Conflict of Interest Statement
Dr. Keszler has received honoraria for lectures and research grant support from Dräger Medical. Dr. Keszler also chairs the Data Safety Monitoring Board of a clinical trial supported by Medipost America and serves on a Data Safety Monitoring Board for a clinical trial sponsored by Chiesi USA Inc. None of the companies had any input into the content of this chapter.
The immature lungs of extremely preterm infants are susceptible to damage from a variety of factors that are potentially modifiable by the use of lung-protective strategies of respiratory support.
Avoidance of mechanical ventilation, optimal delivery room stabilization, and early use of noninvasive respiratory support are important elements in minimizing lung injury.
When mechanical ventilation is needed, it should be used with care and attention to the individual patient’s specific pathophysiology and with a goal of extubation at the earliest opportunity.
Recruitment and maintenance of optimal lung volume is a key element in any lung-protective ventilation strategy, including both conventional and high-frequency ventilation.
Volume-targeted ventilation maintains more stable tidal volume and minute ventilation, shortens the duration of mechanical ventilation, and is associated with a decrease in both lung and brain injury.
Despite appropriate emphasis on noninvasive respiratory support when feasible, mechanical ventilation (MV) remains a mainstay of therapy in extremely preterm infants. Although it is undoubtedly lifesaving, invasive MV has many untoward effects on the brain and the lungs, especially in the most immature infants. The endotracheal tube (ETT) acts as a foreign body, quickly becoming colonized and acting as a portal of entry for pathogens, increasing the risk of ventilator-associated pneumonia and late-onset sepsis. For these reasons, avoidance of MV in favor of noninvasive respiratory support is considered one of the most important steps in preventing neonatal morbidity. When MV is required, the goal is to wean the patient from invasive ventilation as soon as feasible to minimize ventilator-associated lung injury (VALI). Although VALI is a key element in the pathogenesis of bronchopulmonary dysplasia (BPD), many other factors play an important role in its pathogenesis, including the intrauterine environment (inflammation and infection), postnatal infection, oxidative stress, antenatal and postnatal nutritional deprivation, presence of patent ductus arteriosus, and excessive fluid administration.
Ventilator-Associated Lung Injury
Many terms have been used to describe the mechanism of injury in VALI. Barotrauma refers to damage caused by inflation pressure. The conviction that pressure is the major determinant of lung injury has caused clinicians to focus on limiting inflation pressure, often to the point of precluding adequate ventilation. However, there is convincing evidence that high inflation pressure by itself, without correspondingly high tidal volume ( V T ), does not result in lung injury. Rather, injury related to high inflation pressure is mediated through tissue stretch resulting from excessive V T or from regional overdistention when ventilating lungs with extensive atelectasis. Dreyfuss and colleagues demonstrated more than 20 years ago that severe acute lung injury occurred in small animals ventilated with large V T , regardless of whether that volume was generated by positive or negative pressure. In contrast, animals exposed to the same high inflation pressure but with an elastic bandage over the chest and abdomen to limit the delivered V T showed substantially less acute lung injury. Similarly, Hernandez et al. demonstrated that animals exposed to pressure as high as 45 cm H 2 O did not show evidence of acute lung injury when their chest and abdomen were enclosed in a plaster cast. Volutrauma refers to injury caused by overdistention and excessive stretching of tissues, which leads to disruption of alveolar and small airway epithelium, resulting in acute edema, an outpouring of protein-rich exudate, and release of proteases, cytokines, and chemokines. This in turn leads to activation of macrophages and invasion of activated neutrophils. Collectively, this latter process is referred to as biotrauma . Another important concept is that of atelectrauma , or lung damage caused by tidal ventilation in the presence of atelectasis. Atelectrauma causes lung injury via several different mechanisms. The portion of the lungs that remains atelectatic experiences increased surfactant turnover and has high critical opening pressure. There are shear forces at the boundary between the aerated and atelectatic parts of the lung, leading to structural tissue damage. Ventilation of injured lungs using inadequate end-expiratory pressure results in repeated alveolar collapse and expansion with each inflation, a process that leads to lung injury quite rapidly. Perhaps most importantly, when a portion of the lungs is atelectatic, the gas entering the lungs will preferentially distend the aerated portion of the lung, which is more compliant than the atelectatic lung with its high critical opening pressure. This fact is evident from Laplace’s law and corroborated by experimental evidence, showing that the most injured portion of the lung was the aerated nondependent lung. This maldistribution of V T leads to overdistention of that portion of the lungs and regional volutrauma. Thus VALI is initiated by some form of biophysical injury, which in turn triggers a release of mediators and activated leukocytes, leading to biotrauma and initiating the complex cascade of lung injury and eventual repair. A schematic representation of the cycle of VALI is illustrated in Fig. 18.1 .
How Can VALI Be Reduced?
As is evident from the prior discussion, the process of lung damage from MV is multifactorial and cannot be linked to any single variable. Consequently, any approach to reducing lung injury must be comprehensive and begin with the initial stabilization of the infant in the delivery room. Because some degree of impairment of normal pulmonary development (i.e., arrest of alveolarization) is probably inevitable when an extremely preterm fetus is suddenly thrust into what by fetal standards is a very hyperoxic environment and must initiate air breathing with incompletely developed lungs, it is unlikely that improved respiratory support and avoidance of MV can completely prevent impairment of lung structure and function. However, optimal respiratory and general supportive care can minimize the superimposition of VALI on this developmental arrest and together with optimal nutrition can facilitate lung growth and repair.
Delivery Room Stabilization
The time immediately after birth when air breathing is initiated in a structurally immature surfactant deficient lung is recognized as a critical time during which the process of lung injury and subsequent repair may be triggered in a matter of minutes. Moments after birth, the newborn must rapidly clear lung fluid from the airways and terminal air spaces, aerate the lungs, and sustain a functional residual capacity, thus facilitating a dramatic increase in pulmonary blood flow. Vigorous full-term infants are able to achieve this critical transition quickly and effectively, but this is much more problematic in very preterm infants who often fail to generate sufficient critical opening pressure to achieve adequate lung inflation because of their limited muscle strength, excessively compliant chest wall, limited surfactant pool, and incomplete lung development. The excessively compliant chest wall of the preterm infant fails to sustain the lung aeration that may have been achieved spontaneously or with positive-pressure ventilation. For the same reasons, these infants may be unable to generate sufficient negative intrathoracic pressure to effectively move lung fluid from the air spaces to the interstitium, lymphatics, and veins. Subsequent tidal breathing, both spontaneous and that generated by positive-pressure ventilation, occurs in lungs that are still partially fluid-filled and incompletely expanded. This situation leads to maldistribution of V T to a fraction of the preterm lung, a phenomenon that can generate excessive tissue stretching even when the V T is in an appropriate range that is generally considered safe.
The use of positive end-expiratory pressure (PEEP) and/or continuous positive airway pressure (CPAP) during the initial stabilization of preterm infants compensates for the excessively compliant chest wall and surfactant deficiency by stabilizing alveoli during the expiratory phase and has been shown to help establish functional residual capacity. Both Neonatal Resuscitation Program (NRP) and International Liaison Committee on Resuscitation (ILCOR) guidelines give a qualified endorsement to the use of PEEP/CPAP but cite a lack of high-quality randomized controlled trials (RCTs) to make a stronger recommendation. However, the physiologic rationale and experimental evidence from preclinical studies is so persuasive that this practice has become the de facto standard of care in much of the developed world and thus a RCT would be very difficult to undertake at this point. However, provision of end-expiratory pressure alone may not sufficiently address the inadequate muscle strength of the preterm infant or help clear lung fluid sufficiently rapidly to avoid regional volutrauma and atelectrauma, which can occur in minutes. Owing to the much greater viscosity of liquid compared with air, resistance to moving liquid through small airways is much greater than that for air, making the time constants required to move fluid through the airways much longer. These considerations support the concept that a prolonged (aka “sustained”) inflation applied soon after birth should be more effective in clearing lung fluid in the first minutes of life than the typical short inflations used during positive-pressure ventilation. Theoretically, rapid and effective lung recruitment that results in even distribution of V T immediately after birth should reduce VALI. Some evidence supports the theoretical advantages of sustained inflation in extremely preterm infants, but there is also potential for harm if excessive pressure is applied to relatively compliant lungs, resulting in lung overdistention. At present, the evidence that this intervention can reduce VALI remains inconclusive. Clear evidence regarding the optimal way to deliver a sustained ventilation is also lacking. Therefore routine use of the procedure cannot be recommended pending the outcome of an ongoing controlled clinical trial.
Noninvasive Respiratory Support
Avoiding MV is widely believed to reduce iatrogenic lung injury largely based on early cohort comparisons of CPAP versus MV, which suggested a dramatic reduction in the incidence of BPD. However, a series of more recent randomized trials showed a much smaller impact of noninvasive respiratory support. A meta-analysis of four recent trials that enrolled nearly 2800 preterm infants showed that BPD rates alone were not significantly different between infants randomly assigned to MV and those assigned to nasal CPAP (32.4% vs. 34.0%). However, the more important combined outcome of death or BPD (death and BPD are competing outcomes) showed a nearly 10% reduction (relative risk [RR] 0.91, 95% confidence interval [CI] 0.84–0.99) with a number needed to treat of 25. There was also a significant decrease in the duration of MV and a trend toward shorter duration of supplemental oxygen with early CPAP in two of the trials.
Nasal intermittent positive-pressure ventilation (NIPPV), if delivered effectively, may augment an immature infant’s inadequate respiratory effort without the complications associated with endotracheal intubation. This approach benefits from avoiding the use of an ETT, thus reducing the incidence of VALI and ventilator-associated pneumonia, and avoiding the contribution of postnatal inflammatory response to the development of BPD. A meta-analysis of several small single-center studies concluded that NIPPV was superior to CPAP in preventing extubation failure, especially when NIPPV was synchronized with the infant’s respiratory effort. However, a more recent large multinational randomized trial in extremely low-birth-weight infants failed to substantiate any benefits, showing no reduction in BPD, mortality, or the combined outcome. Lack of synchronization might be responsible for lack of benefit in some of these studies. Current evidence suggests that unsynchronized NIPPV rarely generates a measurable V T , likely because the vocal cords are closed except during active inspiratory effort. The cyclic inflation, however, does result in a higher mean airway pressure and it is likely that this higher distending pressure accounts for some of the observed benefits of NIPPV. The benefit of a higher continuous distending pressure may be more evident in infants with residual lung disease who require higher oxygen supplementation. This was evidenced in a small clinical trial during the extubation phase where a CPAP of 8 cm H 2 O was more effective in reducing extubation failure than 5 cm H 2 O. Nasal high-frequency oscillatory ventilation (HVOF) has also been explored by several groups, but as of this writing, the suggestion of efficacy is based only on case reports and small series. Noninvasive neurally adjusted ventilatory assist uses the electrical activity of the diaphragm to trigger ventilator inflations and is thus not affected by leakage of a noninvasive interface; it may thus be the best tool to deliver noninvasive ventilation in small preterm infants. However, definitive studies are still lacking.
Mechanical Ventilation Strategies
The goal of MV is to maintain acceptable gas exchange with a minimum of adverse effects and to wean the patient from invasive support as expeditiously as possible. There are many choices of devices and modes of ventilation with limited high-quality data to guide the clinician’s choice. Because of the wide range of clinical conditions, weights, and gestational ages of neonatal patients, there are no simple rules regarding indications for intubation and initiation of MV. “Standard” indications for intubation and initial ventilator settings often recommended in texts on this subject have limited utility. Instead, the choice of modalities of support and ventilation strategies should be guided by the specific underlying pathophysiologic considerations. Ventilators are nothing but tools in our hands that we need to use thoughtfully to optimize outcomes.
Basic Modes of Synchronized Ventilation
Despite the relative paucity of high-quality evidence, the question is no longer whether to use synchronized versus unsynchronized ventilation, but rather which modality of synchronization is optimal. Synchronization of ventilator inflations with the infant’s spontaneous breaths allows the clinician to avoid or minimize sedation/muscle relaxation and to maximize the patient’s own spontaneous respiratory effort. Encouraging spontaneous breathing during MV has clear advantages, but it makes managing MV more challenging for the clinician. The optimal use of assisted ventilation requires a clear understanding of the complex interaction between the awake, spontaneously breathing infant and the various modalities of synchronized ventilation discussed in depth in Chapter 16 . A key ingredient is an appreciation of the additive nature of the patient’s inspiratory effort and the positive pressure generated by the ventilator. The magnitude of the V T is determined by the transpulmonary pressure, the sum of the negative inspiratory effort of the infant, and the positive inflation pressure from the ventilator. Because the spontaneous effort is often sporadic and highly variable in a preterm infant, the resulting transpulmonary pressure and V T typically fluctuate substantially.
In the following text, we briefly review the common modes of synchronized ventilation, the important advantages of volume-targeted ventilation and the importance of the optimizing lung inflation.
Synchronized Intermittent Mandatory Ventilation
Synchronized intermittent mandatory ventilation (SIMV) is a basic synchronized mode that provides a user-selected number of inflations in synchrony with the infant’s breathing. If no spontaneous effort is detected during a trigger window, a mandatory inflation is delivered. Spontaneous breaths in excess of the set ventilator rate receive no support. In small preterm infants, this results in uneven V T s and high work of breathing because of the high airway resistance of the narrow ETT, hindered by the limited muscle strength and mechanical disadvantage of the infant’s excessively compliant chest wall. SIMV thus is not the optimal mode in this population. SIMV allows the operator to set the ventilator rate as well as inflation pressure and PEEP. Weaning is accomplished by gradual lowering of both rate and inflation pressure.
In neonatal applications, assist control (A/C) is a time-cycled, pressure-controlled mode that supports every spontaneous breath, thus resulting in more uniform V T and lower work of breathing than SIMV. The goal is for the infant and the ventilator to work together with every breath, resulting in lower ventilator pressure. A backup ventilator rate provides a minimum rate in case of apnea and should be set just below the infant’s spontaneous rate, usually at 40 inflations/min. A backup rate that is too low will result in excessive fluctuations in minute ventilation and oxygen saturations during periods of apnea. Because the infant controls the ventilator rate, gradual withdrawal of support is accomplished by lowering the peak inflation pressure, reducing the support given to each breath and thus encouraging the infant to gradually take over the work of breathing. Depending on the device used, some vigilance may be necessary to avoid autotriggering. Because ventilators do not commonly have a user-set upper limit to the cycling frequency, there is a risk of hyperventilation owing to autotriggering. This is seldom a problem when specialty neonatal ventilators with effective leak compensation are used, but even with those devices, water condensation in the expiratory limb can cause autotriggering.
Pressure-support ventilation (PSV) functions differently in adult and neonatal ventilators, a situation that frequently causes confusion. In specialty neonatal ventilators, PSV is a pressure-controlled mode that supports every spontaneous breath just like A/C but is flow-cycled. Flow cycling means that inflation is terminated when inspiratory flow declines to a preset threshold (usually 5%–20% of peak flow), eliminating inspiratory hold (prolonged inflation time) and thus providing more complete synchrony. Avoidance of prolonged inflation time likely results in less fluctuation in intrathoracic and intracranial pressure that is a consequence of infants exhaling during inspiratory hold. Additionally, PSV automatically adjusts inspiratory time ( T I ) in response to the changing lung mechanics of the patient. Changing from time-cycled A/C to PSV often results in a shorter T I and therefore lower mean airway pressure ( P AW ). Adjustment to PEEP may be needed to maintain P AW and avoid atelectasis. A substantial leak around the ETT may affect flow cycling. To prevent excessive inspiratory duration during PSV the time limit should be set appropriately. Similar to A/C, the risk of autotriggering is also present during PSV, especially when universal ventilators not specifically designed for neonates with uncuffed ETTs are used.
Similar to A/C, a backup rate will maintain a minimum inflation rate. In most devices PSV can also be used to support spontaneous breathing between low-rate SIMV to overcome the problems associated with inadequate spontaneous respiratory effort and high ETT resistance, or in a fully spontaneous mode (CPAP + PSV). When used with SIMV or with CPAP, PSV does not provide a backup mandatory rate, so a reliable spontaneous respiratory effort is required. Weaning from PSV when it is used as a primary mode is accomplished in the same way as for A/C. When used in conjunction with SIMV, both the SIMV inflation rate and peak inflation pressure (PIP) should be lowered, leaving the infant increasingly breathing spontaneously with only a modest level of PSV, at which point extubation should be attempted.
Choice of Assisted Ventilation Modes
To date, only three small randomized trials compared A/C and SIMV. These trials did not show clear advantages of one mode over the other with regard to weaning and duration of MV, though a recent meta-analysis by Greenough et al. found a strong trend favoring A/C with a shorter duration of weaning (mean difference −42.4 hours, 95% CI −94.35 to 9.60). In the absence of large randomized trials to provide the necessary evidence base and establish the superiority of one mode or the other, the choice between A/C and SIMV—the two most widely used modalities of synchronized ventilation—remains a matter of training, personal preference, or habit. Valid physiologic considerations and short-term studies indicate that modes that support every spontaneous breath are preferable in small preterm infants breathing through narrow ETTs. Documented benefits of A/C over SIMV include smaller and less variable V T , less tachypnea, more rapid weaning from MV, and smaller fluctuations in blood pressure. Despite indications that SIMV does not provide optimal support in extremely low-gestational-age newborns, many clinicians continue to use it both in the acute phase of illness and during weaning from MV, based on the common assumption that both rate and pressure must be weaned before extubation. Preference for the lower inflation rate of SIMV appears to be based on the superficially plausible assumption that a smaller number of inflations is inherently less damaging. This concept ignores the fact that the slower ventilator rate is accomplished at the expense of a larger V T , a variable that is clearly more injurious than ventilator rate (e.g., consider high-frequency ventilation) and is contradicted by available evidence from both preclinical and clinical studies. Many clinicians also mistakenly believe that assisting every breath prevents respiratory muscle training. This concern reflects a failure to understand that the infant contributes an increasing amount of inspiratory effort to the transpulmonary pressure during weaning, progressively taking over a greater proportion of the work of breathing as the ventilator inflation pressure is decreased, resulting in effective training of the respiratory muscles. As weaning progresses, the inflation pressure is decreased to the point where it only overcomes the added resistance of the ETT and circuit, at which point the infant can be extubated.
Pressure-controlled ventilation became the standard approach to MV in newborns because early attempts at volume-controlled ventilation using equipment available at the time in infants with uncuffed ETTs were disappointing. Pressure-controlled ventilation remains the dominant mode of ventilation in the neonatal intensive care unit because of its simplicity, ability to ventilate despite a large ETT leak, and improved intrapulmonary gas distribution owing to a decelerating gas flow pattern. The fear of pressure as the major culprit in lung injury is deeply ingrained and many clinicians continue to believe that directly controlling PIP is important, despite unequivocal evidence that volume, not pressure, is the key element in VALI. The danger of using pressure control is that V T is not directly controlled and changes when lung compliance is altered. Consequently, minute ventilation may change substantially without any adjustment of ventilator settings with lung volume recruitment or surfactant administration, resulting in hyperventilation and volutrauma. V T may become insufficient with decreased lung compliance, increased airway resistance, or absence of the patient’s spontaneous respiratory effort and lead to hypercapnia, tachypnea, increased work of breathing and oxygen consumption, agitation, fatigue, atelectasis/atelectrauma, and acidosis. Rapid, shallow breathing leads to inefficient gas exchange owing to increased dead space/ V T ratio. It is therefore evident that relatively tight control of V T is highly desirable and for this reason, volume-controlled ventilation remains the standard of care in adult and pediatric ventilation.
There are many ways to regulate V T delivery during MV. Modern ventilators now make it possible to use volume-controlled ventilation in newborns by allowing for measurement of exhaled V T at the airway opening, so that manual adjustment of set V T at the ventilator end of the patient circuit can be made to achieve a desired exhaled V T . More convenient are volume-targeted modes that are modifications of pressure-controlled ventilation that automatically adjust inflation pressure and/or time to achieve a target V T . It is likely that the key benefit of volume-targeted ventilation (VTV) rests in the ability to regulate V T , regardless of how that goal is achieved. With V T as the primary control variable, inflation pressure is automatically reduced as lung compliance and patient inspiratory effort improve, resulting in real-time weaning of pressure. Although volume targeting can also be achieved by close monitoring of V T and manual changes of PIP, VTV results in more timely adjustments. Volume guarantee (VG) is the most extensively studied form of VTV and the basic control algorithm is increasingly being adopted by ventilator manufacturers. Benefits documented in two recent meta-analyses that encompassed several different modalities of VTV include significant decrease in the rate of BPD, pneumothorax, severe intraventricular hemorrhage, and periventricular leukomalacia, as well as less hypocapnia and shorter duration of MV ( Table 18.1 ).