Bacterial translocation occurs early in cirrhosis and triggers a selective inflammatory response

Background Experimental data suggest that bacterial translocation (BT) promotes systemic inflammation, portal hypertension, and circulatory dysfunction in advanced chronic liver disease (ACLD). Methods Patients with ACLD undergoing hepatic venous pressure gradient (HVPG) measurement and absence of acute decompensation or infections were included (n = 249). Serum biomarkers of BT (lipopolysaccharide [LPS], lipoteichoic acid [LTA], bacterial DNA [bactDNA]), systemic inflammation and markers of circulatory dysfunction were assessed. T-cell subsets in intestinal biopsies (n = 7 ACLD, n = 4 controls) were analyzed by flow cytometry. Results Patients had a median HVPG of 18 (12–21) mmHg and 56% had decompensated ACLD. LPS (0.04 [0.02–0.06] vs. 0.64 [0.30–1.06] EU/mL), LTA (4.53 [3.58–5.97] vs. 43.2 [23.2–109] pg/mL), and detection of bactDNA (≥ 5 pg/mL; 5% vs. 41%) were markedly higher in patients with ACLD than healthy controls (n = 40; p < 0.001) but were similar between different clinical stages of compensated and decompensated ACLD and displayed no meaningful correlation with HVPG and systemic hemodynamics. TNF-α and IL-10 correlated with LPS (Spearman’s rs = 0.523, p < 0.001/rs = 0.143, p = 0.024) but not with LTA. Presence of bactDNA was associated with higher LPS (0.54 [0.28–0.95] vs. 0.88 [0.32–1.31] EU/mL, p = 0.001) and TNF-α (15.3 [6.31–28.1] vs. 20.9 [13.8–32.9] pg/mL). Patients with ACLD exhibited a decreased CD4:CD8-ratio and increased TH1-cells in the intestinal mucosa as compared to controls. During a median FU of 14.7 (8.20–26.5) months, bacterial antigens did not predict decompensation or liver-related death (in contrast to HVPG, IL-6, and MAP) as well as infections at 24 months. Conclusion BT occurs already in early ACLD stages and triggers a systemic inflammatory response via TNF-α and IL-10. Interestingly, BT markers showed no clear correlation with portal hypertension and circulatory dysfunction in patients with stable ACLD. Clinical trial number NCT03267615. Supplementary Information The online version contains supplementary material available at 10.1007/s12072-023-10496-y.


Measurement of hepatic venous pressure gradient
Measurement of hepatic venous pressure gradient (HVPG) were performed following a standard operating procedure in fasting condition [1]. Briefly, under local anaesthesia and ultrasound guidance, a catheter introducer sheath was placed in the right internal jugular vein. Subsequently, a balloon catheter was advanced into inserted into a large hepatic vein via the inferior vena cava under fluoroscopic guidance. Correct and sufficient wedge position of the catheter was ensured by injecting contrast media while the inflated balloon was obstructing the outflow of the cannulated hepatic vein. At least three measurements of free and wedged hepatic vein pressure were performed to assess HVPG.

LPS measurement
A quantitative chromogenic limulus amoebocyte lysate (LAL) test (BioWhittaker, Nottingham, UK) was followed to evaluate LPS levels. Due to LPS ubiquity, samples and reagents were handled in an airflow chamber and processed with pyrogen-free material tested by manufacturers. E. coli lyophilized endotoxin (22 UE/ml) provided by the kit was used to set standard endotoxin concentrations ranging from 5.0 UE/ml to 0.1 UE/ml. To verify the lack of product inhibition by plasma protein, a dilution/heating inactivation protocol was followed prior to endotoxin measurement. A pooled E. coli endotoxin spike solution (0.4 UE/ml) was prepared with serum samples. Dilutions ranging from 1/2 to 1/20 were performed over spiked and unspiked serum samples. All test samples were then incubated at 60°C during 30 min [2]. The LAL test was performed after this period. The non-inhibitory dilution was established when the difference between spiked and unspiked endotoxin values was equal to the known concentration of the spike ±25%, as detailed by the manufacturer. Final sample dilutions used were 1/10 (spike recovery after correction of dilution: 0.34 UE/ml). All samples were tested in triplicate and read at 405 nm in a Epoch 2 microplate reader (Agielnt, Santa Clara, CA).

Statistical analysis
Continuous variables are reported as mean ± standard error of the mean or median with interquartile range (IQR) and categorical variables are displayed as absolute (n) and relative

Definition of liver-related events
The endpoint "first/further decompensation or liver-related death" was determined as the first occurrence and/or aggravation of variceal bleeding, ascites, hepatic encephalopathy (HE), or incidence of liver-related death within the follow-up period. The timepoint of HVPG measurement was defined as baseline, and the time to the closest event of interest was determined. Patients were censored at last clinical contact, transplantation, or at 24 months of no event occurred within 24 months of follow-up. The third large-volume paracentesis within 6 months was defined as worsening of ascites in patients without refractory ascites who had already received treatment for ascites at the timepoint of HVPG measurement.
Admission for overt HE was defined as worsening of HE in patients who had already received HE treatment and/or mild HE but had no record of overt HE prior to HVPG measurement.