INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Mechanical ventilation can injure the lung either directly or by inducing an inflammatory response that then injures the lung (1). Ventilation of adult lungs with volumes above total lung capacity or below functional residual capacity causes lung injury and initiates an inflammatory cascade resulting in release of proinflammatory cytokines (2, 3). Injured adult lung can be a source of systemic inflammation leading to multiple organ dysfunction (1). Different styles of ventilation can minimize ventilator-induced lung injury. Ventilation of patients with acute respiratory distress syndrome (ARDS) with higher positive end-expiratory pressure (PEEP) and lower tidal volumes improves outcomes (4).

The preterm lung is uniquely susceptible to injury with the initiation of ventilation after birth because potential lung gas volumes are small, surfactant may be deficient, the lung matrix is not fully developed, and the airspaces contain residual lung fluid (5). Lung injury in the preterm newborn can occur from birth if resuscitation is initiated with large tidal volumes (6, 7). Infants destined to develop bronchopulmonary dysplasia have increased granulocytes and proinflammatory cytokine levels in airway samples (8, 9), but there is minimal information about how proinflammatory indicators appear or progress in the preterm lung with mechanical ventilation. The immune system of the preterm lung differs from the adult lung because the fetal lung contains almost no neutrophils or mature macrophages (10) and the innate host defense proteins SP-A and SP-D that modulate inflammatory responses are low (11, 12). The production of proinflammatory cytokines in the lung has not been well studied in the ventilated preterm animal model (13). Ventilation without PEEP resulted in the highest levels of proinflammatory mediators in isolated adult rat lungs (3). Previously we found that ventilation of the surfactant-treated preterm sheep lung without PEEP for 7 h results in poor compliance and oxygenation, and loss of static lung volumes (14). However, the possible inflammatory consequences of initiation of ventilation were not evaluated. Therefore, we asked if initiation of ventilation with different levels of PEEP for 2 h or 7 h would result in an inflammatory response and the expression of proinflammatory cytokines in the preterm lung.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Delivery and Ventilation of Lambs

Preterm lambs were delivered by cesarean section as previously described (6, 14). Pregnant ewes at 126-132 d gestation (term 150 d) were anesthetized prior to delivery of the preterm lambs. Before the first breath, each preterm lamb was treated with a recombinant surfactant protein-C (rSP-C) (Venticute; Byk Gulden, Konstanz, Germany) containing synthetic surfactant at a dose of 100 mg lipid/kg body weight (15). This surfactant is free of endotoxin and is as effective as natural surfactant in treatment of respiratory distress syndrome (RDS) in preterm animals (15) and in adult lung injury models (16).

The lambs were randomized to PEEP values of 0, 4, or 7 cm H₂O and were ventilated for 2 h or 7 h with time-cycled, pressure-limited ventilators (Sechrist Industries, Anaheim, CA). Subsequently, only the fraction of inspired oxygen (FI_O2) and peak inspiratory pressure (PIP) were changed in order to maintain the arterial PO₂ in the range of 100 to 200 mm Hg and the Pa_CO2 between 50 and 60 mm Hg for 2 h or 7 h of ventilation. Six fetal lambs that were not ventilated were used as the unventilated control group.

Lung Gas Volumes and Lung Processing

The maximal lung volumes were measured at 40 cm H₂O pressure (6). Alveolar washes were performed on the isolated left lung (17) and the aliquots were used for determinations of total protein, neutrophil counts, and hydrogen peroxide activity. Protein was determined using a modified method of Lowry and coworkers (18). Differential cell counts were performed on cytospin preparations after staining with Diff-Quik (Scientific Products, McGaw Park, IN). Hydrogen peroxide was measured by the commercial Bioxytech H₂O₂-560 assay (OXIS International, Portland, OR).

RNA Isolation and RNase Protection Assay

Total RNA was isolated from lung tissue from right lower lobe (19) and 10 µg of total RNA was used for multiprobe RNase protection assays using previously characterized sheep specific cytokine probes (20). Tissues from lambs (n = 6-7/group) ventilated for 7 h were from the previous study by Michna and coworkers (14). The protected fragments were resolved on polyacrylamide urea gels, visualized by autoradiography, and quantified on a PhosphorImager by means of ImageQuant v1.2 software (Molecular Dynamics, Sunnyvale, CA).

Lung Morphometry

Four animals in each group ventilated for 2 h were studied. The right upper lobe was inflation fixed and 5-µm sections were stained with hematoxylin and eosin on polysine-coated slides. The respiratory parenchyma was classified into three categories: 1 = collapsed alveoli, 2 = distended alveoli, and 3 = overdistended alveoli. Morphometric measurements of the proportion (percentage fractional areas) were performed in a blinded fashion using point counting (21) and were calculated as percentage fractional area = (n/N) × 100, where n = number of points in each category and N = total number of points counted for each field. Three fields from each of two sections per animal were analyzed.

Data Analysis

Results are given as mean ± SEM. Analysis of variance (ANOVA) was used for comparison of differences between groups with Student-Newman-Keuls test used for post-hoc analysis. Two-tailed unpaired t tests were used for two groups comparison. Significance was accepted at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Description of Lambs

Thirty-four lambs at 126-132 d gestational age were randomized to PEEP of 0, 4, and 7 cm H₂O. Control group consisted of six fetal lambs that were not ventilated. There were no differences in body weights (Table 1) and cord blood pH between the groups.

Respiratory Outcomes

The Pa_CO2 values were in the range of 50-65 mm Hg and the tidal volumes (VT) were 9-11 ml/kg for both the 2 h and the 7 h ventilation groups (Figure 1). The ventilatory pressures required to achieve the desired tidal volume were higher (p < 0.05) for the lambs ventilated with 0 PEEP (Figure 2A). The mean Pa_O2/FI_O2 ratios were 1.5- to 3.5-fold higher for 4 and 7 PEEP groups compared with 0 PEEP (p < 0.05) (Figure 2B). One animal included in the study from the 2 h ventilation group ventilated with 4 PEEP developed pneumothorax toward the end of the study.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Sequential measurements of PaCO2 and the tidal volumes. The ventilatory goals were to keep PaCO2 between 50 and 60 mm Hg, and the tidal volumes (VT)  10 ml/kg. There were no differences between the groups. The filled symbols indicate combined PaCO2 values and VT for both 2 h and 7 h ventilation groups.

View larger version (25K): [in this window] [in a new window]   Figure 2.   (A-C ) Ventilatory pressures measured as peak inspiratory pressure minus positive end-expiratory pressure (PIP  PEEP), PaO2/FIO2 ratios, and lung gas volumes at 40 cm H2O (V40) 2 h and 7 h. The ventilatory pressures required to achieve the target PaCO2 were lower for the 4 and 7 PEEP groups. The PaO2/FIO2 ratios were higher for the 4 and 7 PEEP groups. V40 with 0 PEEP had consistently lower lung volumes relative to 4 and 7 PEEP after both 2 h and 7 h ventilation. *p < 0.05 versus 4, 7 PEEP.

Maximal lung volumes measured at 40 cm H₂O (V₄₀) were lower with 0 PEEP than 4 and 7 PEEP after 2 h ventilation, and similar differences were measured after 7 h ventilation (Figure 2C).

Indicators of Lung Inflammation

The amounts of protein (mg/kg) in alveolar wash fluid were 2- to 4-fold higher in all ventilated animals at 2 h and 7 h relative to unventilated fetal controls (p < 0.05) (Figure 3). The total protein was higher for the 0 PEEP and 7 PEEP groups than for the 4 PEEP group after 7 h ventilation (p < 0.05).

View larger version (12K): [in this window] [in a new window]   Figure 3.   Total protein in alveolar wash fluid after 2 h and 7 h ventilation. All ventilated animals had increased protein relative to unventilated fetal controls. After 7 h of ventilation, protein in alveolar wash fluid significantly increased in the 0 and 7 PEEP groups relative to the 4 PEEP group. tp < 0.05 versus all groups of ventilated animals; *p < 0.05 versus 4 PEEP at 7 h.

Few neutrophils were detected in the alveolar wash fluid from the unventilated lambs as shown in Figure 4. Neutrophil counts in the alveolar wash fluid were significantly elevated (2.5- to 6-fold) in all ventilated animals relative to unventilated fetal controls by 2 h (p < 0.05) (Figure 4A). Animals ventilated with 0 and 7 PEEP had significantly higher neutrophils in the alveolar wash fluid than those in the 4 PEEP group. Total hydrogen peroxide activity (µmol/kg body weight) in alveolar cells was significantly higher in all ventilated groups, however the different PEEP used for ventilation did not alter the amount of hydrogen peroxide (Figure 4B).

View larger version (14K): [in this window] [in a new window]   Figure 4.   Neutrophil counts (A) and total H2O2 activity (B) in alveolar wash fluid after 2 h ventilation. All ventilated groups had higher neutrophil counts relative to fetal controls. The 0 and 7 PEEP groups had elevated neutrophil counts relative to the 4 PEEP group. Total H2O2 activity in alveolar washes was increased in ventilated groups compared with unventilated fetal control. tp < 0.05, control group versus all ventilated groups; *p < 0.05 versus 4 PEEP.

The steady-state mRNA levels for the proinflammatory cytokines interleukin (IL)-1, IL-6, IL-8, and tumor necrosis factor- (TNF-) were very low in unventilated fetal controls (Figures 5 and 6). All ventilated groups had elevated levels of cytokine mRNA relative to the unventilated controls (p < 0.05). In animals ventilated with 0 PEEP IL-1 mRNA increased 24-fold and IL-6 mRNA increased 20-fold compared with 7- to 8-fold increase in the 4 PEEP group (Figure 5). Both IL-1 and IL-6 were elevated in the 0 PEEP group relative to the 4 PEEP group at 2 h and 7 h ventilation. IL-1 and IL-6 mRNA expression decreased 2- to 3-fold from 2 h to 7 h in all PEEP groups. IL-8 and TNF- were modestly increased (2- to 4-fold) by ventilation (Figure 6). The IL-8 mRNA level was significantly higher in 0 PEEP relative to 4 PEEP after 2 h ventilation. There was a tendency toward increased TNF- expression after 7 h ventilation in all PEEP groups.

View larger version (17K): [in this window] [in a new window]   Figure 5.   IL-1 and IL-6 mRNA levels after 2 h and 7 h ventilation. IL-1 and IL-6 mRNAs were elevated in all ventilated groups relative to fetal control groups. Both IL-1 and IL-6 mRNA were significantly increased in 0 PEEP relative to 4 PEEP after both 2 h and 7 h ventilation. IL-1 and IL-6 mRNA decreased between 2 h and 7 h. tp < 0.05, control groups versus all ventilated groups, *p < 0.05 versus 4 PEEP, and #p < 0.05 versus 7 h ventilation. The inset shows representative RNase protection assay for IL-1 and IL-6 after 2 h ventilation.

View larger version (22K): [in this window] [in a new window]   Figure 6.   IL-8 and TNF- mRNA levels after 2 h and 7 h ventilation. IL-8 was increased in all ventilated groups relative to fetal controls. IL-8 was also significantly elevated in 0 PEEP compared with 4 PEEP after 7 h ventilation. TNF- was increased in all ventilated groups after 7 h ventilation relative to fetal control groups. The different PEEP levels did not influence TNF- mRNA levels. tp < 0.05 control groups versus all ventilated groups. *p < 0.05 versus 4 PEEP. **p < 0.05 control groups versus groups ventilated for 7 h. The insets show representative RNase protection assays for IL-8 and TNF- after 2 h ventilation.

Morphometry

Figure 7A shows representative sections of lung parenchyma illustrating 1 = collapsed, 2 = distended, and 3 = overdistended alveoli. The percentage fractional area of collapsed alveoli was 38.5% for 0 PEEP compared with 10% for 4 PEEP and 5% for 7 PEEP indicating that ventilation with 0 PEEP resulted in an atelectatic lung (Figure 7B). The percentage fractional area of highly distended alveoli was 61% for 7 PEEP compared with 9% for 0 PEEP demonstrating that 7 PEEP (p < 0.05) caused more distension of alveoli. Inflammation as indicated by large numbers of granulocytes or tissue edema was not evident for any of the groups.

View larger version (46K): [in this window] [in a new window]   View larger version (24K): [in this window] [in a new window]   Figure 7.   Morphology of the lungs. (A) Representative sections that were scored as 1 = collapsed. 2 = distended, and 3 = overdistended alveoli. All the panels were photographed at the same magnification. Original magnification: ×230. Scale bars = 100 µm. (B) Percentage fractional areas of alveolar inflation. The 0 PEEP group had more collapsed alveoli compared with 4 and 7 PEEP. 7 PEEP showed more overdistended alveoli compared with 0 PEEP. * p < 0.05 versus 4 and 7 PEEP; tp < 0.05 versus 4 and 7 PEEP.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

To study the effects of ventilator-induced lung injury in the preterm lung, surfactant-treated preterm lambs were ventilated for 2 h and 7 h with PEEP of 0, 4, and 7 cm H₂O. We found that initiation of ventilation resulted in increased protein and activated granulocytes in the alveolar wash fluid and in increased expression of proinflammatory cytokine IL-1, IL-6, and IL-8, and TNF- mRNAs in the fetal lung tissue. Mechanical ventilation of preterm lung resulted in large increases in IL-1 and IL-6 mRNA (8- to 24-fold) and more modest increases of IL-8 and TNF- mRNA. Use of 4 PEEP minimized the proinflammatory response to ventilation in the premature lung. Animals ventilated with 0 and 7 PEEP had higher protein leaks, elevated neutrophils, and increased mRNA expression of proinflammatory cytokines relative to those ventilated with 4 PEEP. Ventilation with 0 PEEP caused the most severe injury as evident by the highest levels of biological markers of lung injury.

Ventilator-induced lung injury in adults is associated with decreased compliance, accumulation of neutrophils, increased protein leak, and increased levels of proinflammatory cytokines (2). Ventilation of the adult lung without PEEP results in a loss of static lung volume and accumulation of granulocytes in the lungs (22). We have previously shown that use of PEEP preserved static lung volumes and improved oxygenation after 7 h of ventilation (14). In the present study we found similar results within 2 h of ventilation.

The immature immune system of a preterm fetal lung is characterized by few macrophages and low or absent granulocytes (10). The granulocytes of the preterm also have deficiencies in adherence, deformation, chemotaxis, and respiratory burst (23). There are few cells with poor inflammatory potential in the preterm lung and a blunted or delayed response might be anticipated. However, within 2 h of initiation of ventilation, activated neutrophils that produced hydrogen peroxide were recovered in the alveolar wash fluid, indicating recruitment and activation of these inflammatory cells. Carlton and coworkers (24) previously observed granulocyte recruitment to the nonsurfactant treated lungs of preterm ventilated lambs using high tidal volume ventilation strategy. Our study differs in that we used a strategy of surfactant treatment and lower tidal volume ventilation to minimize lung injury. Nevertheless neutrophil accumulation occurred in the lungs. Increased protein leak in these animals can be explained by neutrophil-induced increased vascular permeability as neutrophil-depleted lambs have minimal protein leaks (24). Ventilator-induced lung injury can be a vicious cycle with inhibition of surfactant function secondary to protein leak, leading to atelectasis and thereby more injury. These results demonstrate that ventilation of preterm lung even with a minimally injurious strategy causes acute protein leak and early recruitment of activated neutrophils.

Proinflammatory cytokines contribute to pathogenesis of ventilator-induced injury in adult lung by inducing other proinflammatory mediators, causing sequestration and accumulation of neutrophils and enhancing vascular permeability (25). We chose to assay the proinflammatory cytokines IL-1, IL-6, IL-8, and TNF- because they are increased in adult lung injury models (4), are early responders to injury, and the presence of these proinflammatory mediators in airway samples from ventilated preterm infants increases the risk of developing bronchopulmonary dysplasia (8). The preterm lamb lung responded to mechanical ventilation with increased expression of these cytokine mRNAs within 2 h as shown in adult animals (3). However, in this experiment, the magnitude of response of individual cytokine mRNAs was different. The low induction of IL-8 in the present study is in contrast to markedly elevated levels seen in adult lung injury models (2, 3). TNF- expression in our study differs from adult lung injury models in that the TNF- mRNA increased at a later time (7 h) compared with an early response seen in adults. Our observation of a modest increase in expression of TNF- mRNA after ventilation is consistent with low levels of TNF- seen in airway specimens of preterm infants in response to different injuries (26, 27). These different responses of IL-8 and TNF- may be attributed to the immaturity of the fetal immune system.

To determine whether different ventilation strategies changed the expression of these biological markers of lung injury, we used different levels of PEEP. Our data indicate that 0 and 7 PEEP are more injurious than 4 PEEP. However, 0 and 7 PEEP will cause lung injury by different mechanisms. The histology/morphometry is consistent with observations in adult animal models that with 0 PEEP the lung was being repetitively opened and closed leading to increased neutrophil infiltration, surfactant inactivation, and progressive lung injury (28). Ventilation with 7 PEEP will cause overdistension with adverse hemodynamic effects and stress failure of pulmonary capillaries causing injury (3). Use of PEEP in the physiological range of 3-4 cm H₂O may protect the lung from injury as it minimizes the inflammatory response caused by ventilation. Expression of IL-1 and IL-6 mRNA decreased between 2 h and 7 h in all ventilated animal groups. Currently there is no information about the kinetics of expression of these cytokines in ventilated preterm animals. In contrast TNF- had a modest initial expression followed by a slight increase by 7 h of ventilation. These expression patterns are different from adult lung injury models (3) and may be due to the immature immune system variable responses to different stimuli and the biological role of cytokines in fetal development.

In conclusion, our study demonstrates that the initiation of mechanical ventilation in a preterm lung is injurious and that different ventilatory strategies influence the injury response. We tried to minimize lung injury by using low VT, administering surfactant before the first breath, and maintaining higher Pa_CO2. Nevertheless indicators of injury still increased with ventilation. Our observation that mechanical ventilation from birth induces proinflammatory cytokines is provocative from the clinical perspective because increased cytokine levels occur in bronchoalveolar lavage of infants who are at risk of developing BPD (8). This suggests that these cytokines may contribute to the pathogenesis of chronic lung disease in ventilated preterm infants. Epidemiological studies correlate the avoidance of intubation and mechanical ventilation with large reductions in the incidence of BPD and IVH (29). Van Marter and coworkers identified the initiation of ventilation as a major risk factor for development of chronic lung disease (30). Our results in preterm lambs and the available clinical data support approaches that minimize the mechanical ventilation of the preterm infant. Mechanical ventilation may be necessary in the preterm to sustain life but may also cause or sustain lung injury and have long-term consequences. The preterm lung, despite being immunologically immature, can mount a proinflammatory cytokine response to a ventilator strategy designed to minimize injury.