A 2-year follow-up study in children with positive coeliac disease-specific serum antibodies and architecturally normal small intestinal mucosa

Abstract

Background and aim. In recent years the presence of antiendomysium EMA) and anti tissue transglutaminase 2 (anti-TG2) with a normal duodenal ucosa has been defined as potential coeliac disease (CD). Aim of our study as to characterise the natural history of potential CD. Patients.Two cohorts of children were recruited: 30 in the Pediatric Unit f Cava de’ Tirreni Hospital and 32 in the Department of Pediatrics at the niversity Federico II in Naples. Forty-two patients were Marsh T0 and 20 1. In all serum anti-TG2 and/or EMA were positive. All carried HLA DQ2 nd/or DQ8. The mean age was 6.7 years (range: 1.6–14.6). Methods. Growth and nutritional parameters, serology, autoimmunity, aecal calprotectin and intestinal permeability were assessed every 6 months. atients with persistent EMA and/or anti-TG2 even if not symptomatic were e-biopsied after 2 years. Results. The incidence of potential CD was 9.5–12%. Eight patients ere symptomatic and began a gluten-free diet. Symptoms resolved in 3/8 atients. All other patients had a normal daily gluten intake.13/62 potential D patients were first-degree relatives, 12/62 were affected by autoimmune iseases and 3/62 were asymptomatic. In all other cases gastrointestinal omplaints resolved with appropriate medical therapy. 33/54 patients entered the follow-up. After 2 years 17/33 patients were erologically negative, but two developed autoimmune disorders (diabetes, hyroiditis). During the follow-up biopsies were taken from 7 patients all MA and/or anti-TG2 positive. In 4 patients the second biopsy after 2 years nd in one patient the third biopsy after 5 years was still normal. Two patients eveloped partial villous atrophy, one at the second biopsy after 2 years and nother at the third biopsy after 5 years. Clinical and nutritional evaluation as normal in all patients except in one of the two patients who developed illous atrophy. In 9 patients we correlated faecal calprotectin, intestinal pereability and serology. Faecal calprotectin and intestinal permeability were ignificantly correlated (R= 0.71). Interestingly, in patients with positive MA and anti-TG2 >10 UA/ml (nv < 5) intestinal permeability and faecal alprotectin were abnormal; furthermore, these patients remained serologially positive during the follow-up and one developed partial villous atrophy. n patients with negative EMA and anti-TG2 < 10 intestinal permeability and aecal calprotectin were normal; during the follow-up they remained seroogically negative. Conclusions. Incidence of potential CD is increasing in the last years 3.8–17%). This condition is mainly seen in first-degree relatives (20%) and atients with autoimmune disorders (19%). In our cohort EMA and anti-TG2 eturned to normal in 50% of cases, but 6% developed other autoimmune iseases. Duodenal mucosa can remain normal for several years despite he presence of serum EMA and anti-TG2. Finally, the presence of both MA and anti-TG2 in addition to abnormal intestinal permeability, clinical ymptoms and faecal calprotectin seems to be strongly predictive of the volution to overt coeliac disease.