Pulmonary vascular disease and pulmonary hypertension contribute to the pathophysiology and outcomes of infants with bronchopulmonary dysplasia. Data are extremely limited regarding many aspects of pulmonary hypertension in bronchopulmonary dysplasia, including the need to learn more about its natural history and prevalence, mechanisms that cause pulmonary hypertension or contribute to progressive disease, and the relative risks and benefits of current therapeutic strategies. Although new therapies are now available for the treatment of pulmonary hypertension, their role in the clinical care of severe bronchopulmonary dysplasia, and improving long-term outcomes requires more thorough investigation.
Bronchopulmonary dysplasia (BPD) is the chronic lung disease of infancy that occurs in premature infants after oxygen and ventilator therapy for acute respiratory disease at birth. As first characterized by Northway and colleagues nearly 45 years ago, BPD originally described severe respiratory morbidity and high mortality in relatively late-gestation preterm infants, which was largely due to the lack of surfactant therapy and insufficient neonatal ventilator care in that era (Figure 1).1 Improved obstetrical and neonatal care over time has increased survival of even the smallest of immature newborns, but BPD persists as a major problem, occurring in an estimated 10,000-15,000 infants per year in the US alone. This has important health care implications, as infants with BPD require prolonged NICU courses; have frequent readmissions during the first 2 years after NICU discharge for respiratory infections, asthma, and related problems; and have persistent lung function abnormalities and exercise intolerance as adolescents and young adults.1–3
Chest x-ray showing severe BPD.
The overall incidence of BPD has not declined over the past decade,8 but the respiratory course and number of infants with severe BPD have clearly changed with current clinical practice. Infants with chronic lung disease after premature birth have a different clinical course and pathology than had been traditionally observed in infants dying with BPD during the pre-surfactant era.3–7 The classic progressive stages of disease, including prominent fibro-proliferative changes that first characterized BPD, are often absent now, and the disease has changed to being predominantly defined as a disruption of distal lung growth, referred to as the “new BPD.”5 In contrast with the past, the “new BPD” often develops in preterm newborns who may have required minimal or even no ventilator support and relatively low inspired oxygen concentrations during the early postnatal days.6,7 The “new BPD” is likely the result of disrupted antenatal and postnatal lung growth, leading to persistent abnormalities of lung architecture and function. The implications of how these changes in BPD alter long-term pulmonary outcomes remain uncertain.
Although marked improvements in care have led to milder respiratory courses for most preterm infants, infants with BPD can still develop severe chronic respiratory failure with marked cardiopulmonary impairment. Despite these striking changes in the nature and epidemiology of BPD over time, pulmonary hypertension (PH) continues to contribute significantly to high morbidity and mortality in BPD, and is present early in the course of disease.8,9 Even the original descriptions of BPD noted striking pulmonary hypertensive vascular remodeling in severe cases, and that the presence of PH beyond 3 months of age was associated with high mortality (40%).10 Now in the “post-surfactant era,” late PH continues to be strongly linked with poor survival in the “new BPD,” with a recent report suggesting mortality rates of nearly 70% for infants with severe PH.11 Recently, with more active testing with echocardiograms, PH is being recognized more frequently even in preterm infants with relatively mild lung disease. This article briefly discusses the pathogenesis, diagnostic approach, and current therapies for preterm infants with BPD and PH.
In addition to the adverse effects of PH on the clinical course of infants with BPD, the lung circulation is further characterized by abnormal (or “dysmorphic”) growth of the pulmonary circulation, including a reduction of small pulmonary arteries and an altered pattern of distribution within the lung interstitium.12–14 This reduction of alveolar-capillary surface area impairs gas exchange, which increases the need for prolonged oxygen and ventilator therapy, causes marked hypoxemia with acute respiratory infections and exercise, and increases the risk for developing severe PH. Experimental studies have further shown that early injury to the developing lung can impair angiogenesis, which further contributes to decreased alveolarization and simplification of distal lung airspace (the so-called “vascular hypothesis”).15 Thus, abnormalities of the lung circulation in BPD are not only related to the presence or absence of PH, but more broadly, pulmonary vascular disease after premature birth as manifested by decreased vascular growth and structure also contributes to the pathogenesis and abnormal cardiopulmonary physiology of BPD.
From the earliest descriptions of BPD, PH and cor pulmonale have long been recognized as being associated with high mortality in BPD.10,16,17 Early injury to the lung circulation leads to the rapid development of PH after premature birth.18 Abnormalities of the pulmonary circulation in BPD include increased vascular tone and vasoreactivity, hypertensive remodeling, and decreased growth (Figure 2).18 Physiologic abnormalities of the pulmonary circulation in BPD include elevated pulmonary vascular resistance (PVR) and abnormal vasoreactivity, as evidenced by the marked vasoconstrictor response to acute hypoxia.19,20 Cardiac catheterization studies have shown that even mild hypoxia can cause marked elevations in pulmonary artery pressure in some infants with BPD, including infants with only modest basal elevations of pulmonary hypertension.19 Increased pulmonary vascular tone contributes to high PVR even in older children with BPD without hypoxia, suggesting that abnormal vascular function persists even late in the course.21 Reduced vascular growth (angiogenesis) also limits vascular surface area, causing further elevations of PVR, especially in response to high cardiac output with exercise or stress. The ability of the lung to achieve normal gas exchange requires ongoing growth and maintenance of an intricate system of airways and vessels, including the establishment of a thin yet vast blood-gas interface.
In animal models, disruption of angiogenesis during lung development not only increases the risk for PH, but it also impairs alveolarization.22,23 Inhibition of lung vascular growth during a critical period of postnatal lung growth impairs alveolarization, suggesting that endothelial-epithelial cross-talk, especially via vascular endothelial growth factor signaling, is critical for normal lung growth following birth and if disrupted, contributes to BPD.15,22,23 Clinically, reduced vascular surface area implies that even relatively minor increases in left-to-right shunting of blood flow through a patent foramen ovale, atrial septal defect (ASD), or patent ductus arteriosus may induce a far greater hemodynamic injury in infants with BPD than in infants with normal lung vascular growth. Prominent bronchial and other systemic-to-pulmonary collateral vessels are also found in morphometric studies of infants with BPD, and can be readily identified in many infants during cardiac catheterization. Although these collateral vessels are generally small, large collaterals may contribute to significant shunting of blood flow to the lung, causing edema and need for higher FiO2.
In addition to high pulmonary vascular tone, abnormal vasoreactivity, hypertensive vascular remodeling, and decreased surface area, left ventricular diastolic dysfunction (LVDD) can also contribute to high pulmonary artery pressure in infants with BPD.24 Up to 25% of BPD infants with PH who were evaluated by cardiac catheterization had hemodynamic signs of LVDD in one retrospective study.25 Some infants with LVDD present with persistent requirements for frequent diuretic therapy to treat recurrent pulmonary edema, even in the presence of only mild PH.
Although the exact incidence of PH in BPD is uncertain, marked PH beyond the first few months of life has been recently associated with a 47% mortality within 2 years after diagnosis.15 PH is not only a marker of more advanced BPD, but high PVR also causes poor right ventricular function, impaired cardiac output, limited oxygen delivery, increased pulmonary edema, and possibly a higher risk for sudden death. PH in BPD is increasingly recognized in preterm infants with lower mortality risk, and retrospective studies have noted PH in roughly 25%-37% of infants with BPD.26,27 These retrospective studies are somewhat limited, however, by the selective assessment of PH by echocardiogram. The diagnosis of PH and other cardiovascular complications in infants with BPD can be difficult because clinical signs and symptoms of PH can be subtle or overlap with respiratory signs. Based on strong correlations between PH and survival in BPD,14,15 early detection of PH may provide helpful prognostic information and lead to the earlier application of more aggressive respiratory support, cardiac medications, vasodilators, and surgical or interventional cardiac catheterization procedures in order to improve late outcomes. Prospective data regarding the precise incidence and natural history of PH in BPD are lacking, and most information on diagnostic and therapeutic strategies are based on clinical observations, rather than rigorous, randomized clinical trials.
Lung histology illustrating hypertensive pulmonary vascular remodeling in severe BPD.
In general, we recommend early echocardiograms for the diagnosis of PH in preterm infants with severe respiratory distress syndrome (RDS) who require high levels of ventilator support and supplemental oxygen, especially in the setting of oligohydramnios and intrauterine growth restriction (IUGR). Infants with more severe prematurity (<26 weeks) are at highest risk for late PH. Similarly, infants with a particularly slow rate of clinical improvement, as manifested by persistent or progressively increased need for high levels of respiratory support, should be assessed for PH. In the setting of established BPD, preterm infants who at 36 weeks post conceptual age still require positive pressure ventilation support, are not weaning consistently from oxygen, have oxygen needs at levels disproportionate to their degree of lung disease, or have recurrent cyanotic episodes warrant screening for PH or related cardiovascular sequelae. Other clinical markers often associated with more severe disease include feeding dysfunction and poor growth, recurrent hospitalizations, and elevated PaCO2. High PaCO2 is a marker of disease severity and reflects significant airway obstruction, abnormal lung compliance with heterogeneous parenchymal disease, or reduced surface area, and is an indication for PH screening. Another strategy would be to use echocardiograms to screen every patient at 36 weeks of age who is diagnosed with moderate or severe BPD, but how often PH would be missed in patients with milder BPD is uncertain.
Serial electrocardiograms may have inadequate sensitivity and positive predictive value for identification of right ventricular hypertrophy (RVH) as a marker of PH. Some patients can have significant RVH and PH despite minimal or normal electrocardiogram findings. As a result, we recommend serial echocardiograms for screening for PH in patients with BPD. Estimated systolic pulmonary artery pressure (sPAP) derived from the tricuspid regurgitant jet (TRJV) measured by echocardiogram has become one of the most utilized findings for evaluating PH. Past studies have shown excellent correlation coefficients (r values between 0.93-0.97) when compared with cardiac catheterization measurements in children less than 2 years old with congenital heart disease.28,29 However, these studies evaluated echocardiogram and cardiac catheterization performed simultaneously under the same hemodynamic conditions, and the utility of echocardiograms in predicting disease severity as applied in the clinical setting is less clear.
A recent study examined the utility of echocardiogram assessments of PH in infants with BPD with subsequent cardiac catheterization measurements of pulmonary artery pressure.24 Systolic pulmonary artery pressure could be estimated in only 61% of studies, and there was poor correlation between echocardiogram and cardiac catheterization measures of sPAP in these infants. Echocardiogram estimates of sPAP correctly identified the presence or absence of PH in 79% of these studies, but the severity of PH was correctly assessed in only 47% of those studies. Seven of 12 children (58%) without PH by echocardiogram had PH during subsequent cardiac catheterization. In the absence of a measurable TRJV, qualitative echocardiogram findings of PH, including right atrial enlargement, right ventricular hypertrophy, right ventricular dilation, pulmonary artery dilation, and septal flattening, either alone or in combination have relatively poor predictive value. Factors associated with chronic lung disease, specifically marked pulmonary hyperinflation, expansion of the thoracic cage, and alteration of the position of the heart, adversely affect the ability to detect and measure TRJV.30 Thus, as used in clinical practice, echocardiography often identifies PH in infants with BPD, but estimates of sPAP were not obtained consistently and were often not reliable for determining disease severity. Other measures of right ventricular strain and PH, including AT/ET ration and the Tei index, could be helpful in the absence of a measurable TRJV, but have not been fully evaluated in infants with BPD. Despite its limitations, echocardiography remains the best available screening tool for PH in BPD patients.
In patients with PH by echocardiogram, we generally recommend cardiac catheterization for patients with BPD who: 1) have persistent signs of severe cardiorespiratory disease or clinical deterioration not directly related to airways disease; 2) are suspected of having significant PH despite optimal management of their lung disease and associated morbidities; 3) are candidates for chronic PH drug therapy; 4) have unexplained, recurrent pulmonary edema; and others. The goals of cardiac catheterization are to: assess the severity of PH; exclude or document the severity of associated anatomic cardiac lesions; define the presence of systemic-pulmonary collateral vessels, pulmonary venous obstruction, or left heart dysfunction; and to assess pulmonary vascular reactivity in patients who fail to respond to oxygen therapy alone. Other critical information can be acquired during cardiac catheterization that may significantly aid in the management of infants with BPD. In particular, assessment of shunt lesions, including the degree of shunting as determined by the pulmonary to systemic flow ratio may help determine the timing of shunt closure. In some infants, closure of a congenital heart lesion may prevent the ability of the right sided circulation to “pop off” in times of acute increases in pulmonary pressure and resistance. The presence, size, and significance of bronchial or systemic collateral arteries; determining the presence of pulmonary artery stenosis; and structural assessments of the pulmonary arterial and venous circulation by angiography are among several key factors that may affect cardiopulmonary function. A recent report highlighted the importance of pulmonary vein stenosis or veno-occlusive disease in premature infants.31 Most importantly, elevated pulmonary capillary wedge or left atrial pressure may signify left sided systolic or diastolic dysfunction. LVDD can contribute to PH, recurrent pulmonary edema, or poor inhaled nitric oxide (iNO) responsiveness in infants with BPD, and measuring changes in pulmonary capillary wedge pressure and left atrial pressure during acute vasoreactivity testing may help with this assessment.
The initial clinical strategy for the management of PH in infants with BPD begins with treating the underlying lung disease. This includes an extensive evaluation for chronic reflux and aspiration, structural airway abnormalities (such as tonsillar and adenoidal hypertrophy, vocal cord paralysis, subglottic stenosis, tracheomalacia, and other lesions), assessments of bronchoreactivity, improving lung edema and airway function, and others. Periods of acute hypoxia whether intermittent or prolonged are common causes of persistent PH in BPD.32 Brief assessments of oxygenation (“spot checks”) are not sufficient for decisions on the level of supplemental oxygen needed. Targeting oxygen saturations to 92%-94% should be sufficient to prevent the adverse effects of hypoxia in most infants, without increasing the risk of additional lung inflammation and injury. A sleep study may be necessary to determine the presence of noteworthy episodes of hypoxia and whether hypoxemia has predominantly obstructive, central, or mixed causes.
Additional studies that may be required include flexible bronchoscopy for the diagnosis of anatomical and dynamic airway lesions (such as tracheomalacia) that may contribute to hypoxemia and poor clinical responses to oxygen therapy. Upper gastrointestinal series, pH or impedance probe, and swallow studies may be indicated to evaluate for gastro-esophageal reflux and aspiration that can contribute to ongoing lung injury. For patients with BPD and severe PH who fail to maintain near normal ventilation or require high levels of FiO2 despite conservative treatment, consideration should be given to chronic mechanical ventilatory support. Despite the growing use of pulmonary vasodilator therapy for the treatment of PH in BPD, data demonstrating efficacy are extremely limited, and the use of these agents should only follow thorough diagnostic evaluations and aggressive management of the underlying lung disease. Current therapies used for PH therapy in infants with BPD generally include iNO, sildenafil, endothelin receptor antagonists, and calcium channel blockers (CCB).
Inhaled nitric oxide causes selective pulmonary vasodilation20 and improves oxygenation in infants with established BPD.33 Although long-term iNO therapy has been used in BPD infants, especially for those who require continued mechanical ventilator support, efficacy data are not available. Although iNO for PH therapy is often initiated at doses of 10-20 ppm, most patients subsequently tolerate weaning of the iNO dose to a range of 2-10 ppm. The lower dose may further enhance ventilation-perfusion matching, allowing for better oxygenation at lower FiO2.
Sildenafil, a highly selective type 5 phosphodiesterase (PDE-5) inhibitor, augments cyclic GMP content in vascular smooth muscle, and has been approved for adults with PH alone34 and in combination with standard treatment regimens.35,36 Studies of sildenafil therapy in children with PH have been limited, but include a demonstration of its efficacy in the treatment of persistent PH of the newborn,37 and its safety and possible efficacy during long-term therapy in older children with PH.38 By prolonging cGMP levels during iNO induced vasodilation, PDE-5 inhibitors may be useful to augment the response to iNO therapy or to prevent rebound PH after abrupt withdrawal of iNO. In a study of 25 infants with chronic lung disease and PH (18 with BPD), prolonged sildenafil therapy as part of an aggressive program to treat PH was associated with improvement in PH by echocardiogram in most (88%) patients without significant rates of adverse events.39 Although the time to improvement was variable, many patients were able to wean off mechanical ventilator support and other PH therapies, especially iNO, during the course of sildenafil treatment without worsening of PH. The recommended starting dose for sildenafil is 0.5 mg/kg/dose every 8 hours. If there is no evidence of systemic hypotension, this dose can be gradually increased over 2 weeks to achieve desired pulmonary hemodynamic effect or a maximum of 2 mg/kg/dose every 6 hours.
Bosentan, a nonselective endothelin receptor antagonist, is commonly used in older patients with PH. A retrospective study suggested that bosentan may be safe and effective for the treatment of PH in children as young as 9 months,40 but data are limited to case reports regarding its use in BPD infants. Monthly liver function testing is required to monitor for hepatotoxicity. Calcium channel blockers benefit some patients with PH, and short-term effects of CCB in infants with BPD have been reported.41,42 Nifedipine can acutely lower pulmonary artery pressure and PVR in children with BPD; however, some patients were acutely hypoxemic during this study, and the effects of nifedipine on pulmonary artery pressure were not different from the effects of supplemental oxygen alone. In comparison with an acute study of iNO reactivity in infants with BPD, the acute response to CCB was poor and some infants developed systemic hypotension. We generally use sildenafil or bosentan for chronic therapy of PH in infants with BPD.
Intravenous prostacyclin analogue (PGI2; epoprostenol, treprostinil) therapy has been used extensively in older patients with severe PH, and has been shown to improve survival of patients with advanced disease. PGI2 has been used in some infants with BPD and late PH,43 but concerns regarding its potential to worsen gas exchange due to increased ventilation-perfusion mismatching in the setting of chronic lung disease and systemic hypotension have limited its use in this setting. Although another stable PGI2 analogue, iloprost, is available for inhalational use, the need for frequent treatments (6-8 times daily) and occasional bronchospasm may be significant factors restricting its use in the setting of BPD.44
In addition to close monitoring of pulmonary status, infants with BPD and PH should be followed by serial echocardiograms, which should be obtained at least every 2-4 weeks with the acute initiation of therapy and at 4- to 6-month intervals with stable disease. Abrupt worsening of PH may reflect several factors, including the lack of compliance with oxygen therapy or medication use, but can be related to the progressive development of pulmonary vein stenosis or veno-occlusive disease). Repeat cardiac catheterization may be indicated for patients being treated for PH with vasodilator therapy who experience clinical deterioration, worsening PH by echocardiogram, or when echocardiogram measurements fail to provide adequate hemodynamic assessment of sicker patients. Frequently pulmonary vasodilators are continued through the winter cold and flu season, as patients with mild disease may have a severe PH exacerbation with a respiratory illness. We recommend weaning medications with serial normal or near-normal echocardiogram findings, and that the addition of biomarkers, such as pro-NT brain natriuretic peptide levels, may be useful for long-term follow-up.
In summary, pulmonary vascular disease and PH contribute to the pathophysiology and outcomes of infants with BPD. Data are extremely limited regarding many aspects of PH in BPD, including the need to learn more about its natural history and prevalence, mechanisms that cause PH or contribute to progressive disease, and the relative risks and benefits of current therapeutic strategies. Although new therapies are now available for the treatment of PH, their role in the clinical care of severe BPD and improving long-term outcomes requires more thorough investigation.