Crohn's Disease Features in Anastomotic Biopsies from Patients With and Without Crohn's Disease: Diagnostic and...

Endoscopic evidence of disease activity is a critical predictor of clinical relapse in patients with Crohn's disease (CD), and histologic disease activity is evolving as a similarly important end point for patient management. However, classical morphologic features of CD may overlap with postoperative inflammatory changes, confounding the evaluation of anastomotic biopsies. There is a clear unmet need for better characterization of diagnostic and clinically significant histologic features of CD in these surgically altered sites. We evaluated ileocolonic and colocolonic/rectal anastomotic biopsies performed at 3 academic institutions in patients with and without CD. The biopsies were blindly assessed for CD histologic features and correlated to clinical and endoscopic characteristics. In CD patients, the presence of each feature was correlated with the subsequent clinical exacerbation or relapse. We obtained anastomotic biopsies from 208 patients, of which 109 were operated on for CD and 99 for another indication (neoplasia [80%], diverticular disease (11%), and other [9%]). Mean time since surgery was 10 years (0-59; 14 years for CD [1-59], 6 years for non-CD [0-33]). Endoscopic inflammation was noted in 52% of cases (68% for CD and 35% for non-CD). Microscopic inflammation was present in 74% of cases (82% for CD and 67% for non-CD). Only discontinuous lymphoplasmacytosis (P < .001) and pyloric gland metaplasia (P = .04) occurred significantly more often in CD patients. However, none of the histologic features predicted clinical disease progression. In subset analysis, the presence of histologic features of CD in nonanastomotic biopsies obtained concurrently in CD patients was significantly associated with relapse (P = .03). Due to extensive morphologic overlap between CD and postoperative changes and the lack of specific histologic features of relapse, biopsies from anastomotic sites are of no value in predicting clinical CD progression. Instead, CD activity in biopsies obtained away from anastomotic sites should be used for guiding endoscopic sampling and clinical management.