Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

A diagnosis of celiac disease is made based on clinical, genetic, serologic, and duodenal morphology features. Recent pediatric guidelines, based largely on retrospective data, propose omitting biopsy analysis for patients with concentrations of IgA against tissue transglutaminase (IgA-TTG) >10-fold the upper limit of normal (ULN) and if further criteria are met. A retrospective study concluded that measurements of IgA-TTG and total IgA, or IgA-TTG and IgG against deamidated gliadin (IgG-DGL) could identify patients with and without celiac disease. Patients were assigned to categories of no celiac disease, celiac disease, or biopsy required, based entirely on antibody assays. We aimed to validate the positive and negative predictive values (PPV and NPV) of these diagnostic procedures.We performed a prospective study of 898 children undergoing duodenal biopsy analysis to confirm or rule out celiac disease at 13 centers in Europe. We compared findings from serologic analysis with findings from biopsy analyses, follow-up data, and diagnoses made by the pediatric gastroenterologists (celiac disease, no celiac disease, or no final diagnosis). Assays to measure IgA-TTG, IgG-DGL, and endomysium antibodies were performed by blinded researchers, and tissue sections were analyzed by local and blinded reference pathologists. We validated 2 procedures for diagnosis: total-IgA and IgA-TTG (the TTG-IgA procedure), as well as IgG-DGL with IgA-TTG (TTG-DGL procedure). Patients were assigned to categories of no celiac disease if all assays found antibody concentrations <1-fold the ULN, or celiac disease if at least 1 assay measured antibody concentrations >10-fold the ULN. All other cases were considered to require biopsy analysis. ULN values were calculated using the cutoff levels suggested by the test kit manufacturers. HLA typing was performed for 449 participants. We used models that considered how specificity values change with prevalence to extrapolate the PPV and NPV to populations with lower prevalence of celiac disease.Of the participants, 592 were found to have celiac disease, 345 were found not to have celiac disease, and 24 had no final diagnosis. The TTG-IgA procedure identified patients with celiac disease with a PPV of 0.988 and an NPV of 0.934; the TTG-DGL procedure identified patients with celiac disease with a PPV of 0.988 and an NPV of 0.958. Based on our extrapolation model, we estimated that the PPV and NPV would remain >0.95 even at a disease prevalence as low as 4%. Tests for endomysium antibodies and HLA type did not increase the PPV of samples with levels of IgA-TTG ≥10-fold the ULN. Notably, 4.2% of pathologists disagreed in their analyses of duodenal morphology-a rate comparable to the error rate for serologic assays.In a prospective study, we validated the TTG-IgA procedure and the TTG-DGL procedure in identification of pediatric patients with or without celiac disease, without biopsy. German Clinical Trials Registry no.: DRKS00003854.

Original publication




Journal article



Publication Date





410 - 419.e17


Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics, Medical Faculty of the University and University Hospital, Leipzig, Germany.


Duodenum, Humans, Celiac Disease, GTP-Binding Proteins, Transglutaminases, Immunoglobulin A, Immunoglobulin G, Gliadin, Autoantibodies, Biopsy, Serologic Tests, Sensitivity and Specificity, Prospective Studies, Predictive Value of Tests, Child, Child, Preschool, Infant, Europe, Female, Male