Pulmonary Vasodilator therapy in Persistent Pulmonary Hypertension of the Newborn (PPHN)

Introduction

Persistent pulmonary hypertension of the newborn (PPHN) is a clinical syndrome characterized by sustained elevation of pulmonary vascular resistance (PVR) after birth, resulting in extra-pulmonary right-to-left shunting and severe hypoxemia. PPHN affects approximately 2 infants per 1000 live births, and the highest incidence occurs in late preterm infants (34–36 weeks of gestational age) at 5.4 per 1000 live births compared to term infants at 1.6 per 1000 live births [1]. Inhaled nitric oxide (iNO), a potent pulmonary vasodilator, had a transformational impact on the management of infants with PPHN. However, up to 30% of PPHN infants fail to respond to iNO and face increased risk of complications, including the need for extracorporeal membrane oxygenation (ECMO) to survive [2]. Improved knowledge of pathobiology of PPHN combined with the critical need to improve outcomes in iNO non-responsive infants has led to the development and clinical application of alternative vasodilators. However, lack of high-quality evidence in this area presents a clinical dilemma and therapeutic challenge to the bedside clinician. This review focuses on optimizing the response to iNO and use of alternative pulmonary vasodilators in the management of PPHN. We will summarize the literature on the use of other pulmonary vasodilators, suggest a management approach based on our experience and outcomes, and discuss challenges involved in future randomized clinical trials in this field.

Background

PPHN is often a complication of many lung diseases and systemic disorders that lead to elevated pulmonary vascular resistance (PVR) [1, 2]. The fetal pulmonary circulation is characterized by elevated pulmonary vascular resistance and low pulmonary blood flow [3, 4]. At birth, pulmonary vascular resistance decreases exponentially with subsequent rise in lung blood flow to facilitate gas exchange during postnatal life [3, 4]. This transition at birth in response to oxygen and other birth-related stimuli involves structural reorganization of the pulmonary vessel walls, recruitment of intra-acinar arteries, and gradual vascular remodeling [4–6]. Elevated pulmonary vascular resistance in infants suffering from PPHN can be a consequence of a plethora of neonatal disorders that lead to vasoconstriction, medial hypertrophy of the pulmonary arteries (PAs), or decreased blood vessel density due to failure of angiogenesis. The conditions that lead to high PVR are listed in Box 1.

Box 1: Etiology of PPHN in neonates.

Maladaptation of pulmonary vasculature (abnormal, ‘reactive’ pulmonary vasoconstriction)

○ parenchymal lung diseases, such as meconium aspiration syndrome (MAS), respiratory distress syndrome (RDS) and pneumonia

○ in response to systemic disorders, such as hypothermia, sepsis, fetal hypoxia/distress, hypercapnia, acidosis, and hyper-viscosity

○ toxic/pharmacological exposure in utero (maternal SSRI use)

Maldevelopment of pulmonary vasculature (remodeling of pulmonary vasculature)

○ in utero closure of ductus arteriosus (maternal cyclooxygenase inhibitor use)

○ sustained pulmonary over-circulation in congenital heart disease with large left-to-right shunts

○ intra-uterine growth restriction

○ Genetic/chromosomal anomalies (Trisomy 21, alveolar-capillary dysplasia, surfactant protein deficiency)

Underdevelopment of pulmonary vasculature (hypoplastic pulmonary vessels; ↓cross-sectional area)

○ Congenital Diaphragmatic Hernia

○ Pulmonary hypoplasia (premature prolonged rupture of membranes, oligohydramnios and anhydramnios, space-occupying lesions in the chest).

Endogenous vascular signals

The endothelial cells (ECs) and smooth muscle cells (SMCs) in the pulmonary blood vessels together mediate the transitional changes in lung vasculature at birth. The endothelium plays a key role in generating vasodilators and limiting vasoconstrictors [6–8], thus mediating an immediate, local alteration of vascular tone [6] (Figure 1). NO, the key endogenous vasodilator is released by endothelial nitric oxide synthase (eNOS) [6] when stimulated by oxygen [7, 8]. NO diffuses to smooth muscle cells to stimulate soluble guanylyl cyclase (sGC) which produces cyclic guanosine monophosphate (cGMP) [4, 9] to induce vascular smooth muscle relaxation [10, 11] (Fig 1). Phosphodiesterase (PDE)-5 degrades cGMP [4], and is expressed at high levels in lung tissue, increasing as the fetus approaches term gestation [12]. The disruption in the NO-cGMP pathway in PPHN has been studied in animal models, and a decrease in eNOS expression and activity occurs following prenatal ligation of the ductus arteriosus [13–15]. This decreased expression and activity of eNOS is also seen in neonates suffering from PPHN [16]. Prostaglandins, particularly PGI₂ and prostaglandin E₂ (PGE₂), are important pulmonary vasodilators produced by EC. They induce SMC relaxation by increasing cyclic adenosine monophosphate (cAMP) levels in SMC. Tight control of cAMP levels is mediated by Phosphodiesterase (PDE)-3, which rapidly breaks down cAMP. Downregulation of PG and cAMP signaling contributes to increased PVR in the PPHN model [17]. Endothelin-1 (ET-1) is released by EC and induces SMC constriction and proliferation [18, 19] and promotes release of the vasoconstrictor, thromboxane [20]. ET-1 levels are elevated in PPHN infants [21]. Thus, the altered balance of vasoactive mediators with increased levels of vasoconstrictors and decreased vasodilators contributes to the pathogenesis of PPHN. Current therapies for PPHN aim to overcome these signaling alterations by increasing cAMP and cGMP levels or antagonizing ET-1 receptors, as summarized in Fig 2.

Figure 1.

Mechanisms of endothelium-dependent pulmonary vasodilation and vasoconstriction. Oxygen, lung distension, shear stress, ATP, & VEGF activate endothelial nitric oxide synthase (eNOS) and cyclooxygenase directly or indirectly. Release of NO and prostacyclin (PGI2) leads to activation of soluble guanylate cyclase (sGC) and adenylate cyclase (AC), respectively, in vascular smooth muscle cells (SMC) with generation of cGMP and cAMP, respectively. Activation of the corresponding protein kinases G and A (PKG and PKA) by cyclic nucleotides leads to decreased calcium influx and subsequent SMC relaxation. Phosphodiesterase (PDE)-5 and 3 breakdown cGMP and cAMP, respectively, to limit the duration of vasodilation. Two important vasoconstrictor pathways are conversion of PGH2 to thromboxane A2 (TxA2) by thromboxane synthase and synthesis and release of endothelin-1 (ET-1).

Figure 2.

Emerging targets and therapies for PPHN. L-citrulline mediated activation of NO and soluble guanylate cyclase (sGC) activators and phosphodiesterase (PDE) 5 inhibitors increase cGMP levels. Rho-kinase inhibitors and specific endothelin receptor – A antagonists (Sitaxsentan and Ambrisentan) reverse vasoconstriction. PPAR γ agonists and antioxidants potentially reverse remodeling of pulmonary arteries. These potential new therapies require further evaluation and clinical trials in PPHN. Reproduced from Lakshminrusimha S, Mathew B, Leach CL. Pharmacologic strategies in neonatal pulmonary hypertension other than nitric oxide. Semin Perinatol. 2016 Apr;40(3):160–73.

Bronchopulmonary Dysplasia (BPD)-Related Pulmonary Hypertension

Pulmonary hypertension affects 16–25% of infants with BPD and increases the risks of mortality and long-term complications [22, 23]. The mechanisms of pulmonary hypertension in BPD (BPD-PH) are unclear, and appear to differ from the key alteration of vasoconstrictor-vasodilator imbalance seen in PPHN. The major alteration in BPD-associated pulmonary hypertension is impaired vascular development due to exposure of the preterm, saccular stage lung to several postnatal exposures like hyperoxia, intermittent hypoxia, barotrauma, inflammation, hemodynamic stress from PDA and elevated pulmonary artery pressure [22]. These changes disrupt normal growth factor expression and signaling pathways, resulting in impaired gas diffusion, abnormal vascular remodeling, and simplification of pulmonary vascular tree from vascular growth arrest [22, 24]. Unfortunately, given the unclear pathogenesis of BPD-associated pulmonary hypertension and paucity of randomized clinical trials, there is a lack of evidence-based management strategies for BPD-PH.

APPROACH TO MANAGEMENT OF PPHN

General Approach

Since the clinical manifestation of PPHN is hypoxemic respiratory failure (HRF), the goal of overall management is to maintain adequate oxygen delivery to vital organs and tissues. The initial steps should focus on providing adequate oxygen to maintain appropriate saturations, adequate ventilation and lung recruitment, appropriate fluid resuscitation and hemodynamic support. Intubation and mechanical ventilation for alveolar recruitment and early surfactant administration are essential initial stabilization steps prior to considering specific pulmonary vasodilator therapies. Neonates in whom HRF persists with an OI of >15 (except CDH for reasons addressed in that section) despite establishing adequate ventilation and circulatory resuscitation are candidates for a trial of pulmonary vasodilator therapy.

Pulmonary Vasodilators used in the NICU

Oxygen and iNO are the widely used and selective pulmonary vasodilators for infants with PPHN and should be used as first-line therapies. There are several other adjunctive therapies at various stages of investigation that will be summarized in detail in the sections below.

OXYGEN

Although the use of oxygen to correct hypoxemia and minimize hypoxic pulmonary vasoconstriction are important clinical goals in managing neonates with PPHN, maintaining higher than normal blood oxygen tension/content does not lead to additional pulmonary vasodilation and may be potentially harmful [25]. Excess oxygen administration can worsen vasoconstriction by generation of reactive oxygen species in the pulmonary arteries which can render other vasodilators like iNO ineffective [25]. We recommend avoiding hyperoxia and targeting goal oxygen saturations of 92–97% and/or paO2 60–90. Hypoxemia in neonates is often secondary to VQ mismatch from parenchymal lung disease; lung recruitment strategies should be adopted in place of reliance primarily on supplemental oxygen.

INHALED NITRIC OXIDE

Inhaled Nitric Oxide is the only FDA approved pulmonary vasodilator widely accepted as standard of care in the management of PPHN in developed countries [26]. Several large randomized clinical trials have demonstrated that iNO therapy decreases the need for ECMO/risk of mortality in full term and late preterm (≥34 weeks’ gestation) infants with severe hypoxic respiratory failure and pulmonary hypertension [27–29]. Inhaled NO gas reaches alveolar space rapidly and diffuses to the vascular smooth muscle of the adjacent pulmonary artery causing relaxation by increasing the intracellular cGMP levels. Key facts and considerations are summarized below.

What makes iNO a selective and ideal pulmonary vasodilator?

Once NO diffuses from alveolar space into the lumen of pulmonary artery, it is rapidly bound and inactivated by Hb, limiting its effect to the pulmonary circulation. Since NO gas enters and dilates pulmonary vessels selectively in the ventilated segments of the lungs, it promotes ventilation/perfusion match and thereby improves oxygenation in neonates with parenchymal lung disease. The effect of iNO on pulmonary circulation is also not limited by the presence of extra-pulmonary right–left shunts, which often lead to hypotension with intravenous vasodilators.

What is the optimal timing for starting iNO?

We recommend initiation of iNO therapy when hypoxic respiratory failure progresses and OI reaches 15–20 on at least two blood gases, based on a clinical trial of iNO for infants in moderate degree of HRF [30]. Initiation at this OI decreases the progression of HRF and reduces the need for ECMO/incidence of death compared to initiation at OI>20 and decreases the length of stay [30].

What are the right doses and weaning strategy for iNO?

Previous RCTs have shown that the ideal starting dose for iNO is 20ppm; higher starting doses do not increase the response to iNO, while potentially increasing the incidence of methemoglobinemia and NO2 exposure [29]. It is important to wean high FiO2 levels prior to weaning iNO dose. Weaning of iNO as oxygenation improves is well tolerated with reductions in doses from 20 to 10, 5 and by 1 ppm decrements below 5 ppm, as shown in a previous study that demonstrated the safety of this approach [31].

Considerations prior to starting iNO

Prior to starting any pulmonary vasodilator, it is essential to ensure optimum lung recruitment via CPAP or mechanical ventilation and surfactant administration if there is evidence of parenchymal lung disease [30]. Adequate ventilation is necessary because inhaled NO gets preferentially distributed to the ventilated segments of the lung, resulting in increased perfusion of the ventilated segments, thereby optimizing the ventilation-perfusion matching. Since several cyanotic heart diseases can mimic PPHN, an echocardiogram should be performed to confirm the diagnosis and rule out cyanotic heart disease. Inhaled NO is contra-indicated in congenital heart diseases with ductal-dependent systemic blood flow and in Total Anomalous Pulmonary Venous Return (TAPVR) where dilation of pulmonary vessels in the presence of venous obstruction can worsen the pulmonary hypertension.

What is the role of iNO in Congenital Diaphragmatic Hernia (CDH)?

Despite widespread use of iNO in neonates with CDH during initial stabilization, randomized controlled trial data and large retrospective studies do not show evidence of improved survival and decreased need for ECMO [32, 33]. The questionable benefit and high cost of iNO therapy should prompt clinicians to use this therapy cautiously, ensuring careful patient selection with echocardiographic measures of pulmonary hypertension and normal left ventricular function [34]. In the presence of systolic dysfunction of left ventricle, the rapid rise in pulmonary venous return can overburden a dysfunctional left heart and worsen cardiopulmonary status. iNO treatment did not improve oxygenation in patients with LV systolic dysfunction and in this specific subgroup, was associated with greater ECMO rate in a single-center study [34]. We recommend that in instances where iNO is initiated based on echo parameters of PH or in acute hypoxemic crisis to stabilize for transport and/or ECMO, there should be ongoing evaluation of the response and need for continued iNO treatment.

Other pulmonary vasodilators

Pulmonary vasodilators acting via the cGMP pathway

SILDENAFIL (PDE-5 inhibitor)

Nitric oxide exerts its vasodilator effect primarily through soluble guanylate cyclase (sGC) and the second messenger cGMP. Oxidative stress in PPHN oxidizes sGC and decreases cGMP production, and stimulates PDE5 to enhance breakdown of cGMP, thus rendering iNO ineffective [35, 36]. Alternate agents have been investigated for neonates unresponsive to iNO. Among them, PDE-5 inhibitors such as sildenafil have been studied in a few RCTs. Sildenafil increases cGMP levels by preventing its breakdown by endogenous PDE-5, resulting in pulmonary vasodilation. Small, randomized trials of sildenafil performed in resource-constrained settings where iNO and ECMO were unavailable, reported improved oxygenation and decreased mortality with enteric administration of sildenafil [37–39]. A pilot randomized controlled trial of enteric sildenafil (1–2mg/kg every 6 hours) in Colombia in a setting where ECMO was unavailable showed that it improves oxygenation in neonates with severe PPHN compared to placebo-treated infants. Reported as a proof-of-concept study, this RCT was halted early after 5/6 infants in the placebo group died compared to 1/7 infants in the sildenafil group [37]. Improvement in oxygenation occurred in the sildenafil-treated infants 6–12 hours after the first dose. Systemic hypotension was not observed with enteric sildenafil. Similar results were noted with an RCT in Mexico [38]. This study reported that sildenafil decreased the mortality risk significantly from 40% in the placebo group to 6% in the sildenafil-treated neonates. These RCTs show that sildenafil, used as the primary therapy for PPHN, is effective and safe in resource-limited settings lacking access to iNO.

A small pilot RCT comparing enteric sildenafil + iNO to placebo + iNO did not see improvement in oxygenation in the sildenafil + iNO group [39]. An open label dose escalation trial of IV sildenafil in 36 neonates (29 receiving iNO) showed decreased OI starting 4 hours after the administration of the drug. A loading dose of sildenafil 0.4mg/kg over 3h (0.14mg/kg/h), followed by 0.07mg/kg/h (or approximately 1.6mg/kg/day) continuous infusion provided the intended therapeutic levels as well as clinical benefit [40]. A phase-3 randomized, placebo-controlled trial of IV sildenafil in PPHN investigating short- and long-term outcomes at 12–24 months in a group of PPHN neonates already receiving iNO therapy has completed the initial phase. The study found that IV sildenafil (0.1mg/kg over 30 mins followed by 0.03 mg/kg/hr i.e. lower doses than the above open label trial) as an additive therapy to iNO did not reduce the treatment failure rate of need for additional vasodilator/ECMO/death or duration of iNO therapy, compared to placebo [41]. Infants treated with IV sildenafil were more likely to experience hypotension. The results of phase B of this trial, the neurodevelopmental outcomes at 12–24 months of age are awaited at this time (NCT01720524). Based on the data available, sildenafil appears to be an effective alternative to iNO in resource constrained areas as primary therapy when iNO is not available. The benefit of addition of sildenafil to infants already on iNO therapy remains unclear. Based on current evidence and our experience, hypotension is more likely with IV than enteric sildenafil, though the latter may have unpredictable absorption.

Pulmonary vasodilators acting via the cAMP pathway

PROSTAGLANDINS

Prostacyclin (PGI2) is an arachidonic acid metabolite that stimulates adenyl cyclase in vascular smooth muscle cells causing an increase in intracellular cAMP and vasodilation in systemic and pulmonary circulations. There are 3 commercially available prostacyclin or its analog formulations.

(1). EPOPROSTENOL (Flolan)

Epoprostenol (Flolan): is a prostacyclin available in IV or aerosolized formulations. There are no RCTs to demonstrate its efficacy and safety profile in neonates with PPHN, and evidence is limited to retrospective reports. Because of its very short half-life (~6 minutes), it must be administered in a continuous IV or inhaled form. The available evidence is briefly described below.

Epoprostenol (FLOLAN) (Intravenous):

A prospective case series of 8 consecutive infants showed that IV prostacyclin (epoprostenol) improved ECHO parameters of PAP and oxygenation within a median of 87 hours of administration [42]. Hypotension, resulting from non-selective vasodilation from IV prostacyclins, can be managed with volume expanders and vasopressor medications. A retrospective review of prostacyclin use in critically ill infants with Pulmonary Hypertension showed acceptable safety and tolerability though the authors did not investigate its ability to modulate pulmonary hypertension severity [43]. There are three major limitations to IV prostacyclin: (1) Hypoxemia remains a concern in infants with parenchymal lung disease with intravenous administration of epoprostenol, due to continued or increased VQ mismatch from global rather than selective pulmonary vasodilation. However, based on available evidence in the neonatal population, this complication risk remains low. Additionally, given the short half-life of epoprostenol, if appropriate dosing and monitoring are followed, adverse effects can resolve quickly after discontinuation of the medication. (2) Parenteral administration can lead to systemic hypotension in the presence of right to left shunts across PFO or PDA. (3) The alkaline pH leads to drug compatibility issues, requiring a dedicated line (peripheral or preferably central), which can be challenging for an acutely ill neonate on multiple IV drips and limited access. In order to overcome these limitations of IV epoprostenol, administration via inhaled route is a preferred alternative.

Epoprostenol (FLOLAN) (Inhaled):

Aerosol administration is given through a nebulizer connected to the breathing circuit of both conventional and high frequency ventilators. The intravenous formulation of Flolan is dissolved in 20 ml of manufacturer’s diluent (a glycine buffer, pH 10). The effect of such alkaline pH on the neonatal respiratory tract is unknown. Using continuous nebulization at a dose of 50 ng/kg/min, diluted to a volume of 8ml/hr, Kelly et al reported improved oxygenation in 4 infants with PPHN unresponsive to iNO, although one neonate with alveolar capillary dysplasia subsequently deteriorated [44]. Several other case reports and pilot studies reported improved oxygenation status in neonates with aerosolized epoprostenol [45–47]. In a recent retrospective review of 43 critically ill neonates with refractory PPHN, use of inhaled epoprostenol administered via Aeroneb nebulizer was associated with improvement in OI and FiO2 after 12-hours of treatment [48]. Infants with CDH and meconium aspiration were the best responders. None of the infants experienced hypotension with inhaled PGI2 though most were on concomitant inotropes. However, a rebound effect was observed at the end of the continuous nebulization with an increase in OI which was not sustained at 4 hours after the end of treatment [48]. The limitation of these studies is that PPHN infants show labile oxygenation, and it is unclear whether the changes in PaO2 or OI occurred in response to treatment or as natural course of their underlying illness. Appropriate monitoring of systemic blood pressure and oxygen saturation should be in place to monitor the rebound effects of interrupted drug delivery.

(2). ILOPROST

Iloprost, a synthetic analog of prostacyclins, is similar to epoprostenol, but can be given as intermittent nebulizations six to nine times a day due to longer half-life (20–30 minutes) and greater duration of pulmonary vasodilator effect (1–2 hrs). A vibrating mesh nebulizer can be integrated into the inspiratory limb of ventilator circuit as proximal as possible to the endotracheal tube. Nine prospectively enrolled term neonates with PPHN received inhaled iloprost at 1–2μg/kg every 3–4 hours in a center lacking iNO availability [49]. Decreased FiO2 and improved oxygen saturation were noted in 8/9 patients with no adverse effects reported. A retrospective comparative study of 47 neonates with PPHN in a center without access to iNO, HFV and ECMO showed that iloprost (1–2.5μg/kg every 2–4 h) was more effective than oral sildenafil in the time to improvement in OI, duration of drug therapy, mechanical ventilation and inotropic support [50]. However, as with any nebulizer device, some uncertainty will exist regarding the amount of prostacyclin effectively delivered into the alveolar space. DiBlasi et al performed an in-vitro study using neonatal test lung model to measure the amount of drug delivered via modern vibrating mesh nebulizers [51]. They found satisfactory iloprost delivery both in conventional and high-frequency ventilation with greater drug delivery when the nebulizer was placed in a proximal position (between patient and circuit) than the distal position, and better delivery with HFOV than conventional ventilation. Evidence of iloprost use in critically ill preterm infants is limited to case reports [52–54] with hypotension noted with IV iloprost but not with inhaled administration.

Based on the available evidence and our experience, aerosolized iloprost or epoprostenol are preferred over IV forms to avoid systemic hypotension and to leverage the pulmonary vasodilator effects in the ventilated segments of the lung. Inhaled prostacyclins overcome the limitation of right to left extrapulmonary shunts (PFO, PDA) which interfere with the delivery of agents to pulmonary circulation. Inhaled route also avoids the need for a dedicated central line required for any IV prostacyclin due to alkaline pH.

(3). TREPROSTINIL

Treprostinil is a stable tricyclic analogue of prostacyclin, which compared to epoprostenol, has a longer half-life, and fewer side effects. It provides an option of a pump for continuous subcutaneous infusion, avoiding need for central line, as well as means of transitioning to home therapy. Evidence in neonates is limited to small case series or retrospective reviews [55], though a pilot Phase 2 RCT (NCT02261883) is currently underway. A recent case series provides guidance into transitioning from epoprostenol to Treprostinil [56]. Turbenson et al described 5 patients (three CDH, one 24-week GA preemie with late pulmonary hypertension, one PPHN treated with iNO and sildenafil) that were initiated on IV epoprostenol, escalated to target dose of epoprostenol and transitioned to IV Treprostinil, followed by switch to subcutaneous Treprostinil [56]. All patients survived to hospital discharge and were sent home on SC Treprostinil with minimal adverse effects. SC Treprostinil has also been used in CDH patients with chronic pulmonary hypertension, resulting in decreased severity of PPHN by ECHO and BNP measurements [57, 58].

(4). BERAPROST

Beraprost sodium is an oral prostacyclin formulation shown to improve pulmonary hypertension in adults [59, 60] and children with congenital heart disease [61, 62]. The only evidence in neonates is a case series of 7 infants with PPHN; beraprost has improved the OI but also decreased systemic blood pressure [63]. Further studies to evaluate the appropriate dose to minimize the risk of systemic hypotension are warranted.

(5). PROSTAGLANDIN E1 (ALPROSTADIL)

PGE1 is typically used in infants with ductal-dependent congenital heart disease. In a small phase I/II open-label clinical trial of 21 infants with PPHN, an inhaled formulation of prostaglandin has (alprostadil) improved oxygenation without any adverse events [64]. An RCT has attempted to investigate this further but was terminated due to low enrollment [65]. PGE1 can improve right ventricular function by maintaining the patency or reopening a closed ductus arteriosus to provide an outlet to decompress the strained right ventricle and assist systemic blood flow in the presence of LV dysfunction seen in CDH. However, these effects can be inconsistent and potentially increase right-to-left shunt with worsening hypoxemia. In the short term, it may help RV adaptation to high PVR while other agents work to lower pulmonary pressures.

PDE-3 INHIBITORS

MILRINONE

Milrinone inhibits PDE3 in vascular smooth muscle cells and cardiac myocytes, leading to increased cAMP expression. Increased cAMP in VSM causes vasorelaxation by improving calcium uptake into the sarcoplasmic reticulum while in the cardiac myocytes, improved calcium uptake leads to increased contractility. Milrinone uniquely induces pulmonary vasodilation, enhances systolic myocardial contractility (inotropy) without increasing myocardial oxygen demand, and promotes diastolic myocardial relaxation (lusitropy). The use of milrinone in refractory PPHN has been supported by several small studies, mostly case series [66–70]; however, evidence from RCTs is lacking. An open label study in 11 term infants using a bolus of 50 mcg/kg followed by an infusion of 0.33–0.99 mcg/kg per min has reported decreases in OI, lactate levels and base deficit [67]. A small RCT performed in a single center, resource-limited setting reported that oral sildenafil plus intravenous milrinone infusion led to a quicker and longer lasting effect on improving pulmonary artery systolic pressure than either of the agents used alone [71]. Milrinone use may be considered in iNO non-responders before considering ECMO, with close monitoring of cardiovascular response using clinical parameters and if available, functional echocardiography. The inotropic and lusitropic effects of milrinone may be particularly beneficial in neonates with CDH and left ventricular dysfunction. An RCT evaluating the role of Milrinone in improving oxygenation in CDH infants is ongoing (NCT02951130).

Pulmonary vasodilators acting via the Endothelin pathway

ENDOTHELIN ANTAGONISTS

BOSENTAN

Bosentan is an Endothelin-1 antagonist of both endothelin A and B receptors. A small, single-center, randomized study by Mohamed et al in a center lacking iNO and ECMO showed improved oxygenation in the bosentan, compared to placebo group (80% vs 20%) [72]. These data and some other reports suggest that it may have a role in the management of PPHN in resource-constrained settings without access to iNO [73–76]. A multicenter, randomized, double-blind, placebo-controlled trial (FUTURE-4 study) assessed the effects of bosentan (2mg/kg twice daily via nasogastric tube) as adjuvant therapy in 21 neonates with respiratory failure already receiving iNO [77]. Owing to slow recruitment, the trial was terminated early. In contrast to the observations by Mohamed et al, FUTURE-4 trial showed no additional benefit when given as an adjuvant therapy in terms of time to weaning off iNO, oxygenation status or need for ECMO. Enteral absorption was slow and unpredictable and took five days to reach steady state. Though no safety concerns were reported in the above trials, elevation of transaminases has been reported with bosentan use in adult patients with pulmonary hypertension. Anemia and peripheral edema were reported in the FUTURE-4 study. Given the limited evidence of clinical benefit and unreliable pharmacokinetic profile with oral administration, we recommend the cGMP and cAMP targeted therapies as first or second line options for PPHN.

STEROIDS

Glucocorticoids are frequently used in critically ill PPHN neonates due to their potent anti-inflammatory and vasopressor effects, reducing right to left shunting. Glucocorticoids increase cGMP levels by normalizing sGC and PDE5 activity and decrease oxidant stress in lambs with PPHN [78–81]. A small randomized 3-arm trial evaluating short courses of IV methylprednisolone, inhaled budesonide and placebo in meconium-aspiration suggested improvements in the duration of oxygen therapy, radiological clearance of lungs and length of stay [82]. The risk benefit profile should always be considered due to previously reported concerns for neurodevelopment in premature infants when steroids are used early in life.

A PHYSIOLOGY-BASED APPROACH TO PULMONARY VASODILATORS

Some key factors, based on our experience, that may optimize outcomes in PPHN are shown in Fig 3. We recommend a proactive approach to hypoxemic respiratory failure in any late preterm or term infant needing >40% FiO2 with radiographic findings of parenchymal lung disease. This involves early surfactant and lung recruitment maneuvers like conventional mechanical or high frequency ventilation. Early initiation of iNO is important to optimize the oxygenation response and to avoid the need for high ventilator settings or FiO2 or other vasodilator therapies. If there is partial or no response to iNO, we initiate iloprost since it can be given as intermittent nebulizations, while carefully monitoring for hypotension and recognizing its short-half life. If there is a response to iloprost nebulizations, we transition to IV or subcutaneous Treprostinil. All IV prostacyclins require a dedicated venous line which is often challenging, hence our preference for subcutaneous Treprostinil. If there is evidence of left ventricular dysfunction, milrinone is recommended. IV fluids, pressors and steroids are administered to optimize blood pressure especially when multiple adjuvant therapies are needed. We prefer enteral over IV sildenafil due to lower risk of hypotension. Weaning process involves weaning oxygen first, before iNO and is based on pre-ductal pulse oximetry and not solely based on paO2 from post-ductal umbilical artery blood gas monitoring. The Goldilocks’ principle of ‘just the right amount of” oxygen to minimize hypoxic pulmonary vasoconstriction and avoiding hyperoxia lung injury should be adopted by all members of the medical team. Recommendations for vasodilators used in PPHN are summarized in Table 1.

Figure 3.

Suggested approach and timing of interventions for the management of HRF/PPHN. It is important to consider cardiopulmonary system to be one fully integrated unit and optimize lung recruitment, pulmonary vasodilation and cardiac function to facilitate successful transition. The algorithm focuses on vasodilator agents and is not meant to be inclusive of all therapies used in the management of PPHN.

Table 1.

Summary of Dosages & practical considerations for Pulmonary Vasodilator use in PPHN

DRUG CATEGORY	ADMINISTRATION (Route/Dose)	MECHANISM OF ACTION	USE IN PPHN
Oxygen	Goal SpO2 92–97% or paO2 60–90 mm Hg	Enhances NO release from endothelium, activates K+ channels in SMC	First line of treatment Avoid hypoxia or hyperoxia
(NO)
Inhaled nitric oxide (iNO)	Inhalation: Start at 20 ppm, wean gradually 20➔ 10➔ 5➔ 4➔ 3➔ 2➔ 1➔0.5➔off	Generated within pulmonary endothelial cells, diffuses to smooth muscle cells, ↑cGMP ➔ vasodilation	• Standard treatment • Rapid onset, selective pulmonary vasodilator, improves V/Q match • Optimize lung recruitment/ventilation prior to iNO
PHOSPHODIESTERASE INHIBITORS
Sildenafil	IV: loading 0.14 mg/kg/h for 3 h followed by 0.07 mg/kg/h PO/NG: 0.5–2mg/kg/dose Q6–8 h	Inhibition of phosphodiesterase-5 enzyme (responsible for cGMP degradation), ↑cGMP ➔ vasodilation	• May potentiate nitric oxide • Safe and easy to administer • May worsen oxygenation due to vasodilation of unventilated areas of the lung
Tadalafil	PO/NG: 1mg/kg/day once daily	Similar to sildenafil	Similar to sildenafil
Milrinone	Term: IV: loading: 50 mcg/kg over 60 min; maintenance: 0.25–0.75 mcg/kg/min OR continuous at 0.25–0.75 mcg/kg/min Preterm GA< 30 weeks: IV: Loading 50mcg/kg/min over 3 h; maintenance: 0.2 mcg/kg/min	Inhibition of phosphodiesterase-3 enzyme (responsible for cAMP degradation), ↑cAMP ➔ vasodilation	• May potentiate action of prostaglandins • Should be strongly considered if diminished RV function
PROSTAGLANDINS
PGI 2	Epoprostenol (Flolan) IV: 2–5 ng/kg/min, increments of 2–5 ng/kg/min Inhalation: 50ng/kg/min continuous Iloprost (Inhaled): 0.5–2 mcg/kg/dose Q2–4 h	Produced from arachidonic acid, increases cAMP in pulmonary vascular smooth muscle ➔ vasodilation	• May enhance NO action • Non-specific pulmonary vasodilator • May cause systemic hypotension (IV>inhalation) • IV formulation needs dedicated line due to incompatibility with most medications/fluids • Avoid abrupt discontinuation if using continuous Epoprostenol infusion/inhalation • Inhaled route is desirable
PGE1	IV: 0.01–0.1 mcg/kg/min Inhalation: 100–300 ng/kg/min	Similar to PGI2	Similar to PGI2
ENDOTHELIN RECEPTOR INHIBITOR
Bosentan	Oral: 1–2 mg/kg twice daily	Non-specific antagonist of endothelin A and B receptors	• Limited proven efficacy • LFTs should be monitored

POTENTIAL FUTURE THERAPEUTIC APPROACHES

Some newer investigational drugs that target a range of cellular mechanisms are worth mentioning, although their potential application to neonatal care remains unexplored. L-citrulline is converted to L-arginine which is a key substrate for enzyme eNOS that produces NO. Although L-citrulline has been shown to decrease PVR and augment functional capacity in adults with PAH [83], there are no clinical studies examining its effects in neonates. Since sGC is the downstream target for NO, sGC stimulators or activators (riociguat, a stimulator and cinaciguat, an activator) have been investigated as pulmonary vasodilators. Riociguat is used in adult PH but remains to be studied in pediatric and neonatal PH [84]. Cinaciguat induces pulmonary vasodilation in newborn lambs with PPHN in the setting of oxidative stress [35, 85]. However, both riociguat and cinaciguat require further investigation on efficacy and long-term effects in neonatal population prior to clinical use. PPARγ agonists (Rosiglitazone) regulate smooth muscle cell proliferation as well as smooth muscle vasodilation through inhibition of Rho-kinase. In a rat PH model, activation of PPARγ has decreased RV pressures and vascular remodeling, suggesting its potential role in the management of neonatal PH [86]. Activation of AMP Kinase with metformin ameliorated pulmonary hypertension and improved angiogenesis in the fetal lamb model of PPHN [87].

Pharmacotherapy in BPD-PH

Current guidelines for BPD-PH recommend the initiation of targeted therapy with pulmonary vasodilators in infants with sustained PH after optimization of underlying respiratory and cardiac disease [88]. Although these medications are widely used in this population, there is paucity of data on their safety and efficacy, and effects on long-term outcomes for infants treated with these medications. In the absence of RCT data, use of PH targeted medications in infants with BPD is based on expert opinion and clinical experience, emphasizing the need for their comprehensive evaluation in PH centers. A brief review of Pharmacotherapy in BPD-PH is summarized below along with Table 2.

Table 2.

Pharmacotherapy of pulmonary hypertension in BPD

Names	Dose/titration	Side effects	Comments
Sildenafil phiosphodiesterase-5 inhibitor	PO: 1 mg/kg q 6–8 h; start with low dose (0.3–0.5 mg/kg/dose) and increase gradually to 1 mg/kg/dose as tolerated; slower as outpatient. Maximal dose of 10 mg q 8 h per EMA guidelines for infants. Intravenous:0.25–0.5 mg/kg/dose q 6–8 h (titrate slowly and administer over 60 min.	Hypotension, GER, irritability (headache), bronchospasm, nasal stuffiness, fever, rarely priapism	Monitor for adverse effects, lower the dose or switch to alternate therapy if not tolerated
Bosentan (Endothelin receptor antagonist)	1 mg/kg PO q 12 h as starting dose; may increase to 2 mg/kg BID in 2–4 wk, if tolerated and liver enzymes stable.	Liver dysfunction especially during viral infections, VQ mismatch, hypotension, anemia (edema and airway issues rare in infants)	Monitor LFTs monthly (earlier with respiratory infections); monitor CBC quarterly. Teratogenicity precautions for caregivers
Inhaled Iloprost	2.5–5 mcg every 2–4 h. Can be given as continuous inhalation during mechanical ventilation. Can titrate dose from 1–5 mcg and frequency from every 4 h to continuous.	Bronchospasm, hypotension, ventilator tube crystallization and clogging, pulmonary hemorrhage, prostanoid side effects (GI disturbances), may be teratogenic to caregivers	Need close monitoring for clogged tubing, may need further dilution. May need bronchodilators or inhaled steroid pretreatment with bronchospasm.
Intravenous Epoprostenol (Flolan)	Start at 1–2 ng/kg/min, titrate up slowly every 4–6 h to 20 ng/kg/min; need to increase dose at regular intervals because of tachyphylaxis. Further increases as guided by clinical targets and avoiding adverse effects.	Hypotension, VQ mismatch, GI disturbances. Needs dedicated line, very short half-life with high risk for rebound PH with brief interruption of therapy; line related complications include infection, clogging, breaks in line, thrombosis, arrhythmia)	Monitor closely if added to other vasodilator therapies, such as milrinone; careful attention to line care is essential.
Treprostinil (Remodulin) IV or Subcutaneous	Start at 2 ng/kg/min and titrate every 4–6 h up to 20 ng/kg/min, then slow increase dose as tolerated (dose often 1.5–2 times greater than equivalent epoprostenol dose, if switching medications)	SQ: local site pain; IV: similar risks as with epoprostenol, but treprostinil has a longer half- life, which reduces risk for severe PH with interruption of infusion	Site pain managed with local and systemic measures
Milrinone (IV) (phosphodiesterase-3 inhibitor)	0.15–0.5 mcg/kg/min – lower dosage range when used with other vasodilators	Arrhythmogenic; systemic hypotension and high risk for decreased myocardial perfusion; caution with renal dysfunction	May need to add a pressor, such as vasopressin, to mitigate effects of decrease in systemic pressures.

BID, twice a day; CBC, complete blood count; EMA, European Medicines Agency; GER, gastroesophageal reflux; GI, gastrointestinal; IV, intravenous; kg, kilogram; LFT, liver function tests; mcg, microgram; ng, nanogram; PO, oral; SC, subcutaneous; SR, sustained release; VQ, ventilation-perfusion. Reproduced from Krishnan U, Feinstein JA, Adatia I, Austin ED, Mullen MP, Hopper RK, Hanna B, Romer L, Keller RL, Fineman J, Steinhorn R, Kinsella JP, Ivy DD, Rosenzweig EB, Raj U, Humpl T, Abman SH; Pediatric Pulmonary Hypertension Network (PPHNet). Evaluation and Management of Pulmonary Hypertension in Children with Bronchopulmonary Dysplasia. J Pediatr. 2017 Sep;188:24–34.e1

Inhaled Nitric Oxide:

iNO may be beneficial in improving oxygenation during acute deterioration of oxygenation in the setting of worsening BPD-PH, in addition to other cardiorespiratory support. A dose of 10–20 PPM may be used for acute PH crises and weaned after stabilization as tolerated. There is no evidence supporting longer term use of iNO in infants with BPD-PH.

Sildenafil:

Sildenafil is widely used in the infants with BPD-PH when it persists despite optimizing ventilation and addressing the potential contributing factors (infection, aspiration, airway pathology and intermittent hypoxia). A study of 25 infants with BPD-PH showed hemodynamic improvement in 88% of patients receiving long-term sildenafil with no safety concerns reported [89]. Several other retrospective reviews have demonstrated improvement in echocardiographic measurements of PH after sildenafil initiation [90–93]. Whether these improvements were due to sildenafil treatment or evolution of PH over time is difficult to discern in these observational studies with no controls. A recent meta-analysis of chronic sildenafil use in preterm infants with BPD-PH reported an overall mortality of 29.7% per year and demonstrated improvements in estimated pulmonary arterial pressure and respiratory severity scores in patients on sildenafil [94]. The 2012 US FDA warning has been discussed in reviews of BPD-PH and in expert panel recommendations that highlight the importance of optimizing the management of contributing factors and evaluation of vasoreactivity by cardiac catheterization if long-term therapy with sildenafil is being considered [88].

Prostacyclins

Iloprost:

There are no large studies of iloprost use in BPD-PH patients. In a single-center retrospective study of BPD-PH, its use was discontinued in 17% of patients, most commonly due to increased oxygen requirement [95]. There was higher mortality in the group that received iloprost. However, this was attributed to the underlying severity of illness and its use in acutely deteriorating patients, and not to the medication itself [95]. Given its rapid onset of action, ease of administration and discontinuation, it may be useful in a select group of patients in acute pulmonary hypertensive crisis. If there is a response, infants can be transitioned to parenteral prostacyclins [95].

Epoprostenol:

There is limited evidence on continuous prostacyclin use in BPD-PH. A case report demonstrated improvement in pulmonary artery pressures, quality of life, and eventual discontinuation of home ventilation in a child with BPD-PH treated with IV epoprostenol [96] and another case report demonstrated improvement in RV systolic pressures [97]. However, intravenous access can be a significant challenge in premature infants. Hence, subcutaneous treprostinil has been used at some experienced PH centers.

Treprostinil:

Subcutaneous treprostinil via a continuous infusion pump was successfully used in 5 patients with BPD-PH as a long term therapy to safely deliver prostacyclin while avoiding the need for central venous access [98]. There was improvement in echocardiographic measures of PH, decreased need for respiratory support and minimal pump-site specific reactions (pain, redness, swelling) [98]. However, this therapy needs a dedicated PH team, committed outpatient follow up program for dose titration and home nursing support along with extensive education, counseling and support of caregivers.

Bosentan:

Bosentan has been used as an add-on therapy in BPD-PH. A case series of 6 patients with BPD-PH demonstrated improvement in respiratory and hemodynamic status over the course of 2.1–2.9 years, 4 of whom were also on sildenafil [99]. Liver function should be monitored cautiously, as 2 of 6 infants had elevated liver enzymes. The long-term efficacy and safety of bosentan from these limited case series remain unknown.

Complexities of PPHN clinical research

Despite decades of work to elucidate molecular pathways and to understand mechanisms using excellent animal models of PPHN, there is paucity of level 1A evidence in human neonates for various treatment modalities commonly used in bedside management of PPHN. iNO and ECMO are the only clinically proven, effective, life-saving treatment options in severe cases of PPHN. More recent RCT data provides convincing level 1A evidence that early surfactant administration not only in RDS but for all late preterm and term infants with parenchymal lung disease and hypoxic respiratory failure significantly decreases the need for ECMO [2, 30, 100, 101]. Evidence also suggests that high-frequency ventilation is superior to conventional ventilation for alveolar recruitment in parenchymal lung disease associated with PPHN [102].

In 30–40% cases of PPHN, iNO is either clinically ineffective or provides only transient improvement. Hence, there is an urgent need to conduct high quality RCTs evaluating the efficacy of additional pulmonary vasodilator therapies in iNO resistant PPHN. Majority of pulmonary vasodilator adjuvant therapies have either not been tested in large multicenter clinical trials, or those RCTs have been conducted in low resource settings lacking iNO and ECMO, limiting their generalizability. Many of these studies included small numbers, heterogenous populations with different etiologies of PPHN and lack long-term follow up. The reasons for lack of larger RCTs in PPHN are several: (1) PPHN is a rare disease affecting only 1 in 500 infants. Hence, it is difficult to recruit adequate number of patients at single centers and requires multicenter trials which presents logistical challenges. (2) Presentation of severe PPHN is often unexpected in an otherwise uncomplicated pregnancy and may not be known early enough (antenatal/early postnatal) for informed consent. Within a narrow timeframe, discussing the study and consenting parents who are frequently at a distant delivery hospital is challenging. (3) Since iNO is the standard of care and available in developed countries, it is not possible to randomize and compare adjuvant therapies versus iNO. (4) iNO has decreased the mortality and ECMO rates remarkably in severe PPHN. Further reductions in mortality or ECMO use would need substantially higher number of infants to be adequately powered to detect the effect size for traditional outcome of ECMO/mortality. Alternate study endpoints need to be used to test efficacy of other therapies. (5) Parental reluctance and physician bias to use therapies off label rather than enroll infants into a complex RCT that may randomize the infants into placebo group. The combination of situational, logistical, financial and ethical considerations in a rare and critically ill patient population challenges the successful recruitment of study subjects into otherwise much needed trials.

CONCLUSION

In the absence of RCT evidence of efficacy of adjuvant pulmonary vasodilator therapies, clinicians are faced with difficult decisions while managing critically ill PPHN infants. As an alternative to traditional RCTs with equal allocation to placebo and treatment groups, alternative trial designs with adaptive and response adaptive clinical trials have been suggested by expert panels. Such trial designs to conduct pragmatic trials in the rare disease population of infants will improve the quality of evidence and allow for timely conduct of trials in this population. Until such randomized trial data is available, clinicians should adopt a physiology based approach backed by knowledge of pathophysiology, drug pharmacology and regular bedside assessment of response to these adjuvant therapies.

Key Points.

• ➢Up to one-third of newborns with PPHN do not respond to Inhaled Nitric Oxide
• ➢Advances in understanding of PPHN pathobiology in animal models has provided the basis for alternative therapeutic targets in iNO non-responders
• ➢Efficacy of other vasodilators is limited to observational studies
• ➢Randomized Clinical Trials from resource-constrained settings without access to iNO/ECMO supports use of some alternate vasodilators. Benefit of such therapies in infants already on iNO remains unclear
• ➢Conduct of traditional randomized trials in PPHN is challenging due to low incidence of PPHN and low enrollment

Synopsis:

Inhaled Nitric Oxide (iNO) therapy had a transformational impact on the management of infants with PPHN. iNO remains the only approved pulmonary vasodilator for PPHN; yet 30–40% of patients do not respond or have incomplete response to iNO. Lung recruitment strategies with early surfactant administration and high frequency ventilation can optimize the response to iNO in the presence of parenchymal lung diseases. Alternate pulmonary vasodilators are used commonly as rescue, life-saving measures, though there is lack of high-quality evidence supporting their efficacy and safety. This article reviews the available evidence and future directions for research in PPHN.

Best Practices.

What is the current practice?

• Inhaled Nitric Oxide is the only approved pulmonary vasodilator for PPHN in the US and European Union.

• About 30–40% patients do not respond or have incomplete response to iNO, necessitating invasive and expensive treatment modalities like ECMO.

• Alternate pulmonary vasodilator therapies are used commonly as rescue life-saving measures, though there is a lack of high-quality data supporting their efficacy and safety.

• Such alternate therapies (sildenafil, prostacyclins) may be effective and safe in resource-limited settings lacking access to iNO.

What changes in current practice are likely to improve outcome?

• Early surfactant administration and high frequency ventilation in conjunction with iNO are beneficial for optimal lung recruitment and V-Q matching in PPHN with parenchymal lung disease.

• Avoidance of hyperoxia and maintaining gentle ventilation are lung protective strategies that will maximize the benefit from pulmonary vasodilators.

• Echocardiography is essential to rule out cyanotic congenital heart disease in babies suspected of having PPHN, prior to consideration of vasodilator therapies.

• Functional echocardiography to assess hemodynamics and impact of therapies at the bedside may support better clinical decision making individualized to the infant’s pathophysiology.

• Clinician awareness of mechanism, pharmacokinetics and side-effect profile of adjuvant vasodilator therapies is crucial prior to their application at the bedside.

Summary statement

• An integrated approach at the bedside to optimize cardiopulmonary support and to minimize injury to the lungs and systemic organs is needed to optimize outcomes for neonates with PPHN. The evidence-based approach presented in this review has led to dramatic reductions in ECMO use for neonates with PPHN over the last 22 years since the approval of iNO for PPHN. Further refinements are needed to test newer therapies using novel trial designs in this rare disease population.