Severe eosinophilic infiltration in colonic biopsies predicts patients with ulcerative colitis not responding to medical therapy

Aim Eosinophils are potent proinflammatory cells that are involved in the pathogenesis of ulcerative colitis (UC). We evaluated the infiltration of eosinophils into the lamina propria in patients with active and inactive ulcerative colitis (UC) and investigated its clinical significance, among other variables, in predicting the outcome of medical treatment in active disease. Method We studied colorectal biopsy specimens from 18 UC patients with disease in long-standing remission, from 22 patients with active disease who responded to therapy (12 with complete response and 10 with partial response) and from 10 patients who were nonresponders. Demographic information was obtained at baseline, and clinical, endoscopic and laboratory data were obtained at baseline and 12 weeks post-treatment. We evaluated five histological features: mucosal ulceration; mucosal erosions; crypt abscesses; cryptitis; and eosinophilic infiltration of the lamina propria. The severity of these lesions was graded as: none or minimal; mild; moderate; or severe. Statistical analyses were performed between responders and nonresponders for differences in demographic, clinical, laboratory, endoscopic and histological parameters. Results Laboratory, endoscopic and histological parameters were significantly improved after treatment only in the complete responders group. Analyses of baseline data revealed no significant differences in parameters between complete or partial responders and nonresponders, except for a less severe eosinophilic infiltration of lamina propria in complete responders (P < 0.05). Multiple logistic regression analysis showed that severe eosinophilic infiltration in colonic biopsies was the most significant predictor of poor response to medical therapy. Conclusion Assessing the severity of eosinophilic infiltration in the lamina propria of colonic biopsies in patients with ulcerative colitis could be a valuable predictive tool of response to medical therapy.