Introduction
Infant chronic lung disease, or bronchopulmonary dysplasia (BPD), is a debilitating lung disease that occurs in premature infants. Since the nineteen seventies, advances in neonatal intensive care have led to the survival of smaller and extremely immature infants at a critical stage of lung development. As the survival of very preterm infants improved, the rate of BPD has increased. Survivors with BPD have serious consequences ranging from chronic cardiopulmonary impairment to growth failure, developmental delay, and impaired social functioning of the patient and family. Lengthy hospitalizations, persistent respiratory illness, pulmonary hypertension, delayed growth and development, and poor long-term neurodevelopmental outcomes are common in this population. Currently there is no definitive treatment or prevention for BPD. Many clinical practices currently used in the care of these infants are inadequately studied to ensure safety or efficacy, with potentially serious consequences. The lack of an evidence base has led to large practice variations between neonatal intensive care units and within individual centers. Therefore management of infants with BPD, especially the ones with severe disease, can be extremely challenging to clinicians. This chapter will focus on the respiratory management of patients with established BPD.
Epidemiology, Pathophysiology and Diagnosis of Bronchopulmonary Dysplasia
Despite the efforts to reduce premature birth, there are still over 75,000 infants born at less than 32 weeks’ postmenstrual age (PMA) each year in the United States. Over 50,000 are very low birth-weight (VLBW) infants, born with birth weight <1500 g. BPD is the most prevalent sequela of preterm birth, affecting 10,000-15,000 infants annually in the United States. The Israeli, Canadian, and Japanese Neonatal Networks reported rates of BPD in VLBW infants of 13.7%, 12.3%, and 14.6%, respectively.
Traditionally, the diagnosis of BPD has been based on the presence of chronic respiratory insufficiency with persistent oxygen requirement and abnormal chest radiograph at 1 month of age or at 36 weeks’ PMA. Over the years, many diagnostic criteria for BPD have been developed. Currently the most widely adopted criterion is the severity-based National Institutes of Health (NIH) consensus definition. In this definition, the diagnosis of BPD is first based on treatment with >21% oxygen for at least 28 accumulative days after birth, and then the severity of BPD is based on oxygen level and respiratory support level at 36 weeks’ PMA. However, there is significant center-to-center variability in oxygen use due to the ongoing controversy of appropriate oxygen saturation limits for preterm infants. To decrease this variability, Walsh et al. have proposed a “physiologic” definition of BPD that utilizes an oxygen reduction test to determine oxygen dependency at 36 weeks’ PMA in infants receiving ≤30% supplemental oxygen. Depending on the diagnostic criteria used, the rate of BPD in extremely preterm infants of 22 to 28 weeks’ gestation in the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network centers varies from 68% by NIH consensus definition, to 42% defined as supplemental oxygen use at 36 weeks’ PMA, to 40% by the physiologic definition.
The etiology of BPD is multifactorial, and many potential risk factors for BPD have been identified. Some commonly discussed risk factors for BPD are listed in Table 35-1 . The most important risk factor for developing BPD is extreme prematurity. The incidence of BPD in VLBW infants ranges between 15% and 65%, and this incidence increases as the gestational age (GA) decreases. In a large cohort study of over 15,000 infants born between 22 and 29 weeks’ GA with birth weight of 401 to 1500 g, Lapcharoensap et al. reported an overall rate of BPD among survivors to 36 weeks’ PMA of 33.1%. However, rates of BPD were as high as 80.7% for infants born at <750 g and 93.8% for those with GA of <24 weeks.
| Prenatal risk factors | Intrauterine growth restriction |
| Lack of antenatal corticosteroids | |
| Maternal chorioamnionitis, smoking, or preeclampsia | |
| Genetic predisposition | |
| Risk factors at birth | Low gestational age |
| Low birth weight | |
| Male gender | |
| Lower level of neonatal intensive care at birth hospital | |
| Lower Apgar scores | |
| Perinatal asphyxia | |
| Postnatal risk factors | Mechanical ventilation |
| Supplemental oxygen | |
| Patent ductus arteriosus | |
| Sepsis and systemic inflammatory response | |
| Gastroesophageal reflux |
Northway et al. first described BPD in a group of preterm infants who died after receiving mechanical ventilation and oxygen therapy for respiratory distress syndrome (RDS) in 1967. This was later named as the “classic” or the “old BPD,” as it occurred mainly in relatively large premature newborns (born at 30-34 weeks’ GA) and with pathological features characterized by extensive heterogeneous lung injury with alternating areas of atelectasis, cystic changes and fibrosis, pulmonary artery muscularization, and severe large airway injury. Over the past 40 years, there have been significant advancements in the care of premature infants, including the routine use of surfactant replacement and antenatal steroids as well as the introduction of gentler ventilation modalities. As a result, smaller and more immature infants born at late canalicular or early saccular stages of lung development are surviving. These infants are born several weeks before alveolarization begins, and their lungs are therefore extremely prone to injury. However, with the advances in clinical care, many of these premature infants have a much milder clinical course, and the concept of “new BPD” was proposed in the 1990s. The effects of various injuries on the developing lung give rise to a pathologic picture characterized by impaired alveolar and pulmonary vascular development but less heterogeneous lung injury. These injurious stimuli include inflammation, oxidative stress, ventilator-induced lung injury, infection, drugs, and other factors such as maternal smoking. In recent years, BPD has become a much less common event in infants born at over 30 weeks’ GA and much of the discussion has focused on the new BPD. However, with the prolonged survival of the smallest and sickest premature infants, we have seen a slow increase in the number of infants with extremely severe BPD. At autopsy, the lung histology of those infants who die with the extremely severe lung disease displays a mixed feature of severely delayed lung development and significant lung injury, features seen in both the old and the new BPD ( Fig. 35-1 ).
Clinical Presentation of Established Bronchopulmonary Dysplasia
With the widespread use of antenatal steroids and postnatal surfactant, many small premature infants exhibit a milder course of BPD. These infants often start out with only minimal or mild RDS requiring a low level of respiratory support and then display deterioration in lung function with increased respiratory support and/or oxygen requirements within a few days or weeks after birth. The typical radiographic changes in these patients are usually mild diffuse haziness that persists over time and the pathologic findings are those of typical “new” BPD. With proper nutritional and respiratory support, prevention and control of infection, and other management, such as control of pulmonary overflow through persistent patent ductus arteriosus (PDA), most of these infants will demonstrate slow but steady improvement in their lung function and radiographic changes. After a variable period of time, they can be weaned off respiratory and oxygen support.
Despite the therapeutic improvements, however, we still see infants with severe BPD (sBPD). Using data from the NICHD Neonatal Research Network Generic Database, Natarajan et al. found that 537 of 1159, or 46.3%, infants with birth weights of 401 to 1000 g (born between January 1, 2006, and June 30, 2007) still required mechanical ventilation or continuous positive airway pressure (CPAP) or supplemental oxygen with an effective FiO 2 >30% at 36 weeks’ PMA. Many of these infants require high levels of respiratory support and high concentrations of inspired oxygen from the first week of life. Their initial postnatal courses are frequently complicated by severe RDS, pneumothorax, pulmonary interstitial emphysema, and pulmonary hemorrhage. Some infants may have a milder initial course but have significant deterioration after infections or due to pulmonary edema secondary to PDA. As described earlier in this chapter, the pathologic feature of their lung disease is a combination of delayed lung development and severe injury.
Since the establishment of the Newborn and Infant Chronic Lung Disease Program at the Children’s Hospital of Philadelphia in September 2010, we have treated over 200 premature infants with sBPD. The majority of these patients are older former premature infants with extremely severe lung disease that results in high rates of mortality and long-term ventilation needs. However, although all patients have the same diagnosis of sBPD, they can have very different clinical presentations. This is probably due to different lung developmental stages at which the injuries occurred and the interaction between injuries and host response. We have observed three major phenotypical variants—namely, (1) severe lung parenchyma disease, (2) pulmonary vascular disease, mainly manifests as pulmonary hypertension (PH), and (3) severe airway disease, as the leading features of sBPD ( Fig. 35-2 ). In addition, each of these phenotypes has various subtypes, and an individual patient may have overlapping clinical features from different phenotypes.
Severe Lung Parenchyma Disease as the Leading Feature of Severe Bronchopulmonary Dysplasia
The underlying pathologic changes in these patients may range from homogeneous alveolar simplification ( Fig. 35-3, A ) to heterogeneous micro-/macrocystic changes with areas of fibrosis and/or atelectasis. The radiographic appearance therefore varies from generalized parenchymal opacification to a bubble-like pattern to mixed areas of hyperinflation, cystic changes, and opacifications ( Fig. 35-3, B ).
Pulmonary Hypertension as the Leading Feature of Severe Bronchopulmonary Dysplasia
PH often complicates the course of BPD and contributes to late morbidity and mortality during infancy. Some patients with relatively mild lung disease develop PH, and other infants with severe lung disease may or may not have PH. Nonetheless, this group of patients presents with significant or worsening PH as their key clinical feature in addition to various degrees of lung parenchyma disease. The pulmonary vasculature of these patients is not only underdeveloped and hyperactive but is also undergoing remodeling. Clinically, these infants often present with chronic respiratory insufficiency with oxygen dependency, intermittent cyanotic or life-threatening episodes (“BPD spells”) when agitated, and poor growth. These symptoms are typically not seen until 3 to 4 months after birth, when the patient starts to “outgrow” his or her own pulmonary vascular supply. At this time, significant pulmonary vascular “pruning” can be seen in these patients. As the long-term morbidity and mortality of patients with PH are extremely high, it is very important to identify these infants early so that we can deliver adequate respiratory and other support and provide appropriate follow-up for this population.
Airway Disease as the Leading Feature of Severe Bronchopulmonary Dysplasia
These patients have striking airway abnormalities in addition to varying degrees of lung disease. The central and upper airways abnormalities—for example, glottic/subglottic stenosis and tracheal stenosis—are well-known complications from tracheal intubation and prolonged mechanical ventilation. However, the importance of tracheobronchomalacia and acquired tracheomegaly is less recognized. BPD spells in many patients are caused by airway collapse instead of PH crisis, and therefore the treatment strategy is vastly different from that for someone with PH. Patients with BPD often have asthma-like symptoms, many due to increased small airway reactivity. However, chronic wheezing in some of these patients may be associated with airway obstruction due to malacia and therefore may be unresponsive to bronchodilator therapy.
Physiologic Basis for Respirtory Support in Infants with Established Bronchopulmonary Dysplasia
Ventilatory Control in Infants with Bronchopulmonary Dysplasia
Carotid body function plays an important role in the normal ventilatory response to hypoxia or hyperoxia. The perinatal environment has been shown to affect the development of normal carotid body morphometry and function. Supplemental oxygen at birth profoundly blunts future carotid body development. In rodent BPD models, perinatal hyperoxia causes impaired oxygen sensitivity, carotid body hypoplasia, and decreased total afferent neuron number. Infants with BPD have been shown to have decreased response to both hypoxia-induced increase in ventilation and hyperoxia ventilation depression compared to premature infants who did not need mechanical ventilation or supplemental oxygen. In infants with BPD, ventilation control dysfunction, in combination with decreased efficiency in gas exchange, muscle immaturity, and increased work of breathing, may contribute to both difficulties with oral feeding and disordered breathing during sleep, with more central and obstructive apneas in these infants. Because these infants have been exposed to prolonged hypercarbia, the control of P co 2 is reset, and they can appear comfortable at higher than normal P co 2 levels.
Pulmonary Mechanics in Infants with Bronchopulmonary Dysplasia
Many methods have been used to evaluate pulmonary mechanics in infants with established BPD. For example, resistance and compliance can be measured using an esophageal pressure catheter or the single-breath occlusion method, plethysmography or nitrogen washout and gas dilution methods can be used to measure functional residual capacity (FRC), and the rapid thoracic compression method has been used in the measurement of forced flows. Each of these methods has its own advantages and limitations, and accurate measurements in the unstable infants can be difficult to achieve. In general, infants with established BPD have been found to have decreased compliance, increased resistance with decreased conductance, reduced FRC, and reduced forced flows. These abnormalities may improve over time in the first 3 years of life, but many persist until adolescence and even young adulthood. Many modern ventilators have resistance, compliance, and pressure–volume (P-V) loops displayed. Although the resistance and compliance values may not be absolute, they can be used to trend the patient. Monitoring the changes in resistance and compliance while adjusting ventilator settings at the bedside can be a valuable tool.
In many patients with severe BPD, there are heterogeneous cystic changes with areas of atelectasis in the lung. Ventilation in these patients is not uniform throughout the lung fields, and the time constant is different in different parts of the lung. The respiratory system mechanics in these patients are therefore better explained by a two-compartment (fast and slow) model, rather than a linear one-compartment model. With similar compliance, the slow compartment has a much higher resistance and therefore results in significantly longer time constants compared to the fast compartment. The utility of the two-compartment model in severe BPD has been explained in detail by Castile and Nelin. In addition, current data suggest that infants with established BPD mainly have obstructive rather than restrictive disease and small airways are the primary contributor to the obstruction. Follow-up studies have demonstrated that this obstructive airway disease with air trapping persists over time despite continued increase in the total lung capacity over the first 2 years of life. These changes in lung mechanics need to be taken into consideration when deciding the best respiratory support strategies in the patients with severe BPD.
Respiratory Management in Infants with Established Bronchopulmonary Dysplasia
Over the years, multiple strategies have been tried to prevent BPD with variable success. The major changes in the respiratory support strategies include an attempt to redefine the goals for “adequate gas exchange,” allowing for permissive hypercarbia and permissive hypoxemia, and the widespread use of noninvasive ventilation. These strategies are aimed at minimizing exposure to mechanical ventilation whenever possible. However, in the population with established sBPD, the focus of respiratory support is no longer on the prevention of BPD, as in the newly born premature infants, but rather on how to provide adequate support and minimize V/Q mismatch to promote lung growth while preventing further lung injury.
Noninvasive ventilation
In recent years, noninvasive respiratory support modalities, including nasal intermittent positive-pressure ventilation (NIPPV), nasal CPAP (NCPAP), or high-flow nasal cannula (HFNC), have been used as first-line respiratory therapies in premature infants after birth and following mechanical ventilation. Current pooled data from multiple randomized trials have shown a significant benefit for the combined outcome of death and BPD at 36 weeks’ corrected gestation in premature infants treated with NCPAP, with a number needed to treat of 25. Other studies have demonstrated that noninvasive support modalities such as NIPPV or HFNC may have similar effects compared to NCPAP. With these data, there was widespread use of noninvasive respiratory strategies as a way to avoid or limit duration of intubated respiratory support, and this has also been quickly extended beyond the immediate neonatal period.
Despite a lack of convincing data demonstrating the efficacy of noninvasive respiratory support modalities in infants with established BPD, many neonatologists will extubate or be reluctant to reintubate infants with chronic respiratory insufficiency and keep them on high levels of noninvasive support. In infants with typical “new” BPD who have mild respiratory insufficiency, noninvasive support may be able to adequately support these infants, and over time they can gradually wean off support. Unfortunately, in infants with sBPD, prolonged periods of undersupport may bring severe consequences, including poor growth (both somatic and alveolar/pulmonary vascular growth) and persistent V/Q mismatch contributing to lung injury and the development of PH. Figure 35-4 shows the computed tomography (CT) scan of a 6-month infant born at 26 weeks’ GA. She received about 6 weeks of mechanical ventilation followed by 4 months of noninvasive support. Despite chronic respiratory failure with P co 2 in the 80- to 100-mm Hg range, she was maintained on noninvasive support with chronic diuretics and systemic steroids. Physicians were reluctant to change her from 7-L/min HFNC to intubated mechanical ventilation when she was unable to maintain oxygen saturation above 70%, with P co 2 over 110 mm Hg, persistent tachycardia over 200, evidence of PH, and severe growth failure at 52 weeks’ PMA. Many neonatologists think that weaning the patient to lower respiratory support modalities—namely, CPAP or HFNC—is being successful. However, this may not be true in this group of patients. Although this may be an extremely severe case, it reflects the current trend of relying on noninvasive support and a fear of intubation and mechanical ventilation. Indeed, we have been seeing more patients with sBPD transferred to our center on very high levels of noninvasive support, such as NCPAP >10 cm H 2 O or HFNC >5 L/min or high settings of NIPPV. In a retrospective cohort study of infants who were referred to our center with severe BPD between 2010 and 2013, we found 45% (32 of 71 patients) were on noninvasive respiratory support when they were transferred to our center at 40 to 45 weeks’ PMA. In these patients, 28% either died (9%) or required tracheostomy placement for long-term respiratory support (19%). Fifty-three percent of patients discharged without tracheostomy were discharged on supplemental oxygen. These data highlight the fact that prolonged noninvasive support may not necessarily translate into better pulmonary outcome in infants with severe BPD.
The mechanism of action for noninvasive support may be related to its ability to provide continuous positive pressure and to flush the dead space of the nasopharyngeal cavity, hence improving alveolar ventilation. The goal of noninvasive support in infants with established BPD needs to be providing adequate support to minimize V/Q mismatch and promote growth rather than the sole purpose of avoiding intubation. Therefore some authors have advocated using more objective tools such as esophageal and gastric pressure monitoring to help titrate CPAP pressure. Although there is no consensus on the best methods of titrating the noninvasive support levels, the cardiorespiratory status, overall health, and tolerance to activities, as well as growth, need to be closely monitored and taken into consideration during the duration of noninvasive support. In patients who are not adequately supported by noninvasive support methods, mechanical ventilation (MV), via either endotracheal tube or tracheostomy, needs to be considered.
Mechanical Ventilation
Despite the advancements in respiratory care, a subset of infants with sBPD continues to require prolonged MV. Severe ventilator-dependent BPD is uncommon in most delivery centers but is not rare in many major referral centers. The BPD Collaborative Group reported data from eight U.S. academic centers and showed that 28% of infants with sBPD were on invasive MV at a mean PMA of 47 weeks (range 36 to 86 weeks). Currently there is a dearth of evidence from clinical trials to guide the optimal ventilator management in patients with established BPD. Similar to noninvasive support, the goal of MV should be improving V/Q matching and promoting optimal growth. MV strategies therefore need to be selected based on the lung physiology and pathologic changes of each patient. Identifying phenotypical presentation of sBPD and determining the underlying lung pathology may therefore be the first important step in determining the appropriate management strategy in each patient.
Conventional Mechanical Ventilation
Intermittent mandatory ventilation with time-cycled, pressure-limited ventilation has been the main mode of conventional MV for many years. Advances in ventilation techniques in recent years include volume-targeted ventilation, patient-triggered ventilation including synchronized intermittent mandatory ventilation (SIMV), assist/control ventilation, pressure support ventilation (PSV), and flow-cycled ventilation (mainly used in PSV). In addition, real-time graphic monitoring is now available in the newer ventilators, enabling clinicians to visualize respiratory mechanics on a breath-to-breath basis. Unfortunately, most of the ventilation data for preterm infants, especially high-quality randomized trials, is concentrated in the early postnatal period with RDS, and there is no clear evidence for an optimal ventilator strategy in infants with established BPD. Because of the lack of evidence in this population, our group evolved the following care strategies after caring for several hundred infants with sBPD since 2010. Summarized below are three guiding principles to use when trying to provide optimum ventilator support for these patients:
- 1.
The main goal of MV in this population is to provide sufficient support that the patient needs, rather than weaning.
- 2.
Make sure the alveoli are well recruited—that is, use an open lung strategy.
- 3.
Ensure adequate expiration to minimize air trapping.
Here we use the SIMV + PSV mode as an example to discuss the methods we can use to follow the above guiding principles. Main parameters to adjust in this mode include tidal volume (V T ) or peak inflating pressure (PIP) to achieve the targeted V T , mandatory ventilator rate, inspiration time (i-time), positive end-expiratory pressure (PEEP), and pressure support (PS). The targets and strategies used to set the ventilator parameters are summarized in Table 35-2 . Other ventilator modes may also be used successfully in infants with sBPD; however, the same guiding principles and targets should be followed.
| Target | Strategies |
|---|---|
| Establish optimum lung volume |
|
| |
| |
| Promote even distribution of ventilation |
|
| |
| |
| Maintain open airway |
|
|
When adjusting the ventilator parameters to achieve these targets, it is important to remember there is interplay of the targets. Therefore adjusting one or two parameters may produce a profound impact in one area but may not result in overall improvement. In the following paragraphs, we will discuss each key parameter in more detail.
Setting the target tidal volume
Because volutrauma has been associated with the development of lung injury, volume-targeted ventilation has been advocated in neonatal MV in recent years. In the early postnatal days, reported benefits of volume-targeted ventilation included tighter V T and carbon dioxide control, fewer pneumothoraces, fewer days of ventilation, reduction in severe intraventricular hemorrhage, and, most important, decreased death or BPD. As patients with sBPD often have marked variability in compliance and resistance over time, they may benefit from volume-targeted or patient-initiated pressure-regulated and volume-controlled ventilation to ensure delivery of adequate V T with the least pressure. Unfortunately, currently there is no specific evidence to guide the use of volume-targeted ventilation in patients with sBPD. In our experience, these patients may need much higher V T s (8 to 12 mL/kg) compared to younger preterm infants. High enough V T , in conjunction with adequate support during spontaneous breath, can ensure adequate minute ventilation and improve a patient’s comfort, which in turn will contribute to improvement in the increased work of breathing and tachypnea often seen in this population.
Owing to issues of high airway resistance and distended trachea with prolonged intubation, many infants with sBPD have significant leak around the endotracheal tube (ETT), often over 50%, and the amount of leak varies from inflation to inflation. This poses challenges to effective volume ventilation. In the presence of ETT leak, the gas leaving the lung most closely represents the V T that entered the lung. Targeting the expired V T may be the best way to control the delivered V T in these patients. Luckily many newer ventilators now have the flow sensor at the “Y” connector and are able to measure and display the V T in and out of the baby. This improvement enables the ventilator to provide better leak compensation, and clinicians are able to achieve tighter control of the expired V T . However, in cases of severe leak and when the newer ventilators are not available, volume ventilation may not be feasible. In these cases, patients may need high PIP settings in the 30- to 40-cm H 2 O range, given their stiff lungs with poor lung compliance.
Finding appropriate rate and inspiration time
The majority of patients with sBPD have heterogeneous lung disease (see Fig. 35-3, B ), with both collapsed and overinflated areas in their lungs, causing significant maldistribution of ventilation. As discussed earlier, the lung mechanics in these patients are better explained by a two-compartment model with different time constants in different parts of the lung. To ensure adequate gas exchange and emptying of the slower compartment, which contributes to the majority of the exhaled V T , we elect to use a low rate and long inspiratory time strategy. This strategy has been used successfully in several centers that treat infants with sBPD. Patients with sBPD often breathe fast with a short i-time. Setting a long i-time during mandatory ventilator breaths therefore is needed to ensure air entry into the slow compartments. However, these slow compartments also need long expiration time for alveolar pressure to equilibrate with upper airway pressure. Therefore, a slow rate allowing enough time during expiration for the slow compartments to empty is critical in minimizing gas trapping. To ensure an overall slow respiratory rate and a good composite inspiratory-to-expiratory (I:E) ratio, we advocate using a slow ventilator rate (10 to 20) with adequate V T and PS. Adding adequate PS to the spontaneous breath will help prevent underventilation of the fast compartment. The combined effort of improving ventilation in both the slow and the fast compartments often results in improved minute ventilation and patient comfort, which in turn helps to slow down the breathing rate and further minimize air trapping.
The slow-rate, long i-time ventilation plus adequate PS strategy may be an effective way of ventilating the majority of infants with sBPD. In a patient with uniform lung disease, however (see Fig. 35-3, A ), who has a homogeneous hazy chest radiographic appearance and underlying pathologic feature of generalized alveolar simplification, respiratory insufficiency is probably due to decreased alveoli surface area rather than having a maldistribution of ventilation. The lung compliance would be fairly consistent throughout the lung fields, and the time constant is usually relatively short. These patients may do better with faster rate and shorter i-time.
Setting optimum PEEP
Setting an appropriate PEEP is an important component of ventilator management. An appropriate level of PEEP can increase FRC, promote alveoli recruitment, reduce work of breathing, and improve V/Q matching. Animal studies have suggested that very low PEEP will lead to impaired gas exchange and increased risk of lung injury, whereas open lung ventilation improves gas exchange and attenuates secondary lung injury. Major concerns about high PEEP level mainly come from the worry that high PEEP may decrease tidal and minute ventilation; impair expiration, causing gas trapping; and impair venous return, resulting in decreased cardiac output. Randomized clinical trials comparing different PEEP levels have been performed in both adults and neonates with acute RDS but not in infants with BPD.
Paradoxically, increased PEEP may be indicated when overexpansion of the lungs is observed. This is contrary to common practice but is based on sound pathophysiologic principles. As described earlier, infants with established BPD have been found to have decreased lung compliance, increased resistance, reduced FRC, and obstructive lung disease. In addition, many patients with sBPD have issues with tracheobronchomalacia, resulting in dynamic airway collapse. These airway and pulmonary mechanical characteristics put them at increased risk of developing inadvertent or intrinsic PEEP (PEEP i ). When the set ventilator PEEP is less than PEEP i , the nonparalyzed infant must first overcome the imposed elastic load of the PEEP i before any inspiratory flow can be generated. This means that the infant often cannot generate enough inspiratory flow to trigger the ventilator in the normal respiratory cycle, resulting in ineffective inspiratory efforts, loss of patient–ventilator synchrony, air hunger, and excessive respiratory work. This may also be the source of some BPD spells (desaturation episodes), as the infant’s ineffective efforts cause greater air hunger and hypoxemia. The poorly supported floppy airways of infants with sBPD are susceptible to collapse in the later phase of exhalation as lung volume decreases, especially when the infant is agitated.
It is imperative that an individual level of PEEP be established for each patient and changes made as the disease changes. We have learned that this individualized PEEP level can be found based on possible underlying pathology and ventilator P–V curves of each patient. Finding this optimum PEEP may help break the cycle of alveolar collapse and airway instability. Figure 35-5 shows the admission chest radiograph of an infant with sBPD who was transferred to our center for management of ventilator failure. She had persistent hyperinflation of her lungs despite decreasing PEEP to 3 cm H 2 O and required 100% oxygen for several weeks prior to her transfer. Based on P-V curve changes with different levels of PEEP, we determined that this patient required a PEEP of 14 to 15 cm H 2 O (see Fig. 35-5 , right). A bedside bronchoscopy demonstrated that her bilateral bronchus completely collapsed when PEEP was decreased to less than 8 cm H 2 O. Her oxygenation improved dramatically, and the lung hyperinflation gradually decreased with the higher PEEP. We have attempted using a PEEP grid to identify optimum PEEP in each patient. Figure 35-6 demonstrates the identification of peak and plateau pressure, peak flow, and volume under a PEEP of 12 cm H 2 O during a PEEP grid testing. Compliance and resistance under different levels of PEEP (5-18 cm H 2 O) were calculated based on the identified values. Airway malacia can be documented using full-inflation and end-exhalation controlled-ventilation chest CT or bedside flexible bronchoscopy. In addition, PEEP level can be titrated at the bedside with the use of bronchoscopy by applying a stepwise increase/decrease in PEEP to the airways and directly visualizing and determining the effect of increased PEEP on airway collapse. The challenge in using these methods is that the patient needs to be quiet and often sedated to obtain an accurate value. PEEP requirements in these patients are also dynamic and may vary from day to day, during agitation, or when the disease process changes. The optimal PEEP is determined by the interplay between the severity of airway collapse or tracheobronchomalacia and the severity of parenchymal lung disease.
