Pressure- versus volume-limited sustained inflations at resuscitation of premature newborn lambs

Initial resuscitation of preterm infants aims to establish a functional residual capacity (FRC) and facilitate initiation of gas-exchange within the immature lung. However, the initiation of ventilation after preterm birth may be a critical period of susceptibility for the development of lung and brain injury [1‐4].

Sustained inflation at birth is practiced in some centers for early establishment of FRC [5, 6]. A sustained inflation is recommended for the initial ventilation of apneic term and preterm infants in the recent European Resuscitation Council Guidelines [7]. An initial inflation sustained for 20 s fully aerates the preterm rabbit lung prior to the onset of tidal ventilation [8]. A sustained inflation also facilitates establishment of pulmonary blood flow immediately after birth and improves cerebral blood flow stability in preterm lambs compared to preterm lambs resuscitated without a sustained inflation [9]. The optimal way to deliver a sustained inflation is unknown.

Current neonatal resuscitation guidelines published by the European Resuscitation Council and American Heart Association, suggest that the initial inflations should be given by constant application of a predetermined inflation pressure [7, 10]. However, the lung volume achieved with a set pressure is dependent upon the mechanics of the respiratory system, the maturational stage of lung development and the volume of lung liquid remaining within the air spaces. Thus, application of a constant pressure may have variable efficacy in establishing a functional residual capacity. Acute over-distension resulting from a high sustained inflation volume delivered with a constant pressure could have injurious effects on the preterm lung and brain, whereas a low inflation volume would be ineffective in aerating the fluid-filled lung.

An alternative approach to sustained inflations would be to target delivery of a defined volume for the initial inflation. Whereas a defined initial delivered volume would potentially achieve more consistent volume inflation of the non-aerated lung, the inflation pressure required to achieve the predetermined volume would be variable and could result in exposure of the immature lung to potentially injurious high static inflating pressures, and possibly pneumothoraces.

We aimed to understand whether delivery of sustained inflations should be pressure- or volume-limited. Specifically, we asked how the method of sustained inflation influenced the homogeneity of aeration, the consistency of the functional residual capacity achieved immediately at the end of the sustained inflation, and the up-regulation of early markers of lung injury. We hypothesized that a sustained inflation that targets a preset delivered volume/kg birth weight will provide a more consistent FRC and more homogeneous aeration than is achieved using a pressure-limited sustained inflation, potentially reducing lung injury.

All experimental procedures were approved by the animal ethics committee of The University of Western Australia, in accordance with the National Health and Medical Research Council (Australia) Australian code of practice for the care and use of animals for scientific purposes (7^th Edition, 2004).

Sustained inflations may promote rapid establishment of FRC during the initial resuscitation and ventilation of preterm and term infants [8, 26]. However, the most effective and least injurious way to deliver a sustained inflation is not well understood. Three randomized controlled trials investigated the use of pressure limited SIs in preterm infants [27‐29]: only the study by te Pas and colleagues showed improved clinical outcome, including decreased need for intubation in the first 72 h, shorter duration of ventilatory support and reduced bronchopulmonary dysplasia. No studies investigating the use of volume limited sustained inflations in the delivery room are reported, likely due primarily to the limitations of current delivery room resuscitation devices. We investigated whether a pressure- or volume-limited sustained inflation was more beneficial for the establishment of FRC, improving aeration homogeneity and reducing lung inflammation and injury. We observed worse arterial oxygenation at 15 min in volume-limited sustained inflations, but no differences in regional or total end expiratory lung volumes, lung injury, or cerebral oxygenation variables between the two sustained inflation strategies.

The volume delivered to a preterm lung exposed to a predetermined pressure is determined by the lung compliance, as well as the duration of the inflation relative to the time-constant (resistance x compliance) of the respiratory system. The large volume of fetal lung fluid present in the preterm ovine airways at birth results in significantly elevated respiratory resistance compared to the air-filled lung, and consequently a prolonged time-constant of the respiratory system at delivery. Giving the same duration and depth of suctioning of the intubated lamb airway equalized the residual lung fluid volume between all lambs. Thus, it is likely that the predominant factor determining the variability of the delivered tidal volume at the end of the sustained inflation was lung compliance. While the mean tidal volume delivered by the sustained inflation was similar between strategies, considerably more variation in the sustained inflation volume was observed in PressSI lambs compared to the VolSI groups. Although VolSI lambs should have had no variability in sustained inflation volume, some of the lambs within the VolSI group did not reach the targeted 15 mL/kg as the peak PIP was limited to 50 cmH₂O to minimize pneumothoraces. The target of 15 mL/kg was chosen as this was demonstrated previously to be the functional residual capacity of lambs of this gestation [9] and was also based on our best approximation of the likely average volume delivered during the PressSI group in preliminary studies.

Our finding of increased variability in delivered volume using a pressure-limited SI at initiation of ventilation is consistent with recent reports of variable tidal volumes (0-30 mL) achieved with pressure-limited resuscitation of preterm infants [30‐32]. Whereas these measured tidal volumes in preterm infants were confounded by the influence of facemask leak, our measurements were free of leak due to the use of tracheal tubes with inflated cuffs. Given the known association between serial high tidal volumes at initiation of ventilation and development of lung injury in the newborn lung [3, 33‐36], we hypothesized that the use of a set sustained inflation pressure may inadvertently initiate lung injury as a result of high delivered volumes in preterm subjects with compliant lungs.

In contrast, a pressure-limited sustained inflation may be ineffective in achieving a functional increase in lung volume in the presence of poorly compliant lungs. We showed that a volume-limited sustained inflation overcomes the impact of compliance, to achieve a target inflation volume more reliably. However, the more consistent volume delivered with a volume-limited sustained inflation was achieved at the cost of more variable and on average 10 cmH₂O higher peak inspiratory pressures than were required during pressure-limited sustained inflation. Thus, a critical question is whether excessive inflation due to volume stretch, or excessive pressure to achieve appropriate inflation, is the most injurious strategy. Barotrauma is also known to cause injury in preterm lungs leading to BPD [37], however it is unlikely that barotrauma was responsible for the difference in oxygenation and early fall in FRC in the VolSI group as evidence by the pattern of pressure and volume delivery for each group (Figure 6): despite higher peak pressures in the VolSI group, peak pressures were achieved only momentarily with a fall in sustained inflation pressure after the first five seconds with lower median pressure over the course of the sustained inflation in VolSI group. The fall in inflation pressure may have contributed to the impaired oxygenation and early (<2 min) fall in the end-expiratory volume noted in the VolSI group on analysis of EIT recordings. Whereas no further volume recruitment occurred after the first 5 s in the VolSI group, recruitment was achieved more slowing and was ongoing in the PressSI group at the end of the 20 s sustained inflation period. for the remaining 15 s), the pattern of pressure and volume delivery to the lung differed between the two groups (Figure 6). Volume recruitment was achieved slowly with the PressSI group whereas the rapid volume delivery in the VolSI group resulted in very brief exposure to high inflation pressure with a subsequent fall off in inflation pressure. This pattern of rapid inflation (as opposed to rapid then sustained pressurization), may explain the subsequent fall in end-expiratory volume (EEV) noted on EIT analysis. However, this fall in EEV within the VolSI group may simply be due to a reduction in pressure of ~10 cmH₂O when the modality was switched to volume guarantee targeting 7 ml/kg. An alternative approach to volume-limited SI that aimed to achieve target volume at 20 s rather than at 5 s may have resulted in lower peak pressures but more sustained pressure throughout the recruitment process. Arguably, a target inflation volume delivered more slowly would achieve more homogeneous distribution with a lower risk of overdistension of fast time-constant lung units.

Both SI strategies significantly elevated HR and arterial pressure above the fetal values by the end of the sustained inflation (20 s) suggesting that both SI strategies were suitable at transitioning the fetal circulation to that of the newborn as shown previously [9, 38]. Interestingly, the VolSI group elevated HR and arterial pressure faster (by 5 s) than the Press SI group (Figure 3). This likely occurred due to the higher lung volume of the lung during the first 10 s in VolSI lambs, which would clear lung liquid faster, increase aerated regions of the lung, establish functional residual capacity and trigger the hemodynamic transition at birth [8, 26, 38]. It is unlikely that the more rapid hemodynamic transition observed in the VolSI group would infer long-term benefits.

The 131 d lamb is approximately similar to 34-36 w GA in the human infant – a group that is still prone to respiratory distress due to surfactant deficiency. We demonstrated previously that naïve (non-steroid) lambs at 130-133 d gestation are surfactant deficient [39] with a negligible pool of saturated phosphatidyl choline and with little chance of postnatal survival without antenatal exposure to corticosteroids and significant postnatal intervention. The initiation of resuscitation of preterm lambs with even a few serial high volume breaths is sufficient to initiate a pulmonary and systemic inflammatory response, potentially leading to adverse long-term consequences of neonatal lung disease [3, 33‐36]. Given the short study period, we investigated only early markers of lung inflammation and injury to determine whether volume- or pressure-limited sustained inflations initiated lung inflammation and injury. Total protein content in the bronchoalveolar lavage fluid is a measure of lung injury [40]. We observed no differences in total protein content between SI strategies. EGR-1, CTGF and CYR-61 mRNA expression are sensitive early markers of the degree of lung injury, increasing within 15 min after lung injury induced by ventilation of preterm lambs, with amplified expression in response to injurious ventilation [24]. Our failure to observe differences in mRNA levels of any these early-response genes suggests that there was no difference between pressure- or volume-limited sustained inflations in early markers of lung inflammation or injury. It is possible that the 15 min exposure was too short to affect a demonstrable rise in mRNA expression of any early response markers. However, longer duration studies are also compromised for assessment of lung injury, as it is difficult to delineate the inflammatory response to the sustained inflation from the inflammatory response to the subsequent ventilation strategy. We did not collect lung for histological assessment of injury, but given the short duration of ventilation, it is unlikely that structural differences would be evident by 15 min.

We demonstrated the efficacy of volume-limited sustained inflations for lung recruitment in preterm lambs. Sustained inflation procedures resulted in significant variability in either the delivered volume (PressSI) or pressures (VolSI). The VolSI strategy increased heart rate and arterial pressure sooner than the PressSI strategy, but had worse oxygenation at 15 min. There were no differences observed in physiological variables after 30 s or early markers of lung inflammation and injury between different sustained inflation strategies. While our findings suggest that both volume- and pressure-limited sustained inflations will obtain similar outcomes in preterm populations, the results need to be verified in preterm infants once devices suitable for delivery of volume-limited sustained inflations are available for infants.