Diagnostic yield of endoscopic markers for celiac disease

In the setting of open access endoscopy, the recognition of suggestive endoscopic features in the duodenum can select patients with probability of celiac disease (CD). This could add to the current efforts to increase the diagnostic rate of this disease. The aim of this study was to evaluate the diagnostic accuracy of these markers for CD in an adult population undergoing endoscopy, without a prior serological testing. Over a period of 3 years, between June 2012 and 2015, all the patients who underwent upper gastrointestinal endoscopy and presented one or more of the endoscopic markers consistent with CD, or those suspected for CD, irrespective of the presence of these markers, were included. Sensitivity, specificity, positive and negative predictive values were calculated for these markers in CD diagnosis. Among the 182 patients, 56.04% were females, with a mean age of 47.6 ± 13.9 years. 20/182 (10.99%) had a final diagnosis of CD. The presence of any endoscopic marker had a high sensitivity (95%) and a negative predictive value (98.41%). Bulb atrophy and reduced folds in the descending duodenum had a low diagnostic accuracy, while scalloping, mosaic pattern and fissures were highly specific for CD (98.77%, 99.38% and 98.77%) and their presence greatly increased the probability of CD, with a positive likelihood ratio of 24.3, 24.3 and 12.15, respectively. A wide set of endoscopic markers, including the duodenal bulb, were evaluated in this study. Our results showed that the endoscopy with a careful examination of the duodenum is a sensitive indicator for CD. CD = celiac disease, GI = gastrointestinal, VA = villous atrophy, NSAID = nonsteroidal anti-inflammatory drug, Sn = sensitivity, Sp = specificity, PPV = positive predictive value, NPV = negative predictive value, AUC = area under the curve, ROC = receiver operating characteristics, WLE = white light endoscopy, NBI = narrow band imaging, tTG = tissue transglutaminase, EMA = anti-endomysial antibodies.