Authors: Anton H. van Kaam
Categories: Review, Bronchopulmonary dysplasia, High-frequency ventilation, Lung protective ventilation, Preterm infants, Volume guarantee, Volume-targeted ventilation
Source: Neonatology
Doi: 10.1159/000539346
Despite the increasing use of non-invasive support modalities, many preterm infants still need invasive mechanical ventilation. Mechanical ventilation can lead to so-called ventilator-induced lung injury, which is considered an important risk factor in the development of bronchopulmonary dysplasia. Understanding the concepts of lung protective ventilation strategies is imperative to reduce the risk of BPD.
Overdistension, atelectasis, and oxygen toxicity are the most important risk factors for VILI. A lung protective ventilation strategy should therefore optimize lung volume (resolve atelectasis), limit tidal volumes, and reduce oxygen exposure. Executing such a lung protective ventilation strategy requires basic knowledge on neonatal lung physiology. Studies have shown that volume-targeted ventilation (VTV) stabilizes tidal volume delivery, reduces VILI, and reduces BPD in preterm infants with respiratory distress syndrome. High-frequency ventilation (HFV) also reduces BPD although the effect is modest and inconsistent. It is unclear if these benefits also apply to infants with more heterogeneous lung disease.
Understanding basic physiology and the concept of ventilator-induced lung injury is essential in neonatal mechanical ventilation. Current evidence suggests that the principles of lung protective ventilation are best captured by VTV and HFV.
Keywords: Volume-targeted ventilation, High-frequency ventilation, Preterm infants, Bronchopulmonary dysplasia, Lung protective ventilation, Volume guarantee
Newborn infants, especially those born prematurely, are at an increased risk of respiratory failure. The main causes for respiratory failure are a compromised lung function or an impaired control of breathing. As a result, many preterm infants require respiratory support to ensure adequate gas exchange. Respiratory support can be provided non-invasively or invasively. During non-invasive respiratory support, a continuous distending pressure (CDP) and/or supplemental oxygen is delivered via a nasal interface to the lungs. The infant remains in control of breathing and delivers a large part of the work. If this is insufficient, invasive respiratory support should be started following placement of an endotracheal tube. During invasive support a ventilator takes over a large part of the work and control of breathing. Although invasive mechanical ventilation can be life saving, it also increases the risk of (secondary) lung injury and the development of bronchopulmonary dysplasia (BPD). As BPD has serious long-term consequences for pulmonary and neurological development, avoiding or reducing lung injury during mechanical ventilation should have a high priority in ventilation management.
Knowledge on the mechanisms by which mechanical ventilation causes lung injury, also referred to as ventilator-induced lung injury (VILI), has been mainly obtained from pre-clinical studies using different animal models. These studies identified several risk factors for VILI. First, a high end-inspiratory lung volume will cause overdistension of airways and sacculi/alveoli. Not the pressure but the delivered volume is the cause of overdistension and for this reason this risk factor is also referred to as volutrauma. Although volutrauma is most often caused by applying high tidal volumes, it can also be caused by delivering smaller tidal volumes on top of a high end-expiratory lung volume or functional residual capacity as both will result in end-inspiratory overdistension [1]. The second important risk factor is the presence of atelectasis during mechanical ventilation, also referred to as atelectotrauma. The presence of atelectasis has two important consequences. First, some of the sacculi/alveoli will open during the inspiratory phase when a higher peak inflation pressure is applied to the lungs and collapse during the expiratory phase when a lower positive end-expiratory pressure (PEEP) is applied. This repetitive opening and collapsing results in lung injury [2]. Second, some of the poorly compliant lung regions will remain closed during both the inspiratory and expiratory phase of mechanical ventilation. As a result, the applied tidal volume will be redistributed to the open lung units, resulting in regional overdistension (volutrauma) [3]. Finally, ventilating with high fractions of inspired oxygen (FiO2) can cause oxidative stress, which also results in lung injury [4]. This is especially true in preterm infants who have impaired abilities to increase antioxidant enzymes in response to hyperoxia [5].
The risk factors mentioned above will result in an increase in alveolar-capillary permeability with leakage of fluids and proteins into the alveolar space, surfactant inactivation, and will trigger an inflammatory response (biotrauma) [6, 7]. Interestingly, the condition of the lungs at the start of mechanical ventilation will impact the severity of VILI. Animal studies have shown that a compromised surfactant system and exposure to pre- or postnatal inflammation will augment VILI [8–10]. As these conditions are often present in preterm infants, this makes this population extremely vulnerable for VILI.
Both animal and human studies have shown that mechanical ventilation not only results in lung injury but may also cause damage to the developing brain. Animal studies showed that on a microscopical level, brain injury already occurs within several days of mechanical ventilation [11]. Studies in preterm infants have shown that the protracted ventilation is an increased risk factor for neurodevelopmental impairment and that this risk is proportional to the duration of mechanical ventilation [12]. Part of this association between (duration of) mechanical ventilation and neurodevelopmental impairment is probably mediated by BPD as this complication is associated with both (prolonged) mechanical ventilation and adverse neurodevelopmental outcome [13].
Based on the known risk factors for VILI, we can define the basic principles of a lung protective ventilation strategy. First, we should avoid delivering high tidal volumes as these can result in saccular/alveolar overdistension. Although it remains unclear what the optimal tidal volume is in preterm infants, there is general consensus that the tidal volume should probably not exceed 7 mL/kg unless ventilation is seriously compromised due to an increase in anatomical or alveolar dead space. Second, if there is atelectasis this should be resolved by recruitment, thereby minimizing repetitive opening and collapsing of lung units and regional overdistension. Third, high FiO2 leading to oxidative stress should be avoided as much as possible. Finally, if surfactant deficiency is present this should be treated with exogenous surfactant.
It is important to emphasize that lung physiology may be compromised in different ways depending on the underlying lung disease that causes respiratory failure. It is especially important to assess if the underlying lung disease is characterized by atelectasis because only then a recruitment procedure is justified. Performing a recruitment procedure in the absence of atelectasis will lead to overdistension and aggravate VILI.
Lung recruitment requires specific knowledge of lung physiology. First, the concept of intrapulmonary right-to-left shunt is used to guide the process of lung recruitment. Intrapulmonary right-to-left shunt is characterized by perfusion of non-ventilated lung units (atelectasis), a condition that results in hypoxemia. Opening collapsed lung units by increasing airway pressure will reverse intrapulmonary right-to-left shunt, improve oxygenation, and allow for a reduction in the fraction of inspired oxygen (FiO2) needed to reach normal oxygenation. In this way, the FiO2 becomes a reflection of the degree of atelectasis present in the lungs, with a low FiO2 (<0.30) indicative of optimal recruitment. Second, based on the law of Laplace, the pressure needed to open up a sacculus/alveolus is higher than the pressure needed to keep it open. This means that once the lung is recruited, stabilization can be achieved with lower airway pressure [14]. Lowering the airway pressures after the lungs are recruited is therefore imperative to avoid overdistension.
It has been suggested that lung protective ventilation should start in the delivery room as soon as the baby is born. At that time, the lungs are still filled with fluid, which needs to be cleared and replaced by air. It is important to acknowledge that this condition at birth is different from infants already several hours old, who have a compromised (air-filled) functional residual capacity due to alveolar/saccular collapse. Nevertheless, both animal and human studies have indicated that the lungs are vulnerable for VILI shortly after birth [15, 16]. However, the ideal ventilatory support strategy to limit lung injury in the delivery room is still unclear. Based on the potential advantages of lung recruitment in patients with respiratory distress syndrome (RDS), several studies have investigated the efficacy and safety of lung recruitment within the first minutes after birth to clear lung fluid and establish an air-filled functional residual capacity. A systematic review of these studies showed that applying a so-called sustained inflation to recruit the lungs at birth does not lead to improved outcome and could potentially be harmful [17]. Therefore, a sustained inflation is currently not recommended. Different ventilation modalities can be used to apply a lung protective ventilation strategy when infants with respiratory failure in the NICU require invasive mechanical ventilation.
Historically, newborn infants have been ventilated with time-cycled pressure-limited ventilation. During this modality, the delivered pressure above PEEP is constant and the delivered tidal volume is highly variable. More advanced ventilators capable of measuring small tidal volumes via a flow sensor placed at the proximal end of the patient circuits made it possible to introduce volume-targeted ventilation (VTV) in neonatal care. During VTV, the ventilator targets a preset tidal volume and the necessary pressure amplitude to deliver this volume varies breath by breath. VTV therefore has the potential to minimize delivery of excessive tidal volumes, which can lead to volutrauma. Basically, there are two modalities that qualify as VTV volume-controlled ventilation and volume guarantee (VG) ventilation. VG ventilation is the most commonly used VTV mode in newborn infants. During VG ventilation, the patient receives pressure-limited inflations and the delivered expiratory tidal volume is measured after each inflation. If the tidal volume deviates from the preset target tidal volume, the ventilator adjusts the peak inflation pressure breath by breath in an attempt to match the preset tidal volume as much as possible. Studies in preterm infants have shown that VG ventilation stabilizes the delivered tidal volume and reduces exposure to excessive tidal volumes [18]. Furthermore, the inflammatory response in the lungs is dampened compared to regular pressure-limited ventilation [19]. A meta-analysis of mostly small randomized controlled trials comparing different modes of VTV to pressure-targeted ventilation showed a reduction in time on mechanical ventilation, pneumothoraces, hypocarbia, and the combined outcome death or BPD, all favoring the VTV mode [20]. It is important to acknowledge that all of these trials were performed in preterm infants with RDS, which is considered a relatively homogeneous lung disease. How these results translate to older preterm infants with more heterogeneous lung disease is unclear and needs further research.
An alternative strategy to limit VILI is to accept high carbon dioxide levels (permissive hypercapnia) in an attempt to limit the delivered tidal volume. Several randomized controlled trials have compared permissive hypercapnia to targeting normal carbon dioxide levels in ventilated preterm infants. None of these trials was able to show a clear benefit of permissive hypercapnia on the outcome BPD [21]. A more in-depth analysis of the results showed that the contrast in carbon dioxide levels between the groups was considerably smaller than intended and the delivered tidal volume was similar in both groups, which may explain the negative results of these trials.
As previously mentioned, atelectotrauma can also contribute to VILI. Setting what is considered a “safe” tidal volume between 4 and 7 mL/kg can still lead to regional overdistension if atelectasis is present in the lungs. Despite the fact that animal studies have shown that resolving atelectasis via a recruitment maneuver and stabilizing unstable lung units with higher levels of PEEP attenuates lung injury, this has not been substantiated in human studies [22]. In fact, to date only one study has investigated the impact of lung recruitment with higher PEEP levels in preterm infants with RDS. This study showed improved oxygenation in the recruitment group, but the small sample size did not allow for meaningful analysis of more important clinical outcomes [23].
Finally, conventional mechanical ventilation requires synchronization between the spontaneous efforts of the infants and the mechanical inflations of the ventilator. A systematic review of trials comparing synchronized to non-synchronized mechanical ventilation showed that synchronized ventilation reduced the duration of ventilation and the risk of air leaks, although there was no effect on BPD [24].
The basic principles of high-frequency ventilation (HFV) are shown in Figure 1. During HFV, a CDP is applied to the lungs and superimposed on this CDP small pressure swings (delta pressure) are delivered in a frequency of 10–15 Hz, resulting in oscillatory volumes between 1.5 and 2.5 mL/kg. Although these small volumes limit the risk of volutrauma, animal studies have shown that the full benefit of HFV on VILI reduction will only be reached if atelectasis is also resolved by performing a recruitment maneuver and stabilizing the open lung units with the lowest possible CDP [25]. In contrast to conventional mechanical ventilation, the physiological principles of lung recruitment can be successfully applied in preterm infants with RDS and more mature infants with heterogeneous lung disease [26, 27].
Fig. 1. Basic principles of HFV. A continuous distending pressure (CDP) is applied to the lungs and on top of this pressure swings are delivered with a preset delta pressure. Each oscillation has an inspiration time (Ti) and expiration time (Te) and the sum of these is the cycle time. As indicated in the figure, at a frequency of 10 Hz (600/min) the cycle time will be 0.1 s.
Despite the promising animal data, randomized controlled trials comparing HFV to conventional mechanical ventilation in preterm infants with RDS have shown, at most, a modest and inconsistent reduction in the BPD, in favor of HFV [28]. A more in depth analysis of these trials revealed considerable trial differences in patient characteristics and the ventilation strategies during both HFV and CMV [29]. Based on these differences, it has been suggested that HFV is probably most effective when applied to preterm infants with more severe RDS, when combined with an open lung ventilation strategy, and when applied during the entire ventilation period [30, 31]. Similar to VG ventilation, the effect of HFV on lung protection has so far only been studied in preterm infants with RDS and not in older preterm infants with more heterogeneous lung disease.
The association between invasive mechanical ventilation and the development of BPD has been well established and studies have shown that the odds for developing BPD already increase if the duration of invasive mechanical ventilation exceeds 24 h [32]. So the best way to avoid VILI and reduce the risk of BPD is to manage preterm infants with (imminent) respiratory failure on non-invasive support. In case non-invasive management fails and the infant needs to be intubated, a policy of aggressive weaning and extubation should be pursuit. There are several interventions, such as treatment with caffeine or postnatal steroids, that can improve the chances of successful weaning and extubation [33, 34]. Pro-active non-invasive respiratory management reduces the number of babies needing invasive mechanical ventilation and in those that do need to be intubated, the duration of mechanical ventilation is reduced [35]. In terms of outcome, these infants have a reduced risk of developing BPD and have improved neurodevelopmental outcome.
As mentioned previously, the focus of lung protective ventilation research has almost exclusively been on preterm infants with RDS. However, with new and improved performance of non-invasive respiratory support modalities and the introduction of less-invasive surfactant administration, the majority of preterm infants are no longer intubated and mechanically ventilated during the first 72 h of life. Instead, infants are intubated after the first week of life for impaired control of breathing or systemic inflammatory diseases (sepsis, necrotizing enterocolitis). At that time, the underlying lung disease is often heterogeneous in nature, which increases the risk of regional overdistension. Future studies should assess if the benefits of VTV and HFV are also present in these underlying lung conditions.
Unlike HFV, optimization and stabilization of end-expiratory lung volume using lung recruitment and higher levels of PEEP during conventional mechanical ventilation has been poorly studied. Future studies need to investigate the efficacy and safety of such a strategy.
Most modern ventilators allow for automated control of oxygen delivery based on the measured oxygen saturation. Cross-over studies have shown that this results in better oxygen targeting with less time spent in hyperoxia [36]. Future studies should investigate if this also results in less oxidative stress and, more importantly, a reduction in BPD and retinopathy of prematurity.
BPD remains an important complication of preterm birth and VILI is an important risk factor. Although preterm infants are increasingly managed on non-invasive support, many still require intubation and mechanical ventilation at some point during their admission. The duration of mechanical ventilation should be reduced as much as possible and during mechanical ventilation a lung protective strategy should be pursued. Such a lung protective strategy should optimize lung volume by resolving atelectasis and avoid overdistension caused by high tidal volumes. Current evidence suggests that VTV and HFV result in some degree of lung protection, at least in preterm infants with RDS. Future studies are needed to assess the efficacy and safety of these modalities in infants with more heterogeneous lung disease.
Dr. van Kaam received financial support for research projects, travel expenses, and lecture fees from Chiesi Pharmaceuticals and Vyaire Medical.
This study was not supported by any sponsor or funder.
Anton van Kaam wrote the manuscript.
This study was not supported by any sponsor or funder.