Objective The anti-endomysial antibody (EMA) immunofluorescence test is recommended by some clinical guidelines as an adjunct to immunoglobulin A anti-tissue transglutaminase (TTG) antibody in coeliac disease (CD) diagnosis. However, dual testing can complicate interpretation. We assessed the diagnostic value of EMA alongside TTG. Method We retrospectively analysed patients tested for TTG and EMA between 2018 and 2019 at a single laboratory. 587 patients were categorised into two groups: TTG 7–10 U/mL (‘borderline’) and TTG >10 U/mL. CD was confirmed by duodenal biopsy; if unavailable, TTG ≥100 U/mL (10×upper limit of normal) was diagnostic, serologically. Results Biopsies were performed in all borderline cases (n=41). Among 23 EMA-negative patients, 17 (74%) had biopsy-confirmed CD. 12 of 18 (67%) EMA-positive patients had CD. EMA showed poor sensitivity (0.4138) and specificity (0.5000). Fisher’s exact test found no association between EMA and biopsy results (p=0.7341), indicating EMA’s limited diagnostic value in borderline cases. In the TTG >10 U/mL group (n=546), 523 were EMA-positive (511 with CD, 98%). Of 23 EMA-negative patients, 14 (61%) had CD. Despite a statistically significant association with CD (p<0.0001), EMA added minimal clinical value (sensitivity 0.9733; specificity 0.4286). Conclusion EMA had low sensitivity in borderline cases and added no diagnostic value when TTG was positive. We recommend discontinuing EMA testing in clinical practice. The high rate of CD in borderline cases suggests that current TTG diagnostic thresholds need reconsideration. Prospective, multicentre replication using standardised assays is required to evaluate whether biopsies are warranted at lower TTG values than currently recommended.
Yau et al. (Fri,) studied this question.