Abstract
Although intrapulmonary fibrin deposition is a pathognomonic feature of acute lung injury, it remains uncertain whether thrombin inhibitors affect clinically important outcomes. We hypothesized that both heparin and antithrombin (AT) concentrate improve gas exchange during experimental respiratory distress syndrome. We also tested whether combination therapy is more beneficial than monotherapy. Forty-eight newborn piglets were randomized within 12 litters to one of four groups in a factorial design: (1) AT; (2) heparin; (3) AT plus heparin; (4) untreated control animals. After lung lavage and 4 h of barovolutrauma, mechanical ventilation was continued for 24 h during which ventilator pressures and inspired oxygen were adjusted to maintain normal blood gases. The arterial/ alveolar oxygen tension ratio (a/A ratio) and the ventilator efficiency index (VEI) at 18 and 24 h were compared by repeated measures analysis of variance (ANOVA). In contrast to our hypothesis, only heparin improved gas exchange, and we found little evidence of an interaction with AT. The a/A ratio was 0.48 ± 0.27 (mean ± SD) in the presence of heparin versus 0.33 ± 0.26 in its absence; p = 0.01. Corresponding VEI was 0.30 ± 0.12 versus 0.25 ± 0.14; p = 0.04. Hyaline membrane formation was also decreased in heparin-treated animals (p = 0.02).
Intrapulmonary accumulation of fibrin and increased procoagulant activity of bronchoalveolar lavage (BAL) fluid are pathognomonic features of acute lung injury (1-4). The presence of fibrin implies that clotting occurs in the injured lung. Activation of the coagulation system has been documented in both newborn and adult patients with respiratory distress syndrome (5-7). Several groups of investigators have suggested that thrombin formation and fibrin deposition in the lung may exacerbate pulmonary hypertension (8), increase endothelial permeability (9), inactivate surfactant (10), and amplify lung inflammation (11) and fibrosis (12, 13).
We have previously shown that the thrombin inhibitors antithrombin (AT) and hirudin reduce intrapulmonary accumulation of fibrinogen and decrease procoagulant activity of BAL fluid during experimental acute lung injury caused by pulmonary overdistension (14). However, the excessive tidal volumes which were used throughout those studies to generate and sustain lung injury made it impossible to determine whether the reduced potential for intrapulmonary fibrin deposition was associated with improvements in gas exchange. Moreover, hirudin caused unacceptable bleeding side effects, which discouraged us from testing its use any further (14). Recently, Cox and coworkers reported that systemic heparin administration after smoke inhalation injury increased oxygenation and reduced lung water in sheep (15). We therefore hypothesized that both heparin and AT improve gas exchange in newborn piglets with experimental acute lung injury. Moreover, because heparin accelerates the inhibition of thrombin and other serine proteases by AT, we hypothesized further that combination therapy with these two agents is more beneficial than monotherapy.
Newborn Yorkshire piglets, less than 72 h and of either sex, were obtained from a closed herd with excellent health status (Arkell Swine Research Center, University of Guelph, Ontario, Canada). All animal procedures were carried out in accordance with the guidelines of the Canadian Council on Animal Care and the regulations under the Province of Ontario “Animals for Research Act.” The experimental protocol was approved by the Animal Research Ethics Board at McMaster University.
Piglets were weighed and anesthetized with an intraperitoneal injection of sodium pentobarbital (30 mg/kg; MTC Pharmaceutics, Cambridge, ON, Canada). Endotracheal 3.5-Fr tubes (Concord/Portex, Keene, NH) were placed via tracheostomies and connected to an infant ventilator (Model 102; Healthdyne Inc., Marietta, GA). Initial ventilator settings were chosen to ensure physiologic gas exchange: peak inspiratory pressure (PIP) of 1.5 kPa (15 cm H2O), positive end-expiratory pressure (PEEP) of 0.3 kPa (3 cm H2O), inspiratory time (IT) of 0.5 s, respiratory rate (RR) of 30 breaths/min. A humidified gas mixture (warmed to 38° C) of 40% oxygen in nitrogen was delivered at 8 L/min. A 3.5-Fr Argyle umbilical vessel catheter (Sherwood Medical, St. Louis, MO) was inserted into the left carotid artery. The catheter was secured at 5 cm, perfused with normal saline (1.5 ml/h; Baxter Corp., Toronto, ON, Canada), and used for blood sampling and continuous systemic blood pressure monitoring. The umbilical vein was cannulated with a second 3.5-Fr Argyle umbilical vessel catheter for the administration of maintenance fluids (5% dextrose at 80 ml · kg−1 · d−1; Baxter Corp.), prophylactic ampicillin (100 mg · kg−1 · d−1; Novopharm Ltd., Toronto, ON, Canada), and gentamicin (5 mg · kg−1 · d−1; Schering Canada Inc., Pointe Claire, PQ, Canada). To prevent tension pneumothorax, bilateral 10-Fr trochar catheters (Sherwood Medical, St. Louis, MO) were inserted into the chest, advanced 2 cm, and connected to Heimlich chest drain valves (Becton Dickinson Co., Lincoln Park, NJ). After these procedures, piglets were allowed to recover in prewarmed isolettes for 1 h.
Throughout the entire study period, the piglets were regularly monitored for maintenance of blood pressure, heart rate, serum glucose, and oxygen saturation (Ohmeda Biox 3700 Pulse Oximeter; Boc Health Care, Louisville, CO). A rectal infant servo control (ISC) probe was inserted and core body temperature was maintained at 39° C. Suprapubic bladder punctures were performed every 6 h. Additional sodium pentobarbital (16 mg/kg) was given intravenously as required to maintain adequate sedation. Intravenous pancuronium bromide (0.2 mg/kg; Abbott Laboratories Ltd., Sainte Laurent, PQ, Canada) was administered at regular intervals to eliminate spontaneous breathing. If needed, sodium bicarbonate (4.2%) was given to correct metabolic acidosis.
Treatment groups. At the beginning of each experiment, four littermates were randomized (using a random number table) to receive human AT concentrate (Kybernin; Behringwerke AG, Marburg, Germany), heparin (Hepalean; Organon Teknika Inc., Toronto, ON, Canada), AT plus heparin, or untreated controls. An intravenous AT loading dose of 250 U · kg−1 was followed by intravenous injections of 100 U · kg−1 every 6 h. Heparin was administered as a continuous intravenous infusion at a rate of 30 U · kg−1 · h−1, after an initial bolus of 50 U · kg−1.
Bronchoalveolar lavage and high-pressure ventilation. To induce lung injury, all piglets received bronchoalveolar lavage followed by 4 h of high-pressure ventilation according to our previously published protocol (16). Briefly, 35 ml/kg of normal saline, warmed to 37° C, was instilled by gravity through the endotracheal tube and connecting tubing. The saline was allowed to dwell in the lungs for 5 s before it was recovered over 20 s by passive drainage. Approximately 80% of the total volume instilled was retrieved. After each lavage, PIP was increased by 0.1 kPa (1 cm H2O) to compensate for decreases in compliance. Each piglet was lavaged six times at approximately 5-min intervals, to allow SaO2 to recover to values > 90% in a fraction of inspired oxygen (Fi O2 ) of 1.0. Three lavages were performed while the piglets were lying on the right side and three while lying on the left side. High-pressure ventilation was commenced immediately after the lavage procedure. PIP was increased to 3.9 kPa (40 cm H2O). PEEP was reduced to 0.2 kPa (2 cm H2O), RR was set at 20 breaths/min, while the IT was increased to 1.2 s. To maintain arterial CO2 tension (PaCO2 ) between 3.3 and 4.0 kPa (25 and 30 mm Hg), the dead space volume of the ventilator circuit was increased. Humidified pure oxygen was delivered at a flow rate of 15 L/min. During this 4-h period of mechanical ventilation at high peak pressures, the animals were nursed for 2 h each on either side.
Rescue ventilation. At the end of the 4-h period of pulmonary overdistension, rescue ventilation was initiated and maintained for 24 h. Ventilatory settings were modified as follows: PEEP was increased to 0.39 kPa (4 cm H2O), IT 0.5 s, and RR 30 breaths/min. The extra dead space was removed. Arterial blood gas samples were taken every 6 h. PIP and Fi O2 were adjusted to achieve normocarbia and normoxemia. During this period of rescue ventilation, the animals were nursed on either side, alternating every 6 h.
Hematologic studies. Complete blood counts were performed at baseline in EDTA-anticoagulated blood samples using a Coulter counter (Coulter Electronics, Miami, FL). Coagulation tests were performed in platelet-poor citrated plasma samples which had been stored at −70° C for batch assaying. AT activity was determined using a chromogenic substrate specific for factor Xa (Coamate; Chromogenix, Mölndal, Sweden).
BAL was performed immediately after each animal had been killed by an intravenous injection of sodium pentobarbital. With the piglet held in an upright position, the entire lung was lavaged in situ with 5 ml/kg of sterile 0.9% saline. The animal was then placed in the supine position and a suction catheter was advanced to the tip of the endotracheal tube. Through careful aspiration, approximately 80% of the instilled fluid volume was recovered. BAL samples were centrifuged at 3,000 rpm for 10 min, and the supernatant was stored at −70° C until assayed. The procoagulant activity of the BAL sample (BAL plasma recalcification time) was measured as described by O'Brodovich and coworkers (17). In order to quantitate the amount of exogenous AT that could be detected in BAL fluid, an ELISA for human AT was performed using a commercially available kit (Affinity Biologicals, Hamilton, ON, Canada). Qualitative analyses for the human AT in BAL were carried out by Western immunoblots, using standard methods (18). The capture and detecting antibodies for the ELISA and the probing antibody for the immunoblots (also from Affinity Biologicals) do not cross-react with porcine AT. Transfers were to immobilon membrane after sodium dodecyl sulfate–polyacrylamide gel electropheresis (7.5% separating gel). Probing with the alkaline phosphatase-linked detecting antibody was in buffered skim milk while the blot was developed using nitroblue tetrazolium/bromo- (chloro)indolylphosphate (NBT/BCIP) substrates.
Lung function. Expiratory tidal volumes were determined throughout the experiment using a neonatal volume monitor (Bear Medical Systems, Riverside, CA). Blood gases were measured in heparinized arterial samples (AVL 995 pH/Blood Gas Analyser; AVL Scientific Corporation, Roswell, GA). The ventilator efficiency index (VEI) (19) and the ratio of arterial to alveolar oxygen tension (a/A ratio) were computed at 18 and 24 h of rescue ventilation. The VEI is a measure which relates alveolar ventilation to respirator input in the absence of spontaneous breathing.
Lung uptake of circulating 125I-fibrinogen. Human fibrinogen was labeled in the regional radiopharmacy with 125I as for our previous studies. Five μCi of 125I-fibrinogen was injected intravenously into each animal at baseline. At autopsy, the right lung was selectively perfused with 60 ml of normal saline through its pulmonary artery to clear 125I-fibrinogen from the intravascular space (2). The lung was then isolated and counted for 5 min in a standardized manner using a large-diameter, thin-window, thallium-activated sodium iodide crystal (Harshaw Chemical Co., Solon, OH). The sample-to-detector distance was large (40 cm) to minimize any sample size effects. A low background environment was created by the shadow shield of a whole body counter. The multichannel analyzer was set to detect the 28 KeV photons emitted by 125I. The lung count was related to the total body count obtained in the same manner after death and before autopsy.
Histologic examination. The left lung was immersed in 10% buffered formalin for 72 h. The lung was sliced and samples from all lobes were embedded in paraffin. Samples were selected to provide a view of the lung from the hilum to the pleura. Sections cut at 5 μm were stained with hematoxylin/eosin or phosphotungstic acid/hematoxylin. An independent pathologist (D.d.S), who was unaware of the treatment allocation, performed the histologic examination, using the scoring system described by Tsuno and coworkers (20). Eleven features of lung pathology (emphysematous change, interstitial congestion, alveolar hemorrhage, alveolar neutrophil infiltration, alveolar macrophage proliferation, alveolar type II pneumocyte proliferation, interstitial lymphocyte infiltration, interstitial thickening, hyaline membrane formation, interstitial fibrosis, and organization of alveolar exudate) were graded as negative = 0, slight = 1, mild = 2, moderate = 3, and severe = 4. The analysis was focused on the scores for interstitial congestion, alveolar hemorrhage, alveolar neutrophil infiltration, and hyaline membrane formation. In addition, a summary score was computed for each piglet, for which all 11 subscores were used.
Heparin and AT were allocated in a factorial arrangement within litters, the four treatment combinations being assigned at random among four piglet littermates. The general analytic approach used was repeated measures analysis of variance (ANOVA) (21) with “litter” being the random factor and “treatment combination” a fixed within-litter factor. The factorial design allowed the three degrees of freedom between treatment comparison to be split into single degree of freedom comparisons representing the separate effects of heparin and AT, and a test of their interaction. The latter test examined whether the separate effects of heparin and AT were additive. Because a/A ratio and VEI were measured at two time points (18 h and 24 h), their analyses could be conducted separately at each time point but also combining the data over time points. The ANOVA for the combined analysis includes tests of the effect of time and time × treatment interaction.
Although treatments were initially balanced within litters, the early death of some piglets produced missing values for a/A ratio and VEI. Statistical computation thus utilized BMDP5V (22) which caters to unbalanced repeated measures ANOVA. All reported p values are two-tailed.
A total of 48 piglets from 12 litters were studied. Mean birth weights (± SD) were comparable among the four groups: control animals 1,500 ± 280 g, AT 1,590 ± 280 g, heparin 1,640 ± 370 g, AT plus heparin 1,610 ± 300 g, respectively. Six piglets died before the completion of the 24-h period of rescue ventilation; two animals each in the control and heparin groups, and one animal each in the two remaining treatment groups.
Table 1 summarizes important hematologic parameters for each of the four treatment groups at baseline, and at the end of each experiment. The activated partial thromboplastin time (aPTT) was not prolonged by the heparin dose used in these experiments. Plasma AT activity was characteristically low at baseline in all four groups; it was raised to supraphysiologic levels by treatment with AT concentrate (Table 1). Exogenous (i.e., human) AT was recovered in the BAL fluid of piglets that had been allocated to AT infusions. BAL plasma recalcification time was shortest in untreated animals, whereas group means were similar in all three treatment groups (Table 1).
| Time | Control | Antithrombin | Heparin | Antithrombin and Heparin | p Value | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Platelets, × 109 L | ||||||||||
| Baseline | 303 ± 86.9 | 343 ± 55.5 | 313 ± 50.8 | 289 ± 122.3 | 0.265† | |||||
| End of experiment | 225 ± 78.9 | 265 ± 111.3 | 204 ± 96.1 | 259 ± 125.9 | 0.308† | |||||
| aPTT, s | ||||||||||
| Baseline | 28.9 ± 5.25 | 28.0 ± 4.87 | 29.3 ± 5.57 | 29.3 ± 4.12 | 0.816† | |||||
| End of experiment | 27.1 ± 2.71 | 27.3 ± 2.17 | 27.7 ± 4.46 | 28.2 ± 3.88 | 0.778† | |||||
| Plasma antithrombin | ||||||||||
| Baseline | 0.54 ± 0.11 | 0.59 ± 0.10 | 0.54 ± 0.09 | 0.58 ± 0.13 | 0.891† | |||||
| End of experiment | 0.64 ± 0.11 | 2.43 ± 0.81 | 0.51 ± 0.11 | 2.16 ± 0.50 | < 0.0001† | |||||
| BAL plasma recalcification time, s | ||||||||||
| End of experiment | 49.7 ± 8.65 | 62.4 ± 20.23 | 60.9 ± 17.9 | 57.9 ± 10.75 | 0.108† | |||||
| BAL antithrombin, μg/ml | ||||||||||
| End of experiment | < 0.5 | 30.1 (1.7, 120.9) | < 0.5 | 17.7 (0.1, 69.0) | 0.354‡ |
Mean expiratory tidal volumes (± SD) after 24 h of rescue ventilation were: control animals 10.4 ± 2.68 ml/kg; AT 13.3 ± 3.81 ml/kg; heparin 13.2 ± 3.38 ml/kg; AT plus heparin 12.9 ± 4.93 ml/kg. The results for the two measures of gas exchange are summarized in Table 2. For each outcome measure the table contains the mean and between-piglet standard deviation (SD) for each of the four randomized groups. The right-hand panel of the table contains the p values from the unbalanced repeated measures ANOVA which includes the test of the average effects of heparin and AT, and the interaction. For a/A ratio and VEI separate analyses are provided at 18 h of rescue ventilation, 24 h, and the combined 18 and 24 h data.
| Outcome Measure | Time (h) | Control (mean ± SD) | AT (mean ± SD) | Heparin (mean ± SD) | Hep + AT (mean ± SD) | p Values | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AT | Heparin | Interaction | ||||||||||||||
| a/A Ratio | 18 h | 0.31 ± 0.23 | 0.41 ± 0.34 | 0.47 ± 0.32 | 0.48 ± 0.22 | 0.42 | 0.13 | 0.41 | ||||||||
| 24 h | 0.28 ± 0.29 | 0.32 ± 0.16 | 0.47 ± 0.34 | 0.51 ± 0.26 | 0.68 | 0.04 | 0.97 | |||||||||
| Combined | 0.30 ± 0.25 | 0.37 ± 0.28 | 0.47 ± 0.33 | 0.49 ± 0.23 | 0.34 | 0.01 | 0.57 | |||||||||
| VEI | 18 h | 0.26 ± 0.14 | 0.29 ± 0.18 | 0.32 ± 0.13 | 0.31 ± 0.11 | 0.88 | 0.17 | 0.56 | ||||||||
| 24 h | 0.20 ± 0.10 | 0.25 ± 0.10 | 0.28 ± 0.13 | 0.28 ± 0.11 | 0.58 | 0.14 | 0.57 | |||||||||
| Combined | 0.24 ± 0.13 | 0.27 ± 0.15 | 0.30 ± 0.13 | 0.30 ± 0.11 | 0.60 | 0.04 | 0.39 | |||||||||
The data for a/A ratio at 24 h are presented graphically in Figure 1 for illustrative purposes. Compared with control with a mean a/A ratio of 0.28, the addition of heparin increased the mean to 0.47, a change of 0.19. In the presence of AT, the addition of heparin also increased the mean a/A ratio by 0.19. The fact that the heparin effect was equal with and without AT indicates that no interaction was present (i.e., the lines in Figure 1 are approximately parallel) and thus that the effects of heparin and AT are additive. The mean increase with AT over control and in the presence of heparin was only 0.04. As seen in the right-hand panel of Table 2, the formal test at 24 h of the average heparin effect was 0.04, the average AT effect 0.68, and the interaction 0.97.
Fig. 1. a/A ratio after 24 h of rescue ventilation in the four treatment groups: the presence of heparin improves the mean a/A ratio, while the addition of AT has little effect. The two lines in the above diagram are approximately parallel, suggesting a lack of interaction between heparin and AT.
[More] [Minimize]Although showing slightly more evidence of interaction, the 18-h a/A ratio data also support a larger effect of heparin than AT. The analysis of the combined 18- and 24-h data thus has a stronger p value of 0.01 associated with the heparin effect, the AT effect remains unimpressive statistically (p = 0.34), and there is little evidence of interaction (p = 0.57). Although not included in the table, the p values for the treatment × time interactions were firmly nonsignificant, indicating that the treatment effects were reasonably consistent at each time point.
The VEI data exhibit similar patterns to those of the a/A ratio. At both 18 and 24 h of rescue ventilation, the observed improvement in mean VEI is larger with heparin than AT; again there is little real evidence of an interaction. The heparin effect is statistically marginal when considered separately at 18 and 24 h but does attain conventional significance in the combined analysis.
A substantial fraction of the total body count of 125I-fibrinogen was located to the tissues of the right lung in all animals, suggesting endothelial permeability to large molecules (Table 3). We have previously shown that no lung uptake of this tracer was detectable in four unventilated healthy newborn piglets 24 h after intravenous injection of 125I-fibrinogen under identical experimental conditions (14). In the present study, the lung uptake of 125I-fibrinogen tended to be lower in animals receiving heparin and/or AT than in control littermates, although neither treatment effect was statistically significant with the present sample size (Table 3).
| Outcome Measure | Control (mean ± SD) | AT (mean ± SD) | Heparin (mean ± SD) | Hep + AT (mean ± SD) | p Values | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| AT | Heparin | Interaction | ||||||||||||
| Lung uptake of 125I-fibrinogen, % | 19.2 ± 10.4 | 14.0 ± 9.0 | 13.0 ± 11.9 | 10.3 ± 9.8 | 0.10 | 0.08 | 0.69 | |||||||
| Interstitial congestion | 2.42 ± 1.00 | 1.92 ± 0.67 | 1.83 ± 0.83 | 2.00 ± 0.60 | 0.50 | 0.24 | 0.17 | |||||||
| Alveolar hemorrhage | 2.33 ± 0.89 | 1.67 ± 0.98 | 1.75 ± 0.62 | 1.58 ± 0.79 | 0.13 | 0.18 | 0.29 | |||||||
| Alveolar neutrophil infiltration | 2.08 ± 0.67 | 1.75 ± 0.75 | 1.58 ± 1.08 | 1.58 ± 0.79 | 0.49 | 0.18 | 0.42 | |||||||
| Hyaline membrane formation | 3.25 ± 0.87 | 2.83 ± 1.47 | 2.17 ± 1.34 | 2.50 ± 0.90 | 0.85 | 0.02 | 0.23 | |||||||
| Lung histology: summary score | 1.99 ± 0.71 | 1.77 ± 0.81 | 1.44 ± 0.89 | 1.55 ± 0.63 | 0.79 | 0.09 | 0.34 | |||||||
Within each treatment group, there was considerable variability between animals in the histologic response to the injury protocol. Figure 2 illustrates the observed spectrum of histologic lung injury. Despite this variability, we were able to detect a beneficial effect of heparin on hyaline membrane formation (Table 3). As suggested by the fixed effects p values in Table 3, there was also a trend toward lower and therefore better histology summary scores in heparin-treated animals.
Fig. 2. Observed spectrum of histologic lung injury in four untreated control animals, ranging from slight to moderately severe injury: the individual histologic summary scores were 0.64 (left upper panel ), 1.64 (right upper panel ), 2.45 (left lower panel ), and 2.91 (right lower panel ). Sections were stained with hematoxylin/eosin.
[More] [Minimize]In contrast to our prior hypothesis, heparin, but not AT, improved gas exchange in newborn piglets with experimental acute lung injury, and there was little evidence of an interaction between the two agents. Heparin also decreased the histologic severity of hyaline membrane formation; the latter is a pathognomonic feature of respiratory distress syndrome. However, both thrombin inhibitors appeared to reduce the potential for intrapulmonary fibrin deposition equally well, because we observed similar effects of heparin and AT on the lung uptake of circulating 125I-fibrinogen and on the procoagulant activity of BAL fluid.
These findings raise the question whether the observed benefits of heparin were due to the well-characterized anticoagulant properties of the drug, or whether they were the result of nonanticoagulant activities. Increasingly, investigators have been exploring the therapeutic potential of heparin beyond its traditional use as an anticoagulant, with particular emphasis on its role as an inhibitor of inflammation (23). It has been suggested that the anti-inflammatory properties of heparin depend on the degree of sulfation of the heparin molecule (24), and that they are independent of its anticoagulant properties (24-26).
It is important to stress that the present studies were not designed to explore those anti-inflammatory actions of heparin; our objective was to examine the role of coagulation and its inhibition during acute lung injury. To our surprise, we found little evidence for the importance of anticoagulation; even the fact that formation of fibrin-containing hyaline membranes was reduced by heparin does not necessarily attest to an anticoagulant drug effect: recently, Rimensberger and coworkers have shown that experimental hyaline membrane formation can be attenuated by pretreatment with a specific inhibitor of neutrophil adhesion (27).
Both inflammation and coagulation are characteristic events during acute lung injury and therapeutic benefits might be expected from the modulation of either or both phenomena: A large body of experimental evidence suggests that neutrophils that migrate into the alveolar space during lung injury may further damage the lung tissues through release of oxyradicals, proteases such as elastase, leukotrienes, and platelet activating factor. Intrapulmonary thrombin formation and fibrin deposition may exacerbate pulmonary hypertension (8), endothelial permeability (28), surfactant dysfunction (10), lung inflammation (11), and fibrosis (12, 13). Despite this strong biologic rationale, controlled experiments in intact animals and human patients with acute lung injury have yielded conflicting evidence for the therapeutic efficacy of anticoagulant agents.
Heparin reduced lung water and prevented respiratory insufficiency during experimental glass bead embolism in dogs (28). This beneficial effect of heparin on lung fluid balance, however, was not confirmed by Binder and coworkers in sheep (29). On the other hand, heparin prevented human leukocyte elastase-induced acute lung injury in hamsters (30), and also improved oxygenation after smoke inhalation in sheep (15). Forty infants were enrolled in the only controlled clinical trial of heparin versus placebo in mechanically ventilated infants with respiratory distress syndrome (31). Interestingly, infants who received heparin required significantly shorter periods of artificial ventilation. However, this trial was conducted in the presurfactant era and had a number of methodologic limitations, including inadequate statistical power, and postrandomization withdrawals of patients from the primary analysis (31).
Hirudin, an AT-independent and selective inhibitor of thrombin, reduced protein-rich pulmonary edema and pulmonary vascular resistance in a porcine model of endotoxin-induced acute respiratory distress syndrome (32). Hirudin also decreased intrapulmonary accumulation of fibrinogen and procoagulant activity of BAL fluid during acute lung injury caused by pulmonary overdistension in newborn piglets (14).
Finally, purified concentrates of the serine protease antithrombin have been reported to prevent arterial hypoxemia in a sheep model of Escherichia coli–induced acute lung injury (33). We have previously shown that AT also reduces the lung uptake of circulating labeled fibrinogen and the procoagulant activity of BAL fluid in experimental lung injury caused by pulmonary overdistension (14). Although different procedures were used to induce acute lung injury in the current series of experiments, a similar effect of AT on those two outcomes was observed. In addition, we recovered exogenous human AT in the BAL fluid of those animals which received the concentrate; yet gas exchange was not improved by AT therapy. This lack of a clinically important benefit of AT concentrate is consistent with the results of a recently completed placebo-controlled trial of AT therapy in premature infants with respiratory distress syndrome: This trial showed conclusively that AT therapy does not improve gas exchange or reduce the duration of mechanical ventilation and oxygen therapy (34).
The prospects for the therapeutic efficacy of heparin and other sulfated glycosaminoglycans appear more promising; a growing body of evidence suggests that such efficacy may be derived from anti-inflammatory rather than anticoagulant properties (30, 35). In the latter studies, aerosolized rather than systemic heparin was used, a route of administration which may be preferable in critically ill subjects who are at high risk of bleeding side effects. Although the activated partial thromboplastin time was not prolonged in heparin-treated piglets, neonatologists and other critical care physicians would likely be reluctant to use 30 units/kg/h of intravenous heparin in their patients. The feasibility and efficacy of treatment with aerosolized heparin should therefore be examined in suitable animal models. Until such data are available we strongly discourage the use of heparin in newborn infants with acute lung injury.
The authors gratefully acknowledge the excellent technical assistance of Ms. Maryann dela Cruz, Ms. Patsy Vegh, Mr. Les Barry, and the statistical support of Mr. Gary Foster, MSc. Antithrombin concentrate (Kybernin) was a generous gift from Behringwerke AG, Germany.
This work was supported by a grant from the Medical Research Council of Canada: No. MT-12163 (B.S. and D.d.S.).
| 1. | Bachofen A., Weibel E. R.Alterations of the gas exchange apparatus in adult respiratory insufficiency associated with septicemia. Am. Rev. Respir. Dis.1161977589615 |
| 2. | Sitrin R. G., Brubaker P. G., Fantone J. C.Tissue fibrin deposition during acute lung injury in rabbits and its relationship to local expression of procoagulant and fibrinolytic activities. Am. Rev. Respir. Dis.1351987930936 |
| 3. | Idell S., James K. K., Levin E. G., Schwartz B. S., Manchanda N., Maunder R. J., Martin T. R., McLarty J., Fair D. S.Local abnormalities in coagulation and fibrinolytic pathways predispose to alveolar fibrin deposition in the adult respiratory distress syndrome. J. Clin. Invest.841989695705 |
| 4. | Viscardi R. M., Broderick K., Sun C. C., Yale-Loehr A. J., Hessamfar A., Taciak V., Burke K. C., Koenig K. B., Idell S.Disordered pathways of fibrin turnover in lung lavage of premature infants with respiratory distress syndrome. Am. Rev. Respir. Dis.1461992492499 |
| 5. | Carvalho A. C., DeMarinis S., Scott C. F., Silver L. D., Schmaier A. H., Colman R. W.Activation of the contact system of plasma proteolysis in the adult respiratory distress syndrome. J. Lab. Clin. Med.1121988270277 |
| 6. | Schmidt B., Vegh P., Weitz J., Johnston M., Caco C., Roberts R.Thrombin/antithrombin III complex formation in the neonatal respiratory distress syndrome. Am. Rev. Respir. Dis.1451992767770 |
| 7. | Brus F., van Oeveren W., Okken A., Oetomo S. B.Activation of the plasma clotting, fibrinolytic, and kinin–kallikrein system in preterm infants with severe idiopathic respiratory distress syndrome. Pediatr. Res.361994647653 |
| 8. | Seeger W., Neuhof H., Hall J., Roka L.Pulmonary vasoconstrictor response to soluble fibrin in isolated lungs: possible role of thromboxane generation. Circ. Res.621988651659 |
| 9. | Malik A. B., Horgan M. J.Mechanisms of thrombin-induced lung vascular injury and edema. Am. Rev. Respir. Dis.1361987467470 |
| 10. | Seeger W., Elssner A., Gunther A., Kramer H. J., Kalinowski H. O.Lung surfactant phospholipids associate with polymerizing fibrin: loss of surface activity. Am. J. Respir. Cell Mol. Biol.91993213220 |
| 11. | Garcia J. G., Perlman M. B., Ferro T. J., Johnson A., Jubiz W., Malik A. B.Inflammatory events after fibrin microembolization. Am. Rev. Respir. Dis.1371988630635 |
| 12. | Kuhn C., Boldt J., King T. E. J., Crouch E., Vartio T., McDonald J. A.An immunohistochemical study of architectural remodeling of connective tissue synthesis in pulmonary fibrosis. Am. Rev. Respir. Dis.140198916931703 |
| 13. | Tani K., Yasuoka S., Ogushi F., Asada K., Fujisawa K., Ozaki T., Sano N., Ogura T.Thrombin enhances lung fibroblast proliferation in bleomycin-induced pulmonary fibrosis. Am. J. Respir. Cell Mol. Biol.519913440 |
| 14. | Schmidt B., Davis P., La Pointe H., Monkman S., Coates G., deSa D.Thrombin inhibitors reduce intrapulmonary accumulation of fibrinogen and procoagulant activity of bronchoalveolar lavage fluid during acute lung injury induced by pulmonary overdistension in newborn piglets. Pediatr. Res.391996798804 |
| 15. | Cox C. S. J., Zwischenberger J. B., Traber D. L., Traber L. D., Haque A. K., Herndon D. N.Heparin improves oxygenation and minimizes barotrauma after severe smoke inhalation in an ovine model. Surg. Gynecol. Obstet.1761993339349 |
| 16. | Kirpalani H., Abubakar K., Nahmias C., deSa D., Coates G., Schmidt B.[18F]Fluorodeoxyglucose uptake in neonatal acute lung injury measured by positron emission tomography. Pediatr. Res.411997892896 |
| 17. | O'Brodovich H., Berry L., D'Costa M., Burrows R., Andrew M.Influence of fetal pulmonary epithelium on thrombin activity. Am. J. Physiol.2611991L262L270 |
| 18. | Towbin H., Gordon T.Immunoblotting and dot immunobinding—current status and outlook. J. Immunol. Methods721984313340 |
| 19. | Notter R. H., Egan E. A., Kwong M. S., Holm B. A., Shapiro D. L.Lung surfactant replacement in premature lambs with extracted lipids from bovine lung lavage: effects of dose, dispersion technique, and gestational age. Pediatr. Res.191985569577 |
| 20. | Tsuno K., Miura K., Takeya M., Kolobow T., Morioka T.Histopathologic pulmonary changes from mechanical ventilation at high peak airway pressures. Am. Rev. Respir. Dis.143199111151120 |
| 21. | Winer, B. J., D. R. Brown, and K. M. Michels. 1991. Statistical Principles in Experimental Design, 3rd ed. McGraw-Hill, New York. 497–580. |
| 22. | Dixon, W. J., editor. 1988. BMDP Statistical Software Manual. University of California Press, Berkeley. 1081–1114. |
| 23. | Lane D. A., Adams L.Non-anticoagulant uses of heparin. N. Engl. J. Med.3291993129130 |
| 24. | Matzner Y., Marx G., Drexler R., Eldor A.The inhibitory effect of heparin and related glycosaminoglycans on neutrophil chemotaxis. Thromb. Haemost.521984134137 |
| 25. | Bazzoni G., Beltran-Nuñez A. B., Mascellani G., Bianchini P., Dejana E., del Maschio A.Effect of heparin, dermatan sulfate, and related oligo-derivatives on human polymorphonuclear leukocyte functions. J. Lab. Clin. Med.1211993268275 |
| 26. | Nelson R. M., Cecconi O., Roberts W. G., Aruffo A., Linhardt R. J., Bevilacqua M. P.Heparin oligosaccharides bind L- and P-selectin and inhibit acute inflammation. Blood82199332533258 |
| 27. | Rimensberger P. C., Fedorko L., Cutz E., Bohn D. J.Attenuation of ventilator-induced acute lung injury in an animal model by inhibition of neutrophil adhesion by leumedins (NPC 15669). Crit. Care Med.261998548555 |
| 28. | Malik A. B., van der Zee H.Time course of pulmonary vascular response to microembolization. J. Appl. Physiol.4319775158 |
| 29. | Binder A. S., Nakahara K., Ohkuda K., Kageler W., Staub N. C.Effect of heparin or fibrinogen depletion on lung fluid balance in sheep after emboli. J. Appl. Physiol.471979213219 |
| 30. | Rao N. V., Kennedy T. P., Rao G., Ky N., Hoidal J. R.Sulfated polysaccharides prevent human leukocyte elastase-induced acute lung injury and emphysema in hamsters. Am. Rev. Respir. Dis.1421990407412 |
| 31. | Göbel U., von Voss H., Jürgens H., Petrich C., Pothmann R., Sprock I., Lemburg P.Efficiency of heparin in the treatment of newborn infants with respiratory distress syndrome and disseminated intravascular coagulation. Eur. J. Pediatr.13319804749 |
| 32. | Hoffman H., Siebeck M., Spannagl M., Weis M., Geiger R., Jochum M., Fritz H.Effect of recombinant hirudin, a specific inhibitor of thrombin, on endotoxin-induced intravascular coagulation and acute lung injury in pigs. Am. Rev. Respir. Dis.1421990782788 |
| 33. | Emerson T. E. J., Fournel M. A., Redens T. B., Taylor F. B. J.Efficacy of antithrombin III supplementation in animal models of fulminant Escherichia coli endotoxemia or bacteremia. Am. J. Med.87198927S33S |
| 34. | Schmidt B., Gillie P., Mitchell L., Andrew M., Caco C., Roberts R.A placebo-controlled randomized trial of antithrombin therapy in neonatal respiratory distress syndrome. Am. J. Respir. Crit. Care Med.1581998470476 |
| 35. | Ahmed T., Garrigo J., Danta I.Preventing bronchoconstriction in exercise-induced asthma with inhaled heparin. N. Engl. J. Med.32919939095 |
Dr. B. Schmidt is a Career Investigator of the Heart and Stroke Foundation, Ontario, Canada.
Presented in part at the Annual Meeting of the Society for Pediatric Research in Washington, DC, 1997.
