Main

After premature birth, respiratory failure is the result of both structural and functional immaturity. Surfactant deficiency is a primary defect in HMD and RDS, and treatment with exogenous surfactant can cause dramatic improvement in oxygenation(1). However, exogenous surfactant results in suboptimal responses in up to 50% of human newborns with RDS, suggesting that other mechanisms may contribute to failure of postnatal adaptation after premature birth(1). Human neonates with severe or fatal RDS often have pulmonary hypertension and poor gas exchange(2,3). Mechanisms that lead to an increase in PVR and the role of pulmonary hypertension in outcome are uncertain.

Premature delivery and ventilation of the ovine fetus has provided a useful animal model for the study of HMD and RDS(4). Severe HMD in premature lambs is characterized by elevated PVR, impaired gas exchange, pulmonary edema, and lung inflammation(4–6). Vasoactive mediators, such as endogenous nitric oxide or ET-1, may contribute to changes in PVR in the fetal lung. Inhaled nitric oxide improves oxygenation, increases pulmonary blood flow without increasing pulmonary edema, lowers PVR, and decreases lung neutrophil accumulation in severe experimental HMD(5,6); however, the role of other vasoactive substances in the pathophysiology of HMD is unknown. ET-1 is a potent vasoconstrictor peptide which modulates basal PVR in the normal late gestation ovine fetus and contributes to high PVR after chronic intrauterine pulmonary hypertension(7–10). ET-1 levels are elevated in the human fetal circulation at 18-24 wk of gestation in comparison with maternal levels(11). However, it is uncertain whether ET-1 contributes to high PVR in experimental HMD after delivery of severely premature lambs.

We hypothesized that ET-1 contributes to high PVR in the prematurely delivered lamb and that its effects are mediated through the ETA receptor. Therefore, using a model of HMD and acute lung injury caused by delivery in the premature fetal lamb, we studied circulating ET-1 levels during prolonged mechanical ventilation with hyperoxia and assessed the effects of a selective ETA receptor blocker on pulmonary hemodynamics in this model of severe HMD.

METHODS

Surgical preparation and physiologic measurements. The following methods have been previously described(5,6,9,10). All procedures and protocols were previously reviewed and approved by the Animal Care and Use Committee at the University of Colorado Health Sciences Center. Ten mixed breed (Columbia-Rambouillet) pregnant ewes at 125 d of gestation (term = 147 d) were fasted 24 h before surgery. Ewes were sedated with i.v. pentobarbital sodium (2-4 g) and anesthetized with 1% tetracaine hydrochloride (3 mg) by lumbar puncture. Ewes were kept sedated with pentobarbital but breathed spontaneously throughout the surgery. Under sterile conditions, a uterine incision was made, and the left forelimb of the fetal lamb was delivered. Polyvinyl catheters were advanced into the ascending Ao and superior vena cava after insertion into the axillary artery and vein. A left axillary to sternal thoracotomy exposed the heart and great arteries. A catheter was inserted into the MPA by direct puncture through purse string sutures. Catheters were guided into position with a 14 or 16 gauge i.v. placement unit (Angiocath; Travenol, Deerfield, IL). Catheters were secured by tightening the purse string suture as the introducer was withdrawn. The MPA catheter was inserted between the ductus arteriosus and the pulmonic valve. A LA catheter was inserted in the medial portion of the left atrial appendage. An ultrasonic flow transducer (6 mm, Transonic Systems Inc., Ithaca, NY) was placed around the LPA to measure blood flow to the left lung. The flow transducer cables were attached to an internally calibrated flowmeter (Transonics, Ithaca, NY) for continuous measurements of LPA flow. The absolute values of flows were determined from phasic blood flow signals obtained during baseline periods, as previously described(8,12,13). A correction factor between end-diastolic flow and the internally calibrated zero point on the Transonics flowmeter was added to the mean flow on the Transonics flowmeter. The value obtained from this method correlates with previously determined measures of LPA flow in the late gestation ovine fetal lung(12). Calculation of resistances are reported as left lung PVR (mm Hg mL-1 min-1 = mean MPA pressure - LAP/LPA flow). Study measurements included PAP, AoP, LAP, LPA flow, and arterial blood gas tensions. The Ao, MPA, and LA catheters were connected to a computer driven system to record pressures and flow (Biopac, Santa Barbara, CA). The pressure transducer was calibrated with a mercury column manometer. Heart rate was determined from phasic pulmonary blood flow tracings. Blood samples for pH, PaCO2, and PaO2 were drawn from the Ao and measured at 39.5°C with a Radiometer OSM-3 blood gas analyzer and hemoximeter (Radiometer, Copenhagen).

Ventilator Study (Fig. 1). After stabilization of physiologic parameters (baseline), pancuronium was administered to the fetus (0.1 mg/kg), the fetal head was exteriorized, and a tracheotomy was performed with placement of an endotracheal tube (3.0-mm internal diameter). All animals were treated with exogenous surfactant(Infasurf, kindly provided by E. A. Eagan, M.D.) at an estimated dose of 3 mL/kg (105 mg of phospholipid/kg) before the first breath. Mechanical ventilation was initiated with a continuous-flow, time-cycled, pressure-limited, neonatal ventilator at the following settings: peak inspiratory pressure, 35 cm H2O; positive end expiratory pressure, 6 cm H2O; rate, 30 breaths/min; inspiratory time, 1.0 s; and FIO2, 1.00. After 30 min of mechanical ventilation, the umbilical cord was ligated, and 20 mL/kg normal saline was infused. BQ 123 (1 mg/h; n = 5) or 1% DMSO (n = 5) were continuously infused in the MPA. A 5% dextrose solution was infused to provide 15 mL/h crystalloid and 1 mg kg-1 h-1 pentobarbital.

Figure 1
figure 1

Ventilator algorithm during the 8-h ventilator study.

Full size image

Animals were ventilated for 8 h. Mechanical ventilator settings were modified during the course of studies based on results of preductal arterial blood gas samples. The management strategy was guided by target PaCO2 and PaO2 values, with guidelines for ventilator adjustment as described in Figure 1. Arterial blood gas tensions were performed 15-20 min after ventilator changes. For statistical analysis, values for blood gas tensions were recorded every hour after the onset of ventilation. No animal received bicarbonate before 4 h of ventilation. Normal saline (10 mL/kg) was infused for a mean AoP < 30 mm Hg. Dopamine was infused at a rate of 10 µg kg-1 min-1 for a mean AoP < 30 mm Hg unresponsive to normal saline infusion.

Drug preparation. BQ 123 (Alexis Biochemicals, San Diego, CA) was dissolved in DMSO (1 mg in 300 µL). This solution was diluted in 8 mL of normal saline and directly infused into the MPA over 8 h at a rate of 1 mL/h. BQ 123 is a selective ETA receptor antagonist without other known effects on the pulmonary vasculature(17). DMSO(1% in normal saline) was infused at the same rate into control animals.

ET-1 plasma assay. To determine the effect of ventilation on circulating ET-1 levels, 2-mL samples were drawn from the aortic catheter at baseline, and at 1, 4, and 8 h after the onset of ventilation. The samples were collected in tubes of EDTA and spun down immediately at 3000 rpm in a refrigerated centrifuge. Plasma was separated and frozen at -70°C. Plasma levels of ET-1 were measured using a commercial enzyme immunoassay kit (R& D Systems, Inc., Minneapolis, MN). The assay was performed according to the instructions provided with the assay.

Data analysis. Data are presented as means ± 1 SEM. Statistical analysis was performed with the Super ANOVA software package(Abacus Concepts, Berkeley, CA). Comparisons were made using univariate repeated measures analysis of variance by linear contrast analysis. p < 0.05 was considered significant.

RESULTS

Hemodynamics and gas exchange. There was no difference between the control group and the group treated with BQ 123 at baseline before the onset of ventilation with regards to any hemodynamic variable. After delivery, MPA pressure fell more rapidly in animals treated with BQ 123, but was otherwise not different from controls (Fig. 2;p < 0.05). In contrast to the progressive fall in LPA blood flow during ventilation in control animals, BQ 123 treatment caused sustained increases in LPA flow from 2 to 8 h after delivery (Fig. 2;p < 0.05). PVR was lower in animals treated with BQ 123(Fig. 2;p < 0.05). AoP was not different between the control and the treatment group. At baseline AoP in the control group was 44 ± 1 mm Hg and was 38 ± 3 after 8 h of ventilation. The AoP in the BQ 123 group was 47 ± 3 mm Hg at baseline and was 35 ± 4 after 8 h of ventilation. The ratio of MPA to Ao pressure was lower in the group treated with BQ 123 at 1h (0.87 ± 0.08 versus 0.68 ± 0.08 mm Hg; control versus BQ 123; p < 0.05) and 5 h (0.98 ± 0.03 versus 0.91 ± 0.04 mm Hg; control versus BQ 123; p < 0.05) after delivery. There was no difference in PaCO2 between the two groups; however, BQ 123 treatment caused a sustained increase in PaO2 (Fig. 3; 53± 14 mm Hg, control versus 174 ± 71 mm Hg, BQ 123, p < 0.05 at 8 h). The arterial pH was higher in the BQ 123 treatment group at 3 and 4 h after delivery (Fig. 3; p < 0.05). There was no difference between the groups with regard to peak inspiratory pressure or mean airway pressure during the 8-h ventilator study(Table 1). Positive end expiratory pressure was higher in the control group at 1 h, whereas the ventilator rate was lower in the control group at 5 h. Oxygenation index was higher in the control group than the BQ 123 group at 3, 6, and 7 h after the onset of ventilation.

Figure 2
figure 2

Hemodynamic effects of prolonged ETA receptor blockade with BQ 123 on mean PAP, LPA blood flow, and left lung PVR after acute delivery and ventilation of premature lambs. BQ 123(1.0 mg/h) lowered mean PAP at 1 h after delivery and caused a sustained increase in LPA flow and fall in PVR after delivery.

Full size image
Figure 3
figure 3

Effects of BQ 123 treatment on arterial blood gas tensions for 8 h after acute delivery and ventilation of premature fetal lambs. There was no difference in PaCO2 between the two groups; however, BQ 123 treatment caused a sustained increase in PaO2 from 6 to 8 h after the onset of ventilation. The arterial pH was higher in the BQ 123 treatment group at 3 and 4 h after delivery.

Full size image
Table 1 Ventilator settings during 8-h delivery study
Full size table

There was no difference in the weight of the animals after the ventilator study (2713 ± 110 g, control versus 2668 ± 143 g, BQ 123). One animal in the control group died 7 h after delivery and one animal in the BQ 123 group died 6 h after delivery. There was no difference in the average amount of sodium bicarbonate given between the two groups (17.6± 7.6 mEq/animal, control versus 7.2 ± 4.5 mEq/animal, BQ 123, p = 0.27). There was no difference in the number of animals receiving dopamine infusion between the two groups (3/5, control versus 2/5, BQ 123). LAP was not different between the two groups at baseline (2.8 ± 0.7 mm Hg, control versus 2.0± 0.8, BQ 123), and during the entire study period.

ET-1 plasma levels. Circulating ET-1 increased in the control group during the 8 h after delivery (Fig. 4;p < 0.05). There was no difference in plasma ET-1 at baseline between the two groups (0.7 ± 0.1 pg/mL, control versus 0.5± 0.1 pg/mL, BQ 123; p < 0.05). There was no increase in plasma ET-1 levels at 1 or 4 h after delivery in the group treated with BQ 123(0.7 ± 0.1 pg/mL, 1 h; 1.2 ± 0.5 pg/mL, 4 h). Due to a technical error, the samples at 8 h in the BQ 123 group were not able to be analyzed.

Figure 4
figure 4

Effects of delivery of premature fetal lambs on plasma ET-1 levels. Circulating ET-1 increased in the control group at 8 h after delivery (n = 4 for each data point). BL, baseline.

Full size image

DISCUSSION

We hypothesized that in premature lambs high PVR and poor gas exchange were due to ET-1 pulmonary vasoconstriction, mediated by increased ETA receptor activity. We found that plasma ET-1 levels increased after delivery and ventilation of premature fetal lambs, and that BQ 123, an ETA receptor antagonist, increased left pulmonary artery blood flow, decreased PVR, and improved oxygenation after premature delivery and prolonged ventilation. These studies suggest that ET-1 contributes to the hemodynamic abnormalities in this model of pulmonary hypertension and severe HMD.

ET-1 is a potent vasoactive peptide with mitogenic effects on vascular smooth muscle, and is produced primarily by the vascular endothelium in the normal lung circulation(14). In the normal fetal lung, ET-1 is present and contributes to high PVR(10,15–17). We have previously shown that chronic intrauterine pulmonary hypertension causes the loss of ETB-mediated vasodilation, progressive ETA-mediated vasoconstriction, and increased lung ET-1 content(7). Furthermore, ET-1 contributes to a model of pulmonary hypertension caused by ligation of the ductus arteriosus in the late gestation ovine fetus. The ETA receptor antagonist, BQ 123, attenuated the in utero increases in PAP, the abnormal transition of the fetus with pulmonary hypertension after delivery, as well as the right ventricular hypertrophy and smooth muscle changes in chronic pulmonary hypertension caused by ligation of the ductus arteriosus in fetal lambs(9).

The physiologic role of ET-1 in the normal ovine fetal lung has been controversial. ET-1 is present in the perinatal lung(15) and is vasoactive in the fetus(8,10,19). Brief infusion of ET-1 causes potent vasodilation acutely(18–21); however, with prolonged infusion, hypertension prevails(18). However, exogenous infusion of ET-1 may not accurately describe the hemodynamic effects of endogenous production of ET-1 in the fetal lung. Evidence suggests that ET-1 acts as a local autocrine and paracrine factor rather than a circulating hormone, because secretion of ET-1 by endothelial cells is polar and directed abluminally toward the interstitial region(22). Some studies of exogenous infusion of ET-1 have emphasized that the major effect of ET-1 in the normal late gestation fetal lung is vasodilation and that the majority of ET-1 receptor activation in the ovine fetal lung are the ETB1 receptors(20), which mediate only vasodilation(10). In contrast, several studies suggest that the ETA receptors play an important role in mediating vasoconstriction in the late gestation ovine fetus(8,10,17). Intrapulmonary infusion of big-ET-1, the precursor to ET-1, causes progressive and sustained pulmonary hypertension without even transient vasodilation(10), suggesting that stimulation of endogenous ET-1 may have very different effects than brief exogenous infusions of ET-1. Recent studies have shown that combined ETA and ETB receptor blockade with Ro 47-0203 does not change the increase in pulmonary blood flow or decrease in pulmonary vascular resistance with in utero oxygen ventilation, suggesting that endogenous ET-1 activity does not play a major role in the increased pulmonary blood flow during the normal transitional circulation at birth(23).

These findings are interesting as little is known of the role of ET-1 and its receptors in pulmonary vasoregulation in the premature fetus and in this model of severe HMD. Past studies of this experimental model of severe HMD suggests that high PVR is partly due to structural changes and an imbalance in production and responsiveness to vasodilator stimuli(4,5,6,24). Pulmonary immaturity leads to surfactant deficiency and respiratory failure, which in part may be improved by exogenous surfactant therapy(1). However, some newborns born prematurely may continue to deteriorate and develop pulmonary hypertension(2,24). Inhaled nitric oxide improves oxygenation, increases pulmonary blood flow without increasing pulmonary edema, lowers PVR, and decreases lung neutrophil accumulation in severe experimental HMD(5,6). The present study suggests that ET-1 also contributes to the development of pulmonary hypertension in experimental HMD. The mechanism by which BQ 123 improved oxygenation after delivery and ventilation of the premature lamb remains uncertain. We speculate that improved pulmonary blood flow and diminished PVR leads to decreased right to left shunting within the lungs or the foramen ovale, thus improving oxygenation. During the ventilator study the pulmonary to aortic pressure ratio was lower in the BQ 123 group at 1 and 5 h, and tended to be lower during the remainder of the study. Likewise, improved right ventricular function by reducing right ventricular afterload may lead to improved cardiac output and arterial and mixed venous oxygen tension.

Although ET-1 was originally thought to contribute only to vasoconstriction and smooth muscle proliferation, accumulating evidence indicates that ET-1 also contributes to inflammation and acute lung injury(25). ET-1 may act as a cytokine(26), promote neutrophil adhesion(27), and increase lung vascular permeability(28). Similarly, the pathophysiology of HMD is also characterized by increased lung vascular permeability and lung neutrophil accumulation(5), as well as altered production of inflammatory cytokines(29). Circulating ET-1 levels are found in humans with acute lung injury and the adult respiratory distress syndrome(30,31). The role of ET-1 in acute lung injury in this model of severe respiratory failure caused by premature delivery and HMD in the ovine fetus remains uncertain.

In summary, chronic blockade of the ETA receptor with BQ 123 lowered PVR, enhanced vasodilation, and improved oxygenation at delivery in the premature lamb. Circulating ET-1 increased after delivery of these animals with severe HMD. These findings support the hypothesis that increased ET and enhanced ETA receptor activity contributes to high PVR in severe experimental HMD. We conclude that ET contributes to pulmonary hypertension in severe HMD.