Pediatric Pulmonology 41:844–854 (2006)

Effects of Antenatal Corticosteroid Treatment onPulmonary Ventilation and Circulation in Neonatal

Lambs With Hypoplastic Lungs

Keiji Suzuki, MD, Stuart B. Hooper, PhD, Megan J. Wallace, PhD,Megan E. Probyn, PhD, and Richard Harding, PhD, DSc*

Summary. Our aim was to determine whether antenatal corticosteroids improve perinatal

adaptation of the pulmonary circulation in lambs with lung hypoplasia (LH). LH was induced in 12

ovine fetuses between 105 and 140 days gestation (term �147 days); in 6 of these the ewe was

given a single dose of betamethasone (11.4 mg im) 24 hr before delivery (LHþB). All lambs,

including a control group (n¼6), were delivered at�140 days and ventilated for 2 hr during which

arterial pressures, pulmonary blood flow (PBF), and ventilating pressure and flow were recorded.

During ventilation, respiratory system compliance was lower in both LHþB and LH groups than in

controls. Pulmonary vascular resistance (PVR) was lower in LHþB lambs than in LH lambs and

similar to controls; PBF was reduced in LH lambs but was restored to control levels by

betamethasone. Themeandensityof small arteries of LHþB lambswas similar to that of LH lambs

(P¼0.06) and lower than in controls; the thickness of the media of small pulmonary arteries from

LHþB lambswas similar to that in LH lambs and thicker than in controls. VEGFmRNA levelswere

not different between groups. PDGFmRNA levels in LHþB lambswere higher than in LH lambs; a

similar trend (P¼0.06) was seen for PECAM-1. SP-C mRNA levels were greater in both LH and

LHþB lambs than in controls. Effects of betamethasone were greater on indices of pulmonary

circulation than ventilation. We conclude that a single dose of maternal betamethasone 24 hr prior

to birth has significant favorable effects on the postnatal adaptation of the pulmonary circulation in

lambs with LH. Pediatr Pulmonol. 2006; 41:844–854. � 2006 Wiley-Liss, Inc.

Key words: lung hypoplasia; fetus; betamethasone; pulmonary blood flow; newborn;

ventilation.

INTRODUCTION

Fetal lung hypoplasia (LH) is present in about 1 in 1,000human live births,1–3 and is a major cause of neonatalmortality and morbidity.4–6 LH is associated with variousfetal conditions such as oligohydramnios (with or withoutruptured membranes), intrathoracic space occupyinglesions, and neuromuscular compromise.7 At present,treatments for LH are limited, although our recent study,showing that the pulmonary circulation is particularlyaffected in experimental LH8 suggests that improving thepulmonary circulation would be necessary to improveoutcomes for infants with LH. Antenatal corticosteroidsare known to enhance the maturation of the fetal alveolarepithelium, surfactant production, and lung structure.9,10

There is growing evidence that they can also affectpulmonary vascular development such as the regulation ofvasoactive mediators,11,12 remodeling of vascular walls,and maturation of capillary loops.13 In addition to thesechanges in the pulmonary vasculature, corticosteroidshave beneficial effects on the pulmonary circulation in theperinatal transition period.14–16

It is possible that corticosteroid treatment couldfacilitate the perinatal adaptation of infants with LH by

improving the pulmonary circulation. A study in fetal ratswith LH due to experimentally induced CDH showed thatdexamethasone reduced thickening of the pulmonaryarteries.17 Other studies have found that maternaldexamethasone treatment restored the CDH-inducedreductions in pulmonary eNOS protein18 and PDGFexpression19 as well as reversing the increase inangiotensin-converting enzyme activity.20 However,administration of dexamethasone to pregnant rats withCDH fetuses did not change pulmonary VEGF expressionrelative to untreated CDH fetuses.21

Department of Physiology, Monash University, Victoria, Australia.

Grant sponsor: National Health and Medical Research Council of Australia.

*Correspondence to: Prof. Richard Harding, Department of Physiology,

Monash University, Victoria 3800, Australia.

E-mail: richard.harding@med.monash.edu.au

Received 29 November 2005; Revised 6 March 2006; Accepted 7 April

2006.

DOI 10.1002/ppul.20453

Published online 18 July 2006 in Wiley InterScience

(www.interscience.wiley.com).

� 2006 Wiley-Liss, Inc.

Although beneficial effects of corticosteroids on thepulmonary circulation have been shown in lungs renderedhypoplastic due to CDH, it is possible that the pulmonarycirculation in LH due to other, more common causes, suchas oligohydramnios resulting from ruptured membranes,may not be similarly affected. There is evidence that theremay be different mechanisms of remodeling of thepulmonary vasculature between LH with and withoutCDH. For example, oligohydramnios-induced LH in ratsis associated with decreased VEGF mRNA levels22

whereas in LH induced by CDH, pulmonary VEGFmRNA, and VEGF protein were increased.21 A recentstudy of human LH showed no change in VEGF mRNAlevels in patients with LH due to CDH and other causes.23

This study, however, showed that other factors mediatingpulmonary vascular growth (von Hippel–Lindau protein,hypoxia-inducible factor-1a) were differently expressedin the lungs of CDH-induced LH and LH due to othercauses.23 As CDH affects the lungs much earlier ingestation than ruptured membranes, it is likely to havedifferent effects on the pulmonary vasculature.

To date there have been no studies examining the effectsof antenatal corticosteroids on the physiological proper-ties of the pulmonary circulation during ventilation, andtheir relation to structural and biochemical changes, innewborns with LH not secondary to CDH. Therefore, ouraims were to determine the effects of antenatal transpla-cental corticosteroid exposure on respiration and thepulmonary circulation in ventilated newborn lambs withLH induced by chronically reduced lung expansion inutero and to relate these effects to morphological andbiochemical alterations in lung structure, growth factors,and indices of lung maturation.

MATERIALS AND METHODS

In a previous study,8 we compared a group of newbornlambs with experimentally induced LH with a group ofcontrol lambs. In the present study, we report data from athird group of animals in which LH was induced as in ourprevious study; however, these animals were exposed to asingle dose of antenatal transplacental corticosteroids(LHþB group, n¼ 6). We have compared this LHþBgroup with the other two groups of animals (untreated LHgroup, n¼ 6; control group, n¼ 6) described previously.8

The protocols for the fetal surgery, delivery and study ofnewborn lambs, and the procedures for physiological,histological, and biochemical analyses in LHþB lambswere similar to those used for control and LH lambs.8

Fetal LH was induced as described previously.8 Briefly,a tracheal tube and two amniotic fluid drainage catheterswere implanted in the fetus at�105 days of gestation. Thefetal tracheal tube acted as a tracheo-amniotic shuntallowing fetal lung fluid to drain into the amniotic cavity,eliminating the physiological resistance of the larynx;

amniotic fluid was allowed to drain via two short cathetersbetween the amniotic sac and the maternal peritonealcavity. Amniotic and fetal lung fluids were drainedpassively from the time of surgery until delivery at�140 days of gestation (term �147 days).

In the LHþB animals, a single dose of betamethasone(Celestone/ChronodoseTM, Schering-Plough Pty Ltd.,Australia, 11.4 mg) was injected intramuscularly to theewe 24 hr before elective delivery. The dose ofbetamethasone and the route of injection were inaccordance with the National Institute of Health con-sensus on antenatal corticosteroid treatment.24 Weadministered a single dose of betamethasone, instead oftwo doses, in order to avoid preterm delivery of lambs.

At 140 days’ gestation, all ewes and fetuses wereanesthetized with halothane (3%) in oxygen. The fetalhead and chest were exteriorized and catheters insertedinto a carotid artery and jugular vein; the existing trachealtube was replaced by an endotracheal tube for ventilationafter birth. The trachea was ligated around the endo-tracheal tube to avoid any gas leakage during ventilation.After creating a left thoracotomy, an ultrasonic flow probe(5–6SBTM, Transonic Systems, Inc., Ithaca, NY) wasplaced around the left pulmonary artery and a non-occlusive catheter was inserted into the main pulmonaryartery. The rib cage and skin were sutured closed beforethe lambs were delivered, weighed, and laid in the rightlateral posture under a radiant heater. Amniotic fluid wascollected before and after delivery of the fetus and thevolume measured. Once delivered, the lambs wereventilated and monitored for 2 hr with a constant flow,time-cycled, pressure controlled ventilator (Babylog 8000PlusTM, Drager Medizintechnik, Lubeck, Germany). Toeliminate spontaneous breathing, lambs were sedated withalphaxalone/alphadolone acetate (SaffanTM, PitmanMoore, NSW, Australia, i.v.); ventilatory and circulatorydata were continuously recorded using a digital dataacquisition system (Powerlab/8spTM, ADInstruments,NSW, Australia). The recorded variables were: fractionof inspired oxygen (FiO2), ventilating (i.e., proximalairway) pressure (Paw), air flow and tidal volume,systemic artery pressure (SAP), pulmonary artery pres-sure (PAP), and left pulmonary artery blood flow (PBF).Ventilatory pressure, flow, and volume signals obtainedfrom the ventilator were recorded digitally. The left atrialpressure was assumed to be 5 mm Hg.25,26

The initial ventilator settings were: flow 6 L/min; FiO2

1.0; ventilation rate 60 breaths/min; inspiration time0.5 sec; peak inspiratory pressure (PIP) 25 cm H2O; end-expiratory pressure (EEP) 5 cm H2O. These settings wereadjusted every 5–15 min according to the arterial bloodgas tensions and tidal volume (Vt), aiming for targetvalues of: PaCO2 35–45 mm Hg; PaO2 60–100 mm Hg;Vt 5–10 ml/kg body weight. The PIP was adjusteddepending on Vt; ventilation rate was then adjusted

Corticosteroid Treatment of Fetal Lung Hypoplasia 845

depending on PaCO2 and FiO2 was adjusted depending onthe PaO2. The EEP was kept at 5 cm H2O at least for thefirst hour; after this, if PaO2 exceeded 100 mm Hg on aFiO2 of 0.4, EEP was decreased in 1 cm H2O steps.Alveolar-arterial difference in PO2 (AaDO2), ventilatoryefficiency index (VEI), and respiratory system compliance(Crs) were calculated as described previously.8 After 2 hrof ventilation, lambs were painlessly killed by a lethaldose of sodium pentobarbitone (1.5 g i.v.).

Lung Tissue Analysis

At necropsy, the lungs and heart were removed andweighed. After ligating the left main bronchus, the leftlung was removed and portions frozen in liquid nitrogenand stored at �708C for biochemical analysis; otherportions were removed for dry weight analysis. The rightlung was fixed via the trachea at 20 cm H2O with 4%paraformaldehyde. After 3–5 days of fixation, the rightlung volume was determined using the Cavalierimethod.27 Morphometric analysis of lung tissue wasperformed as in our previous study.8 Pulmonary DNA andsoluble protein concentrations in lung tissue weredetermined using established fluorometric DNA andcolorimetric protein assays.28

Total mRNA was extracted and Northern blot analysesperformed for surfactant protein (SP)-C, as an index ofalveolar epithelial type-ll cell number and maturity;vascular endothelial growth factor (VEGF), a potentangiogenic factor; platelet derived growth factor (PDGF),a growth factor with angiogenic properties; plateletendothelial cell adhesion molecule-1 (PECAM-1), amember of the immunoglobulin superfamily which hasbeen shown to be an important scaffolding moleculeinvolved in several signaling pathways including angio-genesis and control of apoptotic events. For Northern blotanalysis, total RNA was extracted from the lung tissue29

and the extracted RNA was electrophoresed in a gel,transferred to a nylon membrane and hybridized withradio-labeled ovine cDNA probes. The VEGF probe wasgenerated by us using primers that were designed using thepublished sequence for ovine VEGF30 to amplify a 373 bpproduct specific for all five isoforms of VEGF. Thesequence of the PCR product has 100% homology with thepublished sequence for ovine VEGF.30 Specific mRNAtranscripts were detected using a phosphoimager and thedensity of the radioactive bands was quantified usingImageQuantTM software (University of Virginia, www.itc.virginia.edu/achs/). Blots were stripped and rehybridizedwith a radioactively labeled cDNA probe for ovine 18SrRNA. The density of each specific transcript wascorrected for the density of the 18S rRNA band to correctfor minor loading differences between lanes. The valueswere then expressed as a percentage of the mean value ofthe control group run on the same blot.

Statistical Analysis

Data from the LHþB group were compared withdata from the LH group to determine the effects ofbetamethasone; data from the control group was used todetermine whether betamethasone had restored pulmonaryparameters to control levels. When comparing the in uterodata, a one way ANOVA was used to compare treatmentgroups. One way ANOVAs were also used to comparegestational ages at surgery and delivery, body weight andamniotic fluid volume at delivery, wet and dry lung weights,pulmonary DNA and protein contents and concentrationsand SP-C, VEGF, PECAM-1, and PDGF mRNA levels aswell as morphometric parameters. Two way repeatedmeasures ANOVAs were used to compare blood gas andventilatory parameters from 5 to 120 min after delivery aswell as blood flow parameters and pulmonary vascularresistance (PVR) between 15 and 120 min. Wheresignificant treatment effects were found (P< 0.05), a leastsignificant difference post hoc test was used to detectdifferences between individual groups and time points.

RESULTS

Gestational Ages and Birth Weights

The gestational ages of the LHþB fetuses at the time ofsurgery and treatment onset, their gestational age atdelivery and their birth weights were not different fromthose of the control or LH fetuses (Table 1). Amniotic fluidvolume at delivery was decreased by 66% in LHþBanimals compared to controls (P< 0.01) but was notdifferent from that in LH lambs (Table 1). No overt fetalanomalies were observed.

Lung Growth

Table 2 shows data relating to lung growth in the threegroups. The wet lung/body weight ratio was reduced by

TABLE 1— Gestational age (GA) at Treatment Onset andDelivery, and Body Weights and Amniotic Fluid Volume atDelivery in the Three Treatment Groups: Controls, LungHypoplasia (LH), and Lung Hypoplasia Treated withBetamethasone (LHþB)

Variable Control LH LHþB

GA at treatment

onset (days)

105.8� 1.4 105.5� 2.5 104.3� 0.4

GA at delivery

(days)

139.8� 0.2 140.0� 0.5 140.3� 0.2

Birth weight (kg) 3.66� 0.29 3.33� 0.18 3.76� 0.23

Amniotic fluid

volume (ml)

467� 47 183� 69* 158� 77*

Values are mean� SEM.

*Indicates significant difference (P< 0.05) from the control group.

846 Suzuki et al.

33% in LHþB lambs compared with controls (P< 0.01)but was not different from that in LH lambs. The dry lung/body weight ratio in LHþB lambs was also lower thanthat of controls (P¼ 0.02) and was not different fromvalues in LH lambs. Similarly, total lung DNA content(mg/kg body weight) in LHþB lambs was 29% lowerthan in controls (P< 0.01) but was not different from thatin LH lambs. Total lung protein content (g/kg bodyweight) in LHþB lambs was 52% lower than in controls(P< 0.01) but was not different to LH values.

Lung volume at 20 cm H2O (ml/kg) in LHþB lambswas lower than in controls (P< 0.05) but was higher thanin LH lambs (P< 0.001). Crs adjusted for body weight(Crs/BW) increased significantly with time in all groups.At 2 hr after delivery, Crs/BW in LHþB lambs wassimilar to values in LH lambs but lower (28%) than incontrols over the study period (P< 0.01; Table 2). Crsadjusted for wet or dry lung weights was similar in thethree groups (data not shown).

Arterial Blood Gases and Ventilator Settings

Arterial pH (7.44� 0.01 at 2 hr), PaCO2

(37.6� 1.1 mm Hg at 2 hr), and PaO2 (69.2� 5.8 mm Hgat 2 hr) in LHþB lambs were not different from values ineither the control or LH lambs. The hematocrit at 2 hr afterbirth was not different between the groups. The FiO2

required to maintain the target PaO2 was reduced morerapidly in LHþB lambs than in LH lambs (P< 0.05).Over the entire 2 hr, FiO2 in LHþB lambs was notdifferent to that in controls; however, values in bothLHþB and control groups were lower (P< 0.02 for both)than in LH lambs (Fig. 1A).

During the first hour of ventilation, the EEP was5 cm H2O in all groups. However, at 2 hr, the EEP wassignificantly lower in LHþB lambs (EEP 3 cm H2O)and controls (EEP 3 cm H2O) than in LH lambs (EEP

5 cm H2O) (P< 0.01; data not shown). In contrast, therewas no difference in PIP between groups during the entire2 hr (data not shown). During the 2 hr ventilation period,mean Paw progressively decreased in all groups. Duringthe last 60 min, values in LHþB lambs were not differentto values in control or LH lambs; however, mean Paw washigher in LH lambs than in controls (P< 0.02; Fig. 1B).

Oxygenation and Ventilation

AaDO2 decreased with time in all groups. Over the firsthour, AaDO2 values in the LHþB lambs were notdifferent from controls whereas values in the LH groupwere significantly greater than in control and LHþBgroups (P< 0.05), indicating less efficient oxygenation inLH lambs (Fig. 1C). Ventilator efficiency index (VEI)significantly increased with time in all groups over the 2 hrstudy period; over the last 30 min, values in LHþB andLH groups were significantly lower (P< 0.04 for both)than in controls (Fig. 1D).

Pulmonary Blood Flow

Prior to delivery, mean PBF adjusted for lung weight(meanPBF/LW) tended to be different between groups(P¼ 0.08) with values in LHþB lambs being higher thanin LH lambs but not different to values in controls(Fig. 2A). During the 2 hr ventilation period, values ofmeanPBF/LW in LHþB lambs were significantly greater(P< 0.005) than in LH lambs and not different to values incontrols (Fig. 2A).

Prior to delivery, mean PBF in LHþB lambs, both raw(meanPBF) and adjusted for body weight (meanPBF/BW), was not different to that in control or LH lambsalthough values in LH lambs were lower than in controls(P< 0.02). During the 2 hr ventilation period, meanPBFand meanPBF/BW tended to increase with time, withvalues in LHþB lambs being similar to those of controls

TABLE 2— Lung Growth and Lung Mechanics Parameters in the Control, Lung Hypoplasia(LH), and Lung Hypoplasia Treated with Betamethasone (LHþB) Groups

Variable Control LH LHþB

Wet lung weight (g/kgBW) 29.8� 2.7 21.7� 1.3* 19.6� 2.1*

Dry lung weight (g/kgBW) 3.72� 0.19 2.67� 0.10* 2.84� 0.35*

Lung DNA content (mg/kgBW) 185.5� 12.5 145.0� 6.8* 131.5� 11.5*

Lung protein content (mg/kgBW) 1383� 249 918� 46 682� 57*

DNA concentration (mg/g dryLW) 50.4� 4.1 54.8� 2.8 47.9� 4.8

Protein concentration (mg/g dryLW) 361� 46 347� 23 250� 27

Lung volume (ml/kgBW) 38.0� 2.4 23.2� 1.2* 31.4� 2.2*,**

Respiratory system compliance at 2 hr

(ml/cm H2O/kg BW)

0.57� 0.04 0.40� 0.03* 0.41� 0.03*

Wet lung weight, dry lung weight, lung DNA content, lung protein content, lung volume (at 20 cm H2O

fixation pressure) and lung compliance at 2 hr are adjusted for body weight (BW). DNA and protein

concentrations are expressed in relation to dry lung weight (LW). Values are mean� SEM.

*Indicates significant difference (P< 0.05) from the control group.

**Indicates significant difference (P< 0.05) between LH and LHþB groups.

Corticosteroid Treatment of Fetal Lung Hypoplasia 847

but higher than in LH lambs (P< 0.05). At 2 hr, meanPBFin LHþB lambs (271� 32 ml/min) was no different tovalues in control lambs (306� 26 ml/min) and signifi-cantly greater than values in LH lambs (171� 31 ml/min;P< 0.04). When adjusted for body weight, there was atendency for meanPBF/BW in LHþB lambs (73� 10 ml/min/kg) to be higher than in LH lambs (49� 7 ml/min/kg,P¼ 0.08) but values were not different to that in controls(85� 9 ml/min/kg).

Arterial Pressures

Prior to delivery, there were no differences in meanpulmonary arterial pressure (PAP) between groups(Fig. 3A). Over the 2 hr ventilation period, mean PAP inLHþB lambs was not different to values in controls or LHlambs, although values were significantly greater(P¼ 0.01) in LH lambs than in controls (Fig. 3A). Priorto delivery, mean systemic arterial pressure (SAP) in

LHþB lambs was higher (P¼ 0.03) than in LH lambs butnot different to controls; values in controls were notdifferent to those in LH lambs (Fig. 3B). Over theventilation period, there was a strong tendency (P¼ 0.07)for these relationships to be maintained (Fig. 3B). Beforebirth, the ratio of mean PAP to mean SAP (PAP/SAP)was not different between groups; however after birth,this ratio was significantly lower (P¼ 0.04) in LHþBlambs than in LH lambs and not different to that of controls(Fig. 3C).

Pulmonary Vascular Resistance

Before birth, PVR adjusted for lung weight(PVR�LW) was not different between groups; however,there was a tendency (P¼ 0.10) for values in LHþB andcontrol groups to be lower than values in the LH group(Fig. 2B). Over the 2-hr study period, PVR�LW inLHþB lambs was significantly lower than in LH lambs

Fig. 1. Changes over the 2 hr postnatal ventilation period in (A) fraction of inspired oxygen (FiO2),

(B) mean airway pressure (meanPaw), (C) alveolar-arterial difference of oxygen tension (AaDO2),

and (D) ventilatory efficiency index (VEI) in the control (circles), lung hypoplasia (LH; triangles),

and lung hypoplasia with betamethasone (LHþB; squares) groups. Data are mean�SEM.

848 Suzuki et al.

(P< 0.01) and similar to values in controls (Fig. 2B). At2 hr, the mean value for PVR�LW in LHþB lambs was56% lower than in LH lambs and not different from that incontrols.

Before birth, PVR in LHþB lambs (0.34�0.11 mm Hg �ml�1 �min) was not different to PVR incontrol (0.22� 0.06 mm Hg �ml�1 �min) or LH lambs(0.53� 0.07 mm Hg �ml�1 �min); however, values incontrol lambs were significantly lower than in LH(P< 0.02). PVR adjusted for body weight (PVR�BW)was not different between groups. After birth, PVR andPVR�BW decreased with time in all groups. Over the2 hr ventilation period, PVR and PVR�BW in LHþBlambs were not different to values in controls and weresignificantly lower than in LH lambs (P< 0.01 for both

Fig. 2. Values in the immediate predelivery period and over the

2 hr ventilation period in (A) mean left pulmonary artery blood

flow adjusted for lung weight (mean PBF/LW), and (B) left

pulmonary vascular resistance adjusted for lung weight

(PVR�LW) in the control (circles), lung hypoplasia (LH;

triangles), and lung hypoplasia with betamethasone (LHþB;

squares) groups. Data are mean�SEM.

Fig. 3. Values in the immediate predelivery period and

over the 2 hr ventilation period in (A) mean pulmonary

arterial pressure (PAP), (B) mean systemic arterial pressure

(SAP), and (C) the ratio of mean PAP to mean SAP in the

control (circles), lung hypoplasia (LH; triangle), and lung

hypoplasia with betamethasone (LHþB; square) groups. Data

are mean�SEM.

Corticosteroid Treatment of Fetal Lung Hypoplasia 849

PVR and PVR�BW). At 2 hr, the mean valuefor PVR�BW in LHþB lambs (0.44�0.04 mm Hg �ml�1 �min � kg) was 42% lower than in LHlambs (0.76� 0.06 mm Hg �ml�1 �min � kg; P< 0.001)and was not different from that in controls (0.32�0.04 mm Hg �ml�1 �min � kg); PVR showed the samerelationship, with values in LHþB lambs (0.118�0.007 mm Hg �ml�1 �min) being lower than those in LHlambs (0.234� 0.032 mm Hg �ml�1 �min; P¼ 0.01)but no different to that in controls (0.089�0.012 mm Hg �ml�1 �min).

Lung Morphometry

Table 3 shows pulmonary morphometric data.

Air Space and Terminal Bronchioles

Air space fraction in the lungs of LHþB lambs was notdifferent from values in LH or control lambs. The meanterminal bronchiole density in LHþB lambs (number perfield) was not different from that in LH lambs but was 48%higher than in controls. The volume density of terminalbronchioles in LHþB lambs was not different from that inLH lambs but was 82% higher than in controls.

Volume Densities of Pulmonary Arteries

The volume density of arteries in the lungs of LHþBlambs was not different to that in LH or control lambs. Thevolume density of arteries in lung tissue tended to bedifferent between groups (P¼ 0.06) with the mean valuein LHþB lambs being higher than in LH lambs but lowerthan in controls.

Small arteries with a diameter of 25–75 mm aresometimes referred to as pulmonary ‘resistance arteries.’

The mean volume density of these arteries (both in lungand lung tissue) in LHþB lambs was not different fromthat in LH or control lambs. In these small arteries, theratio of media thickness to the diameter in LHþB lambswas similar to the value in LH lambs and 13% greater thanin control lambs (P< 0.01). The volume density of thelumen of the small pulmonary arteries (25–75 mm) in thelungs of LHþB lambs was similar to that in LH lambs and37% lower (P¼ 0.01) than in controls; similarly, thevolume density of the lumen of these small arteries in lungtissue in LHþB lambs did not differ from that in LHlambs and was 39% lower than in controls (P< 0.02).

Pulmonary Veins

The volume density of pulmonary veins in the lungs andin lung tissue was not different between groups.

mRNA Expression

Total mRNA was successfully extracted and Northernblot analysis performed for four LH, four control, and fiveLHþB lambs. SP-C mRNA levels in LHþB lambs(151� 17% of control) were similar to those in LH lambsand values in these groups were higher (P< 0.03 for both)than in controls (Fig. 4A). The 3.7 kB VEGF transcriptwas clearly recognizable in the Northern blots; there wasno difference in mRNA levels of this VEGF transcriptbetween LHþB, LH, or control lambs (Fig. 4B). PDGFmRNA levels in LHþB lambs (110� 5% of control)were significantly higher (P< 0.01) than in LHlambs and were similar to values in controls (Fig. 4C).PECAM-1 mRNA levels tended to be different betweengroups (P¼ 0.06), with values in LHþB lambs beinghigher than in LH lambs but similar to those in controls(Fig. 4D).

TABLE 3— Morphometric Data on Pulmonary Air Space Fraction, Arteries and Veins in the Control, Lung Hypoplasia (LH),and Lung Hypoplasia Treated with Betamethasone (LHþB) Groups

Variable Control LH LHþB

Air space fraction in the lung 0.75� 0.02 0.75� 0.02 0.75� 0.01

Mean terminal bronchiole density (number/6.245 mm2 field) 1.60� 0.18 2.12� 0.14* 2.37� 0.18*

Volume density of terminal bronchioles in the lung (�10�3) 13.8� 2.1 30.4� 2.3* 25.0� 2.8*

Volume density of arteries in the lung (�10�3) 3.82� 0.37 2.70� 0.35 3.47� 0.42

Volume density of arteries in lung tissue (�10�3) 14.8� 1.1 10.5� 0.9 14.0� 1.6

Volume density of arteries (25–75 mm diameter) in the lung (�10�3) 1.47� 0.18 1.05� 0.08 1.13� 0.13

Volume density of arteries (25–75 mm diameter) in lung tissue (�10�3) 6.03� 0.93 4.17� 0.35 4.49� 0.47

Ratio of medial thickness to diameter of arteries of 25–75 mm diameter 0.20� 0.00 0.23� 0.01* 0.23� 0.01*

Volume density of the lumen of arteries (25–75 mm diameter) in the lung (�10�3) 0.54� 0.07 0.31� 0.03* 0.34� 0.04*

Volume density of the lumen of arteries (25–75 mm diameter) in lung tissue (�10�3) 2.20� 0.33 1.23� 0.12* 1.34� 0.12*

Volume density of veins in the lung (�10�3) 6.65� 1.26 3.76� 0.90 6.11� 0.86

Volume density of veins in lung tissue (�10�3) 26.3� 5.1 14.0� 2.8 24.5� 3.8

Air space fraction in the lung and volume density of terminal bronchioles, pulmonary arteries, and veins in the lung are expressed as values relative to

the lung (tissue and airspace). Volume densities of pulmonary arteries and veins in the lung tissue are expressed as values relative to the lung tissue

(tissue alone). Mean terminal bronchiole density is expressed as the number of bronchioles in a 100� microscopic field of 6.245 mm2. Volume

density has no units (m3/m3). Values are mean� SEM.

*Indicates significant differences (P< 0.05) between the control and LH or LHþB groups.

850 Suzuki et al.

DISCUSSION

The major finding of this study was that a singleantenatal betamethasone injection, given only 24 hr beforedelivery, had dramatic effects on the pulmonary circula-tion before and after birth. As a result, the impairedperinatal adaptation of the pulmonary circulation in LHlambs was restored to a state that was similar to that ofcontrols. These functional changes were associated withincreased mRNA levels of PDGF. Our findings on theeffects of antenatal corticosteroids on the pulmonarycirculation in hypoplastic lungs are consistent with thoseof a recent study in normal ovine fetuses.31

The recommended regimen of antenatal betamethasoneis two injections 24 hr apart.24 As we found that thisregimen caused a high rate of premature delivery in ourlate gestation sheep we gave a single injection 24 hr beforedelivery. It is possible that the use of two injections 24 hr

apart would have had a greater effect than the singletreatment.

Lung Growth

Betamethasone treatment in the presence of LH did notinfluence fetal pulmonary wet weight, dry weight, or DNAcontent compared to lambs with LH alone. It is notsurprising that these measures of lung size were unaffectedin view of the short interval between treatment anddelivery. Lung volume (at 20 cm H2O) after fixation wasincreased by betamethasone treatment compared tountreated LH lambs but still remained lower than incontrols; however there was no difference in Crs betweenLH and LHþB lambs. Hence, assuming that relative lungcompliance, as determined by fixative, is similar to that inthe gas-filled lung, the greater lung volume at 20 cm H2Oin LHþB compared to LH lambs may indicate an

Fig. 4. mRNA levels of (A) surfactant protein-C (SP-C), (B) vascular endothelial growth factor

(VEGF; 3.7 kB transcript), (C) platelet-derived growth factor (PDGF), and (D) platelet endothelial

cell adhesion molecule-1 (PECAM-1) in control (black bars), lung hypoplasia (LH; open bars), and

LH with betamethasone (LHþB; hatched bars) groups. Values are expressed as a proportion of

the 18S mRNA levels and relative to the mean values of the control group. Data are mean�SEM.

* indicates significant differences (P<0.05) from the control group. # indicates significant

differences (P<0.05) between the LH and LHþB groups.

Corticosteroid Treatment of Fetal Lung Hypoplasia 851

increase in functional residual capacity (FRC). It is wellknown that FRC is mainly dependent on the stabilizationof alveoli by pulmonary surfactant activity, if the totalvolume and structure of the alveoli are the same. We foundthat betamethasone treatment did not further increase SP-C mRNA expression over that in LH animals; however, itis possible that corticosteroids affected other stages of thesurfactant system such as phospholipid synthesis9,32 oractivation.33 In support of the present findings, antenatalcorticosteroid treatment is associated with increasesin lung volume and compliance in preterm newbornlambs.10,34

Oxygenation and Ventilation

There have been few studies examining the effects ofcorticosteroids on respiratory parameters in animals withLH. One study reported that antenatal betamethasoneimproved PaO2, PaCO2, and lung compliance in lambswith CDH.35 Another showed that antenatal cortisoltreatment improved PaO2, lung compliance and lungglycogen, and protein levels in lambs with CDH.36 Ouruntreated LH lambs showed a delayed reduction in AaDO2

during the first hour after birth, compared to controls;however, with betamethasone treatment the time course ofchange in AaDO2 was identical to that in controls. Thiseffect of betamethasone could be related to the decrease inPVR which resulted in increased PBF and a possibledecrease in right to left shunting through the ductusarteriosus in LHþB lambs. It may also be related to anincrease in lung volume which could have stabilizedalveolar recruitment and decreased intra-pulmonaryshunts. It is also possible that a decrease in interstitialfluid (and protein) caused a more efficient diffusion ofoxygen through the gas exchange surface.

In contrast to AaDO2, VEI and Crs were not altered bybetamethasone treatment. This suggests that pulmonaryventilation may be entirely dependent on lung size and thatthe betamethasone treatment had no effect on themechanical properties of the lung. Efficiency of CO2

elimination is determined by alveolar ventilation. BecauseVEI was not different between LH and LHþB lambs,we speculate that the physiological dead space was alsothe same in these groups, assuming that they had the samerate of CO2 production. The reduction in Crs seen at 2 hrwas proportional to the reduction in lung weights and lungDNA content (�30% decrease). These results areconsistent with the findings of a study of infants withLH following preterm rupture of membranes, whichdemonstrated a correlation between VEI and the severityof LH.37

Pulmonary Circulation

We found that betamethasone reduced the elevatedPVR seen in LH lambs towards the level seen in controls.

Indeed the time trends of PVR adjusted for lung weight(PVR�LW) after birth in LHþB lambs were similar tothose in controls. These findings were associated with anincrease in PBF and a reduction in PAP/SAP in the LHþBlambs. No previous study has examined the effects ofcorticosteroid treatment on the pulmonary circulation inthe presence of LH induced by oligohydramnios. How-ever, a study of fetal lambs with normal lung developmentreported that corticosteroid treatment enhanced norepi-nephrine but not acetylcholine or hyperoxia-mediatedpulmonary vasodilation.14 An in vitro study showed that18 hr after betamethasone treatment, pulmonary arteriesfrom preterm lambs showed increased PGE2 and iso-prenalinemediated relaxation.16 A study in preterm fetallambs found that corticosteroid treatment led to anaugmented nitric oxide-mediated relaxation that wasgreater in pulmonary veins than in arteries.15 Each ofthese effects would be expected to improve pulmonaryblood flow (PBF) after birth in the presence of acompromised pulmonary circulation.

It would have been beneficial if we could havemeasured left atrial pressure for the estimation of PVR;however, we were to do this for technical reasons. As thereare currently no data available on left atrial pressure in LHanimals, we assumed the left atrial pressure to be 5 mm Hg,which is similar to values seen in human adults25 andnormal fetal or newborn lambs prior to ventilation.26

While it is possible that decreased PBF in LH may causelower left atrial pressure than normal, it is also possi-ble that compromised left ventricular function in LHlambs may cause higher left atrial pressure. We mea-sured central venous pressure as an index of right atrialpressure and found no difference between the three groups(�4 mm Hg), suggesting that right atrial pressure is notaltered by LH. As there still exists some shunt flowbetween the right and left atrium through the foramenovale in the perinatal period, we did not expect to have alarge difference in left atrial pressure between the threegroups.

PVR is determined by the physical properties of bloodvessels, blood, and blood–vessel wall interactions.38 Inthe isolated perfused lungs of newborn lambs, the totalPVR is approximately evenly distributed between arteries,capillaries, and veins.39,40 We found that in LH lambs,betamethasone led to a trend for an increased volumedensity of large arteries (diameter >75 mm), but not smallarteries (25–75 mm), toward the levels observed in controllambs. Analysis of capillary volume would have beeninformative, but could not be performed for technicalreasons. The morphometric changes observed in largepulmonary arteries (�30% increase) in LHþB lungswere unlikely to have been great enough to explain thechanges in PVR observed in the LHþB group (�55%decrease at 2 hr). Although we did not find a difference inmedial wall thickness or volume density of the lumen of

852 Suzuki et al.

the small pulmonary arteries, it is possible that the smoothmuscle was more relaxed in vivo, with a lower ‘functional’resistance. Indeed, it has been shown that corticosteroidsfacilitate relaxation of smooth muscles of pulmonaryarteries and veins.15,16

Lung Morphometry

The lack of effect of corticosteroid treatment ondensities of air-space elements of the hypoplastic lung isconsistent with our present physiological data, whichshowed no differences in lung mechanics. Similarly, wefound no effect of betamethasone on the volume density ofsmall pulmonary arteries, or in their medial wall thickness,in LH lambs. However, betamethasone tended to increasethe volume density of large arteries (diameter >75 mm) inlung tissue towards the levels in control lambs. Themorphometric changes observed in pulmonary arteries donot appear great enough to explain the changes induced inPVR by steroid treatment. It is possible that in vivo, evenwith a thicker medial smooth muscle layer, the pulmonaryarteries in LHþB lambs had decreased smooth muscletone and were more dilated than vessels in untreated LHlungs.

mRNA Expression

We found that SP-C mRNA in LH lungs was not alteredby corticosteroid treatment. SP-C levels in both LH andLHþB lambs were greater than in controls, which is to beexpected as LH increased the relative proportion ofalveolar epithelial type ll cells in the fetal lung.41 Manystudies have demonstrated increased pulmonary surfac-tant production following antenatal corticosteroids, espe-cially in preterm animals. One study found an increase inSP-C mRNA expression in an ovine model of LH and afurther increase in cortisol-treated fetuses with LH.41

Differences between this study and the present findingscould be related to different kinds and doses ofcorticosteroids, different treatment protocols, or differentgestational ages of the lambs.

We studied the expression of PDGF, PECAM-1, andVEGF as all are thought to play a role in the formation andintegrity of blood vessels. While betamethasone increasedPDGF mRNA levels in LH lambs, and tended to increasePECAM-1 levels, it did not alter VEGF mRNA levels.There have been no comparable studies which haveexamined the effects of corticosteroids on VEGF or PDGFexpressions in animals with LH. Only one study in a ratmodel of CDH reported that VEGF expression wasincreased in CDH-induced LH and this increase persistedafter corticosteroid treatment.21 Another study using a ratmodel of CDH showed that the decrease in pulmonaryPDGF expression observed in the presence of CDH-induced LH was restored by corticosteroids;19 this findingis consistent with the results of the present study. The

increased PDGF expression in LHþB lambs, relative tountreated LH lambs, may be related to the trend for anincrease in the density of arteries. There have been noprevious studies of the effects of corticosteroids onPECAM-1 expression in LH lungs, with or without CDH.

CONCLUSIONS

In our ovine model of moderate LH, the reduction inlung size (�30%) and ventilatory efficiency (�30%) werenot affected by a single antenatal betamethasone injection.In contrast, the greatly increased PVR in LH lambs(�140%), which was more than proportional to thereduction in lung size, was significantly reduced bybetamethasone. Impaired pulmonary circulation is acrucial factor in the postnatal maladaptation of neonateswith LH. Our data show that even a single dose of maternalbetamethasone 24 hr prior to birth, with no significanteffects on pulmonary ventilation, has significant favorableeffects on the adaptation of the pulmonary circulation inneonates with LH.

ACKNOWLEDGMENTS

The authors are grateful for the expert assistance of Dr.Megan Cock, Alex Satragno, and Sally Gregg.

REFERENCES

1. Moessinger AC, Santiago A, Paneth NS, Rey HR, Blanc WA,

Driscoll JM. Time-trends in necropsy prevalence and birth

prevalence of lung hypoplasia. Pediatr Perinat Epidemiol 1989;

3:421–431.

2. Knox WF, Barson AJ. Pulmonary hypoplasia in a regional

perinatal unit. Early Hum Dev 1986;14:33–42.

3. Thurlbeck WM. Prematurity and the developing lung. Clin

Perinatol 1992;19:497–519.

4. Kilbride HW, Thibeault DW. Neonatal complications of preterm

premature rupture of membranes. Pathophysiology and manage-

ment. Clin Perinatol 2001;28:761–785.

5. Kilbride HW, Yeast J, Thibeault DW. Defining limits of survival:

lethal pulmonary hypoplasia after midtrimester premature rupture

of membranes. Am J Obstet Gynecol 1996;175:675–681.

6. Smith NP, Jesudason EC, Losty PD. Congenital diaphragmatic

hernia. Paediatr Respir Rev 2002;3:339–348.

7. Harding R, Albuquerque C. Pulmonary hypoplasia: role of

mechanical factors in prenatal lung growth. In: Gaultier C,

Bourbon JR, Post M, editors. Lung Development. Oxford, UK:

Oxford University Press; 1999. p 364–394.

8. Suzuki K, Hooper SB, Cock ML, Harding R. Effect of lung

hypoplasia on birth-related changes in the pulmonary circulation

in sheep. Pediatr Res 2005;57:530–536.

9. Ballard PL. Hormones and lung maturation. Monogr Endocrinol

1986;28:1–354.

10. Ikegami M, Polk D, Jobe A. Minimum interval from fetal

betamethasone treatment to postnatal lung responses in preterm

lambs. Am J Obstet Gynecol 1996;174:1408–1413.

11. Grover TR, Ackerman KG, Le Cras TD, Jobe AH, Abman SH.

Repetitive prenatal glucocorticoids increase lung endothelial

nitric oxide synthase expression in ovine fetuses delivered at

term. Pediatr Res 2000;48:75–83.

Corticosteroid Treatment of Fetal Lung Hypoplasia 853

12. Asoh K, Kumai T, Murano K, Kobayashi S, Koitabashi Y. Effect

of antenatal dexamethasone treatment on Ca2þ-dependent

nitric oxide synthase activity in rat lung. Pediatr Res 2000;48:

91–95.

13. Vyas J, Kotecha S. Effects of antenatal and postnatal corticoster-

oids on the preterm lung. Arch Dis Child 1997;77:F147–

F150.

14. Deruelle P, Houfflin-Debarge V, Magnenant E, Jaillard S, Riou Y,

Puech F, Storme L. Effects of antenatal glucocorticoids on

pulmonary vascular reactivity in the ovine fetus. Am J Obstet

Gynecol 2003;189:208–215.

15. Zhou HY, Gao YS, Raj JU. Antenatal betamethasone therapy

augments nitric oxide-mediated relaxation of preterm ovine

pulmonary veins. J Appl Physiol 1996;80:390–396.

16. Gao YS, Tolsa JF, Shen H, Raj JU. A single dose of antenatal

betamethasone enhances isoprenaline and prostaglandin E-2-

induced relaxation of preterm ovine pulmonary arteries. Biol

Neonate 1998;73:182–189.

17. Taira Y, Miyazaki E, Ohshiro K, Yamataka T, Puri P. Adminis-

tration of antenatal glucocorticoids prevents pulmonary artery

structural changes in nitrofeninduced congenital diaphragmatic

hernia in rats. J Pediatr Surg 1998;33:1052–1056.

18. Okoye BO, Losty PD, Fisher MJ, Wilmott I, Lloyd DA. Effect of

dexamethasone on endothelial nitric oxide synthase in experi-

mental congenital diaphragmatic hernia. Arch Dis Child 1998;78:

F204–F208.

19. Oue T, Shima H, Taira Y, Puri P. Administration of antenatal

glucocorticoids upregulates peptide growth factor gene expres-

sion in nitrofen-induced congenital diaphragmatic hernia in rats. J

Pediatr Surg 2000;35:109–112.

20. Okoye BO, Losty PD, Fisher MJ, Hughes AT, Lloyd DA.

Antenatal glucocorticoid therapy suppresses angiotensin-convert-

ing enzyme activity in rats with nitrofen-induced congenital

diaphragmatic hernia. J Pediatr Surg 1998;33:286–290.

21. Oue T, Yoneda A, Shima H, Taira Y, Puri P. Increased vascular

endothelial growth factor peptide and gene expression in

hypoplastic lung in nitrofen induced congenital diaphragmatic

hernia in rats. Pediatr Surg Int 2002;18:221–226.

22. Savich RD, Mandell E. VEGF mRNA expression is decreased in

the hypoplastic fetal rat lung due to oligohydramnios. Pediatr Res

2004;55:503A–503A.

23. de Rooij JD, Hosgor M, IJzendoorn Y, Rottier R, Groenman FA,

Tibboel D, de Krijger RR. Expression of angiogenesis-related

factors in lungs of patients with congenital diaphragmatic hernia

and pulmonary hypoplasia of other causes. Pediatr Dev Pathol

2004;7:468–477.

24. National Institute of Health Consensus Development Panel.

Antenatal corticosteroids revisited: repeat courses—National

Institutes of Health Consensus Development Conference State-

ment, August 17–18, 2000. Obstet Gynecol 2001;98:144–

150.

25. West JB. Pulmonary blood flow and metabolism. In: West JB,

editor. Physiological basis of medical practice, 12th edition. New

York: Williams & Wilkins; 1990. p 529–537.

26. Teitel DF, Iwamoto HS, Rudolph AM. Changes in the pulmonary

circulation during birth-related events. Pediatr Res 1990;27:372–

378.

27. Michel RP, Cruz-Orive LM. Application of the Cavalieri principle

and vertical sections method to lung: estimation of volume and

pleural surface area. J Microsc 1988;150:117–136.

28. Joyce BJ, Louey S, Davey MG, Cock ML, Hooper SB, Harding R.

Compromised respiratory function in postnatal lambs after

placental insufficiency and intrauterine growth restriction. Pediatr

Res 2001;50:641–649.

29. Chomczynski P, Sacchi N. Single-step method of RNA isolation

by acid guanidinium thiocyanate phenol chloroform extraction.

Anal Biochem 1987;162:156–159.

30. Cheung CY, Brace RA. Ovine vascular endothelial growth factor:

nucleotide sequence and expression in fetal tissues. Growth

Factors 1998;16:11–22.

31. Houfflin-Debarge V, Deruelle P, Jaillard S, Magnenant E, Riou Y,

Devisme L, Puech F, Storme L. Effects of antenatal glucocorti-

coids on circulatory adaptation at birth in the ovine fetus. Biol

Neonate 2005;88:73–78.

32. Frank L, Lewis PL, Sosenko IRS. Dexamethasone stimulation of

fetal-rat lung antioxidant enzyme-activity in parallel with

surfactant stimulation. Pediatrics 1985;75:569–574.

33. Rebello CM, Ikegami M, Polk DH, Jobe AH. Postnatal lung

responses and surfactant function after fetal or maternal

corticosteroid treatment. J Appl Physiol 1996;80:1674–1680.

34. Rebello CM, Ikegami M, Ervin MG, Polk DH, Jobe AH. Postnatal

lung function and protein permeability after fetal or maternal

corticosteroids in preterm lambs. J Appl Physiol 1997;83:213–218.

35. Kapur P, Holm BA, Irish MS, Sokolowski J, Patel A, Glick PL.

Lung physiological and metabolic changes in lambs with

congenital diaphragmatic hernia after administration of prenatal

maternal corticosteroids. J Pediatr Surg 1999;34:354–356.

36. Schnitzer JJ, Hedrick HL, Pacheco BA, Losty PD, Ryan DP,

Doody DP, Donahoe PK. Prenatal glucocorticoid therapy reverses

pulmonary immaturity in congenital diaphragmatic hernia in fetal

sheep. Ann Surg 1996;224:430–437.

37. Suzuki K. Respiratory characteristics of infants with pulmonary

hypoplasia syndrome following preterm rupture of membranes: a

preliminary study for establishing clinical diagnostic criteria.

Early Hum Dev 2004;79:31–40.

38. Fineman JR, Soifer SJ, Heymann MA. Regulation of pulmonary

vascular tone in the perinatal period. Annu Rev Physiol 1995;57:

115–134.

39. Tod ML, Sylvester JT. Distribution of pulmonary vascular

pressure as a function of perinatal age in lambs. J Appl Physiol

1989;66:79–87.

40. Raj JU, Chen P. Microvascular pressures measured by micro-

puncture in isolated perfused lamb lungs. J Appl Physiol 1986;61:

2194–2201.

41. Flecknoe SJ, Boland RE, Wallace MJ, Harding R, Hooper SB.

Regulation of alveolar epithelial cell phenotypes in fetal sheep:

role of cortisol and lung expansion. Am J Physiol 2004;287:

L1207–L1214.

854 Suzuki et al.