Logo: to the web site of Uppsala University

BACKGROUND: Bacterial translocation from the gut has been suggested to induce a systemic inflammatory response syndrome (SIRS) and organ dysfunction. The liver has a pivotal role in eliminating circulating bacteria entering from the gut. We investigated whether pre-existing inflammation affects hepatic bacterial elimination.

METHODS: Fifteen anaesthetised piglets were infused with E. coli in the portal vein for 3 h. The naive group (n = 6) received the bacterial infusion without endotoxin exposure. SIRS (SIRS group, n = 6) was induced by endotoxin infusion 24 h before the bacterial infusion. For effects of anaesthesia, controls (n = 3) received saline instead of endotoxin for 24 h. Bacterial counts and endotoxin levels in the portal and hepatic veins were analysed during bacterial infusion.

RESULTS: The bacterial killing rate was higher in the naive group compared with the SIRS group (p = 0.001). The ratio of hepatic to portal venous bacterial counts, i.e. the median bacterial influx from the splanchnic circulation, was 0.06 (IQR 0.01-0.11) in the naive group and 0.71 (0.03-1.77) in the SIRS group at 3 h, and a magnitude lower in the naive group during bacteraemia (p = 0.03). Similar results were seen for hepatic endotoxin elimination. Peak log tumour necrosis factor alpha was higher in the naive 4.84 (4.77-4.89) vs. the SIRS group 3.27 (3.26-3.32) mg/L (p < 0.001).

CONCLUSIONS: Our results suggest that hepatic bacterial and endotoxin elimination is impaired in pigs with pre-existing SIRS while the inflammatory response to bacterial infusion is diminished. If similar mechanisms operate in human critical illness, the hepatic elimination of bacteria from the gut could be impaired by SIRS.

BACKGROUND: Sepsis-induced immunosuppression impairs bacterial clearance and increases mortality. Liver and spleen macrophages are an essential part of the mononuclear phagocyte system and crucial for bacterial elimination. Systemic inflammation hampers bacterial clearance in the liver. We hypothesized that immunosuppression due to systemic inflammation might lead to decreased bacterial clearance by the spleen.

METHODS: Anesthetized pigs were subjected to an E. coli infusion for three hours in an intensive care setting. The Naive group only received the bacterial infusion (n = 10). The ETX group (n = 10) received an endotoxin infusion for 24 h, inducing systemic inflammation before the E. coli infusion. The Control group (n = 3) received saline instead of endotoxin for 24 h to study the effects of anesthesia alone. Bacterial counts and endotoxin levels were analyzed during the E. coli infusion. The levels of IL-6 and TNF were analyzed, and the piglets' physiological response was evaluated.

RESULTS: There was no difference in the bacterial counts in the artery, splenic vein or hepatic vein. However, the splenic venous to arterial bacterial counts ratio at 1-3 h was lower in Naive compared to ETX group (0.54 (0.34-1.07), 0.54 (0.20-0.85), 0.52 (0.21-0.64) vs. 0.77 (0.61-1.06), 0.85 (0.63-0.89), 0.74 (0.53-0.88); p < 0.01), but there was no difference in the animals' blood ex vivo bactericidal capacity. There was no difference in endotoxin clearance in the spleen between the groups. The peak log IL-6 and TNF levels in response to the E. coli infusion were higher in the Naive compared to the ETX group (3.40 ± 0.41 vs. 2.94 ± 0.43 pg × mL^-1 and 3.78 ± 0.68 vs. 2.23 ± 0.36 pg × mL^-1; p < 0.05 and p < 0.001, respectively). The respiratory, circulatory and metabolic response to the E. coli infusion was dampened in the animals pre-exposed to endotoxin.

CONCLUSION: The splenic bacterial clearance is impaired by pre-existing systemic inflammation, while differences in endotoxin elimination were not detected. The inflammatory and physiological response to bacteremia is diminished during ongoing inflammation. Since the animals' ex vivo bactericidal capacity was not affected by pre-existing inflammation, our results suggest that inherent mechanisms in the spleen were involved.

Introduction

The liver is essential for bacterial elimination, preventing intestinal bacteria from entering the systemic circulation, and for producing inflammatory cytokines. Glucocorticoids have been reported to exert heterogeneous effects on bacterial clearance in sepsis. Using a porcine model of gram-negative abdominal sepsis, we investigated how portal venous administration of budesonide, a glucocorticoid with extensive hepatic first-pass metabolism, affects hepatic bacterial and endotoxin elimination as well as the systemic inflammatory response, in comparison with systemic administration and no treatment.

Methods

The Portal Steroid-Sepsis (Sep-Port, n = 8) and Systemic Steroid-Sepsis (Sep-Syst, n = 8) groups were administered budesonide in the portal vein or systemically, followed by an E. coli infusion for 3 h in the portal vein. The Septic Controls (Sep-Ctrl, n = 8) received saline instead of budesonide. Non-septic Controls (NSep-Port, n = 3) were treated only with portal budesonide. Portal, arterial, and hepatic venous bacterial counts were analyzed hourly during the bacterial infusion. The levels of endotoxin and inflammatory cytokines were measured.

Results

There was no difference in hepatic/portal venous bacterial count ratios. However, the arterial and hepatic venous bacterial counts were higher in the Sep-Syst compared to Sep-Port group (p < 0.001 and p < 0.01, respectively), while microbiological findings were similar in the Sep-Port and Sep-Ctrl groups. Hepatic endotoxin elimination did not differ between the groups. IL-10 levels were higher in the Sep-Port compared to the Sep-Syst group at 1 h (p < 0.01), and IL-6 levels were lower in the Sep-Port compared to the Sep-Ctrl after the bacterial infusion (p < 0.05).

Conclusions

In this experimental sepsis model, hepatic bacterial elimination was unaffected by portal or systemic budesonide, whereas systemic administration was associated with increased systemic bacterial levels. Endotoxin clearance was unaffected by budesonide. Portal budesonide elicited a more pronounced anti-inflammatory response compared to systemic administration. These findings suggest that hepatic exposure to budesonide may modulate the inflammatory response while limiting adverse effects on systemic bacterial clearance.

Editorial Comment

Glucocorticoids are widely used in sepsis management, yet their effects on bacterial clearance is incompletely understood. In this porcine model of gram-negative abdominal sepsis, portal venous delivery of budesonide (which has ~90% hepatic first-pass metabolism) preserved bacterial elimination and enhanced the early anti-inflammatory response, without the increase in systemic bacterial levels seen with systemic administration. These results demonstrate that route of delivery as a potentially important determinant of the risk-benefit balance of glucocorticoid therapy in sepsis, and support further investigation of hepatic-targeting strategies.