Main

The outcomes for preterm infants with RDS have improved during the last 30 y by innovations in neonatal care that include mechanical ventilation for infants, antenatal glucocorticoids, and surfactant treatments (1–3). Before these therapies were commonly available, Gregory et al.(4) described the use of a simple device to provide CPAP as a way to maintain lung gas volumes in preterm infants with RDS. The use of CPAP as a primary therapy for RDS has become less frequent in many centers because of mechanical ventilators, although it is used to facilitate extubation and to treat apnea of prematurity (5). The major pulmonary morbidity resulting from the treatment of RDS is BPD, which is frequent in infants with birth weights less than 1 kg (6). The frequency of BPD varies among neonatal units, and its occurrence has been associated by multivariant analysis with intubation and ventilation on the first day of life (7, 8). In several epidemiologic studies, avoidance of mechanical ventilation and the increased use of CPAP have been associated with a decrease in BPD (9, 10). Therefore, there is a renewed interest in the use of CPAP to facilitate the initiation of spontaneous breathing and to decrease the need for mechanical ventilation of preterm infants.

BPD develops in very preterm infants because lung injury is superimposed on an immature lung in the saccular stage of development (11, 12). The initial inflammation and subsequent repair processes interfere with normal lung development and result in a variable pattern of alveolar simplification, fibrosis, and chronic inflammation (13, 14). We found that ventilation of preterm lambs from birth initiates an inflammatory response characterized by recruitment of neutrophils to the lungs, increased hydrogen peroxide, and proinflammatory cytokine expression (15). The mechanisms by which CPAP might protect the preterm lung from injury have not been investigated in the human or in animal models. Therefore, we evaluated the effects of CPAP in comparison with mechanical ventilation on the lungs of preterm lambs.

METHODS

Animal model.

The animal protocols were approved by the animal use committees of the Western Australia Department of Agriculture and The Cincinnati Children's Hospital Research Foundation. We tried to use CPAP to assist spontaneous breathing in preterm lambs delivered by cesarean section, but the lambs did not breathe spontaneously. To better prepare the lambs to breathe, we treated Merino ewes carrying twins at 132–134 d gestational age with an inhibitor of progesterone synthesis (epostane; Sanobi-Synthelabo, Malvern, PA, U.S.A.) and 6 mg betamethasone (Celestone Chronodose; Schering-Plough, New South Wales, Australia) to induce preterm labor (16, 17). The epostane (20 mg) was dissolved in 70% ethanol and given to the ewe by i.v. injection (17). The lambs delivered spontaneously approximately 36 to 50 h later. We cut the cord and sampled cord arterial blood when possible. The lambs randomized to study without breathing received a lethal dose of 50 mg/kg pentobarbital by i.v. injection immediately after birth and before breathing was initiated. The other lambs were dried, intubated with a cuffed 4.5-mm endotracheal tube, and randomized to CPAP or mechanical ventilation. CPAP was delivered using a disposable bubble CPAP device provided by Fisher-Paykel (Auckland, New Zealand) using humidified 40% oxygen and with the CPAP set to 5 cm H2O. Mechanical ventilation was with 40% humidified oxygen using a Bourns BP-100 infant ventilator using a rate of 40 breaths/min, an inspiratory time of 0.7 s, a positive end-expiratory pressure of 4 cm H2O, and a peak inspiratory pressure adjusted to target a Pco2 of 40 mm Hg (18). Tidal volumes were monitored continuously with Florian monitors (Acutronic Medical Systems AG, Hirzel, Switzerland). The animals on CPAP received no sedation, and the ventilated animals received ketamine (10 mg/kg, i.m.) as needed to suppress spontaneous breathing. Umbilical arterial catheters were used to sample blood, and body temperature was maintained with radiant warmers and heat lamps.

Processing of lungs.

After 2 h of spontaneous breathing with CPAP or mechanical ventilation, lambs were given 100% oxygen, and, concurrently with a lethal dose of pentobarbital, the endotracheal tube was clamped for 3 min to permit oxygen absorption (18). The thorax of the lamb was opened, the lungs were inflated with air to 40 cm H2O pressure for 1 min, and the maximum lung volume was recorded (18). The pressure measured with a water manometer was sequentially lowered to 20, 15, 10, 5 and 0 cm H2O, and lung volumes were recorded after 30 s at each pressure. Volumes were corrected for the compliance of the system and were expressed as milliliters per kilogram body weight. Tissue from the right lower lobe was immediately frozen in liquid nitrogen for RNA isolation. An alveolar wash of the left lung was performed with 0.9% NaCl at 4°C, and the lavage was repeated five times (19). Alveolar wash samples were pooled, and aliquots were saved for measurement of Sat PC, total protein, cell number, and differential count.

Alveolar wash samples were centrifuged at 500 ×g for 10 min, and the pellet was suspended in PBS. Cells were stained with trypan blue and counted. Differential cell counts were performed on cytospin preparations after staining with Diff-Quick (American Scientific Products, San Diego, CA, U.S.A.). The activation state of the alveolar wash cells recovered by centrifugation was assessed without stimulation by measuring hydrogen peroxide using an assay based on the oxidation of ferrous iron (Fe2+) to ferric iron (Fe3+) by hydrogen peroxide (Bioxytech H2O2-560 Assay; 0XIS International, Portland, OR, U.S.A.) (20). The total protein in the alveolar wash was measured (21), and wet to dry weight ratios were determined after drying pieces of the right lower lobe at 80°C for 96 h.

Cytokine mRNA.

Total RNA was isolated from tissue from the right lower lobe and from cell pellets of the alveolar wash samples by guanidinium thiocyanate-phenol-chloroform extraction. Ribonuclease protection assays were performed with total RNA from lung tissue and cell pellets (22). Briefly, RNA transcripts of ovine interleukins (IL-1β, IL-6, IL-8, IL-10), TNF-α, and ovine ribosomal protein L32 as a reference RNA were synthesized with [32P]uridine triphosphate (Life Sciences Products, Boston, MA, U.S.A.) using SP6 or T7 polymerase (Ribonuclease Protection Assay 111; Ambion, Austin, TX, U.S.A.). Aliquots of 10 μg of RNA were incubated with excess radiolabeled probes for cytokines and L32 at 55°C for 18 h, and single-stranded RNA was digested with RNase A/RNase Ti (Ambion). Protected fragments were electrophoresed on a 6% polyacrylamide-urea sequencing gel and visualized by autoradiography. Densities of the protected bands were quantified on a PhosphorImager (Molecular Dynamics Inc., Sunnyvale, CA, U.S.A.), using ImageQuant Software (Molecular Dynamics).

Saturated phosphatidylcholine.

Sat PC was isolated from chloroform-methanol (2:1) extracts of alveolar washes by neutral alumina column chromatography after exposure of lipid extracts to osmium tetroxide (23). Sat PC was quantified by phosphorus assay (24).

Histologic scoring of lung tissue.

The right upper lobe was inflation fixed with 10% formalin maintained at 30 cm H2O pressure. Paraffin tissue sections of 5-μm thickness, stained with hematoxylin and eosin, were graded for the degree of inflammation by scoring three sections from each animal in a blinded fashion (25). The score was 0 if there were no inflammatory cells in tissue or airspaces; 1 if a few cells were identified; 2 if moderate cell infiltration was present; and 3 if there were a large number of inflammatory cells in airspaces and tissue. Airspaces and tissue were assessed separately. Average scores were calculated for each animal.

Data analysis.

Results are presented as mean ± SEM. ANOVA followed by the Tukey multiple comparison procedure was used for comparisons of the three groups. Two-tailed unpaired t tests were used for two group comparisons. Significance was accepted at p < 0.05.

RESULTS

Lung function.

There were no differences in cord blood pH for the three groups of animals (Table 1). The CPAP animals were smaller than the animals randomized to mechanical ventilation. One of the control lambs was heavily meconium stained, and another lamb was not successfully intubated at birth and was asphyxiated. These animals are not included in the analyses. The lambs on bubble CPAP initiated spontaneous ventilation after birth, had minimal retractions, were not apneic, and maintained a mean respiratory rate of approximately 50 breaths/min. The Pco2 at 15 min of age was 70 ± 2 mm Hg, and the Pco2 decreased to approximately 60 mm Hg for the rest of the 2-h study (Fig. 1). The low pH values reflected the elevated Pco2 values in the CPAP group. The Pco2 values for the ventilated animals averaged 40 mm Hg by design, and these ventilated lambs had higher pH values than the animals treated with CPAP. The ventilated lambs received 24 ± 2 cm H2O PIP at 30 min of age and 19 ± 1 cm H2O PIP at 2 h of age. The calculated mean airway pressures were 12 cm H2O at 30 min and 11 cm H2O at 2 h of age. The average tidal volume for the 2 h of mechanical ventilation was 9.6 ± 0.3 mL/kg, and the tidal volume did not vary with age after birth. Oxygenation in both groups of lambs receiving 40% oxygen was quite variable and not different between groups. The average value for the CPAP group was 75 ± 13 mm Hg and 107 ± 44 mm Hg for the ventilated group at 2 h.

Table 1 Description of animals
Full size table
Figure 1
figure 1

Blood pH and Pco2 values for the cord blood and for 120 min after birth. The pH and Pco2 values were different between the ventilation and CPAP groups at all times except for the cord blood values at 0 time (p < 0.05). SEM indicators are contained within the data points when no SEM is given on the graphs.

Full size image

Pressure-volume curves and surfactant.

The deflation limbs of the pressure-volume curves for the lambs treated with CPAP or mechanical ventilation were parallel, but the lung volumes were higher for the CPAP animals (Fig. 2). The amount of Sat PC recovered by alveolar wash was 5.2 ± 1.0 μmol/kg for the unventilated lambs, 7.7 ± 0.9 μmol/kg for the CPAP animals, and 7.1 ± 1.8 μmol/kg for the ventilated animals. The amounts of Sat PC in alveolar washes were not different for the CPAP and ventilated groups, but both these groups had higher values than the unventilated lambs (p < 0.05).

Figure 2
figure 2

Deflation pressure-volume curves for the CPAP lambs and the ventilated lambs. All volumes between 10 and 40 cm H2O pressure are higher for the CPAP lungs than for the ventilated lungs (p < 0.05).

Full size image

Indicators of injury.

There were no differences in lung wet to dry weight ratios or total protein recovered by alveolar washes among the lambs that were not ventilated, the mechanically ventilated lambs, and the CPAP lambs (Table 2). The numbers of lymphocytes and monocytes in alveolar lavages also were similar for the three groups (Fig. 3A). However, neutrophils were not detected in alveolar washes from the unventilated animals and were 6.6 times higher in the mechanically ventilated lambs than in the CPAP lambs (Fig. 3B). The cells from the alveolar washes of the unventilated lambs and the CPAP lambs contained similar amounts of hydrogen peroxide, and the hydrogen peroxide content was higher for the mechanically ventilated lambs than for the CPAP lambs. The mRNA expression in lung tissue for IL-1β, IL-6, and IL-8 was not different among the groups (Table 2), and no mRNA was detected for IL-4, IL-10, or TNF-α in the lung tissue. There also was no significant increase in IL-1β, IL-6, or IL-8 mRNA in cell pellets recovered from the alveolar washes for the CPAP or mechanically ventilated lambs relative to the unventilated control lambs, although values tended to be higher for the mechanically ventilated lambs. There was not severe edema or cellular infiltrates in any of the lungs. The mean histologic score of 1.2 ± 0.2 for the ventilated lungs indicating a few inflammatory cells were consistently found in the tissue and airspaces was increased relative to the value of 0.8 ± 0.1 for the unventilated lungs (p < 0.2 by t test). The CPAP lambs had a histologic score of 0.9 ± 0.1, which was not different from the unventilated lungs.

Table 2 Indicators of lung injury
Full size table
Figure 3
figure 3

Lymphocytes and monocytes in alveolar washes (A), neutrophils in alveolar washes (B), and hydrogen peroxide (H2O2) in cells from alveolar washes (C). Lymphocyte and monocyte numbers were equivalent for the three groups. There were fewer neutrophils in the alveolar washes of the CPAP lambs than the mechanically ventilated lambs. Less H2O2 was in the cells from the alveolar washes of the CPAP lambs than the ventilated lambs (p < 0.05 vs ventilation).

Full size image

DISCUSSION

CPAP is used primarily in conjunction with surfactant treatments to decrease the need for mechanical ventilation or to facilitate early extubation from mechanical ventilation (5, 26). CPAP will increase the functional residual capacity of the lung and may increase respiratory drive. The avoidance of intubation in the delivery room and the use of CPAP were associated with a decrease in BPD in several clinical reports (9, 10, 27). By multivariant analyses, intubation and mechanical ventilation of preterm infants is the major predictor of subsequent BPD (8). The mechanisms responsible for the possible effects of CPAP to decrease BPD have not been evaluated. A possible mechanism is the avoidance of aggressive initiation of gas exchange with high tidal volumes and inadvertent hyperventilation that occurs in intubated infants (28, 29). CPAP also protects the infant from standard mechanical ventilation and inadvertent over or under ventilation as well as airway injury and colonization from the endotracheal tube. These benefits are consistent with the concepts that BPD results from repetitive inflammatory stresses on the preterm lung (30, 31). The initiation of ventilation, especially with high tidal volumes, results in the recruitment of activated granulocytes to the preterm sheep lung (32). Mechanical ventilation of the surfactant-treated preterm baboon lung results in persistent elevations of inflammatory cells and proinflammatory cytokines in the lungs (30). The effects of CPAP on neonatal transition or on inflammation in the lung have not been evaluated.

We failed in initial attempts to study CPAP in the preterm sheep because the lambs would not breathe consistently after delivery by cesarean section. Attempts at pharmacologic stimulation at delivery with intravascular injections of catecholamines, cortisol, and caffeine did not help. Therefore, we thought that the process of labor and spontaneous delivery might prepare the lamb to breathe spontaneously. To mimic the clinical situation we gave antenatal betamethasone and an inhibitor of progesterone synthesis to induce preterm labor (16, 17). The ewes delivered without incident, and the preterm lambs had normal cord pH values and began breathing spontaneously.

We intubated the lambs to avoid problems with secretions and position-dependent delivery of the CPAP to assure effective delivery of the CPAP. We used a commercial version of the bubble CPAP system used by Dr. Jen Wung at Columbia (33). We used 5 cm H2O pressure because most clinical reports use this pressure, although there is minimal information about the optimal pressures for CPAP in infants (4, 9, 26). The chests of the lambs visibly vibrated using a gas flow of 10 L/min. The lambs breathed comfortably and were not distressed despite their self-regulation of Pco2 to approximately 60 mm Hg. The lambs did not try to normalize Pco2 values and thus had a respiratory acidosis. Infants managed by avoidance of intubation in the delivery room and with CPAP also have higher Pco2 values than is generally considered normal in the first minutes to hours of life (26, 27). We limited this initial experiment to 2 h because we did not know how long the animals would survive on CPAP and because indicators of lung injury are measurable within 2 h of preterm birth (34). Longer experiments might better resolve differences between lambs supported with mechanical ventilation and CPAP.

We used conventional ventilation with a target Pco2 of 40 mm Hg for the ventilated lambs. This Pco2 was chosen as a normal value because we did not know how the CPAP lambs would regulate Pco2. The animals had RDS because they were ventilated with PIPs of 19 ± 2 cm H2O at 2 h of age and had low pool sizes of Sat PC (35). The tidal volumes given to the lambs were higher than generally recommended for preterm infants, but this tidal volume is no more injurious than high-frequency oscillation in preterm lambs (36). The higher tidal volume is needed to normalize Pco2 in lambs in part because these animals have higher dead space than infants. Interesting comparisons for the future would be to compare lung injury in ventilated and CPAP lambs using equivalent Pco2 targets or mean airway pressures. The CPAP lambs and mechanically ventilated lambs had the same low alveolar pool sizes of surfactant lipid. This measurement indicates equivalent degrees of lung maturation in both groups. The difference in the amount of surfactant lipid between the unventilated lambs and the other groups resulted from secretion with breathing (37).

Despite equivalent surfactant pool sizes, the CPAP group had fewer neutrophils, less hydrogen peroxide, and larger lung gas volumes than did the mechanically ventilated lambs. There were no differences in proinflammatory cytokine expression between the ventilated groups, although the mechanically ventilated group tended to have higher cytokine mRNA expression than did the CPAP group. This lack of difference may result from the small number of lambs in each study group. The findings of more neutrophils and hydrogen peroxide is consistent with the initiation of more injury with mechanical ventilation as both indicators are early changes that occur with ventilator-induced lung injury (38). The CPAP lambs and conventionally ventilated lambs had similar wet to dry lung weights, amounts of protein in alveolar lavages, and surfactant pool sizes. Nevertheless, lung compliance as evaluated by the pressure-volume curves was better for the CPAP lambs. Possible explanations are that inflation was more uniform and complete in the lambs exposed to 2 h of CPAP or that CPAP preserved surfactant function better than did mechanical ventilation. Ventilation can degrade surfactant function (39), and an unanswered question is whether CPAP more effectively preserves surfactant function.

The amount of inflammation that occurred with the initiation of ventilation by mechanical ventilation is less than we previously reported in more immature preterm lambs that were treated with surfactant (15). A probable explanation is that endogenous surfactant more effectively minimizes regional overdistension than does surfactant treatment because of the uniformity of distribution of the endogenous surfactant. The antenatal glucocorticoids also may have suppressed the inflammatory response of the fetal lung. Another possibility is that the higher Pco2 and associated respiratory acidosis may have minimized inflammation in the CPAP group. Although using much higher Pco2 values, Laffey et al.(40) found that elevated CO2 protected the isolated and perfused lung from reperfusion injury, and Broccard et al.(41) demonstrated similar effects on inflammation with ventilation-induced injury. This study demonstrates that CPAP can be studied in preterm lambs and indicates that CPAP induces less of an initial inflammatory response in the preterm lung than does mechanical ventilation to normal Pco2 values.