Diagnostic role of upper gastrointestinal endoscopy in pediatric inflammatory bowel disease.

BACKGROUND: Discrimination between ulcerative colitis (UC) and Crohn disease (CD) may be difficult on ileo-colonoscopy alone because of a lack of definitive lesions. Retrospective studies show upper gastrointestinal endoscopy may be helpful in confirming diagnosis in such cases.

AIMS: To prospectively determine importance of upper gastrointestinal endoscopy in diagnosis of inflammatory bowel disease (IBD) and assess factors predictive of upper gastrointestinal involvement in IBD.

METHODS: All pediatric patients were enrolled prospectively and consecutively over a 2-year period and investigated with an ileo-colonoscopy and barium meal follow-through. Children with procto-sigmoiditis, later confirmed histologically to be typical of UC, were excluded from the study. The remainder underwent upper gastrointestinal endoscopy. The protocol and methodology were determined a priori.

RESULTS: 65 children suspected of IBD underwent colonoscopy. Of the total, 11 had recto-sigmoiditis with typical macroscopic appearances of UC; once this was confirmed on histology these patients were excluded from the study. Of the 54 children (males, 31; median age, 11.1 years) remaining, 23 were initially diagnosed with CD on ileo-colonoscopy and 18 (33%) were diagnosed with UC. The diagnosis remained ambiguous in 13 (six colonic, four ileo-colonic, three normal colon) on clinical, radiologic and histologic grounds. Upper GI endoscopy helped to confirm CD in a further 11 (20.4%). Two patients were diagnosed with indeterminate colitis. Upper gastrointestinal inflammation was seen in 29 of 54 (22 CD; 7 UC ). Epigastric and abdominal pain, nausea and vomiting, weight loss and pan-ileocolitis were predictive of upper gastrointestinal involvement (P < 0.05). However, 9 children with upper gastrointestinal involvement were asymptomatic at presentation (31%). Overall upper gastrointestinal tract inflammation was most common in the stomach (67%), followed by the esophagus (54%) and duodenum (22%).

CONCLUSIONS: Upper gastrointestinal tract endoscopy should be part of the first-line investigation in all new cases suspected of IBD. Absence of specific upper gastrointestinal symptoms do not preclude presence of upper gastrointestinal inflammation.