Antiendomysium versus antigliadin antibodies in screening the general population for coeliac disease

Background: It has recently been shown that mass screening for coeliac dise ase, using either the serum antigliadin (AGA) or antiendomysium antibodies (EMA) as screening test, can detect large numbers of cases that had escaped clinical diagnosis. The influence of the diagnostic algorithm on the resul ts of the coeliac screening has not yet been evaluated. Our aim was to comp are the validity of the AGA and the EMA protocols in 2096 students living i n northwest Sardinia, who took part in a serologic screening for coeliac di sease. Methods: The sample included 2096 of 2345 eligible students (89%) ag ed 11-15 years who underwent serum IgG AGA, IgA AGA, and IgA EMA determinat ions. Total serum IgA level was measured in sera showing isolated IgG AGA p ositivity. Subjects showing at least one of the following: a) EMA positivit y, b) IgA AGA positivity, or c) IgG AGA positivity and IgA deficiency (<5 m g/dl) were asked to submit to a small-intestinal biopsy. Results: The preva lence of coeliac disease was 19 (16 showing typical enteropathy, 1 potentia l case, and 2 known cases) of 2096 (0.91%; 95% confidence interval = 0.50-1 .31). Seventeen small-intestinal biopsy specimens were needed to confirm 16 cases of manifest coeliac disease (positive predictive value (PPV) = 94%) by the EMA protocol, whereas the AGA protocol required 21 biopsy specimens for 12 cases of coeliac disease (PPV = 57%). None of six IgA-deficient, IgG AGA-positive cases detected by the AGA protocol also had coeliac disease. Conclusions: The EMA protocol is superior to the AGA protocol for mass scre ening of coeliac disease because of higher sensitivity, decreased need for intestinal biopsy, and possibility to detect potential cases of coeliac dis ease.