Mo1680 Assessing Crohn's Disease Inflammatory Biomarker Diagnostic Accuracy Using Ileocolonoscopy or a Combined Ileocolonoscopy-CTE Score in the Embark Study

Abstract

Background: Crohn's disease (CD), a transmural inflammation of the bowel wall, presents most commonly in the terminal ileum and ascending colon, resulting in challenges in the accurate diagnosis of intestinal inflammation. Non-invasive biomarkers are attractive alternatives to ileocolonoscopy (Ico) to monitor inflammatory CD activity. We assessed the performance of two well-characterized biomarkers, fecal calprotectin and serum CRP, for diagnosis of active mucosal inflammation as measured by Ico alone or a score deduced from combination of Ico and CT enterography (CTE). Methods: 153 CD patients enrolled in the EMBARK study underwent Ico and a subset of patients (n=65) also underwent CTE. Previous intestinal resection rates and disease activity were similar between the groups (resection in 52% and median HBI of 8 in both groups). Both Ico and CTE were scored separately by central read then a novel scoring system that incorporates disease activity visible by either Ico or CTE was developed by both central readers to jointly assess disease activity in IcoCTE patients. Disease severity and length scores were assigned for individual bowel segments based on ulcer size, hyperenhancement and wall thickening. ELISA was used to measure CRP and calprotectin. Results: Active CD was observed in 22% (34/153) of patients by Ico alone using ulcerated area ≥2 and 34% (52/153) of patients using SES-CD>4. CRP was significantly higher in patients using either definition of CD activity, while calprotectin was increased only when SES-CD>4 was used to identify active CD. In the cohort of patients where both Ico and CTE were performed, more patients were observed to have active CD by Ico-CTE score ≥4 (42 of 65, 65%) than by Ico alone using ulcerated area (9 of 65, 14%) or SES-CD>4 (13 of 65, 20%). Patients with ileal disease by Ico-CTE were likely to have missed CD activity by Ico alone (12 out of 22, 55%). Calprotectin was significantly different at the 1% level between active and inactive CD patients when the combined IcoCTE was used (p=0.004) but not by Ico alone using ulcerated area (p=0.26) or SES-CD≥4 (p=0.06). Use of the combined Ico-CTE score also increased specificity of calprotectin in detecting inflammation using a 100 μg/g cutoff. Calprotectin performed similarly in detecting CD activity in the ileum as ileocolonic or colonic disease (77/79 vs. 68/85 sensitivity/ specificity) measured by Ico-CTE. Conclusion: The use of a novel Ico-CTE scoring system increases the detection of mucosal inflammation, particularly in patients with ileal disease. Calprotectin had increased performance in active CD patients when mucosal activity was measured by a combination of Ico and CTE. Ico alone is likely to underestimate the number of active CD patients and therefore the performance of calprotectin as a biomarker of inflammation.