Pulmonary arterial hypertension in children · Pulmonary arterial hypertension in children A. Widlitz, R.J. Barst Pulmonary arterial hypertension in children. A. Widlitz, R.J. Barst

SERIES 0ADVANCES IN PATHOBIOLOGY, DIAGNOSIS, AND TREATMENT OFPULMONARY HYPERTENSION0Edited by A.T. Dinh-Xuan, M. Humbert and R. NaeijeNumber 5 in this Series

Pulmonary arterial hypertension in children

A. Widlitz, R.J. Barst

Pulmonary arterial hypertension in children. A. Widlitz, R.J. Barst. #ERS JournalsLtd 2003.ABSTRACT: For physicians to admit that a group of patients remains for whom nocure is available in modern medicine is intellectually unsatisfying. Pulmonary arterialhypertension is a rare condition. Because the symptoms are nonspecific and the physicalfinding can be subtle, the disease is often diagnosed in its later stages. The naturalhistory of pulmonary arterial hypertension is usually progressive and fatal.At the 1998 Primary Pulmonary Hypertension World Symposium, clinical scientists

from around the world gathered to review and discuss the future of pulmonary arterialhypertension. Bringing together experts from a variety of disciplines provided theopportunity for a better understanding of the pathology, pathobiology, risk factors,genetics, diagnosis and treatment for pulmonary arterial hypertension.Remarkable progress has been made in the field of pulmonary arterial hypertension

over the past several decades. The pathology is now better defined and significantadvances have occurred in understanding the pathobiological mechanisms. Risk factorshave been identified and the genetics have been characterised. Advances in technologyallow earlier diagnosis as well as better assessment of disease severity. Therapeuticmodalities such as new drugs, e.g. epoprostenol, treprostinil and bosentan, and surgicalinterventions, e.g. transplantation and blade septostomy, which were unavailable severaldecades ago, have had a significant impact on prognosis and outcome. Thus, despite theinability to really cure pulmonary arterial hypertension, therapeutic advances over thepast two decades have resulted in significant improvements in the outcome for childrenwith various forms of pulmonary arterial hypertension.This review of pulmonary arterial hypertension will highlight the key features of

pulmonary hypertension in infants and children and the current understanding ofpulmonary arterial hypertension with specific recommendations for current practice andfuture directions.Eur Respir J 2003; 21: 155–176.

Dept of Pediatrics, Columbia Uni-versity College of Physicians andSurgeons, New York, NY, USA.

Correspondence: R.J. Barst3959 BroadwayBHN-2New YorkNY10032USAFax: 1 2123054429E-mail: [email protected]

Keywords: Paediatricspulmonary arterial hypertensionpulmonary heart disease

Received: September 24 2002Accepted: October 3 2002

Until recently the diagnosis of primary pulmo-nary hypertension was virtually a death sentence.This was particularly true for children, in whom themean survival was v1 yr. This bleaker outlook forchildren compared to adults was underscored bythe data in the Primary Pulmonary HypertensionNational Institutes of Health Registry [1]. In thisRegistry, the median survival for all of the 194patients was 2.8 yrs, whereas it was only 10 monthsfor children. Significant progress in the field ofpulmonary hypertension has occurred over the pastseveral decades. Advances in technology have alsoallowed a better diagnosis and assessment of the

disease severity with treatment now available thatimproves quality of life and survival [2–4]. Never-theless, extrapolation from adults to children is notstraightforward for at least several reasons: 1) theanticipated lifespan of children is longer; 2) childrenmay have a more reactive pulmonary circulationraising the prospect of greater vasodilator respon-siveness and better therapeutic outcomes [5]; and 3)despite clinical and pathological studies suggestingincreased vasoreactivity in children, before the adventof long-term vasodilator/antiproliferative therapy, thenatural history remained significantly worse forchildren compared to adult patients [1, 6].

Previous articles in this Series: No. 1: Humbert M, Trembath RC. Genetics of pulmonary hypertension: from bench to bedside. Eur Respir J2002; 20: 741–749. No. 2: Galiè N, Manes A, Branzi A. The new clinical trials on pharmacological treatment in pulmonary arterialhypertension. Eur Respir J 2002; 20: 1037–1049. No. 3: Chemla D, Castelain V, Hervé P, Lecarpentier Y, Brimioulle S. Haemodynamicevaluation of pulmonary hypertension. Eur Respir J 2002; 20: 1314–1331. No. 4: Eddahibi S, Morrell N, d9Ortho M-P, Naeije R, Adnot S.Pathobiology of pulmonary arterial hypertension. Eur Respir J 2002; 20: 1559–1572.

Eur Respir J 2003; 21: 155–176DOI: 10.1183/09031936.03.00088302Printed in UK – all rights reserved

Copyright #ERS Journals Ltd 2003European Respiratory Journal

ISSN 0903-1936

Definition and classification

The definition of primary pulmonary hyperten-sion in children is the same as for adult patients. Itis defined as a mean pulmonary artery pressureo25 mmHg at rest or o30 mmHg during exercise,with a normal pulmonary artery wedge pressure andthe absence of related or associated conditions. Theinclusion of exercise haemodynamic abnormalities inthe definition of pulmonary arterial hypertension isimportant since children with pulmonary arterialhypertension often have an exaggerated response ofthe pulmonary vascular bed to exercise as well as inresponse to hypoventilation compared with adults.Not uncommonly, children with a history of recurrentexertional or nocturnal syncope have a resting meanpulmonary artery pressure of only y25 mmHg thatmarkedly increases with modest systemic arterialoxygen desaturation during sleep, as well as withexercise.

In 1998, at the Primary Pulmonary HypertensionWorld Symposium, clinical scientists from aroundthe world proposed a new diagnostic classificationsystem (table 1). This classification system cate-gorises pulmonary vascular disease by commonclinical features. This classification reflects the recentadvances in the understanding of pulmonary hyper-tensive diseases as well as recognising the similaritybetween primary pulmonary hypertension and pulmo-nary hypertension of certain other causes. Thus, inaddition to primary pulmonary hypertension (bothsporadic and familial), pulmonary arterial hyper-tension related to the following: congenital systemicto pulmonary shunts; collagen vascular disease; portalhypertension; human immunodeficiency virus infec-tion; drugs and toxins (including anorexigens); andpersistent pulmonary hypertension of the newborn, isclassified with primary pulmonary hypertension aspulmonary arterial hypertension. This classificationseparates these cases of pulmonary arterial hyper-tension from pulmonary venous hypertension, pul-monary hypertension associated with disorders ofthe respiratory system and/or hypoxaemia, pulmonaryhypertension due to chronic thrombotic and/orembolic disease, as well as pulmonary hypertensiondue to disorders directly affecting the pulmonaryvasculature. This new diagnostic classification pro-vides rationale for considering many of the thera-peutic modalities that have been demonstrated to beefficacious for primary pulmonary hypertension forchildren who have pulmonary arterial hypertensionrelated to these other conditions. Because the cause(s)of primary pulmonary hypertension, as well aspulmonary arterial hypertension related to other con-ditions, remains unknown or at least incompletelyunderstood, the various treatment modalities used forpulmonary arterial hypertension have been based onthe pathology and pathobiology of the pulmonaryvascular bed. The pathology remains central to theunderstanding of the pathobiological mechanisms. Asinsight is advanced into the mechanisms responsiblefor the development of pulmonary arterial hyperten-sion, the introduction of novel therapeutic modalities(alone and in combination) will hopefully increase the

overall efficacy of therapeutic interventions for pulmo-nary arterial hypertension.

Persistent pulmonary hypertension of the newbornis a syndrome characterised by increased pulmonaryvascular resistance, right-to-left shunting and severehypoxaemia [7]. Persistent pulmonary hypertensionof the newborn is frequently associated with pulmo-nary parenchymal abnormalities including meconiumaspiration, pneumonia or sepsis, as well as occurringwhen there is pulmonary hypoplasia, maladaptationof the pulmonary vascular bed postnatally as a resultof perinatal stress or maladaptation of the pulmonaryvascular bed in utero from unknown causes. In someinstances there is no evidence of pulmonary paren-chymal disease and the "injury" that is the "trigger" ofthe pulmonary hypertension is unknown. Persistent

Table 1. – Diagnostic classification of pulmonaryhypertension

Pulmonary arterial hypertensionPrimary pulmonary hypertensionSporadicFamilial

Related to:Collagen vascular diseaseCongenital systemic-to-pulmonary shuntsPortal hypertensionHIV infectionDrugs/toxinsAnorexigensOther

Persistent pulmonary hypertension of the newbornOther

Pulmonary venous hypertensionLeft-sided atrial or ventricular heart diseaseLeft-sided valvular heart diseaseExtrinsic compression of central pulmonary veinFibrosing mediastinitisAdenopathy/tumours

Pulmonary veno-occlusive diseaseOther

Pulmonary hypertension associated with disorders of therespiratory system and/or hypoxiaChronic obstructive pulmonary diseaseInterstitial lung diseaseSleep-disordered breathingAlveolar hypoventilation disordersChronic exposure to high altitudesNeonatal lung diseaseAlveolar-capillary dysplasia

Pulmonary hypertension due to chronic thrombotic and/orembolic diseaseThromboembolic obstruction of proximal pulmonaryarteriesObstruction of distal pulmonary arteriesPulmonary embolism (thrombus, ova and/or parasites,foreign material)In situ thrombosisSickle cell disease

Pulmonary hypertension due to disorders affecting thepulmonary vasculatureInflammatorySchistosomiasisSarcoidosisOther

Pulmonary capillary haemangiomatosis

HIV: human immunodeficiency virus.

156 A. WIDLITZ, R.J. BARST

pulmonary hypertension of the newborn is almostalways transient [8], with infants either recoveringcompletely without requiring chronic medical therapyor dying during the neonatal period despite maximalcardiopulmonary therapeutic interventions. In con-trast, patients with pulmonary arterial hypertensionwho respond to medical therapy appear to needtreatment indefinitely. Some infants with persistentpulmonary hypertension of the newborn may have agenetic predisposition to hyperreact to pulmonaryvasoconstrictive "triggers" such as alveolar hypoxia. Itis possible that in some neonates the pulmonaryvascular resistance may not fall normally after birth,although the diagnosis of persistent pulmonaryhypertension of the newborn is not made during theneonatal period and subsequently pulmonary hyper-tension is diagnosed as the pulmonary vasculardisease progresses. Pathological studies examiningthe elastic pattern of the main pulmonary artery [9,10] suggest that primary pulmonary hypertension ispresent from birth in some patients, although it isacquired later in life in others.

Whether pulmonary hypertension is due to increasedflow or resistance (table 2) depends on the cause ofthe pulmonary hypertension (table 3). By definition,hyperkinetic pulmonary hypertension refers to pulmo-nary arterial hypertension from congenital systemic topulmonary communications with increased pulmo-nary blood flow, e.g. ventricular septal defect orpatent ductus arteriosus. Pulmonary venous hyper-tension is caused by disorders of left heart filling, e.g.mitral stenosis, pulmonary venous obstruction or leftventricular failure. Unless left heart obstruction ordysfunction is causing pulmonary venous hyper-tension, the pulmonary arterial wedge pressure isnormal. Pulmonary vascular disease related to con-genital heart disease (Eisenmenger9s syndrome) isthought to develop after a hyperkinetic period ofnormal pulmonary vascular resistance and increasedpulmonary blood flow. With pulmonary venoushypertension, as seen with mitral stenosis or leftventricular dysfunction, pulmonary artery pressuremay vary from one child to another with the same

elevations of pulmonary venous pressure accountedfor by differences in pulmonary arterial vasoreac-tivity. Many different congenital heart defects areassociated with an increased risk for the developmentof pulmonary vascular disease. Approximately one-third of patients with uncorrected congenital heartdisease will die from pulmonary vascular disease [11].It is not known why some children with the sameunderlying congenital heart defect develop irreversiblepulmonary vascular obstructive disease in the firstyear of life and others maintain "operable" levels ofpulmonary hypertension into the second decade andbeyond. In many children whose congenital heartdisease is diagnosed late in life, an important anddifficult decision is necessary to determine whether thepatient is "operable" or has "irreversible" pulmonaryvascular disease. In the past, this evaluation of opera-bility has used anatomical criteria based on micro-scopic findings from lung biopsies to aid in thedetermination [12]. More recently, new approachesto the evaluation of operability and perioperativemanagement have allowed for surgical "corrections"in patients who present late in life with elevatedpulmonary vascular resistance. The assessment of sur-gical operability requires an accurate determinationof the degree of pulmonary vasoreactivity or rever-sibility. It is important to predict whether the elevatedpulmonary vascular resistance will respond favour-ably to pharmacological vasodilatation. In the pastseveral years, studies with inhaled nitric oxide andintravenous epoprostenol have proven useful in thepreoperative evaluations, as well as in the treatmentof postoperative patients with elevated pulmonaryvascular resistance [13–19]. If a patient with elevatedpulmonary vascular resistance is being considered forsurgery there is an increased risk of postoperativepulmonary hypertensive crises. Thus, knowing ifthe pulmonary circulation will respond favourably toinhaled nitric oxide or intravenous prostacyclin willhelp in guiding the management of this potentiallylife-threatening complication [14, 20].

Although misalignment of the pulmonary veinswith alveolar capillary dysplasia is often diagnosedas persistent pulmonary hypertension of the newborn,it is a separate entity, i.e. a rare disorder of pulmo-nary vascular development that most often is diag-nosed only after an infant has died from fulminantpulmonary hypertension [21]. Features that will oftenalert clinicians to the possibility of alveolar capillarydysplasia include association with other nonlethalcongenital malformations, the late onset of presenta-tion (especially after 12 h) and severe hypoxaemiarefractory to medical therapy. Infants most oftenpresent with severe pulmonary arterial hypertension

Table 2. – Classification of pulmonary hypertension

Type Classification

Hyperkinetic P=F6rPulmonary vascular obstructionor pulmonary venous hypertension

P=f6R

P: increased pulmonary artery pressure; F: high pulmonaryblood flow; r: normal total pulmonary resistance; f: normalpulmonary blood flow; R: high total pulmonary resistance.

Table 3. – Causes of pulmonary hypertension

ReversibleHyperkinetic Congenital systemic-to-pulmonary shunt, e.g. VSD, PDAPulmonary venous hypertension Pulmonary venous obstruction, cor triatriatum, mitral stenosis, left ventricular failure

IrreversiblePulmonary vascular obstruction Primary pulmonary hypertension, Eisenmenger9s syndrome

VSD: ventricular septal defect; PDA: patent ductus arteriousus.

157PULMONARY ARTERIAL HYPERTENSION IN CHILDREN

with transient responses to inhaled nitric oxide,which subsequently require increases in the dose ofinhaled nitric oxide as well as transient responsesafter intravenous epoprostenol is added to the inhalednitric oxide, with virtually all infants subsequentlydying within the first several weeks of life. The onlycase with longer survival that the authors are awareof was an infant who presented at 6 months of agewith what was initially thought to be overwhelmingpneumonia requiring maximal cardiopulmonary sup-port, including extracorporeal membrane oxygena-tion. The infant was initially diagnosed as havingprimary pulmonary hypertension, although upon fur-ther review of the open lung biopsy, alveolar capillarydysplasia was diagnosed with a very heterogeneousappearance on the biopsy. The infant significantlyimproved with inhaled nitric oxide and intravenousepoprostenol while awaiting heart-lung trans-plantation; she was subsequently weaned off extra-corporeal membrane oxygenation as well as offmechanical ventilation and inhaled nitric oxide. Shehad marked clinical and haemodynamic improvementon chronic intravenous epoprostenol and continuedchronic intravenous epoprostenol until she was nearly4 yrs of age, at which time after acquiring a respi-ratory tract infection she rapidly deteriorated and died(unpublished data). Post-mortem examination con-firmed the diagnosis of alveolar capillary dysplasiawith a very heterogeneous involvement in the pulmo-nary parenchyma (consistent with the late presenta-tion as well as significant palliative response withintravenous epoprostenol). This variability in clinicalseverity and histopathology is consistent with themarked biological variability that occurs in manyforms of pulmonary arterial hypertension. Whenpulmonary hypertension results from neonatal lungdisease such as meconium aspiration, the pulmonaryvascular changes are most severe in the regions of thelung showing the greatest parenchymal damage.

Congenital heart disease is the most commoncause of pulmonary venous hypertension in childrendue to total anomalous pulmonary venous returnwith obstruction, left heart obstruction or severe leftventricular failure. The lungs of those born with leftinflow obstruction show pronounced thickening inthe walls of both the arteries and the veins; and theoutcome depends on the results of the surgicalintervention. Pulmonary veno-occlusive disease has adistinct pathological feature of uniform fibrotic occlu-sion of peripheral small venules [22]. Although rare, itdoes occur early in childhood and has been reportedin familial cases [23]. Progressive long-segmentpulmonary vein hypoplasia leading to pulmonaryvenous atresia is another uniformly fatal conditionpresenting in infancy with severe pulmonary venoushypertension.

Epidemiology

The frequency of pulmonary arterial hypertensionin children as well as in adults remains unknown.Estimates of the incidence of primary pulmonaryhypertension ranges from one to two new cases per

million people in the general population [24].Although the disease is rare, increasingly frequentreports of confirmed cases suggest that more patients(both children and adults) have pulmonary arterialhypertension than had been previously recognised. Onoccasion, infants who have died with the presumeddiagnosis of sudden infant death syndrome have hadprimary pulmonary hypertension diagnosed at thetime of post-mortem examination. The sex incidencein adult patients with primary pulmonary hyperten-sion isy1.7:1 females:males [25], similar to the currentauthors9 experience with children, 1.8:1 with no signi-ficant difference in the younger children comparedwith the older children.

Natural history

Primary pulmonary hypertension historically exhi-bited a course of relentless deterioration and earlydeath. Unfortunately, the data available for childrenis much less than for adult patients (due to thedecreased occurrence of primary pulmonary hyper-tension in children compared with adults). Severallarge survival studies of primarily adult patientswith primary pulmonary hypertension conducted inthe 1980s have provided a basis of comparison forsubsequent evolving therapeutic modalities. Theseretrospective and prospective studies have yieldedquite uniform results: adult patients with primarypulmonary hypertension who have not undergonelung or heart/lung transplantation have had actuarialsurvival rates at 1, 3 and 5 yrs of 68–77, 40–56 and22–38%, respectively [1, 26–28]. However, there issignificant biological variability in the natural historyof the disease in both adults and children, with somepatients having a rapidly progressive downhill courseresulting in death within several weeks after diagnosisas well as instances of survival for at least severaldecades.

Pathobiology

By definition, the aetiology(ies) of primary pulmo-nary hypertension are unknown, e.g. primarypulmonary hypertension is also referred to as "unex-plained" or "idiopathic" pulmonary hypertension. Inyoung children, the pathobiology suggests failure ofthe neonatal vasculature to open and a striking reduc-tion in arterial number. In older children, intimalhyperplasia and occlusive changes are found in thepulmonary arterioles as well as plexiform lesions.Despite significant advances (during the past severaldecades) in the understanding of the pathobiology ofprimary pulmonary hypertension, the mechanism(s)which initiate and perpetuate the disease processremain(s) speculative. Adults with primary pulmo-nary hypertension often have severe plexiform lesionsand what appears to be "fixed" pulmonary vascularchanges. In contrast, children with primary pulmo-nary hypertension have more pulmonary vascularmedial hypertrophy and less intimal fibrosis andfewer plexiform lesions. In the classic studies by


WAGENVOORT and WAGENVOORT [29] in 1970, medialhypertrophy was severe in patients v15 yrs of ageand it was usually the only change seen in infants.Among the 11 children v1 yr of age, all had severemedial hypertrophy, yet only three had intimal fibro-sis, two with minimal intimal fibrosis and one withmoderate intimal fibrosis and none had plexiformlesions. With increasing age, intimal fibrosis andplexiform lesions were seen more frequently. Thesepost-mortem studies suggested that pulmonary vaso-constriction, leading to medial hypertrophy, mayoccur early in the course of the disease and mayprecede the development of plexiform lesions andother fixed pulmonary vascular changes (at least insome children). Furthermore, these observations mayoffer clues to the observed differences in the naturalhistory and factors influencing survival in childrenwith primary pulmonary hypertension comparedwith adult patients. Younger children in generalappear to have a more reactive pulmonary vascularbed relative to both active pulmonary vasodilatationas well as pulmonary vasoconstriction, with severeacute pulmonary hypertensive crises occurring in res-ponse to pulmonary vasoconstrictor "triggers" moreoften than in older children or adults. Thus, basedon these early pathological studies, the most widelyproposed mechanism for primary pulmonary hyper-tension until the late 1980s and early 1990s waspulmonary vasoconstriction [5, 30–32]. Subsequentstudies have identified potentially important struc-tural and functional abnormalities; whether these per-turbations are a cause or consequence of the diseaseprocess remains to be elucidated. These abnormalities

include imbalances between vasodilator/antiproliferativeand vasoconstrictive/mitogenic mediators, defects inthe potassium channels of pulmonary artery smoothmuscle cells and increased synthesis of inflammatorymediators which cause vasoconstriction as well asenhanced cell growth (fig. 1) [33–40].

The molecular process(es) behind the complexvascular changes associated with primary pulmonaryhypertension include the following: the description ofphenotypical changes in endothelial and smoothmuscle cells in hypertensive pulmonary arteries, therecognition that cell proliferation contributes to thestructural changes associated with the initiation andprogression of primary pulmonary hypertension (aswell as apoptosis contributing to hypertensive pulmo-nary vascular disease) and the role of matrix proteinsand matrix turnover in vascular remodelling and theimportance of haemodynamic influences on the dis-ease process. It is now clear that gene expression inpulmonary vascular cells responds to environmentalfactors, growth factors, receptors, signalling pathwaysand genetic influences, which can interact with eachother. Examples of effector systems controlled bygene expression include the following: transmem-brane transporters, ion channels, transcription factors,modulators of apoptosis, kinases, cell-to-cell inter-active factors, e.g. intergrins and membrane receptors,mechanotransducers, extracellular matrix turnoverand growth factors/cytokines and chemokine net-works. By identifying causes of molecular process(es)that are linked to epidemiological risk factors, as wellas developing molecular, biochemical and physiologi-cal tests to monitor and diagnose primary pulmonary

��

��

��

��

��

��

�� !"

��

��

��

#��

$��

Fig. 1. – Schematic representation of the factors influencing the imbalance between relaxation and contraction in primary pulmonaryhypertension. ATP: adenosine triphosphate; ADP: adenosine diphosphate; EDRF: endothelium-derived relaxing factor; PGI2:prostacyclin; ET-1: endothelin-1. Reprinted from [42] with permission from Elsevier Science.


hypertension, novel treatment strategies based onestablished pathobiological mechanisms will increasethe overall efficacy of therapeutic interventions forprimary pulmonary hypertension. Although manyimportant physiological processes have been identifiedfrom descriptive studies from patients (suggestingpossible pathobiological mechanisms in the develop-ment of primary pulmonary hypertension, fig. 2),whether these observations are a cause or a conse-quence of the disease remains unclear.

The vascular endothelium, an important sourceof locally active mediators that contribute to thecontrol of vasomotor tone and structural remodelling,appears to play a crucial role in the pathogenesisof primary pulmonary hypertension [42]. A numberof studies have suggested that imbalances in theproduction or metabolism of several vasoactivemediators produced in the lungs may be importantin the pathogenesis of primary pulmonary hyper-tension. An imbalance may exist in the vasoconstrict-ing and vasodilator mediators as well as substancesinvolving control of pulmonary vascular tone. Theseinclude increased thromboxane and decreased pros-tacyclin [33, 34], increased endothelin and decreasednitric oxide [35–37, 40], as well as other vasoactivesubstances yet to be described. Thromboxane andendothelin are vasoconstrictors as well as mitogens; in

contrast, prostacyclin and nitric oxide are vasodilatorswith antiproliferative effects. Other factors may alsobe involved such as serotonin, platelet derived growthfactor, angiotensin or the loss of pulmonary vascularprostacyclin or nitric oxide synthase gene expression.Vasoconstrictors may also serve as factors or cofac-tors that stimulate smooth muscle growth or matrixelaboration. It appears likely that endothelial injuryresults in the release of chemotactic agents leading tomigration of smooth muscle cells into the vascularwall. In addition, this endothelial injury, coupled withexcessive release of vasoactive mediators locally,promotes a procoagulant state, leading to furthervascular obstruction. The process is characterised,therefore, by an inexorable cycle of endothelial dys-function leading to the release of vasoconstrictiveand vasoproliferative substances, ultimately progres-sing to vascular remodelling and progressive vascularobstruction and obliteration. In addition, defects inthe potassium channels of pulmonary resistancesmooth muscles may also be involved in the initiationand/or progression of primary pulmonary hyper-tension; inhibition of the voltage regulated (Kv)potassium channel has been reported in pulmonaryartery smooth muscle cells from patients with pri-mary pulmonary hypertension [43]. Whether a geneticdefect related to potassium channels in the lung vessels

��%��

��

��

��

'

(��)*��

��

��+��*�)+��*�

��

��,

��-��*��

��*��+��

��%��,��*��+��+��-��.

��

��,��/��

��

01��

��+��

��

(��

��

� �!�

��!��!��+��-�

��

$��*��

Fig. 2. –Possible pathogenesis of primary pulmonary hypertension. BMPR2: bone morphogentic protein receptor-2; TGF: transforminggrowth factor; PAH: pulmonary artery hypertension. Reprinted from [41] with permission from Elsevier Science.


in primary pulmonary hypertension patients lead-ing to vasoconstriction is relevant to the develop-ment of primary pulmonary hypertension in somepatients remains unknown. However, these studiessuggest that primary pulmonary hypertension is adisease of "predisposed" individuals, in whom vari-ous "stimuli" may initiate the pulmonary vasculardisease process. Whether or not vasoconstriction isthe initiating event in the pathobiology of primarypulmonary hypertension, it is an important compo-nent in the pathophysiology of the disease (in at leasta subset of patients). There may also be subsets ofpatients in whom endothelial dysfunction is theinitiating "injury" as opposed to vasoconstriction.Regardless, pulmonary vasoconstriction is an exacer-bating factor in the disease progression. Exaggeratedepisodes of pulmonary vasoconstriction in suscep-tible individuals can damage the pulmonary vascularendothelium, resulting in further alterations in thebalance between vasoactive mediators. Coagulationabnormalities may also occur, initiating or furtherexacerbating the pulmonary vascular disease [44, 45].The interactions between the humoral and cellularelements of the blood on an injured endothelial cellsurface result in remodelling of the pulmonary vas-cular bed and contribute to the process of pulmonaryvascular injury [46, 47].

Migration of smooth muscle cells in the pulmonaryarterioles occurs as a release of chemotactic agentsfrom injured pulmonary endothelial cells [48]. Endo-thelial cell damage can also result in thrombosis in situ,transforming the pulmonary vascular bed from itsusual anticoagulant state (owing to the release ofprostacyclin and plasminogen activator inhibitors) toa procoagulant state [49]. Elevated fibrinopeptide-Alevels in primary pulmonary hypertension patientsalso suggests that in situ thrombosis is occurring [50].Further support for the role of coagulation abnor-malities at the endothelial cell surface comes from thereports of improved survival in patients treated withchronic anticoagulation [2, 26, 51].

Pathophysiology

Although the histopathology in children with pri-mary pulmonary hypertension is often qualitativelythe same as that seen with adult patients, the clinicalpresentation, natural history and factors influencingsurvival may differ. These differences appear to bemost apparent in the youngest children. Children alsoappear to have differences in their hemodynamicsparameters at the time of diagnosis compared withadult patients [52]. The increased cardiac index inchildren (as opposed to adults) may reflect an earlierdiagnosis and why children tend to have a greateracute vasodilator response rate with acute vasodilatortesting than adults. The findings of higher heart ratesand lower systemic arterial pressures in children is notunexpected. In order to understand the signs andsymptoms of pulmonary arterial hypertension, onemust first briefly review normal physiology of thepulmonary circulation. The pulmonary vascular bednormally has a remarkable capacity to dilate and

recruit unused vasculature to accommodate increasesin blood flow. In pulmonary arterial hypertension,however, this capacity is lost, leading to increases inpulmonary artery pressure at rest and further eleva-tions in pulmonary artery pressure with exercise. Inresponse to this increased afterload, the right ventriclehypertrophies. Initially, the right ventricle is capableof sustaining normal cardiac output at rest, but theability to increase cardiac output with exercise isimpaired. As pulmonary vascular disease progresses,the right ventricle fails and resting cardiac outputdecreases. As right ventricular dysfunction progresses,right ventricular diastolic pressure increases andevidence of right ventricular failure, the most ominoussign of pulmonary vascular disease, manifests. Alth-ough the left ventricle is not directly affected by thepulmonary vascular disease, progressive right ventri-cular dilatation can impair left ventricular filling,leading to moderately increased left ventricular enddiastolic and pulmonary capillary wedge pressures.Dyspnoea, the most frequent presenting complaint inadults with primary pulmonary hypertension as wellas in some children with primary pulmonary hyper-tension, is due to impaired oxygen delivery duringphysical activity as a result of an inability to increasecardiac output in the presence of increased oxygendemands. Syncopal episodes, which occur more fre-quently with children than with adults, are oftenexertional or postexertional, and imply a severelylimited cardiac output, leading to diminished cerebralblood flow. Peripheral vasodilatation during physicalexertion can exacerbate this condition.

From the pathobiology and pathophysiology ofprimary pulmonary hypertension, the two mostfrequent mechanisms of death are progressive rightventricular failure and sudden death, with the formeroccurring far more often [1]. With progressive rightventricular failure, the scenario, described above,leads to dyspnoea, hypoxaemia and a progressivedecrease in cardiac output. Pneumonia may be fatalas a result of alveolar hypoxia causing furtherpulmonary vasoconstriction and an inability to main-tain adequate cardiac output, resulting in cardiogenicshock and death. When arterial hypoxaemia andacidosis occur, life-threatening arrhythmias maydevelop. Postulated mechanisms for sudden deathinclude brady- and tachyarrhythmias, acute pulmonaryemboli, massive pulmonary haemorrhage and suddenright ventricular ischaemia. Haemoptysis appears tobe due to pulmonary infarcts with secondary arterialthromboses.

Diagnosis and assessment

Although the diagnosis of primary pulmonaryhypertension is one of exclusion, it can be madewith a high degree of accuracy if care is taken toexclude all likely related or associated conditions. Athorough and detailed history and physical examin-ation, as well as appropriate tests, must be performedto uncover potential causative or contributing fac-tors, many of which may not be readily apparent.Questions should be asked about family history:


pulmonary hypertension, connective tissue disorders,congenital heart disease, other congenital anomalies,and early unexplained deaths. If the family historysuggests familial primary pulmonary hypertension,careful screening of all family members is recom-mended. It is reasonable to consider a transthoracicechocardiogram in all first degree relatives of thepatient diagnosed with primary pulmonary hyper-tension (at the time of his/her diagnosis), as well as atany time symptoms consistent with pulmonary arterialhypertension occur, and subsequently every 3–5 yrs inasymptomatic family members (per the World HealthOrganization (WHO) World Symposium on PrimaryPulmonary Hypertension recommendations 1998)[53]. Although it is reasonable to believe that initiatingtherapy in "asymptomatic" siblings with familialprimary pulmonary hypertension may improve out-come, this remains to be proven. Additional issues toaddress include a carefully detailed birth and neonatalhistory, a detailed drug history, prolonged medicationhistory including psychotropic drugs and appetitesuppressants, exposure to high altitude or to toxiccooking oil [54, 55], travel history and any history offrequent respiratory tract infections. Problems withcoagulation should also be queried. The answers tothese questions may offer clues to a possible "trigger"

or the "injury" initiating the development of thepulmonary arterial hypertension.

The diagnostic evaluation in children suspectedof having primary pulmonary hypertension is similarto that of adult patients (fig. 3), although certainconditions, e.g. chronic thromboembolic disease orobstructive sleep apnoea, seen not uncommonly inadult patients, rarely occur in children. A functionalassessment is also helpful using the modified NewYork Heart Association functional classification, i.e.WHO functional classification, for patients withpulmonary hypertension (as shown in table 4) [53].

Cardiac catheterisation acute vasodilator testing

Although noninvasive tests are useful in theevaluation of suspected pulmonary arterial hyper-tension, cardiac catheterisation confirms the diag-nosis. In adults, haemodynamic values obtained atcardiac catheterisation can also be used to predictsurvival, although this has not been validated inchildren. Cardiac output can accurately be measuredby the thermodilution technique in patients withoutsignificant shunting through a patent foramen ovale,however it remains controversial whether using the

.��+��

��+��+��+��

��-��+��-��+�

2��-��,��+��+��3�+��

�+��+��

��+��

��-��+�3�$��3�4.5��

2��-��,��+��+��3�+��+��+��

�-��

��-��+��+��+��+��

��+��-��+��+��-��-��-��--��'

��,��+��,��+��-��-��*��

��,��*��

6��2� ��-��+�

��--��

��)��--��-��+��

��+��'

��,��+��*��*��-��*��3��

��)��,��,�+��+��

+��-��+�'

��,��+��3�$�3�+��-�-��

��*��+��+��3��+��-��+��

�*��+!��-��-��

6��

6��

6��

2�

2�

2�

Fig. 3. –Diagnostic evaluation in children suspected of having primary pulmonary hypertension. CPET: cardiopulmonary exercise testing;WHO: World Health Organization; V9/Q9: ventilation perfusion ratio; CT: computed tomography.


thermodilution technique is accurate in the face ofsignificant pulmonary or tricuspid regurgitation. TheFick method is used with measured oxygen consump-tion if there is any concern regarding pulmonary ortricuspid regurgitation as well as if there is a patentforamen ovale. Because of the increased risk ofcardiac catheterisation in patients with severe pulmo-nary vascular disease, special precautions should betaken during cardiac catheterisation, particularly withchildren who may have a more reactive pulmonary

vascular bed than adult patients, i.e. prone to acutepulmonary hypertensive crises: adequate sedation tominimise anxiety without depressing respiration andprevention of hypovolaemia and hypoxaemia areimportant issues that need to be addressed. Asshown in the treatment algorithm guideline (fig. 4),it was recommended that all children undergo acutetesting at the time of the initial right heart catheter-isation with a short-acting vasodilator to determinevasodilator responsiveness. Unfortunately, there are

Table 4. – World Health Organization class

Class I Patients with pulmonary hypertension but without resulting limitations of physical activityOrdinary physical activity does not cause undue dyspnoea or fatigue, chest pain or near syncope

Class II Patients with pulmonary hypertension resulting in slight limitation of physical activityThey are comfortable at restOrdinary physical activity causes undue dyspnoea or fatigue, chest pain or near syncope

Class III Patients with pulmonary hypertension resulting in marked limitation of physical activityThey are comfortable at restLess than ordinary activity causes undue dyspnoea or fatigue, chest pain or near syncope

Class IV Patients with pulmonary hypertension with inability to carry out any physical activity without symptomsThese patients manifest signs of right heart failureDyspnoea and/or fatigue may even be present at restDiscomfort is increased by any physical activity

��+��

�*��+��*��*�)��*��-7

��

��*��+�+��

��*��

��+��8 2��+��

�-��

��!�� -��(��+�+��

�� (

��

9��-��)��-��+��

��+��8 2��+��

��-��3��(�+�+��

��-��

��

9��-��)��-��+��

��-��3��

��!��

��

��-��

Fig. 4. –Treatment algorithm for primary pulmonary hypertension prior to novel therapeutics. IV: intravenous. #: responder to acutevasodilator testing defined as a significant fall in mean pulmonary arterial pressure, e.g. o20%, without a fall in the cardiac output.


no haemodynamic or demographic variables thataccurately predict whether or not a child will respondto acute vasodilator testing. Although the younger thechild is at the time of diagnosis, the more likely it isthat the child will respond to acute testing, there is awide spectrum of variability (fig. 5) [56]. Children withsymptoms for several years suggestive of severepulmonary vascular compromise may manifest nearcomplete reversibility with chronic oral calciumchannel blockade therapy (as discussed below), whileothers with a brief duration of symptoms may havewhat appears to be irreversible disease. These obser-vations underscore the marked biological variabilityin the time course of primary pulmonary hypertensionand serve to emphasise the need to individualise thetherapeutic approach for each child. The followingvasodilators are recommended for acute vasodilatordrug testing: intravenous epoprostenol sodium, doserange 2–12 ng?kg-1?min-1 (although higher dosesmay be needed with children for acute vasodilatordrug testing compared with adult patients), half-life2–3 min; inhaled nitric oxide, dose range 10–80 partsper million, half-life 15–30 s; and/or intravenous adeno-sine 50–200 ng?kg-1?min-1, half-life 5–10 s; inhalediloprost, dose range 14–17 mg, half-life of 20–30 min.Patients who are responsive to acute vasodilatortesting are likely (although not always) to have afavourable response with acute calcium channelblocker testing and subsequently chronic oral calciumchannel blockade therapy [2, 56]. Although there isno consensus regarding the definition of an acutevasodilator response, an acceptable response would bea significant reduction in mean pulmonary arterypressure, e.g. at least 20%, with no change or anincrease in cardiac output. Patients who do notmanifest a response to acute vasodilator challengeare unlikely to have clinical benefit from chronic oralcalcium channel blockade therapy. Furthermore,acute deterioration and decompensation may occur

with empiric calcium channel blockade therapy [57,58].

Clinical presentation

The presenting symptoms in children with primarypulmonary hypertension may differ when comparedwith adult patients. Infants with primary pulmonaryhypertension often present with signs of low cardiacoutput, e.g. poor appetite, failure to thrive, lethargy,diaphoresis, tachypnea, tachycardia and irritability. Inaddition, infants and older children may be cyanoticwith exertion, owing to right-to-left shunting througha patent foramen ovale. Children without adequateshunting through a patent foramen ovale can alsopresent with syncope, which is more often effort-related in children than in adult patients. After earlychildhood, children present with similar symptoms asadults. In the older children, the most commonsymptoms are exertional dyspnoea and, occasionally,chest pain. Clinical right ventricular failure is rare inyoung children, occurring most often in childrenw10 yrs of age with severe long-standing primarypulmonary hypertension. The interval between onsetof symptoms and time of diagnosis is usually shorterin children than in adults, particularly in thosechildren who present with syncope. Some infantshave crying spells, presumably as a result of chest painthat cannot be otherwise verbalised. In addition toexertional or postexertional syncope, children also notuncommonly present with seizures resulting fromexertional syncope as well as exaggerated pulmonaryvasoconstriction with mild systemic arterial oxygendesaturation during sleep (particularly in the earlymorning hours). Children of all ages also commonlycomplain of nausea and vomiting (reflecting lowcardiac output). Chest pain or angina results fromright ventricular ischaemia and is often underappre-ciated. Peripheral oedema is generally a reflection ofright ventricular failure and is more likely to beassociated with advanced pulmonary vascular disease.

Physical examination

Many of the physical findings in children are typicalof any patient, child or adult with pulmonary arterialhypertension, whether or not the pulmonary arterialhypertension is primary or related to other conditions.An increased pulmonic component of the second heartsound is usually audible; however, a right-sided fourthheart sound is not heard as often in children as it is inadults. Children may have distortion of their chestwall, secondary to severe right ventricular hypertro-phy. Tricuspid regurgitation is very common, whereaspulmonary insufficiency is heard less often. Clinicalsigns of right-sided heart failure, e.g. hepatomegaly,peripheral oedema and acrocyanosis are rare in youngchildren, particularly in those who present withsyncope. Clubbing is not a typical feature of primarypulmonary hypertension, although on rare occasionsit has been observed in patients with long-standing

:;

;

?;

@;

�;

";

;

��+��A

;B� �B? ?B= =B: :B"; ";B"� "�B"? "?B"= ��-��-��+��

Fig. 5. –Primary pulmonary hypertension: response to acute vaso-dilator drug testing by age. The younger the child at the time oftesting, the greater the likelihood of eliciting acute pulmonaryvasodilatation (pv0.005). Reproduced with permission fromLippincott Williams and Wilkins [56].


disease who develop chronic hypoxaemia secondary toright-to-left shunting via a patent foramen ovale.

Treatment

Although there is no cure for primary pulmonaryhypertension, nor is there a single therapeuticapproach that is uniformly successful, therapy hasimproved dramatically over the past several decades,resulting in sustained clinical and haemodynamicimprovement, as well as increased survival in childrenwith primary pulmonary hypertension [56]. An over-view of the current approach (in 2001) to treatment isshown in figure 4. Noninvasive studies obtained priorto initiating therapy, as well as periodically thereafter,may be useful in guiding changes in therapeuticregimens, particularly in light of recent advanceswith various novel therapeutic agents. Serial follow-upis necessary to maintain an "optimal" therapeuticregimen for children based on overall risk/benefitconsiderations.

General measures

The paediatrician plays an invaluable role in thecare of children with pulmonary arterial hypertension.Since children often have a more reactive pulmonaryvascular bed than adult patients, any respiratory tractinfection that results in ventilation/perfusion mis-matching from alveolar hypoxia can result in a cata-strophic event if not treated aggressively. Annualinfluenza vaccination as well as pneumococcal pneu-monia vaccination are recommended unless there arecontraindications. The current authors recommendthat children with pneumonia be hospitalised for theinitiation of antibiotic therapy, with antipyretics admi-nistered for temperature elevationsw38uC (101uF) tominimise the consequences of increased metabolicdemands. Children may also require aggressive the-rapy for acute pulmonary hypertensive crises occur-ring with episodes of pneumonia or other infectiousdiseases. The present authors have seen the adultrespiratory distress syndrome occur not uncommonlyin children with pulmonary hypertension followingviral infections, requiring aggressive cardiopulmonarysupport including inhaled nitric oxide, intravenousepoprostenol and, on rare occasion, extracorporealmembrane oxygenation, in addition to conventionalmaximal cardiopulmonary support to recover. Dietand/or medical therapy should be used to preventconstipation, since Valsalva manoeuvres transientlydecrease venous return to the right side of the heartand may precipitate syncopal episodes.

Anticoagulation

Consideration for chronic anticoagulation inchildren with primary pulmonary hypertension isbased on studies in adult primary pulmonary hyper-tension patients [2, 26, 51]: histopathology demon-strating thrombotic lesions in small pulmonary

arteries in the majority of adult patients with primarypulmonary hypertension and biochemical data con-sistent with a hypercoagulable state in some patients.The rationale for anticoagulation, to prevent second-ary thrombosis in situ from occurring which canexacerbate the underlying pulmonary vascular disease,is certainly logical in states with low pulmonaryblood flow, e.g. right heart failure. Whether or notsecondary thrombosis in situ is a significant exacer-bating factor in patients with a normal resting car-diac output is unknown. Clinical data supporting thechronic use of anticoagulation is limited but sup-portive. Warfarin has been shown to be associatedwith improved survival in two retrospective studiesand one prospective study; all three were with adultpatients only [2, 26, 51]. The optimal dose of warfarinin these studies was not determined, although therange of anticoagulation that is usually recommendedis to achieve an international normalised ratio (INR)of 1.5–2; however, certain clinical circumstances, e.g.positive lupus anticoagulant, positive anticardiolipinantibodies, Factor V Leiden and/or Factor II 20210Avariant [59–62], as well as documented chronic thrombo-embolic disease, may require dose adjustment to main-tain a higher INR, as well as dose adjustment tomaintain a lower INR for patients at a higher riskof bleeding, e.g. patients with significant thrombo-cytopenia. Whether or not chronic anticoagulation isefficacious as well as safe with a low risk/benefit profilefor children with pulmonary hypertension remains tobe determined. The current authors9 approach hasbeen to anticoagulate children with right heart failure,as with adult patients. In children who are extremelyactive, particularly toddlers, unless there is severepulmonary vascular disease, the authors recommendmaintaining an INRv1.5 and in some situations mini-dose oral anticoagulation with warfarin that may notsignificantly increase the INR from baseline at all.Antiplatelet therapy with aspirin or dipyridamole doesnot appear to be effective in areas of low flow wherethrombosis in situ is known to occur. Parents shouldbe advised to avoid administering other medicationsthat could interact with the warfarin unless the drug/drug interactions are known and the dose of thewarfarin is adjusted as needed. Frequent adjustmentsin warfarin dosage may be necessary when right-sidedheart failure is present; heparin may be preferred inthese children. As in the authors9 approach with adultpulmonary hypertension patients, if anticoagulationwith warfarin is contraindicated or dose adjustmentsare difficult, either intravenous heparin to achieve anactivated partial thromboplastin time of 1.3–1.5 timescontrol or low molecular weight heparin at a dose of0.75–1 mg?kg-1 by subcutaneous administration onceor twice daily may be reasonable alternatives,although the long-term side effects of heparin, suchas osteopenia and thrombocytopenia, may be trouble-some. To date there are no studies comparing thesafety and efficacy of anticoagulation with warfarinand heparin; however experimental studies byTHOMPSON et al. [63] in vivo and BENITZ et al. [64]in vitro suggest that vascular smooth muscle growthinhibitors, such as heparin, may be useful in pre-venting pulmonary vascular disease.


Calcium channel blockade

Calcium channel blockers are a chemically hetero-geneous group of compounds that inhibit calciuminflux through the slow channel into cardiac andsmooth muscle cells. Their usefulness in pulmonaryhypertension is believed to be based on the ability tocause vasodilatation of pulmonary vascular smoothmuscle. The rationale for this is based on histo-pathological studies that demonstrated an apparent

progression in vascular remodelling from medialhypertrophy to plexiform arteriopathy [29], althoughthis is now controversial with current speculationthat there may be subsets of patients with a vaso-constrictive component initiating the process in somepatients and a proliferative initiation in other pati-ents. The recent discovery of the primary pulmonaryhypertension-1 (bone morphogenic protein receptor-2)(fig. 6) [65, 66] gene also suggests that primarypulmonary hypertension is a proliferative disorder

��

'

''

'

'

��C�?��:

��?

9�#0"��

��+��

*��

��.

Fig. 6. – Signalling pathways of the bone morphogenetic protein receptor (BMPR)2. In the extracellular space, bone morphogenticproteins (BMPs) bind directly to the BMPR2 on the cell membrane. The bioavailability of BMPs is regulated by the presence of BMPR2receptor antagonists such as noggins, chordins and differential screening-selected gene (DAN). The binding of ligands to BMPR2 leads tothe recruitment of BMPR1 to form a heteromeric receptor complex at the cell surface. This results in the phosphorylation and activationof the kinase domain of BMPR1. The activated BMPR1 subsequently phosphorylates and activates cytoplasmic signalling proteinsSmads. Phosphorylated Smads bind to the common mediator Smad 4 and the resulting Smad complex moves from the cytoplasm into thenucleus and regulates gene transcription. Other downstream signalling pathways that can be activated following the engagement ofBMPR1 and BMPR2 by BMPs include cell-type dependent activation of p38 mitogen activated protein kinase (p38 MAPK) and proteinkinase A (PKA). In addition, the cytoplasmic tail of BMPR2 has been shown to interact with the LIM motif-containing protein kinase 1(LIMK1) that is localised in the cytoskeleton. Germline mutations of the gene encoding BMPR2 underlie primary pulmonaryhypertension (PPH) that is characterised by the abnormal proliferation of pulmonary vascular cells. However, the specific cytoplasmicproteins and nuclear transcription factors that are involved in the development of PPH have not been identified. The molecularmechanism of how defects of BMPR2 lead to primary pulmonary hypertension remains to be elucidated.


and that there may be a role for both pulmonary vaso-dilators and antiproliferative agents in the treatmentof primary pulmonary hypertension. Regardless, sincemedial hypertrophy is presumably associated withvasoconstriction and plexiform lesions are associatedwith fixed pulmonary vascular disease if a significantportion of the pulmonary vascular bed has a reactivecomponent, whether or not this is early in the courseof the disease (when fewer vessels are affected with"fixed" obstruction), chronic vasodilator therapy withan oral agent such as calcium channel blockers shouldbe efficacious. In the 1970s, the advent of vasodilatordrugs to treat systemic hypertension led to vasodilatortherapy for primary pulmonary hypertension.

Although only a minority (y20%) of adult patientswith primary pulmonary hypertension will respondwith chronic oral calcium channel blockade [2], e.g.documented by improvement in symptoms, exercisetolerance, haemodynamics (via reduction in pulmo-nary artery pressure and an increase in cardiac output)and survival, a significantly greater percentage ofchildren are acute responders (40%) and can beeffectively treated with chronic oral calcium channelblockade [56]. In addition, although most studies haveused calcium channel blockers at relatively high doses,e.g. long-acting nifedipine 120–240 mg daily or amlo-dipine 20–40 mg daily (for children as well as adults,i.e. children tolerating and appearing to need a highermg?kg-1 dose than adults), the optimal dosing forpatients, both children and adults with primary pulmo-nary hypertension, is uncertain. Children with noevidence of an acute haemodynamic response (duringacute vasodilator testing including testing with cal-cium channel blockers) are unlikely to benefit fromchronic therapy. Because of the frequent reportingof significant adverse effects with calcium channelblockade in these nonresponders, including systemichypotension, pulmonary oedema, right ventricularfailure and death, it is not recommended that calciumchannel blockers be used in patients in whom acuteeffectiveness has not been demonstrated.

Serial re-evaluations

Serial re-evaluations, including repeat acute vaso-dilator testing, to maintain an optimal chronictherapeutic regimen is essential to the care of childrenwith primary pulmonary hypertension. In the presentauthors9 experience, acute responders who are treatedwith chronic oral calcium channel blockade therapycontinue to do exceedingly well as long as they remainacutely reactive with vasodilator testing at repeatcardiac catheterisation. In contrast, children whohave initially been acute responders and treated withchronic calcium channel blockade, who no longerdemonstrate active vasoreactivity on repeat testing,more often than not subsequently deteriorate clini-cally and haemodynamically despite continuation ofchronic calcium channel blockade therapy; they arethen switched to chronic intravenous epoprostenolwith improvement with chronic epoprostenol therapysimilar to that seen with children who are initiated

on epoprostenol because they are nonresponders attheir initial evaluation.

Prostaglandins

The use of prostacyclin (epoprostenol) or a pros-tacyclin analogue for the treatment of primarypulmonary hypertension is supported by the demon-stration of an imbalance in the thromboxane toprostacyclin metabolites in patients with primarypulmonary hypertension [34], as well as the demon-stration of a reduction in prostacyclin synthase in thepulmonary arteries of primary pulmonary hyper-tension patients [39]. Although chronic intravenousepoprostenol does improve exercise tolerance, haemo-dynamics or survival in patients with primarypulmonary hypertension, its mechanism(s) of actionremains unclear [3, 4, 67–69]. In addition to loweringpulmonary artery pressure, increasing cardiac outputand increasing oxygen transport, based on studiesdemonstrating that the lack of an acute response toepoprostenol does not preclude a chronic beneficialresponse, an effect on reversing pulmonary vascularremodelling with chronic intravenous epoprostenol israised [56]. The development of tolerance to the effectsof intravenous epoprostenol remains unknown; somechildren appear to need periodic dose escalation,which may represent tolerance and/or disease progres-sion. In addition, the optimal dose of intravenousepoprostenol is also unclear. The starting dose, similarto that in adult patients, is 2 ng?kg-1?min-1 withincremental increases most rapid during the firstfew months after epoprostenol is initiated. Althoughthe mean dose in adult patients at 1 yr is y20–40 ng?kg-1?min-1, the mean dose at 1 yr in children,particularly young children, is closer to 50–80 ng?kg-1?min-1, with a significant patient variability withregards to the optimal dose.

Prostacyclin analogues

Prostacyclin analogues (oral, inhaled, subcuta-neous) were clinically developed to provided a lessinvasive treatment alternative for patients withpulmonary arterial hypertension, with continuousintravenous epoprostenol having significant adverseevents, including serious adverse events due to thedelivery system, e.g. sepsis, paradoxidal emboli as wellas temporary interruption of the epoprostenol whichcan result in catastrophic pulmonary hypertensivecrises.

Oral prostacyclin analogue

Beraprost sodium is an oral prostacyclin analogue.It is chemically stable and an orally active prosta-glandin I2 derivative with a substantially longer half-life than epoprostenol (half-life: 1.11¡0.1 h; peakplasma level is obtained within 2 h; time to maximalconcentration: 1.42¡0.15 h). The drug has similarproperties to epoprostenol (prostacyclin), e.g. increases


red blood cell flexibility, decreases blood viscosity,inhibits platelet aggregation, produces vasodilatationand decreases platelet adherence to the endothelium aswell as disaggregating platelets that have alreadyclumped. Its potency is y50% of epoprostenol. Pre-liminary data from Japan suggests that beraprost isefficacious in improving haemodynamics and survivalin primary pulmonary hypertension patients [70, 71].Although these studies were open-label, uncontrolledclinical trials, they provided proof of concept for furtherclinical investigation. More recently, a 12-week double-blind, randomised, placebo-controlled trial in Europedemonstrated improved exercise capacity and qualityof life in adults with pulmonary arterial hypertension,although a longer study (1 yr only) demonstratedtransient improvement in exercise capacity at 3 and 6months that disappeared by 9 and 12 months.

Inhaled prostacyclin analogue

Although inhaled iloprost, a stable syntheticanalogue of prostacyclin, has been shown to beefficacious in improving haemodynamics and exercisecapacity in pulmonary arterial hypertension patients[72–76], the clinical experience with children isextremely limited. It has a similar molecular structureto prostacyclin and works though prostacyclin recep-tors present on vascular endothelial cells. Iloprost ismore stable than epoprostenol and can be stored atroom temperature without needing protection fromlight [74]. It has a biological half-life of 20–30 min[77]. The acute effects of inhaled nitric oxide withaerosolised iloprost have been evaluated in childrenwith pulmonary hypertension associated with con-genital heart defects and it was demonstrated thatboth agents appear to be equally efficacious [78]. Thisopens the door for consideration of using aerosolisediloprost to treat children with various forms ofpulmonary arterial hypertension including primarypulmonary hypertension.

Subcutaneous prostacyclin analogue

Treprostinil sodium is a chemically stable prosta-cyclin analogue which also shares the same pharma-cological actions as epoprostenol. Previous studieshave demonstrated similar acute haemodynamiceffects to intravenous epoprostenol in patients withprimary pulmonary hypertension, however, trepros-tinil is stable at room temperature and neutral pHand has a longer half-life of y3 h when deliveredsubcutaneously [79]. Similar to epoprostenol, trepros-tinil is infused continuously, through a portablepump, however, it is administered subcutaneously.Double-blind, randomised, controlled trials demon-strated improved exercise capacity, clinical signs andsymptoms as well as haemodynamics in patientswith pulmonary arterial hypertension includingchildren with pulmonary arterial hypertension [80].With regard to adverse effects, the risk of centralvenous line infection is eliminated when treprostinil isused subcutaneously. Although no serious adverse

events related to treprostinil or its delivery systemhave been reported, there are side-effects with thistherapeutic modality; the most common side-effectattributed to chronic subcutaneous treprostinil the-rapy is discomfort at the infusion site [81].

Endothelin receptor antagonists

Endothelin-1, one of the most potent vasoconstric-tors identified to date [82], has been implicated in thepathobiology of pulmonary arterial hypertension,thus providing the rationale for endothelin receptorantagonists as promising drugs for the treatment ofpulmonary arterial hypertension. Plasma endothelin-1levels are known to be increased in patients withprimary pulmonary hypertension and correlate inver-sely with prognosis [83]. There are at least twodifferent receptor subtypes: endothelin (ET)A recep-tors are localised on smooth muscle cells and mediatevasoconstriction and proliferation while ETB recep-tors are found predominantly on the endothelin cellsand are associated with endothelium dependentvasorelaxation through the release of vasodilators,e.g. prostacyclin and nitric oxide, but are alsoresponsible for the clearance of endothelin, as wellas initiating vasoconstriction (on smooth muscle cells)and bronchoconstriction (fig. 7). The oral dualendothelin receptor antagonist bosentan has beenshown to improve exercise capacity, quality of life, aswell as cardiopulmonary haemodynamics in patientswith pulmonary arterial hypertension [85, 86].Although these studies were primarily carried out inadult patients, the present authors anticipate thatbosentan will also be efficacious in children. Clinicalinvestigation is also underway, evaluating whether aselective ETA receptor antagonist such as sitaxsentanwill be more or less efficacious than a dual endothelinreceptor antagonist. In an open-label, uncontrolledstudy of 20 patients including children, the selectiveETA receptor antagonist sitaxsentan improved exer-cise capacity and haemodynamics [87]. A 12-week,double-blind, placebo-controlled trial has recentlybeen completed (results pending).

9��

(��

��!"

�$� &-!��!"

��&

��!" 25� ��

��

��&

��

(��

(��+��

Fig. 7. –Endothelin system in vascular disease. ET: endothelin;Big-ET-1: proendothelin-1; ECE: endothelin-converting enzyme;NO: nitric oxide; PGI2: prostacycin. Reproduced with permissionfrom Elsevier Science [84].


Both ETA and ETB receptors play a fundamentalrole in pulmonary vasoconstriction, inflammation,proliferation of smooth muscle cells, fibrosis andbronchoconstriction. In addition, because ETB recep-tors may be upregulated on smooth muscle cells incertain pathological conditions, as well as crosstalkoccurring between ETA and ETB receptors, whetherdual endothelin receptor blockade or selective ETAreceptor blockade will prove more efficacious remainsto be elucidated. Adverse events associated withendothelin receptor antagonists include an increasein liver function, which may require a decrease in doseor discontinuation of therapy, teratogenicity as wellas possibly irreversible male infertility. Furthermore,whether the addition of endothelin receptor antagonists

to chronic oral calcium channel blockade therapy,continuous intravenous epoprostenol or a prostacy-clin analogue, will increase the overall efficacy andrisk/benefit profile for treating children with pulmo-nary arterial hypertension remains to be elucidated.

Nitric oxide

Nitric oxide is an inhaled vasodilator that exertsselective effects on the pulmonary circulation (fig. 8)[89]. Nitric oxide activates guanylate cyclase inpulmonary vascular smooth muscle cells, whichincreases cyclic guanosine monophosphate (GMP)and decreases intracellular calcium concentration,

�� #� ��

��

25

5�

5�

5�5�

5�

5�

25 25

25

25

25 25 25 25

��*��

2515�

��

��

��

�$. #

.�5�

1 25��.�25�.�25�!�/�25@!

9�

9!��-��251��C25�D

Fig. 8. –Nitric oxide (NO) is endogenously formed from L-arginine after acetylcholine (ACh) binds to muscarinic receptors (M) on theintact endothelium and stimulates endothelial nitric oxide synthase (NOS). NO exists as a gas and can be delivered to alveoli. It diffusesacross alveolar membrane to closely adjacent vessels. Constricted pulmonary vessels dilate as the result of increased intracellular cyclicguanosine monophosphate (cGMP) production in smooth muscle cells. NO is immediately inactivated by haemoglobin with the formationof methaemoglobin (met Hb), nitrosylhaemoglobin (NO Hb), nitrates (NO3

-) and nitrites (NO2-) limiting its activity to the pulmonary

circulation. GTP: guanosine triphosphate; O2: oxygen; PA: pulmonary artery; LA: left atrium. Reprinted from [88] with permission fromElsevier Science.


thereby leading to smooth muscle relaxation. Wheninhaled, the rapid combination of nitric oxide andhaemoglobin inactivates any nitric oxide diffusing intothe blood, preventing systemic vasodilatation. Nitricoxide is therefore a potent and selective pulmonaryvasodilator when administered by inhalation. Inhalednitric oxide has been demonstrated to be safe andefficacious in the treatment of persistent pulmonaryhypertension of the newborn [90, 91]. Despite thedifferences between primary pulmonary hypertensionand persistent pulmonary hypertension in the new-born, these studies suggest that inhaled nitric oxidemay be useful in the treatment of primary pulmonaryhypertension as well as other forms of pulmonaryarterial hypertension. Although there is considerableexperience with the use of inhaled nitric oxide as ashort-term treatment for pulmonary arterial hyper-tension in a variety of clinical situations, the role ofinhaled nitric oxide as a chronic therapy for pulmo-nary arterial hypertension remains under clinicalinvestigation [92]. Analogous to the suggestion thatthe improved survival in some children on long-termepoprostenol therapy is unrelated to its acute effectsand may be related to antiproliferative effects onsmooth muscle and/or anti-aggregatory effects onplatelets, nitric oxide may also have antiproliferativeeffects on smooth muscle and inhibit platelet adhe-sion. It is possible that in suitable children, a low-doseinhalation of nitric oxide may become a useful agentin the treatment of pulmonary arterial hyperten-sion, both by reversing the vasoconstrictor componentwhen present as well as encouraging the regression ofremodelling which obliterates the pulmonary vascularbed. Employing such therapeutic strategies for nitricoxide as well as epoprostenol challenges the presumedirreversibility and lethality of pulmonary arterialhypertension. This seems particularly true for infantswith pulmonary arterial hypertension who have sub-stantial capacity for smooth muscle regression, alveolargrowth and angiogenesis. Despite this rationale, clinicaltrials are needed to evaluate the safety and potentialtoxicity, as well as efficacy of chronic inhaled nitricoxide.

Phosphodiesterase inhibitors

Additional investigations evaluating nitric oxidepotentiating compounds, such as type-5 phosphodies-terase inhibitors, e.g. sildenafil, are also being per-formed [93, 94]. These strategies are aimed at raisingcyclic GMP to potentiate pulmonary vasodilatation inresponse to nitric oxide, as well as to facilitatewithdrawal of nitric oxide. Phosphodiesterase type-5inhibitors may be particularly beneficial in conjunc-tion with chronic inhaled nitric oxide, where inad-vertent withdrawal of nitric oxide may otherwise leadto a precipitous rise in pulmonary arterial pressure.Intravenous, long acting oral or aerosolised formsof phosphodiesterase type-5 inhibitors all have ther-apeutic appeal. Carefully designed studies of its pos-sible therapeutic use alone, and in combination withinhaled nitric oxide, appear warranted. Preliminary

safety and efficacy trials are underway with due con-cern for potential toxicity in the paediatric population.

Elastase inhibitors

COWAN and co-workers [95, 96] have suggested thatincreased activity of an elastolytic enzyme may beimportant in the pathophysiology of pulmonaryvascular disease. A cause and effect relationshipbetween elastase and pulmonary vascular disease issupported by studies reporting the reversal ofadvanced pulmonary vascular disease induced bymonocrotaline in rats. Part of the novelty of theseobservations is the complete regression of patho-physiological findings in the treatment group, eventhough treatment began after the disease was faradvanced. The presumed mechanism of an apoptoticcascade with regression of smooth muscle hyper-trophy and neointimal proliferation has excitingtherapeutic implications and human trials are beingconsidered. However, one must bear in mind that themonocrotaline model does not exhibit identicalvascular pathology to the human condition androdent models of successful treatment of vasculardisease have not always borne similar success inanalogous human diseases. Moreover, there was noapparent growth of new blood vessels despite thevascular recovery, possibly limiting angiogenesispotential. However, what seems clear from thesestudies is that the presumed irreversibility and fatalityof progressive pulmonary arterial hypertension nowneeds further assessment.

Gene therapy

With the identification of a primary pulmonaryhypertension gene in selected families, attention hasfocused on gene replacement therapy linked to themutated 2q33 chromosome. An alternative therapeu-tic approach has been to induce the overexpression ofvasodilator genes, notably endothelial nitric oxidesynthase and prostacyclin synthase. Patients withprimary pulmonary hypertension have been shownto have deficiencies in prostacyclin synthase, nitricoxide production as well as other endothelial func-tions. Given the preliminary clinical observations thatexogenous administration of chronic nitric oxide orepoprostenol may have salutary effects on evenadvanced forms of the disease, it seems meritoriousto pursue these alternative modes of drug delivery.Coupled with advances in the understanding of thegenetic predisposition associated with at least some, ifnot all, patients with pulmonary arterial hypertension,it seems likely that an important gene deliverytreatment for this disease may be close at hand.

Oxygen

Some children, who remain fully saturated whileawake, demonstrate modest systemic arterial oxygendesaturation with sleep, which appears to be due to


mild hypoventilation [97]. During these episodes,children may experience severe dyspnoea, as well assyncope with or without hypoxic seizures. Desatura-tion during sleep usually occurs during the earlymorning hours and can be eliminated by usingsupplemental oxygen. The current authors also recom-mend that children have supplemental oxygen avai-lable at home for emergency use, even if they do notuse it on a routine basis. Children should also betreated with supplemental oxygen during any signifi-cant upper respiratory tract infections if systemicarterial oxygen desaturation occurs, even if the child istreated at home with oral antibiotics. If more thanmild oxygen desaturation occurs, the child should betreated in a hospital setting. Children with desatura-tion due to right-to-left shunting through a patentforamen ovale usually do not improve their oxygensaturation with supplemental oxygen. Although a smallstudy of children with Eisenmenger9s syndrome demon-strated improved long-term survival with supplementaloxygen [98], a more recent study with 23 adult patientswith Eisenmenger9s syndrome did not observe anysignificant improvement in survival with nocturnaloxygen [99]. Whether children with Eisenmenger9s syn-drome benefit from treatment with nocturnal oxygenremains unclear. Supplemental oxygen may also reducethe degree of polycythaemia in children with signifi-cant right-to-left shunting via a patent foramen ovale.Similar to the experience with adult patients, somechildren experience arterial blood oxygen desaturationwith exercise as a result of increased oxygen extractionin the face of fixed oxygen delivery. These childrenmay benefit from ambulatory supplemental oxygen.In addition, children with severe right ventricular fail-ure and resting hypoxaemia, resulting from a markedlyincreased oxygen extraction, should also be treatedwith continuous oxygen therapy.

Additional pharmacotherapy: cardiac glycosides,diuretics, antiarrhythmic therapy, inotropic agentsand nitrates

Although controversy persists regarding the valueof digitalis in primary pulmonary hypertension [100],the present authors believe that children with right-sided heart failure may benefit from digitalis, inaddition to diuretic therapy. Diuretic therapy mustbe initiated cautiously, since these patients appear tobe extremely dependent on preload to maintain opti-mal cardiac output. Despite this, relatively high dosesof diuretic therapy are commonly needed.

Although malignant arrhythmias are rare in pulmo-nary hypertension, they require treatment if docu-mented. Atrial flutter or fibrillation often precipitatesan abrupt decrease in cardiac output and clinicaldeterioration once atrial systole is lost. As opposed tohealthy children, in whom atrial systole is responsiblefor y25% of the cardiac output, atrial systole inchildren with primary pulmonary hypertension oftencontributes as much as 70% of the cardiac output.Therefore, aggressive treatment of atrial flutter offibrillation is advised. The treatment of children with

clinically significant supraventricular tachycardias aswell as frequent episodes of nonsustained ventriculartachycardia and complex ventricular arrhythmias isrecommended, but it should probably be avoided forthe treatment of lesser grades of arrhythmia.

There are no studies on the usefulness of inter-mittent or continuous treatment with inotropic agents.The present authors occasionally add dobutamine foradditional inotropic support to continuous intra-venous epoprostenol for a child with severe rightventricular dysfunction who is awaiting transplanta-tion. Children have occasionally also benefited fromshort-term inotropic support during acute pulmonaryhypertensive crises to augment cardiac output duringa period of increased metabolic demands.

Oral and sublingual nitrates have been used totreat some children with primary pulmonary hyper-tension, although the experience with these agentsremains somewhat limited. Children who complain ofchest tightness, or pressure, or vague discomfort thatis responsive to sublingual nitroglycerin should betreated with chronic oral nitrates as well as sublingualnitroglycerin.

Atrial septostomy

Children with recurrent syncope or severe rightheart failure have a very poor prognosis [1, 101].Exercise-induced syncope is due to systemic vaso-dilatation, with an inability to augment cardiac outputto maintain cerebral perfusion pressure. Childrenwith pulmonary arterial hypertension and recurrentsyncope are unable to adequately shunt through apatent foramen ovale [102]. If right-to-left shuntingthrough an interatrial communication is present,cardiac output can be maintained or increased ifnecessary. In addition, right-to-left shunting at theatrial level alleviates signs and symptoms of rightheart failure by decompression of the right atrium andright ventricle. Increased survival has been reported inprimary pulmonary hypertension patients with apatent foramen ovale, although there has been somecontroversy regarding this [103]. Patency of the fora-men ovale may improve survival if it allows sufficientright-to-left shunting to occur to maintain cardiacoutput, as is evidenced by systemic arterial oxygendesaturation at rest and/or during exercise. Althoughthere is a worldwide experience inw100 patients, theprocedure should still be considered investigational.Successful palliation of symptoms with atrial septos-tomy has been reported in several series [104–108].In the current authors9 experience, patients withpulmonary arterial hypertension with recurrent syn-cope and/or right heart failure significantly improveclinically as well as haemodynamically following atrialseptostomy; e.g. patients experience no further syn-cope after the procedure and signs and symptoms ofright-sided heart failure improve. Although systemicarterial oxygen desaturation decreases, cardiac outputand oxygen delivery improve through right-to-leftshunting at the atrial level. The authors have alsoobserved an improvement in survival at 1 and 2 yrs;


e.g. survival rates of 87 and 76%, respectively,following atrial septostomy compared with standardtherapy (64 and 42% at 1 and 2 yrs, respectively) [107].Thus, although atrial septostomy does not alter theunderlying disease process, it may improve quality oflife and represent an alternative for selected patientswith severe primary pulmonary hypertension. Alth-ough this invasive procedure is not without risk, theindications for the procedure include: recurrent syn-cope and/or right ventricular failure despite maximalmedical therapy as well as a bridge to transplantationif deterioration occurs despite maximal medical the-rapy. Closure of the atrial septal defect can beperformed at the time of transplantation.

Transplantation

Heart/lung, single lung and bilateral lung transplan-tation have been performed successfully for childrenand adults with pulmonary arterial hypertension.Since 1981, over 1,500 patients have undergonetransplantation for pulmonary arterial hypertensionworldwide. The operative mortality ranges between 16and 29% and is affected by the primary diagnosis. The1-yr survival is between 70 and 75%, 3 yr survival55–60% and 5 yr survival 40–45% [109, 110]. Forpaediatric lung and heart/lung recipients, the datafrom the registry of the International Society forHeart and Lung Transplantation demonstrates thatcurrent survival isy65% at 2 yrs andy40% at 5 yrs[111].

Transplantation should be reserved for patientswith pulmonary arterial hypertension who haveprogressed in spite of optimal medical therapy. Asprogress is made in the medical management ofpulmonary arterial hypertension, the indications fortransplantation will evolve. The appropriate time forevaluation and listing for transplantation remainsproblematic. The course of the disease and the waitingtime must be taken into account. Timing a referral fortransplantation depends upon the patient9s prognosiswith optimal medical therapy, the anticipated waitingtime for transplantation in the region and the expectedsurvival after transplantation. Ideally, children shouldbe listed when their probability of 2-yr survivalwithout transplantation is f50%. There are severaltransplantation options. Acceptable results have beenachieved with heart/lung transplantation, bilaterallung transplantation and single lung transplantation.While there are advantages and disadvantages to eachoperation, there is currently no consensus

Documents

Pulmonary arterial hypertension in children · Pulmonary arterial hypertension in children A. Widlitz, R.J. Barst Pulmonary arterial hypertension in children. A. Widlitz, R.J. Barst