Antiendomysium antibodies and coeliac disease: solved and unsolved questions. An Italian multicentre study

A total of 3783 subjects were enrolled to compare IgA and IgG gliadin antibodies (AGA) with IgA endomysium antibodies (EMA) in coeliac disease (CD). Among 688 children with untreated CD EMA were positive in 93.8%, IgA AGA in 84.9% and IgG AGA in 90.2%. AGA, but not EMA, sensitivity decreased with age. EMA were present in 3.8% of control subjects, IgA AGA in 14.9% and IgG AGA in 34.3%. Follow-up of 5 of 39 EMA-positive controls showed flat mucosa. Combined determination of EMA and AGA showed an increased predictive value: if EMA and AGA were both positive, the mucosa was flat in 99.1%, if both were negative, the mucosa was normal in 99.1%. After a gluten-free diet (GFD), IgA-AGA disappeared first. Among 21 patients not on a strict GFD and in 194 coeliac patients after challenge, EMA, but not AGA, were always positive. Among 67 first-degree relatives of coeliacs, the positive predictive value of EMA was 90.6%, IgA AGA 74.3% and IgG AGA 44.6%. In conclusion, EMA screening is an excellent test for the diagnosis and follow-up of CD, and for identification of its silent and latent forms.