755 Histology and Endoscopy Discordance in Inflammatory Bowel Disease Varies by Phenotype and Location of Disease

Abstract

INTRODUCTION: Mucosal healing is currently the ultimate target in the treatment of inflammatory bowel disease (IBD). It has become increasingly evident that mucosal healing assessed during endoscopy may not imply histological healing. It is important to define which patients are more likely to have underlying microscopic inflammation in the setting of a normal appearing mucosa in order to tailor treatment approaches and minimize unnecessary biopsies. METHODS: Patients were included if they have a diagnosis of either Crohn's or ulcerative colitis, based on standard criteria and had provided consent to have clinical data and tissue collected as part of our IRB-approved IBD biorepository. Endoscopies were performed as part of routine clinical care and biopsies taken and endoscopy findings recorded in a protocolized fashion. The endoscopist rated each area biopsied as “no inflammation,” “mild,” “moderate” and “severe” inflammation. Clinical pathology results were reviewed from the matching endoscopic area. Multiple univariate and multivariate analyses with ordinal logistic regression were used to examine the relationship of endoscopic and histologic scores. RESULTS: There is an association between inflammation on endoscopy and inflammation on biopsy (OR, 2.97; 95% CI, 2.75-3.18; P < 0.0001). There is an incremental effect of the probability of having an inflamed biopsy as the degree of inflammation on endoscopy increases (mild inflammation OR, 2.50; P < 0.0001; moderate OR 4.28; P < 0.0001; severe OR 4.81; P < 0.0001). A subgroup of patients had normal appearing endoscopies but inflamed biopsies (9% of the sample). Sub group analysis revealed that biopsies taken from the neo-terminal ileum compared to native terminal ileum were two times more likely to be discordant on histology (OR 2.01; 95% CI, 1.27-3.16; P = 0.0027). CONCLUSION: The treatment goal for patients with IBD is mucosal healing. The definition of mucosal healing varies and encompasses both macroscopic and microscopic healing. We describe an overall good correlation between endoscopic inflammation and histologic inflammation. These findings are important as they may lead to a decrease in the number of biopsies taken in endoscopically normal appearing tissues. Interestingly, the neo-TI in CD patients was most frequently discordant with normal appearing endoscopy yet inflammation microscopically. Longer follow up studies are needed to determine whether patients with microscopically inflamed neo-TI have a worse outcome than those without inflammation.