The significance of intraepithelial lymphocytes in the evaluation of duodenal mucosal pathology spectrum in coeliac disease

Aim: The aim of our study was to evaluate the presence of γδ+ IELs in the duodenal mucosa using TCRγδ monoclonal H-41 antibody and to determine diagnostic value of γδ+ IELs in coeliac disease (CD) with clinical and serological diagnostic difficulties. We further aimed to improve the accuracy and reliability of histopathological diagnosis of CD using discriminant analysis by combining the number of IELs determined in H&E, CD3 and TCRδ stained sections with morphometric and histopathological data. Materials and methods: The study group consisted of 167 patients who were diagnosed as CD at the Departments of Adult and Paediatric Gastroenterology, Ankara University School of Medicine, between 2010 and 2021. "Coeliac" group comprised of 117 patients with active coeliac disease (29 type 1, 29 type 2, 39 type 3) and 20 CD patients on gluten-free diet while non-celiac group consisted of 24 control cases evaluated as normal and 26 non-coeliac IELosis cases with increased IELs despite having normal villus/crypt architecture in the biopsy. Archival H&E and CD3 stained sections of duodenal biopsies of these patients were reevaluated, and all biopsies with a diagnosis of CD were reclassified according to the updated Marsh classification. Paraffin sections of duodenal biopsies were stained with anti-TCRδ antibody, a γδ T cell marker. Quantitative and morphometric evaluations were made under the light microscope and also on digital slides scanned at the digital pathology unit in the Department of Pathology established with the BAP project of AU BAP A14023003. Statistical analyses were performed using "SPSS for Windows 11.5" package program, and p<0.05 was accepted as statistically significant. Results: The number of TCRγδ+ IELs and the ratio of TCRγδ+ IELs to the total number of IELs in the "coeliac" group were significantly higher than in the "non-coeliac" group. The number of TCRγδ+ IELs showed higher performance in differentiating "coeliac" from "non-coeliac" groups from each other compared to the number of IELs detected in CD3 and H&E stained sections. The result of ROC analysis performed to determine cut-off points for distinguishing the "celiac" group from the "non-celiac" group with the highest sensitivity and specificity, revealed that the number of γδ+ IELs could be used as a discriminating criterion. The cut-off value for the number of γδ+ IELs was determined to be 9.5 yielding a sensitivity of 84.54% and a specificity of 84%. Conclusions: We observed that the number of TCRγδ+ IELs was significantly higher in patients with active coeliac disease and coeliac patients under gluten-free diet compared to non-coeliac patients, and also that the monoclonal H-41 antibody could specifically detect γδ+ IELs. We, therefore, suggest that the evaluation of γδ+ IELs using TCRδ antibody together with CD3 immunohistochemistry will enable pathologists to accurately diagnose coeliac diease, particularly when serological and histopathological findings are borderline. We, hereby, propose a 'histopathological algorithm' to be used in pathology practice for the diagnosis of coeliac disease.