Mechanical ventilation of preterm babies increases survival but can also cause ventilator-induced lung injury (VILI), leading to the development of bronchopulmonary dysplasia (BPD). It is not known whether shear stress injury from gases flowing into the preterm lung during ventilation contributes to VILI.
Preterm lambs of 131 days’ gestation (term = 147 d) were ventilated for 2 hours with a bias gas flow of 8 L/min (n = 13), 18 L/min (n = 12) or 28 L/min (n = 14). Physiological parameters were measured continuously and lung injury was assessed by measuring mRNA expression of early injury response genes and by histological analysis. Control lung tissue was collected from unventilated age-matched fetuses. Data were analysed by ANOVA with a Tukey post-hoc test when appropriate.
High bias gas flows resulted in higher ventilator pressures, shorter inflation times and decreased ventilator efficiency. The rate of rise of inspiratory gas flow was greatest, and pulmonary mRNA levels of the injury markers, EGR1 and CTGF, were highest in lambs ventilated with bias gas flows of 18 L/min. High bias gas flows resulted in increased cellular proliferation and abnormal deposition of elastin, collagen and myofibroblasts in the lung.
High ventilator bias gas flows resulted in increased lung injury, with up-regulation of acute early response genes and increased histological lung injury. Bias gas flows may, therefore, contribute to VILI and BPD.
Citation: Bach KP, Kuschel CA, Hooper SB, Bertram J, McKnight S, et al. (2012) High Bias Gas Flows Increase Lung Injury in the Ventilated Preterm Lamb. PLoS ONE 7(10): e47044. doi:10.1371/journal.pone.0047044
Editor: Peter Rosenberger, University of Tübingen, Germany
Received: May 13, 2012; Accepted: September 10, 2012; Published: October 8, 2012
Copyright: © Bach et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The authors thank the Auckland Medical Research Foundation (grant number 81502, http://www.medicalresearch.co.nz/), the Health Research Council of New Zealand (grant number 06/082; http://www.hrc.govt.nz/), the National Research Centre for Growth and Development in New Zealand (grant number (grant number 06018FB; http://www.nrcgd.org.nz/), and the National Health and Medical Research Council of Australia (http://www.nhmrc.gov.au) for their financial support. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Preterm birth is the greatest cause of morbidity and mortality in newborn infants and rates of preterm birth are increasing worldwide. Up to 70% of babies born before 26 weeks’ gestation develop bronchopulmonary dysplasia (BPD), a debilitating lung disease that results in abnormal lung development and which is usually defined as an ongoing requirement for supplemental oxygen at 36 weeks’ postmenstrual age , . Histologically, BPD is characterized by poor alveolarisation , , abnormal elastin deposition , , , fibrosis , , mesenchymal cell hyperplasia ,  and abnormal capillary growth . Although the causes of BPD are unknown, the majority of infants with BPD received mechanical ventilation, which is considered to be a major contributing factor through the induction of ventilator-induced lung injury (VILI) to the immature lung .
Mechanisms underlying VILI are thought to include barotrauma , , volutrauma , , atelectotrauma  and biotrauma . Although this knowledge has led to advances in perinatal respiratory care , , , the incidence of BPD following extremely preterm birth has not decreased .
The potential role of rheotrauma in VILI, related to high gas flows in the delicate preterm airways, has not yet been investigated. Mechanical conventional ventilation involves forcible injection of gas into the lungs, achieved in most neonatal ventilators by a continuous or variable bias gas flow within the ventilator circuit. During inspiration, the expiratory valve closes and pressure builds in the ventilator circuit providing a pressure gradient for gas to flow into the lung. Neonatal ventilators applying a continuous bias gas flow are commonly set to flows of 6–12 L/min, but there is little evidence to support the use of these flows . Furthermore, it is often assumed that bias gas flow is directly related to inspiratory flow in the pulmonary airways. However, many factors influence inspiratory flow, including the type of airflow (laminar vs turbulent), the pressure gradient between the ventilation circuit and the airways as well as the resistance and compliance of the respiratory system.We have previously demonstrated that bias gas flow in term healthy lambs is inversely related to inflation time and that optimal ventilator efficiency was obtained at bias gas flows much lower than currently used in many neonatal intensive care units . It is not known, however, whether bias gas flows also affect lung injury or pulmonary dynamics in the very immature and poorly compliant lung. We report here that high bias gas flows affect cardiorespiratory parameters, decrease ventilation efficiency, result in histological changes consistent with lung injury and upregulate mRNA levels of the early injury response genes EGR1, CYR61 and CTGF which are known to play a pivotal role in adult human lung injury , ,  and which reliably indicate VILI in preterm lambs .
Materials and Methods
In vitro study.
The inter-relationships between gas flow into an artificial lung and (1) bias gas flow in the ventilator circuit, (2) respiratory system compliance and (3) respiratory system resistance, were examined using a mechanical ventilator (Babylog8000plus, Dräger Medical, Lübeck, Germany). Three test lungs of differing compliances (0.4, 0.8 and 1.3 mL/cmH2O) and four different airway resistances (no endotracheal (ET) tube and ET tube sizes of 2.5, 3.0 and 4.0) were used. Peak inspiratory pressure (PIP) was set at 35 cm H2O. Airway pressure and gas flow into the test lungs were measured directly from the ventilator and recorded continuously at 4 different bias gas flows (4, 8, 18 and 28 L/min).
Ethics statement: Experiments were approved by the University of Auckland Animal Ethics Committee, approval number AEC03/2006/R450. Ewes received betamethasone (Schering-Plough, Wellington, New Zealand, 11.4 mg intramuscularly) 48 and 24 h prior to delivery of lambs via hysterotomy at 131–133 days’ gestation (term = 147 days). Lambs (weight 4.8±0.1 Kg) were intubated with a size 4.0 cuffed endotracheal tube (Mallinckrodt, Athlone, Ireland) and ventilation was commenced with randomly assigned bias gas flows of 8 (n = 13), 18 (n = 12) or 28 L/min (n = 14). Umbilical arterial and venous catheters were placed and anesthesia induced with alfaxalone (JUROX Pty. Ltd, Rutherford, Australia, 4 mg/Kg) and maintained with midazolam (Roche, Auckland, New Zealand, 0.4 mg/Kg) and ketamine (Parnell Laboratories, Auckland, New Zealand, 4 mg/Kg). Muscle relaxation was maintained with pancuronium (AstraZeneca, North Ryde, Australia, 0.1 mg/Kg, intravenous).
Lambs were ventilated for 2 h (Babylog8000plus, Dräger Medical, Lübeck, Germany) with heated (37 °C) and humidified gas in pressure support ventilation + volume guarantee (PSV+VG) mode. In this mode, inflation is terminated when the inspiratory flow falls to <15% of the peak inspiratory flow, indicating that the lungs are nearly fully inflated. Settings were: tidal volume 7 mL/Kg; positive end expiratory pressure 6 cmH2O; maximum positive inspiratory pressure (PIP) 50 cmH2O; maximum inflation time (Ti) 0.8 s, and frequency 40 breaths/min (increased if PaCO2>60 mmHg or pH <7.25). Supplemental oxygen was administered if SaO2 was <90%. Pneumothoraces were promptly drained percutaneously.
Inspiratory and expiratory flow (MLT300L pneumotachograph; ADinstruments, Dunedin, New Zealand), blood pressure, oxygen saturation and ventilator parameters (Ventview, Dräger Medical) were measured and recorded continuously. Ventilator efficiency index (VEI) was calculated as 3800/[(PIP-PEEP)×ventilator rate×PaCO2]  and oxygenation index (OI) as mean airway pressure (cmH2O)×FiO2 (%)]/PaO2 (mmHg). Arterial blood gases were measured every 15 minutes. Full blood counts were performed at 0, 30, 60 and 120 min after commencing ventilation. Data from lambs that died during the experimental period were excluded from these analyses (numbers analysed: 8 L/min, n = 11; 18 L/min, n = 11; 28 L/min, n = 14).
At 2 hours lambs were euthanized with pentobarbitone (Provet NZ Pty Ltd, Auckland, New Zealand, 90 mg/Kg). Additional non-ventilated lambs (n = 8) were euthanized prior to delivery to act as unventilated controls. Lung tissue was collected as described previously .
mRNA levels of early growth response factor 1 (EGR1), cysteine-rich 61 (CYR61) and connective tissue growth factor (CTGF) in lung tissue were measured by quantitative real-time PCR as described previously, utilising the delta-delta CT method normalised to 18S rRNA as a housekeeping gene and expressed relative to mean levels of the gene of interest in control tissue . PCR data were available for all 8 control lambs and all experimental lambs for which physiological data were analysed (8 L/min, n = 11; 18 L/min, n = 11; 28 L/min, n = 14).
Lung tissue was prepared for histological and immunohistochemical analysis as described previously . Five lambs were randomly selected from each experimental group for histological analysis; all 8 control lambs were analysed. Elastin was identified using Hart’s resorcin-fuchsin stain (12) and collagen with Gordon and Sweet’s reticulum stain. Myofibroblasts, which play a role in connective tissue fibre deposition and secondary septal crest formation, were identified by immunohistochemical staining of alpha smooth muscle actin , . Proportion of tissue stained positive for elastin, collagen and alpha smooth muscle actin, percentage of lung occupied by tissue and secondary septal crest density were calculated in 5 random images per section as described previously  by a single observer (KPB) blinded to the experimental group. Proliferating cells, staining positive for Ki67 , were counted on 10 images per section and expressed as a proportion of the total number of cells to determine a labelling index.
Data that were not normally distributed (EGR1 and CTGF mRNA levels) were transformed by taking the cube root to approximate a normal distribution prior to analysis with parametric techniques. Data are represented as mean ± SEM. For the in vitro study, a non-parametric multivariate analysis was used to analyse the effects of bias gas flow, airway resistance and respiratory system compliance on gas flow into the artificial lung. Physiological data from the animal study were analysed in 10 minute epochs, commencing 5 minutes before each blood gas measurement, and were analysed by ANOVA, including analysis of effect of time. Histological data were analysed by nested ANOVA and mRNA levels by one-way ANOVA. The Tukey post-hoc test was used when appropriate. Data were analysed using JMP7 (SAS Institute Inc., Cary, NC, USA) or SPSS 14.0 (SPSS Inc., Chicago, IL, USA) and significance taken at p<0.05.
In vitro Study
All three variables (bias gas flow, airway resistance and lung compliance) significantly and independently affected the flow of gas into the test lung. Using the artificial lung with a compliance of 0.8 mL/cmH2O and minimal resistance, increasing the bias gas flow from 4 to 8, 18 and 28 L/min increased the average rate of pressure rise from 63.8 cmH2O/sec to 93.2, 157.0 and 161.4 cmH2O/sec, respectively (Fig. 1A). Thus, 200 msec after the onset of inflation, airway pressure was increased to 26%, 50%, 88% and 95% of the peak inflation pressure (PIP; 35 cmH2O), respectively, at the 4 different bias gas flows (Fig. 1A). As a result, airflow into the lungs increased from 3.5 to 5.0, 8.5 and 10.0 L/min, respectively (Fig. 1B) at these different bias gas flows.
(A) Effect of bias gas flow on the pressure wave in the ventilation circuit; dotted line indicates 200 msec after inflation onset and values indicate the pressure at this point, expressed as a percentage of the peak pressure. (B) Effect of lung compliance on gas flow into the lung at different bias gas flows. (C) Relationship between airway resistance and gas flow into the lung, measured at 3 different lung compliances and 4 different bias gas flows.
Reducing the compliance of the test lung by half (to 0.4 mL/cmH2O) reduced airflow into the lung by 30–37% to 2.4, 3.5, 5.7 and 6.3 L/min at ventilator bias flows of 4, 8, 18 and 28 L/min (p<0.0001; Fig. 1B), whereas increasing the compliance to 1.3 mL/cmH2O increased the airflow into the lung by 22–27% compared with the lung with a compliance of 0.8 mL/cmH2O, to 4.3, 6.1, 10.8 and 12.2 L/min at bias flows of 4, 8, 18 and 28 L/min, respectively (Fig. 1B). Similarly, the relationship between airway resistance and airflow into the lungs was highly significant, following a logarithmic curve (p<0.0001; Fig. 1C).
Effect of Bias Gas Flow on Cardiorespiratory Parameters
Three animals died from pneumothoraces during the 2-hour ventilation period, two from the 8 L/min group and one from the 18 L/min group. Physiological data are therefore presented for 11 lambs in the 8 L/min group, 11 in the 18 L/min group and 14 in the 28 L/min group.
The peak inspiratory gas flows measured at the wye piece by pneumotachograph in the 8, 18 and 28 L/min bias gas flow groups were 14.9±1.7, 20.8±1.7 and 28.2±1.4 L/min, respectively, confirming that bias gas flow does not accurately reflect gas flow into the lung. As the lambs were ventilated in the PSV+VG mode, in which inspiration ceases when flow reduces to <15% of peak inspiratory flow, the inflation time decreased with increasing bias gas flow (table 1). Peak inspiratory pressure was lowest, whereas tidal volumes and ventilator efficiency index were highest (table 1), in lambs ventilated with a bias gas flow of 8 L/min. Surprisingly, the rate of rise of inspiratory flow was greatest, the expiratory time was shortest and the expiratory flow was greatest in lambs ventilated at 18 L/min (table 1). Pulmonary compliance was similar for all animals (1.8–1.9 mL/cmH2O) but pulmonary resistance was highest, and tidal volume lowest in lambs ventilated at 28 L/min (table 1). To maintain pCO2 levels during the experiment, the ventilator rate was gradually increased in all animals, independently of bias gas flow (p<0.05; data not shown), resulting in a decrease in expiratory time. However, the PaCO2 was highest in lambs ventilated at 18 L/min (table 2), despite these lambs also having significantly higher ventilator rates and expiratory airflows (table 1).
Heart rate, but not blood pressure, was higher in the 18 L/min group compared with both other groups (table 2). PaO2 (table 2), but not mean airway pressure (table 1) or oxygenation index, was also highest in lambs ventilated at 18 L/min, although FiO2 (table 2) tended to be higher in these lambs. Pneumothorax rates tended to be higher in lambs ventilated at 18 L/min (66%) compared with lambs ventilated at 28 L/min (43%) and 8 L/min (31%) (table 2). Blood white cell and neutrophil counts were not different between groups, although blood white cell count at 60 and 120 minutes were decreased in all three groups compared with baseline values (p<0.01; data not shown).
Effect of Bias Gas Flow Rate on mRNA Levels of Acute Response Genes and on Pulmonary Histology
Mechanical ventilation caused significant up-regulation of CYR61, CTGF and EGR1 mRNA levels (control, n = 8; 8 L/min, n = 11; 18 L/min, n = 11; 28 L/min, n = 14), but the levels of all 3 genes were highest in lambs ventilated at 18 L/min, consistent with the physiological data. CTGF and EGR1 mRNA levels were ~200% greater than those seen in lambs ventilated at 8 L/min (p<0.05; Fig. 2), whereas mRNA levels of CYR61 were only 35% higher and this was not statistically significant.
mRNA levels, expressed as fold change relative to mRNA levels in age-matched un-ventilated control tissue (n = 8), of CTGF (A), EGR1 (B) and CYR61 (C) in lung tissue of lambs ventilated with a ventilator bias gas flow of 8 (n = 11), 18 (n = 11) or 28 L/min (n = 14). * p<0.05 vs. flow 18 L/min, § p<0.05 vs. control tissue and §§ p<0.01 vs. control tissue.
Quantitative histological analysis of the lung (control, n = 8; 8 L/min, n = 5; 18 L/min, n = 5; 28 L/min, n = 5) revealed that the percentage of space occupied by tissue was significantly reduced in lambs ventilated at 8 and 28 L/min compared with unventilated control lambs. On the other hand, collagen density, measured as a percentage of tissue area, was increased in lambs ventilated at 8 and 28 L/min compared to control lambs (table 3), which may be related to the reduction in tissue area observed in these lambs.
Ventilation at 18 L/min had the most marked effects on the quantitative histological analysis of lung tissue, with decreased elastin density compared with control lambs and lambs ventilated at 28 L/min, decreased αSMA staining density compared with all other groups and increased proportion of total lung space occupied by tissue compared with lambs ventilated at 8 and 28 L/min (table 3). The proportion of Ki67 positive cells (labelling index) was significantly lower in lambs ventilated at 8 vs 28 L/min (table 3). There was no significant difference amongst groups in number of secondary septal crests (data not shown).
Qualitative analysis demonstrated that, in control tissue, elastin and αSMA staining was predominantly located at the tips of the secondary septal crests. Following only 2 h of ventilation, septal crests appeared thicker and shorter and increased elastin and αSMA staining was seen in the alveolar wall with differentiated myofibroblasts more randomly distributed throughout the interstitium (Fig. 3). These changes were most apparent in lambs ventilated at flows of 18 and 28 L/min (Fig. 3). Collagen fibres, seen as thick parallel fibres in the airway wall in control tissue, appeared more tortuous following ventilation with a fine meshwork of fibres in the thickened interstitium. Again, these changes were most apparent in lambs ventilated at 18 and 28 L/min (Fig. 3).
Photomicrographs of controls (n = 8) are shown in the first row (A, E, I), of lambs ventilated at 8 L/min (n = 5) in the second row (B, F, J), 18 L/min (n = 5) in the third row (C, G, K) and 28 L/min (n = 5) in the fourth row (D, H, L). Columns demonstrate elastin (stained black with Hart’s resorcin stain; A–D), differentiated myofibroblasts (stained brown using immunohistochemistry; E–H), and collagen type I and III fibres (stained black with Gordon-Sweet’s stain; I–L). Bar 10 µm. Arrows (↑) demonstrate secondary septal crests with elastin (A) or myofibroblasts (E) visible at the tip. Solid arrowheads (▴) demonstrate abnormal deposition of elastin and open arrowheads (Δ) demonstrate thickened interstitium (C, D, H) containing a finer meshwork of collagen fibres in tissue ventilated at 18 and 28 L/min (K, L).
This study has investigated the effects of ventilator bias gas flows on gas flow into the lung, ventilator efficiency, cardiopulmonary responses and lung injury in a preterm lamb paradigm. Although bias gas flow influences the rate of gas flow into the lung, it is only one of a number of determinants of inspiratory flow, with both respiratory system compliance and resistance also having highly significant effects. We report that high bias gas flows, resulting in higher inspiratory gas flows, have adverse effects on ventilatory parameters, increase mRNA levels of early response genes and result in histological evidence of lung injury. These findings occurred after only 2 hours of ventilation and it remains to be determined whether adverse effects would increase with increasing duration of ventilation.
Many factors are known to contribute to the development of VILI in the preterm lung, including barotrauma , , volutrauma , , atelectotrauma  and oxytrauma , . As volutrauma is considered to be more injurious than barotrauma , , we used volume guarantee ventilation to ensure delivery of a consistent tidal volume independent of changes in compliance or resistance. Compliance was not different amongst groups, although resistance was lowest in lambs ventilated at 8 L/min and highest in lambs ventilated at 28 L/min, most likely secondary to less turbulence with lower flows. Lambs ventilated at 18 and 28 L/min had lower tidal volumes (by 0.3 mL/Kg; less than 5% of the set tidal volume) but required higher PIP values than lambs ventilated at 8 L/min. Apart from flow-related parameters (including the inflation time, which is closely correlated to flow) , ventilator rate was the only ventilatory parameter different between lambs ventilated at 18 and 28 L/min; lambs ventilated at 18 L/min were ventilated at higher rates in response to higher PaCO2 levels. As there is a pattern of greater lung injury in animals ventilated at 18 L/min compared with lambs ventilated with 28 L/min, despite similar tidal volumes and pressures, it appears that the differences in degree of injury are unrelated to either barotrauma or volutrauma.
A change in the shape of the pressure wave may contribute to lung injury by altering the rate of lung inflation . A slower pressure rise can be achieved by altering various ventilatory parameters , , , including reducing the bias gas flow, which reduces gas flow into the lung by reducing the pressure gradient between the ventilator circuit and the lung throughout most of inflation (Fig. 1A). However, it should also be noted that ET tube resistance and even respiratory system compliance influences the rate of pressure increase, presumably due to their independent effects on gas flow into the lungs (Fig. 1B & 1C). A decreased bias gas flow, with its attendant altered pressure wave pattern, has been shown to protect against lung injury in adult sheep ventilated at high pressures (PIP = 50 cmH2) . However, there are no studies in neonates reporting advantages of one pressure wave pattern over another.
Lambs ventilated at 18 L/min had the most marked injury, despite having lower inspiratory flows than lambs ventilated at 28 L/min. Possible reasons for this include a contribution from hyperoxia-induced trauma , , as the FiO2 and PaO2 were highest at 18 L/min, and higher ventilatory rates secondary to higher PaCO2 values, a finding which may also have contributed to the trend in a higher rate of pneumothoraces in these lambs.
However, the rates of increase in both inspiratory and expiratory flows were highest in lambs ventilated with 18 L/min bias gas flow, rather than in those ventilated with 28 L/min. The finding of most marked lung injury and highest mRNA levels of the acute response genes in lambs ventilated with 18 L/min, therefore, supports the hypothesis that lung injury may be related to rate of rise of flow and/or expiratory flow.
The lower rate of rise of inspiratory flow in lambs ventilated at 28 L/min compared with those ventilated at 18 L/min is most likely due to turbulent gas flow in the endotracheal tube at a flow of 28 L/min. Flow of gas at 37°C and at a barometric pressure of 760 mmHg becomes turbulent through a size 4.0 endotracheal tubes once the flow exceeds 18 L/min . Thus, at the highest flow of 28 L/min the flow pattern will have been turbulent, paradoxically leading to a lower rate of increase in flow into the airways. Furthermore, a turbulent flow pattern leads to a larger pressure drop from the wye piece to the endotracheal tube . Thus, even though PIP values measured by the sensor at the wye piece were similar in lambs ventilated with flows of 18 and 28 L/min, the pressure in the endotracheal tube is likely to have been higher at 18 L/min when flow is more laminar than at 28 L/min. Lambs ventilated at 18 L/min also had the greatest expiratory flow. The reason for this is not clear, but may result from higher expiratory flow in the endotracheal tube creating a Venturi-like effect at the intersection between the comparatively narrow tube and wider ventilator circuit. It could also be due to reduced pulmonary compliance as a result of greater lung injury. We did not measure dynamic pulmonary compliance, but there was no difference in effective pulmonary compliance between lambs ventilated at 18 and 28 L/min. However, it is unlikely that the system is sensitive enough to detect the changes in compliance that could lead to these changes in flow.
Heart rate was significantly higher in animals ventilated at 18 L/min. As tidal volume in animals ventilated at 18 L/min was not different from animals ventilated at 28 L/min, this is unlikely to be hyperexpansion leading to altered diastolic filling of the heart. It may reflect higher PaCO2 values in this group which, despite increasing the ventilator rate, remained above 60 mmHg. Since the animals were muscle relaxed, they were not able to increase their breathing rate spontaneously to compensate for higher PaCO2, but the heart rate may have increased via the chemoreceptor reflex in an attempt to eliminate surplus arterial CO2 .
The early response genes CTGF, CYR61 and EGR1 have recently been identified as potentially useful biomarkers of VILI in preterm newborn lambs . They are also up-regulated in lungs of adults with chronic obstructive pulmonary disease and pulmonary fibrosis , , , in response to stimuli promoting fetal lung growth in sheep  and in response to hyperoxia in mice . CTGF and CYR61, members of the CCN family of cysteine-rich proteins, promote cell proliferation and differentiation, tissue regeneration and synthesis of extra-cellular components, including collagen, and are involved in embryonic angiogenesis , . CTGF is immunolocalised to activated fibroblasts and type II alveolar epithelial cells  whereas CYR61 is an extracellular matrix associated protein ,  and is up-regulated when subjected to cyclic mechanical stretch . CYR61 mediates expression of αSMA in smooth muscle cells through cytoskeletally based mechano-transduction .
Expression of the transcription factor EGR1 is rapid and of short duration after acute tissue injury , hypoxia  and pneumonectomy , although a prolonged expression has been found in lung tissue from patients suffering from advanced stage emphysema . Thus, EGR1 is involved in both immediate early responses and more chronic lung destruction.
We report an up-regulation of mRNA levels of these three genes in the lung after 2 hours of ventilation, which is consistent with our previous study demonstrating that levels are elevated within 15 minutes and remain elevated after 2 hours of injurious ventilation in preterm lambs . In our study, highest levels (~10 fold of control values) were measured in animals ventilated at 18 L/min. In both studies, up-regulation of these genes occurred shortly after the onset of ventilation and higher expression levels were associated with ventilation using larger volumes or higher bias gas flows. Therefore, altered levels of these early response genes likely reflect the level of injury in the preterm lung and might potentially predict the level of injury and risk of BPD development in preterm infants.
The histological evidence of lung injury that we report is consistent with the gene expression data. Secondary septal crest formation is an important part of normal lung development, during which the distal airsacs subdivide to form alveoli. An arrest of alveolarisation is a primary characteristic of BPD ,  and is associated with abnormal deposition of elastin , ,  and collagen fibres , , decreased secondary septal crest densities  and increased mesenchymal cell proliferation , .
Collagen and elastin fibres provide the stabilising framework for the lung  and both are important for secondary septal crest formation and airway branching , . Elastin is essential for lung recoil  and is deposited primarily at the tip of secondary septal crests forming an elastin ring around the alveolar opening. Abnormal elastin deposition (increased deposition and altered spatial pattern) is a well described feature of VILI leading to BPD-like changes in lung tissue architecture , , . Allison et al suggested that VILI mechanically disrupts immature secondary septal crests, causing them to flatten, with the consequence that pre-existing elastin fibres located at the tip of secondary crests are re-absorbed into the primary alveolar wall giving the appearance of abnormal deposition .
We describe altered deposition of both elastin and collagen following ventilation, which was most marked at higher flows. Quantitative analysis demonstrated that elastin and collagen densities were lowest after ventilation at 18 L/min, which may reflect disruption of fibres and re-absorption into the primary alveolar wall. Alternatively, it is possible that the altered deposition, with less discrete areas of staining, resulted in an underestimation of the degree of staining by the image analysis software.
The pattern of differentiated myofibroblasts, detected by αSMA staining, followed that of elastin. That is, most myofibroblasts were located at the tip of secondary septal crests in control animals but were more randomly distributed throughout the interstitium following ventilation, consistent with the hypothesis of re-absorption of septal crests into the primary wall . However, we did not find decreased secondary septal crest density in this study, possibly because of the short duration of the study and the possibility that a greater proportion of the septal crests are more mature and robust compared with the very immature lungs described previously . The appearance of shorter and thicker septa after ventilation, especially at the higher flows, might reflect early lung injury and inflammation, as described in baboons with BPD . An actual decrease in density in our study might have been found after a longer duration of ventilation.
Increased cellular proliferation is another characteristic feature of BPD, and has been described in baboons with BPD  and following in-utero ventilation of 110 day lamb fetuses . Recruitment of inflammatory cells to the lungs, with a consequent fall in circulating neutrophil count , has also been reported with ventilation resulting in an increased DNA concentration in the lung . Our findings of increased numbers of proliferating cells, increased space occupied by tissue in the lung and reduced circulating white blood cells are consistent with these reports.
The histological changes that we describe here after only two hours of ventilation are consistent with changes found in babies with BPD , ,  and animals with VILI , , . Our results indicate that lower bias gas flows, even for a short duration of ventilation, may potentially reduce lung injury in preterm infants. We previously demonstrated flows as low as 5 L/min to be safe and efficient in healthy term lambs. Furthermore, optimal ventilator efficiency occurred at flows between 1.5–3 L/Kg.min (bodyweight ~ 6 Kg) . Lambs of our breed are ~ 4.5–5 Kg at 131 days’ gestation; therefore, we chose 8 L/min to be the lowest flow in this study. The Dräger Babylog 8000plus ventilator can provide continuous bias gas flows from 2–30 L/min. However, at flows of 30 L/min, the flow sensor measurements are inconsistent and we therefore used a maximum flow of 28 L/min. It remains to be determined if further reducing the bias gas flow will result in a further beneficial effect on lung injury whilst still maintaining effective ventilation. Randomised studies in preterm babies will be necessary to determine whether bias gas flows contribute to lung injury in humans.
High ventilator bias gas flows contribute to more severe lung injury, alter ventilator parameters and up-regulate biomarkers of VILI in the preterm lamb compared to lower gas flows. Paradoxically, lung injury at the highest bias gas flow (28 L/min) was less than at 18 L/min, probably reflecting turbulent gas flow at the highest flows which may mitigate some of the adverse effects of fast laminar flow. These data demonstrate that ventilator bias gas flow is important in the development of VILI. More research needs to be done to examine the effect that rheotrauma has on the more immature lung and to determine whether bias gas flows may contribute to long-term lung injury, such as BPD.
The authors thank technicians at both Ngapouri Sheep Research Station, Reporoa, New Zealand and The Ritchie Centre, Monash Institute of Medical Research, Melbourne, Australia for their help during the experiments and analysis of the study and Dräger Medical Australia for the loan of the ventilator.
Conceived and designed the experiments: FHB KPB CAK SBH. Performed the experiments: FHB KPB CAH JB SM SEP VAZ SJF MHO MJW. Analyzed the data: KPB FHB SBH MJW. Contributed reagents/materials/analysis tools: FHB SHB SJF MJW. Wrote the paper: KPB SBH FHB.
- 1. Jobe AH, Bancalari E (2001) Bronchopulmonary dysplasia. Am J Respir Crit Care Med 163: 1723–1729. doi: 10.1164/ajrccm.163.7.2011060
- 2. Bancalari E, Claure N, Sosenko IR (2003) Bronchopulmonary dysplasia: changes in pathogenesis, epidemiology and definition. Semin Neonatol 8: 63–71. doi: 10.1016/s1084-2756(02)00192-6
- 3. Albertine KH, Jones GP, Starcher BC, Bohnsack JF, Davis PL, et al. (1999) Chronic lung injury in preterm lambs. Disordered respiratory tract development. Am J Respir Crit Care Med 159: 945–958. doi: 10.1164/ajrccm.159.3.9804027
- 4. Coalson JJ, Winter VT, Siler-Khodr T, Yoder BA (1999) Neonatal chronic lung disease in extremely immature baboons. Am J Respir Crit Care Med 160: 1333–1346. doi: 10.1164/ajrccm.160.4.9810071
- 5. Pierce RA, Albertine KH, Starcher BC, Bohnsack JF, Carlton DP, et al. (1997) Chronic lung injury in preterm lambs: disordered pulmonary elastin deposition. Am J Physiol 272: L452–460.
- 6. Thibeault DW, Mabry SM, Ekekezie II, Zhang X, Truog WE (2003) Collagen scaffolding during development and its deformation with chronic lung disease. Pediatrics 111: 766–776. doi: 10.1542/peds.111.4.766
- 7. Allison BJ, Crossley KJ, Flecknoe SJ, Davis PG, Morley CJ, et al. (2008) Ventilation of the very immature lung in utero induces injury and BPD-like changes in lung structure in fetal sheep. Pediatr Res 64: 387–392. doi: 10.1203/pdr.0b013e318181e05e
- 8. Webb HH, Tierney DF (1974) Experimental pulmonary edema due to intermittent positive pressure ventilation with high inflation pressures. Protection by positive end-expiratory pressure. Am Rev Respir Dis 110: 556–565.
- 9. Slutsky AS (1999) Lung injury caused by mechanical ventilation. Chest 116: 9S–15S. doi: 10.1378/chest.116.suppl_1.9s-a
- 10. Dreyfuss D, Saumon G (1992) Barotrauma is volutrauma, but which volume is the one responsible? Intensive Care Med 18: 139–141. doi: 10.1007/bf01709236
- 11. Hernandez LA, Peevy KJ, Moise AA, Parker JC (1989) Chest wall restriction limits high airway pressure-induced lung injury in young rabbits. J Appl Physiol 66: 2364–2368.
- 12. Speer CP (2006) Inflammation and bronchopulmonary dysplasia: A continuing story. Semin Fetal Neonatal Med 11: 354–362. doi: 10.1016/j.siny.2006.03.004
- 13. Roberts D, Dalziel SR (2006) Antenatal corticosteroids for accelerating fetal lung maturation for women at risk of preterm birth. Cochrane Database Syst Rev: CD004454.
- 14. Crowther CA, Haslam RR, Hiller JE, Doyle LW, Robinson JS (2006) Australasian Collaborative Trial of Repeat Doses of Steroids Study Group. Neonatal respiratory distress syndrome after repeat exposure to antenatal corticosteroids: a randomised controlled trial. Lancet 367: 1913–1919. doi: 10.1016/s0140-6736(06)68846-6
- 15. Sinn JKH, Ward MC, Henderson-Smart DJ (2002) Developmental outcome of preterm infants after surfactant therapy: systematic review of randomized controlled trials. J Paediatr Child Health 38: 597–600. doi: 10.1046/j.1440-1754.2002.00061.x
- 16. Gerhardt T, Claure N, Bancalari E (2008) Role of pulmonary function testing in the management of neonates on mechanical ventilation. In: Bancalari E, Polin RA, editors. The newborn lung: neonatology questions and controversies. 1st ed. Philadelphia: Saunders Elsevier. 419–448.
- 17. Bach KP, Kuschel CA, Oliver MH, Bloomfield FH (2009) Ventilator gas flow rates affect inspiratory time and ventilator efficiency index in term lambs. Neonatology 96: 259–264. doi: 10.1159/000220765
- 18. Zhang W, Yan SD, Zhu A, Zou YS, Williams M, et al. (2000) Expression of Egr-1 in late stage emphysema. Am J Pathol 157: 1311–1320. doi: 10.1016/s0002-9440(10)64646-9
- 19. Ning W, Li CJ, Kaminski N, Feghali-Bostwick CA, Alber SM, et al. (2004) Comprehensive gene expression profiles reveal pathways related to the pathogenesis of chronic obstructive pulmonary disease. Proc Natl Acad Sci U S A 101: 14895–14900. doi: 10.1073/pnas.0401168101
- 20. Pan LH, Yamauchi K, Uzuki M, Nakanishi T, Takigawa M, et al. (2001) Type II alveolar epithelial cells and interstitial fibroblasts express connective tissue growth factor in IPF. Eur Respir J 17: 1220–1227. doi: 10.1183/09031936.01.00074101
- 21. Wallace MJ, Probyn M, Zahra VA, Crossley KJ, Cole TJ, et al. (2009) Early biomarkers and potential mediators of ventilation-induced lung injury in very preterm lambs. Respir Res 10: 19. doi: 10.1186/1465-9921-10-19
- 22. Leslie KO, Mitchell JJ, Low RB (1992) Lung myofibroblasts. Cell Motil Cytoskeleton 22: 92–98. doi: 10.1002/cm.970220203
- 23. Leslie KO, Mitchell JJ, Woodcock-Mitchell JL, Low RB (1990) Alpha smooth muscle actin expression in developing and adult human lung. Differentiation 44: 143–149. doi: 10.1111/j.1432-0436.1990.tb00547.x
- 24. Spitzer AR, Greenspan JS, Fox WW (2003) Positive-pressure ventilation; pressure-limited and time-cycled ventilation. In: Goldsmith JP, Karotkin EH, editors. Assisted ventilation of the neonate. Philadelphia: Saunders. 149–169.
- 25. Greenough A (1988) The premature infant's respiratory response to mechanical ventilation. Early Hum Dev 17: 1–5. doi: 10.1016/s0378-3782(88)80051-3
- 26. Simbruner G, Gregory GA (1981) Performance of neonatal ventilators: the effects of changes in resistance and compliance. Crit Care Med 9: 509–514. doi: 10.1097/00003246-198107000-00002
- 27. Boros SJ, Bing DR, Mammel MC, Hagen E, Gordon MJ (1984) Using conventional infant ventilators at unconventional rates. Pediatrics 74: 487–492. doi: 10.1203/00006450-198404001-01764
- 28. Rich PB, Reickert CA, Sawada S, Awad SS, Lynch WR, et al. (2000) Effect of rate and inspiratory flow on ventilator-induced lung injury. J Trauma 49: 903–911. doi: 10.1097/00005373-200011000-00019
- 29. Kamlin CO, Davis PG (2003) Long versus short inspiratory times in neonates receiving mechanical ventilation. Cochrane Database Syst Rev: CD004503.pub004502.
- 30. Jarreau PH, Louis B, Dassieu G, Desfrere L, Blanchard PW, et al. (1999) Estimation of inspiratory pressure drop in neonatal and pediatric endotracheal tubes. J Appl Physiol 87: 36–46.
- 31. Seeley RR, Stephens TD, Tate P (1992) Respiratory system. In: Allen D, editor. Anatomy and Physiology. St. Louis: Mosby-Year Book. 735–770.
- 32. Sozo F, Wallace MJ, Zahra VA, Filby CE, Hooper SB (2006) Gene expression profiling during increased fetal lung expansion identifies genes likely to regulate development of the distal airways. Physiol Genomics 24: 105–113. doi: 10.1152/physiolgenomics.00148.2005
- 33. Perkowski S, Sun J, Singhal S, Santiago J, Leikauf GD, et al. (2003) Gene expression profiling of the early pulmonary response to hyperoxia in mice. Am J Respir Cell Mol Biol 28: 682–696. doi: 10.1165/rcmb.4692
- 34. Kubota S, Takigawa M (2007) CCN family proteins and angiogenesis: from embryo to adulthood. Angiogenesis 10: 1–11. doi: 10.1007/s10456-006-9058-5
- 35. Chen CC, Mo FE, Lau LF (2001) The angiogenic factor Cyr61 activates a genetic program for wound healing in human skin fibroblasts. J Biol Chem 276: 47329–47337. doi: 10.1074/jbc.m107666200
- 36. Zhou D, Herrick DJ, Rosenbloom J, Chaqour B (2005) Cyr61 mediates the expression of VEGF, alpha-v Integrin and alpha-actin genes through cytoskeletally-based mechanotransduction mechanisms in bladder smooth muscle cells. J Appl Physiol 98: 2344–2354. doi: 10.1152/japplphysiol.01093.2004
- 37. Sakamoto S, Yokoyama M, Aoki M, Suzuki K, Kakehi Y, et al. (2004) Induction and function of CYR61 (CCN1) in prostatic stromal and epithelial cells: CYR61 is required for prostatic cell proliferation. Prostate 61: 305–317. doi: 10.1002/pros.20098
- 38. Yan SF, Fujita T, Lu J, Okada K, Shan Zou Y, et al. (2000) Egr-1, a master switch coordinating upregulation of divergent gene families underlying ischemic stress. Nat Med 6: 1355–1361.
- 39. Landesberg LJ, Ramalingam R, Lee K, Rosengart TK, Crystal RG (2001) Upregulation of transcription factors in lung in the early phase of postpneumonectomy lung growth. Am J Physiol Lung Cell Mol Physiol 281: L1138–L1149.
- 40. Young CD, Moore GW, Hutchins GM (1980) Connective tissue arrangement in respiratory airways. Anat Rec 198: 245–254. doi: 10.1002/ar.1091980211
- 41. Wright C, Strauss S, Toole K, Burt AD, Robson SC (1999) Composition of the pulmonary interstitium during normal development of the human fetus. Pediatr Dev Pathol 2: 424–431. doi: 10.1007/s100249900145
- 42. Heine UI, Munoz EF, Flanders KC, Roberts AB, Sporn MB (1990) Colocalization of TGF-beta 1 and collagen I and III, fibronectin and glycosaminoglycans during lung branching morphogenesis. Development 109: 29–36.
- 43. McGowan SE (1992) Extracellular matrix and the regulation of lung development and repair. FASEB J 6: 2895–2904.
- 44. Coalson JJ, Winter VT, Gerstmann DR, Idell S, King RJ, et al. (1992) Pathophysiologic, morphometric, and biochemical studies of the premature baboon with bronchopulmonary dysplasia. Am Rev Respir Dis 145: 872–881. doi: 10.1164/ajrccm/145.4_pt_1.872
- 45. Carlton DP, Albertine KH, Cho SC, Lont M, Bland RD (1997) Role of neutrophils in lung vascular injury and edema after premature birth in lambs. J Appl Physiol 83: 1307–1317.
- 46. Margraf LR, Tomashefski JF Jr, Bruce MC, Dahms BB (1991) Morphometric analysis of the lung in bronchopulmonary dysplasia. Am Rev Respir Dis 143: 391–400. doi: 10.1164/ajrccm/143.2.391
- 47. Thibeault DW, Mabry SM, Ekekezie II, Truog WE (2000) Lung elastic tissue maturation and perturbations during the evolution of chronic lung disease. Pediatrics 106: 1452–1459. doi: 10.1542/peds.106.6.1452