Search in the Abstract Database

Abstracts Search 2014

* = Presenting author

P220. Faecal calprotectin and ileal Crohn's disease: correlation with a small bowel MRI score for disease activity

F. Fascì-Spurio1,2, N.A. Kennedy1,3, L. Wong4, P. MacLean4, J. Satsangi1,3, S. Glancy4, C.W. Lees1,3, 1Western General Hospital, Gastrointestinal Unit, Edinburgh, United Kingdom, 2Sapienza University, Gastroenterology Unit, Rome, Italy, 3University of Edinburgh, Gastrointestinal Unit, Centre for Genomic and Experimental Medicine, Edinburgh, United Kingdom, 4Western General Hospital, Department of Radiology, Edinburgh, United Kingdom

Background

Small bowel MRI (SBMRI) is the current standard for assessing ileal inflammation in Crohn's disease. Faecal calprotectin (FC) is closely correlated with colonic inflammation, but is thought to be of less utility in ileal disease. Interpretation of existing data linking FC with SBMRI findings have been confounded by the presence of colonic inflammation. We therefore aimed to ascertain how FC best reflects MRI findings exclusively in the small bowel.

Methods

150 SBMRI studies with matched FC results (±30 days) were identified from the Edinburgh FC Register (2008–12; n = 18,138). Scans were entered into an anonymous ‘teaching’ list on PACS and each re-read independently by 2 expert GI radiologists blind to all clinical and lab data. Technical, quality and disease parameters were recorded onto standard proformas. Scans rated by one or other radiologist as being of poor quality were excluded (n = 31/150). 7/13 disease parameters were excluded due to poor inter-observer variability (Cohen's kappa <0.5). A 6 item simple MRI score (range 0–10) was derived from assessment of the worst segment (bowel wall thickness, oedema, and relative enhancement, mesenteric oedema and pre-stenotic dilatation) plus total disease extent (overall kappa = 0.85). For comparisons with FC, studies where the radiologists reported upper GI or colonic inflammation were excluded (27/119).

Results

150 SBMRI scans were re-evaluated from 123 patients with purely ileal Crohn's (Montreal L1, n = 109; L3 + previous panproctocolectomy, n = 14; 65% female; median age at MRI 45 years (IQR 32–56); median follow-up 34 months (IQR 25–44). The median (IQR) FC was 80 µg/g (20–142) where SBMRI demonstrated no active ileal disease (simple MRI score = 0, n = 38), 198 µg/g (120–444) for mild to moderate (1–6, n = 30) and 398 µ/g (168–771) for severe disease (>6, n = 24) (p < 0.001). ROC analysis showed an AUC of 0.81 (0.72–0.90) for FC which was significantly higher than for CRP (0.65 [0.53–0.77], p = 0.020) (Figure 1).

Figure 1. ROC curves for abnormal ileum on small bowel MRI scan against various laboratory parameters.

Conclusion

FC correlates closely with SBMRI findings in ileal Crohn's disease and outperforms other laboratory tests. In future, following validation, we will derive clinical useful MRI and FC cut-offs that predicate on important patient outcomes.