Search in the Abstract Database

Abstracts Search 2019

P282 Combination of biomarkers reflecting type IV collagen degradation and citrullinated vimentin predicts response to adalimumab with high diagnostic accuracy, in patients with Crohn’s disease

J. H. Mortensen*1, M. A. Karsdal1, H. Grønbæk2, C. L. Hvas2, A. Dige2, T. Manon-Jensen1

1Nordic Bioscience, Biomarkers and Research, Herlev, Denmark, 2Aarhus University Hospital, Hepatology and Gastroenterology, Aarhus, Denmark


In inflammatory bowel diseases (IBD), up to 40% of patients do not respond to biologic treatment, eg, anti-TNFα antibodies. A personalised medicine approach may facilitate the best possible treatment option for IBD patients. Currently, no biomarkers have sufficient sensitivity to separate responders from non-responders within the first weeks of anti-TNFα therapy, which limits the personalised medicine approach for IBD patients. We investigated serum biomarkers that reflect basement membrane degradation (C4M: MMP mediated degradation of type IV collagen) and citrullinated vimentin (VICM: activated macrophages), and their ability to predict response to anti-TNFα treatment in Crohn’s disease.


This was a single-centre cohort study. We measured clinical response to adalimumab at Week 8 after treatment induction in 22 patients with Crohn’s disease, using the Harvey–Bradshaw Index (HBI). Response was defined as clinical remission (HBI<5) at Week 8. ELISA was applied to quantify the degradation of type IV collagen (C4M) and macrophages activity (VICM). Inflammation was estimated by C-reactive protein (CRP). The biomarkers were combined in a backwards multi-variate regression model to increase the prediction value for non-response to anti-TNF.


At baseline, C4M serum levels was significantly higher in non-responders compared with responders (AUC: 0.81 [CI: 0.58–1.00], p = 0.027). VICM serum levels were not significantly different at baseline between responders and non-responders but was modulated in patients who responded to anti-TNF and was significantly lower at Week 1 compared with non-responders (AUC=0.89 [CI: 0.69–1.00], p = 0.007). CRP did not demonstrate any predictive value at baseline (AUC=0.65 [CI: 0.42–0.89], p = 0.301) or Week 1 (AUC=0.66 [CI: 0.38–0.94], p = 0.282).

Abstract P281 –Figure 1. The figure depicts C4M, VICM, and CRP’s ability to predict response at baseline or Week 1 after adalimumab treatment.

C4M and VICM were included in the final model. The combination of C4M and VICM increased the predicted value to identify patients that do not respond to anti-TNF treatment (AUC=0.94 [CI: 0.75–1.00], p = 0.005), with an odds ratio of 22 (CI: 2.70–313).


The combination of baseline serum levels of C4M and Week 1 serum levels of VICM demonstrated high accuracy to predict who will respond to anti-TNFα treatment in Crohn’s disease, and was superior to CRP. Thus, baseline levels of C4M in combination with Week 1 levels of VICM may be used to predict response to anti-TNFα and may therefore aid in a more personalised treatment approach.