Audit of the Clinical Utility of Antibodies to Endomysium and Gliadin as Markers of Coeliac Disease

Abstract

The diagnosis of coeliac disease is dependent on the demonstration of gluten sensitive small bowel enteropathy.' Circulating antibody markers for coeliac disease are a useful adjunct in the selection of patients for biopsy as well as in subsequent monitoring of compliance with a gluten free diet.2 As a result of increased awareness of the protean manifestations of coeliac disease and the consequences of misdiagnosis, immunology laboratories are receiving an increasing number of requests for coeliac disease antibody screening. There is much information on the clinical utility of antibody markers in selected patient populations attending gastroenterology clinics. However, there is limited information on the performance of coeliac antibodies for unselective screening of hospital patients. We report the results of an audit project conducted over a period of 18 months in a teaching hospital immunology laboratory. METHODSOver an 18-month period we audited a consecutive unselected group of 480 patients referred for a coeliac disease screen. Demographic and clinical details including history of abdominal pain, diarrhoea, weight loss, arthralgia, family history of coeliac disease and IgA deficiency were requested by questionnaire. The gold standard for the diagnosis of coeliac disease was taken as a small bowel biopsy demonstrating villous atrophy.Endomysial antibodies (EMA) were detected by standard indirect immunofluorescence performed on unfixed sections of monkey oesopha- RESULTSFrom the following specialties 275/480 questionnaires were returned (response rate 57.3%): paediatrics 137; gastroenterology 33; haematology 38, and general medicine 67. Of the 275 patients 72 (56 adults, 16 children) underwent small intestinal biopsy yielding 28 biopsy-proven cases of coeliac diseases. The 203 patients who did not undergo biopsy included 185 patients negative for both EMA and AGA and 18 patients positive for only AGA. Table 1 shows the sensitivity, specificity and positive and negative predictive values of our results (based on 72 cases biopsied) in comparison to two other larger studies based on 144 and 96 patients, respe~tively.~,~ Despite the use of highly selected patient groups, these studies were chosen since they reported on the clinical utility of both antibody isotypes (IgG and IgA) of AGA and EMA. DISCUSSIONOverall, our results confirm the value of IgA EMA as a highly sensitive and specific marker of untreated coeliac disease. The high positive 436