Faecal calprotectin, an alternative marker to estimate cumulative inflammatory burden in Ulcerative Colitis

Introduction: Long-term Ulcerative Colitis (UC) increases the risk of colonic dysplasia and colorectal cancer (CRC). We aim to establish whether real-world faecal calprotectin (FCP) data can be used to estimate the cumulative inflammatory burden (CIB) and identify those at risk of dysplasia and CRC.

Methods: Patients with left sided or extensive UC of >8 yrs duration, with >1 endoscopy and >1 serial FCP value (from 2005) were extracted from the NHS Lothian IBD registry. Patients with PSC were excluded. CIB scores based on histology (CIB(H)) or FCP (CIB(FCP)) were calculated based on the method proposed by Choi et al [1]. Patients were categorised into three groups; IBD-associated dysplasia and CRC (IBD-D/CRC n=15), only sporadic adenomas (n=29, excluded from further analysis) and patients who did not develop any type of dysplasia (n=220). To give a more accurate estimation of cumulative inflammation FCP levels were defined as low (n=73) or high (n=162). High CIB(FCP) is equivalent to 5 yrs of a continuous FCP value of ≥250µg/g, a surrogate marker of chronic active inflammation.

Results: A defined cohort of 264 patients (146 males), with a median age 36 (IQR 27.1-46.9) were included. Using the CIB(H) score, patients with no dysplasia (n=220) had a median score of 4.7 (IQR 2.7-7.9), compared with patients with IBD-D/CRC (n=15) who had a score of 5.4 (3.5-8.2) (p=0.4405, unpaired two-tailed t-test). The median CIB (FCP) score for patients with no dysplasia was 1804 (883-3689), compared with patients with IBD-D/CRC who had a median score of 2256 (1593-3848) (p=0.4835). The correlation between the two types of CIB scores in identifying risk of IBD-D/CRC was weak (Spearman’s rho=0.296 (p<0.001). In this cohort neither score was able to predict IBD-D/CRC. To give a more accurate estimation of CIB(FCP) the rates of IBD-D/CRC were stratified by FCP levels over time; 73 patients (31.1%) had low CIB(FCP) and 162 (68.9%) had a high CIB(FCP). Using this model, 1/73 (1.4%) in the low CIB(FCP) group and 14/162 (8.6%) in the high group had IBD-D/CRC (p=0.0417, Fishers Exact Test), figure 1. Figure 1: low/high CIB(FCP)

Conclusions: This uncontrolled cohort study suggests that serial faecal calprotectin measurements can be used to estimate the cumulative inflammatory burden. Patients with chronic UC who have a high CIB (FCP) score were more likely to develop IBD-D/CRC, but there was only weak correlation between CIB(H) and CIB(FCP). Further data is required to validate our findings.