Authors: Srinivasan Mani (1Clinical Assistant Professor, Section of Neonatology, Department of Pediatrics, The University of Toledo/ProMedica Russell J. Ebeid Children’s Hospital, Toledo, OH 43606.), Hussnain Mirza (2Associate Professor, Section of Neonatology, Department of Pediatrics, Advent Health for Children/UCF College of Medicine. Orlando FL, USA. 32408.), James Ziegler (3Pediatric Cardiologist, Division of Cardiovascular Diseases, Department of Pediatrics, Hasbro Children’s Hospital/Brown University, Providence Rhode Island 02903, USA), Praveen Chandrasekharan (4Associate Professor of Pediatrics, Division of Neonatology, Department of Pediatrics, Jacobs School of Medicine & Biomedical Sciences. State University of New York at Buffalo. NY, USA. 32408.)
Categories: Article, Pulmonary Hypertension, Chronic lung disease, Preterm infants
Source: Clinics in perinatology
Authors: Srinivasan Mani, Hussnain Mirza, James Ziegler, Praveen Chandrasekharan
Pulmonary hypertension (PH) in preterm neonates has multifactorial pathogenesis with unique characteristics. Premature surfactant-deficient lungs are injured following exposure to positive pressure ventilation and high oxygen concentrations resulting in variable phenotypes of PH. The prevalence of early PH is variable and reported to be between 8 – 55% of extremely preterm infants. Disruption of the lung development and vascular signaling pathway could lead to abnormal pulmonary vascular transition. The management of early PH and the off-label use of selective pulmonary vasodilators continues to be controversial. Preterm infants present with critical hypoxemia and with pre and post-ductal saturation differences in the presence of a patent ductus arteriosus with right-to-left or bidirectional shunting. The current management strategies for early PH include lung recruitment, optimizing blood gas parameters with mechanical ventilator management, oxygen supplementation, maintaining systemic blood pressures, and pulmonary vasodilator therapy. PH in preterm infants is associated with high mortality and morbidity. Timely diagnosis and physiology-based management strategy may improve the clinical outcomes of extremely preterm infants with PH. A consensus is needed to define different phenotypes of PH in preterm infants that will provide an opportunity to test different treatment strategies suitable for each phenotype. With the advancement in ‘omics’ in the future, we could identify the variable phenotypes of preterm pulmonary hypertension and cater to precise treatment.
Pulmonary hypertension (PH) in preterm infants has unique characteristics because of the interconnections between alveolar and vascular immaturity.^1^ The Panama classification of pediatric PH highlighted the interplay of pathological insults on the growing lung, chromosomal or genetic syndromes, and developmental lung abnormalities in the pathogenesis.^2^ The National Heart, lung, and blood institute (NHBLI) workshop in 2012 identified the lack of phenotyping as the significant barrier in research to advance our understanding of pediatric PH.^3^ The ensuing research endeavors have advanced our understanding of PH phenotypes in preterm infants. The NHLBI workshop was reconvened in 2017 to enhance insights into pulmonary vascular disease (PVD) through a precision medicine approach and to develop interventions based on novel phenotypes.^4^ Recently, different phenotypes of early PH in preterm infants have received increased attention.^5,6^
Echocardiographic evidence of PH without hypoxic respiratory failure (HRF) in preterm infants < 72 h of life is considered a physiologic variant,^7^ whereas echo evidence of PH after 72 h of life is termed delayed pulmonary vascular transition (PVT). Persistent pulmonary hypertension of the newborns (PPHN), the most common phenotype of PVD in term infants, also occurs in preterm infants.^8^ It usually presents before 72 h of life with HRF. These infants are at high risk of developing myocardial ischemia and cardiac dysfunction due to elevated pulmonary vascular resistance (PVR) and diminished pulmonary blood flow (PBF).
Early PH in preterm infants can be defined as echocardiographic evidence of elevated pulmonary artery pressure (PAP) with or without HRF in the early days of life. Early PH can be differentiated from delayed PVT if a preceding echocardiogram documented a normal PVT. Early PH is associated with BPD or death,^5,9,10^ late PH,^11^ poor growth and neurodevelopment.^12^
Per Ohm’s law, PAP is based on the PVR, PBF, and pulmonary capillary wedge pressure [PAP= (PVR X PBF) + PCWP]. High PBF due to a significant left to right shunt, e.g., PDA, can increase PAP even with normal PVR.^13,14^ Higher left atrial pressure (LAP) due to mitral stenosis or left ventricular (LV) dysfunction may also significantly contribute to PH due to high PCWP.^15^ Both of these variants can also present as early PH.
This narrative review aims to summarize the current evidence on pathophysiology, diagnosis and management of early PH and other close phenotypes (Box 1) of PVD in preterm infants, identify knowledge gaps, and suggest future directions for research.
The transition of pulmonary and systemic circulatory circuits from functionally parallel to series is a critical event after birth triggered by removal of the placenta and transition of the gas exchange to the lungs. This transition alters the pulmonary circulation from high PVR and low PBF in the fetus to low PVR and increased PBF. This transition that starts at birth is not completed until the factors that aid in the typical fall in PVR act together as programmed.
A study that evaluated changes in the human fetal weight-indexed pulmonary and systemic vascular resistances (RPi and RSi) using doppler ultrasound found that between gestational age 20 and 30 weeks, there was a 1.5-fold decrease in the RPi.^23^ The fall in the PVR observed from the canalicular to the saccular stage of lung development was probably due to the increased density of the lung vasculature associated with lung growth. RPi-to-RSi ratio decreased significantly during the same period but did not change significantly from 30 to 38 weeks. This correlates with an increase in PVR from 30 weeks to term seen in lamb studies due to increased sensitivity of the pulmonary vasculature to in-utero hypoxemia.^24^
Precisely regulated changes occur in the production of endogenous vasodilator and constrictor mediators at different gestational ages.^25^ Endothelin −1 (ET-1) produced by endothelin converting enzyme-1 (ECE-1) expressing endothelial cells is the predominant vasoconstrictor in the fetus which maintains high basal PVR.^26^ Lamb studies have shown that expression of ECE-1 falls just before birth leading to reduced ET-1 mediated pulmonary vasoconstriction.^27^ A human study investigating the expression of ET-1 and its receptors in fetal pulmonary necropsy samples found that ET-1 and the endothelin receptor-A which induces vasoconstriction, were strongly expressed and stable at various gestational ages and infancy. However, the expression of the endothelin receptor-B, which induces vasodilation, was weak in early gestation, then increased markedly in mid-gestation and remained high for the rest of gestation.^28^
Extraluminal pressure exerted by the lung fluid in the fetus contributes to the increase in PVR by mechanical compression. Lamb studies done to increase the intraluminal pressure by tracheal obstruction (TO) showed a marked reduction in pulmonary blood flow (PBF) at seven days of TO. Pressure equalization by draining the lung fluid increased PBF.^29^ Studies have shown that phasic changes in the intraluminal pressure caused by accentuated fetal breathing movement decrease the PVR and increase PBF compared to the preceding apneic episodes.^30^ So, the egress of the lung fluid and the establishment of rhythmic breathing movements play a role in the drop in PVR after birth.
Inherent nitric oxide from endothelial cells, endothelial nitric oxide synthase (eNOS) and inducible nitric oxide synthase (iNOS) are vital in pulmonary vasodilatation and fall in PVR at birth. The rhythmic distension of the airway epithelium and increase in oxygen tension at birth stimulates the production of iNOS.^31^ Fetal lamb studies have shown that mechanical ventilation with 100% oxygen induces pulmonary vasodilation at birth by modulating the non-ATP-sensitive potassium channels.^32^ In premature animal models, endogenous nitric oxide (eNO) only partially modulates the pulmonary vasodilatation. However, inhaled nitric oxide (iNO) was considered a potent vasodilator that increases the pulmonary blood flow more than 100% oxygen.^33^ Thus multiple factors work in a coordinated fashion to bring about a normal decrease in PVR initiated at birth.
Sixty percent of preterm births are associated with chorioamnionitis.^34^ Studies have demonstrated that chorioamnionitis and funisitis lead to proinflammatory activation of the placental endothelium resulting in an altered angiogenic potential in the lungs.^35,36^ An immunohistochemical study of the human preterm placentas showed that heat shock protein 70, a marker for cellular stress, is overexpressed in chorioamnionitis, followed by an increased gene transcription of nuclear factorkB leading to activation of inflammatory cytokines. Further, proangiogenic factors like VEGF, VEGF R-1, and R-2 were decreased in the placental endothelium.^35^ These changes may explain the increased association between chorioamnionitis and early PH in preterm infants.
Postnatal stressors like mechanical ventilation, hyperoxia, and poor nutrition synergize the prenatal inflammation initiated by intra-uterine infection (figure 1). Preterm lambs exposed to intraamniotic Escherichia coli endotoxin antenatally and mechanically ventilated after delivery showed increased mononuclear infiltrate and elevated IL6 and IL 8 gene expression in lungs compared to lambs exposed to saline antenatally.^37^ Recently, a retrospective clinical study found that in addition to histologic chorioamnionitis, lower birth weight (fetal growth restriction - FGR) and invasive mechanical ventilation for ≥ 14 days are other significant factors associated with PH in the very preterm population.^38^
Preterm premature rupture of membranes (PPROM) in the mid-trimester leading to prolonged oligohydramnios, a risk factor associated with preterm delivery can induce pulmonary hypoplasia. In a study of experimental oligohydramnios in rats, authors demonstrated that oligohydramnios induced a decrease in platelet-derived growth factors (PDGF) A and B and elastin synthesis.^39^ In a similar lab experiment, oligohydramnios caused a decrease in transforming growth factor (TGF) β1 and collagen synthesis.^40^ The defective synthesis of elastin and collagen affecting alveolar development has been proposed as the mechanism for pulmonary hypoplasia.^41^ In a mouse model, oligohydramnios affected the differentiation of type 1 lung epithelial cells leading to defective angiogenesis.^42^ Inhibition of angiogenesis leads to decreased alveolarization and hypoplastic lungs.^43^
In several clinical studies, oligohydramnios was observed to be a risk factor for both early PH and late PH associated with BPD.^11,44^ A recent meta-analysis to identify the risk factors associated with PH in preterm infants diagnosed within two weeks of life by clinical and echocardiographic criteria found that oligohydramnios and small for gestational age (SGA) are strong predictors of early PH in preterm infants (figure 1).^45^
Fetal growth restriction or SGA is associated with approximately 20% of preterm deliveries. SGA status at birth is associated with early and late PH in preterm infants. In a sheep model of experimental placental insufficiency, authors demonstrated that FGR decreased pulmonary vessel density, affected vascular development and caused pulmonary artery endothelial cell (PAEC) dysfunction.^46^ Observational clinical studies support these findings.^45^
The fetal lungs do not participate in gas exchange. Fetal pulmonary vasoconstriction and the resulting high PVR shunt the blood away from the lungs. High PVR in utero is maintained by several factors such as compression of small pulmonary arteries by fluid-filled alveoli, lack of rhythmic lung distension, relative hypoxemia, presence of thick muscular layer in the small pulmonary arteries in the fetus, and vasoconstrictor mediators like endothelin-1 and thromboxane counterbalancing the vasodilatory mediators like prostacyclin and endogenous nitric oxide to maintain a prominent constrictor tone. Pulmonary blood flow is approximately 13% of combined ventricular output at 20 weeks gestation and increases to 25–30% at 30 weeks due to pulmonary vascular growth, and decreases again to 21% at 38 weeks due to developing pulmonary vascular reactivity to low fetal PO2.^23,47^
The fetus is in a state of relative hypoxemia with a PaO2 of 25–28 mm Hg in the aorta compared to the maternal uterine artery (PaO2 90–100 mm Hg). However, the fetus maintains a constant blood oxygen content to maintain cerebral and tissue oxygen delivery adequate for brain and somatic growth. Reactive pulmonary vasculature and increasing fetal hemoglobin with the advancing pregnancy are essential factors that help the fetus regulate cerebral oxygen delivery.
Fetal placental gas exchange must switch to pulmonary respiration in the neonate after birth, an essential step for postnatal survival. Lung aeration is the chief trigger leading to decreased PVR and increased PBF that can maintain the left ventricular preload after the cord clamping (figure 2). The next step in the extra-uterine hemodynamic transition is flow reversal in the ductus arteriosus followed by physiological closure.
PVR drops secondary to lung aeration, oxygen-triggered vasodilation, rapid involution of medial smooth muscle, and thinning of small pulmonary arteries and loss of fetal lung liquid.^48^ Inspired air creates surface tension in the alveoli leading to alveolar recoil, which increases capillary/alveolar wall transmural pressure and subsequently the capillary diameter, leading to a decrease in PVR.^49^ The decrease in PVR and the associated increase in PBF are not spatially related to lung aeration,^50^ resulting in ventilation/perfusion (V/Q) mismatch in the atelectatic portions of the lung. This is a common scenario during the transition of preterm birth resulting in a more gradual decrease in PVR compared to term infants emphasizing the importance of providing end-expiratory pressure and supplemental oxygen to improve V/Q matching.^51^
In an apneic preterm infant, the expected increase in PBF does not occur. In this scenario, if the umbilical cord is clamped and cut, the left ventricular preload is reduced because of the markedly reduced pulmonary venous return.^49^ This predisposes apneic infants to a double hit of hypoxia and ischemia which could possibly be avoided by delayed cord clamping until lung aeration is achieved.^52^ However, high quality clinical evidence supporting such intact cord ventilation practice is lacking and is currently being addressed by several well-designed randomized controlled trials.^53,54^
Early PVD in preterm infants can present as different phenotypes such as PPHN, delayed PVT, and early PH (Box 1). PPHN is caused by aberration in normal neonatal cardio-pulmonary transition resulting in significantly higher PVR, decreased pulmonary blood flow, and evolving cardiac dysfunction. These infants present with HRF that may improve with cardiotropic support and specific pulmonary vasodilator treatment. PPHN, also referred to as acute PH of newborns, can lead to progressive PVD and chronic PH.
Delayed PVT is diagnosed when preterm infants have echocardiographic evidence of elevated PAP without HRF. There is a randomized clinical trial (NCT#03576885) actively enrolling preterm infants to determine if early treatment with iNO can improve the clinical outcomes of extremely preterm infants with delayed PVT. Infants with early PH may have echo evidence of increased PVR that can present with or without HRF. At present, there is no evidence of improved long-term clinical outcomes if these infants are treated with specific pulmonary vasodilators such as iNO, but iNO is commonly used for severe HRF as there are few alternate approaches.
Preterm infants with delayed PVT or early PH may have echo evidence of increased PBF secondary to a significant left to right post tricuspid shunt like PDA (Flow related PH). These infants may require cardiotropic support to maintain systemic blood pressure, optimized ventilation and supplemental oxygen.
During hospitalization, preterm infants may develop intermittent or transient PH secondary to underlying etiologies like sepsis, pneumonia, necrotizing enterocolitis, sub-optimal ventilation, or inadequate recruitment of the lungs (atelectasis or hyperinflation). These infants may respond to the pulmonary vasodilators like iNO in the acute phase of illness along with corrective measures to treat underlying etiology.
The prevalence of early PH in preterm infants is variable, ranging between 8%^55^ to 55%^5^ based on gestational age or birth weight cut-off chosen for prematurity and timing of echocardiography. A case-control study of infants < 34 weeks reported a prevalence of 3.3%. A prospective study of preterm infants (n=277) with birth weight 500 – 1200 grams reported a 42% incidence of early PH based on the echocardiographic assessment.^11^ A retrospective single-center cohort study from the Netherlands reported a 26% prevalence of early PH in infants born < 30 weeks or with birth weight < 1000 grams.^56^ A recent prospective cohort study from the same center with the same inclusion criteria reported a 55% prevalence of early PH.^5^ The vast difference in the reported prevalence of the early PH could be explained by the difference in the timing of the screening echocardiogram. In the absence of any standard guidelines, echocardiographic screening for early PH was performed at different timepoints ranging between 3 days^5^ to 6 weeks^9^ of life.
The prenatal risk factors (figure 1) associated with neonatal PH include maternal diabetes,^57^ absence of use of maternal antibiotics,^57^ higher placental weight,^58^ oligohydramnios,^44^ PPROM ≥ 28 days,^44^ and PPROM at < 26 weeks.^44^ The postnatal risk factors associated with neonatal PH are male gender,^44^ lower gestational age,^58^ small for gestational age,^45^ and lower APGAR scores.^58^
Genetics plays a role in the development of pediatric PH. More than one-third of the cases of pediatric PH are associated with known genes with a significant contribution from de novo variants.^59^ Data specific to the neonatal PH is limited. A family-based candidate gene study examined 32 single nucleotide polymorphisms (SNPs) in 12 genes and found that PPHN was associated with genetic variants in corticotropin-releasing hormone receptor 1 and CRH-binding protein.^60^ A Chinese cohort study including 41 preterm and 74 late preterm and term infants found 3 SNPs (rs192759073, rs1047883, and rs2229589) in carbamoyl-phosphate synthetase1 and 1 SNP (rs1044008) in neurogenic locus notch homolog protein 3 (NOTCH3) gene were associated with PPHN.^61^ A similar study with 112 late preterm and term infants found rs2070699 SNP in the endothelin 1 gene was associated with PPHN.^62^
Hypoxemic respiratory failure (HRF) in neonates is an umbrella term comprising various disease processes leading to hypoxemia and the need for mechanical ventilation. Most studies use a threshold of ≥ 60% supplemental oxygen concentration to define hypoxemic respiratory failure.^63^ Compared to term infants, preterm babies are at higher risk of death due to HRF.^64^ Severe respiratory distress syndrome, congenital pneumonia, sepsis and PPHN are the most common etiologies for HRF in preterm infants.
Acute lung injury, often a precedent for acute PH in newborns, causes the following pulmonary vascular changes (figure 3) - endothelial dysfunction, pulmonary vascular occlusion, increased vascular tone, extrinsic vessel occlusion, and vascular remodeling.^65^ Endothelial dysfunction is an important precursor for the disease progressing to chronic PH leading to oxidative stress, uncoupling of eNOS and increased endothelin production (figure 3). Acute lung injury propagates endothelial dysfunction through - endothelial cell-leukocyte interactions, endothelial cell-platelet-neutrophil interactions, and structural changes between endothelial cells.^66^ These pathophysiological interactions lead to intravascular coagulation and inflammation at the microvascular endothelial level. The ensuing chronic inflammation leads to smooth muscle hypertrophy and neomuscularization of previously non-muscular vessels within the pulmonary vascular bed and adventitial thickening.^67^
The clinical features observed in preterm infants with early PH are non-specific (figure 4). Preterm infants with early PH may be asymptomatic or present with HRF. They require prolonged ventilator days with a median of approximately four weeks.^55^ Infants with early PH can present with respiratory distress out of proportion to the severity of the parenchymal lung disease as assessed by the radiological parameters and ventilator requirements.
A difference in the pre-and post-ductal oxygen saturation can be seen by pulse-oximetry. Differential oxygen saturation indicates right to left shunt across the patent ductus arteriosus. Sometimes the oxygen saturation can be low in all four limbs without a pre/post ductal difference. This condition can be seen if the extracardiac shunt is insignificant compared to the intracardiac right-to-left shunt across the PFO. This variant of early PH has been reported as early PH without ductal shunt.
Natriuretic peptide levels rise due to the atrial strain (atrial natriuretic peptide or ANP) or ventricular strain (B-type natriuretic peptide or BNP). In adults, these markers correspond well with cardiac function and can be used to identify cardiac disease. However, reference ranges for these cardiac biomarkers are not well established in cardiac conditions affecting the extremely preterm infants.^68,69^ Some studies show the significance of these biomarkers to determine the cardiac dysfunction^70^ or the significance of a PDA.^71^ NT-proBNP and BNP may have a role in screening and monitoring of infant with late PH associated with BPD^72^ but no correlation was identified between early PH and these biomarkers.^73^
A single-center prospective observational study studied the biological samples of infants born at a GA of 23w0d to 28w6d. The urine samples were analyzed for nitric oxide metabolites and plasma samples for markers of oxidant stress. Plasma amino acid profile was quantified. The study found that the high urine nitric oxide metabolite at day of life (DOL) 3 was associated with an increased risk of PH at 36 weeks. The high plasma citrulline level at DOL 7 was associated with a lower risk of PH.^74^
A multicenter proof of concept study was done in the US using a discovery and a validation cohort aimed to quantify and compare levels of potential biomarkers in neonates with bronchopulmonary dysplasia (BPD), BPD-PH, PH without BPD, and neonates without lung disease at a postmenstrual age (PMA) of 36 weeks. Multiple cytokines, angiopoietin (ANG) 1 and 2, and intracellular chloride channels 1 and 4 proteins were measured using ELISA at 36 weeks PMA. The study found higher levels of ANG 2 and lower levels of ANG 1 associated with infants with BPD alone compared to infants with BPD-PH and no lung disease. The study observed decreased monocyte chemoattractant protein-1 (MCP-1) and increased IL-1β levels in infants with BPD-PH. Similarly, the BPD-PH group had increased levels of IL-6, IL-8, IL-10, and TNF- α compared to infants with no lung disease.^75^ At present, no reliable biomarker of PH in preterm infants is available for clinical application.
Echocardiogram is not the gold standard for the precise evaluation of PH. However, for preterm infants, it is the only feasible test especially in the early days of life. However, it is important to understand the limitations of using echocardiograms for evaluation of PH.^76^ The objectives of performing an echo in a suspected case of PH are 1. Screen for structural malformations 2. Estimate pulmonary artery pressure 3. Assess the cardiac function, estimate cardiac output and blood flow across any shunt like PDA.
After excluding structural heart defects, PAP can be estimated using several echo measures including the tricuspid regurgitation (TR), inter-ventricular septal position, eccentricity index (EI), LV:RV diameter ratio, RV acceleration or ejection time (RVET), pulmonary artery acceleration time (PAAT), pulmonary artery time interval ratios like PAAT:RVET and pressure gradient across a shunt like VSD or PDA etc.^77,78^ RV function is evaluated by measuring the RV size, RV output, tricuspid annular plane systolic excursions (TAPSE) or speckle tracking etc. Similarly, LV function and systemic blood flow can be estimated by measuring the left ventricular output. A detailed discussion on all these echo variables is beyond the scope of this manuscript but we will discuss the most commonly used variables like TR, septal position and pressure gradient across a shunt (figure 4).
The tricuspid regurgitant (TR) jet measured using a doppler, directly estimates the right ventricular pressure (RVp). This is a validated parameter against cardiac catheterization. In the absence of TR jet, pressure gradient across any available post tricuspid shunt (VSD or PDA) can be used to estimate the PAP. A bidirectional or right to left shunt across a PFO/ASD may be due to the PH or secondary to other etiologies like tricuspid stenosis, RV dysfunction, RVOT obstruction or poor compliance of the right atrium. A qualitative assessment for pulmonary hypertension is possible based on the end-systolic septal position at the level of mitral valve chordae in the short axis view (parasternal or sub-costal). Septal flattening is considered subsystemic PH while reverse bowing into the LV is considered systemic or supra-systemic PH (figure 4). The reliability of septal position can be improved by calculating the eccentricity index (EI) that is a quantitative index of eccentric left ventricular shape.^79^ EI is a reliable and quantitative measure of the septal flattening that may reduce inter-observer variability in reporting septal position.^80^
Cardiac catheterization with acute pulmonary vasoreactivity testing is the gold standard for evaluation of PH. However, there is no role of cardiac catheterization in the screening or diagnosis of the early PH in extremely preterm infants due to the invasive nature of the test. Cardiac catheterization should be considered in evaluating PH associated with BPD, especially if there is an inadequate response to treatment or other comorbidities are complicating the clinical evaluation.^76^ If utilized, cardiac catheterization would be done under the baby’s baseline conditions to assess hemodynamics at rest, during increased cardiac output, following hypoxic exposure, and following vasodilator administration (oxygen and inhaled nitric oxide). It would allow gathering baseline data and assessing pulmonary vascular compliance and reactivity. It is important to note that cardiac catheterization is done in an artificial setting, usually with the patient intubated and under general anesthesia and/or high levels of sedation that makes it difficult to fully assess the dynamic pulmonary circulatory changes. Information obtained from cardiac catheterization should be interpreted with these limitations in mind.
Preterm infants with fetal growth restriction (FGR), exposure to perinatal infections and severe RDS are at higher risk for PPHN or acute PH (figure 1). A comprehensive evaluation starts with screening for any underlying etiology like neonatal infections, sub-optimal ventilation or lung recruitment. Optimal clinical management of PH varies based on the underlying etiology, echocardiographic evaluation of hemodynamics and the phenotype of the PVD. However, there are few general principles in the management of pulmonary hypertension including 1. Treatment of the underlying diseases such as RDS and neonatal infections 2. Ensuring optimal lung recruitment 3. Optimal ventilation and oxygenation to break the vicious cycle of hypoxemic pulmonary vasoconstriction and 4. Maintain systemic and pulmonary blood flow by supporting the cardiac output. Pulmonary vasodilators and other interventions can be offered based on the underlying pathophysiology of different phenotypes of the PVD.
Optimizing respiratory support to maintain adequate lung recruitment is very important in managing infants with early PH. Both hyperinflation and atelectasis are associated with increased PVR.^81,82^ We must achieve optimal lung inflation to avoid V/Q mismatch and intrapulmonary shunting. This is a physiological prerequisite for selective pulmonary vasodilators to be efficacious in improving pulmonary blood flow.^78^ Co-existing RDS should be treated with an exogenous surfactant. Appropriate respiratory support should be provided to maintain the blood gases in an acceptable range. There is no preference for any specific ventilation mode, but the goal of ventilation should be to keep the lungs optimally inflated (expanded up to 8 – 9 ribs on chest x ray), arterial carbon dioxide tension (PaCO2) in the 35 – 50 mmHg range and pH between 7.3 to 7.4. Correcting respiratory acidosis and maintaining pH in the normal reference range helps prevent pulmonary vasoconstriction due to hypercarbia or acidosis. Attention should be paid to the peak inspiratory pressures (PIP) needed to achieve these goals in conventional ventilation. Infants with reduced lung compliance or lung hypoplasia requiring unusually high PIP, should be considered for escalation to high-frequency ventilation that may be synergistic with the action of selective pulmonary vasodilators.
An arterial oxygen tension (PaO2) < 50 mm Hg has been shown to contribute to pulmonary vasoconstriction.^83^ Oxygen saturation targets 90 – 97% help to maintain preductal PaO2 in the 50–80 mm Hg range. In this oxygenation range, the benefit of oxygen-induced pulmonary vasodilation is maximal, with minimal risk of injury due to hyperoxia or the oxidative stress. Clinicians should be aware that PaO2 from the right radial arterial line reflects the preductal oxygenation while the umbilical arterial PaO2 indicates the post-ductal oxygenation. If the peripheral arterial line on the right upper extremity is not available, titrating supplemental oxygen based on the right upper extremity SpO2 is a reasonable alternative.
On conventional ventilation, minimum sedation may be required to facilitate adequate gas exchange. However, sedation is more likely required for infants on high frequency ventilation. Use of neuromuscular blockade by vecuronium is not uncommon among the term infants with severe PPHN and HRF.^84^ However, there is no data on the safety or efficacy of paralysis in preterm infants with acute PH and HRF and should be avoided with minor exceptions.^85^ Close attention to systemic blood pressure is important while providing sedation and neuromuscular blockade.
Systemic hypotension in early PH may be secondary to the elevated PVR, increasing the RV afterload and decreased PBF and, in turn, decreased LV preload. Sepsis, intraventricular hemorrhage, sedation, paralysis, use of high mean airway pressure (MAP), adrenal insufficiency and hypovolemia or dehydration can contribute to systemic hypotension. RV dysfunction could coexist and contribute to systemic hypotension.^86^ Uncorrected hypotensive shock worsens hypoxemia causing further pulmonary vasoconstriction, a vicious cycle leading to high mortality. Detecting RV dysfunction is important to guide us in choosing the inotrope.
Dopamine is the most commonly used inotrope for preterm infants with PH and systemic hypotension.^87^ In the dose range of 5 to 10 mcg/kg/min, the inotropic effect of dopamine is predominant due to beta-1 adrenergic receptor stimulation. However, at higher doses (> 10 mcg/kg/min), dopamine stimulates alpha-adrenergic receptors non-selectively, resulting in an increased ratio of pulmonary-to-systemic vascular resistance. Increasing SVR has been described as a strategy to decrease the right to left shunting across the PDA, but is detrimental for the cardiac function. RV afterload is increased due to the increased PVR and decreased right to left ductal shunting. This combination may worsen the myocardial ischemia and RV dysfunction leading to a further decrease in pulmonary blood flow and HRF. While LV preload decreases secondary to worsening PPHN, increased LV afterload secondary to increase in SVR potentiates the LV dysfunction, further decreasing the cardiac output. This vicious cycle can spiral down fast to reach a point where specific pulmonary vasodilators may not be effective to improve cardiac output. So, the use of dopamine especially in the higher doses should be avoided to treat preterm infants with acute PH.^88,89^
If there is no cardiac dysfunction, cardiac output may be low due to the decreased intrathoracic venous return caused by the use of high MAP. To improve intrathoracic venous return, weaning the MAP or PEEP can be helpful. Fluid bolus can also improve the diastolic filling and the cardiac output. In infants with systemic hypotension without evidence of RV dysfunction, vasopressin is an excellent first choice but may be associated with oliguria and hyponatremia.^90^
In the presence of RV dysfunction with systemic hypotension, epinephrine is the first choice due to its positive inotropic effects and the ability to increase the SVR without adversely affecting the ratio of pulmonary-to-systemic vascular resistance. Dobutamine and milrinone are pure inotropes without vasopressor effects, are potential choices in the presence of RV dysfunction without systemic hypotension, and can be used as an adjunct along with a vasopressor. Dobutamine has minimal effect on PVR.^91^ In the setting of septic shock coexisting with early PH, norepinephrine can be considered.^78^
Inhaled nitric oxide (iNO) is the first-line pulmonary vasodilator for treating acute PH in term and near-term infants after achieving adequate lung recruitment.^92^ However, randomized clinical trials that tested the efficacy of iNO therapy for the prevention of BPD in preterm infants did not enroll patients based on the echocardiographic diagnosis of PH.^93–95^ Indiscriminate iNO treatment of preterm infants was not beneficial to decrease the risk of BPD or death. A Cochrane review of 17 RCTs grouped into three categories – early rescue (<3 DOL), early routine use, and late treatment (>5–7 DOL) concluded that early routine use of iNO in preterm infants does not prevent serious brain injury or improve survival without BPD. The meta-analysis found that the late use of iNO to prevent BPD could be effective and further studies were recommended.^96^
The knowledge of early PH in the preterm population and the understanding of various phenotypes was limited during the period of conduct of most of these iNO trials. A single-center retrospective study including preterm infants with HRF treated with iNO (n=93) found that infants exposed to antenatal steroids and PPROM had improved oxygenation with iNO and better survival without IVH/PVL.^97^ A multicenter study of preterm infants (22–28 weeks gestation) with pulmonary hypoplasia found that early treatment with iNO did not improve survival.^94^ A meta-analysis found that African American infants had significantly reduced death or BPD with iNO 49% treated vs. 63% control.^98^ However, a recent neonatal research network cohort study (n=1732) reported no racial difference in the mortality or the neurodevelopmental outcomes among the preterm infants treated by iNO.^63^
The American Heart Association, American Thoracic Society, and the Pediatric Pulmonary Hypertension Network jointly recommend the use of iNO if PH is consistent with the PPHN pathophysiology.^99^ At present, it is unclear if iNO treatment for early PH or delayed PVT in extremely preterm infants can improve the clinical outcomes.^100^ However, given the safety profile of iNO, if echocardiography confirms PPHN phenotype, it may be prudent to conduct a trial of iNO to assess oxygenation response.^101^ Preterm infants with PPROM, oligohydramnios and suspected pulmonary hypoplasia anecdotally appear to demonstrate improvement in oxygenation.^102^
There is a rising trend to use systemic pulmonary vasodilators like milrinone, sildenafil or bosentan to treat PH in the newborns.^103^ However, limited data are available to determine the efficacy or safety of systemic pulmonary vasodilators in preterm infants specially to treat PPHN or early PH.^104–107^ In the settings of atelectasis or heterogenous parenchymal lung disease, systemic pulmonary vasodilators can increase the VQ mismatch by increasing the pulmonary blood flow in the atelectatic or diseased segments of the lung. Milrinone is a nonspecific vasodilator, it can increase pulmonary blood flow by decreasing the PVR but also increases the risk of systemic hypotension and should be used cautiously.^108^ Right to left shunting can worsen across a significant PDA, if systemic hypotension is relatively higher than the decrease in PVR. Use of the systemic pulmonary vasodilators in this high-risk population is limited until the safety and efficacy can be determined in clinical trials.^105^ In resource-limited settings, enteral sildenafil is the mainstay of PPHN management.^109^ In settings that lack access to iNO for the treatment of PH, sildenafil provides mortality benefit. A recent open-labelled clinical trial that compared oral and intravenous sildenafil in the treatment of mild to moderate neonatal pulmonary hypertension including infants > 34 weeks gestation found equal efficacy irrespective of the route of administration. However, the complications such as hypotension and poor cardiac contractility were common with the intravenous administration of sildenafil compared to oral route.^110^
Prostaglandins can decrease the PVR and protect the right ventricular function by ensuring the patency of the ductus arteriosus. However, there is limited evidence to support the prostaglandin infusion to treat preterm infants with PPHN or early PH. Nevertheless, under special circumstances when systemic blood flow is duct dependent due to severe PPHN and RV dysfunction can be benefitted by keeping the ductus open, judicious use of prostaglandin infusion shall be considered.^89^
Hydrocortisone is a well-studied steroid in the management of PPHN in term infants.^111^ Early use of dexamethasone is associated with improved clinical outcomes in term infants with meconium aspiration and PPHN.^112^ There is limited evidence to support the general use of dexamethasone to decrease the PVR in preterm infants^113^ and its use is associated with hypertrophic cardiomyopathy.
Preterm infants with catecholamine-resistant hypotension are usually treated with hydrocortisone that can improve systemic blood pressure without compromising cardiac function, systemic perfusion, cerebral or renal blood flow. Hydrocortisone increases the catecholamine potency by inhibiting its metabolism and improving microcirculation.^78,114,115^ It has been shown to downregulate phosphodiesterase 5 (PDE5) by blocking reactive oxygen species–induced NFκB activity in fetal pulmonary artery smooth muscle cells in a lamb model of PPHN.^116^
Diuretic treatment may be helpful to decrease the volume overload for preterm infants with echo evidence of flow associated PH secondary to a large left to right shunt across the ductus arteriosus or VSD. Shunt management shall be considered while restricting the total daily fluid intake. Pulmonary vasodilator therapy with iNO is contraindicated as it can further increase the pulmonary blood flow by decreasing the PVR. However, in some cases of severe hypoxemia in extremely preterm infants, a combination of different phenotypes of PH (flow-driven by PDA and resistance driven) may complicate management.
Early PH in preterm infants is associated with poor survival and a high risk for BPD or death.^56^ In a South Korean retrospective study including 247 extremely preterm infants, early PH detected with echocardiography at 4–7 DOL was an independent risk factor for death before 36 weeks PMA.^44^ The mortality rate in this study was 28.4% in the early PH group vs. 10.4% in the control group. The risk of BPD-associated PH was 18.9% in the early PH group compared to 5.2% in infants without early PH. In a subgroup of infants with moderate or severe BPD, early PH was an independent predictor of the occurrence of late PH.
These findings were confirmed by a subsequent meta-analysis including preterm infants with early PH (diagnosed in the first two weeks of life) was associated with an increased risk of moderate to severe BPD, in-hospital mortality and late PH.^45^ A prospective, observational study of preterm infants with birthweights between 500 and 1,250 grams (n=221), found that early PH diagnosed by echo at DOL 7 was associated with asthma, reactive airway disease, BPD exacerbation, bronchiolitis, pneumonia and recurrent hospitalization for a respiratory illness during the first two years corrected age.^117^ The study also found that the risk of late respiratory morbidities in these infants with early PVD increased eightfold if they had received mechanical ventilation at DOL 7 compared to controls. Chandrasekharan et al. have reported high rates of mortality or neurodevelopmental impairment among extremely preterm infants with early hypoxemia irrespective of presence of PH .^63^
The technological advances in transcriptomics, metabolomics, and proteomics can be utilized to classify the pathobiological forms of PH and predict response to vasodilator therapy.^118^ Such a multi-omics approach can provide research avenues with potential to enhance the understanding of neonatal PH and its phenotypes. The lung transcriptional framework for type I PAH constructed by the PH Breakthrough Initiative network study is a fine example of such an endeavor.^119^
A consensus-based classification of different phenotypes of PH in preterm infants will provide an opportunity to test class specific treatment strategies. (Box 1). A randomized clinical trial is needed to determine the efficacy of iNO for preterm infants with early PH and related phenotypes. Further studies are needed to assess safety and efficacy of systemic steroids and systemic pulmonary vasodilators to treat PH in preterm infants. Research on biomarkers to diagnose and classify early PH will be valuable.
Pulmonary hypertension in preterm infants is associated with high mortality and morbidity. Disease has a variable pathophysiology based on the different underlying phenotypes. Timely diagnosis and physiology-based management strategy may improve the clinical outcomes of the extremely preterm infants with pulmonary hypertension.