Published ahead of print on July 19, 2007, doi:10.1164/rccm.200701-051OC
© 2007 American Thoracic Society doi: 10.1164/rccm.200701-051OC
Brief, Large Tidal Volume Ventilation Initiates Lung Injury and a Systemic Response in Fetal Sheep1 Division of Pulmonary Biology, Cincinnati Children's Hospital Medical Center, Cincinnati, Ohio; 2 School of Women's and Infants' Health, The University of Western Australia, Perth, Australia; and 3 Department of Pediatrics/Neonatology, Academisch Ziekenhuis Maastricht, Maastricht, The Netherlands Correspondence and requests for reprints should be addressed to Alan Jobe, M.D., Ph.D., Cincinnati Children's Hospital Medical Center, Division of Pulmonary Biology, 3333 Burnet Avenue, Cincinnati, OH 45229-3039. E-mail: alan.jobe{at}cchmc.org
Rationale: Premature infants are exposed to potentially injurious ventilation in the delivery room. Assessments of lung injury are confounded by effects of subsequent ventilatory support. Objectives: To evaluate the injury response to a brief period of large tidal volume (VT) ventilation, simulating neonatal resuscitation in preterm neonates. Methods: Preterm lambs (129 d gestation; term is150 d) were ventilated (VT = 15 ml/kg, no positive end-expiratory pressure) for 15 minutes to simulate delivery room resuscitation, either with the placental circulation intact (fetal resuscitation [ FR]) or after delivery (neonatal resuscitation [NR]). After the initial 15 minutes, lambs received surfactant and were maintained with either ventilatory support (FR-VS and NR-VS) or placental support (FR-PS) for 2 hours, 45 minutes. A control group received no resuscitation and was maintained with placental support. Samples of bronchoalveolar lavage fluid, lung, and liver were analyzed.
Measurements and Main Results: Inflammatory cells and protein in bronchoalveolar lavage fluid, heat shock protein-70 immunostaining, IL-1 Conclusions: Ventilation for 15 minutes with a VT of 15 ml/kg initiates an injurious process in the preterm lung and a hepatic acute-phase response. Subsequent ventilatory support causes further increases in some injury indicators.
Key Words: resuscitation • premature • bronchopulmonary dysplasia • positive end-expiratory pressure • volutrauma
Interventions during the birth transition from fetus to newborn may affect long-term outcomes, particularly for the preterm infant. During resuscitation and stabilization, a depressed infant is exposed to a variety of potentially harmful interventions, such as oxygen, barotrauma or volutrauma, and poor temperature control. Acute injury of the adult lung results from mechanical ventilation–induced volutrauma and biotrauma (1). In the preterm infant, large tidal volumes (VTs) can cause lung injury and may contribute to the progression of acute lung injury to bronchopulmonary dysplasia (2). Large VT ventilation followed by routine ventilatory care causes epithelial damage, which allows proteins to leak into the airspaces and to inhibit surfactant function (3–5). Excessive stretch of the preterm lung also induces proinflammatory cytokines and an influx of inflammatory cells into the airspaces (6, 7). Although the damage may occur with resuscitation at birth, indicators of lung and systemic injury take hours to develop. In contrast to the adult lung (8), there is very little information about how a preterm newborn lung develops an injury response to high VT ventilation. Current guidelines for resuscitation of the preterm infant have no sound experimental basis and have not been tested in randomized clinical trials (9). In the clinical environment, it is not feasible to measure the effects of the elements of the ventilation maneuvers, such as VT, positive end-expiratory pressure (PEEP), and inspiratory time, on the newborn lung. The elements of the ventilation maneuvers are used simultaneously to stabilize the infant. Because the preterm lung is so easily injured (10), the continued ventilatory support that follows the initial ventilatory stabilization may further increase the injury. We describe a fetal sheep model for the evaluation of the elements of the initiation of ventilation separately from the need for continued ventilatory support. By maintaining placental support after the initial ventilation maneuver, indicators of injury can be evaluated without the need for sustained ventilation. By comparing the fetal intervention with similar interventions in newborn lambs, we have characterized the initial injury and measured the effects of continued ventilation on that initial injury.
The online supplement includes additional details on methods used. The animal studies were performed in Western Australia; the Cincinnati Children's Hospital and the Western Australia Department of Agriculture approved the animal use procedures. Time-mated Merino ewes with singletons or twins at 129 days gestation (term, 150 d) were randomized into four groups (Table 1).
Fetal Resuscitation and Placental Support After maternal hysterotomy, the head and chest of fetus were exteriorized, the fetus was intubated, and lung fluid removed. The fetus was ventilated (rate, 30 breaths/min; inspiratory time, 1 s) with heated and humidified air using escalating VT and no PEEP, while the placental circulation remained intact. The target VT was 15 ml/kg at 15 minutes. The fetus was treated with 100 mg/kg surfactant (Survanta; Ross Laboratories, Columbus, OH) and was returned to the uterus. The lamb was delivered for tissue collection 2 hours and 45 minutes after the end of the ventilation period.
Fetal Resuscitation and Ventilatory Support
Neonatal Resuscitation and Subsequent Ventilatory Support
Control Animals
Lung Processing and Bronchoalveolar Lavage Fluid Analysis
Statistics
Description of Lambs The lambs' body weight at delivery was not different between groups (Table 2). The VTs achieved at 5 and 10 minutes into the initial ventilation period were similar for the three ventilated groups. The 15-ml/kg target was achieved for the FR-PS and FR-VS groups but was somewhat lower for the NR-VS group. The pressures required to achieve those VTs were close to 50 cm H2O, and the three groups had similarly low compliances at 15 minutes, indicative of lung immaturity.
The cord blood of the FR-VS lambs had a pH of 7.26 ± 0.01, a PaCO2 of 60 ± 1 mm Hg, and a PaO2 of 22 ± 1 mm Hg at the end of the 15-minute VT intervention, which indicates the lambs had not been oxygenated or hyperventilated as a result of the fetal ventilation. The arterial blood gas values for the NR-VS group at the end of the initial 15-minute ventilation period were pH 7.36 ± 0.03, PaCO2 42 ± 3 and PaO2 35 ± 4 mm Hg. The cord blood gas values at delivery for the FR-PS group were pH 7.17 ± 0.03, PaCO2 73 ± 4 mm Hg, and PaO2 6 ± 1 mm Hg. The respiratory acidosis resulted from anesthesia and surgery before delivery. The ventilated lambs had similar VT at 3 hours (Table 3), but the FR-VS group required lower peak inspiratory pressures to achieve a higher compliance and had better gas exchange than the NR-VS group. Lung function at a VT of 15 ml/kg was better in FR-PS and FR-VS groups than in the NR-VS group. The ventilatory pressures and blood gas values for the 2 hours and 45 minutes of ventilation are presented and discussed in the online supplement.
Protein and Inflammatory Cells in BALF An increase in protein in BALF is a global indicator of lung injury. Total protein in the BALF was fivefold higher in the FR-PS group than in the control group (Figure 1A), and was 11- and 14-fold higher than the control group in the FR-VS and NR-VS groups, respectively. Inflammatory cells in BALF also were greatly increased in FR-PS, FR-VS, and NR-VS groups (Figures 1B and 1C). BALF from control animals contained very few neutrophils. Neutrophil counts were 300-fold higher in the FR-PS group compared with the control group, and were elevated further in the FR-VS and NR-VS groups. Monocyte numbers in BALF were higher in the FR-PS, FR-VS, and NR-VS groups compared with the control group (Figure 1C). In contrast to neutrophil counts, monocyte counts were not higher after ventilatory support than after placental support.
Cytokines and Chemokines in the Lungs IL-1  mRNA increased 30-fold relative to control animals in the FR-PS group (Figure 2A) and was elevated further in the FR-VS and NR-VS groups, but IL-1 mRNA was not significantly higher in FR-VS and NR-VS groups than in the FR-PS group. A similar pattern of mRNA response was seen for MCP-1, although MCP-1 mRNA was significantly higher in the NR-VS group (115-fold higher than in the control group) compared with the FR-PS group (30-fold higher than the control group; Figure 2B). In contrast to IL-1 and MCP-1, IL-6 and IL-8 mRNA was induced only about threefold in the FR-PS group relative to the control group but was 20- to 100-fold higher in the FR-VS and NR-VS groups compared with the control group (Figures 2C and 2D).
Acute-Phase Responses in Lung HSP-70 was higher in the FR-PS group than in the control group, and was further elevated in the lung parenchyma of the NR-VS group (Figure 3). Messenger RNA for SAA is an acute-phase reactant that is expressed in the fetal sheep lung (19). SAA mRNA increased in the lung after the resuscitation maneuver in all groups (Figure 4). The mRNA for TLR4 and TLR2 also was higher in FR-PS, FR-VS, and NR-VS groups compared with the control group (Figures 5A and 5B), and was higher in FR-VS and NR-VS groups compared with the FR-PS group.
Expression of SAA and TLR mRNA in Liver The inducible form of liver SAA was fivefold higher in the FR-PS, FR-VS, and NR-VS groups than in the control group (Figure 4B). TLR4 and TLR2 mRNA in the liver was higher than in the control group in all resuscitation groups (Figures 5C and 5D). TLR4 mRNA was higher in the FR-VS and NR-VS groups than in the FR-PS group.
These translational experiments were targeted at the poorly studied clinical problem of how to optimize initiation of ventilatory support for the preterm infant. The International Guidelines for Neonatal Resuscitation (ILCOR) give essentially no guidance for initial ventilation of the preterm infant. The guidelines simply state that ventilation should be sufficient to achieve a clinical response and that CPAP may be helpful (9). Preterm lungs are easily injured, because the lungs are fluid-filled, an FRC needs to be established, initial inflation is nonuniform and may require long and variable time constants to achieve inflation, and the total lung capacity is low relative to the adult lung (10). Surfactant deficiency compromises the establishment of an FRC and results in the need for higher ventilatory pressures (20). Mechanical ventilation is also complicated by the following: masks or tracheal tubes that frequently leak (21); no practical way to measure VTs; and a high level of anxiety of the resuscitation team, which frequently results in hyperventilation of the preterm infant (22). In one report, skilled clinicians, instructed to follow ILCOR guidelines for resuscitation, used mean VTs of 18 ml/kg and pressures of approximately 40 cm H2O to ventilate preterm lambs (23).
Large VTs injure the preterm lung, as demonstrated experimentally using preterm sheep and rabbits (3, 5, 24), but the mechanisms causing the injury have not been ascertained to the same degree as in the adult lung. Uhlig (25) described the types of injuries caused by lung stretch, such as necrosis, stress failure of epithelial and endothelial barriers, overdistension without tissue injury, and effects on the vasculature. These injuries can result in biotrauma (cytokine release, inflammatory cell recruitment) and pulmonary edema over several hours in the adult lung (26–28). Large VTs delivered to the preterm lung without PEEP cause structural injury to the airspaces and epithelial leaks that result in interference with surfactant function (5). Proinflammatory cytokine mRNA and inflammatory cells appear in the lung if ventilation without PEEP continues for 2 hours (7). These effects are similar to those observed in adult lungs, but it is risky to assume that the injury responses of the preterm lung will parallel those in the adult lung. For example, the sentinel cell for many types of injury in the adult lung is the alveolar macrophage (29). The preterm lung and airspaces contain a few immature monocytes and essentially no mature macrophages (30). Also, the preterm sheep lung does not respond to tumor necrosis factor- We developed this model of fetal resuscitation followed by recovery in utero to separate the injury caused by the initiation of ventilation from subsequent mechanical ventilation required to support the preterm animal. The recovery period of 2 hours, 45 minutes, in utero was enough time for markers of lung injury to increase. The initial resuscitation maneuver was compared between groups that received the same maneuver, either as fetuses or newborns, and then were supported by the placenta or mechanical ventilation for the same period of time, to assess the effects of postnatal ventilation on the initial resuscitation maneuver. The maneuver was intended to be sufficient to achieve lung injury. In previous experiments by our group, preterm fetal lambs ventilated with VTs of about 7 ml/kg for 15 minutes followed by a period of placental support did not demonstrate much injury (34). Although we used higher pressures in these experiments and targeted a VT of 15 ml/kg by 15 minutes, the volumes were less and the pressures were similar to injury models in adult and young animals (1, 35, 36). The fetuses tolerated the 15 minutes of partial exteriorization well. In pilot experiments, we found other fetuses to be healthy at delivery at 24 hours after the fetal ventilation procedure, suggesting that the model may also be good for studying the evolution and resolution of lung injury in the preterm over extended time periods. A target VT of 15 ml/kg is high relative to the 4 to 6 ml/kg used to ventilate the preterm human. Preterm lambs differ from preterm infants because the preterm lambs breathe spontaneously on CPAP with a VT of about 8 ml/kg, and a VT in this range is needed to normalize PaCO2 values in mechanically ventilated preterm lambs (12). Therefore, the injurious VT was only twice the optimal VT for preterm sheep. A doubling of the normal VT would cause minimal injury in the healthy adult lung or in a postnatal rodent with incomplete alveolarization (35). A VT as large as 50 ml/kg for 4 hours did not cause lung injury in healthy pigs (37).
Ventilation of the preterm lungs with a VT of 15 ml/kg and no PEEP caused a large recruitment of monocytes and neutrophils, and an increase in total protein, in the lungs. Expression of the IL-1 The confounding effects of continued ventilation are demonstrated by the further increase in markers of lung inflammation, even after surfactant treatment and normal VT ventilation. The increase in some injury mediators and markers in the FR-VS group relative to the FR-PS group demonstrates the additive effect of subsequent ventilation to resuscitation injury of the preterm lung. The differences in compliance and some cytokines between the FR-VS and NR-VS groups indicate increased injury and suggest that routine clinical management has the potential to be very injurious. The difference between the NR-VS and FR-VS groups, with respect to decreased arterial pH, decreased lung compliance, and increases in IL-6 and IL-8 mRNA after 2 hours and 45 minutes of ventilation, suggests a role of the fetal circulation or placental blood flow in modulating the initial injury. The fetal resuscitation maneuver did not cause increased oxygenation of the cord arterial blood, indicating that pulmonary blood flow remained low. In contrast, oxygenation increased and PaCO2 decreased with the neonatal ventilation maneuver, indicating increased pulmonary blood flow. This increased blood flow may have contributed to the increased injury. The recruitment of neutrophils to the BALF was rapid in these experiments. Neutrophils are the major mediators of the acute inflammatory process in most models of acute lung injury (39). Depletion of systemic neutrophils decreased ventilation- or endotoxin-induced lung injury in the adult lung (40, 41). Similarly, depletion of neutrophils decreased lung vascular injury and edema in preterm lambs (42). Ultimately, it will be important to learn how stretch signals neutrophil recruitment into the fetal lung. In the adult mouse lung, stretch and endotoxin have distinct signaling pathways that each result in neutrophil recruitment (43). The neutrophil sequestration caused by ventilation-induced lung injury is L-selectin dependent but CD-18 independent (44). In contrast, inhibition of intraamniotic endotoxin-mediated neutrophil recruitment to the lungs of fetal sheep with an anti–CD-18 antibody prevents fetal lung inflammation and injury, whereas IL-1–mediated neutrophil recruitment to the fetal sheep lung is CD-18 independent (45). Further research may identify a specific target to minimize neutrophil recruitment caused by stretch to the preterm lung. Together with neutrophil influx, similar increases in BALF monocytes were seen in all groups given the initial resuscitation maneuver. The similar monocyte influx, despite the large increase in MCP-1 in the NR-VS group, suggests that these monocytes are from the lung tissue and not enough time has elapsed for recruitment of systemic monocytes to the airspace. A provocative new result is the induction of mRNA for the acute-phase reactants serum amyloid and TLR2 and TLR4 receptors in both the lung and liver. We previously reported that 6 hours of mechanical ventilation increased SAA expression in the preterm and term lung and liver (16). Intratracheal endotoxin also induced serum amyloid to very high levels in the lung and liver of preterm lambs (16). In this experiment, just 15 minutes of high VT ventilation increased lung and liver serum amyloid expression, and subsequent ventilation did not further increase the response, indicating there is an immediate systemic response to perinatal lung injury. We recently cloned and characterized TLR2 and TLR4 expression in fetal sheep (46). Intraamniotic endotoxin induced TLR2 and TLR4 mRNA in the fetal lung. We now show that the fetal ventilation maneuver also induces these two TLRs in both the lung and liver. Although the TLR signaling pathway is complex, up-regulation of TLR mRNA could result in an increased inflammatory response. Mechanical ventilation that is not normally injurious can amplify mild lung injury caused by endotoxin in sepsis models (47, 48). Ventilation-mediated lung injury also releases proinflammatory mediators and endotoxin into the systemic circulation of adult and newborn animals and causes shock and multisystem organ failure (49–51). This possibility is a real risk for preterm infants because 50% or more of preterm infants born before 30 weeks' gestation have been exposed in utero to infection and/or inflammation (52), which results in inflammation of the fetal lung (53). Our results suggest a new mechanism for a synergistic interaction between mechanical ventilation and a systemic inflammatory response. Mechanical ventilation can simultaneously result in a systemic inflammatory response, as indicated by activation of acute-phase responses in the liver and increases in TLR mRNA, and deliver the mediators to those receptors. These relationships need to be further explored, and our novel fetal sheep model may be ideal for such studies. Our results demonstrate that brief high VT ventilation, similar to what may occur during resuscitation in the delivery room, can injure the preterm lung. Limitations to application of the results to the clinical environment result primarily from the differences between the healthy sheep and the preterm infant. Most preterm infants have been exposed to antenatal corticosteroids, which may inhibit inflammatory responses to ventilation-mediated injury (25). Many of these infants also have been exposed to chorioamnionitis, which could increase the injury response to mechanical ventilation (54). The style of ventilation we used included no PEEP because PEEP levels are poorly controlled in the delivery room and there is no standard for the use of PEEP in the delivery room (55). An inspiratory time of 0.7 seconds was used because there is also no standardized inspiratory time for neonatal resuscitation. The relatively long inspiratory time was used because of flow limitations in this large animal model. Surfactant was given after the initial ventilation period because it is not routine practice to give surfactant to preterm infants before ventilation is established (56). These clinically relevant variables can be tested using this animal model to evaluate the components of resuscitation and better inform the ILCOR guidelines for neonatal resuscitation.
The authors thank M. Kapp and Amy Whitescarver for their expert technical assistance. They thank Ross Laboratories for providing Survanta.
Supported by grant HD-12714 from the National Institute of Child Health and Development; a National Institutes of Health training grant (HD07541) to N.H.H.; a NHMRC Neil Hamilton Fairley postdoctoral fellowship to J.J.P. (no. 139160); an NHMRC R.D. Wright Career Development Award to T.J.M.M.; the Women's and Infants' Research Foundation; and Fisher and Paykel Healthcare, Auckland, New Zealand. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1164/rccm.200701-051OC on July 19, 2007
Conflict of Interest Statement: N.H.H. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. T.J.M.M.'s department has received research grants totaling approximately $75,000 over the previous 3 years, from Fisher and Paykel Healthcare. S.G.K. received a grant from Merck for $50,000 in 2006 for a phase I clinical trial. C.B. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. J.J.P.'s department has received research grants totaling approximately $75,000 over the previous 3 years from Fisher and Paykel Healthcare; J.J.P. was the recipient of an unrestricted education grant from Abbott, Australia, to King Edward Memorial Hospital to assist with conference travel; she is associated with the Women and Infants' Research Foundation, which has received research grant funding from Bunnel, Inc., for research into high-frequency ventilation. G.R.P. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. I.N. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. B.W.K. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript. A.H.J., in collaboration with T.J.M.M. and J.J.P., has received research grants and equipment donation from Fisher and Paykel Auckland, NZ, to support in part this research ( Received in original form January 10, 2007; accepted in final form July 17, 2007
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
AT A GLANCE COMMENTARY
(
$75,000 over 3 yr). 
B and is blocked by steroids. Am J Respir Crit Care Med 2001;163:711–716.