P210. Intestinal mucosal calprotectin (iMC) in ulcerative colitis (UC): measuring disease activity, defining remission and predicting outcome
M. Guirgis1, E. Wendt1, L.M. Wang2, P. Basset3, D. Burger1, A. Kent1, R. Adamson1, S. Travis1, S. Keshav1, 1Translational Gastroenterology Unit, John Radcliffe Hospital, Oxford, United Kingdom, 2Department of Cellular Pathology, John Radcliffe Hospital, Oxford, United Kingdom, 3StatsConsultancy Limited, United Kingdom
Faecal calprotectin is an established marker of disease activity in UC. However, the potential value of intestinal mucosal calprotectin (iMC) is unknown. We compared iMC to clinical, endoscopic and histological disease activity to determine whether iMC predicted steroid use, hospitalisation and colectomy over 5 years.
In 2007, 94 outpatients with varying UC activity were evaluated. Clinical, endoscopic and histological activity were scored using the Simple Clinical Colitis Activity Index (SCCAI), and at sigmoidoscopy, Baron Score and Truelove & Richards' Index for histology. In 2012, data on steroid use, hospitalisation and colectomy were collected retrospectively. iMC was quantified by immunohistochemistry for S100A8/A9 heterodimer on 372 biopsies from 83 patients and 20 controls (normal colon from surgical specimens). 5 high power fields (hpf) in each biopsy were examined to determine the mean number of calprotectin-positive cells per hpf. Statistical methods used were Pearson correlation for activity and Cox Regression with Bonferroni correction for steroid use, hospitalisation and colectomy.
The mean age of patients was 37.5+15, and 79/83 patients had an SCCAI <10. Median follow up was to 51 months (IQR40–55). Mean iMC was lowest in patients with concordance for clinical, endoscopic and histological remission [2.7 cells/hpf (IQR0.5–6.8), compared to >38cells/hpf for active disease (p < 0.001)]. iMC correlated with increasing clinical, endoscopic and histological activity (R = 0.39, 0.51 and 0.59, respectively, all p < 0.001). iMC <10 cells/hpf was associated with 88%, 70% and 63% chance of remaining steroid-free over 1, 3 and 5yrs respectively.
Clinical, endoscopic, and histological scores did not predict hospitalisation or colectomy in this sample, and neither did iMC. In the 44 patients with histological remission, high iMC was associated with >50% chance of requiring steroids over the next 3 years (HR 2.58, CI 1.25–5.30, p < 0.01).
iMC < 10cells/hpf is associated with a significantly increased likelihood of remaining steroid-free over subsequent years, whilst iMC >10cells/hpf is associated with >50% chance of requiring steroids over 3 years, even when there is histological remission. iMC is therefore a potentially useful objective marker of disease activity.