INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Although exogenous administration of surfactant is now a standard therapeutic modality for preterm infants with the neonatal respiratory distress syndrome (1, 2), this treatment is only beginning to be tested in patients with acute respiratory distress syndrome (ARDS) (3). Unfortunately, clinical responses of patients with ARDS to exogenously administered surfactant have been variable (3, 5). Two potential factors that may influence a host's response to exogenous surfactant include the particular surfactant preparation utilized and the method used to deliver the surfactant into the lungs (6). For example, in one of two randomized controlled clinical trials conducted to date, a tracheally instilled natural surfactant preparation (Survanta; Ross Laboratories, Columbus, OH) improved gas exchange and patient survival (5), whereas in the other clinical trial, an aerosolized synthetic product (Exosurf; Glaxo Wellcome, Research Triangle Park, NC) was ineffective (3). Although utilizing natural products would seem reasonable on the basis of these results, the adult patient population that would potentially benefit from exogenous surfactant would require significant quantities of material. This could eventually result in resource limitations. This issue, together with the current expense of exogenous surfactant preparations, provides a good rationale for pursuing the development of synthetic surfactant preparations for ARDS.

Recently, a surfactant based on a recombinant surfactant associated protein-C (rSP-C) was tested in small animals with promising results (7). Before being investigated in clinical trials involving adult patients, however, such a surfactant must be subjected to further preclinical studies involving larger animals in order to address specific factors such as the dose and delivery methods that one would use in these types of patients. The purpose of this study was to test the safety and efficacy of a novel exogenous surfactant preparation composed of a recombinant SP-C protein combined with lipids, which is currently being considered for clinical trials in patients with ARDS. We evaluated three different doses of this surfactant preparation and compared two different delivery methods. One delivery method involved administering the surfactant to each lobe of the lung via a bronchoscope (4), and the other involved administering the surfactant through an endotracheal tube (ET) directly into the lungs (5).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Induction of Lung Injury

Adult sheep (43 ± 1 kg) were anesthetized with 25 mg/kg of ketamine (100 mg/ml) given intramuscularly, and lidocaine (20 mg/ml) was infiltrated subcutaneously at all surgical sites. Vascular access was obtained by inserting femoral artery and internal jugular venous catheters. Arterial blood pressure and heart rate were monitored continuously, and 0.15 M NaCl was infused through the venous catheter at a rate of 5 ml/kg/h throughout the experiment. A tracheotomy was performed and an ET was tied into place. All animals were ventilated with a Servo 900B volume-cycled ventilator (Siemans-Elema, Solna, Sweden). The following parameters were utilized throughout the experimental period; an inspired oxygen fraction (FI_O2) of 1.0, a respiratory rate of 16/min, an inspiratory-time fraction of 25%, a positive end expiratory pressure (PEEP) of 5 cm H₂O, and a tidal volume (VT) of 10 ml/kg.

Lung injury was induced by repetitive, whole-lung saline lavage, as previously described (11, 12). Briefly, the animal was disconnected from the ventilator circuit and 40 ml/kg of warmed (37 °C) 0.15 M sterile NaCl was instilled through the ET tube. After 30 to 45 s the saline was passively drained, with a recovery of each lavage volume of approximately 90 to 95%. This procedure was repeated every 10 min until the Pa_O2 fell below 100 mm Hg. The first two lavages were performed with the animal in the supine position, and subsequent lavages were done with the animal in the prone position to optimize the distribution of injury. A minimum of four lavages were done on each animal, although approximately six to eight washes were usually required.

All animals were ventilated for 60 min after the final lavage, during which time a severe inflammatory component compounded the lung injury, owing to the surfactant deficiency (11). To ensure consistent injury among experimental groups, only those animals with Pa_O2 values between 70 and 120 mm Hg at 60 min after the last previous lavage were subsequently used for evaluating responses to exogenous surfactant. Anesthesia was maintained throughout the experiment via intermittent intravenous infusions of thiopental sodium (15 mg/kg/ dose) and intramuscular ketamine (10 mg/kg/dose).

Surfactant Preparation

The exogenous surfactant preparation used for these experiments was a synthetic product made from a rSP-C protein and lipids. The composition of the material was 2% rSP-C protein in phospholipids (PL) (dipalmitoylphosphatidylcholine [DPPC] and palmitoyloleoylphosphatidylglycerol in a 70:30 [wt/wt] ratio), and 5% palmitic acid. Radiolabeled rSP-C surfactant was prepared to measure the distribution of the surfactant in lung tissue. ¹⁴C-palmitic acid-labeled DPPC with an activity of 15 µCi was dried with 1 mg unlabeled DPPC onto the bottom of a glass tube. The appropriate dose of unlabeled rSP-C surfactant was then added to this tube and the combination was resuspended with glass beads. Unlabeled surfactant was supplied as a lyophylized powder and was suspended in 0.15 M NaCl to the appropriate concentration for treatment.

Surfactant Treatment Groups

Tracheal instillation. The first series of experiments (Phase I) evaluated different doses of the rSP-C surfactant (25, 100, and 200 mg phospholipid/kg body weight) delivered via instillation through the ET tube. Because all three doses were administered in equal volumes (4 ml/kg), the per-dose concentrations of phospholipid differed. The 25 mg/kg dose was delivered at a concentration of 6.25 mg PL/ml, the 100 mg dose at 25 mg PL/ml, and the 200 mg/kg dose at 50 mg PL/ml. For each animal, surfactant was administered according to the same technique that has previously been utilized in this laboratory (11, 12). Briefly, the animal was positioned on the left side at a 45-degree incline, and one-third of the dose of phospholipid to be delivered was instilled in aliquots during the inspiratory phase of ventilation through a size 8 Fr catheter. The catheter was inserted through a side port of the ET tube and positioned just beyond the end of the ET tube in the trachea. The same procedure was performed with the animal positioned on the right side and then in the prone position for the remaining surfactant aliquots. The control group utilized for these series of experiments underwent the same lavage and ventilation procedures as the other animals, but was not given exogenous surfactant.

Bronchoscopic instillation. In a second series of experiments (Phase II), a total dose of 25 mg phospholipid/kg body weight (4 ml/ kg) was delivered via a bronchoscope. In these animals, a bronchoscope was inserted through the side-port adaptor of the ET tube, and aliquots of surfactant were instilled into each lobar bronchus under visual guidance. The calculated volume delivered to each lobe was based on the contribution of each lobe to the whole lung, which was determined from the total protein content and weight of the lobe as reported in previous studies (11, 12). During administration into each lobe, the surfactant was instilled in small aliquots so that the observed reflux up the lobar orifice was minimal. These animals remained in the prone position throughout the administration procedure, and the total time required for administration was recorded. The two other comparison groups studied in this series of experiments included a group given the same surfactant dose of 25 mg/kg (4 ml/kg) via direct tracheal instillation according to a technique identical to that described earlier, and a control group that was not given surfactant but had a bronchoscope inserted into each lobar orifice with only air administered.

Physiologic Measurements

Arterial blood gases including Pa_O2 and Pa_CO2 measurements, as well as pH measurements, were recorded during the induction of lung injury, immediately before treatment, and every 15 to 30 min thereafter. Peak inspiratory pressures (PIP) were also monitored with a Pneumogard pressure transducer (Novometrix Medical Systems, Wallingford, CT) situated at the proximal end of the ET tube and the end of the ventilator circuit. Heart rate and arterial blood pressure were also monitored continuously, and all animals were killed with an overdose of euthanol at 4 h after the end of the surfactant treatment procedure.

Lung Lavage Analysis

After the animals were killed, their lungs were removed from the chest cavity, and whole-lung saline lavage was performed for analysis as previously described (12, 13). A total of three lavage procedures were performed with 0.15 NaCl each, and the total recovered volume was combined and recorded. Aliquots were taken for lipid extraction with chloroform:methanol, and total PL-phosphorus was measured according to the method of Duck-Chong (14). Aliquots were also taken for isolation of surfactant aggregates. Briefly, samples were centrifuged at 150 × g for 10 min to remove cell debris, and the supernatant was spun at 40,000 × g for 15 min (13). The resultant pellet was resuspended, and represented the large surfactant aggregates (LA), whereas the supernatant represented the small surfactant aggregates (SA). Aliquots of both subfractions, as well as the resuspended 150 × g pellet, were analyzed for PL-phosphorus. The percentage of total PL-phosphorus recovered in LA forms (% LA) was also calculated.

Surfactant Distribution

The remaining lung tissue of each lobe was homogenized in saline and aliquots were taken for protein measurements as well as for lipid extraction and radiolabel recovery measurements. The total amount of ¹⁴C-radioactivity recovered in the lung tissue from each lobe was divided by the total recovery of ¹⁴C in the tissue of the whole lung. The radiolabel recovery in the different lobes was standardized by correcting the measured recovery values for the protein values measured for each lobe.

Statistics

All values are presented as mean ± SEM. Comparisons of physiologic data measured over time among groups were made through analysis of variance (ANOVA) for repeated measures, with treatment and time serving as the main effects. A Tukey's honestly significant difference (HSD) test was used for post hoc analysis (p < 0.05). For single between-group comparisons of means, ANOVA with a Tukey's HSD post hoc test was utilized (p < 0.05). Comparisons of values at each time point after treatment with pretreatment or baseline values (time = 0 min) within each group were made with a paired t test.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

A total of 33 animals underwent the initial lavage procedure, although three animals were excluded on the basis of predetermined Pa_O2 inclusion values of 70 to 120 mm Hg after the final lavage and 60 min of ventilation. There were no significant differences among the groups studied in the number of lavages used to induce lung injury or in the amount of saline recovered after each lavage (data not shown). The total number of animals studied per group and the mean Pa_O2, Pa_CO2, and PIP values for animals included in the first series of experiments (Phase I), evaluating the different doses of surfactant, are shown in Table 1. There were no significant differences in these physiologic parameters among groups after the induction of lung injury at 1 h after the final lavage just prior to treatment. There were also no significant differences in hemodynamic variables or in the mean body weights of animals among these groups (data not shown). Similarly, there were no significant differences between these physiologic variables for animals subsequently receiving surfactant for comparison of the two delivery techniques (Phase II, Table 1).

Physiologic Responses

Mean sequential Pa_O2 values for each group during the 4-h period of ventilation after treatment at time = 0 min are shown in Figure 1. Figure 1A shows the dose-response relationship for the rSP-C surfactant preparation. Animals treated with the 100 and 200 mg/kg doses had significant increases in their Pa_O2 values over time, starting at 15 min after treatment, as compared with their respective pretreatment values (p < 0.05). There were no significant differences between these two groups' Pa_O2 values at any time point. Animals given the 25 mg/kg dose also had a significant increase in Pa_O2 values as compared with pretreatment values, starting at 150 min after treatment (p < 0.05). However, Pa_O2 values in this group were significantly lower than those in both the 100 and 200 mg/kg groups from 30 to 240 min after treatment (p < 0.05). Pa_O2 values for all three treatment groups were greater than those for control animals starting at 30 min after treatment (p < 0.05).

View larger version (24K): [in this window] [in a new window]   Figure 1.   Mean PaO2 values of animals treated with rSP-C surfactant. (A) All three doses of the surfactant significantly improved PaO2 values during the monitoring period (p < 0.05 for all three doses versus their respective pretreatment values), although the 100 and 200 mg/kg doses were superior in this regard (p < 0.01 for 100 and 200 mg/kg versus 25 mg/kg and control). There were no significant differences between the 100 and 200 mg/kg doses. (B) Both the bronchoscopic and tracheal instillation techniques for surfactant administration significantly improved PaO2 values as compared with their respective pretreatment values and control values (p < 0.05), although there were no significant differences between the two delivery methods.

Figure 1B shows the mean Pa_O2 values for both the bronchoscopically instilled surfactant treatment group and the tracheally instilled surfactant group given doses of 25 mg/kg. Both surfactant treatment groups had significant increases in Pa_O2 values as compared with their respective pretreatment values, and in both groups these values were greater than control values starting at time = 15 min (p < 0.05). There were no significant differences between these two surfactant treatment groups.

Figure 2 shows the mean Pa_CO2 values for the various groups. For animals given the 100 and 200 mg/kg doses, Pa_CO2 values were significantly lower than the respective pretreatment values at 150, 180, and 210 min after treatment (p < 0.05). Animals given 25 mg/kg had lower posttreatment than pretreatment Pa_CO2 values from 180 to 240 min (p < 0.05). Mean Pa_CO2 values in the 200 mg/kg group were significantly lower than in the other two groups or the control group at the 240-min time point (p < 0.05). There were no significant differences between the tracheal instillation group and the bronchoscopic instillation group at any time point at which Pa_CO2 was measured (Figure 2B).

View larger version (25K): [in this window] [in a new window]   Figure 2.   (A) Mean PaCO2 values for all groups of animals given rSP-C surfactant significantly improved at some point after treatment as compared with the respective pretreatment values, although the improvements were most marked in the 200 mg/kg group. PaCO2 values in this group were significantly lower than in the other groups at the 240-min time point (p < 0.05). (B) There were no significant differences between the bronchoscopic and tracheal instillation groups at any time point of measurement.

Mean PIP values for each group are shown in Figure 3. There was a significant decrease in PIP from 30 to 240 min after treatment as compared to the respective pretreatment values for animals given 100 and 200 mg/kg, and from 45 to 240 min for the 25 mg/kg group (p < 0.05). There were no significant differences between the 100 and 200 mg/kg groups, nor between the bronchoscopically and tracheally instilled groups (Figure 3B). For both series of experiments (Figures 3A and 3B), the surfactant treatment groups had significantly lower PIP values than did their respective control groups (p < 0.05). Moreover, when evaluating the observed changes in Pa_CO2 and PIP values together, it was clear that overall, the surfactant treatment groups had a significant improvement in ventilation efficiency over time as compared with the control groups.

View larger version (28K): [in this window] [in a new window]   Figure 3.   (A) Mean PIP values for the three different doses of surfactant administered. There were significant improvements in PIP values for all three treatment groups over time as compared with the respective pretreatment values and as compared with the control group (p < 0.05), although the improvements were most marked for the higher doses. (B) There were no significant differences between the three groups, nor between the bronchoscopic and tracheal instillation groups.

Surfactant Administration

The total times required to administer the surfactant to the groups of animals receiving the three different doses via direct tracheal instillation were similar, at approximately 10 min each (Table 2, Phase I). On the other hand, when the tracheal instillation technique was compared with the bronchoscopic technique for the 25 mg/kg dose, the times required to deliver the surfactant were significantly different (Table 2, Phase II). For the tracheal instillation technique, the total time required to deliver the surfactant was 11.6 ± 2.5 min, as compared with 24.5 ± 3.3 min for the bronchoscopic instillation technique (p < 0.05).

Surfactant Distribution

The lobar distribution patterns of exogenously administered surfactant for the various treatment groups are shown in Figure 4. These lobar distribution patterns were assessed on the basis of lung tissue samples only, since we have previously shown that the distribution of radiolabeled DPPC in lavage was similar to that in tissue (12). Furthermore, because we expressed the distribution data as lobar recovery of radioactivity relative to total lung recovery, and corrected for the size of the lobe, an ideal distribution pattern would have revealed similar values for all lobes when plotted on a histogram. Although lobar distribution assessments represent relatively crude approximations of the distribution of surfactant within the lung, our results were consistent with those in previous studies showing that surfactant distribution patterns generally depended on the volume of surfactant administered (15, 16). We found no significant differences in surfactant distribution between the three treatment doses of surfactant administered in the same volume directly through the ET tube (Figure 4A). There was also no significant difference in surfactant distribution patterns, at least at the lobar level, with the bronchoscopic and tracheal instillation techniques (Figure 4B).

View larger version (36K): [in this window] [in a new window]   Figure 4.   (A) Lobar distribution patterns of exogenously administered rSP-C surfactant for the three treatment doses tested. If surfactant distribution was completely uniform, all lobes should have had similar values of lobar DPM/mg/total DPM/mg. There were no significant differences in surfactant distribution for the three doses tested. Furthermore, there were no significant differences in surfactant distribution when the surfactant was delivered directly into each lobe via a bronchoscope as compared with its delivery by tracheal instillation (B).

Surfactant Pool Sizes

Total alveolar surfactant pool sizes, as well as the pool sizes of the individual surfactant aggregate subfractions isolated from the crude alveolar lavage (not including the 150 × g pellet), are shown in Table 3. Total alveolar surfactant pool sizes and those of the individual surfactant subfractions varied with the total quantity of surfactant administered (Phase I). Clearly, the greater the amount of surfactant administered, the greater the recovery, although the proportion of the quantity delivered that was recovered in the alveolar lavage varied for the different doses. Moreover, the quantities recovered were all significantly greater than the quantities of surfactant recovered from a group of non-surfactant-treated control animals (2.1 ± 0.1 mg PL/kg body weight). Despite these differences in the total surfactant pool sizes recovered, there were no significant differences in the proportion of the surfactant that existed in large aggregate forms at 4 h after treatment. There were also no differences in this regard between the bronchoscopically and tracheally instilled groups (Table 3, Phase II).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, we utilized the saline lavaged sheep model to evaluate the efficacy of three different doses of a recombinant SP-C-based surfactant preparation. This particular model was used in order to minimize the variability of the alveolar environment with respect to the nature of the underlying injury, and to produce a consistent severity of injury at the time of surfactant administration (11, 12). In addition, we wanted to evaluate exogenous surfactant administration in an animal model that was large enough to reflect the dosing regimens and delivery techniques that one would use in adult patients with ARDS.

Although all three doses of the rSP-C surfactant significantly improved gas exchange, the higher doses were superior. The minimal differences observed with the 100 mg/kg and 200 mg/kg doses may well have been a limitation of the lavage model, since the significant physiologic responses observed for the 100 mg/kg dose did not permit a sufficiently large window for improvement with the 200 mg/kg dose. Nevertheless, the study did show that the results obtained with these doses of rSP-C surfactant were either comparable or superior to those obtained with other modified natural surfactant preparations tested previously in this model (11, 12).

Of particular importance in this study was that these doses were well tolerated and that relatively large doses were required for optimal responses. Several studies have shown that serum proteins leak into the severely injured lungs of patients with ARDS and inhibit surfactant function in a nonstoichiometric manner (17). Because this functional impairment of surfactant can be overcome by higher concentrations of exogenous surfactant, large doses of exogenous surfactant will probably be required for clinical use in patients with severe lung injury.

In the first series of experiments, we intentionally instilled similar volumes of surfactant in order to keep the distribution of the material as consistent as possible between groups. Distribution of surfactant has been shown to influence physiologic outcomes, whereas the effects of different concentrations of surfactant on lung function have not been adequately studied (15, 16). The physiologic differences observed with the different doses of rSP-C surfactant in this study were therefore not due to differences in surfactant distribution, which was confirmed at least at the lobar level (Figure 4A). However, a somewhat surprising finding in the second series of experiments was the absence of significant differences in the lobar distribution of surfactant when it was administered either directly into each lobe under visual guidance or when it was given through the relatively simple technique of tracheal instillation. A total dose of 25 mg/kg was utilized for this comparison of delivery techniques, on the basis of results of the first series of experiments (Phase I), which showed that there was ample room for improvement in lung function with this dose.

Interestingly, previous clinical studies have reported significant improvements in lung function of patients with severe ARDS when surfactant was delivered through a bronchoscope into individual lobes (4, 20). The technique used in these previous studies was similar to the method used in the present study, although none of the clinical studies included a control group of nontreated patients. In addition, there has never been a direct comparison of the bronchoscopic technique with tracheal instillation prior to that in the present study. Our results showed that both techniques were similar with respect to physiologic responses as well as lobar distribution patterns. However, there were two obvious disadvantages of the bronchoscopic method of delivery. First, it required a significantly longer time to administer the same quantity of surfactant than did tracheal instillation, and second, it required the services of an individual experienced with bronchoscopy. These factors alone would limit the widespread clinical use of the bronchoscopic technique. At present, therefore, a tracheal instillation technique similar to that described here may represent the preferred method for administering exogenous surfactant to patients with severe ARDS.

Despite our results, and the considerations described earlier, bronchoscopic instillation may still prove to be a suitable method of surfactant administration in specific clinical situations. For example, in a patient with an obvious nonuniform pattern of lung injury, surfactant may be delivered to specific areas of the lung via the bronchoscope. In addition, certain types of diffuse lung injury, such as with meconium aspiration or gastric aspiration, may require the removal of inhibitory proteins prior to surfactant administration (21). In these situations, lavaging the individual lobes of the lungs with a surfactant suspension via the bronchoscope, with removal of proteins, may prove optimal. These types of delivery techniques need further evaluation in animal models of lung injury before being clinically applied.

The specific surfactant preparation tested in the present study was a mixture of a 34-amino-acid, human recombinant SP-C sequence and PL. This recombinant protein differs from the human SP-C protein in that the cysteine residues in positions 4 and 5 of native SP-C were replaced with phenylalanine, and the methionine in position 32 was replaced by isoleucine. These modifications have been shown to improve the interactions of the SP-C with PL, to increase film stability, and to prevent molecular aggregation, all of which have limited the usefulness of native SP-C (22). Functional studies conducted to date have shown that rSP-C surfactant significantly reduced surface tension values measured in vitro, improved lung function both in preterm animals and in small-rodent models of lung injury (7), and, as shown in the present study, improved lung function in a large animal model of lung injury.

Although we did not directly compare the rSP-C-based product examined in our study with other surfactant preparations currently available for clinical use, there are some data suggesting that this recombinant surfactant preparation is as good as or better than existing products. Davis and colleagues (9), as well as Häfner and colleagues (7, 8, 10), compared various surfactants in different animal models, and found that rSP-C surfactant was as effective as natural surfactant products, as well as a natural sheep surfactant that contained SP-A. Their findings were similar to ours in comparing the efficacy of the rSP-C surfactant in our study with natural surfactant products in the same sheep model of lung injury in previous studies (12). Furthermore, the proportion of rSP-C surfactant remaining in large aggregate forms at 4 h after administration was 60 to 90%, a value greater than that of the other surfactants, which suggests that this surfactant preparation was relatively resistant to conversion into small aggregate forms (12). Previous studies have also shown that a greater conversion of large surfactant aggregates into small aggregate forms in acutely injured lungs was associated with progressive lung dysfunction (23). More extensive metabolic studies of the conversion kinetics of rSP-C surfactant as well as of other surfactant preparations in vivo, and of potential mechanisms responsible for this conversion, are currently underway.

The significance of utilizing an effective recombinant exogenous surfactant preparation for patients with ARDS cannot be overstated. When considering a novel therapeutic strategy for critically ill patients, one must consider the complexity of the intervention as well as its cost. These two factors should be considered before its widespread clinical use is undertaken. The quantities of material that would potentially be necessary could result in resource limitations for natural products, whereas unlimited supplies of a recombinant product would be available. Presumably this would eventually lead to lower costs with use of the proposed intervention on a large scale. In addition, if a treatment modality is to be used extensively by various personnel, the administration technique should be relatively simple and straightforward. The results of our study show that the rSP-C surfactant delivered via tracheal instillation meets the objectives described here.

In summary, we have shown that a recombinant SP-C- based surfactant was well tolerated and effective in improving the lung function of lung-injured adult sheep at the various doses tested. Similar results were obtained in these animals when the surfactant was administered either directly into each lobe of the lung via a bronchoscope or was instilled through an ET. However, the former method had some disadvantages that may limit its clinical usefulness. Based on these and other available preclinical data, we feel that clinical trials testing this surfactant preparation in patients with ARDS are warranted.