P186 Faecal calprotectin (FCal): a valuable non-invasive tool in the management of IBD

FCal emerged as useful tool for IBD management, but varied assay methods, cut-offs, scenarios, phenotypes and populations may influence usefulness. Two substudies were designed: (1) To investigate the value of FCal in mucosal healing (MH) prediction (optimal cut-off, specificity, sensitivity, PPV, NPV) and thresholds for clinical activity and phenotypes and (2) to evaluate the ability of FCal monitoring in IBD in remission to predict relapse. FCal was determined with Bühlmann® ELISA in IBD patients. from a Latin-American centre. Substudy-1 (MH prediction and activity/pattern of IBD): Included 100 IBD patients: (44 UC 56 CD), who underwent routine colonoscopy (VCC) with categorisation by IBSEN score (Frøslie KF, 2007) ‘MH’(scores 0–1) and ‘non-MH’, colleting FCal samples within previous week. Optimal FCal cut-off for ‘MH’prediction (opt-MH cut-off) was calculated (ROC analysis). Substudy-2 (Prediction of relapse): included 50 UC and 50 CD in clinical remission (≥3 months), FCal: basal, ≥biannual, VCC basal/final. Analysis: Kaplan–Meier survival analysis for FCal levels above and below opt-MH cut-off. Mean follow-up 23.0 ± 11.8 months. Global definitions of clinical activity: P.Mayo (UC), HBI (CD), Location/Extent (Montreal). Substudy-1: FCal levels (Mean ± SD) in patients. with ‘MH’ were significant lower vs. ‘Non-MH’: UC (191.3 ± 174.6 vs. 621.1 ± 368.3, p = 0.0001) and CD (237.0 ± 196.9 vs. 618.5 ± 319.3, p < 0.0001) Kruskal–Wallis. Opt-MH cut-off was 242 μg/g, AUC 0.84 (95% CI 0.753–0.906) p = 0.0001, sensitivity: 76.4%, specificity: 84.5%, PPV: 85.7%, NPV: 74.5%. By clinical criteria FCal was lower (p < 0.0001) in remission vs. activity in UC (165.7 ± 14.1 vs. 630.3 ± 349.6) and CD (276.4 ± 250.1 vs. 662.1 ± 289.9), but the cut-off was higher (284 μg/g) than opt-MH cut-off. In endoscopically active CD patients, FCal levels were higher in colonic CD (851.9 ± 232.0) vs. other locations 544.4 ± 313.3 (p = 0.04). Substudy-2: Cumulative probabilities of clinical relapse at 6, 12, 18, 24 months of patients with Fcal ≥ 242 μg/g (n = 34) were 20.6%, 38.2%, 44.7%, 51.6%, and rates with FCal under cut-off (n = 66) were 1.5%, 3.1%, 5.1% and 7.9%, respectively, HR: 14.22 (95% CI 6.18–32.72), p < 0.0001, sensitivity: 85%, specificity 82.7%, PPV: 67.7%, NPV: 93.9%. Globally, relapsed 15 (30%) of UC and 12 (24%) of CD (NS). Clinical relapses with Fcal ≥ 242 were 67.7% vs. 6.1% under cut-off, p < 0.0000001, endoscopic relapses (available in 91 patients) with FCal ≥ 242: 75% (1) Fcal was a good predictor of MH in UC and CD according opt-MH cut-off (242 μg/g), (2) FCal values were significantly lower in remission vs. activity, in UC and CD, but in endoscopically active colonic CD, FCal was higher vs. other locations, (3) FCal showed to be an effective tool to predict relapse for levels above opt-MH cut-off.