Secondary Logo

Journal Logo

Supplement Articles

Autoantibodies and CD

Past and Future of Celiac Antibody Testing

Korponay-Szabó, Ilma R.

Author Information
Journal of Pediatric Gastroenterology and Nutrition: July 2014 - Volume 59 - Issue - p S11-S13
doi: 10.1097/01.mpg.0000450395.68897.c1
  • Free



Gliadin antibody tests have been in clinical use for more than 50 years. Initially, crude gliadin fractions were added to frozen sections of monkey oesophagus where gliadin bound around the epithelial cells in a pemphigus-like pattern (1) and could be detected by fluorescent labels. This test worked reasonably well, because the added gliadin bound to keratinocyte transglutaminase (TG1) was expressed in this location. Later studies showed that gliadins can be used by transglutaminases as substrates that generate deamidated peptides more immunogenic for T lymphocytes. Until the discovery of anti-TG2 antibodies in 1997, serology was not a first-line test, and the diagnosis required until recently histology analysis of a small-intestinal biopsy specimen. Today, anti-TG2 antibodies can be analysed by high-throughput and convenient methods, including enzyme-linked immunosorbent assays (ELISA), rapid tests, and highly specific immunofluorescent methods based on the detection of TG2 antigen in the endomysium, and they are widely used for clinical evaluation, case finding and screening.


Most contemporary commercial coeliac disease (CD) antibody test kits are well optimized, but a significant fraction of interpretation problems may be generated by the inappropriate timing or choice of the performed antibody detection methods. Low quantities or short duration of gluten intake may be a problem in young children, especially in family members of already known patients with CD, where antibodies may not appear in the circulation until the age of 2 to 3 years or even later (2). Furthermore, reduction of gluten consumption during the diagnostic process should be avoided, but often seen when initial positive antibody results are not properly communicated to the patients or obtained by home testing without adequate instructions for use. Delays in appointments with the pediatric gastroenterologist or in performing a biopsy commonly result in significant drops in serum anti-TG2 antibody levels. Such circumstances can often be traced in the history of patients with fluctuating antibody positivity or inconclusive histology results, or in families where parents request to make a noninvasive diagnosis without a small intestinal biopsy. Such events are much more common in everyday practice than in research studies and may lead to the false impression of poor reproducibility of antibody test results. It is also important to consider previous immunomodulatory or immunosuppressive drugs (steroids and others) when antibody tests are ordered in patients with autoimmune diseases, inflammatory bowel diseases, malignant disorders, or posttransplantation (3,4). Immunoglobulin deficiency, either primary or secondary due to renal loss or protein-losing enteropathy, is also a severe challenge for serum celiac antibody detection if both serum immunoglobulin A (IgA) and IgG levels are low. Selective humoral IgA deficiency (total serum IgA<0.05 g/L) is common (1 in 400 in the normal population) but can be easily overcome by ordering IgG class antibody detection, provided IgA deficiency becomes known during the diagnostic process. Low serum total IgA levels are frequently seen in children younger than 3 years of age. Celiac disease is a potent trigger for antibody production in IgA class, and infants with total serum IgA levels as low as 0.04–0.07 g/L have been shown to elaborate IgA class anti-TG2 antibodies (5). It is advisable, however, to perform both IgA and IgG class celiac antibody detection in patients with total serum IgA levels below 0.2 g/L. Sample quality, for example, hemolysis also may constitute a preanalytical problem for serum anti-TG2 antibody detection, because TG2 abundantly present in red blood cells may sequestrate the antibodies before they bind to the test plate in ELISA. Rapid celiac antibody tests designed for using whole blood and using the red blood cell TG2 in the blood as antigen for the antibody detection as in the Biocard series (6), will not work with serum samples that do not contain red blood cell transglutaminase.


CD anti-TG2 antibodies target a complex 3-dimensional surface of the protein specific for CD (7). This epitope surface should be intact, well-shaped, and properly exposed in diagnostic test kits. Most recombinantly produced human TG2 antigens nowadays meet these criteria, although it cannot be excluded that occasionally kits with lower antigen quality could be still found. CD antibodes bind slightly better to the open conformation of TG2 (8), and our studies with conformation-specific monoclonal antibody reagents confirmed that in the well-performing diagnostic TG2 ELISA kits and in tissue substrates used for endomysial antibody detection the TG2 antigen is indeed in an open conformation (unpublished results). TG2-directed antibodies may be produced in low serum concentrations in a number of conditions unrelated to CD (autoimmune disoders, liver and myocardium diseases, tumors and childhood infections (reviewed in (9)). These noncoeliac TG2 antibodies have different epitope specificity and do not react with TG2 in the endomysial test, where the antigen is an extracellular matrix–associated TG2 with less accessible other epitopes. The endomysial antibody test has been shown nearly 100% specific for CD in experienced laboratories (10), and this is why it can be used as a good confirmatory test, despite the fact that it is an observer-dependent method. ELISA tests containing with-type TG2 and mutants of the CD TG2 epitope also can selectively recognize celiac antibodies (7) and may be available in the future to distinguish real celiac antibodies from nonspecific anti-TG2 reactivity. Anti-TG2 tests using IgG antibody recognition may be less optimized than those measuring IgA anti-TG2 antibodies, and despite their high specificity (10), their sensitivity may be variable. Traditional gliadin antibody measuements were difficult to perform due to the water insolubility of gliadins and were neither sufficiently specific nor adequately sensitive. The newer tests using deamidated gliadin peptides (DGP) as antigens have good technical performances, especially if they are run instead of ELISA platforms on automated systems more suitable for small peptide antigens (10). It is true for all celiac antibody tests that some lab-to-lab variations do exist and international quality control programmes often show a few outlyer laboratories. For this reason, not the average of the measured values but the median is representing better the collectively measured values (3).


Because it has been shown that higher serum antibody levels predict better villous atrophy, currently the interpretation of serum antibody results became the most important challenge in CD diagnostics. Which antibody concentrations in serum are then sufficiently high to have high enough probability to predict villous damage and thus serve as basis of the diagnosis of celiac disease without a biopsy and when? First of all, it is essential to verify antibody positivity from a second blood sample. When ESPGHAN recommendations changed in 1990 from the 3-biopsy rule with gluten challenge to the requirement of 1 single biopsy with severe villous damage, this was suggested on the basis of 95% specificity. In recent years, histology has been shown less specific than in 1990, whereas endomysial antibody results had at least 97.2% pooled specificity for villous atrophy and even higher specificity and predictability for the final diagnosis of CD in the long run, even if villous atrophy was not present (10,11). In research studies, also anti-TG2 antibody results exceeding 10 times of the upper limit of normal (>10xULN) provided similar diagnostic reliability (10). Initially, these results were only available for 1 particular commercial test kit, but subsequent studies analyzing big retrospective cohorts arrived to similar conclusions with several major kits used in European laboratories (12). It is, however, important to note, that currently no common standard values exist and result calculations may differ in some kits where no standard curve is applied; therefore, the >10xULN rule would not be applicable. Furthermore, no such predictive values are available for anti-DGP antibodies. Head-to-head direct comparison of antibody values obtained with different anti-TG2 and anti-DGP test kits and their relation to CD diagnosis in prospective studies has only recently started (in the collaborative ProCeDe study). Although antibodies are the critical component for the noninvasive diagnosis suggested in 2012 by the new ESPGHAN guidelines, these guidelines also require that other conditions should be met: the patient should have CD-relevant symptoms, compatible HLA-DQ markers, and a proper evaluation by a pediatric gastroenterologist followed by parental informed consent to a final noninvasive diagnosis for life. At the current stage of medical knowledge, double positivity for both endomysial antibodies and and high titer (>10xULN) anti-TG2 antibodies are required and both seem to be safe, even if lower levels of anti-TG2 and less symptoms may also be predictive enough according to more recent studies (12).


In addition to clarify more precisely numerical characteristics of CD antibody tests and improve reporting of CD antibody results, the production of DGP antibodies should be further explored. Certain DGP and TG2 epitopes are cross-reactive in experimental studies (13), even if this reactivity is not responsible for the whole anti-TG2 reactivity in patient sera. Further, biological role of DGP antibodies is not known. Interestingly, DGP antibodies may be produced in infants even without a coupled anti-TG2 response and CD pathology (14). The most important effects of TG2 antibodies are exerted at the cellular level and tissue-bound antibodies with avidity may be present in the gut even in cases where anti-TG2 antibdies cannot be detected from the circulation (9). Presently, detection of these tissue-bound antibodies requires a frozen biopsy sample, and so this study cannot be performed retrospectively or in non-invasive ways. Further technical improvements in this fields may greatly improve sensitivity and suitability for disease monitoring.


1. Chorzelski TP, Sulej J, Tchorzewska H, et al. IgA class endomysium antibodies in dermatitis herpetiformis and coeliac disease. Ann N Y Acad Sci 1983; 420:325–334.
2. 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.
3. Husby S, Koletzko S, Korponay-Szabó IR, et al. European Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines for the diagnosis of coeliac disease. J Pediatr Gastroenterol Nutr 2012; 54:136–160.
4. Ben-Horin S, Polak-Charcon S, Barshack I, et al. Celiac disease resolution afterallogeneic bone marrow transplantation is associated with absence ofgliadin-specific memory response by donor-derived intestinal T-cells. J Clin Immunol 2013; 33:1395–1402.
5. Korponay-Szabó IR, Dahlbom I, Laurila K, et al. Elevation of IgG antibodies against tissuetransglutaminase as a diagnostic tool for coeliac disease in selective IgA deficiency. Gut 2003; 52:1567–1571.
6. Raivio T, Kaukinen K, Nemes E, et al. Self transglutaminase-based rapid coeliac disease antibodydetection by a lateral flow method. Aliment Pharmacol Ther 2006; 24:147–154.
7. Simon-Vecsei Z, Király R, Bagossi P, et al. A single conformational transglutaminase 2 epitope contributed by three domains is critical for celiac antibody binding and effects. Proc Natl Acad Sci U S A 2012; 109:431–436.
8. Lindfors K, Koskinen O, Kurppa K, et al. Serodiagnostic assays for celiac disease based on the open or closed conformation of the autoantigen, transglutaminase 2. J Clin Immunol 2011; 31:436–442.
9. Szondy Z, Korponay-Szabó I, Király R, et al. Transglutaminase 2 dysfunctions in the development of autoimmune disorders: celiac disease and TG2-/- mouse. Adv Enzymol Relat Areas Mol Biol 2011; 78:295–345.
10. Giersiepen K, Lelgemann M, Stuhldreher N, et al. Accuracy of diagnostic antibody-tests for coeliac disease in children: summary from an evidence report. J Pediatr Gastroent Nutr 2012; 54:229–241.
11. Kurppa K, Räsänen T, Collin P, et al. Endomysial antibodies predict celiacdisease irrespective of the titers or clinical presentation. World J Gastroenterol 2012; 18:2511–2516.
12. Alessio MG, Tonutti E, Brusca I, et al. Correlation between IgA tissuetransglutaminase antibody ratio and histological finding in celiac disease. J Pediatr Gastroenterol Nutr 2012; 55:44–49.
13. Korponay-Szabó IR, Vecsei Z, Király R, et al. Deamidated gliadin peptides form epitopes that transglutaminase antibodies recognize. J Pediatr Gastroenterol Nutr 2008; 46:253–261.
14. Korponay-Szabo IR, Gyimesi J, Koletzko S, et al. Immune response to 100 mg gluten introduced at 4 months of age in children with genetic risk for coeliac disease. 44th Annual Meeting of ESPGHAN, Sorrento, 2011. J Pediatr Gastroenterol Nutr 2011; 52 (Suppl 2):E85.
© 2014 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,