Subjects found to be IgA-deficient and HLA DQ2- or DQ8-positive were also requested to undergo duodenal biopsy. Upper gastrointestinal endoscopy was carried out by a gastroenterologist after informed consent and four biopsies were obtained from different sites of the second and third parts of the duodenum, fixed in 10% buffered formalin NVP-BGJ398 and processed using hematoxylin–eosin staining. Detailed histological evaluation of the biopsies was carried out according to the modified Marsh criteria22 by a pathologist who was blinded both to clinical status and results of screening. Histology showing ‘infiltrative lesion’ with
intra-epithelial lymphocytosis was taken as Marsh I, ‘infiltrative-hyperplastic lesion’ as Marsh II and ‘villous
atrophy’ in addition as Marsh III (partial-IIIa, subtotal-IIIb, total-IIIc). Any first-degree relative who was serology-positive and had Marsh III (villous atrophy) changes on small bowel histology was labeled as a new CD case. Appropriate dietary counseling with the initiation of a gluten-free diet was provided for these relatives and they are currently see more in regular outpatient follow up. Quantitative variables were expressed as median and range. Percentages and proportions were calculated using the standard formulae. The study was approved by our Institutional Ethics and Research Committee. Thirty children (14 boys, median age 9.5; range 3–17 years) with CD were enrolled as index cases. All were IgA-anti-endomysial-antibody-(EMA)-positive, had histology suggestive of CD (Marsh IIIa 12, Marsh IIIb 18) at diagnosis and all had shown a definite response to a gluten-free diet. HLA typing of index CD cases and their first-degree relatives is given in Table 1. There were a total of 94 first-degree relatives (60 parents, 34 siblings) of these index cases and of these, 96.8% were enrolled in the study. Three fathers could not be enrolled because one was not alive, one
was staying abroad and the third was hospitalized for pancreatitis. Of the 91 first-degree relatives evaluated, 57 were parents (27 fathers [median age 38; range 29–53 years], 30 mothers [32 (25–48) years] and 34 were siblings HSP90 [22 brothers [8.5 (1–23) years]; 12 sisters [9.5 (3–24) years]).Among the first-degree relatives, 85.7% were HLA DQ2-positive and 14.3% were DQ2-negative. None were DQ8-positive. The prevalence of DQ2 positivity was similar in parents (86%) and siblings (85.3%) as shown in Table 1. The total IgA level was normal in 89 first-degree relatives and low in two subjects (one father, one sister). Both IgA-deficient first-degree relatives were asymptomatic and HLA DQ2-positive. IgA-tTGA was positive in nine first-degree relatives and of these, six were strongly positive (> 100 U/ml) as shown in Table 2. Symptoms were significantly more common in IgA-tTGA-positive (4/9) first-degree relatives than IgA-tTGA-negative relatives (2/82; P < 0.