INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ventilation strategies that optimize functional residual capacity (FRC) and avoid overdistention of the lungs minimize lung injury (1). Ventilation of the lung from lung volumes below a normal FRC or to volumes that exceed total lung capacity (TLC) cause lung injury and damage the surfactant system (2, 3). Inadequate alveolar expansion depletes lamellar body and alveolar surfactant pools in saline-lavaged and surfactant-treated rabbits (2). High tidal volume (VT) ventilation promotes the conversion of biophysically active large-aggregate surfactant to inactive vesicular surfactant in lung-injured adult rabbits and the rate of conversion increases with severity of injury (3, 4). In contrast, the amount of conversion of the surfactant used for treatment was not increased in preterm lungs injured with high VT (5). Alveolar surfactant function also can be disturbed by protein-rich edema fluid and products of inflammation (6). Ventilation strategy is an important variable for optimizing the clinical response to surfactant treatment of the injured adult lung and the preterm surfactant-deficient lung (2, 3, 7). A less well studied variable is the type of surfactant to be used for surfactant treatment (8). Surfactants that contain lipids only are not very effective at improving compliances or pressure-volume curves of surfactant-deficient preterm rabbits (7, 9). Addition of the surfactant proteins (SP)-B or SP-C (or both) improves lung function if the preterm rabbits are ventilated with 3 to 4 cm H₂O positive end-expiratory pressure (PEEP), but no surfactant response occurs without PEEP (7, 9). The use of PEEP in preterm rabbits also results in striking decreases in the leak of albumin from the vasculature to the lungs (8). We previously evaluated the efficacy of a synthetic surfactant containing recombinant human SP-C (rSP-C) in surfactant-deficient preterm rabbits and lambs (10). Compliances of preterm rabbits treated with rSP-C surfactant and ventilated with 3 cm H₂O PEEP were equivalent to compliances of natural surfactant-treated animals. However, the rSP-C was not effective for a 15-min period of ventilation of preterm rabbits in the absence of PEEP. A similar sensitivity to PEEP was reported by Hawgood and coworkers (11) with other recombinant SP-C-containing surfactants. The effects of different levels of PEEP have been systematically investigated in normal and injured adult lungs (1, 12) but not in surfactant-treated preterm lungs. Therefore, we asked if different levels of PEEP altered the clinical responses of preterm lambs to treatment with rSP-C surfactant and if PEEP influenced the surfactant system.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Surfactant

The rSP-C, with the 34-amino-acid human SP-C sequence altered by replacement of cysteine in positions 4 and 5 with phenylalanine and methionine in position 32 by isoleucine, was expressed in bacteria and purified. The rSP-C surfactant contained 2% rSP-C in phospholipids (dipalmitoylphosphatidylcholine [DPPC] and palmitoyloleoylphosphatidylglycerol in a 70:30 [wt/wt] ratio), plus 5% palmitic acid (Byk Gulden, Konstanz, Germany) (10). The amino acid substitutions stabilize the protein from aggregation and denaturation. The rSP-C surfactant was suspended in 0.9% NaCl to a concentration of 25 mg lipid/ ml with a Pasteur pipette. Sheep surfactant was recovered from fresh saline lavages of adult sheep by a series of centrifugation steps that isolate predominately the large-aggregate, highly surface-active fractions of surfactant (13). This surfactant is very effective when used to treat surfactant-deficient lungs. The sheep natural surfactant was concentrated to 25 mg lipid/ml before treatment.

Preterm Lambs

Delivery and ventilation of lambs. Preterm lambs were delivered by cesarean section as previously described (5). Briefly, each pregnant ewe at 126 d gestation carrying twins or triplets was preanesthetized with ketamine (20 mg/kg intramuscularly) and given spinal-epidural anesthesia (10 ml 2% lidocaine and 0.5% bupivacaine [1:1, vol/vol]). The fetal head was exposed through midline abdominal and uterine incisions, and the lamb was anesthetized with ketamine (10 mg/kg intramuscularly) plus acepromazine (0.1 mg/kg intramuscularly). After local anesthesia to the anterior neck with 2% lidocaine, a tracheal tube with 4.5 mm interior diameter was tied into the trachea, freely flowing fluid from fetal airways was removed by syringe, and the endotracheal tube was clamped (5). The umbilical cord was then cut and the fetus was delivered and weighed. Immediately after birth and before breathing, each preterm lamb was treated with rSP-C surfactant (n = 21) or with natural surfactant (n = 5) at a dose of 100 mg lipid/kg body weight. The surfactant contained 1 µCi of [³H]choline-labeled DPPC (5). The lambs were then randomized to PEEP values of zero, 4 or 7 cm H₂O. The natural surfactant-treated lambs were ventilated without PEEP. The animals were ventilated for 7 h with time-cycled, pressure-limited ventilators (Sechrist Industries, Anaheim, CA) with initial setting of fraction of inspired oxygen (FI_O2) = 1, rate = 40 breaths/min, flow = 8 L/min, and inspiratory time = 0.7 s. Peak inspiratory pressures (PIP) were adjusted to achieve a target VT of 10 ml/kg using continuous VT monitoring (CP-100; Bicore Monitoring Systems, Anaheim, CA). Subsequently, only FI_O2 and PIP were changed in order to maintain the arterial PO₂ in the range of 100 to 200 mm Hg and the Pa_CO2 at ~ 50 mm Hg. A catheter was placed in the distal aorta via an umbilical artery, and 5% dextrose was infused at 100 ml/kg/d. Each lamb received a transfusion of 10 ml/kg cord blood soon after birth. The arterial line was used for blood pressure monitoring and intermittent blood gas analysis. To assess the intravascular-to-interstitial and alveolar albumin leak, 5 µCi [¹²⁵I]albumin was given via the arterial catheter at 6 h of age, and 1 h later a final blood sample was obtained for measurements of hemoglobin and radiolabeled albumin (14). VT were measured every hour with a pneumotachometer, and dynamic compliances were calculated as VT normalized to body weight and divided by the ventilatory pressure (PIP  PEEP) (14). The rate of 40 breaths/min and inspiratory time of 0.7 s were selected to avoid PEEP resulting from inadequate time for expiration (auto-PEEP). The absence of auto-PEEP was verified by clamping the ventilatory circuit distal to the pressure monitor at end expiration and measuring pressure. The volume above FRC at PEEP zero was estimated by measuring with a pneumotachometer the volume of gas that left the lung at end expiration when the ventilator circuit was disconnected. Body temperature was maintained with radiant warmers and heating pads. Supplemental ketamine (10 mg/kg intramuscularly) and acepromazine (0.1 mg/kg intramuscularly) were given as necessary to suppress spontaneous ventilation. After 7 h of ventilation, each animal was ventilated briefly with 100% oxygen, deeply anesthetized with sodium pentobarbital (25 mg/kg intravenously), and the endotracheal tube was clamped for 3 min to permit oxygen absorption. The animal was exsanguinated by cutting the abdominal aorta.

Pressure-volume curves and lung processing. The thorax was opened, lungs were inflated with air to 40 cm H₂O pressure for 1 min, and maximal lung volume was recorded (15). Then the pressure was lowered sequentially to 20, 15, 10, 5, and zero cm H₂O, with lung volumes recorded 30 s after each pressure was reached. Volumes were corrected for the compliance of the system.

The lungs were removed from the thorax and an alveolar wash was performed by filling the lungs by gravity with 0.9% NaCl at 4° C until visually distended, and the alveolar wash was recovered with a syringe (14). The procedure was repeated five times; the five washes were pooled and aliquots were saved for determinations of saturated phosphatidylcholine (Sat PC), recovery of [³H]DPPC, total protein, and [¹²⁵I]albumin. The lungs were homogenized in 250 ml 0.9% NaCl, and aliquots of homogenates were used for measurements of Sat PC, [³H]DPPC, [¹²⁵I]albumin, and total blood content by hemoglobin assay. Sat PC was recovered from chloroform-methanol (2:1) extracts by neutral alumina column chromatography after exposure to osmium tetroxide and was quantified by phosphorus assay (16, 17). Protein was determined using the method of Lowry and coworkers (18).

After an initial centrifugation of alveolar wash at 140 × g for 10 min, large- and small-aggregate surfactant fractions were separated by centrifugation at 40,000 × g for 15 min (19). The supernatant that contained small-aggregate surfactant was used for Sat PC and [³H]Sat PC content measurements. The small-aggregate fraction also was concentrated by ultrafiltration and used for surface tension measurements. The pellet containing large-aggregate surfactant was resuspended in saline and centrifuged at 40,000 × g for 15 min over 0.8 M sucrose. The interface was aspirated, diluted with saline, and again centrifuged at 40,000 × g for 15 min. The large-aggregate surfactant then was resuspended in a small amount of saline, and aliquots were used for Sat PC and [³H]Sat PC content measurements. The large- and small-aggregate surfactant fractions from three lambs for each group were used for minimal surface tension measurements using the captive bubble surfactometer (20). Small-aggregate surfactants were concentrated by ultracentrifugation at 100,000 × g for 48 h. Three-microliter suspensions of large- and small-aggregate surfactant fractions containing 15 µmol Sat PC were injected on the surface of a 25-µl bubble by syringe. Sixty seconds after surfactant application the bubble was oscillated and minimal surface tensions were measured after a change in surface area of 35%.

Data Analysis

Results are given as means ± SEM. Analysis of variance (ANOVA) was used for comparison of differences between the PEEP groups treated with the rSP-C surfactant, and the Student-Newman-Keuls test was used as the discriminating post-test. Comparisons between rSP-C surfactant- and natural surfactant-treated lambs at zero PEEP were made with two-tailed t tests. Significance was accepted at p < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Description of Lambs

Twenty-one lambs at 126 d gestational age were treated with rSP-C surfactant and randomized to PEEP levels of zero, 4, or 7 cm H₂O (Table 1). In addition, five animals were treated with natural surfactant and ventilated with PEEP for comparison with the PEEP zero rSP-C surfactant-treated group. The lambs had similar body weights, cord blood pH, and Pa_CO2 values.

Respiratory Outcomes

Two animals from the PEEP zero rSP-C surfactant-treated group died of pneumothorax before the study was completed. Pa_CO2 and VT were similar for all groups for the 7-h study period (Figure 1). The PIP required to achieve comparable Pa_CO2 and VT values were quite stable for the 7-h ventilation period. Mean PIP values for the zero PEEP rSP-C surfactant and natural surfactant were 29 ± 2 cm H₂O and 23 ± 1 cm H₂O, respectively, and were not different. The values were the same after 6 and 7 h ventilation. Mean PIP values for lambs ventilated with 4 cm H₂O PEEP were 19 ± 1 cm H₂O and were 17 ± 1 cm H₂O for lambs ventilated with 7 cm H₂O PEEP, values that were not different from each other but significantly lower than for the PEEP zero groups. Oxygenation was not stable for the 7-h period of ventilation (Figure 2) and was not different between the groups at 7 h because of the variability. However the mean Pa_O2/FI_O2 ratios for the 7-h period of ventilation of 73 ± 6 for the PEEP zero rSP-C group increased significantly to 186 ± 7 with 4 cm H₂O PEEP (p < 0.05) and increased further to 262 ± 20 with 7 cm H₂O PEEP (p < 0.05 versus both zero and 4 cm H₂O PEEP). The mean Pa_O2/FI_O2 ratios for lambs treated with rSP-C surfactant and natural surfactant and ventilated with PEEP zero were not different. Compliance values were similar for the PEEP zero groups after 1 h of ventilation, and 4 and 7 cm H₂O PEEP resulted in mean increases in compliance of approximately 75% from the PEEP zero groups. The volumes of gas that left the lung on disconnection of the ventilation circuit were zero for the PEEP zero groups, 7.0 ± 0.1 ml/kg for the PEEP 4 group, and 11.5 ± 0.6 ml/kg for the PEEP 7 group.

View larger version (25K): [in this window] [in a new window]   Figure 1.   Sequential measurements of PaCO2, VT, and ventilatory pressure (PIP  PEEP). The ventilatory goals were to keep PaCO2 between 50 and 60 mm Hg using VT of 10 ml/kg, and there were no differences between groups. The ventilatory pressures required to achieve the PaCO2 and VT goals were initially lower for the natural sheep surfactant-treated lambs (PEEP 0-Nat) than for the PEEP zero group. The PEEP 4 and PEEP 7 groups had consistently lower ventilatory pressures. *p < 0.05 versus PEEP 0, tp < 0.05 versus PEEP 4.

View larger version (28K): [in this window] [in a new window]   Figure 2.   Sequential measurements of the ratios of arterial PO2 to inspired O2 fractions (PaO2/FIO2) and dynamic compliances. PaO2/ FIO2 ratios were higher for the PEEP 4 and PEEP 7 groups and for the PEEP zero natural surfactant group than for the PEEP zero rSP-C group at selected times. At 3 and 5 h the PEEP 7 group had higher ratios than all other groups. *p < 0.05 versus PEEP 0, tp < 0.05 versus PEEP 4. Dynamic compliances were significantly higher for animals ventilated with PEEP 4 and PEEP 7. The 1-h compliance also was higher for lambs treated with natural surfactant (PEEP 0-Nat) than rSP-C-treated lambs (PEEP zero). *p < 0.05 versus PEEP 0.

Lung volumes on the deflation limbs of the pressure-volume curves were significantly higher for the lambs ventilated with 4 and 7 cm H₂O PEEP than for animals ventilated without PEEP at all pressures, except zero cm H₂O (Figure 3). Maximal lung volumes measured at 40 cm H₂O were higher with PEEP 7 (49 ± 3 ml/kg) than with 4 cm H₂O PEEP (39 ± 2 ml/kg). The maximal lung volumes after 7 h ventilation using zero PEEP and treatment with rSP-C or natural surfactant were not different. Lung volumes at 5 cm H₂O pressure remained high for the PEEP 4 and PEEP 7 groups, indicating good surfactant function.

View larger version (18K): [in this window] [in a new window]   Figure 3.   Deflation limbs of pressure-volume curves for preterm lambs treated either with rSP-C or natural surfactant. Lung volumes were higher for lambs ventilated with PEEP 4 and 7 relative to the PEEP zero group. There was also a difference in maximal lung volumes between the PEEP 7 and PEEP 4 groups. *p < 0.05 versus PEEP zero, tp < 0.05 versus PEEP 4.

Surfactant Pool Sizes

The alveolar pool of Sat PC for the PEEP zero group was lower (11 ± 5 µmol/kg) than for the PEEP 7 group (22 ± 2 µmol/kg) (Figure 4). The Sat PC pool sizes were similar for the rSP-C- and natural surfactant-treated groups. The recovery of [³H]DPPC in the total lungs was normalized to 100% to evaluate the distribution of the DPPC from the surfactant treatment. The percent recovery of [³H]DPPC in alveolar lavage was significantly higher for the PEEP 7 group than for the PEEP 4 group or the PEEP zero group. Both the natural and rSP-C surfactants used to treat the lambs had greater than 95% of Sat PC recovered as large-aggregate surfactant. The Sat PC recovered in the large-aggregate surfactant from alveolar washes after 7 h ventilation was 28 ± 4% for the PEEP zero rSP-C surfactant group, and this value was less than the recoveries for the other groups (Figure 5). The pool sizes of large-aggregate surfactant estimated by multiplying the percent large aggregates by the pool size increased for all groups relative to the PEEP zero rSP-C surfactant group with the largest pool size measured for the PEEP 7 rSP-C group.

View larger version (31K): [in this window] [in a new window]   Figure 4.   Sat PC pool sizes and recoveries of [3H]DPPC in alveolar washes (AW), lung homogenates (LH), and total lungs (TL). Ventilation with PEEP 7 resulted in a higher recovery of Sat PC in AW. The recovery of [3H]DPPC in the total lung was normalized to 100%. Ventilation with PEEP 7 increased the recovery of [3H]DPPC in AW and decreased recovery in lung tissue relative to PEEP zero and PEEP 4. *p < 0.05 versus PEEP zero, tp < 0.05 versus PEEP 4.

View larger version (14K): [in this window] [in a new window]   Figure 5.   Percent recovery of large-aggregate surfactant (LA) fractions and pool sizes of saturated phosphatidylcholine (Sat PC) in the LA. PEEP increased the percent LA and LA pool sizes relative to PEEP zero. The recovery of LA from natural surfactant-treated lambs (PEEP 0-Nat) also was increased. *p < 0.05 versus PEEP zero.

Radiolabeled Albumin and Protein

The amount of protein in alveolar washes was higher for the PEEP zero group (96 ± 20 mg/kg) than for the PEEP 4 group (44 ± 8 mg/kg) (Figure 6). The PEEP zero group treated with natural surfactant (43 ± 4 mg/kg) also had a lower total protein recovery than the PEEP zero rSP-C surfactant group. The recoveries of [¹²⁵I]albumin in alveolar lavage and total lungs were similar for all groups.

View larger version (16K): [in this window] [in a new window]   Figure 6.   Total protein in alveolar wash (AW) and percent recovery of [125I]albumin in AW and in total lungs (TL) (AW + lung tissue). Total protein in AW was significantly lower for lambs treated with PEEP 4 than PEEP zero. Lambs treated with natural surfactant and PEEP zero (PEEP 0-Nat) also had lower protein recoveries than the PEEP zero rSP-C surfactant-treated group. *p < 0.05. There were no differences in percent recovery of [125I]albumin given at 6 h by intravascular injection in AW or the total lungs among all groups.

Minimal Surface Tension

The rSP-C used for treatment had a minimal surface tension of 1.6 ± 0.5 mN/m with a 35% surface area compression (Figure 7). All the large-aggregate surfactant fractions recovered after 7 h ventilation had similar and low surface tension values. In contrast the small-aggregate fractions had higher minimal surface tension values than the large-aggregate fractions.

View larger version (26K): [in this window] [in a new window]   Figure 7.   Minimal surface tensions of large-aggregate and small-aggregate surfactant fractions measured with the captive bubble method. The rSP-C value is for triplicate measurements of the surfactant used to treat the lambs. The large-aggregate surfactant fractions had lower minimal surface tensions than did the small-aggregate fractions (p < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

These experiments were designed to explore the interaction between surfactant treatment and PEEP levels in the preterm lung. As anticipated, the use of PEEP resulted in improved oxygenation and compliance. The most striking physiologic difference was the large increase in lung gas volumes after 7 h ventilation in the lambs treated with rSP-C surfactant and ventilated with 4 or 7 cm H₂O PEEP. PEEP also protected the alveolar surfactant pool from loss of Sat PC and more importantly preserved the biophysically active large-aggregate surfactant pool. The PEEP 4 and the PEEP 7 groups had 3 and 4 times larger large-aggregate pool sizes than did the PEEP zero group. Therefore, PEEP was critical to the optimization of the surfactant treatment response, to preservation of lung gas volumes, and to maintenance of the function of the alveolar surfactant.

These results in the preterm surfactant-deficient lung are similar to previous reports of injured adult lungs. Ventilation of saline-lavaged lungs from lung volumes below a normal FRC resulted in severe loss of gas volume, cytokine release, and edema (21). Surfactant treatment was effective only if PEEP was used to maintain the FRC with conventional ventilation or if a lung volume recruitment strategy with high mean airway pressures was used for high-frequency oscillation ventilation (HFOV) (22). The high-volume strategy with HFOV was most effective in maintaining alveolar and lamellar body pools in surfactant-treated rabbits with lung injury (2). In the preterm and surfactant-treated lamb lung ventilatory styles using rates of 15 breaths/min and VT of 15 ml/kg, rates of 60 breaths/min and VT of 8 ml/kg, or HFOV did not result in differences in lung physiology, surfactant metabolism, or indicators of lung injury when the different strategies were designed to minimize injury (25). In contrast, Ito and coworkers found that increases in VT but not PEEP in injured rabbit lungs resulted in accelerated conversion of surfactant into the small-aggregate inactive forms (26). Within the limits of PEEP evaluated here, using a constant VT ventilation strategy, loss of total alveolar surfactant and loss of large-aggregate surfactant was minimized by increasing PEEP.

PEEP had a large effect on TLC as measured by pressure- volume curves after 7 h ventilation. FRC was not measured in this experiment. However, we previously reported an FRC of approximately 25 ml/kg for preterm lambs at a similar gestational age that were treated with the same rSP-C surfactant or natural surfactant and ventilated with a PEEP of 3 cm H₂O (10). The gas volume that left the lungs at end expiration on switching from 4 to zero cm H₂O PEEP was 7 ml/kg. Therefore, an estimate of the residual FRC at PEEP zero is about 18 ml/kg. Ventilation with the 10 ml/kg VT results in a total gas volume at end inspiration of 35 ml/kg, a value similar to the maximal lung volume of 39 ml/kg measured at 40 cm H₂O pressure after 7 h of ventilation. For the PEEP 7 group FRC was estimated to be about 30 ml/kg and lung gas volume at end inspiration was about 40 ml/kg. The maximal lung volume measured from the PEEP 7 group was 49 ml/kg. Therefore, the use of PEEP resulted in ventilation of the lambs at relatively high lung volumes that approached but did not exceed maximal lung volumes at end inspiration. Ventilation without PEEP from an estimated FRC of approximately 18 ml/kg would result in a lung volume of 28 ml/kg at end inspiration, a volume that exceeded the maximal volume of 17 ml/kg measured at 40 cm H₂O from the pressure-volume curve. The occurrence of pneumothorax in two of six animals treated with rSP-C surfactant and zero PEEP is not surprising. The alveolar washes from the PEEP zero animals contained more protein than did the alveolar washes from lambs ventilated with PEEP, but the vascular to alveolar leak of [¹²⁵I]albumin was not increased over the 1-h interval from 6 to 7 h of ventilation. Presumably the injury occurred earlier during the ventilation. High VT with initiation of ventilation in fluid-filled fetal lungs can cause lung injury after as few as 6 breaths or within 30 min of ventilation (5, 27). Considering the loss of volume with ventilation with PEEP zero, it is surprising that lung injury as indicated by protein in alveolar washes was not more striking.

An alternate explanation for the large differences in lung gas volumes is that PEEP facilitated the distribution of surfactant to the lungs. However, we treated the animals by mixing surfactant with fetal lung fluid in the airless lung before initiating ventilation, a strategy that will result in a uniform distribution of surfactant that should be independent of the subsequent use of PEEP (15). However, we did not measure the uniformity of surfactant distribution to the lungs.

The natural sheep surfactant resulted in better compliances than did the rSP-C surfactant at 1 h, but subsequent physiological measurements, most importantly the pressure-volume curves, were not different. This result differs from previous reports in ventilated preterm rabbits where natural surfactant that contains SP-A is uniquely capable of improving lung compliance and protecting the lung from injury (7, 28, 29). However, those assessments were over ventilation intervals of 15 or 30 min. With prolonged ventilation, PEEP is necessary to prevent the volume loss after natural surfactant treatment. The natural surfactant did have a higher percent large-aggregate surfactant than did rSP-C surfactant and might perform better with prolonged ventilation with PEEP 0. Overall, the rSP-C surfactant performed extremely well as a treatment for the preterm lamb.

We studied a PEEP of 4 cm H₂O to approximate the value of 3 to 4 cm H₂O used most frequently for surfactant-treated infants with respiratory distress syndrome (RDS). The PEEP of 7 cm H₂O was chosen to be above the normal clinical range. We were surprised that the PEEP 7 group was essentially equivalent to the PEEP 4 group in terms of lung mechanics but PEEP 7 resulted in improved oxygenation and better preservation of active surfactant pools. The estimate of FRC of about 30 ml/kg is approximately 60% of the maximal volume that achieved a high-volume ventilation strategy using conventional ventilation. In patients with acute respiratory distress syndrome, PEEP levels have been selected by evaluating oxygenation or by assessing the opening pressure on inflation pressure-volume curves (30). In infants, inflation pressure-volume curve maneuvers have not been attempted to our knowledge, and oxygenation responses are complicated by variable foramen ovale and ductal shunts. Optimization of PEEP levels have not been rigorously attempted in the surfactant-treated preterm. The optimal PEEP will depend on the type of surfactant used for treatment (8) and will have the combined benefits of decreasing injury, improving clinical responses, and preserving surfactant function.