Abstract
The intraepithelial lymphocyte (IEL) thresholds in diagnosing celiac disease (CD) have limited diagnostic value and low specificity. This study investigated the predictive value of localization-based examination and counting of IEL in diagnosing CD. Duodenal biopsies (n = 418) from pediatric and adult age groups who applied in 2021–2023 were re-examined. The mean number of total and supranuclear lymphocytes per 100 (IEL/100, SupIEL/100) and 20 (IEL/20, SupIEL/20) enterocytes in the villus tip were counted with CD3-stained slides. The SupIEL ratio was obtained by SupIEL/Total IEL. ROC analysis was used to determine thresholds that predicted the diagnosis of CD. The distribution pattern was defined as basal, mixed, and apical, based on the part where the IELs were found in a proportion of ≥0.60 according to the enterocyte nucleus. The most common pattern in the CD group (n = 195) was mixed (n = 127, 65.1%), while the basal pattern (n = 213, 95.5%) constituted the majority of non-CD samples (n = 223), despite some of which had increased IEL (n = 69, 31.0%). All (100.0%) biopsies exhibiting an apical pattern were consistently in the CD group, while 98.6% of those exhibiting a basal pattern were in the non-CD. The positive and negative predictive values of SupIEL/100 > 16, SupIEL/100 ratio > 0.40, SupIEL/20 > 3, and SupIEL/20 ratio > 0.33 were 95.8/95.1, 94.4/95.1, 97.0/98.2, and 75.2/95.8, respectively. In the multivariable logistic regression analyses, the SupIEL/100 ratio > 0.40 was found to be an independent parameter predicting CD. This study first revealed that examining duodenal tissues focused on IEL localization provided a more reliable and accurate prediction of CD than all previously proposed ones.
Get full access to this article
View all access options for this article.