Nitric oxide and beyond: new insights and therapies for pulmonary hypertension

Abstract

Persistent pulmonary hypertension of the newborn (PPHN) contributes significantly to the morbidity and mortality associated with meconium aspiration syndrome. This review article discusses new insights into the vascular abnormalities that are associated with PPHN, including the recent recognition of the importance of oxidant stress in its pathogenesis. Recent data are presented showing that treatment with high oxygen concentrations may increase production of oxygen free radicals. The rationale for the use of inhaled nitric oxide, and strategies for enhancing nitric oxide signaling are discussed. Finally, the rationale for new treatment approaches is reviewed, including inhibition of cyclic guanosine monophosphate-specific phosphodiesterases and scavengers of reactive oxygen species.

Introduction

Neonatal respiratory failure affects 2% of all live births and is responsible for a substantial proportion of neonatal mortality. Although preterm infants are at higher risk of respiratory failure, term and near-term infants account for one-third of the cases.¹ A better understanding of the pathophysiology of hypoxemic respiratory failure is needed to develop more specific and effective therapies. This review will focus on recent progress in our understanding of pulmonary hypertension, which is commonly associated with the severe respiratory failure that accompanies meconium aspiration syndrome (MAS).

Pathophysiology of PPHN

Shortly after birth, the fetus normally undergoes a rapid cardiopulmonary transition to meet the new demands of extrauterine life. However, if pulmonary vascular resistance does not fall, pulmonary blood flow cannot increase, and the result is hypoxemic respiratory failure or persistent pulmonary hypertension of the newborn (PPHN). The incidence of severe PPHN is estimated at 0.2% of live-born term infants. MAS is the most common cause of PPHN, and when present, it can contribute significantly to its morbidity and mortality. Newborns with hypoxemic respiratory failure and/or PPHN are at risk for numerous complications including death, neurological injury and other morbidities.

Persistent pulmonary hypertension of the newborn can generally be characterized as one of three types: (1) the abnormally constricted pulmonary vasculature due to lung parenchymal diseases, such as MAS, respiratory distress syndrome or pneumonia; (2) the lung with normal parenchyma and remodeled pulmonary vasculature, also known as idiopathic PPHN; or (3) the hypoplastic vasculature as seen in congenital diaphragmatic hernia. The most common cause of PPHN is MAS. Infants with MAS will typically fall into the first or second categories, and the most severe cases are probably affected by both parenchymal and vascular disease.

Understanding the pathophysiology of the abnormally remodeled pulmonary vasculature is of utmost importance in directing therapy. As it is not feasible to study the remodeling process in the human infant, much of our current understanding is derived from animal models. One cause of idiopathic PPHN is constriction of the fetal ductus arteriosus in utero because of exposure to nonsteroidal anti-inflammatory drugs during the third trimester. Ductal constriction or ligation can be surgically performed in utero in lambs, leading to rapid antenatal remodeling of the pulmonary vasculature. Findings in PPHN lambs are similar to those observed in human infants, including increased fetal pulmonary artery pressure, pulmonary vascular remodeling and profound hypoxemia after birth.

On the basis of work from animal models, there is strong evidence that disruptions of the nitric oxide (NO)-cyclic guanosine monophosphate (cGMP), prostacyclin-cyclic adenosine monophosphate (cAMP) and endothelin signaling pathways play an important role in the vascular abnormalities associated with PPHN.² The NO-cGMP pathway has been a topic of particularly intense investigation over the past decade. Decreased expression and activity of endothelial NO synthase have been documented in the PPHN lamb model,³ and decreased endothelial NO synthase expression has also been reported in umbilical venous endothelial cell cultures from human infants with meconium staining who develop PPHN.⁴ These important findings were rapidly followed by clinical testing of inhaled NO (iNO) as a therapy for hypoxemic respiratory failure and PPHN.

Inhaled nitric oxide

The primary aim of PPHN therapy is selective pulmonary vasodilatation. Inhaled NO appears to be well suited for this effect: it is a rapid and potent vasodilator, and because NO is a small gas molecule, it can be delivered through a ventilator directly to airspaces approximating the pulmonary vascular bed. Once in the blood stream, NO binds avidly to hemoglobin, limiting its systemic vascular activity and increasing its selectivity for the pulmonary circulation.

Large placebo-controlled trials provided clear evidence that iNO significantly decreases the need for extracorporeal membrane oxygenation (ECMO) support in newborns with PPHN.^5,6 However, it is important to note that up to 40% of infants will not improve oxygenation or maintain a response to iNO, and iNO did not reduce mortality or length of hospitalization. In addition, follow-up studies for 12 to 24 months indicate that iNO does not significantly alter the incidence of chronic lung disease or neurodevelopmental impairment.^7,8 This is an interesting and important observation that may indicate that the underlying disease is associated with early neurological injury. Finally, Konduri et al.^9,10 determined that starting iNO earlier in the disease course (for an oxygenation index of 15 to 25) did not decrease the incidence of ECMO and/or death or improve other patient outcomes, including the incidence of neurodevelopmental impairment.

Following the introduction of high-frequency ventilation (HFV), surfactant and iNO in the early 1990s, the patient demographic of neonatal support with ECMO has changed. Data from the large registry maintained by the Extracorporeal Life Support Organization indicate that the use of these therapies has increased steadily over the past 10 years, accompanied by a greater than 40% reduction in the number of neonates cannulated for ECMO. However, as overall ECMO survival has diminished over the same time period, some physicians have speculated that these new treatment modalities may delay ECMO cannulation and have a negative effect on mortality and morbidity in those infants that continue to require extracorporeal support. Therefore, we recently examined data from the Extracorporeal Life Support Organization registry between 1996 and 2003. We found that NO, HFV and surfactant use were not associated with any adverse outcomes during ECMO, including increased hours on ECMO or increased time to extubation.¹¹ Furthermore, both surfactant and NO use were associated with lower ECMO mortality, and NO use was associated with a decreased risk of cardiac arrest before cannulation. As ECMO is a proven therapy for severe respiratory failure, it is reassuring that these new therapies have not had a negative impact on the most severely affected infants.

New insights into PPHN pathophysiology

As NO is not universally effective, there has been considerable interest in better understanding the biochemical pathways that regulate pulmonary vasoconstriction and remodeling in PPHN. For instance, NO mediates vasodilatation by stimulating soluble guanylate cyclase in vascular smooth muscle cells, which then converts guanosine triphosphate to cGMP (Figure 1). cGMP is the central and critical second messenger that regulates contractility of the smooth muscle cell by modulating the activity of cGMP-dependent kinases, phosphodiesterases and ion channels. The cGMP-dependent or type 5 phosphodiesterase is potentially important, as it can downregulate cGMP concentrations by degrading cGMP to the inactive 5′-GMP. Therefore, there are critical points in the pathway downstream of NO production that serve as attractive targets for manipulating cellular cGMP concentrations. For example, expression and activity of soluble guanylate cyclase are decreased in the abnormally remodeled pulmonary vessels of the PPHN lamb model, which could potentially diminish responses to both endogenous and exogenous NO. This finding would indicate that new compounds that directly stimulate sGC at an NO-independent but heme-dependent site may be helpful, a hypothesis that appears to be promising in preclinical testing.¹² Another potential cause for low cGMP concentrations would be increased expression and/or activity of cGMP-specific phosphodiesterases, which could then be manipulated through use of specific inhibitors.

Figure 1.

Nitric oxide (NO) and prostacyclin signaling pathways in regulation of pulmonary vascular tone. NO is synthesized by NO synthase (NOS) from the terminal nitrogen of L-arginine. NO stimulates soluble guanylate cyclase (sGC) to increase intracellular cGMP. PGH2 is an arachidonic acid (AA) metabolite formed by cyclooxygenase (COX-1) and prostacyclin synthase (PGIS) in the vascular endothelium. Prostacyclin (PGI₂) stimulates adenylate cyclase in vascular smooth muscle cells, which increases intracellular cAMP. Both cGMP and cAMP indirectly decrease free cytosolic calcium, resulting in smooth muscle relaxation. Specific phosphodiesterases hydrolyze cGMP and cAMP, thus regulating the intensity and duration of their vascular effects. Inhibition of these phosphodiesterases with agents such as sildenafil and milrinone may enhance pulmonary vasodilation.

Several new lines of evidence now indicate that oxidant stress is important in the pathogenesis of PPHN. An increase in reactive oxygen species (ROS) such as superoxide and hydrogen peroxide in the smooth muscle and adventitia of pulmonary arteries has been demonstrated in the PPHN lamb model.^13,14 Possible sources of elevated concentrations of ROS include increased expression and activity of NADPH oxidase and a reduction in superoxide dismutase (SOD) activity. PPHN lambs also demonstrate diminished binding of the chaperone protein, heat shock protein 90, to endothelial NO synthase.¹⁵ Decreased heat shock protein 90–endothelial NO synthase interactions lead to an ‘uncoupling’ of NOS activity, which results in decreased synthesis of NO and increased superoxide production. Once present in the lung, elevated concentrations of ROS are believed to play a role in vascular smooth muscle cell proliferation in PPHN, as well as abnormal vascular reactivity.

Finally, current therapeutic practices may have an effect on pulmonary vascular reactivity and remodeling. In particular, the use of oxygen has recently become controversial in numerous settings. While oxygen is a pulmonary vasodilator, the extreme hyperoxia routinely used in PPHN management may be toxic to the developing lung by the formation of ROS. Although hyperoxic ventilation is a mainstay of the treatment of PPHN, we know surprisingly little about what oxygen concentrations will maximize benefits and minimize risk. Superoxide may react with arachidonic acid to increase concentrations of isoprostanes and may also combine with NO to form peroxynitrite. Both are potent oxidants with the potential to produce vasoconstriction, cytotoxicity and damage to surfactant proteins and lipids (Figure 2). New data indicate that even brief periods of ventilation with 100% O₂ (30 min) are sufficient to increase reactivity of pulmonary vessels in normal lambs¹⁶ and to diminish the response of the lung to endogenous and exogenous NO.¹⁷

Figure 2.

Hypothesized role of reactive oxygen species (ROS) and their metabolites in mediating increased pulmonary arterial contractility following exposure to 100% oxygen. Exposure to high oxygen concentrations results in the formation of superoxide anions ( ${\text{O}}_2^{-}$). Superoxide anions combine with nitric oxide (NO) to form peroxynitrite, a potent pulmonary vasoconstrictor. Superoxide dismutase (SOD) enzyme converts superoxide anions to hydrogen peroxide (H₂O₂), also a pulmonary vasoconstrictor. When membrane lipids (arachidonic acid and polyunsaturated fatty acids (PUFA)) are exposed to ROS, such as superoxide anions and hydrogen peroxide or peroxynitrite, a variety of isoprostanes are formed. Isoprostanes are known constrictors of pulmonary arteries. Adapted from Lakshminrusimha et al.¹⁶ with permission.

Alternative and emerging pulmonary vasodilators

An improved knowledge of the biochemical abnormalities responsible for refractory PPHN is leading to a growing list of promising new therapeutic strategies. Many investigators are seeking to enhance cGMP-mediated vasodilation through the use of cGMP-specific phosphodiesterase inhibitors, direct soluble guanylate cyclase activators, scavengers of ROS as well as manipulation of the cAMP pathway.

Similar to cGMP, cAMP also stimulates vasodilatation (Figure 1). One potential approach that might take advantage of this mechanism is using milrinone to inhibit PDE3, the phosphodiesterase that metabolizes cAMP. Milrinone has been shown to decrease pulmonary artery pressure and resistance and to act additively with iNO in animal studies. Recent reports indicate that it may decrease rebound pulmonary hypertension after discontinuation of iNO and may enhance pulmonary vasodilation of infants refractory to iNO.¹⁸ Prostacyclin (PGI₂) stimulates membrane bound adenylate cyclase, increases cAMP and inhibits pulmonary artery smooth muscle cell proliferation in vitro. Although the use of systemic infusions of PGI₂ may be limited by systemic hypotension, inhaled PGI₂ has been shown to have vasodilator effects limited to the pulmonary circulation. Reports in children have been positive, but to date there have been few reports of inhaled PGI₂ use in neonates with PPHN.¹⁹ It is most likely that further investigations will focus on preparations specifically designed for inhalation, such as iloprost.

There has been particular interest in the inhibition of cGMP-metabolizing phosphodiesterase activity, which would theoretically increase cGMP concentrations and result in pulmonary vasodilation and/or increased efficacy of iNO (Figure 1). On the basis of clinical trials, sildenafil, a potent and highly specific phosphodiesterase inhibitor, has recently been approved by the Food and Drugs Administration for the treatment of pulmonary hypertension in adults. In lambs with experimental pulmonary hypertension, both enteric and aerosolized sildenafil dilate the pulmonary vasculature and augment the pulmonary vascular response to iNO. Intravenous sildenafil was found to be a selective pulmonary vasodilator with efficacy equivalent to inhaled NO in a piglet model of meconium aspiration, although hypotension and worsening oxygenation resulted when it was used in combination with iNO.^20,21 Sildenafil may attenuate rebound pulmonary hypertension after withdrawal of iNO in newborn and pediatric patients.²² Use of sildenafil in PPHN has been limited by its availability only as an enteric form, although a recent report indicates that it improved oxygenation and survival in human infants with PPHN compared with placebo.²³ A pilot trial of intravenous sildenafil was recently conducted in newborns with pulmonary hypertension, and data analysis is nearing completion.

New laboratory studies indicate that scavengers of ROS such as SOD may augment responsiveness to iNO. As described above, increased production of superoxide is noted in experimental models of PPHN. As iNO is usually delivered with high concentrations of oxygen, there is the potential for enhanced production of additional oxidants such as peroxynitrite. Superoxide dismutase scavenges and converts superoxide radical to hydrogen peroxide, which is subsequently converted to water by the enzyme catalase. Administration of recombinant human SOD (rhSOD) has been tested in preterm infants without adverse effects and with trends toward decreased pulmonary morbidity. In lambs with pulmonary hypertension, rhSOD was found to dilate the pulmonary circulation and enhance responsiveness to inhaled NO.²⁴ A recent study examined the effects of rhSOD on oxygenation over a 24-h period in ventilated PPHN lambs.²⁵ The results showed that a single dose of rhSOD by itself improved oxygenation to a degree that was similar to iNO (Figure 3). Furthermore, rhSOD treatment appeared to block formation of oxidants such as peroxynitrite and isoprostanes. Thus, an antioxidant therapeutic approach may have multiple beneficial effects: scavenging superoxide may make both endogenous and inhaled NO more available to stimulate vasodilatation and may also reduce oxidative stress and limit lung injury. It is hoped that human trials will begin soon.

Figure 3.

Superoxide dismutase (SOD) and nitric oxide (NO) improve the arterial-to-alveolar oxygen (a/A) ratio to a similar degree. The a/A ratio of four groups of lambs with persistent pulmonary hypertension of the newborn (PPHN) ventilated with 100% oxygen alone (100% O₂), 100% oxygen with rhSOD 5 mg kg⁻¹ administered at birth (100% O₂ SOD), 100% oxygen with rhSOD 5 mg kg⁻¹ administered at 4 h of life (100% O₂ 4h delay SOD) and 100% oxygen and 20 p.p.m. of inhaled NO (iNO; 100%O₂ NO). PPHN lambs ventilated with 100% oxygen were critically ill, and two of these lambs died at approximately 10 and 16 h of age (shown by arrows). The dashed line beyond these points represents mean±s.e.m. for the remaining lambs. *P<0.05 compared with 100% oxygen-alone group. Adapted from Lakshminrusimha et al.²⁵ with permission.