BACKGROUND AND AIMS:
A hallmark of inflammatory bowel disease [IBD] is chronic inflammation, which leads to excessive extracellular matrix [ECM] remodelling and release of specific protein fragments, called neoepitopes. We speculated that the biomarker profile panel for ulcerative colitis [UC] and Crohn's disease [CD] represent a heterogeneous expression pattern, and may be applied as a tool to aid in the differentiation between UC and CD.

METHODS:
Serum biomarkers of degraded collagens I, III-IV [C1M, C3M, and C4M], collagen type 1 and IV formation [P1NP, P4NP], and citrullinated and MMP-degraded vimentin [VICM] were studied with a competitive ELISA assay system in a cohort including 164 subjects [CD n = 72, UC n = 60, and non-IBD controls n = 32] and a validation cohort of 61 subjects [CD n = 46, and UC n = 15]. Receiver operating characteristic curve analysis and logistic regression modelling were carried out to evaluate the discriminative power of the biomarkers.

RESULTS:
All biomarkers were corrected for confounding factors. VICM and C3M demonstrated the highest diagnostic power, alone, to differentiate CD from UC with an area under the curve [AUC] of 0.77 and 0.69, respectively. Furthermore, the biomarkers C1M [AUC = 0.81], C3M [AUC = 0.83], VICM [AUC = 0.83], and P1NP [AUC = 0.77] were best to differentiate UC from non-IBD. The best combinations of biomarkers to differentiate CD from UC and UC from non-IBD were VICM, C3M, C4M [AUC = 0.90] and VICM, C3M [AUC = 0.98] respectively.

CONCLUSIONS:
Specific extracellular matrix degradation markers are elevated in IBD and can discriminate CD from UC and UC from non-IBD controls with a high diagnostic accuracy.