P010 Serum proteomic profiling in Crohn's disease patients undergoing ileocolonic resection reveals discriminative inflammatory markers for endoscopic recurrence. (16th January 2018)

• Record Type: Journal Article
• Authors: Machiels, K; de Bruyn, M; Verstockt, S; Verstockt, B; Van Assche, G; Wolthuis, A; d'Hoore, A; Ferrante, M; Vermeire, S
• Is Part Of: Journal of Crohn's and colitis. Volume 12:Number 1(2018:Jan.)Supplement 1
• Page Start: S096
• Page End: S097
• Publication Date: 2018-01-16
• Subjects: Inflammatory bowel diseases -- Periodicals; 616.344005
• DOI: 10.1093/ecco-jcc/jjx180.137
• Languages: English
• ISSNs: 1873-9946

Abstract

Abstract: Background: Although not curative, intestinal resection is often required in Crohn's disease (CD) patients. Postoperative recurrence (POR) is common and early postoperative endoscopy is considered the gold standard for detection of POR. The recent expansion of proteomics provides an attractive way to identify new biomarkers of early inflammation. We investigated serum inflammatory profiles for their predictive and discriminative value for POR in CD patients undergoing ileocolonic resection. We also studied whether combination of proteomics with metagenomics including our recent association of Fusobacterium spp with POR, may improve diagnostic accuracy. Methods: Serum samples were prospectively collected from 70 healthy subjects (HS) and from 57 CD patients undergoing ileocecal resection with ileocolonic anastomosis before surgery and at month 1, 3 and 6 after surgery. POR - defined by a Rutgeerts score ≥i2 on endoscopy was assessed at month 6. Relative levels of 92 inflammation-related proteins were measured using the Proximity Extension Assay (PEA) (Olink Bioscience, Sweden). Fecal calprotectin (ELISA, Buhlmann) and relative abundance of Fusobacterium spp (16S RNA sequencing) were available in a subset of 47 and 37 patients respectively. Logistic regression and receiver operating characteristic curve analysis were used to evaluate the discriminative power of significant biomarkers. Analyses were conducted in SPSS with false discovery rate (FDR) correction for multiple testing. Results were validated in a second cohort of 41 patients. Results: Comparisons of the serological markers between both patient groups with and without POR were not significantly different after FDR correction at baseline, month1 and month 3. At month 6, CCL3, CCL4, OSM, FGF21, ST1A1 and TGF-ɑ were significantly increased in patients with POR ≥i2 ( N = 29) compared with patients with POR <i2 ( n = 28). A model combining these 6 proteins showed an accuracy of 78.9% and AUC of 0.844 ( p < 0.001). Validation of the model in a second cohort of 41 CD patients (7 <i2, 34 ≥i2) showed an accuracy of 87.8% and AUC=0.962 ( p < 0.001). Calprotectin and Fusobacterium spp each showed lower discriminative power (AUC 0.653, p = 0.072 and AUC 0.629, p = 0.18 resp). Combination with the six serum proteins showed an improved discriminative capacity for Fusobacterium spp (AUC 0.888 → 0.929) but not for calprotectin (AUC 0.801 → 0.804). Conclusions: Using a panel of inflammation-related proteins, we identified a combination of six serological biomarkers able to discriminate early endoscopic POR. These surrogate markers performed better than fecal calprotectin and incorporation of a microbial marker further improved its discriminatory capacity. Quantitative validation is required to test its clinical utility.