Introduction Faecal calprotectin is a sensitive and reliable marker of mucosal inflammation and mucosal healing in IBD. It has always been compared against endoscopic appearance as the presumed gold standard for assessment of disease activity. Several strands of evidence – using magnifying or image-enhanced endoscopy as well as confocal laser endomicroscopy – highlight abnormalities in colonic mucosa reported as ‘normal’ by standard white light colonoscopy (WLC). This may be compounded by the difficulties in bowel preparation in IBD and the non-widespread use of high definition technology (although the latter has not been specifically studied in relation to IBD). We wished to determine whether an elevated FCALP could predict relapse even in the presence of an ostensibly normal WLC.

Method As part of a larger study correlating FCALP to histologic assessment of disease activity, retrospective data was collected for consecutive patients with IBD on stable therapy undergoing colonoscopy for disease assessment. FCALP (Bühlmann ELISA) was collected as close as possible (prior) to the colonoscopy. When the colonoscopic appearances were reported as normal (Mayo endoscopic subscore 0 for ulcerative colitis, UC, and SES-CD = 0 for Crohn’s disease, CD), patients were followed to determine if they relapsed, with time to relapse (or last recorded follow-up) taken from the date of colonoscopy. For the purposes of this study, relapse was defined generally as “continuous or worsening intestinal symptoms requiring an escalation in therapy”. Switching between 5ASA classes was not included in this definition unless the specifically stated to be in response to uncontrolled symptoms.

Results 46 patients with UC and 37 with CD were identified with the above criteria. Median time between FCALP measurement and colonoscopy was 0.70 (max 2.89) and 0.90 (max 1.5) months respectively. Normal FCALP was detected in 16 patients with UC and 12 with CD.

Figures 1 and 2 below show relapse-free survival in those patients with a normal colonoscopy and normal FCALP (<60 mcg/g faeces; green line) compared to those with elevated levels (red line). Median calprotectin in the ‘high’ group was 377 (229–794) in UC and 192 (118–247) in CD. At 12 months, survival proportions were 86% with UC and normal FCALP compared to 34% in those with high FCALP. In CD, these proportions were 50% and 12% respectively.

Conclusion Elevated levels of FCALP predict relapse of IBD even in the presence of a macroscopically normal colonoscopy.

Disclosure of interest None Declared.