Gastric metaplasia: A frequently overlooked feature of duodenal biopsy specimens in untreated celiac disease

Background: Duodenal gastric metaplasia is rarely reported in untreated cel iac disease, although it is seen in 60% to 100% of duodenal biopsy specimen s in nonceliac patients with histologic duodenitis. The low incidence could represent underreporting, a decreased incidence in pediatric patients gene rally, or the more distal sampling site that is customary for most biopsy s pecimens that are obtained to diagnose celiac disease. It could also be a u nique feature of the inflammatory reaction that characterizes this disease. The purpose of this study was to examine the incidence of gastric metaplas ia in duodenal specimens from children with untreated celiac disease with s pecial reference to patient age and biopsy site. Method: Formalin fixed paraffin-embedded specimens of duodenal mucosa were selected from the pathology department's archival material. Sections were e ither stained histochemically or by immunochemical methods, according to an antigen-retrieval protocol. Forty-four duodenal specimens from untreated p atients with celiac disease (n = 22) and control subjects of similar age wi th normal histology (n = 22) were examined. Ten of each were obtained durin g upper endoscopy from the proximal duodenum (proximal site) and 12 of each by Crosby capsule near the ligament of Treitz (distal site). Results: All specimens from patients with celiac disease exhibited marked v illous atrophy. None had been noted to have gastric metaplasia during routi ne examination of sections stained by hematoxylin and eosin. Fifteen (68%) of 22 of the celiac specimens and 2 of 22 (9%) control specimens contained gastric metaplasia, identified as patches of gastric-type cells containing MUC5AC (gastric mucin), pS2 (gastric trefoil factor) and neutral (periodic acid-Schiff-positive) mucin. Five of the seven celiac specimens that had no metaplasia showed increased numbers of goblet cells expressing gastric mar kers. The incidence of gastric metaplasia was not different for endoscopic (70%) or capsule (67%) specimens. Sixty-eight percent (7/11) of patients ag ed less than 3 years had gastric metaplasia. Conclusion: The presence of gastric metaplasia has been previously underrep orted in celiac disease specimens. Detection would be improved by the routi ne use of period acid-Schiff/alcian blue staining. The incidence of gastric metaplasia in celiac disease is not significantly influenced by biopsy sit e or age at time of the biopsy.