P113 Serum adipokines as non-invasive biomarker in Crohn’s disease

Abstract

Abstract Background Adipose tissue wrapping the gastrointestinal tract is a risk factor for disease activity in Crohn’s disease (CD). Indeed, adipokines associated to lipid metabolism can modulate local immune responses. However, few studies have investigated the possible association between adipokines and CD. Here, we aimed to evaluate the role of serum adipokines as possible biomarkers in CD. Methods Serum samples were obtained from 18 patients with endoscopically active CD (aCD), 22 patients with endoscopically quiescent CD (qCD) and 36 non-inflamed healthy controls (HC). Serum leptin, ghrelin, resistin and adiponectin were analysed by Multiplex (Bio-Rad, Hercules, CA) in a Luminex 200 system technology. Adipokines means for each group of patients were compared and adjusted by Bonferroni or Wilcoxon tests. Receiver operating characteristic (ROC) curves and the area under the curve (AUC) were carried on to evaluate the discriminatory capacity of the adipokines levels between patients groups. Furthermore, for those adipokines that showed an AUC > 0.7, a binary logistic regression adjusted by possible confounders (i.e. sex, age and BMI) was performed in order to test their possible association with CD. Results No differences were found in age, sex or BMI among aCD, qCD and HC; there was no correlation either among the adipokines analysed. Means distribution for serum resistin was different among aCD, qCD and HC (p = 0.006). However, only comparisons between HC and aCD groups showed significant differences (p = 0.02) in the post hoc test. Serum leptin, ghrelin and adiponectin did not show differences between means. ROC curve for resistin showed an AUC of 0.78, with a standard error (SE) of 0.067 when HC and aCD groups were analysed (Figure 1). Furthermore, in this case, as sensitivity and specificity for Youden index correspond to the resistin median concentration, this value was selected as a cut-off for the binary logistic regression analysis; thus, odds ratio (OR) along with their 95% confidence interval (CI) analysis of high relative resistin levels (values higher than their median) adjusted by age, sex and BMI yielded a value of 5.46 (1.34–22.14) when aCD group was compared with HC group. Comparison between qCD and aCD showed an AUC of 0.70 with an SE of 0.089. However, ROC curve did not show a good accuracy when HC and qCD groups were analysed (AUC = 0.58; SE = 0.08). ROC curves analysis for leptin, ghrelin and adiponectin did not show enough accuracy to discriminate between groups. Conclusion Resistin may modulate the inflammation state in CD and it is probably associated to activity, being this association independent of sex, age or BMI.