Value of duodenal endoscopic markers of villous atrophy

Several endoscopic aspects of duodenal mucosa have been described in patients with villous atrophy (VA) and celiac disease. We assessed the accuracy of three endoscopic markers in patients referred for duodenal biopsy.

From January 1992 to June 2000 all patients undergoing upper gastrointestinal (UGI) endoscopy by a single operator (n = 4102), had the second part of the duodenum examined, without any magnification technique or addition of vital dye staining, for the presence of three markers of VA, (1) reduction or loss of Kerkring's folds, (2) scalloped configuration of the duodenal folds, and (3) mosaic appearance of the mucosa. Biopsy samples were taken from those in whom the mucosa was abnormal (n = 51), and from those in whom it was normal but where biopsy was indicated (n = 230).

None of the 230 endoscopically normal controls had VA. Seven out of 51 patients (14%) with an abnormal mucosa had a normal histology, the remaining 44 (86%) had some degree of VA confirmed at histology, partial VA in 11 patients (25%) and subtotal or total VA in 33 patients (75%). 63% of the patients with only one endoscopic marker had VA confirmed at histology whereas the coexistence of more than one marker was always associated with histological villous atrophy. The overall sensitivity, specificity, positive and negative predictive values of the endoscopic markers were respectively, 100%, 97%, 86%, and 100%.

The appearance of the second part of the duodenal mucosa as assessed during routine UGI endoscopy, is a sensitive and specific indicator of the presence of VA on biopsy, and should alert the endoscopist to the possibility of a celiac disease.