INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE ACUTE RESPIRATORY DISTRESS SYNDROME (ARDS) is an acute lung injury resulting in respiratory failure and is defined by hypoxemia, nonhydrostatic pulmonary edema, and decreased lung compliance. This syndrome can arise from either direct or indirect pulmonary insults, ultimately resulting in diffuse alveolar damage (2). Although the reported survival of patients with ARDS has varied over the last several years (2), some studies have documented mortality as high as 70% (20). These reports most often include patients with systemic sepsis, which is also the most common predisposing condition leading to ARDS (3).

Studies investigating ventilation-induced lung injury (VILI) observed that some of the alterations of the endogenous surfactant system were similar to those reported for patients with ARDS (12, 26). Specifically, a decrease in the proportion of the superior functioning large surfactant aggregates relative to the functionally inferior small surfactant aggregates has been documented in both situations (13, 15, 27). Moreover, ventilation strategies utilizing relatively low tidal volumes resulted in less in vivo conversion of large aggregates into small aggregates compared with strategies that used higher tidal volumes (12). This decreased conversion was associated with superior oxygenation responses compared with the larger tidal volume strategies. In general, these findings have led to the hypothesis that there is an important relationship between the ventilatory effects on the surfactant system and the progression of lung injury in patients with severe respiratory failure due to ARDS.

Despite the growing concern that mechanical ventilation may contribute to the pathophysiology of acute lung injury (ALI), ventilation remains an important therapeutic intervention for patients with ARDS. It is, therefore, imperative that one attempt to minimize the harmful effects of ventilation while optimizing its beneficial effects. This can only be accomplished with a greater understanding of the mechanisms responsible for the lung damage induced by mechanical ventilation over the course of this disease.

Recently, we characterized an adult rat model of systemic sepsis induced by cecal ligation and perforation (CLP) (17). The relatively mild lung injury observed in these spontaneously breathing septic animals was associated with significant alterations in the endogenous surfactant system. The objective of the present study was to examine the effects of mechanical ventilation on lung function and the endogenous alveolar surfactant system in this clinically relevant model of lung injury. This information will provide insight into mechanisms contributing to the development of ARDS, which will ultimately lead to superior therapeutic strategies for these patients.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal preparation. Adult male Sprague-Dawley rats weighing 350-450 g were acclimatized to the laboratory environment for 1 wk before surgery with free access to food and water. Rats were anesthetized with 4% halothane in oxygen in an anesthetizing box. Once sedation was induced, rats were transferred to a surgical table with a nose cone setup to maintain anesthesia. Two PE-50 monitoring catheters were inserted in the right external jugular vein and right carotid artery, respectively. These lines were routed subcutaneously to the back of the neck and attached to a three-fluid-channel (22-gauge) swivel system. The incision made to insert the catheters was closed with 3-0 silk suture. To induce sepsis and subsequent lung injury, the cecal ligation and perforation (CLP) technique was performed immediately after catheter placement as previously described (17). Briefly, this procedure involved performing a laparotomy, exposing the entire cecum with ligation of the distal one-third of the cecum, and then puncturing the ligated section twice with a 16-gauge needle. Sham control groups consisted of animals undergoing identical anaesthetic procedures and catheter placements for subsequent monitoring, but the laparotomy and CLP procedures were not performed.

1

1

Experimental groups. Both sham and septic animals were randomly assigned after surgery to one of six experimental groups. Two experimental groups consisted of both sham and septic animals that remained spontaneously breathing for 24 h after surgery and then were killed. These groups were designated Sham-Spon and Septic-Spon, respectively. The other four experimental groups consisted of sham and septic animals that were mechanically ventilated with or without positive end-expiratory pressure (PEEP) for 90 min starting at the 22.5-h time point after surgery and then were killed at the 24-h time point. As noted, two different ventilation strategies were utilized in these latter four groups. The Sham-Vent and Septic-Vent groups were ventilated by using a tidal volume of 9 ml/kg body wt, a RR of 62 breaths/min, and 0 cmH₂O of PEEP. The other two groups were ventilated by using identical parameters but with 5 cmH₂O of PEEP. These groups were designated Sham-PEEP and Septic-PEEP.

Mechanical ventilation. A volume-cycled rodent ventilator (Harvard Instruments, St. Laurent, PQ, Canada) was used to ventilate animals involved in this study. PEEP was added to the ventilation system by submerging the expiratory tube in water to a depth of 5 cm. The tidal volume and RR were chosen on the basis of preliminary studies involving normal rats in which arterial PCO₂ (Pa_CO2) values were maintained between 35 and 45 Torr over a 90-min time period. An inspired oxygen fraction (FI_O2) of 50% was delivered to all ventilated animals to maintain adequate arterial PO₂ (Pa_O2) values in the septic groups. To determine the specific effects of the increased FI_O2 in these animals, separate groups of spontaneously breathing sham and septic animals were placed in an airtight chamber at the 22.5-h time point after surgery and exposed to 50% oxygen for 90 min.

Experimental protocol. For all ventilated animals, a sample of arterial blood was obtained for measurements of Pa_O2, Pa_CO2, and lactate levels ~22.5 h after surgery. At the same time, MABP, HR, and RR were recorded. For these ventilated groups, animals were anesthetized in their cage via a slow intravenous injection of pentobarbital sodium until the toe pinch reflex was absent. The rats were then transferred to a surgical table where a tracheotomy was performed. Lidocaine (1%) was administered subcutaneously as local anesthetic, and the trachea was exposed via a midline incision. A 14-gauge angiocatheter was inserted into the trachea and used as an endotracheal tube for ventilation. Animals were then connected to the ventilator, 0.2 ml pancromunium bromide was administered intravenously to eliminate spontaneous breathing, and 0.2 ml bupenoraphen was administered intramuscularly for further analgesia. Additional boluses of both pancromunium bromide and pentobarbital sodium were administered intravenously to maintain paralysis and adequate depth of anesthesia, respectively. At 10, 20, 30, 60, and 90 min after initiation of ventilation, arterial blood samples were obtained, and similar physiological parameters as noted previously were recorded. All mechanically ventilated animals were killed after 90 min of ventilation via an intravenous bolus of pentobarbital sodium and transection of the abdominal descending aorta.

Lung lavage analysis. Immediately after being killed, all animals underwent a whole lung lavage as previously described (13, 17). Briefly, the lungs were infused with sterile 0.15 M saline until fully distended, and the saline was withdrawn and reinfused two more times. This procedure was repeated a total of four times, and the combined volume of the total lavage was recorded. There were no differences in the total volume of saline infused or recovered after the lavage procedure among the six experimental groups.

Phospholipid and protein measurements. To measure the total phospholipid pool size and individual surfactant aggregate phospholipid pool sizes, aliquots from the total lavage, 150-g pellet, large-aggregate, and small-aggregate fractions were extracted by using the method of Bligh and Dyer (4). Phospholipid-phosphorous levels in each of these extracts were determined by using the Duck-Chong phosphorous assay (7). Briefly, 100 µl of 10% magnesium nitrate in methanol were added to the extracted lipids. After being dried, the samples were ashed in a fume hood on an electric rack for ~1 min. After 1 ml of 1 M HCl was added, the samples were reheated on a heating block while covered for 15 min at 95°C. After cooling, a 66-µl aliquot of each sample was added to individual wells of a 96-well plate along with 134 µl of a dye consisting of 4.2% ammonium molybdate in 4.5 M HCl with 0.3% malachite green (1:3 vol/vol). The absorbency of each sample was read at 650 nm by using a MKII Titretek Multiskan ELISA plate reader and compared with reference standards on the same plate.

Phospholipid compositional analysis. Lipids from the total lavage were extracted by using the method of Bligh and Dyer (4), and a cold acetone precipitation was performed to remove neutral lipids. Phosholipids were then separated by thin-layer chromatography as described previously (25). The individual phospholipids were then measured by using the previously described Duck-Chong assay (7).

Surfactant protein analysis. Surfactant-associated proteins A (SP-A) and B (SP-B) were measured in the total lavage obtained from the experimental groups by using ELISAs as previously described (25).

Surface activity analysis. In vitro surface tension measurements of unextracted large aggregate samples recovered from animals in all groups were performed by using a pulsating bubble surfactometer (Electronetics, Amherst, NY) as described by Enhorning (8). Aliquots of the large-aggregate fractions were prepared to obtain a final concentration of 1 mg phospholipid/ml in 0.15 M NaCl and 1.5 mM CaCl₂. The samples were incubated for 90 min at 37°C before analysis. Briefly, a bubble was created in the suspension containing the large surfactant aggregate fraction. After 10 s of adsorption, the bubble was pulsated for a period of 5 min between a maximum radius of 0.55 mm and a minimum radius of 0.44 mm at a rate of 20 pulsations/min and a temperature of 37°C. Pressure was monitored across the air-liquid interface by using a pressure transducer, and surface tension was calculated at the minimum and maximum bubble radii.

Calculations and statistics. Data are expressed as means ± SE. Values between groups were compared using a two-way ANOVA followed by the Student-Newman-Keuls test for multiple comparison. A repeated-measures ANOVA was used to compare measurements within groups over time. A probability level of P < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physiological parameters. Table 1 shows the total number of animals, the mean body weights, and baseline physiological measurements of both sham and septic animals 5 h after recovery from the initial surgery. There were no significant differences in Pa_O2/FI_O2 values or in arterial lactate levels between these two groups at the 5-h time point. Although there were statistical differences noted in MABP, HR, and RR values between the two groups, all values were within the normal physiological range. In general, these data indicate that all animals adequately recovered from the surgical procedures within 5 h.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Oxygenation of sham and septic animals ventilated for 90 min with or without positive end-expiratory pressure (PEEP). PaO2, arterial PO2; FIO2, inspired O2 fraction. There was no change in oxygenation during ventilation of sham animals ventilated with PEEP (Sham-PEEP) or without PEEP (Sham-Vent). Septic animals ventilated without PEEP (Septic-Vent) had a significant deterioration in oxygenation compared with preventilation values (* P < 0.05), whereas septic animals ventilated with PEEP (Septic-PEEP) had a significant improvement in oxygenation compared with preventilation values (* P < 0.05).

Lung lavage and surfactant analysis. Total protein recovered from the lung lavage in the six experimental groups is shown in Fig. 2. There were no significant differences between the two spontaneously breathing groups (Sham-Spon vs. Septic-Spon). Similarly, protein levels among the four ventilated groups were not significantly different; however, they were significantly higher than the spontaneously breathing groups (P < 0.05).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Total protein recovered from lung lavage was increased in all ventilated (Vent) animals compared with spontaneously breathing (Spon) groups (*P < 0.05). There were no differences in total protein recovery between spontaneously breathing sham and septic groups.

View larger version (19K): [in this window] [in a new window]   Fig. 3.   Total surfactant pool sizes in spontaneously breathing and ventilated sham and septic animals were not significantly different; however, there was a trend for lower total surfactant pools in septic animals. Mechanical ventilation did not have an effect on surfactant pools in either sham or septic animals.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Percent large aggregate recovery was increased in the spontaneously breathing septic animals (Septic-Spon) compared with spontaneously breathing sham animals (Sham-Spon) (* P < 0.05). Ventilation without PEEP resulted in a higher percent large aggregate in sham animals compared with spontaneously breathing sham animals (Sham-Vent vs. Sham-Spon) (*P < 0.05), and lower percent large aggregate in septic animals compared with spontaneously breathing septic animals (Septic-Vent vs. Septic-Spon) (# P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mechanical ventilation is utilized to maintain adequate oxygenation in patients with respiratory failure. In fact, this intervention is one of the few supportive therapies available for patients with respiratory failure severe enough to meet the criteria for ALI and ARDS (2). Unfortunately, mechanical ventilation itself has been shown to have deleterious effects on lung function as reviewed by Dreyfuss and Saumon (5). These observations have led to several studies evaluating potential factors contributing to the lung damage induced by this intervention. The mechanical forces attributable to ventilation have been thought to be responsible for the damage incurred. These forces include alveolar overdistension or "volutrauma" as well as repetitive opening and collapse of both small airways and distal lung units that create shear forces (21, 30). The present study evaluated the effects of this latter phenomenon by utilizing ventilation strategies with and without PEEP.

Mechanisms by which alveolar overdistension and airway collapse result in lung dysfunction include increased microvascular permeability, initiation, and/or perpetuation of an inflammatory response within the lung due to physical stretch and alterations of the endogenous surfactant system (5, 12, 23). Although these particular issues have been addressed in previous studies, most were conducted in normal lungs by using relatively nonphysiological ventilation parameters to induce the injury (6, 24, 29, 30). We studied the effects of more conventional ventilatory strategies in a clinically relevant animal model of lung injury, namely systemic sepsis-induced lung injury.

A 90-min period of mechanical ventilation indicated that septic animals associated with a relatively mild lung injury experienced increased susceptibility to the harmful effects of this intervention compared with sham control animals. This was only evident, however, when repeated collapse and reopening of lung units were allowed to occur when PEEP was not utilized (Septic-Vent group). These animals exhibited a marked deterioration in lung function, which was associated with significant alterations of the endogenous surfactant system. These alterations included a shift in the proportion of functionally superior large-aggregate forms to inferior functioning small aggregates, as well as significantly impaired function of the remaining large aggregates compared with large aggregates isolated from septic animals ventilated with PEEP (Septic-PEEP).

Alveolar surfactant can be separated into a surface-active large-aggregate fraction and a less-surface-active small-aggregate fraction. Large aggregates are secreted from alveolar type II cells and are believed to represent a surfactant reservoir for the surface film located at the air-liquid interface of the alveoli. Pulse-chase studies, as well as in vitro surface area cycling experiments, have shown that these large surfactant aggregates are converted into small surfactant aggregates. This conversion process is predominately dependent on two factors: a change in the alveolar surface area and protease activity (10, 13, 22, 28). Increases in either of these two variables can result in an increase in large-aggregate conversion (10, 28). Increases in alveolar surface area occur when patients are mechanically ventilated, because the administered tidal volumes during ventilation are usually greater than those generated during spontaneous breathing during respiratory failure. This factor, together with the inflammatory changes generally observed in injured lungs, can ultimately result in an increased conversion of large aggregates. As the injury progresses, this increased conversion results in changes in aggregate pools that are thought to impact lung function and thus represent an important mechanism contributing to the lung dysfunction observed in severe ARDS (11, 27).

Surfactant changes observed in the sham animals revealed that ventilation without PEEP resulted in an increase in percent large aggregates compared with spontaneously breathing sham animals that was associated with the maintenance of Pa_O2/FI_O2 levels and low airway pressures. These observations are similar to those described by Webb and Tierney (30), where no lung damage was evident in normal, control rats when ventilated for 1 h with inspiratory pressures of 14 and 0 cmH₂O PEEP. The observed increase in percent large aggregates in these lungs were presumably due to an increase in freshly secreted surfactant, because alveolar stretch has been shown to induce surfactant secretion in normal lungs (18). A relatively normal rate of conversion of these large aggregates would also be expected within the air space of normal lungs. However, in a situation of lung injury, as observed in the septic animals in the present study, there was a significant decrease in the percent large-aggregate fraction compared with that of spontaneously breathing septic animals. There was an opposite response compared with sham animals ventilated with the identical strategy. This observation is consistent with previous studies that have implicated a decrease in percent large aggregates as a contributor to lung dysfunction (14, 15, 27). However, these alterations in aggregate forms were not likely fully responsible for the lung impairment observed in these septic animals (Septic-Vent), because the quantity of large aggregates remaining in the lung at death were similar to the ventilated sham group (Sham-Vent). Other factors may, therefore, have contributed to the lung dysfunction observed in the Septic-Vent group. Analysis of the surface tension-reducing function of the remaining large-aggregate forms in the experimental groups revealed that the surface tension values of the large aggregates isolated from the Septic-Vent group was markedly higher compared with those from the other groups. Large aggregates from all sham groups reduced surface tension values below 10 mN/m, indicating excellent biophysical activity, which is presumably adequate for maintaining alveolar stability at end expiration in these animals, even in the absence of extrinsic PEEP. In contrast, large aggregates from the Septic-Vent group had surface tension values >17 mN/m after 100 pulsations. Therefore, the relative change in large-aggregate pools combined with the impairment of large aggregate functional activity over the period of ventilation would contribute to alveolar instability and consequently impact lung compliance (increased PIP values) and oxygenation in the septic group ventilated without PEEP. The mechanism responsible for the inferior surface activity of the isolated large aggregates from this group is unknown at the present time and was surprising considering the similar lipid and protein composition. We speculate that, although the quantitative analysis of SP-A as determined by the ELISA technique revealed no significant difference among the experimental groups, there may well have been conformational and/or monomer or multimer changes that occurred in this protein that would not be reflected by the ELISA measurement. Such functional vs. quantitative abnormalities in SP-A and/or the other surfactant-associated proteins would potentially make these large aggregates more susceptible to conversion and/or protein inhibition within the air space of the injured lung.

The addition of PEEP to the ventilation strategy of septic animals mitigated both the physiological impairment and alterations of surfactant. PEEP prevents the repetitive opening and collapse of alveolar lung units, thereby decreasing shear stresses within the lung, which can damage lung units (30). In the present study, PEEP mitigated both the conversion of large aggregates and the decreased function of these large aggregates in the lungs of septic animals compared with those of similar animals ventilated without PEEP. Similar findings were recently reported by Michna et al. (19) because the addition of increasing amounts of PEEP to surfactant-treated, ventilated preterm lambs resulted in superior lung function as well as preservation of the administered surfactant in large aggregate forms.

Surprisingly, both phospholipid and surfactant-associated protein composition were not altered in either the spontaneously breathing or mechanically ventilated septic animals in the present study. Although previous studies have demonstrated significant alterations in these parameters in patients with severe ARDS (9, 27), the lung injury induced by the CLP procedure in the present study was relatively mild and represented an early stage of lung injury not well characterized in the clinical setting to date. We conclude that the changes in alveolar surfactant may represent an initial response to mechanical ventilation. Furthermore, we suggest that the rapid onset of deleterious alterations of alveolar surfactant that occur in the present study can, during extended ventilation, contribute to the progression to severe lung dysfunction. Thus it is imperative that an optimal ventilation strategy be employed at the start of this intervention.

In summary, we have shown that animals with a systemic sepsis-induced lung injury were more susceptible to the deleterious effects of mechanical ventilation than noninjured animals. Although several mechanisms may contribute to ventilation-induced lung injury, the present study has implicated the repetitive opening and collapse of lung units resulting in alterations of the endogenous surfactant system as at least one potential mechanism. The importance of utilizing adequate levels of PEEP in patients with ARDS is underscored by our results, as is the necessity to implement optimal ventilatory strategies at the onset of this intervention. Because ventilation strategies utilizing smaller tidal volumes are being implemented in patients with ARDS (1), it is important to understand the mechanisms responsible for the physiological differences observed in these studies so that the deleterious effects of mechanical ventilation are minimal.