Introduction
Ileo-colonoscopy remains the gold standard for assessing inflammatory bowel disease (IBD) activity; however, its invasiveness and cost limit repeated use. Current non-invasive biomarkers, including C-reactive protein (CRP) and faecal calprotectin (FC), have suboptimal performance, partly due to poor patient adherence to FC testing. Lipopolysaccharide-binding protein (LBP), an acute-phase protein reflecting microbial translocation and intestinal permeability, has emerged as a potential non-invasive biomarker.
Aims & Methods
Following promising retrospective findings (1), this prospective longitudinal study aimed to evaluate the clinical relevance of lipopolysaccharide-binding protein (LBP) in IBD. Adult patients with Crohn’s disease (CD) or ulcerative colitis (UC) undergoing routine follow-up endoscopy were enrolled. Endoscopic activity was defined as SES-CD ≥4 for CD and Mayo endoscopic subscore ≥1 for UC; remission corresponded to scores below these thresholds. This interim analysis includes baseline and first follow-up visit (V1; 8–16 weeks). Clinical activity was assessed using PRO2, and biological activity using CRP and FC. Serum LBP (ng/mL) was measured by ELISA. Diagnostic accuracy was evaluated using ROC curve analysis, and associations using non-parametric tests and Spearman correlations.
Results
A total of 119 IBD patients were included (63 CD, 56 UC), of whom 23.5% were in remission. At baseline, LBP correlated strongly with CRP (ρ=0.53, p<0.001), moderately with endoscopic activity (ρ=0.28, p<0.01) and PRO2 (ρ=0.29, p<0.01), but not with FC. This correlation with CRP strengthened after exclusion of corticosteroid-treated patients with normal CRP (ρ=0.59, p<0.001), suggesting LBP is not confounded by corticosteroid use. Longitudinally, changes in LBP (ΔLBP) correlated with changes in CRP (ΔCRP) (ρ=0.41, p<0.01), supporting its value for dynamic monitoring. At baseline, LBP levels were significantly higher in patients with active disease compared to those in remission, both in the overall IBD cohort and after stratification by CD and UC (Table). LBP showed discriminative ability for active disease (AUC=0.746), with an optimal cut-off of 6,166 ng/mL (sensitivity 63.4%, specificity 82.1%). Among patients with active disease, LBP levels were higher in CD than in UC at baseline (7622 [6056–9573] vs 6430 [4731–8639], p<0.05) and at V1 (8343 [6537–11262] vs 6023 [3231–8425], p<0.01). A similar, non-significant trend was observed in patients in remission at baseline (Table), but was not maintained at V1, likely due to limited sample size. At V1, LBP levels remained higher in patients with active disease compared to those in remission (7634 [4718–10238] vs 5816 [3365–8067], p<0.05), supporting a persistent correlation between LBP and intestinal inflammation.
Table : Baseline LBP levels (ng/mL) according to disease activity. Data are presented as median [IQR].
| Remission (n=28) | Active disease (n=91) | p-value |
| IBD | 4618 [2692–5944] | 7271 [5191–9081] | <0.001 |
| CD | 5214 [3939–5938] | 7622 [6056–9573] | <0.001 |
| UC | 2990 [2575–5944] | 6430 [4731–8639] | <0.05 |
Conclusion
In this interim analysis, LBP was associated with both endoscopic disease activity and systemic inflammation in IBD, with moderate accuracy for discriminating active disease. Its longitudinal changes support its potential role as a non-invasive biomarker for disease monitoring.
References
(1) Toris, Louison D et al. “Lipopolysaccharide-binding protein in Crohn's disease patients: a promising noninvasive biomarker monitoring disease activity.” European journal of gastroenterology & hepatology Vol. 36,9 (2024): 1093-1100. doi:10.1097/MEG.0000000000002811.