Systematic review: the use of serology to exclude or diagnose coeliac disease (a comparison of the endomysial and tissue transglutaminase antibody tests).

BACKGROUND: With the appreciation of the high prevalence of coeliac disease there is increasing use of serology in screening asymptomatic people and testing those with suggestive features. AIM: To compare the sensitivities and specificities of the endomysial antibody and the tissue transglutaminase antibody tests.
METHODS: Using electronic databases a search was made for relevant papers using the terms tissue transglutaminase and endomysial antibody.
RESULTS: Both the endomysial antibody and tissue transglutaminase antibody have very high sensitivities (93% for both) and specificities (>99% and >98% respectively) for the diagnosis of typical coeliac disease with villous atrophy. Human recombinant tissue transglutaminase performs much better than guinea pig tissue transglutaminase. Review of studies comparing endomysial antibody with human recombinant tissue transglutaminase antibody shows that endomysial antibody more often has a higher specificity and human recombinant tissue transglutaminase antibody more often has a higher sensitivity.
CONCLUSION: The human recombinant tissue transglutaminase antibody is the preferred test for screening asymptomatic people and for excluding coeliac disease in symptomatic individuals with a low pretest probability (i.e. <25%) for coeliac disease. Furthermore, it has a number of practical and financial advantages. If the pretest probability is >25%, biopsy is preferred as the post-test probability of coeliac disease with a negative test is still >2%.