Matthew J. Fabian

Reduced tumor necrosis factor production in endotoxin-spiked whole blood after trauma: Experimental results and clinical correlation

BACKGROUND The overproduction of tumor necrosis factor-alpha (TNF) plays a key role in virtually every experimental model of septic shock, which has led to the development of several therapies that target TNF and other cytokines in clinical sepsis. However, our previous work showed that plasma TNF was reduced, rather than increased, when a septic challenge was administered 3 days after hemorrhagic shock. In this study we compared whole-blood TNF production ex vivo in human beings and animals after trauma. METHODS TNF was measured before and after a 4-hour incubation of whole blood with 0 or 5 micrograms/ml Escherichia coli endotoxin (LPS) at 37 degrees C ex vivo. Samples were obtained from trauma patients with (n = 8) and without (n = 14) sepsis and compared with those obtained in healthy volunteers (n = 11). In parallel experiments in a pig model TNF was measured before and after fluid resuscitation from trauma after an ex vivo (0 or 5 micrograms/ml LPS) or an in vivo (5 micrograms/kg LPS, 30 minutes intravenously) challenge. RESULTS With either an immunoassay or a bioassay in either human beings or pigs before or after trauma, TNF was at or below the threshold of detection, unless the blood sample was spiked with LPS. After spiking, TNF was markedly elevated, but the increment was reduced after trauma. In pigs an LPS challenge in vivo delayed 3 days after trauma evoked an increment in plasma TNF that was blunted compared with that in an uninjured control. This trauma-induced reduction in blood TNF could not be attributed to a simple reduction in the number of monocytes nor to changes in cortisol, nor to increased numbers of neutrophils, whose proteolytic enzymes can impair production or increase the degradation of TNF. Although the plasma concentration of soluble TNF-binding proteins (60 kd) was elevated in nonsepsis (p = 0.0358) and sepsis trauma patients (p = 0.0148), the correlation with TNF production was relatively weak (R2 = 0.260). CONCLUSIONS There was no evidence of TNF overproduction in whole blood after trauma. If these results could be generalized to other tissues, it would be difficult to justify therapeutic targeting of TNF in exaggerated inflammatory response (or septic complications) after trauma.

Acadesine and lipopolysaccharide-evoked pulmonary dysfunction after resuscitation from traumatic shock.

BACKGROUND We have reported that the purine precursor acadesine (AICAR) improved the microcirculation, repleted adenosine triphosphate, and attenuated local and lung neutrophil infiltration after intestinal reperfusion and that it quickly improved systemic hemodynamics after resuscitation from hemorrhagic shock. This study evaluated the therapeutic potential of AICAR after fluid resuscitated trauma. METHODS Anesthetized (fentanyl) mongrel pigs were subjected to tissue injury plus hemorrhage and randomized to receive resuscitation fluids comprised of shed blood plus either lactated Ringers solution (LR) or AICAR (1 or 10 mg/kg bolus + 0.5 mg/kg/min x 30 min). Thereafter either LR or AICAR (1 or 10 mg/kg) was administered at 12-hour intervals for 72 hours. In a smaller series (n = 7) a single bolus (0.5 mg/kg) of the adenosine deaminase inhibitor deoxycoformycin was administered at the time of resuscitation. After 72 hours, and endotoxin challenge (0.5 microgram/kg, lipopolysaccharide [LPS]) was administered. RESULTS At 1 mg/kg (n = 9), AICAR had no obvious effect versus LR (n = 31). At 10 mg/kg AICAR (n = 11), the fluid required to stabilize hemodynamics after trauma was higher (66 +/- 5 versus 52 +/- 3 ml/kg/hr, p = 0.014), but there were fewer deaths 3 days after trauma versus LR (0 of 11 versus 4 of 31, p = 0.210), fewer deaths within 5 hours after LPS administration (3 of 11 versus 16 of 27, p = 0.074), and a longer survival time after LPS administration (4.5 +/- 0.3 versus 3.9 +/- 0.2 hr, p = 0.054). Deoxycoformycin had similar salutary effects on survival after LPS administration. LPS increased protein permeability of pulmonary capillaries, increased peak inspiratory pressures on constant tidal volume, increased dead space ventilation, and caused progressive arterial desaturation on 0.65 FiO2 (all p < 0.05). This pulmonary dysfunction was associated with a compensatory increase in cardiac output, decrease in systemic vascular resistance, increase in O2 consumption, and rise in plasma cortisol level (all p < 0.05). All these changes were blunted or eliminated with 10 mg/kg AICAR. Hematocrit and systemic pressures were maintained relatively constant after LPS administration with fluid resuscitation, but less was required with AICAR versus LR (40 +/- 8 versus 83 +/- 14 ml/kg/hr, p = 0.023). AICAR caused a concentration-related reduction in CD18 expression on LPS-stimulated neutrophils in vitro, but there was no effect versus LR on circulating leukocyte counts in vivo and no effect of AICAR on LPS-stimulated production of tumor necrosis factor in vitro or in vivo. CONCLUSIONS 1. AICAR reduced the pulmonary dysfunction associated with posttrauma endotoxemia but had no effect on circulating leukocytes, so its mechanism could be related to an adenosine-mediated improvement in peripheral perfusion or O2 use. 2. AICAR is a generic compound that is safe and apparently efficacious in human beings, so AICAR prophylaxis could be cost-effectively administered to trauma patients.

The impact of hypercarbia on the evolution of brain injury in a porcine model of traumatic brain injury and systemic hemorrhage

Carbon dioxide is perhaps the most potent available modulator of cerebrovascular tone and thus cerebral blood flow (CBF). These experiments evaluate the impact of induced hypercarbia on the matching of blood flow and metabolism in the injured brain. We explore the hypothesis that hypercarbia will restore the relationship of CBF to metabolic demand, resulting in improved outcome following traumatic brain injury (TBI) and hemorrhage. A behavioral outcome score, hemodynamic, metabolic, and pathologic parameters were assessed in anesthetized and ventilated juvenile pigs. Animals were assigned to either normocarbia or hypercarbia and subdivided into TBI (via fluid percussion) with or without hemorrhage. The experimental groups were TBI; TBI + 40% hemorrhage (40%H); TBI + hypercarbia (CO2); and TBI + 40%H + CO2. Hemorrhaged animals were resuscitated with blood and crystalloid. Hypercarbia was induced immediately following TBI using 10% FiCO2. The normocarbic group demonstrated disturbance of the matching of CBF to metabolism evidenced by statistically significant increases in cerebral oxygen and glucose extraction. Hypercarbic animals showed falls in the same parameters, demonstrating improvement in the matching of CBF to metabolic demand. Parenchymal injury was significantly decreased in hypercarbic animals: 3/10 hypercarbic versus 6/8 normocarbic animals showed cerebral contusions at the gray/white interface (p = 0.05). The hypercarbic group had significantly better behavioral outcome scores, 10.5, versus 7.3 for the normocarbic groups (p = 0.005). The decreased incidence of cerebral contusion and improved behavioral outcome scores in our experiments appear to be mediated by better matching of cerebral metabolism and blood flow, suggesting that manipulations modulating the balance of blood flow and metabolism in injured brain may improve outcomes from TBI.

Hemodynamic actions of acute ethanol after resuscitation from traumatic brain injury.

BACKGROUND The purposes of this study were to determine how clinically relevant levels of acute ethanol (EtOH) influence cerebral perfusion pressure (CPP), cerebral venous O saturation (Scvo ), and systemic hemodynamics after fluid resuscitation from traumatic brain injury (TBI); and to test the hypothesis that the actions of EtOH on these variables are mediated by adenosine. METHODS Anesthetized swine were ventilated (Fio = 0.4) and instrumented. In protocol 1, EtOH (3.5 g/kg, n = 11) or its vehicle (n = 17) was administered orally before TBI + 40% hemorrhage. At 90 minutes post-TBI, resuscitation consisted of shed blood + saline. In protocol 2, either saline (n = 15) or an adenosine-regulating agent (5-amino-4-imidazolecarboxamide riboside) in saline (1 mg/kg bolus + 12 mg/kg/h intravenously [i.v.]) (n = 5), was administered i.v. before TBI + 45% hemorrhage. At 90 minutes post-TBI, resuscitation consisted of saline only (three times shed blood volume). In protocol 3, EtOH was administered i.v. (1 g/kg; 20% vol/vol in saline) followed by either an adenosine receptor antagonist (theophylline, 10 mg/kg) or an adenosine uptake inhibitor (dipyridamole, 0.25 mg/kg). RESULTS In protocol 1, with no EtOH, 11 of 17 (65%) survived post-TBI hypotension. Mean arterial blood pressure, cardiac index, and mixed venous oxygen saturation were stable for 1 hour at 40% to 60% below their respective baselines, whereas lactate increased three- to fourfold (all p < 0.05). After fluid resuscitation, most variables rapidly corrected, but intracranial pressure was increased 10 to 15 mm Hg (p < 0.05). With EtOH, 9 of 11 (82%) survived post-TBI hypotension (p = 0.42 vs. no EtOH). After resuscitation from TBI, there were significant effects of EtOH on systemic hemodynamics (mean arterial pressure, cardiac index, mixed venous oxygen saturation), on CPP, on lactate, and on Scvo at normo- and hypercapnia (all p < 0.05). The data from protocol 2 showed that essentially none of these changes were duplicated with an adenosine-regulating agent. In protocol 3, i.v EtOH produced small but significant changes in Scvo, intracranial pressure, and lactate, at normo-, hyper-, and hypocapnia. Dipyridamole and theophylline tended to have opposite, albeit small and not statistically significant, effects on these variables relative to EtOH alone.(2) (2) CONCLUSION Acute EtOH (200-300 mg/dL) did not increase mortality after TBI + secondary hypotension, as long as cardiopulmonary support was provided. With EtOH, CPP was maintained and cerebral blood flow appeared to be adequate, if not excessive, with respect to cerebral metabolic demand, as judged by changes in Scvo at normo-, hyper-, and hypocapnia. These changes were probably not mediated, but might have been modulated, by increases in endogenous adenosine.

Gastric and extragastric actions of the histamine antagonist ranitidine during posttraumatic sepsis

BACKGROUND Histamine H2 antagonists (e.g., ranitidine) are generally thought to specifically reduce gastric acid secretion and are commonly used for stress ulcer prophylaxis in critically ill patients because of their efficacy and safety profile. A few reports suggest that ranitidine might also bind to extragastric sites and/or act as an immunomodulator. The potential effects on posttraumatic sepsis are unknown. METHODS Mongrel pigs (n = 24) were anesthetized with fentanyl, injured by a 10 kg steel bar dropped from a height of 1 m onto the fleshy portion of the posterior thigh, and then 35% of their blood volume was drained through the arterial catheter. All the shed blood plus two times the hemorrhage volume as lactated Ringers solution was infused after a 1-hour shock period. Either vehicle or ranitidine (1.5 mg/kg) was intravenously administered at the time of resuscitation and every 12 hours thereafter in a blinded fashion. After 72 hours a septic challenge was administered (15 micrograms/kg Escherichia coli lipopolysaccharide [LPS] x 30 min). Serial gastroscopy, gastric pH, hemodynamics, leukocyte counts, cortisol, and tumor necrosis factor were recorded for 180 minutes after LPS. RESULTS Immediately before LPS all hemodynamic variables were identical between treatments, but gastric pH was slightly higher and stress gastritis was marginally lower with ranitidine. LPS caused profound leukopenia and a hyperdynamic circulatory response (i.e., tachycardia, increased cardiac output, and decreased peripheral vascular resistance at relatively constant blood pressure); these changes were not altered by ranitidine. Gastric pH remained elevated after LPS with ranitidine, but LPS-induced gastritis was not modified. Ranitidine delayed the LPS-induced ventilation-perfusion imbalance and attenuated the peak increase in the proinflammatory cytokine, tumor necrosis factor, without altering its antiinflammatory opponent, cortisol. Similar changes were observed in four additional animals treated with cimetidine. The proportion of circulating neutrophils and lymphocytes was slightly altered 180 minutes after LPS, but there was no obvious effect on T lymphocytes in vivo, and no effect on the LPS-induced increase in neutrophil CD18 expression in vitro was seen. CONCLUSIONS (1) Ranitidine increased gastric pH, which blunted the stress gastritis caused by trauma but not that caused by LPS; (2) ranitidine delayed the early LPS-evoked pulmonary changes and reduced the tumor necrosis factor spike, which is consistent with a favorable immunomodulatory action that has been reported in patients who are critically ill or are undergoing an elective abdominal surgical procedure; (3) the mechanism is probably related to H2 receptor antagonism rather than to a nonspecific side effect of ranitidine, which suggests that histamine may have a previously unrecognized role in posttraumatic septic responses; and (4) the site of action is probably not in the heart or peripheral resistance vessels, but salutary effects on circulating lymphocytes or neutrophils cannot be excluded.

Actions of prostaglandin E1 on lipopolysaccharide-evoked responses in vivo and in vitro following resuscitated trauma.

Prostaglandins of the E series (PGE1, PGE2) have well-described immunosuppressive (anti-inflammatory) as well as vasodilator (pro-inflammatory) actions. The net effect on an acute inflammatory response would depend on the dose, timing, and site of action. Egg phosphatidyl liposomes are novel drug delivery vehicles that can alter the in vivo disposition of PGE1. The purpose of this study was to explore the therapeutic potential of PGE1, with or without liposome encapsulation, on the systemic inflammatory response evoked by endotoxin following trauma. Anesthetized pigs received a soft tissue injury + hemorrhage, and fluid resuscitation after 1 h. In one series, whole blood was incubated with PGE1 (0, 40, or 200 μg/mL) and Escherichia coli endotoxin (LPS; 0, 1, 5, or 10 μg/mL) in vitro and neutrophil CD18 adherence receptor density was measured with immunomonitoring. In another series, LPS (5 μg/kg) was administered 3 days following trauma to animals pretreated with either phosphate-buffered saline (PBS) + PGE1, (62 ng/kg/min x 40 min, 2.5 μg/kg total, n = 8), PBS (n = 12), liposomes alone (Lipo, n = 10) or liposome-encapsulated PGE1 (Lipo + PGE, n = 7). This PGE1 dose had minimal effects on blood pressure in baseline conditions. Hemodynamics, cell differential counts, plasma cortisol, and plasma tumor necrosis factor (TNF) were measured for 3 h post-LPS. LPS in vitro caused a dose-related increase in neutrophil CD18 expression that was not altered by < 200 μg/mL PGE1 before or after trauma. LPS in vivo increased pulmonary vascular resistance and heart rate and both were blunted by PGE1: .73 ± .14 vs. .40 ± .06 mmHg/mL/min/kg, PBS vs. PBS + PGE1, ρ = .0167 and 128 ± 7 vs. 93 ± 9 beats/min, PBS vs. PBS+PGE1, ρ = .0020, respectively. In addition, stroke index (and therefore cardiac efficiency) was improved with PGE1 + PBS vs. PBS (ρ = .0024). These cardiovascular effects were eliminated when PGE1 was liposome encapsulated. Plasma TNF was increased to 300–600 pg/mL following LPS and there was no effect of PGE1 or liposomes, but the LPS increased plasma cortisol to 7.8 ± .8 vs. 4.0 ± 1.0 μg/100 mL for PBS vs. PBS + PGE1 (ρ = .0732) and 8.5 ± 2.1 vs. 2.9 ± 1.1 μg/100 mL for Lipo vs. Lipo + PGE1 (ρ = .0131). Conclusions are as follows: 1) PGE1 reduced the LPS-evoked cortisol increase and improved cardiac function, but had no detectable effect on the evoked TNF spike or neutropenia; 2) Liposome encapsulation eliminated the cardiac, but not the cortisol-lowering, effect; 3) the relative lack of PGE1 on LPS-evoked acute inflammation could reflect a desensitization to its immunosuppressive actions. Alternatively, the results are consistent with the interpretation that PGE1 is not a primary regulator for acute inflammation, but is rather one of a myriad of pro- and anti-inflammatory factors that balance the process.

The Impact of Hypercarbia on the Evolution of Brain Injury in the Setting of Traumatic Brain Injury and Systemic Hemorrhage

The Impact of Hypercarbia on the Evolution of Brain Injury in the Setting of Traumatic Brain Injury and Systemic Hemorrhage