The outcomes of neonates after assisted ventilation are highly variable. Even among infants who require minimal respiratory support, predicting long-term lung and developmental outcomes is complex. Whether the result is respiratory distress syndrome (RDS), genetic predisposition, or postnatal management, prematurity remains the strongest predictor of chronic lung disease and developmental delays. In this chapter we will discuss the key drivers for long-term pulmonary morbidity among premature infants, subsequent pulmonary disease, and the associated neurodevelopmental outcomes.
Prematurity, as defined by birth prior to 37 weeks’ gestation, is an important risk factor for the development of lung disease. Despite significant advances in the respiratory care of neonates, bronchopulmonary dysplasia (BPD) remains the most common serious pulmonary morbidity in premature infants. Attempts at eliminating BPD have been largely unsuccessful, and the incidence of BPD is widely variable among sites, even after adjusting for potential risk factors. Among Vermont Oxford Network (VON) sites the rates of BPD range from 12% to over 40% among infants born at less than 32 weeks’ gestation. Data from a 2010 VON report suggest that the incidence of BPD appears to have plateaued and the absolute number of premature infants with BPD may be increasing. A potential cause for the increase may be the improvement in the survival of extremely preterm and extremely low birth-weight (ELBW) infants, leading to an increase in the numbers of preterm infants who survive to be classified as having BPD.
The contemporary definitions, predictors, and outcomes of BPD have changed. BPD, as recognized and described by William Northway, a pediatric radiologist, was based largely on radiographic evidence of lung disease in infants who survived RDS. Chest radiography demonstrated areas of diffuse heterogeneity and coarse opacities in the most severe of cases. Northway recognized that the key component was a history of mechanical ventilation, a technology not available only several years prior to Northway’s observations. The infants were moderately to late premature infants, and the management of these infants included exposure to prolonged mechanical ventilation, high airway pressures, and oxygen. Histologically, characteristic areas of hyperinflation alternating with areas of focal collapse were found. Hyperplasia of the bronchial epithelium was present and extensive fibrosis was also noted.
The current landscape in which BPD occurs is distinct from that of Northway. “Classic” or “old” BPD has now been replaced by a “new” form of the disease. “New” BPD is found primarily among extremely low gestational age neonates, defined as <28 weeks’ gestational age, and ELBW (i.e., <1000 g birth weight) infants with a history of RDS. These infants are almost uniformly exposed to antenatal steroids and frequently treated with exogenous surfactant therapy. Even the most premature infants may receive only limited exposure to conventional mechanical ventilation but may be exposed to other noninvasive respiratory support such as intermittent positive pressure ventilation, continuous positive airway pressure, or high-flow nasal cannula. The trend in respiratory management has resulted in fewer days on positive pressure ventilation, less time on endotracheal intubation, and less exposure to supplemental oxygen. Chest radiography of new BPD is characterized by diffuse hazy opacities and minimal hyperinflation. In addition, the histology of new BPD shows a reduction in alveoli and capillaries, but minimal fibrosis.
Definitions of Bronchopulmonary Dysplasia
The definition and classification of BPD have evolved as the disease itself has changed. The hallmark feature of BPD that remains constant is the receipt of oxygen therapy or positive pressure for a period of time or on a specific day of life. There are three commonly used definitions: (1) receipt of oxygen at 28 days, (2) need for supplemental oxygen at 36 weeks’ postmenstrual age, and (3) the physiologic definition. The first two definitions, though simple, are limited by the various developmental considerations of infants born at different gestational ages (i.e., 23 weeks vs 28 weeks), site variation in defining need for supplemental oxygen (i.e., oxygen targets), or the use of therapies targeted at reducing oxygen requirements (i.e., diuretics, steroids, high-flow nasal cannulae).
In 2000 a workshop to clarify the definition of BPD was held by the National Institute of Child Health and Human Development (NICHD). At that time the NICHD recognized the importance of distinguishing BPD from the large heterogeneous group of chronic lung disease. This workshop proposed a severity-based definition classifying BPD into mild, moderate, or severe based on either postnatal age or postmenstrual age (PMA) ( Table 43-1 ). Mild BPD was defined as a need for supplemental oxygen (O 2 ) for ≥28 days but not at 36 weeks’ PMA or discharge, moderate BPD as O 2 for ≥28 days plus treatment with <30% O 2 at 36 weeks’ PMA, and severe BPD as O 2 for ≥28 days plus ≥30% O 2 and/or positive pressure at 36 weeks’ PMA. The severity-based definition of BPD was validated by Ehrenkranz et al. by comparing it to the other commonly used definitions such as supplemental oxygen at 28 days and at 36 weeks’ PMA. Overall, the NICHD consensus severity-based scale identified infants most at risk for poor pulmonary outcomes as well as neurodevelopment impairment better than the common definitions ( Table 43-2 ).
|Respiratory Support at 28 Days of Age||Respiratory Support at 36 Weeks’ PMA|
|No BPD||Room air||Room air|
|Mild BPD||Respiratory support||Room air|
|Moderate BPD||Respiratory support||Respiratory support (FiO 2 <30%)|
|Severe BPD||Respiratory support||Respiratory support (≥30%)|
|BPD Definition||NICU Infants, n (%; n = 4866)||Follow-up Infants, n (%; n = 3848)||Pulmonary Medications |
(% of Follow-up † )
|Rehospitalized Pulmonary Cause (% of Follow-up † )||RSV Prophylaxis (% of Follow-up † )|
|None||1124 (23.1)||876 (22.8)||27.2||23.9||12.5|
|Mild||1473 (30.3)||1186 (30.8)||29.7||26.7||16.6|
|Moderate||1471 (30.2)||1143 (29.7)||40.8||33.5||19.2|
|Severe||798 (16.4)||643 (16.7)||46.6 ‡||39.4 ‡||28.4 ‡|
∗ Missing data: 28 days–CXR, 17 infants (13 for follow-up cohort); 36 weeks–CXR, 12 infants (8 for follow-up cohort); pulmonary medications, 17; rehospitalizations for pulmonary causes, 35; RSV prophylaxis, 17.
‡ p < 0.0001, § p < 0.001 versus no BPD for the 28 days, 28 days–CXR, 36 weeks, and 36 weeks–CXR definitions, Mantel-Haenszel Π 2 for linear association across the categories of the consensus definition (none to severe), Mantel-Haenszel Π 2 .
The physiologic definition of BPD, as developed by Walsh, defines BPD at 36 weeks’ adjusted age. Unit-specific rates of BPD among premature infants weighing 501 to 1249 g were compared using the traditional oxygen at 36 weeks’ PMA definition (15-66%) and compared to rates of BPD using the physiologic definition (9-57%). The physiologic definition reduces the between-center variability in the diagnosis of BPD and reduces the diagnosis as much as 10% at individual centers. The physiologic definition has also been validated and shown to be independently predictive of cognitive impairment in infants with BPD. The physiologic definition is used in many clinical trials throughout the United States.
Some clinicians have questioned whether all the definitions are needed. There is merit to the severity classification introduced with the National Institutes of Health (NIH) consensus definition, rather than a binary outcome of yes or no. There is also merit to the more objective criteria introduced by the physiologic definition with the room air challenge. The two definitions are easily combined to merge the best of both ( Table 43-3 ).
|Respiratory Support at 28 Days of Age||Respiratory Support at 36 Weeks’ PMA||BPD Definition|
|Room air||Room air||No BPD|
|Any respiratory support||Room air||Mild BPD|
|Any respiratory support||Respiratory support with FiO 2 <0.30||Room air challenge needed||Challenge passed, mild BPD|
|Challenge failed, moderate BPD|
|Any respiratory support||Respiratory support with FiO 2 ≥0.30||Severe BPD|
It is time for careful consideration of the validity, relevance, and limitations of commonly targeted short-term and long-term outcomes in neonatal clinical trials. The ideal primary outcomes are those that are objective rather than subjective, occur with sufficient frequency to offer practical enrollment targets, and are important to the child’s long-term quality of life. RDS mortality, for example, no longer serves as a relevant primary outcome. Death from RDS in the first week or two of life is an infrequent occurrence in contemporary neonatal intensive care units (NICUs). Death before discharge, while clearly objective and relevant, is influenced by comorbidities and ethical considerations and would be an applicable endpoint only for infants with birth weights less than 800 g, as death occurs infrequently in larger preterm infants and is more often attributable to nonrespiratory causes.
Composite outcomes are frequently used in contemporary neonatal clinical trials. The value of such combined outcomes as death or BPD (or its counterpart, survival without BPD) becomes clear when considering an intervention that may increase the percentage of survivors without BPD by increasing the mortality among the sickest or most vulnerable infants. Such composite outcomes also adjust for differences among centers in their willingness to withdraw support from infants with a poor neurodevelopmental prognosis.
Although BPD, as diagnosed near term gestation, cannot be considered a long-term outcome, the incidence of BPD remains a very relevant endpoint for clinical trials of respiratory management in preterm infants. BPD is an important cause of morbidity and mortality, has been associated with prolonged and recurrent hospitalizations, and is linked to higher rates of other serious complications of prematurity. Its incidence is high, with 7000 to 10,000 new cases each year in the United States alone, and has not been reduced by numerous interventions, including surfactant replacement therapy. The prevalence of BPD has actually increased since 2005 as more ELBW infants survive to discharge. Long-term follow-up suggests that, compared with gestational age-matched infants without BPD, infants with BPD have lifelong alterations in lung function and an increased incidence of cerebral palsy and neurodevelopmental delays.
Pulmonary Function Testing and Imaging
In the future it would be desirable to directly assess an infant’s functional respiratory status with bedside pulmonary function tests during the acute hospitalization. Such tests are currently feasible in intubated neonates and in older, nonintubated infants as young as 4 months of age and have revolutionized the care of neonates with cystic fibrosis. Unfortunately, the currently available tests for nonintubated older infants require sedation and are appropriate only for those over 4 months of age. It would be highly desirable to develop such techniques to study nonintubated convalescent premature infants.
The primary pulmonary function abnormality in survivors of preterm lung disease is reduced forced flows and forced expiratory volume in the presence of normal forced vital capacity, which did not normalize later in infancy in infants either with or without BPD. Many investigators have documented limitations in airflow that may, in part, be reversible with bronchodilators. In the original descriptions by Northway, persistent abnormalities were detected into adulthood.
In the modern era, however, most studies suggest that, with growth, pulmonary function can improve or even normalize. Narayanan and colleagues used the magnetic resonance imaging to demonstrate that former infants with BPD who were evaluated at 10 to 14 years of age had alveolar dimensions similar to those of term controls and of preterm infants without BPD. This suggests an ability to normalize alveolarization with continued development.
Sophisticated imaging techniques to study pulmonary structure have improved the care of neonates with cystic fibrosis. The use of advanced imaging systems, such as high-resolution computerized tomography (CT) and dynamic imaging magnetic resonance, has allowed detailed imaging of the airways, has improved understanding of the pathophysiologic derangements in cystic fibrosis, and is beginning to be applied to BPD. Sarria and colleagues used thin-slice high-resolution CT scan to evaluate the airways and lung parenchyma in 38 survivors of preterm birth (25-29 weeks) compared to full-term infants who had sedated CT scans for nonpulmonary reasons. They found structural differences in both airways and parenchyma. Airway changes were the most marked, while reductions in lung volume were shown only among those with moderate or severe BPD. Increased heterogeneity within the parenchyma was seen in all survivors of BPD. Remarkably, they also found a negative impact of exposure to maternal smoking, with smaller airways in both BPD and control infants and toddlers. Additional information on the long-term impact of these airway and parenchymal changes is emerging. Wong and colleagues studied 21 adult nonsmoking survivors of BPD with CT scan. All of those studied had parenchymal abnormalities, with some showing severe emphysematous changes ( Figs. 43-1 and 43-2 ). These changes were correlated with reductions in pulmonary function tests. Taken together, these findings provide cause for concern about the future pulmonary function of survivors.