BACKGROUND: Perfusion failure in the microcirculation, known as "no-reflow phenomenon", due to ischemia-/reperfusion injury might cause severe problems following resuscitation from shock or sepsis. This situation often develops within a few days after the initial injury. The implication of the Platelet Activating Factor (PAF) in the development of microcirculatory disturbances after various conditions of shock, sepsis and ischemia-reperfusion incidents was shown in a variety of studies. We evaluated the role of PAF in the hepatic microcirculation using an intra-vital microscopy (IVM) rat-model of hemorrhagic shock. METHODS: The animals were randomly assigned to shock groups with microscopy after two or five days or corresponding control-groups without shock. After pressure controlled hemorrhagic hypotension of 90 minutes, shock rats were treated with WEB 2086 (1 mg/kg) or saline in a randomized and blinded manner followed by three hours of resuscitation with Ringers. Intravital microscopy of the liver was performed two or five days after shock. RESULTS: Two days after shock the sinusoidal leukocyte flow was significantly improved by WEB 2086. However, after five days no difference was seen between all groups. After two days, but not after five days, the perfusion-index, indicating the proportion of perfused sinusoids to all investigated sinusoids, was significantly increased by WEB 2086 in comparison to the saline treated animals. Hemorrhagic shock induced a long-lasting narrowing of the sinusoids up to five days. The mean sinusoidal diameters were significantly increased by WEB 2086 compared with the Shock group at day two and five. CONCLUSIONS: The results suggest a positive effect of WEB 2086 on the shock-induced microcirculatory dysfunction.

Hepatic microcirculatory disorders such as narrowing of sinusoids after hemorrhagic shock play a major role in the pathogenesis of organ failure. It is known that the balance of vasoactive mediators such as endothelin and nitric oxide (NO) regulate microvascular perfusion, including the diameter of hepatic sinusoids. The present study was designed to evaluate the role of exogenous substitution of NO by S-nitroso-albumin (S-NO-HSA) in the prevention of pathophysiological alterations of hepatic microcirculation. Anesthetized Sprague-Dawley rats were instrumented for invasive hemodynamic monitoring. Hemorrhagic shock was induced by bleeding to a mean arterial pressure (MAP) of 40 mmHg and was maintained for 60 min. Thereafter, the animals were resuscitated with shed blood and Ringer's solution. During the first hour of resuscitation, S-NO-HSA or pure HSA was infused continuously (10 mumol/kg/h) and hepatic microcirculation was detected by intravital epifluorescence microscopy either 5 or 24 h after the insult. Results were compared with a sham-treated group (n = 6-8 per group). Shock-induced microcirculatory narrowing of sinusoids was significantly reduced in the S-NO-HSA group compared with the HSA group both at 5 and 24 h (HSA: 9.3 +/- 0.2 mum; S-NO-HSA: 12.1 +/- 0.2 mum, P < 0.05). Sinusoidal perfusion was significantly higher in the S-NO-HSA group than in the HSA group (HSA: 50,934 +/- 1,382 mum(3)/s; S-NO-HSA: 78,120 +/- 2,348 mum(3)/s, P < 0.05). Reversible leukocyte adhesion to sinusoidal endothelium, an indicator of the inflammatory response, was significantly reduced in the S-NO-HAS-treated group. The findings of this study in a rat model of hemorrhagic shock suggest that NO substitution by S-NO-HSA during resuscitation attenuates both early and late hepatic microcirculatory disturbances as well as the increase in leukocyte adherence.

ObjectiveWe investigated the potential of dietary fish oil containing n-3 polyunsaturated fatty acids to attenuate hepatic injury and mortality rate of rats in response to systemic endotoxemia. DesignProspective, randomized, controlled animal study. SettingUniversity laboratory. SubjectsA total of 43 male Sprague Dawley rats. InterventionsRats were fed either fish oil supplement or regular standard lab chow. After 8 wks of feeding, each diet group was subjected to a single exposure of lipopolysaccharide (Escherichia coli, 10 mg/kg intravenously) or saline. Hepatic microvascular response and liver injury were assessed by in vivo analysis of Kupffer cell phagocytic activity, leukocyte-endothelial cell interaction, nutritive sinusoidal perfusion failure, and parenchymal cell apoptosis (intravital fluorescence epi-illumination technique) as well as bile flow, serum liver enzyme activities, and tissue histomorphology. Measurements and Main ResultsIn animals fed a standard diet, livers at 16 hrs after lipopolysaccharide-exposure exhibited depressed Kupffer cell phagocytic activity, enhanced hepatic microvascular leukocyte activation, leukocytic tissue infiltration, sinusoidal perfusion failure, and parenchymal cell apoptosis. Hepatic microvascular injury was further accompanied by reduced bile flow and enhanced liver enzyme release. The fish oil enriched diet did not significantly change the multiple features of endotoxemia-associated liver injury; however, it maintained arterial blood pressure, systemic leukocyte count, and acid base balance and showed a tendency toward improved survival on lipopolysaccharide exposure with a 16 hr-survival rate of 80% (p = .06 vs. survival rate of 40% in animals fed a regular diet). Moreover, slightly increased serum concentrations of interleukin-10 coincided with enhanced concentrations of interleukin-6 in fish oil fed endotoxemic animals. Healthy, non-lipopolysaccharide-exposed, fish oil fed animals did not differ from those fed with the regular diet, except for dampened Kupffer cell phagocytic activity. ConclusionsFish oil feeding does not protect from local endotoxemia-induced hepatic microvascular dysfunction. However, dietary modulation of inflammatory mediator response by macrophages, constituting an appropriate immune response, could add to the survival advantage seen in fish oil-fed animals on exposure to lipopolysaccharide.

Abstract Background The liver plays a key role in the host defence response after haemorrhagic shock–resuscitation (H/R) and sepsis. Kupffer cells (KCs) have been shown to be a trigger and motor of the subsequent inflammatory response syndrome. This may lead to hepatocellular dysfunction, microcirculatory alterations and liver injury involving, for example, tumour necrosis factor α, interleukin (IL) 1 and IL-6. In a double-blind study the effect of KC blockade with either gadolinium chloride or liposome-entrapped dichloromethylene diphosphonate (DMD) on hepatic microvascular flow after H/R and sepsis was investigated. Methods After pretreatment with intravenous gadolinium chloride 10 mg kg−1, DMD 1 mg kg−1 or saline 24 h before induction of shock, male Sprague-Dawley rats (n = 6–10 per group and time) were subjected to either haemorrhagic shock (mean arterial pressure 40 mmHg) for 60 min followed by resuscitation or lipopolysaccharide (LPS) 1 mg kg−1 intravenously. Microvascular flow was assessed by intravital microscopy of fluorescence-marked leucocytes in liver sinusoids at baseline, and 1, 6 and 12 h after shock induction. Results In saline groups, the mean(s.d.) leucocyte flow was significantly (P < 0·05) higher at 1 h (20 759(2901) μm3 s−1) and 6 h (16 278(2916) μm3 s−1) after H/R as well as at 6 h after LPS (17 661(3949) μm3 s−1) compared with the baseline value (13 509(1580) μm3 s−1). Animals pretreated with gadolinium chloride showed a significant flow increase compared with baseline (11 797(1124) μm3 s−1) at l h following H/R (26 269(5909) μm3 s−1). In DMD-pretreated animals leucocyte flow showed no significant change over time, following either H/R or LPS treatment. However, flow was significantly higher at baseline (18 054(998) μm3 s−1) versus gadolinium chloride and saline groups. In addition, DMD-treated animals showed higher flow values 1 h after LPS challenge (20 665(2337) μm3 s−1) compared with gadolinium chloride (13 110(1224) μm3 s−1) and saline (15 311(800) μm3 s−1) groups. Similarly, at 12 h after H/R the DMD group (21 782(1887) μm3 s−1) had higher flow values than the gadolinium chloride (14 026(1616) μm3 s−1) and saline (15 999(3175) μm3 s−1) groups. Conclusion These results imply a significant influence of KCs on regulation of microvascular perfusion in liver sinusoids under normal conditions as well as after H/R and sepsis. The data indicate differential pathways and effects of blocking KCs by gadolinium chloride and DMD.

Background The generation of iron-dependent toxic oxygen radicals during the initial resuscitation from hemorrhagic shock was shown to be a relevant factor for the initiation of the inflammatory cascade. Therefore, this experimental study was designed to evaluate the effects of a deferoxamine-conjugated hydroxyethyl-starch solution (HES-DFO) on oxygen radical induced injury and microcirculatory alterations in the rat liver compared with resuscitation with regular hydroxyethyl-starch, lactated Ringer's solution (RL), or a gelatin-based solution. Methods After hemorrhage and random assignment to 1 hour of blood-free resuscitation with the aforementioned solutions, hepatic microcirculation and leukocyte adhesion characteristics were assessed by intravital fluorescence microscopy in anesthetized rats. Oxygen radical activity was estimated by determination of glutathione levels in liver homogenate and determination of thiobarbituric acid-reactive substances in plasma as markers of lipid peroxidation. Results Resuscitation by HES-DFO resulted in restoration of hemodynamic parameters compared with gelatin-based solution and HES. The hepatic microcirculation was severely altered 1 hour after resuscitation from shock in all groups indicated by sinusoidal narrowing and reduced sinusoidal blood flow. HES-DFO, however, attenuated leukocyte adhesion and improved velocity index in sinusoids as well as sinusoidal perfusion. The shock-associated generation of oxygen radicals during resuscitation was prevented by HES-DFO as indicated by restored glutathione and reduced thiobarbituric acid-reactive substances. Conclusion The results suggest that HES-DFO effectively reduces oxygen radical formation during the initial resuscitation period, thus, attenuating pathologically enhanced leukocyte adhesion and improving hepatic microcirculation.

Objectives: Rapid post-anaesthetic awakening and low hypnotic potency are two characteristic properties of the new opioid remifentanil. For clinical use remifentanil must be combined with another anaesthetic agent. Propofol is well-established for ambulatory anaesthesia, however, the properties of desflurane (low blood-gas solubility, rapid elimination) suggest this volatile anaesthetic to be a comparable alternative, particularly if rapid awakening is desired. The present study was designed to compare emergence times and haemodynamics for a combination of remifentanil wich hypnotic concentrations of either propofol or desflurane. Methods: Gynaecological patients, scheduled for elective laparoscopy, were studied at random. After oral premedication with diazepam 0.08-0.12mg/kg, anaesthesia was induced identically in both groups: remifentanil bolus (1 mu g/kg), start of remifentanil infusion (0.5 mu g/kg/min), followed by propofol (approx. 2 mg/kg) and cisatracurium (0.1 mg/kg). For maintenance of anaesthesia remifentanil (0.25 mu g/kg/min) was combined with either desflurane (0.5 MAC=3.0vol%) or propofol (6 mg/kg/h). With termination of surgery anaesthetic delivery was discontinued simultaneously and recovery times were recorded. Heart rate and non-invasive blood pressure were recorded at defined points of interest. Results: In total, 40 patients (desflurane n=20, propofol n=20) were studied in comparable groups. For both regimens, emergence after remifentanil-based anaesthesia was remarkably rapid between unconsciousness and complete recovery: In mean only 60 s elapsed from the onset of spontaneous breathing to the moment when patients could clearly state their name. In comparison, recovery times were significantly shorter after remifentanil-desflurane than after remifentanil-propofol: time to spontaneous ventilation 6.4 +/- 2.8 vs. 9.6 +/- 3.9 min (mean +/- SD, p = 0.01); extubation 6.7 +/- 2.8 vs. 9.8 +/- 4.0 min (p = 0.02) and arrival at PACU 11.1 +/- 3.4 vs. 14.7 +/- 4.2 min (p = 0.005). The courses of heart rate (HR) and mean arterial pressure (MAP) were mostly similar in both groups with only minimal or moderate cardiocirculatory reactions during intubation or start of surgery. Conclusions: Remifentanil in combination with either desflurane or propofol, used for general anaesthesia during gynaecological laparoscopy, will facilitate a smooth haemodynamic course as well as a rapid emergence thereafter. Recovery times after remifentanil-based anaesthesia are significantly shorter with 3.0 vol% of desflurane than with 6 mg/kg/h propofol. Thus, desflurane appears to be a well-suited adjunct to remifentanil and an ideal alternative to propofol, if rapid recovery is required. Differences are best explained by the pharmacological properties of both co-anaesthetics and their applied dosages.

Kohlenmonoxid (CO) entsteht als Intermediärstoffwechselprodukt im Hämabbau und kann nach neuesten Erkenntnissen ähnlich wie Stickstoffmonoxid über Aktivierung der Guanylatzyklase vasodilatorische Wirkungen entfalten. Für die endogene Bildung von CO ist ausschließlich die mikrosomale Hämoxygenase (HO) verantwortlich, die als konstitutives Enzym (HO-2) sowie in einer streßinduzierbaren Isoform (HO-1 oder Hitzeschockprotein 32) vorliegt. Hemmung des Hämstoffwechsels nach Schock und Volumentherapie führt nach eigenen Untersuchungen zu einer spezifischen Zunahme des Pfortaderwiderstands, ohne vasokonstriktorische Wirkungen in der systemischen Zirkulation hervorzurufen. In der vorliegenden Arbeit wurde daher das organspezifische Expressionsmuster der streßinduzierbaren Isoform des Enzyms nach hämorrhagischem Schock und Volumentherapie untersucht.

This study was designed to investigate the effects of endogenous adenosine by the use of a novel adenosine kinase-inhibitor (GP-1-515) on pathological hepatic leukocyte adhesion following hemorrhagic shock.Anesthetized rats were prepared for invasive hemodynamical monitoring and subjected to hemorrhagic shock for 60 min. Five hours after adequate resuscitation the hepatic microcirculation was investigated in vivo using epifluorescence microscopy. The adenosine kinase inhibitor GP-1-515 (0.25 mg/kg i.v.) was given in a randomized and blinded manner before shock induction.GP-1-515 had no significant effect on central hemodynamics. Inhibition of adenosine kinase by GP-1-515 significantly reduced shock-induced permanent leukocyte adhesion to sinusoidal endothelium. Narrowing of sinusoids was partially prevented and sinusoidal blood flow was improved by GP-1-515The inflammatory process following hemorrhagic shock reflected by enhanced leukocyte adhesion is effectively attenuated by GP-1-515 due to increased endogenous adenosine levels.

Deferoxamine is known to reduce the iron-dependent generation of toxic oxygen-derived radicals during reperfusion of ischaemic tissue. The present study investigates the antioxidative properties of a deferoxamin-conjugated hydroxyethyl starch solution and its effects on the hepatic microcirculation in a haemorrhagic-shock rat model.Methods: Anaesthetized Sprague-Dawley rats were tracheotomized, prepared for invasive haemodynamic monitoring, and subject to haemorrhagic shock (MAP=40 mmHg during 60 min). The animals were resuscitated blood-free with lactated Ringer's (RILA, n=10), gelatine (GELA, n=10), hydroxyethyl starch (HES, n=10), or deferoxamine-conjugated HES (DFO, n=8) solution (MAP greater than or equal to 70 mmHg). After 1 h of resuscitation the hepatic microcirculation was investigated by intravital microscopy, the glutathione concentration was measured in liver homogenate, and the thiobarbituric acid reactive substances (TBARS) were determined as markers of lipid peroxidation.Results: Resuscitation resulted in restoration of MAP to greater than or equal to 70 mmHg within a short time. The volume required to stabilise the arterial pressure during 1 h of resuscitation was significantly less in the DFO group compared with HES, GELA, and RILA. Significantly higher glutathione levels in liver homogenate as well as decreased TBARS levels were observed in the DFO group. The shock-induced increase of leukocyte adhesion in liver sinusoids was significantly attenuated by DFO.Conclusion: DFO significantly attenuates shock-induced oxidative stress, thereby reducing the early inflammatory reaction and improving the hepatic microcirculation.

Bouma, M. G.; Bauer, C.; Herrmann, I.; van den Wildenberg, F. A.J.M.; Firestein, G. S.; Marzi, I.; Buurman, W. A.

UNLABELLED:Deferoxamine is known to reduce the iron-dependent generation of toxic oxygen-derived radicals during reperfusion of ischaemic tissue. The present study investigates the antioxidative properties of a deferoxamine-conjugated hydroxyethyl starch solution and its effects on the hepatic microcirculation in a haemorrhagic-shock rat model.METHODS:Anaesthetized Sprague-Dawley rats were tracheotomized, prepared for invasive haemodynamic monitoring, and subject to haemorrhagic shock (MAP = 40 mmHg during 60 min). The animals were resuscitated blood-free with lactated Ringer's (RILA, n = 10), gelatin (GELA, n = 10), hydroxyethyl starch (HES, n = 10), or deferoxamine-conjugated HES (DFO, n = 8) solution (MAP > or = 70 mmHg). After 1 h of resuscitation the hepatic microcirculation was investigated by intravital microscopy, the glutathione concentration was measured in liver homogenate, and the thiobarbituric acid reactive substances (TBARS) were determined as markers of lipid peroxidation.RESULTS:Resuscitation resulted in restoration of MAP to > or = 70 mmHg within a short time. The volume required to stabilise the arterial pressure during 1 h of resuscitation was significantly less in the DFO group compared with HES, GELA, and RILA. Significantly higher glutathione levels in liver homogenate as well as decreased TBARS levels were observed in the DFO group. The shock-induced increase of leukocyte adhesion in liver sinusoids was significantly attenuated by DFO.CONCLUSION:DFO significantly attenuates shock-induced oxidative stress, thereby reducing the early inflammatory reaction and improving the hepatic microcirculation.

Background/Aims: Nitric oxide (NO) is an important mediator in the regulation of vascular tone. However no data exist on the physiological role of NO in the regulation of the hepatic microcirculation. This tudy was designed to evaluate the role of NO in the hepatic microcirculation in vivo under physiological conditions.

Adhesion of leukocytes to the vascular endothelium hallmarks a key event in neutrophil-mediated organ injury after ischemia-reperfusion. The autacoid adenosine has been shown to inhibit activated neutrophil function and to interfere with leukocyte-endothelial adherence. Its therapeutic use in ischemia-reperfusion, however, has been limited by severe cardiovascular side effects. We therefore investigated the effects of the adenosine kinase inhibitor GP515 in vivo on hepatic leukocyte-endothelial interactions in a rat model of hemorrhagic hypotension and resuscitation, using intravital microscopy. Rats were pretreated with either GP515 (0.25 mg/kg) or saline in a randomized and blinded manner and subjected to pressure-controlled hemorrhagic hypotension at a mean arterial pressure of 40 mmHg for 60 min followed by 5 h of resuscitation. Five hours after resuscitation in saline-treated animals, firm leukocyte-sinusoidal adhesion was strongly enhanced in the periportal and midzonal sublobular regions, and sinusoidal diameters were also markedly reduced. Compared with saline treatment, GP515 significantly attenuated shock and resuscitation-induced leukocyte adhesion in both sublobular regions. Moreover, although GP515 did not significantly affect macrohemodynamical and hematological parameters, it enlarged narrowed sinusoidal diameters and tended to improve sinusoidal blood flow. We propose that the adenosine-regulating agent GP515 has a therapeutic potential to attenuate ischemia-reperfusion-induced inflammation by capitalizing on the beneficial anti-inflammatory effects of endogenous adenosine.

Defined microcirculatory disturbances have been demonstrated during the first hours after hemorrhagic shock. Since clinical complications due to inflammatory reactions (SIRS, MOF) usually develop within days, our aim was to establish an experimental model of hemorrhagic shock simulating this clinical situation. Therefore, a non-lethal rat model was developed allowing investigation of the hepatic microcirculation by intravital microscopy up to day 4 after shock.Anesthetized Sprague-Dawley rats were instrumented with a tail vein cannula and a carotid artery catheter for measurement of mean arterial blood pressure (MAP) and blood withdrawal. Hemorrhagic shock was induced to a MAP of 40 mmHg for 40 min. Thereafter, animals were adequately resuscitated by shed blood and Ringer's. One, 2, 3, or 4 days after hemorrhagic shuck, the rats were anesthetized again and instrumented prior to evaluation of the hepatic microcirculation using intravital fluorescence microscopy. The systemic hemodynamical changes were comparable in all groups during shock and resuscitation. Investigation of the hepatic microcirculation, however, demonstrated a biphasic decrease of sinusoidal diameters with narrowed sinusoids at days 2 and 3 after shock. Permanent adhesion of leukocytes rose from control values continuously up to day 2 and remained still elevated at day 4. The biphasic pattern of sinusoidal narrowing suggests a secondary vasoconstrictor activity (endothelin activity or Ito-cell priming ?) induced during shock period. Furthermore, the consistent and lasting induction of leukocyte adhesion indicates prolonged expression of adhesion molecules on the sinusoidal endothelial surface, which has been shown to involve ICAM-1 during the first hours after shock. Thus, shuck induces a long-term alteration of the hepatic microcirculation, possibly contributing to an altered systemic immune function.

Kupffer cell activation, pathological leukocyte adhesion and deterioration of sinusoidal perfusion in the liver after hemorrhagic shock have been attributed to the low How situation during shock period. As a consequence, exposure of sinusoidal cells to gut-derived endotoxin has been proposed to cause intrahepatic release of numerous of inflammatory mediators leading to microvascular perfusion alterations. Granulocyte derived bactericidal/permeability increasing protein (BPI) has been recognized as a potent endotoxin binding protein during host defense. Therefore, the aims were to determine the impact of early and late administration of recombinant BPI on hepatic microcirculation after hemorrhagic shock.Methods. Hemorrhagic shock was induced in anesthetized female Sprague Dawley rats by reduction of the mean arterial blood pressure to 40 mmHg for 60 min. Rye hours following resuscitation, the hepatic microcirculation was investigated using epifluorescence intravital microscopy. Sinusoidal perfusion and leukocyte adhesion were evaluated using computer-assisted image analysis. Beside one sham operated control group (n=6) four shock groups (n=8) were studied receiving rBPI(21)(10 mg/kg: XOMA) or placebo (thaumatin) either as pretreatment prior to shock induction or as early treatment al the beginning of resuscitation period. In a separate series, the effects of rBPI(21) on survival and systemic endotoxin levels after 180 min of hemorrhagic shock were studied.Results. While systemic hemodynamical, respiratory and metabolic parameters were comparable in ail shock groups, the hepatic microcirculation revealed a severe reduction of sinusoidal diameters in all shock groups that was not affected by rBPI(21). Pathologically increased leukocyte adhesion in liver sinusoids after shock was significantly reduced by exogenously administered rBPI(21) either as pretreatment or early treatment. Endotoxin plasma levels and survival after 180 min of hemorrhagic shock were markedly reduced by BPI21 administration.Conclusion. Hemorrhagic shock induced a substantial pathological adhesion of leukocytes in the liver sinusoids, which was completely inhibited by administration of the endotoxin-binding rBPI(21). Thus, gut-derived endotoxin after hemorrhagic shock seems to play an important role for the induction of an intrahepatic inflammatory process. It seems likely, that endotoxin stimulates local release of Kupffer cell derived mediators such as TNF or IL-l which are responsible for the observed increase in sinusoidal leukocyte adhesion. The results suggest a further pathophysiological evaluation of the effects of BPI after hemorrhagic shock, gastrointestinal or liver ischemia/reperfusion injury.

OBJECTIVE:To investigate the effects of the recombinant 21-kilodalton N-terminal fragment of recombinant bactericidal and permeability increasing protein (rBPI21) on leukocyte adhesion and the hepatic microcirculation after hemorrhagic shock. DESIGN:Prospective, randomized, blinded, and placebo-controlled experimental study. SETTING:University research laboratory. SUBJECTS:Anesthetized Sprague-Dawley rats, weighing 220 to 250 g. INTERVENTIONS:Rats were subjected to 60 mins of hemorrhagic shock and subsequent resuscitation to sufficiently restore systemic circulation. The microcirculation of the liver was investigated by intravital fluorescence microscopy 5 hrs after hemorrhagic shock. Four shock groups were compared with a sham-control group. Shock groups received either rBPI21 (10 mg/kg) or placebo either before or after shock period. MEASUREMENTS AND MAIN RESULTS:No differences were observed in hemodynamic, respiratory, or metabolic parameters between the shock groups. However, the hepatic microcirculation showed severe deterioration 5 hrs after shock, indicated by significantly narrowed sinusoids in all shock groups compared with controls (8.5 +/- 0.3 microm vs. 10.0 +/- 0.4 pm). Leukocyte adhesion was markedly increased to comparable values in both placebo groups (619 cells/mm2 and 644 cells/mm2; sham, 168 cells/mm2). Neutralization of endotoxin by administration of rBPI21 before or after shock resulted in plain reduction of pathologic leukocyte-endothelial interaction (138 cells/mm2 and 85 cells/mm2). CONCLUSION:The results support the hypothesis that endotoxin induces microcirculatory alterations after shock, and further suggest a potentially beneficial role of rBPI21 in the treatment of posttraumatic endotoxin-induced inflammatory reactions.