Continuous venovenous hemofiltration improves arterial oxygenation in endotoxin-induced lung injury in pigs.

BACKGROUND: Hypoxemia is common in septic acute lung failure. Therapy is mainly supportive, and most trials using specific inhibitors of key inflammatory mediators (ie., tumor necrosis factor alpha, interleukin 1) have failed to prove beneficial. The authors investigated if a nonspecific blood purification technique, using zero-balanced high-volume continuous venovenous hemofiltration (CWH), might improve arterial oxygenation in a fluid-resuscitated porcine model of endotoxin-induced acute lung injury. METHODS: Piglets of both sexes weighing 25-30 kg were anesthetized and mechanically ventilated. After baseline measurements, animals received an intravenous infusion of 0.5 mg/kg endotoxin (Escherichia coli lipopolysaccharide). One hour after endotoxin, animals were randomly assigned to either treatment with CWH (endotoxin + hemofiltration, n = 6) or spontaneous course (endotoxin, n = 6). At 4 h after randomization, animals were killed. Hemofiltration was performed from femoral vein to femoral vein using a standard circuit with an EF60 polysulphone hemofilter. RESULTS: Endotoxin challenge induced arterial hypoxemia, an increase in peak inspiratory pressure, pulmonary hypertension, and systemic hypotension. Treatment with CWH did not improve systemic or pulmonary hemodynamics. However, arterial oxygenation was increased in endotoxin-challenged animals at 5 h after completion of endotoxin infusion, as compared with animals not receiving CVVH (arterialoxygen tension, 268+/-33 vs. 176+/-67 mm/Hg, respectively, P < 0.01). In addition, treatment with CWH attenuated the endotoxin-induced increase in peak inspiratory pressure and increased lung compliance. CONCLUSION: These results suggest that nonspecific blood purification with high-volume CWH improves arterial oxygenation and lung function in endotoxin-induced acute lung injury in pigs, independent of improved hemodynamics, fluid removal, or body temperature.

[Callus development after bored and unbored femoral interlocking nailing. An experimental study]. / Kallusbildung nach aufgebohrter und unaufgebohrter Femurverriegelungsnagelung. Eine experimentelle Studie.

This was a prospective randomized baboon animal study, the study protocol was approved by the ethical committee according to the international guidelines for animal research projects. In 8 animals a midschaft femoral osteotomy was stabilized with reamed femoral interlocking nailing and in 8 animals by unreamed locked femoral nailing. Polychrome sequence bone labeling was done 5 weeks postop. with calcein-green, 8 weeks postop. with xylenol-orange and 10 weeks postop. with oxytetracycline. 10 weeks postop. the animals were sacrificed, the femurs explanted and planimetric and epifluorescence histomorphometric evaluation of serial transverse sections were done. In planimetric histomorphometric evaluation in unreamed femoral nailing a mean endostal callus formation was recorded with 28.0 +/- 9.9 mm2 per section and in reamed femoral nailing with 11.5 +/- 5.0 mm2 (p < 0.001). Periostal callus formation was recorded in the unreamed group with 238.7 +/- 87.1 mm2 per section and in the reamed group with 142.1 +/- 71.9 mm2 (p < 0.001). In epifluorescence histological evaluation endostal as well as periostal callus formation was more extensive and earlier after unreamed than reamed femoral nailing. Endostal callus formation was found in all animals after unreamed femoral nailing, and was present in 2 out of 8 specimen in the reamed group. Also 1 out of 8 animals in the reamed group developed a non-union. Unreamed femoral nailing with low diameter interlocking nails proved to be safe regarding bone healing in this experimental model with obvious advantages both in amount and time course of callus formation compared to reamed femoral nailing. Based on this results unreamed femoral nailing techniques can be recommended for femoral fractures.

Procalcitonin release patterns in a baboon model of trauma and sepsis: relationship to cytokines and neopterin.

OBJECTIVES: Procalcitonin (PCT) has been described as an early, discriminating marker of bacteria-associated sepsis in patients. However, little is known of its source and actions, in part because no appropriate animal models have been available. We tested the hypothesis that plasma PCT increases during various pathophysiological conditions, such as hemorrhagic shock and sepsis, which differ with regard to the degree of associated endotoxemia. We further hypothesized that in sepsis, PCT would be significantly different in survivors vs. nonsurvivors. DESIGN: Prospective, blinded analysis of previously collected plasma of experimental animals. SETTING: Independent nonprofit research laboratory in a trauma hospital and a contract research institute. SUBJECTS: A total of 22 male baboons (17.5-31 kg). INTERVENTIONS: Hemorrhagic-traumatic shock was induced by hemorrhage for up to 3 hrs, reperfusion with shed blood and infusion of cobra venom factor (n = 7). By using a similar experimental setup, severe hyperdynamic sepsis was induced (n = 15) by intravenous infusion of live Escherichia coli (2 x 10(9) colony-forming units/kg) over 2 hrs, followed by antibiotic therapy (gentamicin 4 mg/kg twice a day). MEASUREMENTS AND MAIN RESULTS: Plasma PCT at baseline was barely detectable, but levels increased significantly (p < .05) to 2+/-1.8 pg/mL 2 hrs after the start of reperfusion in the shock group, and to 987+/-230 pg/mL at 4 hrs after E. coli in the sepsis group. Levels were maximal between 6 and 32 hrs and had returned nearly to baseline levels at 72 hrs. Interleukin-6 levels paralleled the course of PCT measurements, whereas a significant increase in neopterin was seen at 24 hrs. PCT levels were approximately three times higher in the sepsis group than in the shock group, corresponding to endotoxin levels (at the end of hemorrhage, 286+/-144 pg/mL vs. 3576+/-979 pg/mL at the end of E. coli infusion; p = .003). PCT levels were significantly different at 24 hrs between survivors (2360+/-620 pg/mL) and nonsurvivors (4776+/-563 pg/mL) in the sepsis group (p = .032), as were interleukin-6 (1562+/-267 vs. 4903+/-608 pg/mL; p = .01) and neopterin/creatinine ratio (0.400+/-0.038 vs. 0.508+/-0.037; p = .032). CONCLUSIONS: PCT is detectable in the baboon as in humans, both in hemorrhagic shock and sepsis. PCT levels are significantly higher in sepsis than in hemorrhage, a finding that is probably related to the differences in endotoxin. The baboon can be used for the study of PCT kinetics in both models; PCT kinetics are clearly different from other markers of sepsis, either IL-6 or neopterin, in both models. There are significant differences between survivors and nonsurvivors in the sepsis model.

Isolation, partial characterization, and concentration in experimental sepsis of baboon lipopolysaccharide-binding protein.

Lipopolysaccharide-binding protein (LBP) is important for mediating host responses to lipopolysaccharide (LPS). The structure and properties of human, rabbit, and murine LBP have been previously described. In this study we partially characterized baboon LBP and investigated its appearance in experimental sepsis. Recurrent bacteremia was induced in baboons by infusion of live Escherichia coli organisms over a 2-hour period at 0, 24, and 48 hours. To assay baboon plasma LBP levels, an enzyme-linked immunosorbent assay with cross-reactive antibodies to human LBP was developed. Control baboon plasma LBP concentrations were 2 to 5 microg/mL. During experimental sepsis, baboon plasma LBP levels increased to between 200 and 350 microg/mL and in parallel with the increase in C-reactive protein levels. Baboon LBP was isolated from acute phase serum by ion-exchange chromatography followed by immuno-affinity chromatography. Its NH2-terminal sequence (XNPGLVARTTNKGLEYSAQE) and its molecular weight (approximately 60 kd) were determined and were proved to be highly homologous to human LBP.

4-Hydroxynonenal as a second messenger of free radicals and growth modifying factor.

Immunohistochemical analysis of the distribution of the lipid peroxidation product 4-hydroxynonenal (HNE) in the brain of baboons exposed to experimental hemorrhagic traumatic shock or sepsis showed that systemic oxidative stress and the thereby generated HNE affect the blood:brain barrier and the regulation of cerebral blood flow determining secondary brain damage. Similarly, HNE was determined during ischemia in the brain blood vessels of rats exposed to ischemia/reperfusion injury of the brain. After reperfusion, HNE disappeared from the blood vessels but remained in neurones and in glial cells. Since HNE modulates cell proliferation and differentiation (including proto-oncogene expression), it is postulated that HNE might have prominent local and systemic effects that are not only harmful but beneficial, too, determining the outcome of various pathophysiological conditions based on oxidative stress.

Anti-L-selectin antibody treatment of hemorrhagic-traumatic shock in baboons.

OBJECTIVES: The adhesion molecule L-selectin plays an important role in leukocyte-endothelium interactions, thereby contributing to inflammatory reactions. We tested the hypothesis that humanized anti-L-selectin antibodies reduce trauma-associated organ damage and mortality. DESIGN: Prospective, randomized experimental study. SETTING: Independent nonprofit research laboratory in a trauma hospital (Ludwig Boltzmann Institute) and a contract research institute (Biocon). SUBJECTS: Twenty-eight male baboons (Papio ursinus), 18 to 29 kg. INTERVENTIONS: Hemorrhagic-traumatic shock was created by complement activation with cobra venom factor, followed by withdrawal of blood to a mean arterial pressure of 35 to 45 mm Hg. Blood and lactated Ringer's solution were reinfused. Animals were randomized to receive either 2 mg/kg humanized anti-L-selectin antibody (HuDREG-55 [Ab]) or placebo (lactated Ringer's solution [LRS]). MEASUREMENTS AND MAIN RESULTS: Treatment with humanized anti-L-selectin antibody decreased mortality (Ab 21% vs. LRS 71%; p = .011) and improved survival time (p = .016). A trend toward reduced organ damage, especially in the adrenal glands (score 1.2 +/- 0.2 placebo vs. 1.0 +/- 0.1 antibody; p = .059) was seen, and at 24 hrs was accompanied by significantly increased mean arterial pressure (Ab 99 +/- 6 mm Hg vs. LRS 79 +/- 8 mm Hg; p = .023), cardiac output (Ab 3.4 +/- 0.2 L/min vs. LRS 2.4 +/- 0.3 L/min; p = .007), core temperature (p = .048), and improved perfusion, with less negative base excess (Ab 2.9 +/- 1.1 vs. LRS 2.1 +/- 1.7; p = .019) and a trend toward less lactate (p = .065). These improvements were accompanied by significantly (p = .006) decreased fluid requirements in the treatment group (Ab 11.7 +/- 2.5 mL/kg/hr vs. LRS 23.0 +/- 2.3 mL/kg/hr). There were also fewer circulating leukocytes (p = .042) in the treatment group at 24 hrs. CONCLUSION: Humanized anti-L-selectin antibody has beneficial effects on survival in a long-term in vivo model of hemorrhagic. traumatic shock.

Bacterial cell envelopes (ghosts) but not S-layers activate human endothelial cells (HUVECs) through sCD14 and LBP mechanism.

Bacterial cell-envelopes (called ghosts) and surface layers (S-layers) are discussed to be used as vaccines and/or adjuvants, consequently it is necessary to find out which immunomodulatory mediators are induced in human cells. The present work focuses on the effects of ghosts (Escherichia coli O26:B6), S-layers (Bacillus stearothermophilus) in comparison with LPS and antibiotic-inactivated whole bacteria (E. coli O26:B6) on human umbilical vein endothelial cells (HUVEC) with regard to the release of interleukin 6 (IL-6) and the expression of surface E-selectin and the role of lipopolysaccharide binding protein (LBP), soluble CD14 (sCD14) and serum for this activation. Endothelial cells responded to ghosts, whole bacteria and LPS with IL-6 release up to 15000 pg/ml and surface E-selectin expression, while in contrast the response to S-layers with IL-6 release up to 500 pg/ml was very weak. Compared to LPS, 10-100-fold higher concentrations of bacterial ghosts and whole bacteria were required to induce the cytokine synthesis and E-selectin expression. IL-6 release and E-selectin expression of HUVECs were reduced in the absence of serum and equivalent to unstimulated samples. We have also studied the role of CD14 and LBP for the activation of endothelial cells using antiCD14 and antiLBP antibodies (Ab). AntiCD14 and antiLBP Ab both inhibited IL-6 release and E-selectin expression in a dose dependent manner after stimulation with ghosts, whole bacteria and LPS but had no effect on S-layers stimulated cells. AntiCD14 Ab inhibited more effectively than antiLBP Ab. These findings suggest that bacterial ghosts but not S-layers activate HUVECs through sCD14 and LBP dependent mechanisms.

Impaired energy metabolism in hearts of septic baboons: diminished activities of Complex I and Complex II of the mitochondrial respiratory chain.

Recent findings support the view that the bioenergetic part of septic organ failure is not caused by insufficient supply of oxygen but by disturbances of the mitochondrial function. Therefore, the aim of the present study was to investigate key enzymes of energy metabolism in septic hearts to answer the question whether or not impairment of mitochondrial or glycolytic enzymes occur under these conditions. For this purpose the well established model of septic baboons was used. Baboons under general anesthesia were made septic by infusion of Escherichia coli. Single challenge with infusion of high amounts of bacteria was compared with a multiple challenge protocol (less bacteria infused). Some animals obtained no E. coli (sham). The hearts of the baboons were removed after 72 h (survival: yes) or after death (survival: no) of the animals, frozen in liquid nitrogen, and stored at -80 degrees C until spectrophotometrical measurement of nine mitochondrial and glycolytic enzymes. A reduction of the activity of NADH:cytochrome-c-reductase (Complex I + III) to 67% and succinate:cytochrome-c-reductase (Complex II + III) to 45% was found in the hearts of surviving animals after infusion of high amounts of bacteria. After multiple challenge with lesser amounts of bacteria, no significant changes in enzyme activity were detectable. After lethal septic shock, activities of Complex I + III (12%) and Complex II + III (13%) as well as of phosphofructokinase (16%) were found to be strongly diminished. Decylubiquinol:cytochrome-c-reductase (Complex III, 59%), cytochrome-c-oxidase (51%), succinate dehydrogenase (60%), glucosephosphate isomerase (61%), lactate dehydrogenase (61%), and citrate synthase (120%) were less or unaffected. Similar but less pronounced effects were found after infusion of lesser amounts of bacteria. By means of inhibitor titrations of succinate: cytochrome-c-reductase, it was shown that the loss of activity is not caused by Complex III but by disturbances in Complex II. It is concluded that E. coli-induced sepsis causes decreased activities of Complex I and Complex II in baboon heart mitochondria in a dose-dependent manner.

Intramedullary pressure and bone marrow fat extravasation in reamed and unreamed femoral nailing.

This study was designed to investigate whether intramedullary pressure and embolization of bone marrow fat are different in unreamed compared with conventional reamed femoral nailing in vivo. In a baboon model, the femoral shaft was stabilized with interlocking nailing after a midshaft osteotomy. Intramedullary pressure was measured in the distal femoral shaft fragment at the supracondylar region. Extravasation of bone marrow fat was determined by the modified Gurd test (range: 0-5) with blood samples from the vena cava inferior. Data were monitored in eight unreamed and eight reamed intramedullary femoral nailing procedures. Intramedullary pressure increased in the unreamed group to 76 +/- 25 mm Hg (10.1 +/- 3.3 kPa) during insertion of 7-mm nails and in the reamed group to 879 +/- 44 mm Hg (117.2 +/- 5.9 kPa) during reaming of the medullary cavity. Insertion of 9-mm nails after the medullary cavity had been reamed to 10 mm produced an intramedullary pressure of 254 +/- 94 mm Hg (33.9 +/- 12.5 kPa) (p < 0.05). Fat extravasation in the unreamed group was recorded with a score of 2.9 +/- 0.4 for the Gurd test during nailing with 7-mm nails, whereas in the reamed group significantly more fat extravasation was noticed during the reaming procedures, with a score of 4.6 +/- 0.1. Liberation of fat during insertion of 9-mm nails after reaming was recorded with a score of 3.5 +/- 0.4. In both groups, a positive correlation of fat extravasation with the rise in intramedullary pressure was found (reamed group: r(s) = 0.868; unreamed group: r(s) = 0.698), resulting in significantly less liberation of bone marrow fat in the unreamed stabilized group than in the reamed control group (p < 0.05). The data indicate that fat embolization during nailing procedures after femoral osteotomy increases with increasing intramedullary pressure and occurs in a lesser degree in unreamed than in reamed intramedullary femoral shaft stabilization.

In vivo femoral intramedullary pressure during uncemented hip arthroplasty.

There is evidence in several animal and human studies that high intramedullary pressure in the femur is of causal significance for bone marrow release into the circulation, causing pulmonary fatty marrow embolization. A previous clinical study provided evidence that in uncemented hip arthroplasty, high intramedullary pressure and subsequent fat embolism with cardiorespiratory deterioration can occur. In this prospective clinical trial, the effect of five surgical techniques on the femoral intramedullary pressure was recorded intraoperatively in 36 patients during uncemented press fit hip arthroplasty. In Group A, the conventional surgical technique (slide hammer and femoral rasps) showed intramedullary hypertension during opening of the femoral canal, femur preparation, and prosthesis insertion. In Group B, a mechanical high frequency vibration rasp was used, instead of the slide hammer, and provided reduction of the intramedullary pressure peaks during opening of the femoral canal but could not prevent intramedullary hypertension during rasping and prosthesis insertion. In Group C, a modified surgical technique to prevent high intramedullary pressure reduced pressure peaks during opening of the femoral canal and resulted in a significant reduction of intramedullary pressure during femur preparation and prosthesis insertion compared with the conventional surgical technique used with Group A. In Group D the results of the modified surgical technique could be improved additionally by using the high frequency vibration rasp, instead of the slide hammer. In Group E conventional surgical technique in combination with a distal venting hole has not proven to be efficient in uncemented hip arthroplasty. Based on the results of this in vivo study, the proposed modified surgical technique in cementless hip arthroplasty can be recommended to avoid high intramedullary pressure peaks, which should minimize the risk of significant bone marrow release into the circulation and the risk for cardiorespiratory deterioration caused by fat embolism.

Protective effect of bactericidal/permeability-increasing protein (rBPI21) in baboon sepsis is related to its antibacterial, not antiendotoxin, properties.

OBJECTIVE AND SUMMARY BACKGROUND DATA: The recombinant fragment of bactericidal/permeability-increasing protein, rBPI21, has potent bactericidal activity against gram-negative bacteria as well as antiendotoxin (lipopolysaccharide [LPS]) action. On the basis of these activities, the authors sought to discover whether rBPI21 would be protective in baboons with live Escherichia coli-induced sepsis and whether the potential protective effects of rBPI21 (together with antibiotics) would be more closely related to its antibacterial or LPS-neutralizing effects. METHODS: In a prospective, randomized, placebo-controlled subchronic laboratory study, the efficacy of rBPI21 or placebo was studied over 72 hours in chronically instrumented male baboons infused with live E. coli under antibiotic therapy. RESULTS: Intravenous rBPI21 attenuated sepsis-related organ failure and increased survival significantly. Bacteremia was significantly reduced in the rBPI21 group at 2 hours after the start of the E. coli infusion, whereas circulating LPS was less affected. The in vivo formation of tumor necrosis factor was significantly suppressed by the rBPI21 treatment regimen. Microcirculation and organ function were improved. CONCLUSIONS: In baboon live E. coli sepsis, the salutary effect of rBPI21 results from a more prevalent antibacterial than antiendotoxin activity.

Anti-TNF treatment of baboons with sepsis reduces TNF-alpha, IL-6 and IL-8, but not the degree of complement activation.

The activation of complement and the release of TNF-alpha, IL-6 and IL-8 are important pathogenic factors behind organ dysfunction in sepsis. The aim of this study was to determine whether infusion of anti-TNF antibodies alters complement activation and plasma concentrations of pro-inflammatory cytokines at high doses of Escherichia coli. Six baboons received intravenously 2 x 10(9) live E. coli bacteria per kg body weight (group 1), in addition five received pretreatment with 1 mg per kg body weight anti-TNF antibodies (group 2), and seven received 5 x 10(8) live E. coli bacteria per kg body weight (group 3). Two hours after the start of infusion of the bacteria, plasma concentrations of C3 activation products, C5a and the terminal SC5b-9 complement complex were increased in groups 1 and 2 (P < 0.05), but there was no significant difference between the groups. At 2 h the levels of TNF-alpha, IL-6 and IL-8 were lower in group 2 compared with group 1 (P<0.05). In group 2 compared with group 1 the TNF-alpha concentrations were, however, higher at 4, 8 and 24 h. The explanation for this phenomenon is probably that TNF-alpha binds to the anti-TNF antibody complex and is released slowly after it has been bound. The study showed that infusion of anti-TNF antibodies reduced the concentrations of TNF-alpha, IL-6 and IL-8, without any detectable influence on complement activation.

Local and systemic reactions after lung contusion: an experimental study in the pig.

The present study was designed to investigate the consequences of isolated unilateral lung contusion on local alveolar and systemic inflammatory responses in an animal model in the pig. Isolated unilateral lung contusion was induced by bolt shot in eight mechanically ventilated animals under general anesthesia (sham: n=4). Plasma and bronchoalveolar lavage fluid were collected during a period of 8 h following lung contusion. Leukocytes, leukocyte neutral protease inhibitor (LNPI), terminal complement complex (TCC), thrombin-antithrombin-complex (TAT) as well as pulmonary microvascular permeability and surfactant function were determined. Within 30 min, lung contusion was found to cause a significant local and systemic increase in TCC and TAT concentrations and a systemic increase in LNPI concentrations. The latter was accompanied by a sequestration of leukocytes in the contused lung. Complement activation and leukocyte sequestration in the contused lung progressively increased during the investigation period. Although surfactant function decreased in the entire lung 30 min after contusion, TCC, TAT, and leukocyte sequestration was unchanged in the contralateral lung. The first indication of an involvement of the contralateral lung was obtained by an increase in leukocyte sequestration 8 h after lung contusion. Unilateral lung contusion initiates an early systemic activation of humoral and cellular defense systems. Involvement of the contralateral lung appears to be a secondary event caused by a systemic inflammatory reaction.

The inflammatory basis of trauma/shock-associated multiple organ failure.

Multiple alterations in inflammatory and immunologic function have been demonstrated in clinical and experimental situations after trauma and hemorrhage, in particular the activation of various humoral (e.g. complement, coagulation) and cellular systems (neutrophils, endothelial cells, macrophages). As a consequence of this activation process there is synthesis, expression and release of numerous mediators (toxic oxygen species, proteolytic enzymes, adherence molecules, cytokines), which may produce a generalized inflammation and tissue damage in the body. Mediators are responsible for ongoing interactions of different cell types and for amplification effects through their networks and feedback cycles, finally leading to a sustained inflammation and multiple organ damage in the body. In the setting of trauma/shock, many activators including bacterial as well as non-bacterial factors may be present that will induce local and systemic inflammatory responses. Although the potential role of bacteria/endotoxin translocation and its clinical relevance remains controversial, many lines of evidence support the concept that the gut may be the reservoir for systemic sepsis and subsequent MOF in a number of pathophysiologic states.

[Endotoxin tolerance and hemorrhagic shock in rats]. / Endotoxintoleranz und hämorrhagischer Schock bei Ratten.

The aim of the presents study was to investigate the protective capacity of endotoxin tolerance in a hemorrhagic shock model in rats. A pretreatment with low dose endotoxin induces a state of tolerance, which is characterized by decreased TNF alpha production in vivo and in vitro upon subsequent high dose endotoxin challenge. This endotoxin tolerance improves survival after hemorrhagic shock from 22.8% in untreated controls to 68.8 in tolerant rats. The protection was accompanied with the appearance of n TNF alpha inhibitory activity in the serum of endotoxin tolerant animals, which might be responsible for the improved survival after hemorrhagic shock.

[Transfusion of plasma of endotoxin tolerant rats improves survival of hemorrhagic shock in the rat model]. / Die Transfusion von Plasma endotoxintoleranter Ratten verbessert die Uberlebensrate beim hämorrhagischen Schock im Rattenschockmodell.

To evaluate the protective effect of endotoxin tolerant plasma after hemorrhagic shock, plasma of endotoxin tolerant animals was used for resuscitation of normal rats. The transfusion of plasma of endotoxin tolerant rats improved the survival rate from 20 to 60 per cent. In the sera of endotoxin tolerant animals a TNF inhibitory activity was detected, which could be transferred by the plasma. This effect may be responsible for the protection to hemorrhagic shock.