Objectives: The goal of our study was to determine whether visual assessment of the esophagus and stomach could predict abnormal histology and determine the frequency of interventions based on biopsies in patients undergoing endoscopy for elevated tissue transglutaminase immunoglobulin A antibody (TTG).
Methods: Pathology records were searched for patients with biopsy performed for elevated TTG. Pathology report, endoscopy report, and follow-up were obtained and slides from the duodenum reviewed. Pathology was considered gold standard for sensitivity and specificity calculations.
Results: 240 patients were included. 215 patients had esophageal biopsies performed. Esophageal endoscopic visual assessment had sensitivity of 47% and specificity of 93% for abnormal histology. 16(7%) patients had therapy or referral related to results and, of these, 6(38%) had visually normal endoscopy. 237 biopsies were performed of stomach. Gastric endoscopic visual assessment had a sensitivity and specificity of 20% and 87%. 24(10%) patients had therapy based on findings and, of these, 12 (50%) had visually normal endoscopy.
Conclusions: Endoscopic assessment of esophagus and stomach has low sensitivity and high specificity for pathologic abnormalities when indication for endoscopy is elevated TTG. When endoscopy is visually normal clinical interventions based on biopsy are rare, and foregoing biopsy may be considered.
Keywords: GI; Pediatric; celiac disease; esophagus; mucosal biopsies; stomach.