it is not more efficient than serum anti-TG2 measurement. Our results constitute additional evidence that the diagnosis of CD cannot be based on a single test but must instead be based on a combination of clinical criteria and serologic, genetic, and histologic tests.
1. Sollid LM. Coeliac disease: dissecting a complex inflammatory disorder. Nat Rev Immunol 2002; 2:647–655.
2. Dieterich W, Ehnis T, Bauer M, et al
. Identification of tissue transglutaminase as the autoantigen of celiac disease
. Nat Med 1997; 3:797–801.
3. Korponay-Szabó IR, Sulkanen S, Halttunen T, et al
. Tissue transglutaminase is the target in both rodent and primate tissues for celiac disease
-specific autoantibodies. J Pediatr Gastroenterol Nutr 2000; 31:520–527.
4. Lerner A, Kumar V, Iancu TC. Immunological diagnosis of childhood coeliac disease: comparison between antigliadin, antireticulin and antiendomysial antibodies. Clin Exp Immunol 1994; 95:78–82.
5. Mankai A, Sakly W, Landolsi H, et al
. Tissue transglutaminase antibodies in celiac disease
, comparison of an enzyme linked immunosorbent assay and a dot blot assay. Pathol Biol 2005; 53:204–209.
6. Sacchetti L, Ferrajolo A, Salerno G, et al
. Diagnostic value of various serum antibodies detected by diverse methods in childhood celiac disease
. Clin Chem 1996; 42:1838–1842.
7. Bottaro G, Volta U, Spina M, et al
. Antibody pattern in childhood celiac disease
. J Pediatr Gastroenterol Nutr 1997; 24:559–562.
8. Sulkanen S, Halttunen T, Laurila K, et al
. Tissue transglutaminase autoantibody enzyme-linked immunosorbent assay in detecting celiac disease
. Gastroenterology 1998; 115:1322–1328.
9. Savilahti E, Perkkio M, Kalimo K, et al
. IgA antigliadin antibodies: a marker of mucosa damage in childhood coeliac disease. Lancet 1983; 12:320–322.
10. Agardh D. Antibodies against synthetic deamidated gliadin peptides and tissue transglutaminase for the identification of childhood celiac disease
. Clin Gastroenterol Hepatol 2007; 5:1276–1281.
11. Liu E, Li M, Emery L, et al
. Natural history of antibodies to deamidated gliadin peptides and transglutaminase in early childhood celiac disease
. J Pediatr Gastroenterol Nutr 2007; 45:293–300.
12. Burgin-Wolff A, Gaze H, Hadziselimovic F, et al
. Antigliadin and antiendomysium antibody determination for coeliac disease. Arch Dis Child 1991; 66:941–947.
13. Lagerqvist C, Dahlbom I, Hansson T, et al
. Antigliadin immunoglobulin A best in finding celiac disease
in children younger than 18 months of age. J Pediatr Gastroenterol Nutr 2008; 47:428–435.
14. Sblattero D, Berti I, Trevisiol C, et al
. Human recombinant tissue transglutaminase ELISA: an innovative diagnostic assay for celiac disease
. Am J Gastroenterol 2000; 95:1253–1257.
15. Collin P, Kaukinen K, Vogelsang H, et al
. Antiendomysial and antihuman recombinant tissue transglutaminase antibodies in the diagnosis of coeliac disease: a biopsy-proven European multicentre study. Eur J Gastroenterol Hepatol 2005; 17:85–91.
16. Fabiani E, Catassi C, International Working Group. The serum IgA class anti-tissue transglutaminase antibodies in the diagnosis and follow up of coeliac disease. Results of an international multi-centre study. International Working Group on Eu-tTG. Eur J Gastroenterol Hepatol 2001; 13:659–665.
17. Marzari R, Sblattero D, Florian F, et al
. Molecular dissection of the tissue transglutaminase autoantibody response in celiac disease
. J Immunol 2001; 166:4170–4176.
18. Korponay-Szabò IR, Halttunen T, Szalai Z, et al
. In vivo targeting of intestinal and extraintestinal transglutaminase 2 by celiac autoantibodies. Gut 2004; 53:641–648.
19. Salmi TT, Collin P, Jarvinen O, et al
. Immunoglobulin A autoantibodies against transglutaminase 2 in the small intestinal mucosa predict forthcoming coeliac disease. Aliment Pharmacol Ther 2006; 24:541–552.
20. Oberhuber G, Granditsch G, Vogelsang H. The histopathology of celiac disease
: Time for a standardized report scheme for pathologists. Eur J Gastroenterol Hepatol 1999; 11:1185–1194.
21. Paparo F, Petrone E, Tosco A, et al
. Clinical, HLA, and small bowel immunohistochemical features of children with positive serum antiendomysium antibodies and architecturally normal small intestinal mucosa. Am J Gastroenterol 2005; 100:2294–2298.
22. Tosco A, Maglio M, Paparo F, et al
. Immunoglobulin A anti-tissue transglutaminase antibody deposits in the small intestinal mucosa of children with no villous atrophy. J Pediatr Gastroenterol Nutr 2008; 47:293–298.
23. Simell S, Hoppu S, Hekkala A, et al
. Fate of five celiac disease
-associated antibodies during normal diet in genetically at-risk children observed from birth in a natural history study. Am J Gastroenterol 2007; 102:2026–2035.
24. Catassi C, Fanciulli G, Appello AR, et al
. Antiendomysium versus antigliadin antibodies in screening the general population for coeliac disease. Scand J Gastroenterol 2000; 35:732–736.
25. Barone MV, Caputo I, Ribecco MT, et al
. Humoral immune response to tissue transglutaminase is related to epithelial cell proliferation in celiac disease
. Gastroenterology 2007; 132:1245–1253.