Dual-sugar tests of small intestinal permeability are poor predictors of bacterial infections and mortality in cirrhosis: A prospective study

摘要

AIM: To prospectively analyze the impact of increased intestinal permeability (IP) on mortality and the occurrence of infections in patients with cirrhosis.METHODS: IP was quantified using the lactulose/mannitol (L/M) test in 46 hospitalized patients with cirrhosis (25 Child-Pugh A/B, 21 Child-Pugh C) and in 16 healthy controls. Markers of inflammation [LPS-binding protein, Interleukin-6 (IL-6)] and enterocyte death [intestinal fatty-acid binding protein (I-FABP)] were determined in serum using enzyme-linked immunosorbent assays. Patients were followed for one year and assessed for survival, liver transplantation, the necessity of hospitalization and the occurrence of bacterial infections. The primary endpoint of the study was defined as differences in survival between patients with pathological and without pathological lactulose/mannitol test.RESULTS: Thirty-nine (85%) patients with cirrhosis had a pathologically increased IP index (L/M ratio > 0.07) compared to 4 (25%) healthy controls (P < 0.0001). The IP index correlated with the Child-Pugh score (r = 0.484, P = 0.001) and with serum IL-6 (r = 0.342, P = 0.02). Within one year, nineteen (41%) patients developed a total of 33 episodes of hospitalization with bacterial or fungal infections. Although patients who developed spontaneous bacterial peritonitis (SBP) (n = 7) had a higher IP index than patients who did not (0.27 vs 0.14, P = 0.018), the baseline IP index did not predict time to infection, infection-free survival or overall survival, neither when assessed as linear variable, as tertiles, nor dichotomized using an established cut-off. In contrast, model for end-stage liver disease score, Child-Pugh score, the presence of ascites, serum IL-6 and I-FABP were univariate predictors of infection-free survival.CONCLUSION: Although increased IP is a frequent phenomenon in advanced cirrhosis and may predispose to SBP, it failed to predict infection-free and overall survival in this prospective cohort study.