P143 Gastrointestinal ultrasound is an accurate marker for endoscopic activity in Crohn’s disease

Abstract

Background Treatment goals for Crohn’s disease (CD) have changed from symptom control to objective endpoints, preferably endoscopic remission. However, as the relative invasiveness of ileocolonoscopy limits repeated examinations, a surrogate marker of endoscopic remission is needed. Gastrointestinal ultrasound (GIUS) is accurate in detecting CD, and we aimed to explore its ability to distinguish between patients with active disease from patients in remission. Methods One hundred and forty-five CD patients scheduled for ileocolonoscopy were prospectively included. The endoscopic disease activity was quantified using the Simple Endoscopic Score for Crohn’s disease (SES-CD), and endoscopic remission was strictly defined as SES-CD=0. Ultrasound remission was defined as wall thickness <3 mm (<4 mm in the rectum). Additionally, SES-CD was compared with colour Doppler, Harvey Bradshaw’s index (HBI), C-reactive protein (CRP) and calprotectin. Twenty-three patients were examined by two investigators for interobserver assessment. Results 102 had active disease and 43 patients were in remission. GIUS yielded a sensitivity of 92.2% and specificity of 86% for wall thickness and sensitivity 66.7% and specificity 97.7% for colour Doppler. For HBI sensitivity was 34.3% and specificity 88.4%, CRP sensitivity 35.7% and specificity 82.9% and calprotectin sensitivity 55.9 % and specificity 82.1%. The interobserver analysis revealed excellent agreement (k=0.90). Conclusion GIUS has a high sensitivity for detecting endoscopic activity, but poorer specificity. In a hospital cohort of patients with Crohn’s disease, the number of invasive procedures could be reduced by reserving ileocolonoscopy for patients with a normal ultrasound for identification of patients in remission.