Diagnostic accuracy of ten second-generation (human) tissue transglutaminase antibody assays in celiac disease

Abstract

Background: Anti-tissue transglutaminase (tTG) assays that use human tTG as antigen have recently become available. We evaluated commercially available assays with human tTG antigen to estimate their diagnostic accuracies and to determine whether they agree sufficiently to be used interchangeably. Methods: Ten commercially available second-generation anti-tTG assays were evaluated. The following populations were studied: celiac disease (CD) patients at the time of diagnosis without (n = 70) or with (n = 5) IgA deficiency; diseased controls (n = 70); and CD patients without (n = 28) or with (n = 2) IgA deficiency during follow-up. All individuals included in the study underwent intestinal biopsy. Technical performance (linearity, interference, precision, correlation, and agreement) and diagnostic accuracy (sensitivity and specificity) were compared. Anti-gliadin and anti-endomysium antibodies were also measured. Results: IgA anti-tTG results correlated well overall, but numerical values differed. Diagnostic sensitivity ranged between 91% and 97% and specificity between 96% and 100%. These were higher than the sensitivity and specificity of the IgA endomysium assay and the IgA gliadin assay. Generally, IgG anti-tTG was less sensitive but more specific than IgG anti-gliadin for the diagnosis of CD in the small group of IgA-deficient patients. Conclusions: Overall diagnostic performance of IgA tTG assays is acceptable and comparable among the different assays, but numerical values differ. Standardization is needed. © 2004 American Association for Clinical Chemistry.