P269 Correlation between soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) expression and endoscopic activity in intestinal Behcet's disease

Abstract

Background: It is increasingly recognised that there is a disconnect between symptoms and underlying disease activity in Crohn’s disease (CD). Endoscopy and MRI are the gold standards for assessment of disease activity in CD. Faecal calprotectin (FC) is established as an excellent non-invasive biomarker of intestinal inflammation. We aimed to assess the relationship between clinical symptoms and FC in patients with CD. Methods: This was a retrospective study of all patients with established CD who had one or more FC performed (from the Edinburgh Faecal Calprotectin Register ECFR). Data were collected by detailed review of electronic and written records. Baseline demographics, biochemical parameters, Harvey Bradshaw Index (HBI) and disease related hospitalisation were recorded for each patient where available. Clinical remission was defined as HBI 4, with 5 as active disease, corresponding to CDAI 150 and >150 respectively. A FC value of 200mg/g was used to indicate active disease. Results: 801 patients with CD were included in the analysis. There were 485 females (61%) with median an age at diagnosis of 28y (IQR 20 42). 552 (69%) had colonic involvement and 188 (23%) were smokers at diagnosis. 1691 FC were recorded with complete HBI. Of the 1071 results where patients were in ‘clinical remission’ with HBI 4, 436 (41%) had a FC of 200mg/g. There was a poor correlation between HBI and FC (Spearman’s rank correlation coefficient = 0.296). Likewise there was poor correlation between HBI and CRP (n = 860, rho = 0.241). In a sub-group of 202 patients with HBI 4 at baseline and follow-up HBI within 12 months, 90 (45%) had index FC 200mg/g. 48/202 (23.76%) had clinical relapse within a year (HBI 5), with 27/90 (18.75%) in the high FC group and 21/112 (30%) in the low FC group (p = 0.069). When stratified by disease location, 25.2% of 143 patients with L2 or L3 disease had clinical flare compared to 20.3% of 59 patients with non-colonic disease (p = 0.59). Symptoms alone did not predict disease progression. Conclusions: Our analysis support the ascertain that patients symptoms are a relatively poor indicator of underlying disease activity in CD. There is a poor correlation with FC. Furthermore, FC alone is a poor predictor of disease flare when this is defined by symptoms alone. Other data presented from this cohort show instead that FC is an excellent predictor of hard end-points pertaining to progression of disease in CD.