Control and management of
pulmonary vascular tone in the newborn

ROBIN STEINHORN, MD

aSpring 2000

At birth, pulmonary  vascular resistance normally falls dramatically as the lung assumes the function of gas exchange. In some newborn infants, this decrease in pulmonary vascular tone does not occur, resulting in persistent pulmonary hypertension of the newborn (PPHN). This syndrome occurs in more than 1 in 1000 live births, complicates more than 10% of neonatal intensive care unit admissions, and results in substantial morbidity and mortality in otherwise normal term infants. It is characterized by marked pulmonary hypertension causing right to left extrapulmonary shunting of blood and hypoxemia. PPHN can occur idiopathically, or it can complicate a wide variety of neonatal cardiorespiratory disorders, including hyaline membrane disease, asphyxia, meconium aspiration, and congenital diaphragmatic hernia.

THE ROLE OF ECMO IN PPHN TREATMENT

Until recently, there was little to offer the infant with PPHN. In the mid 1970s, an adaptation of heart-lung bypass used in the operating room was developed and first applied to a newborn with respiratory failure. Termed ECMO, or extracorporeal membrane oxygenation, catheters were placed into the right atrium to drain deoxygenated blood through a membrane oxygenator and to return it into the right common carotid artery. Working virtually alone in Southern California in 1976, Dr. Robert H. Bartlett had the vision to use ECMO to treat a baby girl with severe meconium aspiration. The nurses, moved by the desperate situation, named her "Esperanza," Spanish for "hope." Not only did the infant survive, Esperanza is now an adult with two children of her own.

Today, ECMO remains a cornerstone of the therapy for persistent pulmonary hypertension of the newborn. Approximately three years ago, results of the United Kingdom Randomized Neonatal ECMO Trial were published in the Lancet. This prospective randomized study offered firm proof that ECMO was beneficial in decreasing mortality without increasing morbidity for infants with severe respiratory failure. Now there are close to 100 ECMO centers in the United States alone and at least another 20 in Europe. Children’s Memorial Hospital’s program is one of the largest in the country; it is under the capable direction of Marleta Reynolds, MD, Pediatric Surgery.

While effective, ECMO is a formidable undertaking. Large catheters are inserted into the jugular vein, and in some cases, carotid arteries of small babies. Cerebral blood flow is inevitably altered, complete heparinization is required, and there are distinct risks of complications such as intracranial hemorrhage and thrombotic events. Round-the-clock monitoring by an ECMO technician and bedside nurse are required. Technical improvements have been and continue to be made, including the use of venovenous catheters and percutaneous cannulation. However, it is important to remember that ECMO by itself is supportive, not specific therapy. It simply allows time for lungs and pulmonary vasculature to rest and heal. The next wave of advances for PPHN will come from understanding what causes persistent pulmonary hypertension, so that specific therapies can be directed to the root of the problem.

In this review, mediators that may play a role in the normal transition and the development of PPHN are discussed. Conventional and new treatments for PPHN, such as inhaled nitric oxide, surfactant, and liquid ventilation are also reviewed.

NORMAL FETAL CIRCULATION

See Table 1 for proposed mediators of vascular resistance in the fetal pulmonary vasculature. Much of our understanding of the regulation of the fetal circulation comes from experiments done in the fetal lamb. During the latter half of gestation, the number of small blood vessels increases 40-fold, while the wet weight of the lung increases 4-fold. Therefore, the number of small blood vessels per unit of lung increases 10-fold. Despite this increase in the number of vessels, high arterial pressure and high vascular resistance characterize the fetal pulmonary circulation, and pulmonary blood flow does not increase during the last trimester. Right ventricular output instead bypasses the lungs through the foramen ovale and ductus arteriosus and is directed to the organ of gas exchange, the placenta. As the fetal pulmonary vascular bed responds readily to vasodilators, the high-resistance, low-flow state is believed to be maintained in large part by active vasoconstriction. Hypoxic pulmonary vasoconstriction, sympathetic tone, norepinephrine, arachidonic acid metabolites, and endothelin are only a few of the potential mechanisms and mediators that have been proposed to produce high fetal vascular tone.

Vasoconstriction in response to low oxygen tension is at least partially responsible for the high pulmonary vascular resistance of the fetus. Hypoxic pulmonary vasoconstriction develops over the period of gestation when the cross-sectional area of the vascular bed is increasing rapidly. Decreasing oxygen tension in the fetus early in gestation (103 days) does not increase pulmonary vascular resistance, but in the near term fetus (132–138 days) it doubles the resistance. Conversely, increasing oxygen tension before 100 days of gestation does not decrease pulmonary vascular resistance, but by 135 days of gestation it decreases resistance markedly and increases pulmonary blood flow to normal newborn levels. While the mechanism by which low fetal oxygen tension regulates pulmonary vascular tone is not known, it may be in part through regulation of activity and gene expression of nitric oxide synthase and/or endothelin. In addition, acute decreases in oxygen tension decrease synthesis of vasodilator prostaglandins such as prostacyclin in pulmonary arteries isolated from fetal lambs.

Schematic diagram showing the proposed mechanism in the production of and response to nitric oxide. Nitric oxide synthase can be stimulated by a number of pharmacologic agents, such as acetylcholine, as well as by birth and oxygen. Endothelial production of NO can be blocked by L-arginine derivatives such as L-NA. Nitric oxide produced from the endothelium, or delivered as an inhaled gas, activates soluble guanylate cyclase and increases cGMP concentrations in vascular smooth muscle. Vascular smooth muscle relaxation occurs in response to cGMP. Cyclic GMP is inactivated by specific phosphodiesterases. Zaprinast and dipyridamole are two inhibitors of the cGMP specific phosphodiesterase, and are possible therapeutic tools for enhancing cGMP responses to nitric oxide. New experimental phosphodiesterase inhibitors, such as E4021, appear even more promising.

THE TRANSITION AT BIRTH

At birth, a dramatic decrease in pulmonary vascular resistance allows half of the combined ventricular output to be redirected from the placenta to the lung. This decrease in pulmonary vascular resistance leads to an 8- to 10-fold increase in pulmonary blood flow. The increase in pulmonary blood flow increases the pressure in the left atrium, thus closing the one-way flap valve of the foramen ovale. Systemic vascular resistance increases at birth, at least in part due to removal of the low vascular resistance bed of the placenta. The largest immediate drop in pulmonary vascular resistance occurs at birth. However, PVR continues to drop over the first several months of life until it reaches the low level normally found in the adult circulation. As pulmonary vascular resistance falls to less than systemic, blood flow through the patent ductus arteriosus initially reverses, followed rapidly by functional closure. The ductus arteriosus may not anatomically close for several months. It is critical that all of the above events occur within the first several hours of life for the normal postnatal pattern of circulation to be established.

The stimuli that seem to be most important in the rapid decrease in pulmonary vascular resistance at birth are the ventilation of the lungs with a gas and raising oxygen tension in the lungs. Fetal lambs exposed to either of the following conditions exhibit a decrease in pulmonary blood flow: in one scenario, the lambs are exposed to intrauterine ventilation without changing the arterial tension of carbon dioxide or oxygen, while in another scenario, the lambs' fetal PaO₂ is increased by administering hyperbaric oxygenation to the ewe.

Prostaglandins have been proposed as important mediators of the transition at birth. Indomethacin and meclofenomate, both inhibitors of prostaglandin synthesis, blunt the decrease in pulmonary vascular resistance noted when the fetal lung is ventilated with a gas. Prostacyclin and its metabolites dilate the fetal pulmonary vascular bed, and prostacyclin synthesis in ovine PA endothelium and vascular smooth muscle increases markedly during late fetal and early newborn life. However, the relative role of prostacyclin during normal transition and PPHN is not clear. Indomethacin only modestly increases pulmonary artery pressure in term lambs at birth, and its effect on the transition to gas exchange by the lungs at birth is variable.

NITRIC OXIDE IN THE MEDIATION PROCESS

The role of endothelial derived nitric oxide (NO) during fetal lung development and transition is currently of great interest. Nitric oxide is produced from the terminal nitrogen of arginine by nitric oxide synthase. All three of the known nitric oxide synthase (NOS) isoforms are present and developmentally regulated in the fetal rat lung. The pulmonary vasodilation produced by acetylcholine in the fetal lamb can be blocked by the arginine analogues L-NA and L-NMMA, and restored by excess L-arginine, indicating that acetylcholine is stimulating endothelial nitric oxide synthase activity. Blocking NOS activity with the arginine analogue LNA immediately prior to delivery of the fetal lamb blunts the increase in pulmonary blood flow and decrease in pulmonary vascular resistance following birth. These findings indicate that production of nitric oxide is almost certainly an important mediator of the normal transition at birth.

Oxygen is an important stimulus for endothelial nitric oxide production during transition. The increased pulmonary blood flow following hyperbaric oxygenation of the fetal lamb is almost completely blocked by inhibition of NO synthase. The mechanisms for the increase in NO release related to increased oxygen tension are not clear. Both basal and stimulated release of endothelial nitric oxide increase when oxygen tension is acutely increased in pulmonary arteries isolated from late gestation fetal lambs, and NO production is attenuated following acute decreases in oxygenation in cultured fetal pulmonary arterial cells. Changes in oxygen tension regulate nitric oxide synthase gene expression in fetal pulmonary arterial cells and human endothelial cells. In addition to these direct effects, oxygen increases red blood cell ATP, which is a fetal pulmonary vasodilator and potential stimulus for endothelial NO production.

Cyclic GMP mediates the vascular smooth muscle relaxation produced by NO. Infusions of cGMP analogues produce prolonged dilations of the pulmonary circulation of the fetal lamb. The magnitude and duration of cGMP effect is regulated by specific phosphodiesterase (PDE) isoforms. Inhibition of cGMP phosphodiesterases dilates the pulmonary vascular bed of normal fetal lambs. These relaxations are blocked following inhibition of NO synthase, suggesting the primary source of cGMP is due to endothelial nitric oxide and soluble guanylate cyclase activity. Phosphodiesterases are further believed to be involved in signaling pathways regulating the interaction between cAMP and cGMP.

SO WHAT CAUSES PPHN?

PPHN occurs when the normal changes described above do not occur. Many newborn infants are born with parenchymal lung disease, which interferes with the normal ventilation and oxygenation of the lungs. Several diseases, such as meconium aspiration syndrome, surfactant deficiency and pneumonia, are commonly associated with PPHN. Inflammatory lipid mediators such as thromboxanes, leukotrienes and platelet activating factor contribute to the development of PPHN associated with group B streptococcal sepsis and pneumonia. All of these diseases produce stimuli such as acidosis, hypoxia, hypercarbia, and endothelin that may abnormally constrict the transitional pulmonary circulation.

Meconium aspiration syndrome is one of the most frequent causes of respiratory failure and persistent pulmonary hypertension in newborns. However, it is interesting to note that respiratory failure develops in fewer than 10% of infants with meconium-stained amniotic fluid. Further, when meconium is instilled into the trachea of experimental animals, only modest pulmonary hypertension develops. While it has been suggested that meconium is not injurious to the lung unless associated with perinatal asphyxia, combining meconium instillation with acute asphyxia in baboons did not worsen pulmonary hypertension. These findings should lead us to conclude that parenchymal disease alone is often not an adequate explanation for the development of PPHN.

In fact, infants may develop PPHN without any of these abnormal stimuli. In these infants, the pulmonary hypertension appears to be caused by a structurally abnormal pulmonary vascular bed. An increase in pulmonary arterial medial smooth muscle and extension of muscle to normally nonmuscular pulmonary arteries has been reported in newborns who died with idiopathic PPHN. Since these changes are observed in at least some infants dying in the first 24 hours of life, this suggests that an altered intrauterine environment may produce structural changes in the pulmonary circulation of the fetus. Chronic intrauterine hypoxia has been hypothesized to produce these changes.

While a number of attempts have been made to create models of chronic hypoxia in the fetus, unfortunately these models have not been developed to the point of determining mechanisms by which pulmonary hypertension develops. In the fetus, the diversion of right ventricular output away from the lungs across the ductus arteriosus may be an important mechanism that protects remodelling of resistance arteries. One of the causes of PPHN in infants is prenatal constriction of the ductus arteriosus due to maternal ingestion of prostaglandin synthesis inhibitors.

PPHN is more common in post-term newborns, which may be due to intrauterine constriction of the fetal ductus arteriosus. Fetal lambs born 7 to 14 days following ductal constriction or ligation have persistent pulmonary hypertension, with many of the physiologic hallmarks of the human syndrome, including pulmonary arterial pressure equal to aortic pressure and hypoxemia unresponsive to ventilation with 100% oxygen. Structural alterations include extension of muscle into the normally nonmuscular distal arteries and the formation of periadventitial fibrosis surrounding the intraacinar arteries. Studies utilizing this model have revealed that critical enzyme pathways are altered in structurally abnormal vessels. Activity, message, and protein content of endothelial nitric oxide synthase are decreased in lung extracts of ligated compared to control fetal lambs. Following ductal ligation, newborn lambs require concentrations as high as 100 ppm NO to maximally decrease pulmonary arterial pressure and pulmonary vascular resistance. This may be due to enhanced endothelin-1 production, and/or alterations in activity and content of soluble guanylate cyclase or phosphodiesterases.

Pulmonary hypertension frequently occurs in infants with congenital diaphragmatic hernia. Hypoplasia of the pulmonary vascular bed, in the form of a reduction of total cross-sectional area of pulmonary arteries and veins, occurs in addition to hypoplasia of the lung itself, probably related to the fact that vessels develop in parallel with the conducting airways. Abnormal muscularization of the pulmonary arteries also occurs. In addition to structural changes, there may be alterations in some of the biochemical mediators of the pulmonary vascular transition. Plasma thromboxane concentrations correlate with pulmonary vascular resistance, and endothelin levels increase during clinical episodes of deterioration of CDH babies. Further research will determine whether these vasoconstrictors have a role in the etiology of the pulmonary hypertension or are synthesized in response to it.

CONVENTIONAL TREATMENT FOR PPHN

Initial therapies are directed toward correcting acidosis, hypoxia, and hypercarbia. When parenchymal lung disease is present, mechanical ventilation with sufficient positive end-expiratory pressure should allow for alveolar recruitment. In some infants with hypoplastic lungs or especially severe parenchymal lung disease, high-frequency ventilation may permit adequate gas exchange with lower tidal volumes and perhaps less barotrauma. However, excessive mean airway pressure may actually be counterproductive, as alveolar overdistension may compromise cardiac output and actually decrease pulmonary blood flow. Sedation with narcotics is often necessary to achieve adequate mechanical ventilation. Muscle paralysis is occasionally used for the same purpose, but often has adverse circulatory effects and may lead to overdistension in some areas of the lung, and alveolar collapse in dependent regions of the lung. Cardiac output is maintained with the use of inotropic agents and judicious use of volume replacement.

Pulmonary artery pressure decreases and oxygenation acutely improves following hyperventilation in human infants, although in some infants the PaCO₂ needs to be reduced to 20 torr and pH increased to as high as 7.6. For this reason, hyperventilation became a popular treatment for PPHN that was not responsive to the above measures. However, it is interesting to note that hyperventilation has never been prospectively demonstrated to reduce the morbidity or mortality for PPHN. In fact, hyperventilation with 100% oxygen has been shown to increase long-term neurodevelopmental disabilities and increase the incidence of sensorineural hearing loss.

NEW PPHN TREATMENTS

It is important to re-emphasize that infants presenting with PPHN have a diverse pathophysiology. It is probably due to this diversity that no one "magic bullet" exists or may ever exist for infants with persistent pulmonary hypertension. It also explains why many newer therapies described below have beneficial effects that are highly disease-specific. A logical treatment strategy for one infant might fail miserably in the next.

Part of the challenge in finding effective vasodilators for PPHN is that agents that decrease pulmonary vascular resistance tend to decrease systemic vascular resistance at the same time. Inhalational nitric oxide provides a unique solution to this dilemma. While nitric oxide dilates both systemic and pulmonary vessels, it has an affinity for hemoglobin more than 1000 times greater than that of carbon monoxide. Administration of exogenous nitric oxide as an inhaled gas should then be uniquely able to selectively exert its effects on the pulmonary circulation, because any excess will be unavailable to the systemic circulation through inactivation by hemoglobin.

Initial reports of small groups of infants with PPHN showed acute improvement in oxygenation following inhaled NO in doses of 6–80 ppm. Two recent multicenter studies of term infants with PPHN demonstrated the important finding of sustained improvement in oxygenation following inhaled nitric oxide, which decreased the need for ECMO. While these results are an exciting advance in the therapeutic approach to PPHN, these studies and others also demonstrated that NO is not universally effective in infants with PPHN, and its effects are not always sustained. This indicates that in some patients, even exogenous nitric oxide may not be able to stimulate or sustain cGMP production over time.

The correct dose of inhaled NO is not known. No dose-response relationship appears to exist in improvement in oxygenation following inhaled NO in doses of 5–80 ppm. This may in part be due to the lack of echocardiographic evidence for pulmonary hypertension in several of the infants studied. However, a recently completed clinical trial evaluated low-dose inhaled NO (20 ppm for up to 24 hours followed by 5 ppm up to 96 hours) for a tightly controlled period of time (less than or equal to 96 hours). Neonates treated with NO had significantly decreased utilization of ECMO with no increase in any morbidities. In fact, infants treated with this low-dose, limited-duration approach less often needed oxygen at 30 days of life than neonates randomized to the control gas (7% vs 20%, p=0.02). These findings imply a low-dose, limited-duration treatment strategy is as effective and potentially safer than higher doses for more prolonged periods. Further direct study of the pulmonary hemodynamic effects of inhaled NO, as well as its effects on oxygenation, will be extremely important in determining the optimal dose.

Potential toxicities of NO are extremely important in considering the ideal NO dose. In the presence of oxygen, NO is rapidly oxidized to nitrogen dioxide which even in very low concentrations can acutely injure the distal airways and alveoli and disrupt the vascular endothelium. Fortunately, significant levels of NO₂ do not appear to be generated during clinical use of NO at concentrations of 20 ppm or less. In many situations where inhaled NO is used, such as inflammatory lung disease and high inspired oxygen concentrations, there may be increased production of superoxide. When NO comes into contact with superoxide, peroxynitrites are formed. Peroxynitrites may damage surfactant associated proteins, inhibit surfactant function, and cause cell damage. Finally, nitric oxide increases cGMP concentrations in platelets as well as in vascular smooth muscle, which can inhibit platelet aggregation and adhesion.

It is widely presumed that when NO is delivered as an inhaled gas, it simply diffuses through the pulmonary interstitium and the vascular adventitia into the vascular smooth muscle cell. Experimental data indicate that despite its extremely small molecular size, nitric oxide does not readily cross the adventitia of larger pulmonary vessels, possibly due to its free-radical nature. Therefore, nitric oxide must be delivered to peripheral lung units to be effective. One strategy that may improve the response to inhaled NO when there is parenchymal lung disease is to deliver it in conjunction with high-frequency oscillatory ventilation at airway pressures sufficient to maintain lung volume.

In some types of parenchymal lung diseases, even more specific approaches may be possible. For instance, surfactant deficiency due to decreased production or to inactivation due to protein leak may be a part of many lung diseases associated with PPHN. Administration of exogenous surfactant has been recently reported to improve oxygenation and reduce pulmonary morbidity and hospitalization time in infants with meconium aspiration syndrome and might enhance the effect of inhaled NO in infants with parenchymal lung disease.

There are also patients who do not respond or who do not sustain their response to nitric oxide due to abnormalities of vascular development prior to or just after birth. Infants with congenital diaphragmatic hernia frequently fall into this category. Remodeled pulmonary vessels appear to have altered content and activity of soluble guanylate cyclase, the target enzyme for NO, within the vascular smooth muscle cell. Potential adjunctive therapies for such patients will need to be targeted at increasing vascular cyclic GMP production to inhaled nitric oxide. It is interesting to note that in addition to inhaled NO, many of the pulmonary vasodilators that have been tried in infants with PPHN may directly activate or enhance soluble guanylate cyclase. The relative intracellular activities of superoxide and superoxide dismutase activity may be a potential pathway for modulating the effects of nitric oxide. By reducing NO clearance by superoxide, superoxide dismutase significantly enhances responses to nitric oxide in vitro, and future studies will determine if this is a useful clinical adjunct as well.

High concentrations of the cGMP specific phosphodiesterase (PDE5) are found in lung homogenates, and levels appear to peak at the time of birth. Phosphodiesterase inhibition may provide an avenue for increasing cGMP concentrations and increasing the efficacy of inhaled nitric oxide in human infants with PPHN. We recently reported that administration of the experimental PDE5 inhibitor zaprinast in combination with a threshold dose of inhaled NO (6 ppm) quadrupled the drop in pulmonary vascular resistance and increase in oxygenation compared to NO alone in the ductal ligation lamb model of PPHN. Therefore, we felt that phosphodiesterase inhibition would allow for administration of lower NO doses, avoiding some of its potential toxicity. Dipyridamole, which has been used for many years in humans, also has significant inhibitory activity against PDE5. Unfortunately, we found that its effects on the systemic circulation will probably limit its clinical applicability. Fortunately, new highly selective PDE5 inhibitors are being developed. We recently reported that a potent experimental PDE5 inhibitor decreased pulmonary artery pressure in an animal model of PPHN. What was surprising in this study was the magnitude of response, which was greater than typically observed with high doses of inhaled NO. We are hopeful that phosphodiesterase inhibitors will eventually provide a safe and effective alternative for the treatment for PPHN.

Acknowledgement: Much of the work reported here was supported by grants to Dr. Steinhorn from the American Heart Association, and National Institutes of Health (HL# 54705). The contributions of Dr. James A. Russell, Sylvia Gugino, and Dr. Frederick C. Morin III are gratefully acknowledged.

FOR FURTHER READING

1. Morin III FC, Stenmark KR: Persistent pulmonary hypertension of the newborn. Am J Respir Crit Care Med 1995;151:2010–32.

2. Kinsella JP, Abman SH: Recent developments in the pathophysiology and treatment of persistent pulmonary hypertension of the newborn. J Pediatr 1995;126:853–864.

3. Roberts JD, Fineman J, Morin III FC, Shaul PW, Rimar S, Schreiber MD, Polin RA, Thusu KG, Zayek M, Zwass MS, Zellers TM, Wylam ME, Gross I, Zapol WM, Heymann MA: Inhaled nitric oxide gas improves oxygenation in PPHN. New Engl J Med 1997;336:605–610.

4. Neonatal Inhaled Nitric Oxide Study Group: Inhaled nitric oxide in full-term and nearly full-term infants with hypoxic respiratory failure. New Engl J Med 1997;336:597–604.

5. Neonatal Inhaled Nitric Oxide Study Group: Inhaled nitric oxide and hypoxic respiratory failure in infants with congenital diaphragmatic hernia. Pediatrics 1997;99:838–845.

6. Steinhorn RH, Millard SL, Morin III FC: Persistent pulmonary hypertension of the newborn: Role of nitric oxide and endothelin in pathophysiology and treatment. Clin Perinatol 1995;22:405–428.

7. Steinhorn RH, Morin III FC, Russell JA: Cyclic GMP production to nitric oxide is reduced in lambs with persistent pulmonary hypertension. Pediatr Res 1995;37:352A.

8. Thusu KG, Morin III FC, Russell JA. Steinhorn RH: The cGMP phosphodiesterase inhibitor zaprinast enhances the effect of nitric oxide. Am J Resp Crit Care Med 1995;152:1605–10.

9. Steinhorn RH, Morin FC III, Fineman JR: Models of persistent pulmonary hypertension and the role of cGMP in pulmonary vasorelaxation. Seminar Perinatol 1997;21:393–408.

10. Hallman M, Bry K: Nitric oxide and lung surfactant. [Review] [101 refs]. Semin Perinatol 1996;20:173–185.

11. Dukarm RC, Russell JA, Morin FC III, Steinhorn RH: The phosphodiesterase inhibitor E4021 selectively dilates the pulmonary circulation. Am J Resp Crit Care Med 1999;858–865.

12. Clark RH, Kueser TJ, Walker MW, et al: Low Dose Nitric Oxide Therapy for Persistent Pulmonary Hypertension of the Newborn. N Engl J Med 2000;342:469–474.

Copyright © 2008 by Children’s Memorial Hospital. All rights reserved.