Cellular changes in association with defense mechanisms in intra-abdominal sepsis

Despite progress, intra-abdominal sepsis is associated with a high morbitity and mortality rate. Although, much effort has been made in basic research, there have not been any therapeutic applications as yet. The peritoneal defense system (innate and specific) represents the first local reaction to inflammation caused by bacterial invasion. It includes lymphatic absorption of bacteria, phagocytosis, entrenchment of inflammation and lymphocyte production (humoral and cellular immune). Also, the fibrin formation and degradation by intraperitoneal activation of coagulation and fibrinolysis plays an important role in this local response. The endotoxin from Gram-negative or exotoxins from Gram-positive bacteria cause the release of proinflammatory cytokines (TNF-alpha, IL-1betha, IL-6) by macrophages. They act as mediators resulting in the initiation of systemic inflammatory response syndrome (SIRS) at first and cellular damage with multiple organ dysfunction syndrome (MODS) ultimately. There are two different, but communicated, functional departments, i.e. peritoneal and systemic compromising the host inflammatory response to bacterial infection. Cytokine production occurs in both of them.     

Granulocyte colony-stimulating factor increases hepatic sinusoidal perfusion during liver regeneration in mice.

Conditioning with granulocyte colony-stimulating factor (G-CSF) promotes liver regeneration in an experimental small-for-size liver remnant mouse model. The mechanisms involved in this extraordinary G-CSF effect are unknown. The aim of this study was to investigate the influence of G-CSF on the hepatic microvasculature in the regenerating liver. The hepatic sinusoidal microvasculature and microarchitecture of the regenerating liver were evaluated by intravital microscopy in mice. Three experimental groups were compared: (1) unoperated unconditioned animals (control; n = 5), (2) animals conditioned with G-CSF 48 h after 60% partial hepatectomy (G-CSF-PH; n = 6), and (3) animals sham conditioned 48 h after 60% PH (sham-PH; n = 6). PH led to hepatocyte hypertrophy and increased hepatic sinusoidal velocity in the sham-PH and G-CSF-PH groups. Increased sinusoidal diameter and increased hepatic blood flow were observed in the G-CSF-PH group compared to the sham-PH and control groups. Furthermore, there was a strong positive correlation between spleen weight and hepatic sinusoidal diameter in the G-CSF-PH group. The increased hepatic blood flow could explain the observed benefit of G-CSF conditioning during liver regeneration. These results elucidate an unexplored aspect of pharmacological modulation of liver regeneration and motivate further experiments.     

Granulocyte colony-stimulating factor (G-CSF) production in hemorrhagic shock requires both the ischemic and resuscitation phase

Granulocyte colony-stimulating factor (G-CSF) is the cytokine that is critical for polymorphonuclear neutrophilic granulocyte (PMN) production as well as being a potent agonist of PMN activation. We have recently reported that in the lung and the liver of rats resuscitated after hemorrhagic shock (HS) G-CSF mRNA expression is induced. It is not known if both phases of HS, the ischemic and the reperfusion phase, are required for G-CSF mRNA induction. The present study was designed to test the hypothesis that the upregulation of G-CSF mRNA expression is the consequence of HS followed by resuscitation and that ischennia alone is insufficient to induce G-CSF mRNA expression in the affected organs. Male Sprague-Dawley rats were subjected to resuscitated and unresuscitated shock protocols of varying severity. Control animals were subjected to anesthesia and all surgical preparations except for hemorrhage. Lungs and livers were isolated and their RNA extracted. Using semiquantitative reverse trans criptase-polymerase chain reaction (RT-PCR), we demonstrated that G-CSF mRNA was induced in the lung and liver of shock animals above the level observed in control animals. Upregulation of G-CSF mRNA relative to controls occurred only in animals undergoing resuscitated HS and not in ones subjected to unresuscitated HS. These results indicate that G-CSF production specific for the hemorrhage component of shock is dependent on resuscitation. As a consequence, the production of this cytokine may be decreased through modifications in the resuscitation protocols.     

Granulocyte macrophage colony-stimulating factor improves survival in two models of gut-derived sepsis by improving gut barrier function and modulating bacterial clearance.

OBJECTIVE: The effect of recombinant murine granulocyte macrophage colony-stimulating factor (rmGM-CSF) on survival and host defense was studied using two clinically relevant models of infection that included transfusion-induced immunosuppression. SUMMARY BACKGROUND DATA: Granulocyte macrophage colony-stimulating factor improves resistance in several models of infection, but its role in transfusion-induced immunosuppression and bacterial translocation (gut-derived sepsis) has not been defined. METHODS: Balb/c mice were treated with 100 ng of rmGM-CSF or placebo for 6 days in a model of transfusion, burn, and gavage, or cecal ligation and puncture (CLP). Translocation was studied in the first model. RESULTS: Survival after transfusion, burn, and gavage was 90% in rmGM-CSF-treated animals versus 35% in the control group (p < 0.001). After CLP, survival was 75% in the rmGM-CSF group versus 30% in the control group (p = 0.01). Less translocation and better killing of bacteria was observed in the tissues in animals treated with rmGM-CSF. CONCLUSION: The ability of rmGM-CSF to improve gut barrier function and enhance killing of translocated organisms after burn injury-induced gut origin sepsis was associated with improved outcome. Granulocyte macrophage colony-stimulating factor also improved survival after CLP.     

Recombinant human granulocyte colony-stimulating factor and Pseudomonas burn wound sepsis

Multiple immune defects have been demonstrated following thermal injury, including defective granulocyte production and function. Recombinant human granulocyte colony-stimulating factor (rhGCSF) is a regulator of the myelopoietic system. The effect of rhGCSF administration on survival and on the myelopoietic system in a murine model of Pseudomonas burn wound sepsis was investigated. Male BDF1 mice that underwent a 15% total body surface area burn injury and burn wound seeding with 1 x 10(8) Pseudomonas aeruginosa organisms demonstrated an improved mean survival time with the subcutaneous administration of 100 ng of rhGCSF twice a day. Mice that underwent a similar thermal injury and burn wound seeding with 3 x 10(7) P aeruginosa organisms demonstrated an augmented myelopoietic response through the administration of rhGCSF, as represented by significantly increased white blood cell count, neutrophil count, splenic weight, femoral marrow cellularity, and femoral marrow granulocyte-macrophage colony-forming cell count. Myelopoietic augmentation through rhGCSF administration may serve to decrease the morbidity of septic events following thermal injury.     

Cardiovascular and metabolic changes in shock and sepsis. Review fo changing concepts.

The commonly accepted sequence of low blood flow, tissue hypoxia, lacticacidosis and death does not apply to all patients dying from shock. The hyperdynamic circulation characteristic of severe sepsis is not likely due to peripheral arteriovenous shunts, since in skeletal muscle at least, capillary blood flow is increased and varies directly with cardiac index. A hyperdynamic circulatory state is seen in many patients with sepsis and may be related to metabolic changes rather than changes in oxygen transport. Skeletal muscle capillary blood flow is increased in fasting normal subjects and septic postoperative patients, both of whom are catabolic. Therefore, elevated blood flow, which is characteristic of severe sepsis, may be a response to or necessary for the catabolism of body protein required for energy production. Profound metabolic abnormalities resulting in rapid catabolism may be responsible for the demise of the septic patient. If this concept of sepsis is accepted, it follows that treatment which heretofore has been aimed at increasing blood flow and blood pressure should be redirected to therapy which provides energy substrates and alters hormonal patterns to favor anabolism.     

Granulocyte colony-stimulating factor (G-CSF) stiffens leukocytes but attenuates inflammatory response without lung injury in septic patients.

OBJECTIVE: To determine whether granulocyte colony-stimulating factor (G-CSF) administration changes leukocyte deformability resulting in lung injury in patients with sepsis. METHODS: Twenty-five consecutive septic patients were divided randomly into two groups. Twelve patients were given recombinant human G-CSF subcutaneously at 2 microg/kg once a day for 5 days (group G). The remaining 13 patients were given sterilized saline as placebo (group N). Leukocyte count; concentrations of C-reactive protein (CRP) and thrombomodulin (TM); respiratory index (RI) and lung injury score (LIS); and APACHE II score and Goris MOF index were determined before and after G-CSF or placebo administration. Leukocyte deformability was observed in a microchannel array etched on a single-crystal silicon tip, which simulates the microvasculature. The number of microchannels obstructed (NOM) by stiffened leukocytes was counted. Transit time (TT), that is, the time taken for 100 microL of whole blood to pass through the microchannel, was determined. RESULTS: G-CSF administration significantly increased leukocyte count and decreased CRP concentration. In group G, both NOM and TT increased significantly 5 days after G-CSF administration; they did not change in group N. However, RI, LIS, and TM did not change, suggesting that no patient developed lung injury. CONCLUSION: G-CSF causes leukocyte stiffness but attenuates inflammatory response without inducing lung injury in septic patients.     

Granulocyte macrophage colony-stimulating factor expression by both renal parenchymal and immune cells mediates murine crescentic glomerulonephritis

GM-CSF has previously been demonstrated to be important in crescentic glomerulonephritis (GN). As both renal parenchymal cells and infiltrating inflammatory cells produce GM-CSF, their separate contributions to inflammatory renal injury were investigated by creation of two different types of GM-CSF chimeric mice: (1) GM-CSF-deficient (GM-CSF-/-)-->wild-type (WT) chimeras with leukocytes that are unable to produce GM-CSF and (2) WT-->GM-CSF-/- chimeras with deficient renal cell GM-CSF expression. Crescentic anti-glomerular basement membrane GN was induced in WT, GM-CSF(-/-)-->WT chimeras, WT-->GM-CSF-/- chimeras, and GM-CSF-/- mice by planting an antigen (sheep globulin) in their glomeruli. WT mice developed severe crescentic GN, whereas GM-CSF-/- were protected from development of disease. Glomerular T cell recruitment, CD40+ glomerular cells, and renal IFN-gamma and TNF expression were similar in both chimeras and WT mice but significantly reduced in GM-CSF-/- mice, indicating that either leukocyte or renal sources of GM-CSF are sufficient to drive these aspects of the inflammatory response. Restricted expression of GM-CSF revealed a major role for renal cell-derived GM-CSF but a minor role for leukocyte-derived GM-CSF in the formation of cellular crescents; glomerular MHC II expression; serum creatinine; and monocyte chemoattractant protein-1, vascular cellular adhesion molecule, and IL-1beta expression. Glomerular macrophage accumulation, proteinuria, and interstitial infiltrate were equivalent in both chimeric groups but intermediate between WT and GM-CSF-/-, indicating that both sources are required for the full development of glomerular injury in crescentic GN.     

Hemodynamic changes during acute elevation of intra-abdominal pressure in rabbits

BACKGROUND: The intra-abdominal pressure (IAP) may be increased during pneumoperitoneum for minimally invasive surgery, after high tension repairs of congenital abdominal wall defects, major abdominal surgery, liver transplantation, abdominal trauma, peritonitis or ileus. The aim of this study was to investigate hemodynamic changes during elevation of IAP using an experimental setting, which mirrors anatomical and physiological conditions of neonates and small infants as closely as possible. METHODS: In five fasted, anesthetized, mechanically ventilated and multicatheterized New Zealand rabbits, the IAP was gradually increased by intra-abdominal infusion of normal saline (total volume 1000 ml). At baseline and after each infusion of 100 ml normal saline cardiac output (CO, transcardiopulmonary thermodilution), pressure in the superior (SVCP) and inferior vena cava (IVCP), mean arterial pressure (MAP), peak airway pressure (PAP) and IAP was recorded. RESULTS: During the study, IAP, SVCP and IVCP increased significantly. IVCP was significantly higher than SVCP from timepoint 200 ml to study end. After abdominal decompression IAP, SVCP and IVCP decreased to baseline levels. Changes in MAP were not significant. CO increased significantly from baseline to timepoint 200 ml (peak value), remained nearly constant until timepoint 800 ml and decreased thereafter until the abdominal infusion ceased. After abdominal decompression CO returned to baseline level. SVCP, IVCP and PAP correlated significantly with IAP (SVCP, r = 0.73; IVCP, r = 0.97; PAP, r = 0.94; P < 0.0001). CONCLUSIONS: The hemodynamic changes caused by increased IAP cannot be recognized by routine monitoring of arterial blood pressure and transcutaneous oxygen saturation. The increase in central venous pressure may be misinterpreted as an elevation of cardiac preload. One major effect of a prolonged increase in IAP is a decreased CO.     

Re-operation for intra-abdominal sepsis. Indications and results in modern critical care setting.

In a 2-year period (1981-1983), 87 abdominal re-explorations (1.6% of total laparotomies) were performed on 77 patients for sepsis in five Downstate hospitals. Fifty-one patients were re-explored solely on clinical grounds, 21 on clinical plus radiographic criteria, four solely on radiographic grounds, and 11 for multiple organ failure. The overall mortality rate was 43%. As expected, the most common laparotomy finding was intra-abdominal abscess (47); other findings included anastomotic leak (14), necrotic bowel (10), evidence of technical error (five), and acalculous cholecystitis (two). The most common clinical findings were localized tenderness, fever, and absent bowel sounds (85%). Fifty-four special studies were performed with an overall accuracy rate of 76%. CAT scans and contrast radiographs were most accurate (92% and 81%) while sonography and gallium scans were less useful (59% and 60%). Seven patients had negative laparotomies. While all were distended and six were febrile, only one patient had focal tenderness. In the 11 patients explored solely for multiple organ failure, six patients had drainable pus despite negative radiographic studies, and two survived. The other five patients had negative laparotomies, and all died. Factors correlated with mortality were age over 50, peritonitis at the primary operation, and multiple organ failure. The approach to these seriously ill patients should be governed by a high index of suspicion. Clinical findings are at least as reliable as sophisticated radiographic modalities of which CAT scan appears to be the most accurate. Re-exploration for multiple organ failure alone will yield a significant group of patients with drainable septic foci and some survivors; thus, exploration for this indication appears to be defensible.     

The requirement for granulocyte-macrophage colony-stimulating factor and granulocyte colony-stimulating factor in leukocyte-mediated immune glomerular injury.

Proliferative glomerulonephritis in humans is characterized by the presence of leukocytes in glomeruli. Granulocyte-macrophage colony-stimulating factor (GM-CSF) and granulocyte colony-stimulating factor (G-CSF) can potentially stimulate or affect T cell, macrophage, and neutrophil function. To define the roles of GM-CSF and G-CSF in leukocyte-mediated glomerulonephritis, glomerular injury was studied in mice genetically deficient in either GM-CSF (GM-CSF -/- mice) or G-CSF (G-CSF -/- mice). Two models of glomerulonephritis were studied: neutrophil-mediated heterologous-phase anti-glomerular basement membrane (GBM) glomerulonephritis and T cell/macrophage-mediated crescentic autologous-phase anti-GBM glomerulonephritis. Both GM-CSF -/- and G-CSF -/- mice were protected from heterologous-phase anti-GBM glomerulonephritis compared with genetically normal (CSF WT) mice, with reduced proteinuria and glomerular neutrophil numbers. However, only GM-CSF -/- mice were protected from crescentic glomerular injury in the autologous phase, whereas G-CSF -/- mice were not protected and in fact had increased numbers of T cells in glomeruli. Humoral responses to the nephritogenic antigen were unaltered by deficiency of either GM-CSF or G-CSF, but glomerular T cell and macrophage numbers, as well as dermal delayed-type hypersensitivity to the nephritogenic antigen, were reduced in GM-CSF -/- mice. These studies demonstrate that endogenous GM-CSF plays a role in experimental glomerulonephritis in both the autologous and heterologous phases of injury.     

Endotoxemia and bacteremia during hemorrhagic shock. The link between trauma and sepsis?

Previous investigations of a treated model of hemorrhagic shock in the rat indicated the frequent occurrence of bacteremia that appeared to derive from the gut. This paper determines the incidence of bacteremia and endotoxemia during the acute shock period and compares this with similar observations in humans in varying degrees of shock. Studies in 26 rats indicated that bacteremia and endotoxemia was present in 50% and 87%, respectively, by the end of 2 hours at a mean arterial pressure of 30 mmHg. Observations in 50 patients admitted to the trauma unit showed that positive bacterial blood cultures were present in 56% when the admission systolic blood pressure was 80 mmHg or less (p less than 0.01 compared with either of the other groups). Endotoxemia was noticed in two such patients. Direct access of bacteria and endotoxin to the blood stream may occur during hemorrhagic or traumatic shock and is the probable cause of subsequent sepsis in traumatized patients when no other source is apparent.     

Unsuspected right ventricular dysfunction in shock and sepsis.

Monitoring of ventricular function by central venous (CVP) and pulmonary wedge pressures (PCW) was compared with ejection fraction and end-diastolic volume (gated pool scan) in patients resuscitated from hypovolemic and septic shock. Sixteen patients were studied within 24 hours of resuscitation and all showed depressed right ventricular ejection (RVEF) and/or an increased end-diastolic volume (RVEDVI). Group I (eight patients, hypovolemia and sepsis) had low RVEF (mean, 0.30), high RVEDVI (mean 129.2 ml/m2), and nearly normal left ventricular function (LVEF 0.63 and LVEDVI 63.6 ml/m2), compared to angiographic normals (RVEF 0.52, RVEDVI 55.8 ml/m2; nL LVEF 0.59, LVEDVI 52.3 ml/m2). Group II (3 patients, all septic) had better RVEF (mean, 0.54) but high RVEDVI (mean, 121.1 ml/m2) with normal LVEF (mean, 0.67) and high LVEDVI (mean LVEDVI 107.2 ml/m2). Group III consisted of five patients (hypovolemia and sepsis) who had biventricular depression (RVEF 0.25 and LVEF 0.29) and elevated EDVI. The mortality rate for group I (25%) was significantly less than for groups II and III (100% and 80%, respectively), and could be correlated with failure to improve RV function. Follow-up studies in ten patients showed improvement in seven which correlated with increased RVEF and reduced RVEDVI. Comparing survivors to non-survivors showed no predictability on the basis of initial studies but a significantly larger RVEDVI and RV stroke work index in non-survivors' follow-up studies. No correlation could be made with left ventricular performance, and there were no correlations between PCWP and LVEDVI or CVP and RVEDVI. A significant negative correlation was seen between RVEF and pulmonary vascular resistance (r = -0.34, p less than 0.05). Both LVEDVI and RVEDVI were correlated significantly with cardiac index and with each other. RV dysfunction occurs after resuscitation of hypovolemia and sepsis without reliable alteration in filling pressure and is likely related to myocardial ischemia as well as increased pulmonary vascular resistance. Survival seems to depend on improvement in RV performance, which can be measured at the bedside by cardiac scintigraphy.     

Correlation of plasma and tissue oxidative stresses in intra-abdominal sepsis

BACKGROUND: The aim of this study was to investigate the correlation between plasma and tissue oxidative stress and the antioxidative response, by measuring malon dialdehyde (MDA) and glutathione (GSH) in late sepsis induced by cecal ligation and perforation. MATERIALS AND METHODS: A prospective, randomized, controlled experimental study in rats was done. Forty rats, weighing 200-250 g, were randomly divided into two groups (n = 20). In group 1, laparotomy was performed under aseptic conditions, and the cecum ligated and perforated. The abdomen was closed. In group 2, sham control, there was only laparotomy. Twenty-four hours later, blood samples were taken by cardiac puncture for plasma MDA and GSH, followed by harvesting of samples from lung, liver, kidney, and heart in both groups. RESULTS: In the liver, lung, plasma, heart, and kidney, MDA concentrations were increased in the sepsis group after 24 h (P < 0.001 for all organ samples). In the same organs, GSH concentrations were decreased by sepsis (P < 0.001 for all organ samples). In both groups, plasma MDA was positively correlated to MDA in heart (r = 0.82, P < 0.001), liver (r = 0.76, P < 0.001), lung (r = 0.78, P < 0.001), and kidney (r = 0.73, P < 0.001). Similarly, plasma GSH was positively correlated to GSH in liver (r = 0.93, P < 0.001), heart (r = 0.86, P < 0.001), lung (r = 0.91, P < 0.001), and kidney (r = 0.79, P < 0.001). CONCLUSIONS: Plasma MDA and GSH were positively correlated with tissue MDA and GSH in intra-abdominal sepsis in a rat model.     

Organ interactions in sepsis. Host defense and the hepatic-pulmonary macrophage axis

Endotoxin (lipopolysaccharide [LPS]) and tumor necrosis factor (TNF-alpha) have been implicated in the pathogenesis of sepsis-induced adult respiratory distress syndrome. To evaluate the possible interaction of the hepatic-pulmonary macrophage axis in the adult respiratory distress syndrome, we compared the kinetics of immunosuppressive prostaglandin E2, TNF-alpha, and interleukin 6 production in LPS-stimulated Kupffer cells and alveolar macrophages (AMs). Interleukin 6 production by Kupffer cells was significantly higher than for equal numbers of AMs. Kupffer cell TNF-alpha levels peaked early before decreasing as regulatory prostaglandin E2 levels rose. In contrast, AM TNF-alpha levels rose sharply and remained significantly higher than for Kupffer cells throughout culture coincident with negligible prostaglandin E2 production. Kupffer cell sequestration of LPS may normally invoke a coordinated cytokine response able to locally induce acute-phase hepatocytes. In hepatic failure, however, LPS spillover to the lung may promote adult respiratory distress syndrome by inducing unregulated AM TNF-alpha production within the pulmonary microenvironment.