Endomysial and tissue transglutaminase antibodies in coeliac sera. A comparison not influenced by previous serological testing

BACKGROUND: Since transglutaminase was shown to be the antigen of endomysial antibodies (EMA), it has become possible to screen for coeliac disease (CD) with an enzyme-linked immunosorbent assay (ELISA) for tranglutaminase antibodies (TTA). However, it is possible that sera used to show that TTA are found in CD were obtained from patients diagnosed because they were positive for EMA. So, a comparison between EMA and TTA has not been possible so far. METHODS: EMA and TTA were tested in sera from 52 controls and 56 untreaded CD patients, who had not undergone serological testing. Samples were tested for TTA with and ELISA kit. Based on the ROC analysis of a pilot study, results were considered as either positive, borderline, or negative. EMA were analysed by indirect immunofluorescence on monkey oesophagus. RESULTS: Forty-nine CD patients were positive for TTA, six borderline, one negative. Forty-four controls were negative, seven borderline, one positive. If we consider borderline results to be positive, sensitivity is 98.2% and specificity 84.6%. EMA were positive in 53 CD patients; the controls were all negative. Performing TTA in all cases and EMA only in the few TTA borderline cases (12.0%) would have a sensitivity of 94.6% and a specificity of 98.1%. CONCLUSIONS: This study is the first to compare TTA with EMA. Due to 100% specificity and high sensitivity, EMA seems to be the most accurate coeliac antibody. Conversely, TTA offer advantages in terms of sensitivity and simplicity. A satisfactory strategy is to use TTA first and then EMA to confirm the borderline results.