Abstract BACKGROUND AND AIMS Chronic inflammation is common in patients with chronic kidney disease (CKD) and is associated with poor outcomes. Although the etiology is only partly understood, the gut–kidney axis is considered an important contributor [1]. Lipopolysaccharide (LPS), also known as endotoxin, is a well-characterized pyrogen found in the exterior cell membrane of most of the gram-negative bacteria. It plays an important role in promoting intestinal inflammatory responses. When absorbed through the intestinal epithelium, LPS induces inflammation by activating macrophages and monocytes. Due to short half-life and semi-quantitative assay characteristics of the Limulus Amebocyte Lysate (LAL) assay, direct quantification of LPS is not suitable to quantify activation of the innate immune system. Lipoprotein-binding protein (LBP), a more stable biomarker, is an acute phase protein produced mostly by the hepatocytes and intestinal epithelial cells and is an essential component of an effective innate immune response to LPS. Circulating LBP significantly enhances the sensitivity of CD14 + cells (mostly monocytes and macrophages) to stimulation by LPS. LBP has been shown to facilitate binding of LPS to CD14 receptor. Levels of LBP peak in serum shortly after endotoxemia and remain increased up to 72 h later. An increased plasma LBP indicates gut epithelial barrier dysfunction [2]. Whether LBP is altered by CKD is not known. METHOD We analyzed the effects of CKD on LBP plasma concentrations. We used samples from the Leuven mild-to-moderate CKD cohort (NCT00441623). To study causality, we used animal models of experimentally induced CKD. To exclude model-related bias, two different rat models of CKD, i.e. 5/6th nephrectomy and adenine supplementation, were used. LBP was measured using commercially available ELISA kits (Hycultbiotech, The Netherlands). RESULTS In a cohort of 460 patients with CKD, we found a significant increase in LBP levels across the different stages of CKD (ANOVA, P: 0.001). When analyzed as a continuous variable, LBP is significantly inversely correlated with eGFR (P < 0.001; Spearman, r: −0.221), and positively correlated with CRP (P < 0.001; Spearman, r: 0.592). During a median follow-up of 56 (IQR, 53–59) months, 70 patients died, with more deaths observed among patients with LBP in higher tertiles [tertiles 1–3: 11, 22, and 37 events, respectively (Fig. 1)]. In univariate Cox proportional hazard analysis, plasma LBP was significantly associated with mortality [hazard ratio (HR) of 1.032; 95% confidence interval (95% CI): 1.014– 1.046; P < 0.001). This association remained significant in multivariate models with adjustment for age, sex and BMI. In two different rat models, induction of CKD resulted in a significant increase in LBP (P < 0.001). In these animals, we observed a significant inverse association between eGFR (and measured creatinine clearance) and LBP concentrations (P < 0.001; Spearman, r: −0.460). CONCLUSION Patients with CKD have higher levels of LBP, with higher levels of plasma LBP present in patients with more advanced CKD. Plasma LBP is an independent predictor of mortality. Experimentally induced CKD equally results in significantly increased LBP. Our data suggest that CKD leads to increased passage of LPS across the intestinal barrier. These findings strengthen the hypothesis of the gut–kidney axis as a source of chronic inflammation.