INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: polymer; surfactant; lung injury

Some forms of acute lung injury (ALI) involve inactivation of pulmonary surfactant where inactivation is defined as the reduction of surface activity when a substance is added to surfactant (1). These types of lung injury may be amenable to surfactant treatment if surfactants that are resistant to inactivation can be developed. A variety of surfactant proteins (SP-A, SP-B, and SP-C) together or singly have been found to reduce inactivation of surfactant caused by a number of substances (4). We have found that addition of nonionic polymers (polyethylene glycol [PEG], polyvinylpyrrolidone, or dextran) prevents and reverses some forms of surfactant inactivation in vitro (8). When polymer/surfactant mixtures were used to treat lung injury in rats caused by intratracheal instillation of meconium, larger improvements in serial measures of oxygenation and lung compliance and improvement in static lung mechanics were found than with surfactant alone (9). In the current study, we have used two additional types of lung injury (lipopolysaccharide [LPS] and hydrochloric acid [HCl]) that more closely resemble acute lung injury/acute respiratory distress syndrome (ALI/ARDS) to determine whether the same differences in response are evident.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Experiments were designed to find whether addition of PEG to Survanta improved the response to treatment after lung injury caused by either HCl or LPS.

Materials

The substances used for instillation were obtained and prepared as follows. HCl (100 mM H⁺, Fisher Scientific, Pittsburgh, PA) was diluted to a final concentration of 78 mM H⁺ (pH 1.12) in 0.9% saline. Escherichia coli LPS (serotype 0111:B4 Lot 39H4103; Sigma, St. Louis, MO) was dissolved in sterile 0.9% saline at a concentration of 10 mg/ml.

Survanta (Ross Laboratories, Columbus, OH) was diluted with 0.9% saline or with distilled water to a concentration of 10 to 12.5 mg/ml, then mixed by gentle repeated inversion. The choice of Survanta in this study was based upon its beneficial effects in ARDS in a clinical study (10) and the fact that it was one of two surfactants available for human use in the United States at the time of these studies. Dry PEG (10 kDa; Sigma, St. Louis, MO) was added to the diluted surfactant in a concentration of 5% (wt/vol), mixed by Vortex, and used within 1 h.

Rat Experiments

Rats were handled in accord with guidelines promulgated by the American Physiological Society. Protocols were approved by the University of California-San Francisco committee on animal research. Male Sprague-Dawley rats of 250 to 350 g were given intraperitoneal anesthesia of 60 to 100 mg/kg pentobarbital, then tracheostomized. A carotid artery catheter was used for blood sampling, blood pressure measurements, and for administering pancuronium (1 mg/kg) and pentobarbital. Flushes of 10 units of heparin per milliliter of saline were used to maintain patency of the catheter. Over the course of the 3-h experiments approximately 2 ml (7 ml/kg) of blood was removed and 10 ml of saline/drug (30 ml/kg) was administered. Blood pressures and tracheal pressures were measured continuously using Viggo-Spectromed transducers (model P23XL; Gould) attached to a recorder (Gould WindoGraf; Gould, Valley View, OH), after calibration using mercury and water manometers. Baseline measurements were recorded over 30 min. During the experiments, repeat intra-arterial doses of pentobarbital (7 mg/kg) were given when systolic blood pressures rose above 125 mm Hg. Doses of pancuronium (0.5 mg/kg) were administered every hour.

The protocol used for HCl lung injury was adapted from Chiumello and colleagues (11). Animals were placed on a Harvard volume-regulated ventilator with the following settings: frequency (f) = 40/min; tidal volume (VT) = 7 ml/kg; positive end-expiratory pressure (PEEP) = 4 cm H₂O; fraction of inspired oxygen (FI_O2) = 1.0 at a flow rate of 0.5 L/min. After animals were in stable condition, HCl, 3.5 ml/kg (pH 1.12), was instilled into the endotracheal tube using a miniature nebulizer (Penn-Century, Philadelphia, PA) that consisted of a 1-ml syringe in combination with a nozzle mounted on the tip of a 5.5-cm-long stainless steel tube. This technique, adapted from van Helden and coworkers, allows a grossly diffuse lung injury (12). HCl was given by rapid aerosol in two bursts with the rat first on one side, then on the other side for the second burst. Each aerosol burst was followed by 5 ml of air. Immediately after the instillation, the rat was shaken vigorously to distribute the acid (11). Furosemide (10 mg/kg) was given intra-arterially 5 min after HCl instillation. Fifteen minutes after the first HCl instillation, a repeat instillation of HCl (2.5 ml/kg) was given, making the total volume of HCl used 6 ml/kg. Fluid, when apparent, was removed by suction from the trachea. Five minutes after the second administration of HCl, surfactant treatment of 4 ml/kg of either Survanta or Survanta + PEG was given to the rats through the tracheal tube. Arterial blood gases, peak inspiratory pressures (peak PI), and systolic blood pressures were recorded at 30 min, 1, 2, and 3 h after treatment. At the conclusion of the experiment, the rat was killed by pentobarbital injection, and pressure-volume determinations were carried out. Samples for determining wet and dry lung weights and histology were taken as described subsequently.

The protocol for LPS injury was adapted from the work of Tashiro and coworkers (13, 14) and from van Helden and coworkers (12). Rats were prepared similarly to those receiving HCl injury. Ventilator settings differed, however, with f = 30/min; peak PI = 25 cm H₂O; PEEP = 5 to 7.5 cm H₂O; and FI_O2 = 1.0 at a flow rate of 0.5 L/min. Therefore, peak PI was not used as a dependent variable in the LPS experiments. Forty mg/kg LPS was aerosolized (as described previously for HCl injury but no air was given and the rats were not shaken) into the lungs at a concentration of 10 mg LPS/ml 0.9% saline. Another one or two doses of 20 mg/kg LPS were administered if Pa_O2 had not fallen below 400 mm Hg by 30 min after the initial instillation. After the first dose of LPS, peak PI was maintained at 20 cm H₂O by adjusting the volume-regulated ventilator, and PEEP was set to zero. When tracheal fluid became apparent, it was aspirated through a fine catheter (13, 14). This ventilation strategy was different from that used for HCl injury and obviates the use of peak PI as a dependent variable. Diuretic was not used after LPS injury. When Pa_O2 decreased below 200 mm Hg, surfactant treatment was given and peak PI was increased to 25 and PEEP to 5 to 7.5 cm H₂O. (Initial experiments were done using PEEP of 7.5 cm H₂O, and subsequent experiments with 5 cm H₂O. Equal numbers of animals in each treatment group received each PEEP level, and no differences were noted within experimental groups receiving higher versus lower PEEP.) Arterial blood gases and systolic blood pressures were recorded at 30 min, 1, 2, and 3 h after treatment.

Deflation Pressure-Volume Determinations

Three hours after treatment, the animal was killed by injection of pentobarbital with the trachea clamped, thereby degassing the lungs. The abdomen and diaphragm were opened and the vena cava cut. Postmortem quasi-static deflation pressure-volume measurements were carried out by first inflating the lungs to a pressure of 35 cm H₂O. After a period of stabilization (approximately 45 s or when < 0.1 ml of air entered in 10 s), the volume was measured to define total lung capacity. Pressure was then reduced in steps of 5 cm H₂O with at least a 10-s stabilization period at each step, and the corresponding volumes measured. Volumes were corrected for compression.

Lung Weights

The lungs were removed, blotted, and weighed. A whole midsagittal slice of right lung was weighed, then placed in an oven at 50° C for 24 h and reweighed to determine the dry/wet lung weight ratios.

Histology

An additional sagittal section of the left lung was placed in 4% formalin, embedded in paraffin, and stained with hematoxylin-eosin. The sections were examined by one of us (B.R.) with particular reference to airway epithelial necrosis, intra-alveolar edema, hyaline membranes, hemorrhage, and recruitment of inflammatory cells to the air spaces. The various changes were graded according to an arbitrary four-grade scale (absent, mild, moderate, or prominent) (9). The examination was performed on coded sections, i.e. without knowledge of the experimental conditions of individual animals.

Measurements of Surface Activity

In vitro measurements for surface activity were made in a pulsating bubble surfactometer (Electronetics, Amherst, NY). A simple modification minimized errors that could have been introduced by wetting of the capillary tube in the sample chamber (15). The surface tension was estimated by the Laplace formula (P = 2ST/r where ST is the surface tension, P the inflating pressure, and r the bubble radius) assuming a spherical bubble shape. Sample chambers were maintained at 37° C throughout the measurements by means of a temperature-controlled water jacket. The device was calibrated by a water manometer and checked by the use of pure fluids with known surface tensions. The bubble was inflated and deflated at a rate of 20 cycles/min. Measures of surface activity were the surface tension 20 s after initial formation of a bubble of standard diameter (adsorption), and the minimal surface tension at the end of the tenth bubble deflation. Measures of surface activity were carried out in the presence of the supernatant from centrifuged (2,000 g_av for 5 min) tracheal fluid from rats injured with HCl and sham-treated with saline (group 3 in Table 1). The tracheal fluid had a concentration of approximately 120 mg protein/ml, and 1.8 mg of protein (15 µl of tracheal fluid) was added to 1 ml of the 1.25 mg/ml Survanta or Survanta + PEG mixtures.

Analysis

Data are expressed as means ± SEM. Serial measurements of arterial blood gases, PI, and pressure-volume curves were analyzed by two-way analysis of variance (ANOVA) modified for repeated measures using SigmaStat software (SPSS; SPSS Science, Chicago, IL). Group differences were analyzed by a two-tailed t test. A p < 0.05 was considered statistically significant and the Bonferroni correction was used where indicated for repeated t tests (16). Histologic differences among groups were evaluated with the chi-square test with Yates' correction.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physiologic Status of Animals After Injury

General characteristics of the five groups of animals after injury with either HCl or LPS are shown in Table 1 and Figure 1. Severe deficits in oxygenation were achieved with both lung injury models, with Pa_O2 < 150 mm Hg despite ventilation with 100% O₂. Before treatment, no significant differences in average blood pressure, degree of acidosis, and oxygenation were found between rats injured with HCl or with LPS. Also, within each injury group, pretreatment measures of those subsequently treated with Survanta or with Survanta + PEG were similar.

View larger version (18K): [in this window] [in a new window]   Figure 1.   Serial measures of carotid artery PaO2 for LPS- and HCl- injured rats are shown. Panel A shows results for groups injured with LPS and then treated with Survanta 50 mg/kg (n = 6) or with Survanta 50 mg/kg and PEG 5% (n = 6). The group treated with Survanta + PEG has oxygenation that is greater than the group treated with Survanta alone (p < 0.02). Panel B shows results for animals injured with HC1 and then treated with Survanta 40 mg/kg (n = 6) or with Survanta 40 mg/kg and PEG 5% (n = 6). The group treated with Survanta + PEG has significantly higher values for PaO2 when compared with animals treated with Survanta (p = 0.002) or with saline (p = 0.001).

Three hours after treatment, five of the 30 animals had died. Deaths were limited to those injured with LPS or to the sham-treated group injured with HCl. On average these three groups (1, 2, 3 in Table 1) were the most ill after treatment as indicated by average values for base deficit and blood pressure.

LPS Injury

Results of experiments with LPS injury are shown in Figures 1A and 2A. After LPS injury, the Survanta + PEG treatment group showed values for Pa_O2 that were greater than those found for the Survanta alone group (p < 0.02) (Figure 1A). Also, quasi-static measures of lung volume during deflation were increased in the Survanta + PEG group compared with those treated with Survanta alone (p = 0.04). No significant differences between treatment groups were found for measures of wet or dry lung weights. Three hours after treatment, the surviving rats treated with Survanta alone tended to have lower arterial blood pH (p < 0.002) (Table 1).

View larger version (19K): [in this window] [in a new window]   Figure 2.   Average deflation pressure-volume relationships for LPS (A) and HCl (B) injuries are shown. Average lung volume for the group treated with Survanta (50 mg/kg) and PEG is significantly higher than that for the group treated with Survanta (50 mg/kg) alone after LPS injury (p = 0.04). For HCl injury, average lung volumes for the group treated with Survanta (40 mg/kg) and PEG are significantly higher than those for the group treated with Survanta (40 mg/kg) alone (p = 0.04).

HCl Injury

In an initial set of experiments, animals injured with HCl were treated with doses of 50 mg/kg of Survanta ± PEG. We found no significant differences between the two treatment groups for serial measures of Pa_O2 or peak PI, or for pressure-volume relationships. We did observe that increased fluid could be obtained by suction from the tracheae of those treated with Survanta + PEG in the hours after treatment compared with those treated with Survanta alone. We hypothesized that the lack of difference between the two treatment groups might be associated with increased pulmonary fluid owing to the osmotic effects of PEG. Therefore, we carried out another group of experiments using Survanta (50 mg/kg) ± PEG in 0.45% saline. Tracheal fluid volume collected over the first h after treatment was 2.0 ± 0.2 ml for the group treated with Survanta + PEG in 0.9% saline; and 1.4 ± 0.2 ml for the group treated with Survanta + PEG in 0.45% saline (p = 0.04). For those treated with Survanta alone in 0.9% saline, tracheal fluid volume was 0.8 ± 0.3 ml compared with none for those treated with Survanta in 0.45% saline. Significant differences were evident between treatment groups (Survanta versus Survanta + PEG, in either 0.9% or 0.45% saline) when Pa_O2 and peak PI were measured at 1 h, but these differences were no longer apparent after 3 h. We believed this was due to the good oxygenation (Pa_O2 > 200 mm Hg) (data not shown) in both the Survanta and the Survanta + PEG treatment groups. Based on these results, we hypothesized that the differences between Survanta and Survanta + PEG might become more evident after treatment with lower doses of surfactant.

Therefore, we conducted another series of experiments using 40 mg/kg Survanta ± PEG diluted to a final concentration of 0.36% saline. Results after treatment showed that the average Pa_O2 was significantly higher in the Survanta + PEG group compared with the Survanta alone group (p = 0.002) or animals that were sham-treated with saline (p = 0.001) (Figure 1B). However, unlike animals injured with LPS, Pa_O2 responses after either treatment in animals injured with HCl were not sustained (Figure 1B). For serial PI measurements, the mean values for the group treated with Survanta + PEG remained significantly lower than the groups treated with Survanta alone (p = 0.03) or sham-treated animals (p = 0.001) as shown in Figure 3. Measurement of quasi-static pressure- volume relationships (Figure 2B) showed significantly higher average lung volumes for the group treated with Survanta + PEG compared with those treated with Survanta alone (p = 0.04) or compared with the sham saline treatment group (p = 0.01). There were no significant differences between treatment groups for measures of wet or dry lung weights.

View larger version (16K): [in this window] [in a new window]   Figure 3.   Serial measures of peak PI after HCl injury are shown. The group treated with Survanta 40 mg/kg with PEG has peak PI that are lower than the group treated with Survanta 40 mg/kg without PEG (p = 0.03).

Pulsating Bubble Surfactometer Measurements

Tracheal fluids from animals injured with HCl were used to study whether differences in resistance to inactivation were evident between Survanta versus Survanta + PEG. PEG added to Survanta improved both adsorption and minimal surface tension. Adsorption at 20 s was inhibited when tracheal fluid was added to Survanta (35 ± 3 mN/m), whereas adsorption of Survanta + PEG in the presence of tracheal fluid was less inhibited (25 ± 0.8 mN/m, p < 0.01) (Figure 4). With added tracheal fluid, the minimal surface tension averaged 23 ± 3 mN/m for Survanta and 0.6 ± 0.4 mN/m for Survanta + PEG (p < 0.0001) (Figure 4).

View larger version (24K): [in this window] [in a new window]   Figure 4.   Average adsorption at 20 s and minimal surface tension at the tenth cycle are shown for mixtures of 1.25 mg/ml Survanta ± PEG and tracheal fluid from animals injured with HCl. The group treated with Survanta + PEG achieved values for adsorption and dynamic surface tension that were significantly (p < 0.01 for both cases) lower than those for Survanta without PEG.

Histology

Histologic findings are shown in Table 2. Evidence of lung injury of varying degree was found in all but one of the animals examined. However, the type of lesions varied depending mainly on whether the animals had been exposed to HCl or LPS. Airway epithelial necrosis, characterized by poor staining of the epithelium or pyknosis, and usually associated with desquamation, was observed in 11 of the 18 animals exposed to HCl and in only one of the animals exposed to LPS. In contrast, hyaline membranes, interstitial and intra-alveolar hemorrhage, and recruitment of inflammatory cells (mainly neutrophils) were significantly more common and prominent in animals exposed to LPS. Some of these animals showed extensive hemorrhagic-inflammatory lung injury. Figure 5 illustrates the histologic differences that were apparent between these two types of lung injury. No significant differences between those treated with Survanta or Survanta + PEG were evident.

View larger version (129K): [in this window] [in a new window]   View larger version (142K): [in this window] [in a new window]   Figure 5.   Representative microphotographs showing (A) extensive loss of nuclear staining in airway epithelium of animal exposed to HCl and (B) hemorrhagic inflammatory lung injury after exposure to LPS. In both panels, bronchioles are indicated with asterisks. Hematoxylin-eosin; original magnification: ×165.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In this study, the addition of polymer (PEG) to Survanta improved oxygenation and mechanical properties of lungs after two types of lung injury: LPS and HCl. PEG added to Survanta also improved resistance to inactivation when tested in a pulsating bubble surfactometer. Histologic changes in the lungs were different between the two types of lung injury, with damage to airway mucosa dominating in animals receiving HCl, and hemorrhagic inflammation in peripheral parenchyma prevailing after exposure to LPS, but differences between treatment groups were not apparent.

Various animal models of lung injury have been reported in the literature. These vary in relevance to human disease, with no animal model of ARDS being completely satisfactory. We chose these two models to study because as much as 30% of ALI/ARDS is associated with aspiration of gastric contents, and 40% with sepsis (17). Also, acid and LPS lung injuries in adult rats offer models that have been useful for measures of physiologic responses to surfactant treatment, inflammation, degree of capillary leak, and histology (18, 19).

LPS lung injury has been studied extensively by Kennedy and coworkers. They found that ALI caused by LPS was associated with surfactant dysfunction, alterations in surfactant composition, and inflammation (18). Tashiro and colleagues found that surfactant replacement improved lung function significantly after LPS injury (13, 14). Limitations of the LPS lung injury model include the high tolerance and the need for large doses of intratracheal LPS in rats to create severe hypoxemia. Mechanisms of resistance to LPS lung injury and normal mechanisms of LPS clearance in the lung have recently been reviewed (20). Also, in our experience the LPS injury model is inconsistent, with some rats deteriorating rapidly and others showing minimal lung dysfunction with the same doses of LPS. Hypotension and acidemia may occur at times without much lung impairment. The degree of hemorrhagic inflammation varied considerably in animals exposed to LPS, irrespective of whether they had been treated with Survanta + PEG or Survanta alone.

Acid lung injury is also an extensively studied model of ARDS. This type of injury (by aspiration of gastric contents) causes ARDS that is associated with diminished gas exchange, low compliance, and increased proinflammatory cytokines (21). The mechanism of respiratory failure after HCl aspiration may be caused by several factors, including damage to the airways and alveolar-capillary membranes, direct damage to endogenous surfactant, or secondary inactivation of surfactant activity caused by capillary-alveolar leakage of plasma (22). Hypotension and acidemia were less evident with HCl injury in our study than with LPS injury. One limitation of the acid injury model is that few investigators find it to be surfactant responsive, and when it is, only if surfactant is given within minutes of the injury as we have done. In our experiments with HCl injury, we used a diuretic before treatment following the model of Sun and associates (23). Unlike the case of LPS injury, improvements in oxygenation in the HCl model were not sustained over 3 h.

These results agree with and extend earlier in vitro and in vivo work from our laboratory in which inactivation is prevented or reversed by addition of PEG, dextrans, SP-A, or polyvinylpyrrolidone to Survanta (8). We found that Survanta + PEG improved lung function after meconium lung injury in rats, with results comparable to those found here after HCl or LPS injury (9). Similar findings have been reported by Kobayashi and colleagues using dextran added to modified surfactant both in vitro and in vivo (24, 25).

Mechanisms for the effects of PEG and other nonionic polymers on lung surfactants remain largely unstudied. With much simpler mixtures of phospholipids, polymers, including PEG, are known to induce aggregation and growth of vesicle size. Depletion forces play the predominant role in this process (26). In preliminary work, we also have found that PEG added to Survanta increases apparent lipid aggregation and improves surface activity of Survanta both in the presence and in the absence of inhibitors (8, 29). We hypothesize that one reason for the beneficial effect of PEG may be its ability to separate inactivating substances from surfactant lipids. Is the polymer effect specific to Survanta? Our preliminary in vitro work with other surfactants shows that PEG has the greatest effect with Survanta and that dextran has the greatest with Curosurf in the presence of inhibitor (30).

In summary, increasing evidence indicates that surfactants vary in susceptibility to inactivation. Surfactants that resist inactivation improve lung function in animal models of ALI to a greater degree than surfactants that are less resistant. If surfactant inactivation is important in causing lung dysfunction in some kinds of ALI and ARDS, then inactivation-resistant surfactants could improve outcomes in these diseases.