Lung vascular development occurs as highly choreographed sequence, regulated by hypoxia inducible factors, vascular endothelial growth factor, nitric oxide, and other transcription factors and mediators.
In addition to arterial vessels, pulmonary veins are now understood to be highly reactive vessels that contribute to the overall regulation of pulmonary vascular resistance in the fetus and newborn.
Antenatal pulmonary vascular development can be disrupted by events such as placental insufficiency, genetic abnormalities such as Down syndrome, prolonged oligohydramnios, and congenital diaphragmatic hernia.
Postnatal development of the lung circulation can be disrupted by numerous stressors such as preterm birth, asphyxia, and hypoxia or hyperoxia.
Numerous questions and controversies remain, including the role of cardiac dysfunction in congenital diaphragmatic hernia and the causes of acute and chronic pulmonary hypertension in the premature infant.
Development of the Fetal Pulmonary Circulation
The development of the pulmonary vasculature during fetal and neonatal life is highly coordinated with airway growth and plays a key role in normal lung development. Compared with adult pulmonary vascular disease, disruption of lung vascular development plays a central role in the pathobiology of pulmonary vascular disease and airway development in the neonate and young infant.
Lung development is classically divided into five overlapping stages in humans and rodents on the basis of gross histologic features. They are termed the embryonic (weeks 4–7 of gestation), pseudoglandular (weeks 5–17), canalicular (weeks 16–26), saccular (weeks 24–38), and alveolar stages (week 36 to infancy). The development of the pulmonary vasculature is closely correlated with and interacts with airway growth. Lung vascularization initially originates in the mesenchyme, distal to the epithelium. In response to epithelial-derived vascular endothelial growth factor (VEGF), the endothelial cells move toward the epithelium, where they form the epithelium-capillary interface needed for gas exchange. Growth of the lung vasculature continues after birth and into adulthood.
Physiology of the Fetal Pulmonary Circulation
Pulmonary hypertension (PH) is a normal physiologic state during fetal life and permits survival on placental support. Fetal pulmonary vascular resistance (PVR) is high in part because of hypoxic pulmonary vasoconstriction ( Fig. 3.1 ). In the fetal lamb, pulmonary arterial blood has a partial pressure of oxygen (P o 2 ) of approximately 18 mm Hg and oxygen saturation of 50%. Because of high PVR, only about 16% of the combined ventricular output is directed to the lungs; the remainder passes through the ductus arteriosus to the descending aorta. The blood is then oxygenated in the placenta and returns to the body through the umbilical vein, with a P o 2 of ∼32 to 35 mm Hg in lambs. The difference in oxygen saturation between the umbilical vein (85%) and umbilical artery (52%) during fetal life is similar to the difference between the pulmonary vein/aorta (95%–100%) and pulmonary artery (60%–70%) in an adult. The fetus thus achieves normal oxygen delivery at the low P o 2 levels needed for normal lung development.
In human fetuses, Doppler studies demonstrate that the pulmonary blood flow is only 13% of combined ventricular output at 20 weeks’ gestation (canalicular stage), representing a nadir during lung development (see Fig. 3.1 ). This finding is largely secondary to the lower cross-sectional area of the very immature pulmonary vascular bed. Furthermore, in fetal lambs at an equivalent point in gestation (∼65% gestation), pulmonary blood flow does not increase in response to hyperoxia and PVR does not increase in response to hypoxia. Similarly, in human pregnancies, maternal hyperoxygenation with face mask oxygen at 20 to 26 weeks’ gestation does not result in pulmonary vasodilation. Birth at this gestational age (23–26 weeks) is associated with a 2% risk of clinical PH , and perhaps explains the high rates of inhaled nitric oxide (NO) therapy (6%–8%) in extremely premature infants.
As the lung develops through the early saccular stage, rapid proliferation of pulmonary vessels and a marked increase in cross-sectional area of the pulmonary vascular bed occurs, which decreases fetal PVR. At the same time, pulmonary vessels become more reactive to vasoconstrictors such as hypoxia and endothelin and vasodilators such as oxygen. The net result is higher PVR and increased reactivity of the pulmonary vasculature. The maternal hyperoxygenation test (administered with 60% oxygen by face mask) has been proposed to measure the ability of fetal pulmonary arteries to vasodilate in response to oxygen in late gestation and predict pulmonary vascular reactivity and survival in congenital diaphragmatic hernia (CDH). Later in fetal life, PVR becomes very sensitive to small changes in P o 2 . In nonhuman primate fetuses, Arraut et al. evaluated the effect of maternal hypoxemia (by administration of 12% oxygen) and hyperoxemia (by administration of 100% oxygen) on the pulsatility index (PI) of the right pulmonary artery. Maternal hypoxemia increased fetal right pulmonary arterial PI by fivefold, suggesting fetal pulmonary vasoconstriction, and maternal hyperoxemia decreased right pulmonary arterial PI by fourfold and increased ductus arteriosus PI. Maternal oxygenation status did not affect the umbilical arterial PI or ductus venosus PI, suggesting that umbilical flow is not influenced by and does not regulate fetal oxygenation. Similarly, Konduri et al. showed that an increase in pulmonary arterial P o 2 of 7 mm Hg resulted in a threefold increase in pulmonary blood flow in fetal lambs. These changes in PVR will determine the distribution of fetal cardiac output and oxygen delivery to the brain and heart. In pathologic conditions such as CDH, reduced pulmonary venous return and left ventricular filling may contribute to left ventricular hypoplasia, emphasizing the important role of pulmonary blood flow during fetal life.
Pulmonary veins have been traditionally regarded as passive conduit vessels, but they are now recognized as reactive vessels that contribute to the overall regulation of PVR. In the fetus, pulmonary veins contribute a significant fraction to total PVR, and they may play a more important role in regulating the fetal and newborn pulmonary circulation than in adults ( Fig. 3.2 ). In perinatal sheep, NO stimulated endogenously by acetylcholine or given exogenously causes greater relaxation and accumulation of cyclic guanosine monophosphate (cGMP) in pulmonary veins than in arteries. At birth, the veins, as well as the arteries, relax in response to NO and dilator prostaglandins, thereby assisting in the fall in PVR ( Fig. 3.3 ). These effects are oxygen dependent and modulated by protein kinase G. In a number of species, including the human, pulmonary veins are also the primary sites of action of certain vasoconstrictors, such as endothelin and thromboxane (see Figs. 3.2 and 3.3 ).
Mediators of Early Pulmonary Vascular Development
The hypoxic conditions of fetal life support the tremendous lung vascular growth that occurs before birth. Hypoxia-inducible factors (HIFs) are regarded as the “master regulators” of the transcriptional response to hypoxia and are involved in angiogenesis, survival, and metabolic pathways ( Fig. 3.4 ). HIFs are heterodimers consisting of oxygen-sensitive α-subunits (HIF-1α, HIF-2α) and constitutively expressed β-subunits. Hypoxia stabilizes the α-subunit, leading to nuclear accumulation and activation of multiple target genes. HIF-1 regulates genes involved in angiogenesis (e.g., VEGF), oxygen transport (e.g., erythropoietin), and energy metabolism (e.g., glycolytic enzymes), among others. HIFs are constitutively expressed in multiple fetal pulmonary cell types, including endothelial, smooth muscle, and epithelial cells. The importance of HIFs in fetal lung development has been demonstrated by studies revealing that deletion of HIF-1 causes embryonic lethality, and deletion of HIF-2 reduces VEGF levels and leads to early death owing to respiratory failure. In the adult lung, hypoxia induces abnormal vascular remodeling, potentially by inducing HIF activity. In contrast, hypoxia is the normal fetal condition and is a required environmental stimulus to sustain normal fetal lung and vascular development.
Although NO is best known for its vasoactive properties, it also plays an important role in the structural development of the pulmonary vasculature. Lung endothelial NOS (eNOS) mRNA and protein are present in early fetal life in rats and sheep and increase with advancing gestation in utero. The expression and activity of eNOS are regulated by multiple factors, including hemodynamic forces, hormonal stimuli (e.g., estradiol), paracrine factors (including VEGF), substrate and cofactor availability, oxygen tension, and others. Numerous studies suggest the importance of NO-cGMP signaling in lung development. Lungs of fetal and neonatal mice deficient in eNOS have reduced alveolarization and vascularization and are more susceptible to the effects of hypoxia on vascular and alveolar growth. Furthermore, mice deficient in soluble guanylate cyclase (sGC), the main target enzyme for NO, have decreased lung volumes and small airways.
VEGF is a key regulator of lung vascular growth and development during fetal and postnatal life. VEGF transcription is regulated by HIF, and its signaling is transduced via two transmembrane tyrosine kinase receptors: VEGFR-2 and VEGFR-1. VEGF ribonucleic acid (RNA) and protein are localized to distal airway epithelial cells, whereas VEGFR-1 and VEGFR-2 messenger RNA (mRNA) expression is localized to the pulmonary endothelial cells closely approximated to the developing epithelium.
The fundamental importance of VEGF for vascular development has been demonstrated by several studies that inactivate or knock out VEGF or its receptors. Each of these produces a lethal phenotype that is characterized by deficient organization of endothelial cells. Furthermore, VEGFR-1 and VEGFR-2 inhibitors (e.g., SU5416) impair alveolar development in fetal and newborn rodent models, producing pathologic findings similar to those seen in clinical bronchopulmonary dysplasia (BPD). Even in adult rats, long-term treatment with SU5416 causes PH and enlarges the air spaces, suggesting that normal VEGF function is required not only for the formation but also for the maintenance of the pulmonary vasculature and alveolar structures well after lung development is completed.
Recent studies suggest that VEGF-induced lung angiogenesis is in part mediated by NO. VEGF inhibition is associated with decreased lung eNOS protein expression and NO production; treatment with inhaled NO improves vascular and alveolar growth after VEGF inhibition. However, in neonatal mice that are eNOS deficient, recombinant human VEGF protein treatment restores lung structure after exposure to mild hyperoxia, suggesting that VEGF operates in part through mechanisms independent of eNOS.
Numerous transcription factors important to lung vascular development have been identified. The forkhead box (Fox) family of transcription factors regulates expression of genes involved in cellular proliferation and differentiation. Newborn mice with low Foxf1 levels die with defects in lung vascularization and alveolarization, and endothelial-specific deletion of Foxf1 produces embryonic lethality, growth restriction, and vascular abnormalities in the lung, placenta, and retina. These findings are directly relevant to human lung development, as Foxf1 haploinsufficiency is found in 40% of infants with alveolar capillary dysplasia (ACD), a lethal disorder of lung vascular development.
Nuclear factor kappa B (NF-κB) is a transcription factor traditionally associated with inflammation, but recent data suggest it may play a very different, protective role in the neonatal lung. Constitutive NF-κB expression is higher in the neonatal lung than in the adult lung and inhibiting it impairs in vitro pulmonary endothelial cell proliferation and angiogenesis. Blocking NF-κB activity during the alveolar stage of lung development in neonatal mice induced alveolar simplification and reduced pulmonary capillary density similar to that observed in BPD, effects that appear to be regulated by VEGFR-2.
Lung endothelial progenitor cells (EPCs) have been recently identified as mediators of lung development, although their mechanistic role is not yet well understood. In rats, microvascular pulmonary endothelial cells proliferate twice as fast as endothelial cells isolated from large pulmonary arteries. These cells, called resident microvascular endothelial progenitor cells (RMEPCs), are highly proliferative and express endothelial cell markers (CD31, CD144, eNOS, and von Willebrand factor) and progenitor cell antigens (CD34 and CD309). Thus the pulmonary microcirculation seems to be enriched with EPCs that support vasculogenesis while maintaining endothelial microvascular functionality.
Resident microvascular EPCs also share features of human cord blood–derived endothelial colony-forming cells (ECFCs). Developing human fetal and neonatal rat lungs contain ECFCs with robust proliferative and vasculogenic potential. The functionality of these cells can be disrupted during or after birth: for instance, human fetal lung ECFCs exposed to hyperoxia in vitro proliferate less and form fewer capillary-like networks. These findings suggest a role for ECFCs in lung repair and if their function is impaired, it could contribute to the arrested alveolar growth after extremely preterm birth. Consequently, in rodents, exogenous administration of human cord blood–derived ECFCs restored alveolar and lung vascular growth in hyperoxic rodents. However, lung engraftment was low, suggesting that the ECFCs support lung growth and repair through paracrine effects.
Mediators of Early Pulmonary Vascular Function
As gestation progresses, NO and cGMP become central to the emergence of pulmonary vascular reactivity. VEGF acutely releases NO and causes rapid pulmonary vasodilation in vivo; conversely, chronic inhibition of VEGF receptors downregulates eNOS and induces PH in the late-gestation fetus. Both findings point to its importance in the development and function of the developing pulmonary vasculature. Inhibition of eNOS increases basal PVR as early as 75% gestation (112 days) in the fetal lamb, indicating that endogenous NOS activity contributes to vasoregulation during late gestation. Pulmonary vasodilation in response to NO (an endothelium-independent mediator) precedes the response to endothelium-dependent mediators such as acetylcholine and oxygen. The response to NO is dependent on activity of soluble guanylate cyclase in the smooth muscle cell (see Fig. 3.3 ). In the ovine fetus, sGC mRNA levels are low during early preterm (126-day) gestation and markedly increase toward the end of third trimester. Low levels of pulmonary arterial sGC activity during late canalicular and early saccular stages of lung development could partly explain the variable response to inhaled NO (iNO) observed in some extremely preterm infants. Intracellular cGMP levels are also tightly regulated by cGMP-specific phosphodiesterase type 5 (PDE5) activity. PDE5 expression and activity increase during late gestation, and it plays a critical role in pulmonary vasoregulation during the perinatal period.
Pulmonary endothelial cells produce the prostaglandin (PG) molecules PGI 2 and PGE 2 , which are both potent vasodilators. Prostacyclin (PGI 2 ) acts on its receptor in the smooth muscle cell to produce cyclic adenosine monophosphate (cAMP), which also mediates smooth muscle cell vasodilation (similar to cGMP, see Fig. 3.3 ). cAMP is inactivated by cAMP-specific phosphodiesterase 3A (PDE3A). In isolated pulmonary resistance vessels of term fetal lambs, the cyclooxygenase inhibitor indomethacin constricts arteries under higher oxygen tensions but has no effect under lower concentrations, suggesting that oxygen may regulate the synthesis or downstream signaling of the dilator prostanoids. Although stimuli such as shear stress induce release of PgI 2 , overall, prostaglandin release appears to play a less important role than NO in regulating fetal and transitional pulmonary vascular tone.
Constrictors also play a role in regulating the pulmonary vascular tone of the fetus. Lipid mediators, such as thromboxane A 2 , leukotrienes C 4 and D 4 , and platelet-activating factor are potent pulmonary vasoconstrictors. Although thromboxane A 2 has been implicated in animal models of group B streptococcal sepsis, it does not appear to influence PVR in the normal fetus. Some data suggest that leukotrienes and platelet-activating factor may influence PVR during fetal life and transition but remain inconclusive. Endothelin-1 (ET-1) is produced by vascular endothelium and acts on the ET-A receptors in the smooth muscle cell to induce vasoconstriction by increasing ionic calcium concentrations. A second endothelial receptor, ET-B, on the endothelial cell stimulates NO release and vasodilation (see Fig. 3.3 ). Prepro-ET-1 mRNA (the precursor to ET-1) has been identified in fetal rat lung early in gestation, and high circulating ET-1 levels are present in umbilical cord blood. Although capable of both vasodilator and constrictor responses, ET-1 appears to primarily act as a pulmonary vasoconstrictor in the fetal pulmonary circulation.
Endogenous serotonin (5-HT) production is another contributor to the high PVR of the fetus. Infusions of 5-HT increase PVR and infusions of ketanserin, a 5-HT 2A receptor antagonist, decrease fetal PVR in a dose-related fashion. Conversely, brief infusions of selective serotonin reuptake inhibitors (SSRIs), such as sertraline and fluoxetine, cause potent and sustained elevations of PVR. Together, these findings suggest that 5-HT causes pulmonary vasoconstriction and contributes to maintenance of high PVR in the normal fetus through stimulation of 5-HT 2A receptors and Rho kinase activation. These findings have important implications for SSRI treatment for maternal depression, as described in later text.
Transitional Circulation and Postnatal Pulmonary Vascular Development
A rapid and dramatic series of circulatory events occur at birth as the fetus transitions to extrauterine life. After birth and initiation of air breathing, a number of mechanisms operate simultaneously to rapidly reduce pulmonary arterial pressure and increase pulmonary blood flow. Of these, the most important stimuli appear to be ventilation of the lungs and an increase in oxygen tension (see Figs. 3.1 and 3.2 ). Pulmonary blood flow increases by eightfold, resolving fetal PH. Clamping of the umbilical cord removes the low-resistance placental circulation, increasing systemic arterial pressure as pulmonary arterial pressure falls. In some infants with in utero adverse events or with abnormalities of pulmonary transition at birth, PH persists into the newborn period, resulting in the syndrome of persistent pulmonary hypertension of the newborn (PPHN).
The vascular endothelium releases vasoactive products that play a critical role in achieving rapid pulmonary vasodilation. Pulmonary endothelial NO production increases markedly at the time of birth. Inhibitors of NOS activity (e.g., nitro- l -arginine) attenuate the decline in PVR after delivery of fetal lambs, suggesting that the release of NO may be responsible for 50% of the rise in pulmonary blood flow at birth. Oxygen is an important catalyst for this increased NO production. In near-term fetal lambs, maternal hyperoxia induced by hyperbaric oxygenation increased pulmonary arterial P o 2 from 19 ± 1.5 to 48 ± 9 mm Hg, and pulmonary blood flow from 34 ± 3.3 to 298 ± 35 mL/kg per minute, a rise that nearly replicates the normal transition and is blocked by pretreatment with NOS inhibitors. However, mice deficient in eNOS can successfully make the transition at birth without evidence of PPHN, suggesting the presence of alternate or compensatory vasodilator mechanisms, such as upregulation of other NOS isoforms or dilator prostaglandins. Interestingly, eNOS-deficient mouse pups develop PH after relatively mild decreases in arterial partial pressure of oxygen (Pa o 2 ) and have higher neonatal mortality when exposed to hypoxia after birth. It is possible that eNOS deficiency alone may not be sufficient for the failure of postnatal adaptation, but that a decreased ability to produce NO in the setting of a perinatal stress such as hypoxia or inflammation may contribute to the development of postnatal PH. Expression of sGC also peaks in late gestation in rats and sheep, which may explain why the immediate response to NO is greater in neonates than in any other reported age group. Pulmonary expression of PDE5 also peaks in the immediate newborn period in sheep and rats (see Fig. 3.4 ). Together, these events appear to create the ability to finely regulate vascular cGMP concentrations in the transitional and early neonatal period.
The arachidonic acid–prostacyclin pathway also plays an important role in the transition at birth. Rhythmic lung distention and shear stress stimulate both PGI 2 and NO production in the late-gestation fetus, although the effect of oxygen tension is predominantly on NO activity. Phosphodiesterase 3A catalyzes the breakdown of cAMP (see Fig. 3.3 ) and appears to create important crosstalk with the cGMP pathway.
Less is known about the pulmonary vascular transition after preterm birth, although similar mechanisms appear to be in effect. In premature lambs at ∼70% of gestation (112–115 days), the pulmonary vasodilator responses to rhythmic distention of the lung or increased Pa o 2 are partly due to stimulation of NO release. In human preterm infants, the decrease in pulmonary arterial pressure after birth is significantly slower compared with term infants, particularly if respiratory distress syndrome also exists. Pulmonary arterial pressure elevations may persist for a number of days in extremely preterm infants. Skimming et al. have questioned whether a “natural” increase in PVR benefits the preterm infant by reducing the ductal steal and stabilizing systemic circulation. However, recent prospective studies clearly indicate that early PH in extremely preterm infants is associated with BPD and late PH, so it is more likely that this indicates abnormal vascular development or function.
After birth, structural development of the lung and its vasculature continues. More than 90% of lung alveolarization occurs postnatally, with a prominent surge between birth and 6 months of age. Similarly, there is marked growth and development of the microvascular network during the alveolarization phase. A double capillary network is characteristic of the fetal and neonatal lung, but as alveolarization progresses, the interalveolar septae thin and the double capillary layer fuses into the single layer characteristic of the mature vasculature. The capillary network continues to expand its surface area through childhood by nearly 20-fold.
Features of Abnormal Pulmonary Vascular Development
The histologic features of early neonatal PPHN have been described in animal models and in fatal cases of PPHN in term infants. In two autopsy series of infants with PPHN, vascular remodeling resulted in muscularization of the smallest arteries (<30 μM external diameter) at the level of the alveolar duct and wall and a doubling of the medial wall thickness of the intraacinar arteries ( Fig. 3.5 ). These findings suggest that structural maldevelopment of the peripheral pulmonary arterial bed begins in utero and does not merely represent a failure of the fetal pattern to regress. Very similar patterns of remodeling are observed in animal models of PPHN, including the lamb model of antenatal ductal ligation.
Thickening of the adventitia is observed in remodeled pulmonary vessels (see Fig. 3.5 ) and likely contributes to pulmonary artery stiffness. The adventitial cells (including fibroblasts, pericytes, progenitor cells, and so on) also appear to be regulators of vascular wall function from the “outside in.” For example, NO is less potent when administered to the adventitial side of vessels. This may be partly due to the presence of constitutively active NADPH oxidase in adventitial cells, which generate superoxide anions that actively scavenge NO.
In contrast to PPHN, after very preterm birth, the developing lung is exposed to an extrauterine environment that disrupts the normal fetal vascular developmental patterns. On histology, the lungs of very preterm infants with BPD display evidence of arrested development, with reduced numbers of both alveoli and intraacinar arteries. The pulmonary circulation in animal models and infants with BPD is characterized by vascular pruning, decreased vascular branching, and altered patterns of vascular distribution within the lung interstitium. Similar to early PPHN, smooth muscle proliferation also extends abnormally into the smaller peripheral arteries. In addition, intrapulmonary bronchopulmonary anastomoses have recently been identified that act as arteriovenous shunts that contribute to hypoxemia. These anastomotic vessels may represent a compensatory mechanism to overcome the reduction in vascular surface area or act as a protective “pop-off” mechanism to reduce the severity of PH and protect the right ventricle.
Signaling abnormalities in the remodeled vasculature include decreased expression of endothelial NOS and reduced urinary levels of NO metabolites. sGC expression and activity is also diminished in animal models of neonatal PH and CDH, which is partly secondary to oxidation of sGC that renders it NO-insensitive. Because NO and cGMP also inhibit vascular smooth muscle growth, it is likely that a combination of diminished eNOS expression, inactivation of sGC, and reduced cGMP levels also contribute to excessive muscularization of pulmonary vessels in PPHN and BPD. Ventilation with high concentrations of inspired oxygen and exposure to reactive oxygen species also decreases cGMP levels through increasing PDE5 activity, effects that appear to be mediated by reactive oxygen species produced from the mitochondria (see Fig. 3.3 ). In fetal lambs with PPHN, pulmonary prostacyclin synthase and PGI 2 receptor (IP) protein levels in the lung are decreased, but levels of adenylate cyclase and PDE3A are not altered.
Circulating levels of endothelin (ET-1), a potent vasoconstrictor and smooth muscle mitogen, are increased in human infants with PPHN, and lung and vascular ET-1 levels are increased in fetal lambs with PPHN. ET-1 also appears to be a marker for chronic PH, in that infants with CDH and poor outcomes have higher plasma ET-1 levels at 2 weeks of age and severity of PH than infants discharged breathing room air. The constrictor effects of endothelin are mediated in part through activation of the RhoA-Rho-kinase (ROCK) pathway. Increased Rho-kinase activity leads to phosphorylation of myosin light-chain kinase, which in turn increases intracellular calcium and causes vascular contraction. The ROCK pathway plays an important role in hypoxic pulmonary vasoconstriction and as a mediator of the impaired angiogenesis and increased contractility associated with chronic fetal PH (see Fig. 3.3 ).
Factors That Disrupt Fetal Pulmonary Vascular Development
A number of gene mutations, including mutations in the gene coding bone morphogenic protein receptor type 2 (BMPR2) and other genes (e.g., CAV1, KCNK3, EIF2AK4), have been identified in adults with PH. In contrast, few genetic mutations have been identified in neonates with PH. Although eNOS mRNA expression was found to be reduced or absent in umbilical venous endothelial cells of infants with PPHN, candidate gene analysis did not identify any polymorphisms of the eNOS gene in infants with PPHN. In that same study, no variants for BMPR2, VEGF, cGMP-specific phosphodiesterase, or other plausible causes of fetal vascular remodeling were found in infants with PPHN. Interestingly, higher rates of genetic variants for cortisol signaling (corticotropin-releasing hormone receptor-1 [CRHR1] and CRH-binding protein) were observed in neonates with PPHN, as well as evidence for functional adrenal insufficiency. The CRH-binding protein decreases bioavailability of CRH and may diminish the activity of the hypothalamic-pituitary-adrenal axis and affect fetal lung functional development or the capacity to adequately transition to ex utero life. Additionally, CRHR1 single-nucleotide polymorphisms are located close to the transcription factor binding site for peroxisome proliferator-activated receptor-gamma (PPARγ), which is an essential regulator of pulmonary arterial smooth muscle cell proliferation and vascular tone.
Other genetic abnormalities of the NO pathway have been associated with PPHN. Endothelial cells generate NO from the precursor l -arginine, an amino acid supplied by the urea cycle. Carbamoyl-phosphatase synthetase catalyzes the first, rate-determining step of the urea cycle. In term neonates with respiratory failure, with and without PPHN, a polymorphism in the rate-limiting enzyme of the urea cycle, carbamoyl-phosphate synthetase-1, was associated with PH, low plasma arginine concentrations, and low plasma NO metabolites.
Children with Down syndrome (trisomy 21) commonly develop PH in association with structural heart defects, but they also have a 10-fold increased risk for idiopathic PPHN. In a Dutch cohort, PPHN was documented in 5.2% of infants with Down syndrome without cardiac disease. In addition, infants with Down syndrome have worse pulmonary artery hypertension in conjunction with anatomic cardiac disease than genetically normal infants with similar lesions and they are more likely to require extracorporeal membrane oxygenation (ECMO) support for PPHN. One recent study showed that 85% of autopsy specimens from children with Down syndrome displayed pulmonary vascular remodeling, suggesting that PH may occur even more commonly than clinically recognized.
Chromosome 21 includes at least three genes with potent antiangiogenic properties that could affect fetal vascular development. One of these genes, endostatin, is a known antiangiogenic factor that downregulates signaling of VEGF, so it is plausible that endostatin upregulation could impair angiogenesis and adversely affect lung structure. A threefold increase of endostatin mRNA expression has been reported in prenatal Down syndrome lungs, along with reduced microvascular density, thickened large and small pulmonary artery walls, and a persistent double capillary network. These results strongly suggest a role for genetically driven antiangiogenic signals in the pathogenesis of impaired lung vascular development and PH in Down syndrome.
Antenatal Ductal Closure
A patent ductus arteriosus is critical for the normal fetal circulation. It directs right ventricular output to the aorta, but it also protects the pulmonary circulation from volume overload and the right ventricle from pressure overload. Either partial or complete ductal ligation in the fetal lamb increases pulmonary arterial pressure without sustained elevation in pulmonary blood flow or in utero hypoxemia. Endothelial dysfunction rapidly emerges and results in poor response to endothelium-dependent vasodilators such as oxygen and acetylcholine, along with decreased expression and activity of pulmonary eNOS. There is also a strong constrictor “myogenic response” that may exist to protect the pulmonary capillary bed from high pulmonary blood flow, and may by mediated by activation of the ROCK pathway. Downstream signaling abnormalities emerge within days of ductal closure, including decreased activity and expression of soluble guanylate cyclase and increased activity of PDE5. After birth the newborns develop severe PH, a model that has been extensively used for preclinical studies to evaluate inhaled NO and other pulmonary vasodilators in PPHN.
Recent epidemiology studies have reported an association between fetal growth restriction and the later development of PH in premature infants with BPD. This association suggests that intrauterine stress could initiate the cascade that results in abnormal pulmonary vascular development and PH in preterm infants. In fetuses with severe fetal growth restriction and absent end-diastolic flow in the umbilical artery, PH with right-to-left shunting across the patent foramen ovale and patent ductus arteriosus was commonly observed.
Animal models have suggested that pulmonary vascular maldevelopment and PH associated with chronic lung disease begin before birth in response to chronic fetal hypoxia. Recent studies examining the placental and cord blood findings in preterm infants in whom PH later developed showed a striking association with placental pathologic changes of maternal vascular underperfusion and decreased villous vascularity. Moreover, cord blood angiogenic factors such as placental growth factor and VEGF-A were decreased in premature infants exposed to placental underperfusion, and these fetal blood markers predicted the subsequent development of PH. It is possible that disruption of placentation and placental vascular perfusion represents failed angiogenesis, as reflected by these biomarkers, and that this abnormal developmental angiogenesis may reflect global abnormalities of vascular signaling that contribute to postnatal vascular disease. Additional preclinical studies suggest that disruption of angiogenesis because of adverse antenatal factors such as chorioamnionitis, preeclampsia, or maternal smoking, can cause pulmonary vascular disease that not only leads to PH but also impairs lung growth and alveolarization.
Oligohydramnios/PH in Preterm Infants
Some preterm infants develop severe early PH after prolonged rupture of membranes with oligohydramnios and some degree of pulmonary hypoplasia. Preterm infants with prolonged rupture of membranes and PH were found to have low tracheal aspirate levels of nitrates/nitrites, suggesting a specific deficiency of NO; these infants responded promptly and dramatically to iNO. In animal studies, the pulmonary circulation of lambs with hypoplastic lungs created by a tracheo-amniotic shunt had significantly increased PVR with high pulmonary artery pressure and reduced pulmonary blood flow. Furthermore, changes in indices of lung ventilation were proportional to the changes in lung size, but accompanied by disproportionate changes in the pulmonary circulation associated with reduced density of pulmonary arterioles. These findings suggest that oligohydramnios-induced pulmonary hypoplasia exerts a selective effect on lung vascular development, and raises important questions about the role of the amniotic fluid in maintaining levels of lung vascular growth factors. Histologic changes of the pulmonary vasculature include reduced volume density of pulmonary arteries and increased acinar arterial wall muscle thickness.
Maternal Drug Exposures
Two classes of medications, nonsteroidal antiinflammatory agents (NSAIDs) and SSRIs, have the most evidence to suggest a direct effect on pulmonary vascular development. An association between PPHN and prostaglandin synthase inhibitor use during late gestation was first reported 40 years ago. Experimental closure or constriction of the ductus arteriosus in fetal lambs produces rapid development of pulmonary vascular remodeling and severe PPHN. Prostaglandins maintain ductal patency in utero and are important mediators of pulmonary vasodilation in response to ventilation at birth. Analysis of meconium from newborn infants with PPHN revealed the presence of NSAIDs in approximately half of the samples, linking antenatal NSAID exposure to PPHN. However, a recent epidemiologic study suggests these relationships are more complex than initially appreciated: an association was found only for aspirin use during the third trimester and PPHN, with no effect of ibuprofen at any point in gestation.
Exposure of pregnant rats to fluoxetine, an SSRI, produces PH, hypoxia, and increased mortality in the pups. In human epidemiology studies, the use of SSRIs during the last half of pregnancy has been associated with an increase in the incidence of PPHN, although recent reports indicate the risk is modest when controlling for maternal depression. In addition, the severity of PPHN has not been well described, and one recent report found that neonatal mortality rates were not higher in SSRI-exposed versus nonexposed neonates with PPHN (3.4% vs. 8.3%, P not significant). The mechanism by which SSRIs induce PPHN remains poorly understood, although a recent study shows that SSRIs induce concentration-dependent constriction of the ductus arteriosus and reduce sensitivity to prostaglandin-induced dilation, and that SSRI-exposed mice exhibit inappropriate ductus arteriosus constriction in utero. These findings are in agreement with a recent report from a large Swedish registry, which suggested that SSRI-exposed infants are more likely to have idiopathic PPHN without lung disease.
Maternal medications also have the potential to reverse pathologic fetal vascular development. In fetal lambs with PPHN induced by ductal ligation, maternal betamethasone reduced oxidative stress and improved the relaxation response to NO donors. Although not specifically examined, lower rates of PPHN could partially explain the benefit of antenatal steroids in late preterm infants. In fetal rats with nitrofen-induced CDH, antenatal administration of sildenafil to the dam improved lung morphology, and reduced pulmonary vascular remodeling and right ventricular hypertrophy. Maternal administration of sildenafil is under investigation for severe fetal growth restriction.