Preterm growth restriction and bronchopulmonary dysplasia: the vascular hypothesis and related physiology

Abstract

Key points

• Approximately 5–10% pregnancies are affected by fetal growth restriction.

• Preterm infants affected by fetal growth restriction have a higher incidence of bronchopulmonary dysplasia.

• The present study is the first to measure pulmonary artery thickness and stiffness.

• The findings show that impaired vasculogenesis may be a contributory factor in the higher incidence of bronchopulmonary dysplasia in preterm growth restricted infants.

• The study addresses the mechanistic link between fetal programming and vascular architecture and mechanics.

Abstract

Bronchopulmonary dysplasia is the most common respiratory sequelae of prematurity and histopathologically features fewer, dysmorphic pulmonary arteries. The present study aimed to characterize pulmonary artery mechanics and cardiac function in preterm infants with fetal growth restriction (FGR) compared to those appropriate for gestational age (AGA) in the early neonatal period. This prospective study reviewed 40 preterm infants between 28 to 32 weeks gestational age (GA). Twenty infants had a birthweight <10th centile and were compared with 20 preterm AGA infants. A single high resolution echocardiogram was performed to measure right pulmonary arterial and right ventricular (RV) indices. The GA and birthweight of FGR and AGA infants were 29.8 ± 1.3 vs. 30 ± 0.9 weeks (P = 0.78) and 923.4 g ± 168 vs. 1403 g ± 237 (P < 0.001), respectively. Assessments were made at 10.5 ± 1.3 days after birth. The FGR infants had significantly thicker right pulmonary artery inferior wall (843.5 ± 68 vs. 761 ± 40 μm, P < 0.001) with reduced pulsatility (51.6 ± 7.6 μm vs. 59.7 ± 7.5 μm, P = 0.001). The RV contractility [fractional area change (28.7 ± 3.8% vs 32.5 ± 3.1%, P = 0.001), tricuspid annular peak systolic excursion (TAPSE) (5.2 ± 0.3% vs. 5.9 ± 0.7%, P = 0.0002) and myocardial performance index (0.35 ± 0.03 vs. 0.28 ± 0.02, P < 0.001)] was significantly impaired in FGR infants. Significant correlation between RV longitudinal contractility (TAPSE) and time to peak velocity/RV ejection time (measure of RV afterload) was noted (r ² = 0.5, P < 0.001). Altered pulmonary vascular mechanics and cardiac performance reflect maladaptive changes in response to utero‐placental insufficiency. Whether managing pulmonary vascular disease will alter clinical outcomes remains to be studied prospectively.

Key points

• Approximately 5–10% pregnancies are affected by fetal growth restriction.

• Preterm infants affected by fetal growth restriction have a higher incidence of bronchopulmonary dysplasia.

• The present study is the first to measure pulmonary artery thickness and stiffness.

• The findings show that impaired vasculogenesis may be a contributory factor in the higher incidence of bronchopulmonary dysplasia in preterm growth restricted infants.

• The study addresses the mechanistic link between fetal programming and vascular architecture and mechanics.

Introduction

Fetal growth restriction (FGR) affects ∼5–10% pregnancies worldwide and such infants have a 10‐ to 20‐fold increased risk of perinatal morbidity and/or mortality such as bronchopulmonary dysplasia (BPD) compared to non‐growth restricted preterm infants (Rosenberg, 2008). Data from our own institution (AS) note that the incidence of FGR at 28 to 32 weeks gestational age (GA) is 27% (Sehgal A, unpublished observations). BPD is the most common and most significant long term respiratory sequelae of prematurity, with an incidence of ∼60% in infants of ≤25 weeks GA (Chow, 2014). The incidence in infants of 28 weeks GA was high (23%), progressively decreasing with GA. The overall incidence in infants 28–32 weeks was 10.6%. A proportion (15–58%) of these infants (<1000 g birthweight and <32 weeks GA, respectively) (Khemani et al. 2007 & Bhat et al. 2012) have associated pulmonary hypertension. The evolving pattern of BPD may imply an important pathogenic role arising from FGR. A recent study noted that, at 27 weeks GA, 25% of infants without FGR developed BPD compared to 60% of infants with moderate FGR and 90% with severe FGR (Bose et al. 2009). Several other large and contemporary studies on infants <32 weeks GA have also identified similar associations (Lal et al. 2003 & Reiss et al. 2003). Therefore, FGR increases the risk for BPD; adverse respiratory outcomes are attributable to the FGR state, which are independent of the degree of prematurity (Bose et al. 2009; Gortner et al. 2011). On a histopathological level, infants with BPD have pulmonary vascular changes with thicker‐walled pulmonary arteries (and a more distal extension of the muscular arterioles) (Bhatt et al. 2001 & Coalson, 2006). Chronic hypoxaemia (as seen in chronic utero‐placental insufficiency) leads to muscularization of precapillary ‘resistance vessels’ by way of proliferation of vascular smooth muscle and the adventitial fibroblasts (Rabinovitch et al. 1979). In turn, disruption of angiogenesis during critical periods of lung growth (fetal programming) can impair alveolarization, contributing to the pathogenesis of BPD (Thebaud & Abman, 2007; Janer et al. 2008).

Abnormal angiogenesis (influenced by vascular endothelial growth factors and their receptors) appears to be a feature in the pathogenesis of BPD. Data from neonatal mice noted that chronic hypoxia during the first 2 weeks of life (a period of lung development corresponding to human fetal lung development during the third trimester) interferes with alveolar and pulmonary artery development and up‐regulates transforming growth factor β (Ambalavanan et al. 2008). The dynamic component of this impaired angiogenesis most prominently affects the ‘resistance’ vessels; the effects of which can be assessed by measuring surrogates of pulmonary vascular resistance (PVR) by ultrasound and thickness of the major pulmonary arteries. The question for clinicians is whether there is a connection between FGR, its effects on pulmonary (vascular) development and the higher prevalence of BPD in human infants. There are no data linking in utero events such as utero‐placental insufficiency with structural pulmonary vascular changes, altered perinatal pulmonary vasoreactivity, or the right heart maladaptive coupling in the preterm FGR infants. Given the available vascular cross‐sectional area is a critical determinant of PVR, it may be surmised that a reduction because of vascular remodelling (chronic hypoxaemia) may necessitate an increase in pulmonary artery pressure [and the consequent right ventricular (RV) hypertrophy] to maintain pulmonary blood flow. The effects of FGR on the systemic vasculature (aorta/carotid artery) in term neonates in the early postnatal period and children are well documented (Skilton et al. 2005; Crispi et al. 2010; Sehgal et al. 2018). These are thicker and stiffer in FGR neonates, with the effect persisting in paediatric age groups (3–6 years) and adolescence. Evidence in these FGR cohorts has indicated altered cardiovascular function arising from suboptimal conditions, particularly chronic hypoxaemia (Crispi et al. 2010 & Sehgal et al. 2018). Impaired cardiac/vascular coupling has been demonstrated previously in the systemic and pulmonary circulation in adults (19–60 years and 51 ± 24 years old and isolated canine ventricles), respectively (Sunagawa et al. 1983; Kelly et al. 1992; Sanz et al. 2012). Arterial stiffening is an important index of disease progression and contributes significantly to increased afterload. Such changes in the pulmonary circulation in human preterm neonates have not been studied previously.

Using high resolution ultrasound, the present prospective study aimed to characterize, in the early newborn period, the pulmonary artery and RV indices in a cohort of very preterm infants with FGR compared to an AGA cohort. Important afterload and cardiac interactions arising from vascular stiffness in FGR infants may underlie the association with BPD because normal angiogenesis is crucial to optimal alveolarization.

Methods

Ethical approval

The present study was approved by the Monash Health Research Ethics Board (approval number 14197B). Informed written parental consent was obtained and the study conformed to the standards set by the latest version of Declaration of Helsinki, except for registration in a database.

Twenty infants aged between 28 and 32 completed weeks of GA and weighing <10th percentile were recruited and were compared with 20 preterm AGA infants of similar GA. Infants aged <28 weeks were excluded because they more probably have a patent ductus arteriosus. Patency of the duct was an a priori exclusion criteria. Infants with perinatal depression (5 min Apgar score <5), congenital malformations/heart disease or chromosomal abnormalities, or those born to diabetic mothers, were excluded. Cohort characteristics were collected. The male/female ratio was comparable (11/9 vs. 12/8 in FGR and AGA cohorts, respectively). All echocardiographic (ECHO) assessments were performed in the second week of life by the same operator who was blinded to the group. The subjects in the control group were age‐matched to the FGR cohort. The infants were followed up for respiratory sequelae. This included total duration of respiratory support (including low flow), need for home oxygen and incidence of BPD. In infants born <32 weeks GA, the Australia and New Zealand Neonatal Network defines BPD as lung disease with ongoing requirement for supplemental oxygen therapy or ventilation support (high‐flow oxygen, continuous positive airway pressure or mechanical ventilation) at 36 weeks post‐menstrual age. Table 1 depicts the ECHO assessments and their interpretation (Evans & Archer, 1999; Dyer et al. 2006; Koestenberger et al. 2011; Czernik et al. 2012; Levy et al. 2015). These were performed using the the Vivid 7 Advantage Cardiovascular Ultrasound System (GE Medical Systems, Milwaukee, WI, USA) with the infant in the supine position. Offline analysis was performed using EchoPAC (GE Ultrasound, Horten, Norway) software without revealing group identity. All Doppler measurements were calculated from an average of three consecutive cardiac cycles, with the angle of insonation being kept to <15°.

Table 1.

Summary of haemodynamic assessments

	Technique	View	Cursor position	Comment
Ventricular function
RV fractional area change	2‐D	Apical four chamber	Include full view of the right ventricle (base to apex)	[(RV four‐chamber area at end‐diastole – RV four‐chamber area at end‐systole)/RV four‐chamber area at end‐diastole] × 100%
Tricuspid annular peak systolic excursion	M‐mode	Apical four chamber	Tricuspid annulus	Measure of longitudinal contractility
Tricuspid MPI	TDI	Apical four chamber	Tricuspid lateral annulus	(IVCT + IVRT)/RVET
Pulmonary vascular dynamics
TPV/RVETc	PWD	Long axis RVOT	Aligned with the flow, sample at tips of pulmonary leaflets	1/(TPV/RVETc) acts as a surrogate for pulmonary resistance
RPA inferior wall thickness in diastole	CCM TDI	Short axis	Perpendicular to RPA	Measured in end diastole
RPA pulsatile diameter	CMM TDI	Short axis	Perpendicular to RPA	Difference between the RPA internal calibre in diastole and systole

PWD, pulse wave Doppler; CMM, colour m‐mode; TPV, time to peak velocity; RVET, right ventricular ejection time; TAPSE, tricuspid annular peak systolic excursion; RV, right ventricular; 2‐D, two dimensional; RVOT, right ventricular outflow tract; TDI, tissue Doppler imaging; IVCT, isovolumic contraction time; IVRT, isovolumic relaxation time.

The inferior wall of the right pulmonary artery (RPA) was assessed as previously described using the short axis view because it aligns perpendicular to the ultrasound beam (Dyer et al. 2006). The probe was aligned parallel to the RPA with gain settings and scaling adjusted to maximize the detail of the arterial walls. Colour m‐mode tissue Doppler imaging (TDI) allowed for higher resolution and sharper edges (Fig. 1). RPA thickness was measured in end‐diastole. Measurements were conducted offline by two independent observers masked to the grouping. The data were analysed by Bland–Altman analysis. An average of two measurements for each observer was taken and then the average of the readings from both observers was used for analysis. Assessment of agreement using the Bland–Altman techniques in readings between the two observers showed a small bias in measurements: 10.3 (95% limits of agreement from –38.87 to 59.47). The intraclass correlation coefficient for overall assessment was 0.93 (95% confidence interval = 0.86–0.96). The RPA thickness was also indexed to the lumen diameter in diastole. Pulse diameter was calculated as the difference in diastole and systole from colour m‐mode images.

Figure 1. Pulmonary artery imaging using 2‐dimensional and colour Doppler.

A, 2‐D image from modified short axis view showing the right pulmonary artery. B, colour Doppler showing the right pulmonary artery. C, m‐mode tissue Doppler of right pulmonary artery. [Color figure can be viewed at wileyonlinelibrary.com]

Statistical analysis

A preliminary hypothesis was that cardiac and pulmonary artery indices in FGR are worse than those of an AGA cohort. Vascular and cardiac function indices were summarized as the mean ± SD. We assessed the effect of group (n = 20 FGR vs. n = 20 AGA) on vascular parameters via general linear regression models. Data were analysed using Stata, version 14 (StataCorp, College Station, TX, USA) and SPSS, version 18 (SPSS Inc., Chicago, IL, USA). Correlations between the vascular and cardiac parameters were assessed by Pearson's correlation. P < 0.05 was considered statistically significant.

Results

The study cohort characteristics are shown in Table 2. The GA in the two groups was comparable (FGR 29.8 ± 1.3 weeks vs. 30 ± 0.9 weeks, P = 0.78), whereas birthweight was significantly lower (FGR 92.3 ± 168 g vs. 140.3 ± 237 g, P < 0.001). Data regarding smoking were not available for most mothers. Preterm premature rupture of membranes (four infants, two in each group) and chorioamnionitis (two infants, one in each group) was seen in some infants. The two infants with chorioamnionitis were amongst the four who had premature rupture of membranes. None of the infants were intubated and mechanically ventilated. The median (interquartile) duration of respiratory support [FGR 37 (14–52) vs. AGA 8 (5–21) days] and the incidence of BPD (FGR 40% vs. AGA 5%) were significantly higher in the FGR infants. The GA at discharge was significantly greater in the FGR cohort (43 ± 3.1 weeks vs. 40.2 ± 2.8 weeks, P = 0.0006). No infants had pulmonary hypertension necessitating nitric oxide therapy and none were administered postnatal steroids. The age at ECHO assessments in the FGR and AGA infants was 10.5 ± 1.3 days vs. 10.3 ± 1.3 days, respectively (P = 0.55). Table 3 depicts cardiac and pulmonary artery properties. The RPA inferior wall was significantly thicker in the FGR infants (Fig. 2). The wall thickness to lumen ratio in diastole (%) was also significantly increased in the FGR cohort (11.1 ± 2 vs. 8.6 ± 1.1, P = 0.009). The pulsatile diameter was significantly lesser in FGR compared to AGA infants (51.6 ± 7.6 μm vs. 59.7 ± 7.5 μm, P = 0.001). The RV fractional area change and tricuspid annular peak systolic excursion (TAPSE) were lower and the TDI myocardial performance index was higher in the FGR infants, indicating lower systolic and diastolic function. The significance persisted after adjustment for GA. Significant correlation was noted between RV longitudinal contractility (TAPSE) and time to peak velocity/RV ejection time (TPV/RVETc) (measure of RV afterload) in the overall cohort (r ² = 0.5, P < 0.001). Figure 3 depicts individual cohort correlations. Lower TPV/RVETc (higher PVR) was associated with lower contractility. Correlation between tissue level performance (reflected by myocardial performance index) and TPV/RVETc was r = –0.42 and P = 0.006, indicating that higher PVR is associated with lower cardiac performance.

Table 2.

Demographics of the study population

Variable	FGR infants, n = 20 Mean ± SD	AGA infants, n = 20 Mean ± SD	P
Mode of delivery (caesarean), n (%)	5 (25)	5 (25)	1
Antenatal steroids, n (%)	19 (95)	20 (100)	0.9
Duration of respiratory support (days)	37 (14, 52)a	8 (5, 21)a	0.01
Total length of hospital stay (days)	73 ± 21	50 ± 16	<0.001
Home oxygen, n (%)	4 (20)	0 (0)	0.1
Bronchopulmonary dysplasia, n (%)	8 (40)	1 (5)	0.019
Ventilation at ECHO assessment
Room air	7	9	0.6
CPAP/high flow	13	11
Pressure (cm H2O or L min−1)b	6 ± 1	6 ± 0.5	0.7
Fraction of inspired oxygenc	0.27 ± 0.01	0.26 ± 0.02	0.6

^aMedian (interquartile), CPAP‐continuous positive airway pressure. ^bFor those on CPAP/high flow. ^cFor those in oxygen.

Table 3.

Echocardiographic assessments

Variable	FGR, n = 20 Mean ± SD	AGA, n = 20 Mean ± SD	Unadjusted difference (95% CI)	P	GA adjusted (95% CI)	P
Heart rate (beats min–1)	147 ± 5	146 ± 3	–1.25 (–4.08;1.58)	0.377	–1.32 (–4.19;1.54)	0.356
RV diastolic area (cm2)	2.31 ± 0.13	2.18 ± 0.19	0.13 (0.03, 0.23)	0.016	0.13 (0.02, 0.23)	0.019
RV systolic area (cm2)	2.02 ± 0.14	1.86 ± 0.18	0.17 (0.06, 0.27)	0.002	0.16 (0.06, 0.27)	0.003
RV fractional area change (%)	28.7 ± 3.9	32.5 ± 3.1	–3.8 (–6.0, –1.5)	0.002	–3.7 (–6.0, –1.4)	0.002
Tricuspid annular peak systolic excursion (mm)	5.2 ± 0.3	5.9 ± 0.7	–0.7 (–1.0, –0.4)	<0.001	–0.7 (–1.0, –0.3)	<0.001
Myocardial performance index	0.35 ± 0.03	0.28 ± 0.02	0.07 (0.05, 0.09)	<0.001	0.07 (0.05, 0.09)	<0.001
TPV/RVETc	0.27 ± 0.04	0.32 ± 0.03	–0.05 (–0.07, –0.03)	<0.001	–0.05 (–0.07, –0.03)	<0.001
RPA inferior wall thickness (μm)	843 ± 68	761 ± 40	82.1 (46.3, 117.8)	<0.001	84.0 (49.5, 118.5)	<0.001

Figure 2. Right pulmonary artery.

Scatterplot illustrating thickness of the right pulmonary artery. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 3. Correlation between right ventricular contractility (TAPSE) and afterload (TPV/RVETc).

A, fetal growth restriction. B, appropriate for gestational age. [Color figure can be viewed at wileyonlinelibrary.com]

Discussion

There is a growing body of literature supporting the idea that FGR is an important factor contributing to the early developmental origins of impaired lung structure and function (Maritz et al. 2004; Morsing et al. 2012; Pike et al. 2012; Briana & Malamitsi‐Puchner, 2013). Reduced fetal growth may be a surrogate for abnormal intrauterine lung development because the factors that control fetal somatic growth may significantly increase the vulnerability to lung injury in such affected fetuses. The chronic insufficiency of oxygen and nutrients impacts the lung parenchyma, airways and vasculature (Maritz et al. 2005; Orgeig et al. 2010; Morsing et al. 2012; Pike et al. 2012) and possibly explains the higher incidence of BPD in such infants. Smooth muscle thickening of neonatal pulmonary vessels and changes in pulmonary vascular reactivity after chronic hypoxaemia support the role of utero‐placental insufficiency. The present study noted an increased duration of respiratory support and incidence of BPD in FGR infants compared to GA‐matched AGA infants.

The present study builds on pre‐existing literature using high‐resolution ultrasound and found that, in preterm FGR infants, the pulmonary vasculature is thicker with reduced pulsatility. In addition, important physiological interactions between pulmonary vascular indices and RV function were noted. The role of the proximal conduit arteries is to dampen the pressure oscillations originating from intermittent RV ejection. In turn, this cushioning capacity is influenced by wall stiffness and distensibility. This concept is similar to the effect of aortic stiffness on the distal arterioles and the subsequent organ damage noted previously. This is the first study to show that proximal pulmonary artery changes are possible contributors to respiratory morbidity in FGR cohort via a possible impact on the distal pulmonary vasculature.

BPD: the vascular hypothesis

Perinatal factors including FGR predict persistent respiratory disease at 1 year of life just as accurately as the diagnosis of BPD (Keller et al. 2017). BPD pathophysiology is multifactorial and FGR severity can affect alveolarization, as well as angiogenesis. The role of arterial stiffness appears to be crucial because the lack of waveform cushioning by the major arteries affects the heart by way of back pressure changes (hypertrophy/dilatation) (known as pulsatile afterload) but, more importantly, exposes the pulmonary resistance vessels to higher pulsatile stress, thereby accelerating microvascular disease (Wang & Chesler, 2011). Stiffness of conduit pulmonary arteries increases distal strain damage (increased smooth muscle cell proliferation, leukocyte adhesion, inflammatory gene expression) (Li et al. 2009). This effect mirrors changes in the systemic circulation where aortic stiffening caused renal arteriolar damage (O‘Rourke & Safar, 2005 & O'Rourke et al. 2007). The deleterious effects of the exposure to such high pulsatile stress on the kidneys (systemic circulation) have been demonstrated previously in children (Takenaka et al. 2005). In the FGR cohort, the urinary microalbumin and albumin–creatinine ratio were noted to be significantly higher at 18 months, indicating microvascular glomerular damage (Zanardo et al. 2011). Our data demonstrated significantly thickened pulmonary vasculature with markedly reduced pulsatility. In terms of dynamics, a lower TPV/RVETc indicated elevated PVR. We postulate that high pulsatile stress may be similarly deleterious to the pulmonary arterioles and lead to microvascular leakage, which may contribute to pulmonary oedema. The resultant decreased compliance may contribute to a longer duration of respiratory support and a higher incidence of BPD in preterm FGR infants. These vascular effects (coupled with the effects on the heart) are possibly related to vascular arterial remodelling.

FGR and alterations in pulmonary vasculature: mechanisms

We now discuss possible mechanisms linking utero‐placental insufficiency and pulmonary vascular changes. Figure 4 summarizes the various mediators of the effects of FGR on the pulmonary vasculature and the putative clinical effects. Human epidemiological and clinical studies and data from animal experiments (sheep; 129–141 days GA) have noted a significant burden of respiratory illnesses following FGR (Morrison, 2008 & Orgeig et al. 2010). The hypoxia signalling cascade regulates normal fetal lung angiogenesis, vascular remodelling, surfactant maturation and alveolarization. Vascular endothelial growth factor expression plays a crucial role (Gebb & Jones, 2003; Groenman et al. 2007; McGillick et al. 2016). In FGR sheep, decreased fetal pulmonary alveolarization and reduced pulmonary vessel density, pulmonary artery endothelial cell function and eNOS signalling have been noted. This adverse impact is indicated by a reduced pulmonary vessel density compared to the AGA cohort in this present study in which the FGR fetuses were created by exposing pregnant ewes to elevated ambient temperatures (40°C for 12 h; 35°C for 12 h) from 33.4 ± 0.3 days GA until 115.3 ± 0.4 days GA (Rozance et al. 2011). Angiogenesis in turn regulates alveolarization during pulmonary development (Jakkula et al. 2000; Le Cras et al. 2002; Stenmark & Abman, 2005; Thebaud & Abman, 2007). Work on the sheep model also noted the effects of FGR on lung architecture that were apparent at 8 weeks after birth, and were still evident at 2 years (Maritz et al. 2005). In neonatal mice, chronic hypoxaemia during the first 2 weeks of life interferes with both alveolar and pulmonary artery development by way of increased endothelial permeability, heightened smooth muscle tone and an enhanced thrombotic state. Of concern, even a short period (2 weeks) of hypoxaemia appears sufficient to induce extracellular matrix thickness in these rodents (Ambalavanan et al. 2008). In summary, studies from various animal models indicate the impact of FGR on pulmonary vasculature.

Figure 4. FGR.

Relevant mediators in FGR. [Color figure can be viewed at wileyonlinelibrary.com]

The perinatal period is also associated with significant elastin production in the pulmonary trunk (Leung et al. 1977) and is particularly sensitive to modulation by hypoxaemia during this time of rapid growth. Reduced elastin content in large pulmonary arteries predisposes to elevated RV pressures and hypertrophy (Shifren et al. 2008). Because the rates of elastin synthesis increase to a maximum in the perinatal period; our cohort of FGR infants born between 28 and 32 weeks may be particularly vulnerable. Thus, in utero disruption in the synthesis and deposition of adequate amounts of elastin and its replacement with collagen (100 times greater stiffness than elastin) in fetal life may lead to low arterial compliance (Martyn & Greenwald, 1997).

Exposure to adverse environmental conditions during certain stages of fetal development may be crucial. The physiological consequences of the changes include increased arterial resistance/stiffness and decreased compliance, contributing to increased RV afterload. These parameters have prognostic implications because RV functional status and cardiac index are strong predictors of survival (Ghio et al. 2001).

Interaction of pulmonary afterload and cardiac function

A recent ECHO study on the effects of hypoxaemia in rats noted pulmonary hypertension, as indicated by decreased pulmonary artery time to peak velocity and increased dilatation of RV diameter compared to age‐ and sex‐matched controls. Pregnant rats were exposed to hypoxia (12% O₂) or normoxia (21% O₂) between days 15 and 21 of pregnancy. Assessments at 12 months of age noted signs of left ventricular hypertrophy, diastolic dysfunction and pulmonary hypertension (Rueda‐Clausen et al. 2009). Prenatal hypoxaemia (in the mouse model exposed to chronic hypoxia for 4 weeks and assessments made at 40 weeks) also led to greater thickness of the muscularis media, which, together with adventitial proliferation, decreased the luminal diameter (Pietra et al. 2004 & Bonnet et al. 2006). This remodelling and the consequent stiffness in turn regulate pressure and flow wave velocities in the pulmonary bed and affect afterload (Milnor et al. 1969 & Weinberg et al. 2004). We found similar effects in the specific parameters. Similar interactions between systemic afterload and cardiac forces have been demonstrated previously in newborns with FGR and 3–18‐year‐old children (Rowland & Gutgesell, 1995 & Sehgal et al. 2018). Such important correlations between pulmonary forces, however, especially in FGR infants, have not been demonstrated before. This afterload/contractility relationship may be ideal for longitudinal studies monitoring disease progression. The altered RV function is probably related to pressure overload and intrinsic myocardial issues as a maladaptive response. These include exposure to increased myocardial workload in utero (Tintu et al. 2009; Verburg et al. 2008; Fouzas et al. 2014), alterations in arterial structure and vascular tone (Rouwet et al. 2002), and altered muscle fibre architecture (Greenbaum et al. 1981). Decreased cardiac sarcomeric proteins, a compensatory increase in glycogen and collagen deposition (Sohn et al. 1997 & Tintu et al. 2009), with interstitial fibrosis are also noted in the presence of FGR (Menendez‐Castro et al. 2011).

The TPV/RVETc ratio is the commonly used Doppler assessment of pulmonary artery dynamics and allows semiquantitative measure of PVR; the lower the ratio, the higher the PVR (ratio <0.1 signifies greater severity). It is a reliable surrogate of RV afterload and PVR (1/TPV: RVETc) (Evans & Archer, 1999 & Howard et al. 2012). Its application in infants with chronic neonatal lung disease previously noted a significant and negative correlation between TPV/RVETc and invasive PVR index (Milnor et al. 1969 & Ziino et al. 2010). Although measured in the main pulmonary artery, TPV/RVETc is more reflective of changes in the pulmonary resistance vessels. Infants in the present study possibly have pulmonary vascular disease rather than frank pulmonary hypertension. Invasive assessments, although providing new insights into the pulmonary capacitance and vaso‐reactivity, are impractical in such small FGR infants. Non‐invasive measures of proximal pulmonary artery compliance in the form of vessel morphometery and beat‐to‐beat pulsatility (pulsatile diameter), are best placed to fill this void.

Although representing a useful vessel for Doppler studies because of its orientation, the main pulmonary artery is not conducive for the above assessments where the vessel needs to be perpendicular to the ultrasound beam. The RPA overcomes this limitation. Using non‐invasive colour TDI ECHO, Dyer et al. (2006) recently investigated diameter and compliance in the RPA in a paediatric population ranging from 0.6 to 17 years of age. The assessments correlated well with invasive i.v. pulmonary artery catheter measurements and patients with pulmonary hypertension had lower compliance (Dyer et al. 2006). The present study is the first assessment of RPA thickness and dynamics in the neonatal age group. It provides useful physiological information in settings where ventricular‐arterial coupling may be of relevance. Compared to conventional ECHO, TDI is a relatively recent addition to the neonatal literature, allowing direct measurement of regional myocardial velocities. It has a greater sensitivity compared to conventional measures and is relatively preload independent (Sohn et al. 1997 & Sehgal et al. 2016). The feasibility of TDI for assessing fetal (30 ± 3 weeks GA and neonatal (term ∼39 weeks GA) cardiac function in FGR cohorts has been reported previously (Comas et al. 2010; Altin et al. 2012; Sehgal et al. 2017). In the present study, we noted impairment of cardiac function in FGR infants compared to GA and post‐natal age matched AGA controls. TAPSE is also a relatively recent addition to RV assessments (Koestenberger et al. 2011). It is a simple and highly reproducible measure of longitudinal RV function and is probably not influenced by imaging artefacts. Its clinical relevance was recently noted in a cohort of infants >35 weeks GA with persistent pulmonary hypertension of the newborn, where the sensitivity/specificity for ECMO/death when TAPSE was <4 mm was 56% and 85%, respectively (Malowitz et al. 2015).

Our data are presented adjusted for GA. Although the effect of prematurity is important, a previous study (Cheung et al. 2004) compared systemic arterial stiffness and blood pressure among children who were born preterm and FGR, or preterm appropriate for GA, or term appropriate for GA. The preterm FGR cohort was delivered at 32.3 ± 2 weeks GA (comparable to our cohort). On assessments performed at 8.2 ± 1.7 years, only children born preterm FGR had increased arterial stiffness and elevated mean blood pressure. Unfortunately, there are no comparable data for the early postnatal period.

The strengths of the present study include focussed cardiac and vascular assessments using conventional and newer but validated TDI ECHO parameters and the demonstration of afterload/cardiac forces coupling. This is a preliminary prospective hypothesis generating study involving a small number of infants. A larger multicentric cohort investigating possible intervention strategies is better suited for achieving a greater understanding on the topic. In summary, the present study builds on the current understanding of BPD as being multifactorial in pathogenesis with FGR as an important aetiopathological factor. The vascular hypothesis that we propose mimics the vascular affliction of the systemic vasculature, as discussed in multiple studies across age groups (newborn to adolescence) in FGR cohorts.

Additional information

Competing interests

The authors declare that they have no competing interests.

Author contributions

The study was performed at Monash Newborn, Monash Children's Hospital, Melbourne, Australia. AS, SMG, SM, BA, SLM and GRP were responsible for conception of the study. SM, BA and SLM were responsible for the design of the study. AS and SMG were responsible for data acquisition and analysis. AS, SMG, SM, BA, SLM and GRP were responsible for data interpretation. AS was responsible for writing the first draft and revising it critically. SMG, SM, BA, SLM and GRP were responsible for revising the draft for important intellectual content. All authors approved the final version of the manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.

Funding

This study was supported by an ANZ Trustees/Equity Trustees Medical Research & Technology in Victoria Grant, a National Health and Medical Research Council and National Heart Foundation of Australia Fellowship [GRP: 1105526], an Australian Research Council Future Fellowship (SLM: FT130100650), a Rebecca L. Cooper Medical Research Foundation Fellowship (GRP) and the Victorian Government's Operational Infrastructure Support Program.

Biography

Arvind Sehgal is a Neonatologist at Monash Childrens Hospital and Professor of Paediatrics at Monash University, Melbourne, Australia. He completed a Fellowship in Neonatal Cardiology at University College Hospital, London, UK followed by a Neonatal Perinatal Fellowship at The Hospital for Sick Children, University of Toronto, Canada. This research involves assessment of pulmonary vasculature in growth restricted preterm infants. Future research will focus on interventions that might alleviate these changes. His current research priorities include collaboration in multicentric randomized controlled trials. The highlight of his career includes completion of a PhD (2016) when working as a full time intensivist and being invited to chair and conducted an Invited Science Symposium at the 2017 Pediatric Academic Societies Annual Meeting in San Francisco.