Search in the Abstract Database

Search Abstracts 2016

* = Presenting author

P244 Is prognostic utility of rapid faecal calprotectin test equal in all inflammatory bowel disease (IBD) subtypes? Retrospective analysis based on endoscopic indices

A. Moniuszko*1, D. Koziel2, S. Gluszek2, G. Rydzewska1, 2

1Central Clinical Hospital of the Ministry of Interior, Department of Gastroenterology with Inflammatory Bowel Diseases Subdivision, Warsaw, Poland, 2Faculty of Medicine and Health Science, Jan Kochanowski University, Kielce, Poland

Background

Faecal calprotectin (FC) has been widely adapted in IBD treatment as a valuable surrogate marker of intestinal inflammation. However, there are contradictory reports concerning the influence of location and extent of disease on diagnostic accuracy of FC in ulcerative colitis (UC) and Crohn’s disease (CD). Several cut-off values for mucosal inflammation have been proposed, ranging from 150 to even 400 μg/g, depending on different endoscopic activity scores adapted.

The aim of the study was to evaluate the influence of disease location on FC levels and to calculate optimal cut-off for CD and UC flare.

Methods

We performed a retrospective analysis of medical records of 140 IBD patients (46 UC, 94 CD) hospitalised in the Department of Gastroenterology in tertiary referral centre in Warsaw in the last 2 years (mean age 35.5 ± 13.6 years, 56,8% women). Patients with clostridium difficile infection, prior abdominal surgery, and celiac disease were excluded. FC was measured by both ELISA and rapid semi-quantitative Quantum Blue® test (100–1800 μg/g, Bühlmann Laboratories) at admission.

Endoscopic activity was assessed by endoscopic Mayo sub-score for UC and Simple Endoscopic Score for Crohn’s Disease (SES-CD). Location and extent of disease were defined according to the Montreal classification.

Results

Quantum Blue® results were initially validated with ELISA test (r = 0.86, p = 0.002). FC highly correlated with endoscopic indices and increased significantly even in presence of the earliest signs of IBD flare compared with remission in CD (SES-CD 4–10 vs 0) (median 252.1 vs 100.0 μg/g, respectively; p = 0.02), in UC (Mayo 1 vs 0, median 323.3 vs 100.0 μg/g, respectively; p < 0.001).

ROC curve analysis showed high specificity and sensitivity of FC in prediction of inflammation in both CD (0.89 and 0.70) and UC (0.89 and 0.60, respectively); calculated cut-offs, 238.5 μg/g for CD (AUC 0.831; 95% CI 0.713–0.949) and 499 μg/g for UC (AUC 0.800; 95% CI 0.657–0.943).

In UC, FC was lower in proctitis than in left-side colitis and pancolitis (median 340.0, 500.0, and 421.5 μg/g, respectively), what was not statistically significant (p >0.05). In CD a trend towards lower FC was observed in isolated small bowel disease, but was also not statistically significant (median L1 195.0; L2 401.0; L3 591.5 and L4 660.0; p > 0.05).

Conclusion

FC greatly increases even in presence of mild endoscopic signs of inflammation. Adaptation of rapid bedside FC test to clinical practice can allow detection of IBD flare in shorter period than standard ELISA method. FC is not influenced significantly by location and extent of disease, which makes this marker a good diagnostic and prognostic tool in everyday practice.