Original Article
Depressed Bronchoalveolar Urokinase Activity in Patients with Adult Respiratory Distress Syndrome
N Engl J Med 1990; 322:890-897March 29, 1990
- Abstract
- Abstract
Abundant deposition of bronchoalveolar fibrin and fibronectin occurs during the exudative phase of the adult respiratory distress syndrome (ARDS), promoting hyaline-membrane formation and subsequent alveolar fibrosis. To explore the mechanisms that account for the persistence of bronchoalveolar fibrin and fibronectin, we compared the activity of urokinase, which is necessary for plasminogen activation and fibrin degradation, in cell-free bronchoalveolar-lavage fluid from 8 patients with ARDS, 9 patients with acute pulmonary diseases other than ARDS, and 10 normal subjects.
The mean level of urokinase activity in the lavage fluid from the patients with ARDS was 0.003 IU per milliliter of fluid (range, 0 to 0.008), which was significantly lower (P = 0.001) than the level in the fluid from either the patients with pulmonary diseases other than ARDS (0.118 IU per milliliter [range, 0.032 to 0.295]) or the normal subjects (0.129 IU per milliliter [range, 0.045 to 0.198]). The lavage fluid from all the patients with ARDS also had antiplasmin activity, which would promote the persistence of fibrin. A true decrease in urokinase activity was confirmed by the failure of the lavage fluid from the patients with ARDS to convert [125l]plasminogen to plasmin. Despite the low urokinase activity, immunochemical assays revealed normal levels of urokinase antigen in the fluid from the patients with ARDS, suggesting the presence of urokinase inhibitors. Inhibitors were demonstrated directly by a fibrin gelunderlay assay that detects complexes of urokinase with inhibitors. Plasminogen-activator inhibitor type 1 was the principal inhibitor identified.
We conclude that increased antifibrinolytic activity due to both urokinase inhibitors and antiplasmins in the bronchoalveolar compartment of patients with ARDS contributes to the formation and persistence of hyaline membranes, a key component of alveolar histopathology in ARDS. (N Engl J Med 1990; 322:890–7.)
Article Activity
- Article
PATHOLOGICAL studies of lung tissue in the initial phases of the adult respiratory distress syndrome (ARDS) reveal diffuse damage to both the alveolar epithelial cells and the capillary endothelium, accompanied by a hemorrhagic intraalveolar exudate rich in platelets, fibrin clotting factors, and plasminogen.1 , 2 Several investigators have noted prominent deposition of fibrin and fibronectin along the alveolar ducts with the incorporation of cellular debris, forming hyaline membranes.1 2 3 4 Given the presence of plasminogen and the normal abundance of functional bronchoalveolar urokinase-type plasminogen activator, one would expect to see rapid fibrinolysis and remodeling of the hyaline membranes, followed by reepithelialization of the alveolar basement membrane. Instead, the alveolar healing process is disordered, partly as a result of the persistence of hyaline membranes rich in fibrin and fibronectin and intraalveolar "fibrin buds."5 Sequential pathological observations by Pratt and Basset et al. indicate that the sites of early fibrin deposition correlate with the subsequent location of the fibrotic process.3 , 4 In vitro data support the notion that fibrin and fibronectin are important substrates for the adherence and proliferation of fibroblasts and epithelial cells.6 , 7 In ARDS, the growth of interstitial fibroblasts and histiocytes into these deposits is followed by the deposition of reticulum and collagen. This process, accompanied by alveolar reepithelialization by cuboidal Type II cells, seems in large part to account for the thickened alveolar septums and dense interstitial fibrosis that are an important pulmonary cause of mortality in this disorder.8
In previous studies of the deposition and resorption of fibrin in patients with chronic inflammatory diseases of the lung, we demonstrated that the bronchoalveolar-lavage fluid recovered from patients with idiopathic pulmonary fibrosis or sarcoidosis had diminished urokinase activity relative to that obtained from normal subjects.9 Because similar pathological events leading to fibrosis have been observed in these more chronic inflammatory disorders, we hypothesized that perturbation in alveolar fibrinolytic activity might be present in ARDS. Therefore, we compared the urokinase activity in cell-free bronchoalveolar-lavage fluid from patients with ARDS with that in fluid from patients with other acute pulmonary diseases and from normal subjects.
Methods
Study Population
Eight patients with ARDS, 9 patients with other acute pulmonary diseases, and 10 normal subjects underwent bronchoalveolar lavage after giving informed consent, as described elsewhere.9 Six of the patients with ARDS were studied at Massachusetts General Hospital. Six of the patients with other acute pulmonary diseases were studied at Henri Mondor Hospital. All the other patients were studied at Brigham and Women's Hospital. The bronchoalveolar lavage was performed according to the same protocol at each hospital. The protocol was approved by an institutional review board at each of the U.S. hospitals and by the ethics committee of the Société de Réanimation de Langue Française.
The patients with ARDS all had hypoxia requiring mechanical ventilation, decreased respiratory-system compliance, and acute diffuse pulmonary infiltrates, but not elevated pulmonary-artery wedge pressure. All these patients required mechanical ventilation with positive end-expiratory pressure (PEEP). The patients with other pulmonary diseases had different acute illnesses, all of which were associated with the presence of acute pulmonary infiltrates and hypoxemia requiring supplemental oxygen therapy. No attempt was made to select these patients on the basis of severity of illness. Rather, the members of this group were selected because they had an acute pulmonary process other than ARDS for which diagnostic fiberoptic bronchoscopy was indicated. For the most part, this group consisted of patients with acute pneumonia, either localized bacterial pneumonia or diffuse involvement with Pneumocystis carinii. Three of these patients (two with bacterial pneumonia and one with cardiogenic pulmonary edema) required mechanical ventilation with PEEP.
The contributing diagnoses in the patients with ARDS and the principal diagnoses in the patients with other pulmonary diseases are listed in Table 1Table 1
Clinical Profiles of the Patients with ARDS and the Patients with Other Acute Pulmonary Diseases, with Differential Cell Counts of Their Bronchoalveolar-Lavage Specimens.*, in addition to demographic data and the intervals between initial intubation and bronchoalveolar lavage. In the two patient groups, antibiotics and a variety of other medications such an antacids and antihypertensive agents were administered as indicated clinically. No single medication other than oxygen was administered to a majority of patients in either group. The mean age of the normal subjects was 31 years (range, 22 to 40). Six were men, and four were cigarette smokers. None of these subjects were taking any medications.With the bronchoscope in a wedged position, all patients and the normal subjects underwent lavage with 30-ml aliquots of normal saline until approximately 100 ml of total fluid was recovered.9 In the patients with radiographic evidence of diffuse involvement and in the normal subjects, either the lingula or the right middle lobe was lavaged. In the patients without ARDS but with localized pneumonia, the involved site was lavaged. The amount of lavage fluid recovered varied between 40 and 60 percent of that administered. The pooled specimens of lavage fluid were centrifuged at 1500×g and the cell-free lavage stored at − 70°C (or transported frozen) before testing. The differential cell counts as determined from Giemsa-stained, centrifuged specimens of the recovered lavage fluid from each patient are shown in Table 1. The lavage fluid obtained from normal subjects was studied by hemocytometry to ensure that the cell counts and percentages of macrophages were within the normal range established for this laboratory.9 , 10
Fibrin Plate Assays
Fibrinolytic activity or the inhibition of fibrinolysis was first assayed in [125I]fibrin-coated microtiter wells, as described elsewhere.10 The end point in this assay is the release of fragments of [125I]plasmin into the incubation mixture. In samples containing no appreciable urokinase activity, the inhibition of plasmin activity was assessed with a modified assay in which, after preincubation at 37°C for 30 minutes, the lavage fluid and 0.02 μg of plasmin (Kabi, Helena Labs, Houston) were added to the wells coated with [125I]fibrin, and the percentage of fibrinolysis after three hours was compared with that in control wells containing buffer and an identical amount of plasmin. The lavage samples from the patients with ARDS were tested empirically over a range of dilutions until the inhibition of plasmin activity was proportional to the volume of lavage fluid added. The total protein concentrations in the lavage fluid were determined by the method of Lowry et al. with use of bovine serum albumin as a standard.11 The results were expressed as micrograms of plasmin inhibited per microgram of total protein in the lavage fluid.
[125I]Plasminogen—Conversion Assay
The [125I] plasminogen—conversion assay of Loskutoff and Edgington was used to assess directly the activity of urokinase in lavage fluid.12 In this assay, the cleavage of plasminogen by urokinase or tissue plasminogen activator to form plasmin can be assessed independently of antiplasmins that may be present. In brief, centrifuge tubes containing 2 to 5 μl of [125I]plasminogen and 25 to 50 μl of lavage fluid (or buffer containing a known amount of urokinase) were incubated at 37°C for four hours. The mixtures then underwent electrophoresis in 10 percent acrylamide slab gels under reducing conditions with a Laemmli buffer system and were analyzed by photofluorography.13 In preliminary experiments using a range of concentrations of purified urokinase, we found that as little as 0.0015 IU of urokinase activity produced clearly visible 25-kd plasmin light chains.
Quantification of Urokinase Antigen
To determine whether the amount of active urokinase in a sample of lavage fluid correlated with the urokinase antigen concentration, we used a modification of the double-sandwich enzyme-linked immunosorbent assay (ELISA) described by Astedt and colleagues.14 The monoclonal and polyclonal urokinase antibodies used in this assay have been characterized elsewhere.14 , 15 The specific rate of absorbance was determined by measuring the total absorbance in the experimental wells and subtracting the absorbance in the control wells containing antibody but no sample (nonspecific background) and that in the wells containing lavage fluid and polyclonal antibody but no monoclonal antibody (nonspecific absorbance). As in the fibrin plate assay, a standard curve was generated for urokinase that was linear from 0.04 to 1.0 IU per milliliter (R2 >0.90) and from which the concentration of urokinase in the lavage fluid was calculated. In the initial experiments, the effect of the inhibition of urokinase by plasminogen-activator inhibitor type 2 (PAI-2) on the detection of urokinase in the ELISA was assessed. The complexing of purified urokinase with PAI-2 (one hour at 22°C) resulted in a reduction in the absorbance of urokinase in this assay by approximately half.
Fibrin Gel-Underlay Assay
The presence of urokinase inhibitors was assayed by the fibrin gel-underlay method of Granelli-Piperno and Reich.16 One-milliliter samples of lavage fluid from normal subjects or patients with ARDS were concentrated to 50 μl and underwent electrophoresis in 10 percent polyacrylamide slab gels, along with molecular-weight standards, urokinase, and tissue plasminogen activator, and were laid on 1 percent agarose gels containing fibrin and plasminogen. In some experiments, duplicate gels were tested from which the plasminogen was omitted. This technique allowed the characterization of the plasminogen activator as being either the tissue type or the urokinase type on the basis of its relative molecular weight; because of the denaturation of the inhibitors of fibrinolysis during the electrophoresis, it also allowed the expression of latent urokinase activity that had been restrained because the urokinase molecules had formed complexes with urokinase inhibitors. The two gel layers were incubated at 37°C and visually inspected for bands of lysis of fibrin at 4 and 24 hours. The apparent molecular weights were calculated from the positions of the stained protein standards.
The two most rapidly acting inhibitors of plasminogen activators are plasminogen-activator inhibitor type 1 (PAI-1), thought to be the main inhibitor secreted by vascular endothelial cells and contained in platelets, and PAI-2, which has been associated predominantly with placental tissue and monocyte- or macrophage-like cells.17 18 19 20 21 The lavage-fluid samples from the normal subjects and the patients with ARDS were immunoadsorbed with monospecific polyclonal antibodies to PAI-1, PAI-2, or fibronectin. Urokinase immunoadsorbed in the form of enzyme—inhibitor complexes was subsequently detected by the fibrin gel-underlay assay. This assay was used instead of reverse fibrin autography18 to characterize urokinase inhibitors, because reverse fibrin autography is less sensitive in the detection of PAI-2 than in that of PAI-1, whereas complexes of urokinase with either inhibitor can be assayed directly by the fibrin gel technique.17 , 18 Immunoadsorption was performed by first removing the endogenous IgG from the lavage fluid with protein A Sepharose beads (two incubations with 10 mg of protein A Sepharose each), then adding 100 to 150 μg of IgG of the respective antibodies, incubating the mixture at 4°C for two hours, and precipitating it with fresh protein A Sepharose. Bound inhibitor, enzyme—inhibitor complexes, or both were eluted from the Sepharose beads by incubation in gel sample buffer at 37°C for two hours, followed by electrophoresis and analysis in the fibrin gel assay exactly as described above.
The goat polyclonal PAI-2 antibody was prepared and characterized by Lecander and Astedt,20 and the rabbit PAI-1 antibody was purchased from American Diagnostica (Greenwich, Conn.). Rabbit antifibronectin IgG (CalBiochem, La Jolla, Calif.) was used as a control. Initial experiments using purified urokinase (Winkinase, courtesy of the American Red Cross), purified PAI-1 (American Diagnostica), and partially purified PAI-2 (Alpha Therapeutic, Los Angeles) indicated that the PAI-1 and PAI-2 antibodies specifically precipitated urokinase—PAI-1 and urokinase—PAI-2 complexes, respectively.
Statistical Analysis
Continuous linear regression analyses were performed, and comparisons between groups were made with the unpaired, two-tailed t-test.22 The analyses were performed with the Stat-View software package (Brainpower, Calabasas, Calif.).
Results
Urokinase Activity in Lavage Fluid
Cell-free bronchoalveolar-lavage fluid normally contains measurable plasminogen-activator activity that is similar to that of urokinase.9 , 23 With respect to urokinase activity, we compared unconcentrated samples of lavage fluid from patients with ARDS, a heterogeneous group of patients with acute pulmonary diseases other than ARDS, and normal subjects. The results are shown in Table 2Table 2
Total Protein and Urokinase Activities in the Bronchoalveolar-Lavage Specimens Studied.. Mean (±SD) functional urokinase activity as measured in the fibrin plate assay was nearly absent in the patients with ARDS (0.003±0.003 IU per milliliter of fluid). By contrast, urokinase activity was easily measurable in the lavage fluid recovered from the patients with other pulmonary diseases (0.118±0.08 IU) and the normal subjects (0.129±0.06 IU). The difference between the patients with ARDS and the other groups was significant (P = 0.001). The values reported here for urokinase activity in the normal subjects are virtually identical to those reported previously from this laboratory.10 The results are shown in Figure 1Figure 1
Comparison of the Activity of Plasminogen Activator in Bronchoalveolar-Lavage Fluid from Patients with ARDS, Patients with Other Acute Pulmonary Diseases, and Normal Subjects.. No plasminogen-independent fibrinolysis was measurable in either the group with ARDS or the normal group — that is, the rate of fibrinolysis in the absence of added plasminogen in the lavage fluid from either group was not higher than that caused by the addition of buffer alone. Three patients in the non-ARDS group had plasminogen-independent fibrinolytic activity that was approximately half the urokinase activity of the same samples. The nature of this activity was not explored.The reductions in urokinase activity in the fibrin plate assay could be due either to the presence of inhibitors of plasmin activity or to a lack of active urokinase. The major antiplasmins derived from plasma, alpha2-antiplasmin and alpha2-macroglobulin, are irreversibly inactivated under acidic conditions that preserve the activity of urokinase and tissue plasminogen activator.24 However, attempts to elicit latent urokinase activity in the lavage fluid from the patients with ARDS by treatment at pH 3 for two hours failed, suggesting an absolute decrease in functional urokinase activity. To investigate this point further, we assessed the ability of the lavage fluid from the normal subjects and the patients with ARDS to cleave [125I]plasminogen to plasmin directly, a proteolytic step independent of antiplasmin activity. The results (Fig. 2Figure 2
Conversion of [125I]Plasminogen to Plasmin in Bronchoalveolar-Lavage Fluid from Normal Subjects but Not Patients with ARDS.) show that the incubation of lavage fluid from the normal subjects with [125I]plasminogen for three hours at 37°C resulted in the appearance of [125I]plasmin (most easily seen as the 25-kd band in lanes 3 and 4). In contrast, none of the samples from the three patients with ARDS tested in the same assay generated plasmin (lanes 5 to 7), nor did the lavage-fluid samples from the other patients with ARDS. Thus, the results of the plasminogen-conversion assay confirmed the lack of active urokinase activity in the lavage fluid from the patients with ARDS previously demonstrated by the fibrin plate assay (Fig. 1). In further studies, the mixtures of [125I]plasminogen and lavage-fluid samples from three normal subjects and three patients with ARDS underwent electrophoresis, and the 25-kd regions of the gels were sliced and counted directly to measure plasminogen conversion. This study confirmed the results of the fibrin plate assay for both groups (data not shown).To assess whether the absence of measurable urokinase activity could be explained by a lack of enzyme, we measured the amount of urokinase antigen in lavage fluid from seven patients with ARDS and eight normal subjects by ELISA. Concurrently, urokinase antigenic activity was detected in all the lavage-fluid samples tested from both the normal subjects and the patients with ARDS (Fig. 3Figure 3
Comparison of Urokinase-Antigen Levels in Bronchoalveolar-Lavage Fluid from Normal Subjects and Patients with ARDS.). Thus, the near absence of urokinase activity in the lavage fluid from the patients with ARDS was not explained by the absence of enzyme in itself. Subsequently, we measured urokinase antigenic activity in lavage fluid from 8 patients in the non-ARDS group as well as an additional 15 normal subjects. The mean amount of urokinase antigen per milliliter of lavage fluid in the patients in the non-ARDS group was 0.25±0.19 IU per milliliter, whereas the mean in all 23 normal subjects was 0.34±0.32 IU per milliliter. The mean value in the seven patients with ARDS was 0.38±0.18 IU per milliliter. There were no statistically significant differences between any of the groups.Inhibitors of Urokinase and Plasmin in Lavage Fluid
The presence of inactive urokinase in the lavage fluid from the patients with ARDS suggested that this fluid contained inhibitors of urokinase or plasmin activity or both. The higher mean total protein concentration in the lavage samples from the patients with ARDS as compared with those from the normal subjects (1.46±1.17 mg per milliliter and 0.12±0.08 mg per milliliter, respectively; P = 0.002) is consistent with the disruption of the integrity of the capillary and alveolar-lining cells that occurs in ARDS, allowing plasma-derived antiplasmins access to the airways and alveoli. This disruption was confirmed by the fibrin plate assays in which plasmin was added directly to the lavage fluid. Functional antiplasmin activity was demonstrable in the lavage-fluid samples from all eight patients with ARDS; as shown in Figure 4Figure 4
Correlation of Plasmin-Inhibitor Capacity with the Total Protein Concentration of Bronchoalveolar-Lavage Fluid from Patients with ARDS., this activity was directly correlated (P<0.001) with the total protein concentrations of the samples of bronchoalveolar-lavage fluid. The nature of these antiplasmins was not explored in detail but was almost certainly due in part to alpha2-antiplasmin. In a manner consistent with the physical properties of alpha2-antiplasmin, this inhibition was partially removable by previous acid treatment of the lavage samples (data not shown).To examine further the possibility that the bronchoalveolar-lavage fluid from the patients with ARDS contained urokinase inhibitors, we used the fibrin gelunderlay technique. This technique has been used by a number of investigators to detect latent urokinase activity, either by dissociating urokinase from a reversibly bound inhibitor or by denaturing a covalently bound inhibitor to allow the expression of bound urokinase.17 18 19 20 21 Electrophoresis was used on equivalent amounts of lavage fluid from eight normal subjects and seven patients with ARDS. A representative overlay is shown in Panel A of Figure 5Figure 5
Demonstration of Urokinase—Plasminogen-Activator Inhibitor Complexes in Bronchoalveolar-Lavage Fluid from Patients with ARDS by Fibrin Gel-Underlay Assay.. The lavage fluid from the patients with ARDS, which contained from 0 to 0.008 IU of urokinase activity per milliliter of fluid, consistently produced lysis bands in the underlay at 55 kd (lanes 2 to 4). This mobility corresponded to that of purified urokinase (lane 6) and was the only form of urokinase activity (lanes 1 and 5) found in the lavage fluid from the normal subjects. The lysis zones in the lavage fluid from the patients with ARDS generally appeared at the same time or earlier than those in the lavage fluid from the normal subjects containing 0.10 to 0.15 IU of urokinase per milliliter. Furthermore, an additional lysis zone consistently appeared at approximately 83 kd in the lavage fluid from all but one of the patients with ARDS. The appearance of this band depended on the addition of plasminogen to the agarose gels. In a few instances, a faint lysis zone at 100 kd was also apparent. These zones were consistent with the presence of complexes of urokinase with either PAI-1 or PAI-2.10 , 25 Neither the 83-kd nor the 100-kd band was ever found after the electrophoresis of lavage fluid from the normal subjects, suggesting much less formation of urokinase—urokinase inhibitor complexes in these samples. In additional control studies, purified tissue plasminogen activator produced a lysis zone between 55 kd and 83 kd, consistent with its molecular weight of 66. Lysis compatible with the presence of tissue plasminogen activator was not found in the studies of the lavage fluid from either the patients with ARDS or the normal subjects.To explore the urokinase complexes further, lavage fluid from each of five patients with ARDS was adsorbed with antibodies (IgG) to PAI-1, PAI-2, or fibronectin, and the antibodies adsorbed to protein A Sepharose beads. The adsorbed material was then eluted from the beads and tested for the presence of urokinase in the fibrin gel-underlay assay. The results of three such experiments are shown in Panel B of Figure 5. The immunoadsorption eluates contained bands in the 55-kd and 83-kd regions, indicating that the antibodies to PAI-1 had removed the urokinase—inhibitor complexes from the original reaction mixtures. The patterns of activity in these eluates were similar to those found in the direct assays of lavage fluid from the patients with ARDS (Panel A). Little or no urokinase activity was recovered with antibodies to either PAI-2 or fibronectin (Panel B). In addition, no urokinase activity was recovered from the lavage fluid from the normal subjects when it underwent the same immunoadsorption procedure, indicating that these three antibodies do not bind free urokinase (data not shown). Because the PAI-1 antibodies removed urokinase from the lavage fluid from the patients with ARDS, the urokinase probably forms complexes with PAI-1 (both reversibly and nonreversibly) in the lavage fluid in such patients.
Discussion
These results provide new insight into the catabolism of the fibrinous exudate that develops in patients with ARDS. We found that the normal pathway of bronchoalveolar fibrinolysis is markedly depressed in such patients during the early exudative phase of their illness. The basic defect is the presence of excess fibrinolytic inhibitors. The loss of urokinase activity in lavage fluid from the patients with ARDS was virtually complete, and the fluid was in fact antifibrinolytic. This was a dramatic reversal of normal physiology, in that the bronchoalveolar surfaces of normal subjects are replete with active urokinase.9 , 23 The finding of urokinase activity in the lavage fluid from the patients with acute pulmonary diseases other than ARDS (Table 2 , Fig. 1) indicates that intraalveolar exudates, pulmonary edema, or intubation and mechanical ventilation do not in themselves result in a marked loss of urokinase activity in lavage fluid. It is likely that the striking decrease in urokinase activity in patients with ARDS is associated with more severe cellular stimulation, cellular injury, or both. The finding of marked reductions in urokinase activity in lavage fluid may not be specific to ARDS. Such reductions had been reported in patients with active alveolitis associated with idiopathic pulmonary fibrosis and to a lesser extent sarcoidosis.9 Thus, in more chronic inflammatory conditions associated with focal alveolar deposits of fibrin and fibronectin, reduced urokinase activity is also demonstrable. The development of an overt antifibrinolytic state is not prominent in either idiopathic pulmonary fibrosis or sarcoidosis but is consistently present in ARDS (Fig. 4), the disorder characterized by the most marked accumulations of bronchoalveolar fibrin.
To explore the mechanism of reduced bronchoalveolar urokinase activity in patients with ARDS, we measured the levels of urokinase antigen in the bronchoalveolar fluid from both normal subjects and patients with ARDS. The immunochemical assays revealed that the amount of urokinase was not significantly lower in the lavage fluid from the patients with ARDS than in that from the normal subjects (Fig. 3). In fact, because our immunoassay was less sensitive for detecting complexes of urokinase with urokinase inhibitors than unbound urokinase, the actual urokinase levels in the lavage fluid from the patients with ARDS may have been higher than those in fluid from normal subjects. In any case, our results establish that lack of urokinase was not in itself the major mechanism for reduced fibrinolytic activity.
Although both urokinase inhibitors and plasmin inhibitors were demonstrable in lavage fluid from the patients with ARDS, we focused primarily on analyzing the urokinase inhibitors, for the following reason. The amount of functional antiplasmin must depend to some extent on the amount of active urokinase, because the rate of formation of plasmin from plasminogen determines the rate of consumption of the antiplasmins. Our data indicate that there is almost no free urokinase in lavage fluid from patients with ARDS, and thus it is not surprising that we found a close correlation between total bronchoalveolar-lavage fluid protein levels and functional antiplasmin activity (Fig. 4). It should be noted, however, that this relation does not apply in normal subjects or apply consistently in patients with other lung diseases. The results listed in Table 2 show that the lavage fluid recovered from several of the patients with pulmonary diseases other than ARDS had total protein values in the range of those of the patients with ARDS, yet fluid from the patients with other pulmonary diseases had easily measurable fibrinolytic activity. Thus, increased levels of total bronchoalveolar protein, although probably associated with increased plasma-derived antiplasmins, are not synonymous with reduced fibrinolytic activity. Indeed, the maintenance of normal bronchoalveolar fibrinolytic activity may account for the rapid clearance of alveolar hemorrhage and fibrinous exudate that occurs after a lung contusion.26 , 27
In the lavage fluid obtained from the patients with ARDS, virtually all the urokinase was in complexes with inhibitors, mainly PAI-1 (Fig. 5). The source or sources of this inhibitor were unclear. The simple diffusion of PAI-1 from the vascular bed, its release from platelets within the alveoli, or both could conceivably account for its presence in the lavage fluid. Previously reported data indicate that PAI-1 is an acute-phase reactant protein.28 However, its local production by alveolar epithelium is also possible. Obviously, other urokinase inhibitors could be important as well, such as protease nexin, which we did not assay specifically.29 The sites that regulate the production of urokinase inhibitors within the lung remain to be defined.
One limitation of this study is that we analyzed only bronchoalveolar-lavage fluid and thus cannot exclude the presence of other mechanisms of local alveolar fibrinolysis in the patients with ARDS that were not detected in the recovered fluid. For example, alveolar macrophages can bind fibrin, and these cells are known to express a cell-associated form of urokinase.10 Tissue plasminogen activator, if present, may be expected to be bound to the fibrin binding to macrophages and thus not found in the lavage fluid. Nonetheless, soluble urokinase accounts for the bulk of the normal bronchoalveolar activity of plasminogen activator, and there can be little doubt that the development of an overt antifibrinolytic state such as we found in the patients with ARDS favors the persistence of bronchoalveolar deposits of alveolar fibrin and fibronectin.
Important functional consequences are likely to follow from the persistence of these deposits. Among the plasma proteins, fibrin monomer and fibrinogen are the most potent inactivators of surfactant,30 , 31 which appears to be present but dysfunctional in patients with ARDS.32 , 33 Moreover, the pattern of deposition of fibrin and fibronectin in patients with ARDS correlates with the sites of subsequent pulmonary fibrosis.1 , 5 Both fibrin and fibronectin are known initiators of the fibrotic process. Fibrin appears to be coated with fibronectin, which then attracts and promotes the proliferation of fibroblasts.5 , 6 , 34 Given the likely importance of persistent bronchoalveolar fibrin and fibronectin, it is possible that treatment to remove these bronchoalveolar deposits may eventually prove beneficial in ARDS, like treatment intended to accelerate the resorption of vascular fibrin in atherosclerotic vascular disease. Now, however, such an approach is clearly premature. Although we have identified a major defect in the alveolar fibrinolytic pathway in patients with ARDS and defined a mechanism for it, at this time we have no data about its time course. Additional studies are needed to define the duration of loss of urokinase activity and to define better the source or sources of the urokinase inhibitors, especially PAI-1, that seem to account for this loss. Nonetheless, the defect demonstrated in this study provides a biochemical basis for exploring new possibilities in the treatment of patients with this severe and destructive acute pulmonary inflammatory process.
Note added in proof: Since this report was submitted for publication, Idell et al.35 have demonstrated reductions in urokinase activity and increases in PAI-1 levels with ARDS.
Supported by grants (HL 36563 and HL 23591) from the Public Health Service and the American Lung Association.
Presented in part at the annual meeting of the American Thoracic Society, Las Vegas, May 8 through 11, 1988.
We are indebted to Ms. Ameeta Singhal for technical assistance and to Drs. Taylor Thompson, John Munger, K. Atassi, and F. Verra for obtaining some of the lavage samples.
Source Information
From the Department of Medicine, Brigham and Women's Hospital (P.B., L.Z., H.A.C.), and the Department of Anesthesia, Massachusetts General Hospital, and Harvard Medical School (K.L., W.Z.), Boston; the Department of Obstetrics and Gynecology, University of Lund, Lund, Sweden (B.A.); and the Henri Mondor Hospital, Créteil, France (F.LeM.). Address reprint requests to Dr. Chapman at the Respiratory Division, Brigham and Women's Hospital, 75 Francis St., Boston, MA 02115.
- References
References
1
Spencer H. Pathology of the lung, excluding pulmonary tuberculosis. 3rd ed. New York: Pergamon Press, 1977:235–40.
2
Bachofen
Web of Science | Medline3
Pratt P. Pathology of adult respiratory distress syndrome. In: Thurlbeck WM, Abell MR, eds. The lung: structure, function, and disease. Baltimore: Williams & Wilkins, 1978:45–7.
4
Basset
Web of Science | Medline5
Fukuda
Web of Science | Medline6
Bitterman
CrossRef | Web of Science | Medline7
Grinnell
CrossRef | Web of Science | Medline8
Zapol
Web of Science | Medline9
Chapman
Web of Science | Medline10
Reilly
Web of Science | Medline11
Lowry
Web of Science | Medline12
Loskutoff
Web of Science | Medline13
Laemmli
CrossRef | Web of Science | Medline14
Astedt
CrossRef | Web of Science | Medline15
Jonsson Berg
CrossRef16
Granelli-Piperno
CrossRef | Web of Science | Medline17
Vassalli
CrossRef | Web of Science | Medline18
Erickson
CrossRef | Web of Science | Medline19
Dosne
CrossRef | Web of Science | Medline20
Lecander
CrossRef | Web of Science | Medline21
Chapman
Web of Science | Medline22
Colton T. Statistics in medicine. 1st ed. Boston: Little, Brown, 1974:219–28.
23
Bowen
Medline24
Chmielewska
CrossRef | Web of Science | Medline25
Sanzo
CrossRef | Web of Science | Medline26
Reynolds
Web of Science | Medline27
Wiot
CrossRef | Web of Science | Medline28
Kruithof
Web of Science | Medline29
Baker
CrossRef | Web of Science | Medline30
Taylor
CrossRef | Web of Science31
Seeger
Web of Science | Medline32
Hallman
CrossRef | Web of Science | Medline33
Mason
Medline34
Dvorak
Full Text | Web of Science | Medline35
Idell
CrossRef | Web of Science | Medline
- Citing Articles (52)
Citing Articles
1
Maneesh Bhargava, Chris Wendt. (2012) Biomarkers in acute lung injury. Translational Research
CrossRef2
Marcel Levi, Tom Poll, Marcus Schultz. (2011) Systemic versus localized coagulation activation contributing to organ failure in critically ill patients. Seminars in Immunopathology
CrossRef3
Ashwini Katre, Carol Ballinger, Hasina Akhter, Michelle Fanucchi, Dae-Kee Kim, Edward Postlethwait, Rui-Ming Liu. (2011) Increased transforming growth factor beta 1 expression mediates ozone-induced airway fibrosis in mice. Inhalation Toxicology 23:8, 486-494
CrossRef4
Kristy A. Bauman, Scott H. Wettlaufer, Katsuhide Okunishi, Kevin M. Vannella, Joshua S. Stoolman, Steven K. Huang, Anthony J. Courey, Eric S. White, Cory M. Hogaboam, Richard H. Simon, Galen B. Toews, Thomas H. Sisson, Bethany B. Moore, Marc Peters-Golden. (2010) The antifibrotic effects of plasminogen activation occur via prostaglandin E2 synthesis in humans and mice. Journal of Clinical Investigation 120:6, 1950-1960
CrossRef5
Marcin Kłak, Nicola Änäkkälä, Wanzhong Wang, Stefan Lange, Ing-Marie Jonsson, Andrej Tarkowski, Tao Jin. (2010) Tranexamic acid, an inhibitor of plasminogen activation, aggravates staphylococcal septic arthritis and sepsis. Scandinavian Journal of Infectious Diseases 42:5, 351-358
CrossRef6
Anil Sapru, Martha A. Q. Curley, Sandra Brady, Michael A. Matthay, Heidi Flori. (2010) Elevated PAI-1 is associated with poor clinical outcomes in pediatric patients with acute lung injury. Intensive Care Medicine 36:1, 157-163
CrossRef7
Iraklis Tsangaris, Argirios Tsantes, Pantelis Bagos, Georgios Nikolopoulos, Christos Kroupis, Petros Kopterides, Ioanna Dimopoulou, Stylianos Orfanos, Aikaterini Kardoulaki, Stavros Chideriotis, Anthi Travlou, Apostolos Armaganidis. (2009) The Effect of Plasma Homocysteine Levels on Clinical Outcomes of Patients With Acute Lung Injury/Acute Respiratory Distress Syndrome. The American Journal of the Medical Sciences 338:6, 474-477
CrossRef8
Yung-Yang Liu, Shuen-Kuei Liao, Chung-Chi Huang, Ying-Huang Tsai, Deborah A. Quinn, Li-Fu Li. (2009) Role for nuclear factor-κB in augmented lung injury because of interaction between hyperoxia and high stretch ventilation. Translational Research 154:5, 228-240
CrossRef9
Arne P Neyrinck, Kathleen D Liu, James P Howard, Michael A Matthay. (2009) Protective mechanisms of activated protein C in severe inflammatory disorders. British Journal of Pharmacology 158:4, 1034-1047
CrossRef10
M.O. Maybauer, S. Rehberg, D.L. Traber, D.N. Herndon, D.M. Maybauer. (2009) Pathophysiologie des akuten Lungenversagens bei Schwerbrandverletzten mit Inhalationstrauma. Der Anaesthesist 58:8, 805-812
CrossRef11
Anil Sapru, Helen Hansen, Temitayo Ajayi, Ron Brown, Oscar Garcia, HanJing Zhuo, Joseph Wiemels, Michael A. Matthay, Jeanine Wiener-Kronish. (2009) 4G/5G Polymorphism of Plasminogen Activator Inhibitor -1 Gene Is Associated with Mortality in Intensive Care Unit Patients with Severe Pneumonia. Anesthesiology 110:5, 1086-1091
CrossRef12
Iraklis Tsangaris, Argiris Tsantes, Stefanos Bonovas, Michalis Lignos, Petros Kopterides, Argiro Gialeraki, Evdoxia Rapti, Stylianos Orfanos, Ioanna Dimopoulou, Anthi Travlou, Apostolos Armaganidis. (2009) The impact of the PAI-1 4G/5G polymorphism on the outcome of patients with ALI/ARDS. Thrombosis Research 123:6, 832-836
CrossRef13
Ricard Ferrer, Mélanie Adda, Ana Navas, Maddalena Pasini, Antoni Artigas. (2009) Blood coagulation and inflammation in acute lung injury. Journal of Organ Dysfunction 5:2, 101-109
CrossRef14
Rui-Ming Liu. (2008) Oxidative Stress, Plasminogen Activator Inhibitor 1, and Lung Fibrosis. Antioxidants & Redox Signaling 10:2, 303-320
CrossRef15
Argirios E. Tsantes, Georgios K. Nikolopoulos, Pantelis G. Bagos, Stefanos Bonovas, Petros Kopterides, Georgios Vaiopoulos. (2008) The effect of the plasminogen activator inhibitor-1 4G/5G polymorphism on the thrombotic risk. Thrombosis Research 122:6, 736-742
CrossRef16
P. Dahlem, W.M.C. van Aalderen, A.P. Bos. (2007) Pediatric acute lung injury. Paediatric Respiratory Reviews 8:4, 348-362
CrossRef17
Lorraine B. Ware, Michael A. Matthay, Polly E. Parsons, B Taylor Thompson, James L. Januzzi, Mark D. Eisner. (2007) Pathogenetic and prognostic significance of altered coagulation and fibrinolysis in acute lung injury/acute respiratory distress syndrome*. Critical Care Medicine 35:8, 1821-1828
CrossRef18
Ali A. El Solh, Goda Choi, Marcus J. Schultz, Lilibeth A. Pineda, Corey Mankowski. (2007) Clinical and hemostatic responses to treatment in ventilator-associated pneumonia: Role of bacterial pathogens*. Critical Care Medicine 35:2, 490-496
CrossRef19
S. Patel, L.R. Berry, A.K.C. Chan. (2007) Covalent antithrombin–heparin complexes. Thrombosis Research 120:2, 151-160
CrossRef20
Hajime Fujimoto, Esteban C. Gabazza, Osamu Taguchi, Yoichi Nishii, Hiroki Nakahara, Nelson E. Bruno, Corina N. D'Alessandro-Gabazza, Michael Kasper, Yutaka Yano, Mariko Nagashima, John Morser, George J. Broze, Koji Suzuki, Yukihiko Adachi. (2006) Thrombin-Activatable Fibrinolysis Inhibitor Deficiency Attenuates Bleomycin-Induced Lung Fibrosis. The American Journal of Pathology 168:4, 1086-1096
CrossRef21
Ali A. El Solh, Milapchand Bhora, Lilibeth Pineda, Alan Aquilina, Laurie Abbetessa, Eileen Berbary. (2006) Alveolar plasminogen activator inhibitor-1 predicts ARDS in aspiration pneumonitis. Intensive Care Medicine 32:1, 110-115
CrossRef22
Steven E. MUTSAERS, Cecilia M. PRELE, Arnold R. BRODY, Steven IDELL. (2004) Pathogenesis of pleural fibrosis. Respirology 9:4, 428-440
CrossRef23
Kathleen A Stringer, John S Dunn, Daniel L Gustafson. (2004) ADMINISTRATION OF EXOGENOUS TISSUE PLASMINOGEN ACTIVATOR REDUCES OEDEMA IN MICE LACKING THE TISSUE PLASMINOGEN ACTIVATOR GENE. Clinical and Experimental Pharmacology and Physiology 31:5-6, 327-330
CrossRef24
Noboru Hattori, Shinya Mizuno, Yuka Yoshida, Kazuo Chin, Michiaki Mishima, Thomas H. Sisson, Richard H. Simon, Toshikazu Nakamura, Masayuki Miyake. (2004) The Plasminogen Activation System Reduces Fibrosis in the Lung by a Hepatocyte Growth Factor-Dependent Mechanism. The American Journal of Pathology 164:3, 1091-1098
CrossRef25
Pierre-François Laterre, Xavier Wittebole, Jean-François Dhainaut. (2003) Anticoagulant therapy in acute lung injury. Critical Care Medicine 31:Supplement, S329-S336
CrossRef26
James A. Russell. (2003) Genetics of coagulation factors in acute lung injury. Critical Care Medicine 31:Supplement, S243-S247
CrossRef27
Steven Idell. (2003) Coagulation, fibrinolysis, and fibrin deposition in acute lung injury. Critical Care Medicine 31:Supplement, S213-S220
CrossRef28
Kazunori Murakami, Roy McGuire, Robert A. Cox, Jeffrey M. Jodoin, Frank C. Schmalstieg, Lillian D. Traber, Hal K. Hawkins, David N. Herndon, Daniel L. Traber. (2003) Recombinant antithrombin attenuates pulmonary inflammation following smoke inhalation and pneumonia in sheep. Critical Care Medicine 31:2, 577-583
CrossRef29
A.-M. B. Münster, L. Rasmussen, J. Sidelmann, J. Ingemann Jensen, B. Bech, J. Gram. (2002) Effects of inhaled plasminogen activator on the balance between coagulation and fibrinolysis in traumatized pigs. Blood Coagulation & Fibrinolysis 13:7, 591-601
CrossRef30
I. Martin, M. Humbert, A. Marfaing-Koka, F. Capron, M. Wolf, D. Meyer, G. Simonneau, E. Anglés-Cano. (2002) Plasminogen activation by blood monocytes and alveolar macrophages in primary pulmonary hypertension. Blood Coagulation & Fibrinolysis 13:5, 417-422
CrossRef31
Steven Idell. (2002) Endothelium and disordered fibrin turnover in the injured lung: Newly recognized pathways. Critical Care Medicine 30:Supplement, S274-S280
CrossRef32
Steven Idell. (2002) The Fibrinolytic Defect in Adult Respiratory Distress Syndrome: A New Therapeutic Opportunity?. Clinical Pulmonary Medicine 9:1, 13-19
CrossRef33
Noboru Hattori, Jay L. Degen, Thomas H. Sisson, Hong Liu, Bethany B. Moore, Raj G. Pandrangi, Richard H. Simon, Angela F. Drew. (2000) Bleomycin-induced pulmonary fibrosis in fibrinogen-null mice. Journal of Clinical Investigation 106:11, 1341-1350
CrossRef34
Thomas H. Sisson, Noboru Hattori, Yin Xu, Richard H. Simon. (1999) Treatment of Bleomycin-Induced Pulmonary Fibrosis by Transfer of Urokinase-Type Plasminogen Activator Genes. Human Gene Therapy 10:14, 2315-2323
CrossRef35
James P. Chen, Dennis W. Rowe, Blaine L. Enderson. (1999) Contrasting Post-Traumatic Serial Changes for D-Dimer and PAI-1 in Critically Injured Patients. Thrombosis Research 94:3, 175-185
CrossRef36
Noboru Hattori, Thomas H. Sisson, Yin Xu, Richard H. Simon. (1999) Upregulation of Fibrinolysis by Adenovirus-Mediated Transfer of Urokinase-Type Plasminogen Activator Genes to Lung Cells in Vitro and in Vivo. Human Gene Therapy 10:2, 215-222
CrossRef37
C. Lardot, Y. Sibille, B. Wallaert,. (1999) Depressed urokinase activity in bronchoalveolar lavage fluid from patients with sarcoidosis, silicosis or idiopathic pulmonary Rbrosis: relationship to disease severity. Biomarkers 4:5, 361-372
CrossRef38
Peter Carmeliet, Désiré Collen. (1998) Development and Disease in Proteinase-Deficient Mice. Thrombosis Research 91:6, 255-285
CrossRef39
Kathleen A Stringer, Brooks M Hybertson, Okyong J Cho, Zoe Cohen, John E Repine. (1998) Tissue Plasminogen Activator (tPA) Inhibits Interleukin-1 Induced Acute Lung Leak. Free Radical Biology and Medicine 25:2, 184-188
CrossRef40
Jean-Philippe Rougier, Sophie Guia, Jacqueline Hagège, Geneviève Nguyen, Pierre M Ronco. (1998) PAI-1 secretion and matrix deposition in human peritoneal mesothelial cell cultures: Transcriptional regulation by TGF-β1. Kidney International 54:1, 87-98
CrossRef41
Sandra E. Juul, Harry S. Nick, Gary A. Visner. (1998) Identification of Urokinase as a Hyperoxia-Inducible Gene. Molecular Genetics and Metabolism 63:4, 295-301
CrossRef42
, A. Artigas, G. R. Bernard, J. Carlet, D. Dreyfuss, L. Gattinoni, L. Hudson, M. Lamy, J. J. Marini, M. A. Matthay, M. R. Pinsky, R. Spragg, P. M. Suter. (1998) The American-European consensus conference on ARDS, Part 2. Intensive Care Medicine 24:4, 378-398
CrossRef43
Satoshi Gando, Takashi Kameue, Satoshi Nanzaki, Tatsuya Hayakawa, Yoshimi Nakanishi. (1997) Increased Neutrophil Elastase, Persistent Intravascular Coagulation, and Decreased Fibrinolytic Activity in Patients with Posttraumatic Acute Respiratory Distress Syndrome. The Journal of Trauma: Injury, Infection, and Critical Care 42:6, 1068-1072
CrossRef44
KAMAL K. SINGHAL, LANCE A. PARTON. (1996) Plasminogen Activator Activity in Preterm Infants with Respiratory Distress Syndrome: Relationship to the Development of Bronchopulmonary Dysplasia. Pediatric Research 39:2, 229-235
CrossRef45
Jeffrey B. Pardes, Hajime Takagi, Theresa A. Martin, M. Sofia Ochoa, Vincent Falanga. (1995) Decreased levels of alpha 1(I) procollagen mRNA in dermal fibroblasts grown on fibrin gels and in response to fibrinopeptide B. Journal of Cellular Physiology 162:1, 9-14
CrossRef46
SØREN K. MOESTRUP, ERIK I. CHRISTENSEN, SØREN NIELSEN, KARL E. JØRGENSEN, SØREN E. BJØRN, HANS RØIGAARD, JØRGEN GLIEMANN. (1994) Binding and Endocytosis of Proteins Mediated by Epithelial gp330. Annals of the New York Academy of Sciences 737:1 Biology of 2-, 124-137
CrossRef47
J. V. Sørensen, H. P. Jensen, H. B. Rahr, L. C. Borris, M. R. Lassen, O. Fedders, J. P. Haase, F. Knudsen. (1993) Haemostatic activation in patients with head injury with and without simultaneous multiple trauma. Scandinavian Journal of Clinical & Laboratory Investigation 53:7, 659-665
CrossRef48
Philippe de Moerloose, Edoardo De Benedetti, Laurent Nicod, Catherine Vifian, Guido Reber. (1992) Procoagulant activity in bronchoalveolar fluids: No relationship with tissue factor pathway inhibitor activity. Thrombosis Research 65:4-5, 507-518
CrossRef49
Joachim Yahalom, Carol S. Portlock. (1991) Autologous bone marrow transplantation. Current Problems in Cancer 15:1, 5-57
CrossRef50
(1990) Aerosolized Amiloride for the Treatment of Lung Disease in Cystic Fibrosis. New England Journal of Medicine 323:14, 996-998
Full Text51
(1990) A Fibrinolytic Defect in the Respiratory Distress Syndrome. New England Journal of Medicine 323:6, 422-422
Full Text52
McDonald, John A., . (1990) The Yin and Yang of Fibrin in the Airways. New England Journal of Medicine 322:13, 929-931
Full Text
- Letters
