Celiac disease (CD) is a gluten-sensitive, immune-mediated chronic enteropathy as defined by characteristic changes seen on intestinal biopsy. The clinical spectrum of CD is wide, including cases with either typical intestinal features (eg, chronic diarrhea, abdominal pain, weight loss) or atypical extraintestinal features (eg, anemia, hypertransaminasemia, enamel dysplasia, neurological disturbances).
The definitive diagnosis of CD is based on a small intestinal biopsy examination, serological tests being considered the obligatory tool for a correct patient referral. Determination of the greatly sensitive and specific anti-tissue transglutaminase (tTG) immumoglobulin (Ig) A antibodies (Abs) is presently the test of choice used for this purpose (1). Anti-endomysium Abs (EMA) also react against tTG, but, although as accurate as anti-tTG, its measurement is observer dependent and therefore subjected to interpretation error. Regarding anti-native gliadin (nGlia) Abs, its determination has shown unsatisfying diagnostic sensitivity and specificity. Despite the accuracy of anti-tTG Abs in the diagnosis of CD, its reduced sensitivity is well known when testing children younger than 2 years (2,3). Because many of these anti-tTG negative children are positive for Abs against nGlia, determination of these Abs has been regarded as advantageous in young children. Age also influences clinical manifestations because a delayed onset of CD with atypical symptoms has been described in older children (5–7 years old) (4). Furthermore, an atypical presentation with a mild symptomatology has been described more frequently in adults than in children (5–8). Determination of Abs of IgA isotype is generally used in routine screening, but performance of IgG assays is indispensable to detect patients with IgA deficiency, a disorder present in approximately 2% of patients with CD (8).
The 2 main CD-associated antigens are functionally related because tTG catalyzes deamidation of nGlia in specific residues, a process that plays an important role in CD pathogenesis (9,10). Determination of Abs reacting against synthetic deamidated gliadin peptides (dGp) has emerged as a new and accurate tool for diagnosing and monitoring CD with a higher sensitivity and specificity than tools directed against the native antigen (11–15). Moreover, some authors have reported that the diagnostic performance of anti-dGp is similar to that of anti-tTG IgA (12,14–20). These studies have been performed mainly in either pediatric or adult CD populations and most of them were based on dual assays that simultaneously detect Abs of IgG and IgA isotypes. Therefore, little information exists about the influence of age at onset and Ab isotype in the accuracy of anti-dGp for CD diagnosis. Also, the correlation between Ab levels and the severity of the histopathologic lesions described for the other CD-associated Abs has not been fully demonstrated for both anti-dGp isotypes (5,12,16,21–26).
The aim of the present work was to analyze the diagnostic performance of anti-dGp of both IgG and IgA isotypes for CD regarding the age at diagnosis. We evaluated the usefulness of anti-dGp IgG for the detection of patients with IgA deficiency.
Patients and Control Group
We retrospectively analyzed serum samples from 100 newly diagnosed patients with CD (mean age 23.4 years, range 0.5–74, female/male 66/34) obtained before biopsy and implementation of a gluten-free diet. All of the patients were positive for anti-tTG IgA, and, on the basis of this selection criterion, CD diagnosis was subsequently confirmed by intestinal biopsy according to the Marsh-Oberhuber 1 to 3 grade classification. For this purpose, 1 to 4 biopsy samples were taken from the duodenum. Lymphocytic infiltration (Marsh 1) was found in 4 patients, crypt hyperplasia (Marsh 2) in 5, partial villous atrophy (Marsh 3a) in 16, subtotal villous atrophy in 57 (Marsh 3b), and total villous atrophy in 18 (Marsh 3c). For the analysis of the association between Ab levels and histopathologic severity, patients were grouped into 2 categories: those with mild lesions (Marsh 1–3a = 25 patients) and those with severe lesions (Marsh 3b–3c = 75 patients).
Also, we analyzed sera from 6 patients newly diagnosed as having CD with IgA deficiency (<0.07 g/L) and positive for anti-tTG IgG (mean age 16.5 years, range 1–35, female/male 4/2), 1 showing Marsh 1 biopsy, whereas villous atrophy was present in the other 5 patients (one 3a and four 3b).
We analyzed serum samples from 100 age-matched patients in the control group with various digestive pathologies other than CD (mean age 27.3 years, range 0.4–84, female/male 53/47). This group included 23 patients with Marsh 0 biopsy, 37 with milk and/or egg allergy, 13 with autoimmune hepatitis, 10 with primary biliary cirrhosis, and 17 with inflammatory bowel disease.
All of the Abs were detected by enzyme fluoroimmunoassays (Phadia, Freiburg, Germany) carried out on an automated Phadia 250 analyzer. Anti-dGp IgG and IgA and anti-nGlia IgA were determined in CD and control sera, whereas detection of IgG Abs against dGp and n-Glia was performed in IgA-deficient sera. Anti-tTG IgA Abs were additionally detected in the control group, no case being positive. The levels of anti-tTG Ab that had been detected at diagnosis in patients with CD were 12 to 10,000 U/mL (positive >7 U/mL) in the normal IgA group and 126 to 526 U/mL (positive >10 U/mL) in the IgA-deficient patients.
The diagnostic accuracy of each assay or the ability to discriminate people with celiac disease from healthy individuals was evaluated using receiver operating characteristic (ROC) curve analysis and considered as the area under the ROC curve (AUC). Sensitivity, specificity, and positive and negative predictive values (PPV and NPV) were calculated using the optimal cutoff points obtained on the basis of the ROC associated Youden index, the threshold value for which the best combination of sensitivity and specificity is provided (sensitivity + specifity−1). Monte Carlo replications were used to compare AUC, sensitivity, specificity, PPV, and NPV. Comparisons between Ab levels were performed using the Mann-Whitney U nonparametric test. Statistics were generated using the free statistical package R.2.10.0 (www.r-project.org). A P value <0.05 was considered to be statistically significant.
High Diagnostic Performance of Anti-dGp IgG
When analyzing the diagnostic performance in the whole CD and control populations, anti-dGp of both isotypes and anti-nGlia IgA displayed extremely high AUC, all being >97% (Table 1). Nevertheless, the diagnostic performances of the 3 assays differed significantly (P = 0.003), the highest being obtained with anti-dGp IgG and the lowest with anti-nGlia IgA. Optimal cutoff points for each test were determined on the basis of the ROC-associated Youden index being 8, 6, and 7.5 U/mL for anti-nGlia IgA, anti-dGp IgA, and anti-dGp IgG, respectively. The cutoffs provided by the manufacturer for all of these assays were 10 U/mL, whereas values between 7 and 10 U/mL were considered equivocal. Sensitivity, specificity, PPV, and NPV were calculated for each Ab using the optimal cutoff points, and afterward these parameters were compared between the 3 assays (Tables 1 and 2). Sensitivity and NPV of the 2 assays using dGp were similar and significantly higher than those obtained with the anti-nGlia IgA test (P = 0.040 and P = 0.035, respectively), whereas specificity and PPV did not statistically differ. As expected from the data obtained, no significant differences were found for any parameter analyzed when only IgA and IgG dGp assays were compared. When only controls with a Marsh 0 biopsy were considered, the specificity of anti-dGp IgG increased up to 100%, whereas that of the IgA isotype remained similar (95.6%). The value of all of the studied diagnostic parameters corresponding to anti-tTG IgA was 100% because all of the patients but none of the controls were positive for these Abs.
Diagnostic Performance of Anti-dGp IgG and IgA Is Influenced by Age at Diagnosis
We studied the diagnostic performance of both deamidated and nGlia assays, dividing patients into 2 groups on the basis of increasing diagnosis age cutoffs. We found that when diagnosis age was 7 years or younger, anti-dGp IgG was the assay that showed the highest values for all of the parameters analyzed, reaching 100% in all of them. Then, we analyzed the diagnostic performance of all of the assays comparing the group of children ages 7 years or younger (43 patients with CD, mean age 2.0, range 0.5–7; 39 controls, mean age 2.8 years, range 0.4–7) with the older individuals (57 patients with CD, mean age 39.1 years, range 9–74; 61 controls, mean age 43.0 years, range 8–84) (Tables 3 and 4). In children ages 7 years or younger, anti-dGp IgA performed worst than anti-dGp IgG for all of the parameters analyzed because none of them reached 100%, this difference being not statistically significant. In contrast, in the older group, anti-dGp IgG was only slightly superior to anti-dGp IgA in AUC, specificity, and PPV. It is of interest to note the lower sensitivity of anti-dGp IgG in the older group (91.2%, 95% confidence interval [CI] 80.7%–91.7%) compared with that obtained in children ages 7 years or younger (100%). This difference was not due to the higher number of patients with CD older than 7 years with minimal intestinal lesions (Marsh type 1 and 2) included in the study (7 patients older than 7 years vs 2 patients 7 years or younger) because all of them were positive for anti-dGp IgG and for anti-dGp IgA. Finally, when comparing deamidated and anti-nGlia IgA assays, the only parameters that showed significant differences were AUC in the older group (P = 0.016) and NPV in both age groups (P = 0.046 and P = 0.024 for children 7 years or younger and older than 7 years, respectively).
Anti-dGp Levels Are Associated With the Severity of Histopathologic Lesions
Patients with severe intestinal lesions showed statistically significant higher levels of anti-dGp IgA and IgG than patients with mild lesions (Mann-Whitney U test: z = −2.77, P = 0.006 and z = −3.11, P = 0.002, respectively) (Fig. 1B and C). Supporting these findings, anti-nGlia IgA and anti-tTG IgA levels were also significantly higher in the group of severe lesions (Mann-Whitney U test: z = −2.66, P = 0.008 and z = −2.47, P = 0.013, respectively) (Fig. 1A and D), although, in the case of anti-tTG IgA, one must take into account the possible bias produced by the criterion used for patients’ selection, that is, positivity for these Abs.
Levels of anti-dGp Abs Are Inversely Associated With Age at Diagnosis
Equation (Uncited)Image Tools
The levels of anti-dGp IgA and IgG in patients with CD were found to be inversely associated with age at diagnosis, children ages 7 years or younger showing significantly higher levels than older patients (Mann-Whitney U test: z = −3.52, P < 0.001; and z = −4.82, P < 0.001, respectively) (Fig. 2). A well-established inverse association of anti-nGlia IgA and anti-tTG IgA levels with age at diagnosis was also observed (Mann-Whitney U test: z = −3.63, P < 0.001; and z = −3.81, P < 0.001, respectively) (Fig. 2).
Equation (Uncited)Image Tools
To analyze this inverse association, patients were next grouped by increasing age at diagnosis (0.5–2, 3–7, 8–15, 16–25, and older than 25 years). Anti-dGp levels of both isotypes were highest in children ages 0.5 to 2 years, decreasing up to the group of 8 to 15 years. No further decrease was observed beyond this age group (Fig. 3). The same behavior was also observed for anti-nGlia IgA and anti-tTG IgA (Fig. 3).
Diagnostic Efficiency of Anti-dGp IgG in IgA-deficient Patients With CD
We also retrospectively compared the diagnostic performance of anti-dGp IgG and anti-nGlia IgG assays in 6 patients with CD with IgA deficiency (3 children ages 10 years or younger and 3 patients 16 years or older). All 6 patients were positive for anti-dGp IgG, whereas anti-nGlia IgG Abs were detected only in 5. The negative anti-nGlia IgG case was an adult patient, the only one with a Marsh type I biopsy. Thus, in spite of the low number of cases analyzed, anti-dGp IgG appeared to perform better than anti-nGlia IgG in diagnosing CD in IgA-deficient patients.
In this work, we have demonstrated the usefulness of detecting anti-dGp IgG Abs for CD diagnosis in children and IgA-deficient patients. In addition to confirming the higher performance of anti-deamidated gliadin Abs compared with that of Abs against the native antigen, our results also show the different behavior of anti-dGp IgG and IgA isotypes according to the age at diagnosis. Thus, when analyzing the whole CD and control populations, anti-dGp IgG showed the highest diagnostic accuracy, evaluated as AUC, specificity, and PPV. This good performance of anti-dGp IgG increased up to 100% for AUC, sensitivity, specificity, PPV, and NPV in children ages 7 years or younger. It must be taken into account that the high performance values detected in this work are, to some extent, influenced by the patients’ selection criterion used (positive for anti-tTG IgA) and the small size of the sample after dividing patients by age. Nevertheless, the high performance of anti-dGp IgG in childhood CD found by us is in agreement with the findings of the few studies that have analyzed the diagnostic accuracy of individual IgG and IgA deamidated assays in pediatric patients (15,23). Indeed, anti-dGp IgG has been reported to perform as well as anti-tTG IgA in children ages 2 years or younger (23). The diagnostic performance of anti-dGp Abs could not be compared with that of anti-tTG IgA in this study because all of the patients with CD were selected for being positive for these Abs. Nevertheless, during routine examination of children suspected to have CD, we have observed that anti-dGp IgG Abs are at least as sensitive as anti-tTG IgA in children ages 7 years or younger because all newly diagnosed anti-tTGA IgA-positive children in this age group were also positive for anti-dGp IgG. Furthermore, anti-dGp IgG was detected 6 months earlier than anti-tTG IgA in a 1-year-old girl. This earlier appearance has also been observed in some pediatric patients (15,27). Anti-dGp usefulness has also been demonstrated in small children with isolated positivity for anti-nGlia IgA, IgG isotype being more sensitive than IgA (28). Thus, although the diagnostic performances of both isotypes did not statistically differ, all of these findings support the additional determination of anti-dGp IgG to that of anti-tTG IgA in children to avoid the low sensitivity of these Abs reported in pediatric patients.
In patients older than 7 years, anti-dGp IgG was only superior to anti-dGp IgA regarding specificity and PPV. Although no conclusive data about the diagnostic performance of both isotypes are available from prospective studies in adolescent and adult patients (older than 14 years), a trend toward a higher specificity has also been observed for anti-dGp IgG (14,20). Indeed, these Abs have been shown to have similar or even higher specificity than anti-tTG IgA (12,14,16,20).
We found that levels of anti-dGp IgG and IgA were associated with the severity of the intestinal damage as it has been described for anti-tTG IgA, EMA, and anti-nGlia IgA Abs (5,21,22). Our results are in agreement with those of earlier studies showing anti-dGp levels of IgG and/or IgA isotypes to be lower in patients with lesser mucosal damage (12,16,23–26). Furthermore, similar to that observed for other CD associated Abs, levels of both anti-dGp isotypes were also inversely associated with age at diagnosis, younger patients (0.5–2 years) showing the highest values (29). This inverse association has also been reported for anti-dGp detected by the dual assay when using an age cutoff of 3 years (24). The decline of Ab levels with increasing age at diagnosis is mainly restricted to children because no further decrease was observed in patients older than 15 years.
Conflicting results have been reported about the usefulness of anti-dGp to confirm CD in cases of minimal intestinal lesions (Marsh 1 and 2). Early studies detected a low sensitivity of the dual anti-dGp assay, whereas a recent report, based on a larger group of patients, has shown this assay to have a higher sensitivity and efficiency than anti-tTG IgA for detecting these types of lesions (16,25,26). In line with this report, the sensitivity of anti-dGp (IgG and IgA) in cases of minimal intestinal lesions found by us is at least similar to that of anti-tTG IgA because all 9 patients with CD without villous atrophy included in this study were positive for both Abs.
The reasons for the different isotype performance of anti-dGp regarding the age at diagnosis are not clear. We can assume that the higher usefulness found for Abs of IgG isotype in children was not because of a weak or delayed total IgA production, because the patients included in the present study were positive for anti-tTG IgA. Although the production of both anti-dGp isotypes has been shown to be correlated, a differential isotype reactivity depending on the number of gliadin-deamidated residues has also been described, anti-dGp IgA requiring a higher number of deamidated residues to react than Abs of IgG isotype (30,31). It could be speculated that the number of deamidated residues is dependent on the time of gliadin exposure, thus explaining the lower performance of anti-dGp IgA in children.
Determination of CD-associated Abs of IgG isotype is indispensable to detect patients with IgA deficiency. Therefore, the high diagnostic accuracy of deamidated IgG assays makes them a potential tool for diagnosing these patients. In our study, the 6 patients with CD with IgA deficiency were positive for anti-dGp IgG irrespective of the age at diagnosis. In agreement with our results, this high diagnostic performance in both pediatric and adult patients has also been reported in a collaborative study and in isolated cases (14,15,23,25,32,33). Thus, anti-dGp IgG determination can facilitate the identification of IgA-deficient patients and eliminate the need for measuring total serum IgA.
In conclusion, our results show that the diagnostic performance of anti-dGp depends on the Ab isotype and on the age at CD diagnosis, anti-dGp IgG being a serological marker at least as useful as anti-tTG IgA for detecting children with CD ages 7 years or younger. Our data also indicate that anti-dGp IgG can improve the diagnosis of patients with IgA deficiency.
The authors thank Phadia for generously providing antibody kits for the present study and Pablo Martínez Camblor from the Biomedical Investigation Office (OIB-FICYT) of the Principality of Asturias for the statistical analysis.
1. Hill ID, Dirks MH, Liptak GS, et al. Guideline for the diagnosis and treatment of celiac disease in children: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr 2005; 40:1–19.
2. Simell S, Kupila A, Hoppu S, et al. Natural history of transglutaminase autoantibodies and mucosal changes in children carrying HLA-conferred celiac disease susceptibility. Scand J Gastroenterol 2005; 40:1182–1191.
3. 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; 47:428–453.
4. Fassano A. Clinical presentation of celiac disease in the pediatric population. Gastroenterology 2005; 128:S68–S73.
5. Vivas S, Ruiz de Morales JM, Fernández M, et al. Age related clinical, serological and histopathological features of celiac disease. Am J Gastroenterol 2008; 103:2360–2365.
6. Kurppa K, Koskinen O, Colin P, et al. Changing phenotype of celiac disease alter long-standing gluten exposure. J Pediatr Gastroenterol Nutr 2008; 47:500–503.
7. Johnson MW, Ellis HJ, Asante MA, et al. Celiac disease in the elderly. Nat Clin Pract Gastroenterol Hepatol 2008; 5:697–706.
8. Barker JM, Liu E. Celiac disease: pathophysiology, clinical manifestations and associated autoimmune conditions. Adv Pediatr 2008; 55:349–365.
9. Van de Wal Y, Kooy Y, van Veelen P, et al. Selective deamidation by tissue transglutaminase strongly enhances gliadin-specific T cell reactivity. J Immunol 1998; 161:1585–1588.
10. Molberg O, Macadam SN, Körner R, et al. Tissue transglutaminase selectively modifies gliadin peptides that are recognized by gut-derived T cell in celiac disease. Nat Med 1998; 4:713–717.
11. Schwertz E, Kahlenberg F, Sack U, et al. Serologic assay based on gliadin-related nonapeptides as a highly sensitive and specific diagnostic aid in celiac disease. Clin Chem 2004; 50:2370–2375.
12. Sugai E, Vázquez H, Nachman F, et al. Accuracy of testing for antibodies to synthetic gliadin-related peptides in celiac disease. Clin Gastroenterol Hepatol 2006; 4:1112–1117.
13. Kaukinen K, Collin P, Laurila K, et al. Resurrection of gliadin antibodies in celiac disease. Deamidated gliadin peptide antibody test provides additional diagnostic benefit. Scand J Gastroenterol 2007; 42:1428–1433.
14. Volta U, Granito A, Fiorini E, et al. Usefulness of antibodies to deamidated gliadin peptides in celiac disease diagnosis and follow up. Dig Dis Sci 2008; 53:1582–1588.
15. Prause C, Ritter M, Probst C, et al. Antibodies against deamidated gliadin as new and accurate biomarkers of childhood coeliac disease. J Pediatr Gastroenterol Nutr 2009; 49:52–58.
16. Vermeersch P, Geboes K, Mariën G, et al. Diagnostic performance of IgG anti-deamidated gliadin peptide antibody assays is comparable to IgA anti-tTG in celiac disease. Clin Chem Acta 2010; 411:931–935.
17. Prince HE. Evaluation of the INOVA diagnostics enzyme-linked immunosorbent assay kits for measuring serum immunoglobulin G (IgG) and IgA to deamidated gliadin peptides. Clin Vaccine Immunol 2006; 13:150–151.
18. Aberg AK, Olcén P. Serologic screening for celiac disease in children: a comparison between established assays and tests with deamidated gliadin-derived peptides plus conjugates for both IgA and IgG antibodies. APMIS 2009; 117:808–813.
19. Ankelo M, Kleimola V, Simell S, et al. Antibody responses to deamidated gliadin peptide show high specificity and parallel antibodies to tissue transglutaminase in developing coeliac disease. Clin Exp Immunol 2007; 150:285–293.
20. Niveloni S, Sugai E, Cabanne A, et al. Antibodies against synthetic deamidated gliadin peptides as predictors of celiac disease: prospective assessment in an adult population with a high pretest probability of disease. Clin Chem 2007; 53:2186–2192.
21. Tursi A, Brandimarte G, Giorgetti GM, et al. Low prevalence of anti-gliadin and anti-endomysium antibodies in subclinical/silent celiac disease. Am J Gastroenterol 2001; 96:1507–1510.
22. Donaldson MR, Firth SD, Wimpee H, et al. Correlation of duodenal histology with tissue transglutaminase and endomysial antibody levels in pediatric celiac disease. Clin Gastroenterol Hepatol 2007; 5:567–573.
23. Basso D, Guarisso G, Fogar P, et al. Antibodies against synthetic deamidated gliadin peptides for celiac disease diagnosis and follow-up in children. Clin Chem 2009; 55:150–157.
24. Parizade M, Bujanover Y, Weiss B, et al. Performance of serology assays for diagnosing celiac disease in a clinical setting. Clin Vaccine Immunol 2009; 16:1576–1582.
25. Volta U, Granito A, Parisi C, et al. Deamidated gliadin peptide antibodies as a routine test for celiac disease. A prospective analysis. J Clin Gastroenterol 2010; 44:186–190.
26. Kurppa K, Lindfors K, Collin P, et al. Antibodies against deamidated gliadin peptides in early-stage celiac disease. J Clin Gastroenterol 2010; 45:673–678.
27. 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.
28. Barbato M, Maiella G, Di Camillo C, et al. The anti-deamidated gliadin peptide antibodies unmask celiac disease in small children with chronic diarrhea. Dig Liver Dis 2011; 43:465–469.
29. Ascher H, Hah-Zoric M, Hanson LA, et al. Value of serologic markers for clinical diagnosis and population studies of coeliac disease. Scand J Gastroenterol 1996; 31:61–67.
30. Marietta EV, Rashtak S, Murray JA. Correlation analysis of celiac sprue tissue transglutaminase and deamidated gliadin IgG/IgA. World J Gastroenterol 2009; 15:845–848.
31. Aleanzi M, Demonte AM, Esper C, et al. Celiac disease: antibody recognition against native and selectively deamidated gliadin peptides. Clin Chem 2001; 47:2023–2028.
32. 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.
33. Villalta D, Tonutti E, Prause C, et al. IgG antibodies against deamidated gliadin peptides for diagnosis of celiac disease in patients with IgA deficiency. Clin Chem 2010; 56:464–468.