Tissue factor in experimental acute lung injury.

Acute lung injury (ALI) is characterized by fibrin deposition in the tissue and vascular spaces. Coagulation is activated after exposure to endotoxin or bacteria, and a procoagulant environment rapidly develops in the vascular, interstitial, and alveolar spaces of the lung. These changes are tissue factor (TF)-dependent and associated with increases in inflammatory cytokines. Procoagulant changes also occur in the lungs of patients with the acute respiratory distress syndrome (ARDS), suggesting that epithelial inflammation activates the extrinsic pathway. Many inflammatory mediators have specific effects on coagulation; however, the role of TF in regulation of pulmonary inflammatory responses is less clear. Here we report initial data on blockade of TF-initiated coagulation in baboons with Escherichia coli sepsis-induced ALI, using active site-inactivated FVIIa (FVIIai ASIS). Treatment with FVIIai prevented plasma fibrinogen depletion and attenuated fibrin deposition in the tissues. The drug also decreased systemic cytokine responses and inflammatory changes in the lung, including neutrophil infiltration, and decreased edema. Coagulation blockade with FVIIai improved lung function by preserving gas exchange and compliance, decreased pulmonary hypertension, and enhanced renal function. These results show that TF-FVIIa complex is an important regulatory site for the pathologic response of the lung to sepsis.

Coagulation blockade prevents sepsis-induced respiratory and renal failure in baboons.

Sepsis-induced tissue factor (TF) expression activates coagulation in the lung and leads to a procoagulant environment, which results in fibrin deposition and potentiates inflammation. We hypothesized that preventing initiation of coagulation at TF-Factor VIIa (FVIIa) complex would block fibrin deposition and control inflammation in sepsis, thereby limiting acute lung injury (ALI) and other organ damage in baboons. A model of ALI was used in which adult baboons were primed with killed Escherichia coli (1 x 10(9) CFU/kg), and bacteremic sepsis was induced 12 h later by infusion of live E. coli at 1 x 10(10) CFU/kg. Animals in the treatment group were given a competitive inhibitor of TF, site-inactivated FVIIa (FVIIai), intravenously at the time of the infusion of live bacteria and monitored physiologically for another 36 h. FVIIai dramatically protected gas exchange and lung compliance, prevented lung edema and pulmonary hypertension, and preserved renal function relative to vehicle (all p < 0.05). Treatment attenuated sepsis-induced fibrinogen depletion (p < 0.01) and decreased systemic proinflammatory cytokine responses, for example, interleukin 6 (p < 0.01). The protective effects of TF blockade in sepsis-induced ALI were confirmed by using tissue factor pathway inhibitor. The results show that TF-FVIIa complex contributes to organ injury in septic primates in part through selective stimulation of proinflammatory cytokine release and fibrin deposition.

Anti-neutrophil chemokine preserves alveolar development in hyperoxia-exposed newborn rats.

Inflammation may contribute to lung injury and impaired alveolar development in bronchopulmonary dysplasia. We treated hyperoxia-exposed newborn rats with antibodies to the neutrophil chemokine cytokine-induced neutrophil chemoattractant-1 (CINC-1) during 95% O2 exposure to reduce adverse effects of hyperoxia-induced inflammation on lung development. Rats were exposed at birth to air, 95% O2, or 95% O2 + anti-CINC-1 (injected on days 3 and 4). Bromodeoxyuridine (BrdU) was injected 6 h before death. Anti-CINC-1 treatment improved weight gain but not survival at day 8. Anti-CINC-1 reduced bronchoalveolar lavage neutrophils at day 8 to levels equal to air controls. Total detectable lung CINC-1 was reduced to air control levels. Lung compliance was improved by anti-CINC-1, achieving air control levels in the 10-microg anti-CINC-1 group. Anti-CINC-1 preserved proliferating cell nuclear antigen expression in airway epithelium despite 95% O2 exposure. BrdU incorporation was depressed by hyperoxia but preserved by anti-CINC-1 to levels similar to air control. Alveolar volume and surface density were decreased by hyperoxia but preserved by anti-CINC-1 to levels equal to air control. Blockade of neutrophil influx in newborns may avert early lung injury and avoid alveolar developmental arrest that contributes to bronchopulmonary dysplasia.

Antibody to intercellular adhesion molecule 1 (CD54) decreases survival and not lung injury in baboons with sepsis.

Neutrophil influx into the lung is an important event in the pathogenesis of acute lung injury in gram-negative sepsis. We hypothesized that administration of a monoclonal antibody to intercellular adhesion molecule 1 (ICAM-1, CD54), a molecule mediating neutrophil adhesion to endothelial cells, would decrease neutrophil sequestration and transmigration in the lung and attenuate lung injury in Escherichia coli sepsis. Sepsis was induced in 12 baboons primed with heat-killed E. coli (1 x 10(9) CFU/kg) 12 h before infusion of live bacteria (1 x 10(10) CFU/kg). Six animals received monoclonal antibody to CD54 (1 mg/kg) intravenously at the time of live E. coli infusion. After 48 h or when blood pressure could not be maintained, tissues were harvested and bronchoalveolar lavage (BAL) samples were obtained. Median survival time was decreased in anti-CD54-treated animals. This group also had decreased mean arterial pressure, increased metabolic acidosis, and decreased urine output. Measures of lung injury including gas exchange, lung lavage protein and lactate dehydrogenase (LDH), lung thiobarbituric acid-reactive species, and lung histology, including alveolar neutrophil volumes, were unaffected by treatment. The effect of anti-CD54 on neutrophil influx into tissues as measured by myeloperoxidase was organ specific. These data show that monoclonal antibody to CD54 does not ameliorate acute lung injury in E. coli sepsis, and septic primates given anti-CD54 have worsened metabolic parameters and decreased survival.

Proinflammatory cytokines increase in sepsis after anti-adhesion molecule therapy.

Cytokine mediators and leukocyte-endothelial cell adhesion molecules are critical and interdependent components of the acute inflammatory response in sepsis. We hypothesized that the administration of monoclonal antibodies to intercellular adhesion molecule-1 (CD54) or E- and L-selectin (CD62E/L) would decrease serum levels of the proinflammatory cytokines interleukin-1beta (IL-1), IL-6, and IL-8 and tumor necrosis factor receptor (TNFR-1) in baboons during sepsis. Adult male baboons received infusions of 1 x 10(9) colony forming units (CFU)/kg heat-killed Escherichia coli (E. coli) followed 12 h later by live E. coli (1 x 10(10) CFU/kg). At the time of live bacterial infusion, six septic animals were treated with a monoclonal antibody to CD54 and six with an antibody to CD62E and L (1 mg/kg). Eight untreated septic animals served as controls. Sequentially drawn serum samples were assayed for IL-1, IL-6, IL-8, and TNFR-1 using enzyme-linked immunoassay (ELISA). Data were compared using Mann-Whitney U tests and Chi-square analyses. Median survival was decreased in both treatment groups compared to controls (P < 0.05). Peak IL-1 level was higher than controls in septic animals treated with anti-CD54 but not anti-CD62E/L (P < 0.05, P = NS, respectively). Elevations in IL-6, IL-8, and TNFR-1 were increased and prolonged in both antibody treated groups compared to controls (P < 0.05). These results provide the first in vivo evidence that leukocyte-endothelial adhesion molecules CD54 and CD62E/L regulate cytokine production in sepsis.

Bacterial priming increases lung injury in gram-negative sepsis.

Sepsis syndrome is a leading cause of acute respiratory distress syndrome (ARDS), but the development of acute lung injury is highly variable for reasons that are poorly understood. We hypothesized that nonlethal systemic exposure to gram-negative bacteria, with its consequent activation of inflammatory processes, would increase functional and structural lung injury on a second exposure to live organisms, as compared with exposure of naive animals. Sixteen adult baboons received 1 to 2 x 10(10) colony-forming-units (cfu)/kg Escherichia coli by intravenous infusion. Eight animals received live bacteria as a single infusion, whereas the other eight received 10% of the total dose as heat-killed organisms (priming dose) 12 h before the live infusion. Pulmonary gas exchange and hemodynamics were monitored for 48 h or until blood pressure could not be maintained. The animals were killed and one lung was processed for electron microscopy and morphometry. Group data were compared through analysis of variance (ANOVA). The systemic circulatory responses to the bacterial challenge were similar, although less severe shock occurred in primed animals. In contrast, primed animals had increased structural damage involving lung epithelium and endothelium, and showed increased cellularity of the interstitium. The morphologic evidence of increased lung injury in septic animals with prior exposure to heat-killed bacteria suggests that prior activation of systemic inflammatory responses is a contributing factor in the pathogenesis of ARDS.

Antibody to E- and L-selectin does not prevent lung injury or mortality in septic baboons.

Recruitment of polymorphonuclear leukocytes (PMN) through upregulation of cellular adhesion molecules is a proposed mechanism of injury in sepsis and acute respiratory distress syndrome (ARDS). We hypothesized that pretreatment of baboons with a monoclonal antibody to human E- and L-selectin (EL-246) during sepsis would decrease PMN influx into tissues and result in less organ injury during gram-negative sepsis. We studied 14 anesthetized, ventilated adult baboons; six animals received 1 mg/kg of EL-246 before infusion of an LD100 of live Escherichia coli and six received the E. coli infusion without antibody therapy. Two other animals received 1 mg/kg of EL-246 intravenously without an infusion of bacteria. Intermittent measurements were made of circulatory pressures, cardiac output, urine output, arterial blood gases, ventilation:perfusion ratio (VA/Q), and hematologic status. The experiments were ended at 48 h or at the time of death. Tissues were harvested for pathology and biochemical measurements. The E. coli infusions were associated with a hyperdynamic state, pulmonary hypertension, systemic hypotension, decreased urine output (UOP), and metabolic acidosis. The antibody partly blocked PMN migration, but there were few significant physiologic or biochemical differences between the EL-246-treated and untreated animals. In the antibody-treated animals, UOP was decreased, metabolic acidosis was worsened, and median survival time was decreased significantly. We conclude that treatment with an antibody to E- and L-selectin in gram-negative sepsis does not improve gas exchange or protect against lung injury, and is associated with decreased survival time in primates.

Aerosolized manganese SOD decreases hyperoxic pulmonary injury in primates. I. Physiology and biochemistry.

Prolonged hyperoxia causes lung injury and respiratory failure secondary to oxidative tissue damage mediated, in part, by the superoxide anion. We hypothesized that aerosol treatment with recombinant human manganese superoxide dismutase (rhMnSOD) would attenuate hyperoxic lung damage in primates. Adult baboons were anesthetized and ventilated with 100% oxygen for 96 h or until death. Six animals were treated with aerosolized rhMnSOD (3 mg . kg-1 . day-1 in divided doses), and six control animals did not receive enzyme therapy. Physiological variables were recorded every 12 h, and ventilation-perfusion ratio relationships were evaluated by using the multiple inert-gas elimination technique. After the experiments, surfactant composition and lung edema were measured. We found that rhMnSOD significantly decreased pulmonary shunt fraction (P < 0.01) and preserved arterial oxygenation (P < 0.01) during hyperoxia. The rhMnSOD increased lung phospholipids, phosphatidylcholine and disaturated phosphatidylcholine, and decreased lung edema in this model. Testing of higher and lower doses of MnSOD (1 and 10 mg . kg-1 . day-1) in two other groups of baboons produced variable physiological protection, suggesting a "window" of effective dosage. We conclude that aerosolized MnSOD (3 mg . kg-1 . day-1) affords significant preservation of pulmonary gas exchange during hyperoxic lung injury.

Aerosolized manganese SOD decreases hyperoxic pulmonary injury in primates. II. Morphometric analysis.

Hyperoxia damages lung parenchyma via increased cellular production of reactive oxygen species that exceeds antioxidant defenses. We hypothesized that aerosolized human recombinant manganese superoxide dismutase (rhMnSOD) would augment extracellular antioxidant defenses and attenuate epithelial injury in the lung during hyperoxia in primates. Twenty-four adult male baboons were anesthetized and mechanically ventilated with 100% oxygen for 96 h. The baboons were divided equally into four groups. Oxygen alone and oxygen plus rhMnSOD given at 3 mg . kg-1 . day-1 were compared to assess efficacy of the drug. Subsequently, aerosolized rhMnSOD was given at 1 or 10 mg . kg-1 . day-1 to study dose effects and toxicity. Quantitative morphometry showed protection of alveolar epithelium from hyperoxia by 3 mg . kg-1 . day-1 rhMnSOD (P < 0.05). In addition, interstitial fibroblast volumes were increased in the treatment group (P = 0.06). This effect appeared greater at the two higher doses of the rhMnSOD. The aerosolized drug was localized to the surface of airways and air spaces and macrophages by immunolabeling studies, suggesting efficacy via physicochemical properties that localize it to cell surfaces or by effects on alveolar macrophage function.

Changes in the lung after prolonged positive pressure ventilation in normal baboons.

PURPOSE: The effects of prolonged positive pressure ventilation on lung ultrastructure are not well defined in primates. This study was designed to measure cardiopulmonary and morphological responses to 4 days of positive pressure ventilation in normal baboons. MATERIALS AND METHODS: Six adult male baboons were mechanically ventilated on air for 96 hours with 2.5 cm positive end-expiratory ventilation and a tidal volume of 12 to 15 mL/kg. Physiological measurements were obtained every 12 hours and serial measurements of ventilation-perfusion (VA/Q) were performed using the multiple inert gas elimination technique. Quantitative morphotometry, lung dry-to-wet ratio, and surfactant analysis were performed at the end of the experiment. RESULTS: Cardiovascular variables, except for a small increase in mean pulmonary artery pressure at 84 and 96 hours, were not significantly affected by positive pressure ventilation. Arterial Po2 decreased, and shunt fraction increased from 0.7% of cardiac output to 5.4% (P < .01). Dispersion of perfusion increased threefold (P < .01), and dispersion of ventilation doubled (P < .01) indicating increased VA/Q mismatch mismatch. Respiratory system compliance decreased by 30% (P < .01). There was no lung edema or change in surfactant composition. Lung morphometry showed increases in polymorphonuclear cells and type II cell volume. Vacuolated endothelial cells and bare basement membrane were observed consistently. CONCLUSION: Four days of positive pressure ventilation decreases lung compliance and worsens gas exchange by increasing shunt and VA/Q mismatch in healthy baboons. These effects are accompanied by only minor ultrastructural changes and mild inflammatory responses in the lung.

VA/Q abnormalities during gram negative sepsis.

Hypoxemia in bacterial sepsis develops by mechanisms which are incompletely understood. In this study, we measured pulmonary gas exchange in eight baboons to determine the causes of hypoxemia after infusion of live Escherichia coli (1 x 10(10) CFU/kg) followed by resuscitation with intravenous fluid. VA/Q distributions were measured periodically using the multiple inert gas elimination technique until death or for a maximum of 42 h. After E. coli infusion, dispersion of perfusion (logSDq) increased rapidly and a transient rise in dead space was observed at 6 h coinciding with systemic hypotension and acidosis. The intrapulmonary shunt developed later and reached 27 +/- 6% at 24 h. PaO2 began to decrease at 12 h and correlated with increases in intrapulmonary shunt and logSDq. There was no evidence of diffusion limitation. Lung edema was mild despite aggressive fluid resuscitation. Morphometric analysis of postmortem lungs revealed dramatic intravascular accumulation of granulocytes. There were increases in arithmetic mean thicknesses of epithelium and interstitium. These data indicate that gram negative sepsis with fluid resuscitation causes progressive hypoxemia, primarily due to the development of intrapulmonary shunt and very low VA/Q regions in the lung. The VA/Q abnormalities occur early and likely reflect ongoing cellular responses in pulmonary vasculature and smaller airways in sepsis.

Fungal infections in lung and heart-lung transplant recipients. Report of 9 cases and review of the literature.

We reviewed the pattern and incidence of fungal infections in patients undergoing lung and heart-lung transplantation at Duke University Medical Center from September 1992 until August 1995, and present here 9 illustrative cases. Of the 73 lung and heart-lung transplant recipients studied, 59 (81%) had positive fungal cultures at some point after transplantation. The cases presented here illustrate that lung transplant recipients are predisposed to a wide variety of fungal infections. The clinical pattern of these infections ranges from asymptomatic to rapidly progressive fatal disease. In addition to the reactivation of previous fungal infections and recent exposure to new environmental sources, the donor lung itself can be the source of fungal infection, as we showed by using molecular epidemiology techniques. Because of the associated morbidity and mortality, efforts should be directed at investigating prophylactic antifungal regimens in lung transplant recipients. Preliminary reports on the use of itraconazole and aerosolized amphotericin B have been encouraging. Prospective randomized studies are needed to assess the safety and cost effectiveness of different regimens. Fungal infections in patients after lung transplantation can significantly impede recovery and lead to substantial mortality.

Ultrastructural changes in skeletal muscle mitochondria in gram-negative sepsis.

Energy metabolism during sepsis is incompletely understood, but alterations in mitochondrial structure and function appear important. We measured time-dependent changes in mitochondrial structure during sepsis using serial skeletal muscle biopsies in anesthetized baboons injected with 10(10) CFU/kg of live Escherichia coli (LD(100)). Skeletal muscle biopsies were taken before bacterial challenge (0 h controls) and at 12 h, 24 h, and death. By qualitative electron microscopy, the organelles became enlarged with distorted cristae and developed electron lucent areas within the matrix. With advanced injury the inner membrane became fragmented. Quantitative morphometric analysis showed a 50% increase in mean cristal membrane surface density by 24 h (p < .05) accompanied by a 100% increase in intermembrane space (p < .01). Matrix volume density decreased progressively (p < .01). These changes in mitochondrial ultrastructure occur within 12 h after the onset of the bacterial insult. This damage, including destruction or reorganization of both membrane and matrix proteins, is severe enough to compromise oxidative metabolism in muscle in Gram-negative sepsis.

Artificial surfactant attenuates hyperoxic lung injury in primates. II. Morphometric analysis.

Diffuse lung injury from hyperoxia is accompanied by low compliance and hypoxemia with disruption of endothelial and alveolar epithelial cell layers. Because both function and content of surfactant in diffuse lung injury decrease in animals and in humans, changes in the extent of injury during continuous hyperoxia were evaluated after treatments with a protein-free surfactant in primates. Ten baboons were ventilated with 100% O2 for 96 h and five were intermittently given an aerosol of an artificial surfactant (Exosurf). Physiological and biochemical measurements of the effects of the surfactant treatment are presented in a companion paper (Y.-C. T. Huang, A. C. Sane, S. G. Simonson, T. A. Fawcett, R. E. Moon, P. J. Fracica, M. G. Menache, C. A. Piantadosi, and S. L. Young. J. Appl. Physiol. 78: 1823-1829, 1995.) After O2 exposures, lungs were fixed and processed for electron microscopy. The cellular responses to O2 included epithelial and endothelial cell injuries, interstitial edema, and inflammation. Morphometry was used to quantitate changes in lungs of animals treated with the artificial surfactant during O2 exposure and to compare them with the untreated animals. The surfactant decreased neutrophil accumulation, increased fibroblast proliferation, and decreased changes in the volume of type I epithelial cells. Surfactant-treated animals also demonstrated better preservation of endothelial cell integrity. These responses indicate ameliorating effects of the surfactant on the pulmonary response to hyperoxia, including protection against epithelial and endothelial cell destruction. Significant interstitial inflammation and fibroblast proliferation remained, however, in surfactant-treated lungs exposed to continuous hyperoxia.

Altered mitochondrial redox responses in gram negative septic shock in primates.

Gram negative sepsis causes changes in oxygen supply-demand relationships. We have used a primate model of hyperdynamic gram negative sepsis produced by intravenous infusion of Escherichia coli (E. coli) to evaluate sepsis-induced alterations in mitochondrial oxidation-reduction (redox) state in muscle in vivo. The redox state of cytochrome a,a3, the terminal member of the intramitochondrial respiratory chain, was assessed in the intact forearm by near-infrared (NIR) spectroscopy. The muscle NIR data were compared to routine measures of oxygen delivery (DO2) and oxygen consumption (VO2). After E. coli infusion and fluid resuscitation, DO2 and VO2 showed minimal changes through 24 hr of sepsis. In contrast, changes in cytochrome a,a3 redox state evaluated by NIR occurred within a few hours and were progressive. Mitochondrial functional responses were correlated with structural changes observed on serial muscle biopsies. Gross morphological changes in muscle mitochondria were present in some animals as early as 12 hr, and, in most animals, by 24 hr. The morphologic changes were consistent with decreases in oxidative capacity as suggested by NIR spectroscopy. The NIR data also suggest that two mechanisms are operating to explain abnormalities in oxygen metabolism and mitochondrial function in lethal sepsis. These mechanisms include an early defect in oxygen provision to mitochondria that is followed by a progressive loss in functional cytochrome a,a3 in the muscle.

Pseudomonas putrefaciens bacteremia.

Pseudomonas putrefaciens is an unusual cause of human disease. Since 1978 only five cases of bacteremia due to this organism have been reported. Within 12 recent months four cases of bacteremia due to P. putrefaciens were seen - two occurred in patients with chronic infections of a lower extremity, one in a patient with neutropenia, and one in a patient with fulminant septicemia and disseminated intravascular coagulation. Two of the patients had prostheses; in neither case did prosthetic infection or prosthetic failure occur. Two syndromes of bacteremic infection with P. putrefaciens are suggested. One syndrome is associated with chronic infection of a lower extremity, is fairly well tolerated, and responds to appropriate antimicrobial agents. The other syndrome is more fulminant and may be associated with severe underlying debility, liver disease, and malignancy. It is not yet known whether this is a meaningful distinction. The significance of the recent increase is the isolation of this organism is not clear at present.