Search in the Abstract Database

Abstracts Search 2019

P601 Differential cytokine profiles and drop of faecal calprotectin for prediction of primary response to infliximab induction therapy in Crohn’s disease

B. Mateos1, E. Sáez1,2, I. Moret1,3, D. Hervás4, L. Tortosa1,3, E. Cerrillo2,3, M. Iborra2,3, M. García2, P. Nos1,2,3, B. Beltrán1,2,3

1IIS Hospital La Fe, Gastroenterology, Valencia, Spain, 2Hospital Universitari i Politècnic La Fe, Gastroenterology, Valencia, Spain, 3CIBEREHD, Madrid, Spain, 4IIS Hospital La Fe, Biostatistics, Valencia, Spain


One third of Crohn’s disease (CD) patients do not achieve a clinical response after the induction therapy with infliximab (IFX). Cytokines emerge as possible biomarkers of response, as they are directly implicated in the pathogenesis of CD. Furthermore, novel cytokines have been described recently (ie, Oncostatin M (OSM)1). Their utility as biomarkers remains to be explored. Response to IFX seems to be well reflected by a drop in faecal calprotectin (FC).2 We aimed to determine plasmatic cytokine profiles of active CD patients that started IFX treatment, their changes after the induction therapy, and their capacity to predict response to IFX.


Twenty-two active CD patients (68% males) receiving an induction therapy of IFX (5 mg/kg weeks 0, 2, 6) were included in the study (45% L1). Peripheral blood samples (for cytokine analysis) and faecal samples (for FC analysis) were collected on weeks 0 and 14. Fifteen cytokines (IL-1β, -2, -6, -7, -8, -10, -12p70, -13, -17, -21, -22, -23, IFNγ, TNFα and OSM) concentrations were measured by Luminex technology. FC concentration was determined by ELISA. Response to IFX was evaluated by the drop of FC based on its logarithm values (Ln FC week 0 – Ln FC Week 14). Other clinical parameters (HBI, CRP) were also considered. R statistical software, random forest predictive model, heatmap graphs and Rho Spearman (R2) were used for data analysis.


FC and HBI median values were 498 μg/g (IQR: 247, 918.5) and 7 (IQR: 5.25, 8) pre-induction; and 104 µg/g (IQR: 29, 767) and 3 (IQR: 1.25, 5) post-induction, respectively. Random forest model showed 10 pre-treatment cytokines on the top plot which were related to response: TNFα, IL-13, OSM, IL-7, IL-10, IL-8, IL-23, IL-17, IL-6 and IL-22. Among these cytokines, TNFα, IL-13 and OSM were statistically significant. Heatmap graphs showed that higher levels of IL-13 pre-treatment, low TNFα levels and the presence of OSM were significantly associated with a better IFX therapy response. The analysis of the cytokines’ networks showed that most important correlations were established between IL-17, IL-1b, IL-2, and IFNγ (R2 = 0.92; 0.82; 0.79; 0.77) where IL-13 was also present (R2 = 0.51). TNFα and OSM belonged to different networks: TNF was associated to IL-8 (R2 = 0.68), and OSM to IL-22 (R2 = 0.67). This is the first study exploring the plasma concentration of OSM and its utility as biomarker in CD.


Determination of IL-13, TNFα, and OSM plasma concentrations could help to predict response to the IFX therapy. Networking analysis supports the idea that cytokines may be analysed in groups instead of individually. IL-13, TNFα and OSM seem to have differential and specific interconnections.


1. West NR, Hegazy AN, Owens BMJ, et al. Oncostatin M drives intestinal inflammation and predicts response to tumor necrosis factor-neutralizing therapy in patients with inflammatory bowel disease. Nat Med 2017;23:579–589.

2. Pavlidis P, Gulati S, Dubois P, et al. Early change in faecal calprotectin predicts primary non-response to anti-TNF therapy in Crohn’s disease. Scand J Gastroenterol 2016;51:1447–1452.