P272. Can fecal calprotectin (FC) predict activity in the MR-enterography in small bowel Crohn's disease (CD)?
E. Cerrillo1, B. Beltrán2, A. Echarri3, J.C. Gallego4, J. Pamies5, M.L. Martinez6, M. Iborra2, G. Bastida2, B. Laiz6, P. Nos2, 1IIS Hospital La Fe, Gastroenterology Unit, Valencia, Spain, 2Hospital La Fe and CIBERehd, Gastroenterology Unit, Valencia, Spain, 3Complejo Hospitalario Universitario de Ferrol, IBD Unit. Gastroenterology, Ferrol, Spain, 4Complejo Hospitalario Universitario de Ferrol, Radiology Department, Ferrol, Spain, 5Hospital La Fe, Radiology Department, Valencia, Spain, 6Hospital La Fe, Clinical Analysis Department, Valencia, Spain
Magnetic resonance imaging (MRI) is becoming an important tool in the assessment of the small bowel Crohn's disease (CD). Among the non-invasive markers, fecal calprotectin (FC) has better correlation with endoscopic and histologic activity than classical serum markers such as C-reactive protein (CRP) in colonic involvement. Aim: To evaluate the correlation between FC levels and disease activity on MRI according to a quantitative Magnetic Resonance Index of Activity (MaRIA), and between FC and a histologic activity score, in small bowel CD.
120 patients (59F, 64M, median age:46; range:18–74) with a proven diagnosis of small bowel CD were analyzed in a prospective database to obtain clinical data, serological and fecal biomarkers (CRP and FC), radiological and histological variables. Clinical activity was evaluated by Harvey–Bradshaw Index (HBI), and FC was measured by ELISA (Calprest®, Eurospital). MRI activity was measured by a quantitative previously validated score (MaRIA). Other MRI parameters evaluated were: wall thickening and length of the affected bowel, presence of edema/ulcers, bowel wall enhancement, curve pattern in dynamic study and presence of abscesses/fistulas. Chiorean index was used when surgical procedure was performed.
According to HBI, 75 patients were in clinical remission and 45 had clinical active disease. MaRIA score was significantly associated with FC levels (p < 0.01). Overall correlation between MaRIA and FC was moderate (Spearman's r = 0.45, p < 0.001). FC discriminates inflammatory activity on MRI with an area under the ROC curve of 0.875 (CI 0.794–0.956, p = 0.001). A cut-off value of 166 µg/g had an 83% sensitivity, 76% specificity, 85% PPV and 73% NPV for the diagnosis of inflammatory activity on MRI. Among other MRI parameters evaluated, FC was also significantly associated with the presence of ulcers (p = 0.017), and the curve pattern in dynamic study, differentiating between fibrotic and inflammatory pattern (p = 0.006). No correlation between FC levels and HBI or CRP was detected. In 28 of these patients a resective surgical procedure was performed (66 small bowel lesions were identified). Surgical pathology activity has a good relationship with FC with significance for moderate (p = 0.03) and severe (p = 0.01) Chiorean index. No relationship was detected for CRP.
FC predicts inflammatory activity on MRI, and it could be useful to select candidate patients to its performance, prior to clinical decision making. Thus, levels of FC (cut-off around 170 µg/g) could be useful in clinical practice to screen for performing radiologic studies and support clinical decision making. FC is the only marker associated with pathologic intestinal damage.