High tissue-transglutaminase antibody level predicts small intestinal villous atrophy in adult patients at high risk of celiac disease

BackgroundDuodenal biopsy may be unnecessary to confirm celiac disease in patients with high tissue-transglutaminase antibody level.AimsTo define a cut-off value of tissue-transglutaminase antibody with high positive likelihood ratio for duodenal atrophy in patients with suspected celiac disease.MethodsWe retrospectively identified 945 patients with suspected celiac disease and classified according to the method used for tissue-transglutaminase antibody assay: Group A (n = 393, Eu-tTG® Eurospital), Group B (n = 263; Eu-tTG® Eurospital) and Group C (n = 289; Celikey® Phadia). Duodenal histology was graded according to Marsh. Sensitivity, specificity, and positive likelihood ratio were used to evaluate cut-off points of tissue-transglutaminase antibody as predictor of villous atrophy.Results100% specificity and ∞ positive likelihood ratio for duodenal atrophy was observed at a cut-off value of tissue-transglutaminase antibody 5 times higher than the upper limit of normal. CD diagnosis was confirmed by concordance with antiendomysial antibodies, and by reduction of t-TG titre in all patients and improvement of duodenal histology in 80% during gluten-free diet.ConclusionsTissue-transglutaminase antibody level 5-folds the upper limit of normal is 100% specific for duodenal atrophy and using this cut-off biopsy could by avoided in 1/3 of patients. Diagnostic criteria of celiac disease in adults need revision.