Early diagnosis of celiac disease in IgA deficient children: contribution of a point-of-care test

BACKGROUND: The serological diagnosis of celiac disease (CD) often relies on the presence of anti-tissue transglutaminase (tTG) IgA autoantibodies. Patients suffering from selective IgA deficiency (IgAD) are often not aware of their IgA deficiency and are tested as CD negative, delaying considerably the diagnosis. The detection of IgG against deamidated gliadin peptides (DGP) has high specificity and better sensitivity than IgG anti-tTG. A multi-analytic lateral-flow immunochromatographic assay (CD-LFIA) based on the detection of IgA and IgG anti-DGP and total IgA was shown to have a good diagnostic accuracy for CD. The aim of this study was to evaluate the clinical accuracy of its use in children suffering from IgAD. METHODS: 45 IgAD children ranging from 1.1 to 17.4 years and suspected of CD or having high CD risk factors were referred from outpatient clinics located in the area of Rhone-Alpes (France) to the Hospices Civils de Lyon, Paediatric Hospital-Gastroenterology-Hepatology- Nutrition Department for further CD investigations. The CD investigations, including the sample collection, were performed within the Paediatric Hospital-Gastroenterology-Hepatology- Nutrition Department, and the serological testing was performed at the Lyon-Sud Hospital-Immunology Laboratory. The diagnosis of CD was based on IgG anti-tTG serology, biopsy results and patient follow-up. The serum samples were retrospectively tested on the CD-LFIA test. RESULTS: A total of eight (8) patients were diagnosed as new CD. All were correctly identified by the CD-LFIA. The test yielded four (4) false positive results. Two patients with positive IgG anti-tTG were negative on CD-LFIA, but were classified as CD negative based on biopsy results and patient follow-up. The remaining 33 patients were found negative by both methods. The specificity and sensitivity of CD-LFIA was of 89.2% [74.6-97.0] and of 100% [63.1-100] respectively. The negative predictive value (NPV) was of 100% [89.4-100], and the Likelihood Ratio for Negative Test (LR-) was of 0 [0.0-0.91]. CONCLUSIONS: CD-LFIA is a useful, non-invasive and rapid tool to rule out CD in primary care paediatric patients having CD-related symptoms and IgAD. Patients having a positive CD-LFIA result could be then readily directed to secondary care setting for further evaluation by standard serology and biopsy.