Differential effect of recruitment maneuvres on pulmonary blood flow and oxygenation during HFOV in preterm lambs

Time to Lung Volume Stability After Pressure Change During High-Frequency Oscillatory Ventilation

Supplemental Digital Content is available in the text.

OBJECTIVES:

Clinicians have little guidance on the time needed before assessing the effect of a mean airway pressure change during high-frequency oscillatory ventilation. We aimed to determine: 1) time to stable lung volume after a mean airway pressure change during high-frequency oscillatory ventilation and 2) the relationship between time to volume stability and the volume state of the lung.

DESIGN:

Prospective observational study.

SETTING:

Regional quaternary teaching hospital neonatal ICU.

PATIENTS:

Thirteen term or near-term infants receiving high-frequency oscillatory ventilation and muscle relaxants.

INTERVENTIONS:

One to two cm H₂O mean airway pressure changes every 10 minutes as part of an open lung strategy based on oxygen response.

MEASUREMENTS AND MAIN RESULTS:

Continuous lung volume measurements (respiratory inductive plethysmography) were made during the mean airway pressure changes. Volume signals were analyzed with a biexponential model to calculate the time to stable lung volume if the model R² was greater than 0.6. If volume stability did not occur within 10 minutes, the model was extrapolated to maximum 3,600 s. One-hundred ninety-six mean airway pressure changes were made, with no volume change in 33 occurrences (17%). One-hundred twenty-five volume signals met modeling criteria for inclusion; median (interquartile range) R², 0.96 (0.91–0.98). The time to stable lung volume was 1,131 seconds (718–1,959 s) (mean airway pressure increases) and 647 seconds (439–1,309 s) (mean airway pressure decreases), with only 17 (14%) occurring within 10 minutes and time to stability being longer when the lung was atelectatic.

CONCLUSIONS:

During high-frequency oscillatory ventilation, the time to stable lung volume after a mean airway pressure change is variable, often requires more than 10 minutes, and is dependent on the preceding volume state.

Impact of delivered tidal volume on the occurrence of intraventricular haemorrhage in preterm infants during positive pressure ventilation in the delivery room

BACKGROUND AND OBJECTIVES

Delivery of inadvertent high tidal volume (V_T) during positive pressure ventilation (PPV) in the delivery room is common. High V_T delivery during PPV has been associated with haemodynamic brain injury in animal models. We examined if V_T delivery during PPV at birth is associated with brain injury in preterm infants <29 weeks' gestation.

METHODS

A flow-sensor was placed between the mask and the ventilation device. V_T values were compared with recently described reference ranges for V_T in spontaneously breathing preterm infants at birth. Infants were divided into two groups: V_T<6  mL/kg or V_T>6 mL/kg (normal and high V_T, respectively). Brain injury (eg, intraventricular haemorrhage (IVH)) was assessed using routine ultrasound imaging within the first days after birth.

RESULTS

A total of 165 preterm infants were included, 124 (75%) had high V_T and 41 (25%) normal V_T. The mean (SD) gestational age and birth weight in high and normal V_T group was similar, 26 (2) and 26 (1) weeks, 858 (251) g and 915 (250) g, respectively. IVH in the high V_T group was diagnosed in 63 (51%) infants compared with 5 (13%) infants in the normal V_T group (P=0.008).Severe IVH (grade III or IV) developed in 33/124 (27%) infants in the high V_T group and 2/41 (6%) in the normal V_T group (P=0.01).

CONCLUSIONS

High V_T delivery during mask PPV at birth was associated with brain injury. Strategies to limit V_T delivery during mask PPV should be used to prevent high V_T delivery.

Effect of body position and ventilation on umbilical artery and venous blood flows during delayed umbilical cord clamping in preterm lambs

OBJECTIVE

While delayed umbilical cord clamping (UCC) is thought to facilitate placental to infant blood transfusion, the physiological factors regulating flow in the umbilical arteries and veins during delayed UCC is unknown. We investigated the effects of gravity, by changing fetal height relative to the placenta, and ventilation on umbilical blood flows and the cardiovascular transition during delayed UCC at birth.

METHODS

Catheters and flow probes were implanted into preterm lambs (128 days) prior to delivery to measure pulmonary, carotid, umbilical artery (UaBF) and umbilical venous (UvBF) blood flows. Lambs were placed either 10 cm below or 10 cm above the ewe. Ventilation commenced 2-3 min before UCC and continued for 30 min after UCC.

RESULTS

Gravity reduced umbilical and cerebral flows when lambs were placed below the midline, but the reduction in UaBF and UvBF was similar. Ventilation during delayed UCC reduced UvBF and UaBF by similar amounts, irrespective of the lamb's position, such that flows into and out of the placenta remained balanced. The effects of ventilation on umbilical flows were much greater than the effects of gravity, but no net placental to lamb blood transfusion could be detected under any condition. Cardiovascular parameters, cerebral oxygen kinetics and final blood volumes were similar in both groups 5 min after UCC.

CONCLUSIONS

Gravity caused small transient effects on umbilical and cerebral flow, but given changes were similar in umbilical arteries and veins, no net placental transfusion was detected. Ventilation during delayed UCC has a markedly greater influence on cardiovascular function in the newborn.

Relationship between Mean Airways Pressure, Lung Mechanics, and Right Ventricular Output during High-Frequency Oscillatory Ventilation in Infants

OBJECTIVE

To characterize changes in lung mechanics and right ventricular output (RVO) during incremental/decremental continuous distending pressure (CDP) maneuvers in newborn infants receiving high-frequency oscillatory ventilation, with the aim of evaluating when open lung maneuvers are needed and whether they are beneficial.

STUDY DESIGN

Thirteen infants on high-frequency oscillatory ventilation were studied with a median (IQR) gestational age of 26¹ (25³-29¹) weeks and median (IQR) body weight of 810 (600-1020) g. CDP was increased stepwise from 8 cmH₂O to a maximum pressure and subsequently decreased until oxygenation deteriorated or a CDP of 8 cmH₂O was reached. The lowest CDP that maintained good oxygenation was considered the clinically optimal CDP. At each CDP, the following variables were evaluated: oxygenation, respiratory system reactance (Xrs), and RVO by Doppler echocardiography.

RESULTS

At maximal CDP reached during the trial, 19 [1] cmH₂O (mean [SEM]), oxygenation markedly improved, and Xrs and RVO decreased. During deflation, oxygenation remained stable over a wide range of CDP settings, Xrs returned to the baseline values, and RVO increased but the baseline values were not readily restored in all patients.

CONCLUSION

These results suggest that Xrs and RVO are more sensitive than oxygenation to overdistension and they may be useful in clinical practice to guide open lung maneuvers.

A comparison of high-frequency jet ventilation and synchronised intermittent mandatory ventilation in preterm lambs

PURPOSE

Synchronised intermittent mandatory ventilation (SIMV) and high-frequency jet ventilation (HFJV) are accepted ventilatory strategies for treatment of respiratory distress syndrome (RDS) in preterm babies. We hypothesised that SIMV and HFJV both facilitate adequate oxygenation and ventilation but that HFJV is associated with less lung injury.

RESULTS

There were no differences in arterial oxygenation or partial pressure of carbon dioxide despite lower mean airway pressure during SIMV for most of the study. There were no consistent significant differences in end systolic and end diastolic PBF, lung injury data and static lung compliance.

METHODS

Preterm lambs of anaesthetised ewes were instrumented, intubated and delivered by caesarean section after intratracheal suction and instillation of surfactant. Each lamb was managed for 3 hr according to a predetermined algorithm for ventilatory support consistent with open lung ventilation. Pulmonary blood flow (PBF) was measured continuously and pulsatility index was calculated. Ventilatory parameters were recorded and arterial blood gases were measured at intervals. At postmortem, in situ pressure-volume deflation curves were recorded, and bronchoalveolar lavage fluid and lung tissue were obtained to assess inflammation.

CONCLUSIONS

SIMV and HFJV have comparable clinical efficacy and ventilator pressure requirements when applied with a targeted lung volume recruitment strategy.

Unraveling the Links Between the Initiation of Ventilation and Brain Injury in Preterm Infants

The initiation of ventilation in the delivery room is one of the most important but least controlled interventions a preterm infant will face. Tidal volumes (V T) used in the neonatal intensive care unit are carefully measured and adjusted. However, the V Ts that an infant receives during resuscitation are usually unmonitored and highly variable. Inappropriate V Ts delivered to preterm infants during respiratory support substantially increase the risk of injury and inflammation to the lungs and brain. These may cause cerebral blood flow instability and initiate a cerebral inflammatory cascade. The two pathways increase the risk of brain injury and potential life-long adverse neurodevelopmental outcomes. The employment of new technologies, including respiratory function monitors, can improve and guide the optimal delivery of V Ts and reduce confounders, such as leak. Better respiratory support in the delivery room has the potential to improve both respiratory and neurological outcomes in this vulnerable population.

Protective ventilation of preterm lambs exposed to acute chorioamnionitis does not reduce ventilation-induced lung or brain injury

Background

The onset of mechanical ventilation is a critical time for the initiation of cerebral white matter (WM) injury in preterm neonates, particularly if they are inadvertently exposed to high tidal volumes (V_T) in the delivery room. Protective ventilation strategies at birth reduce ventilation-induced lung and brain inflammation and injury, however its efficacy in a compromised newborn is not known. Chorioamnionitis is a common antecedent of preterm birth, and increases the risk and severity of WM injury. We investigated the effects of high V_T ventilation, after chorioamnionitis, on preterm lung and WM inflammation and injury, and whether a protective ventilation strategy could mitigate the response.

Methods

Pregnant ewes (n = 18) received intra-amniotic lipopolysaccharide (LPS) 2 days before delivery, instrumentation and ventilation at 127±1 days gestation. Lambs were either immediately euthanased and used as unventilated controls (LPS_UVC; n = 6), or were ventilated using an injurious high V_T strategy (LPS_INJ; n = 5) or a protective ventilation strategy (LPS_PROT; n = 7) for a total of 90 min. Mean arterial pressure, heart rate and cerebral haemodynamics and oxygenation were measured continuously. Lungs and brains underwent molecular and histological assessment of inflammation and injury.

Results

LPS_INJ lambs had poorer oxygenation than LPS_PROT lambs. Ventilation requirements and cardiopulmonary and systemic haemodynamics were not different between ventilation strategies. Compared to unventilated lambs, LPS_INJ and LPS_PROT lambs had increases in pro-inflammatory cytokine expression within the lungs and brain, and increased astrogliosis (p<0.02) and cell death (p<0.05) in the WM, which were equivalent in magnitude between groups.

Conclusions

Ventilation after acute chorioamnionitis, irrespective of strategy used, increases haemodynamic instability and lung and cerebral inflammation and injury. Mechanical ventilation is a potential contributor to WM injury in infants exposed to chorioamnionitis.

Respiratory support for premature neonates in the delivery room: effects on cardiovascular function and the development of brain injury

The transition to newborn life in preterm infants is complicated by immature cardiovascular and respiratory systems. Consequently, preterm infants often require respiratory support immediately after birth. Although aeration of the lung underpins the circulatory transition at birth, positive pressure ventilation can adversely affect cardiorespiratory function during this vulnerable period, reducing pulmonary blood flow and left ventricular output. Furthermore, pulmonary volutrauma is known to initiate pulmonary inflammatory responses, resulting in remote systemic involvement. This review focuses on the downstream consequences of positive pressure ventilation, in particular, interactions between cardiovascular output and the initiation of a systemic inflammatory cascade, on the immature brain. Recent studies have highlighted that positive pressure ventilation strategies are precursors of cerebral injury, probably mediated through cerebral blood flow instability. The presence of, or initiation of, an inflammatory cascade accentuates adverse cerebral blood flow, in addition to being a direct source of brain injury. Importantly, the degree of brain injury is dependent on the nature of the initial ventilation strategy used.

The cardiopulmonary haemodynamic transition at birth is not different between male and female preterm lambs

Males born preterm are at greater risk of illness and death than females, principally due to respiratory disease. Much of the excess morbidity occurs within the first few hours of life. Therefore, the aim of the present study was to investigate whether or not differences in the cardiopulmonary transition soon after birth underlie the increased morbidity in males after preterm birth. Nine female and thirteen male lambs (128±2 days gestation) underwent surgery immediately before delivery for implantation of a pulmonary arterial flow-probe and catheters into the main pulmonary artery and a carotid artery. After birth lambs were ventilated for 30 min (tidal volume 7 mL kg(-1)) while anaesthetised. Arterial pressures and flows were recorded in real time and left-ventricular output measured using Doppler echocardiography. Before birth, fetal cardiopulmonary haemodynamics, arterial blood gases, pH, glucose and lactate did not differ between sexes. Similarly, in the neonatal period there were no significant differences in arterial blood gas status, ventilation parameters, respiratory indices or cardiopulmonary haemodynamics between the sexes. Our data show that the cardiopulmonary transition at birth in ventilated, anaesthetised preterm lambs is not influenced by sex. Thus, the neonatal 'male disadvantage' is not explained by an impaired cardiovascular transition at birth.

Initiation of resuscitation with high tidal volumes causes cerebral hemodynamic disturbance, brain inflammation and injury in preterm lambs

AIMS

Preterm infants can be inadvertently exposed to high tidal volumes (V(T)) in the delivery room, causing lung inflammation and injury, but little is known about their effects on the brain. The aim of this study was to compare an initial 15 min of high V(T) resuscitation strategy to a less injurious resuscitation strategy on cerebral haemodynamics, inflammation and injury.

METHODS

Preterm lambs at 126 d gestation were surgically instrumented prior to receiving resuscitation with either: 1) High V(T) targeting 10-12 mL/kg for the first 15 min (n = 6) or 2) a protective resuscitation strategy (Prot V(T)), consisting of prophylactic surfactant, a 20 s sustained inflation and a lower initial V(T) (7 mL/kg; n = 6). Both groups were subsequently ventilated with a V(T) 7 mL/kg. Blood gases, arterial pressures and carotid blood flows were recorded. Cerebral blood volume and oxygenation were assessed using near infrared spectroscopy. The brain was collected for biochemical and histologic assessment of inflammation, injury, vascular extravasation, hemorrhage and oxidative injury. Unventilated controls (UVC; n = 6) were used for comparison.

RESULTS

High V(T) lambs had worse oxygenation and required greater ventilatory support than Prot V(T) lambs. High V(T) resulted in cerebral haemodynamic instability during the initial 15 min, adverse cerebral tissue oxygenation index and cerebral vasoparalysis. While both resuscitation strategies increased lung and brain inflammation and oxidative stress, High V(T) resuscitation significantly amplified the effect (p = 0.014 and p<0.001). Vascular extravasation was evident in the brains of 60% of High V(T) lambs, but not in UVC or Prot V(T) lambs.

CONCLUSION

High V(T) resulted in greater cerebral haemodynamic instability, increased brain inflammation, oxidative stress and vascular extravasation than a Prot V(T) strategy. The initiation of resuscitation targeting Prot V(T) may reduce the severity of brain injury in preterm neonates.

Inflammation in utero exacerbates ventilation-induced brain injury in preterm lambs

Cerebral blood flow disturbance is a major contributor to brain injury in the preterm infant. The initiation of ventilation may be a critical time for cerebral hemodynamic disturbance leading to brain injury in preterm infants, particularly if they are exposed to inflammation in utero. We aimed to determine whether exposure to inflammation in utero alters cardiopulmonary hemodynamics, resulting in cerebral hemodynamic disturbance and related brain injury during the initiation of ventilation. Furthermore, we aimed to determine whether inflammation in utero alters the cerebral hemodynamic response to challenge induced by high mean airway pressures. Pregnant ewes received intra-amniotic lipopolysaccharide (LPS) or saline either 2 or 4-days before preterm delivery (at 128 ± 1 days of gestation). Lambs were surgically instrumented for assessment of pulmonary and cerebral hemodynamics before delivery and positive pressure ventilation. After 30 min, lambs were challenged hemodynamically by incrementing and decrementing positive end-expiratory pressure. Blood gases, arterial pressures, and blood flows were recorded. The brain was collected for biochemical and histological assessment of inflammation, brain damage, vascular extravasation, hemorrhage, and oxidative injury. Carotid arterial pressure was higher and carotid blood flow was more variable in 2-day LPS lambs than in controls during the initial 15 min of ventilation. All lambs responded similarly to the hemodynamic challenge. Both 2- and 4-day LPS lambs had increased brain interleukin (IL)-1β, IL-6, and IL-8 mRNA expression; increased number of inflammatory cells in the white matter; increased incidence and severity of brain damage; and vascular extravasation relative to controls. Microvascular hemorrhage was increased in 2-day LPS lambs compared with controls. Cerebral oxidative injury was not different between groups. Antenatal inflammation causes adverse cerebral hemodynamics and increases the incidence and severity of brain injury in ventilated preterm lambs.

High positive end-expiratory pressure during high-frequency jet ventilation improves oxygenation and ventilation in preterm lambs

Increasing positive end-expiratory pressure (PEEP) is advocated to recruit alveoli during high-frequency jet ventilation (HFJV), but its effect on cardiopulmonary physiology and lung injury is poorly documented. We hypothesized that high PEEP would recruit alveoli and reduce lung injury but compromise pulmonary blood flow (PBF). Preterm lambs of anesthetized ewes were instrumented, intubated, and delivered by cesarean section after instillation of surfactant. HFJV was commenced with a PEEP of 5 cm H2O. Lambs were allocated randomly at delivery to remain on constant PEEP (PEEPconst, n = 6) or to recruitment via stepwise adjustments in PEEP (PEEPadj, n = 6) to 12 cm H2O then back to 8 cm H2O over the initial 60 min. PBF was measured continuously while ventilatory parameters and arterial blood gases were measured at intervals. At postmortem, in situ pressure-volume deflation curves were recorded, and bronchoalveolar lavage fluid and lung tissue were obtained to assess inflammation. PEEPadj lambs had lower pressure amplitude, fractional inspired oxygen concentration, oxygenation index, and PBF and more compliant lungs. Inflammatory markers were lower in the PEEPadj group. Adjusted PEEP during HFJV improves oxygenation and lung compliance and reduces ventilator requirements despite reducing pulmonary perfusion.

Elective high-frequency oscillatory versus conventional ventilation in preterm infants: a systematic review and meta-analysis of individual patients' data

BACKGROUND

Population and study design heterogeneity has confounded previous meta-analyses, leading to uncertainty about effectiveness and safety of elective high-frequency oscillatory ventilation (HFOV) in preterm infants. We assessed effectiveness of elective HFOV versus conventional ventilation in this group.

METHODS

We did a systematic review and meta-analysis of individual patients' data from 3229 participants in ten randomised controlled trials, with the primary outcomes of death or bronchopulmonary dysplasia at 36 weeks' postmenstrual age, death or severe adverse neurological event, or any of these outcomes.

FINDINGS

For infants ventilated with HFOV, the relative risk of death or bronchopulmonary dysplasia at 36 weeks' postmenstrual age was 0.95 (95% CI 0.88-1.03), of death or severe adverse neurological event 1.00 (0.88-1.13), or any of these outcomes 0.98 (0.91-1.05). No subgroup of infants (eg, gestational age, birthweight for gestation, initial lung disease severity, or exposure to antenatal corticosteroids) benefited more or less from HFOV. Ventilator type or ventilation strategy did not change the overall treatment effect.

INTERPRETATION

HFOV seems equally effective to conventional ventilation in preterm infants. Our results do not support selection of preterm infants for HFOV on the basis of gestational age, birthweight for gestation, initial lung disease severity, or exposure to antenatal corticosteroids.

FUNDING

Nestlé Belgium, Belgian Red Cross, and Dräger International.

Intrauterine inflammation causes pulmonary hypertension and cardiovascular sequelae in preterm lambs

Chorioamnionitis increases the risk and severity of persistent pulmonary hypertension of the newborn in preterm infants. Exposure of preterm fetal lambs to intra-amniotic (IA) lipopolysaccharide (LPS) induces chorioamnionitis, causes hypertrophy of pulmonary resistance arterioles, and alters expression of pulmonary vascular growth proteins. We investigated the cardiopulmonary and systemic hemodynamic consequences of IA LPS in preterm lambs. Pregnant ewes received IA injection of LPS ( n = 6) or saline (controls; n = 8) at 122 days gestation, 7 days before exteriorization, instrumentation, and delivery of the fetus with pulmonary and systemic flow probes and catheters at 129 days gestation. Newborn lambs were ventilated, targeting a tidal volume of 6–7 ml/kg and a positive end-expiratory pressure (PEEP) of 4 cmH2O. At 30 min, all lambs underwent a PEEP challenge: PEEP was increased by 2 cmH2O at 10-min intervals to 10 cmH2O and then decreased similarly to 4 cmH2O. Ventilation parameters, arterial blood flows, and pressures were recorded in real-time for 90 min. LPS lambs had higher total protein in bronchoalveolar lavage fluid ( P < 0.002), increased medial thickness of arteriolar walls ( P = 0.013), and right ventricular hypertrophy ( P = 0.012). Compared with controls, LPS lambs had worse oxygenation ( P < 0.001), decreased pulmonary blood flow ( P = 0.05), and higher pulsatility index ( P < 0.001) and pulmonary ( P < 0.001) and systemic arterial pressures ( P = 0.005) than controls. Intra-amniotic LPS increased right-to-left shunting across the ductus arteriosus ( P = 0.018) and decreased left ventricular output ( P < 0.001). We conclude that inflammation and pulmonary remodeling induced by IA LPS adversely alters pulmonary hemodynamics with subsequent cardiovascular and systemic sequelae, which may predispose the preterm lamb to persistent pulmonary hypertension of the newborn.

Dynamic changes in the direction of blood flow through the ductus arteriosus at birth

Major cardiovascular changes occur at birth, including increased pulmonary blood flow (PBF) and closure of the ductus arteriosus (DA), which acts as a low resistance shunt between the fetal pulmonary and systemic circulations. Although the pressure gradient between these circulations reverses after birth, little is known about DA blood flow changes and whether reverse DA flow contributes to PBF after birth. Our aim was to describe the changes in PBF and DA flow before, during and after the onset of pulmonary ventilation at birth. Flow probes were implanted on the left pulmonary artery (LPA) and DA in preterm fetal sheep (n = 8) approximately 3 days before they were delivered and ventilated. Blood flow was measured in the LPA and DA, before and after umbilical cord occlusion (UCO) and for 2 h after ventilation onset. Following UCO, DA flow decreased from 534 +/- 57 ml min(1) to 237 +/- 29 ml min(1) which reflected a similar reduction in right ventricular output. Within 5 min of ventilation onset, PBF increased from 11 +/- 6 ml min(1) to 230 +/- 13 ml min(1) whereas DA flow decreased to 172 +/- 54 ml min(1); negative values indicate reverse DA flow (left-to-right shunting). Reverse flow through the DA contributed up to 50% of total PBF at 30 min and a decrease in this contribution accounted for 71 +/- 13% of the time-related decrease in PBF after birth. DA blood flow is very dynamic after birth and depends upon the pressure gradient between the pulmonary and systemic circulations. Following ventilation, reverse DA flow provided a significant contribution to total PBF after birth.

Cardiovascular and pulmonary consequences of airway recruitment in preterm lambs

Increases in positive end-expiratory pressure (PEEP) improve arterial oxygenation in preterm infants, but the effects on cardiopulmonary hemodynamics are understood poorly. We aimed to determine the effect of increased PEEP on cardiopulmonary hemodynamics and to compare measurements from indwelling flow probes with Doppler echocardiography. Preterm lambs (129 ± 1 days) were ventilated initially with a tidal volume of 7 ml/kg and 4 cmH2O of PEEP. In ramp lambs ( n = 7), PEEP was increased by 2-cmH2O increments to 10 cmH2O and then in decrements back to 4 cmH2O. PEEP was unchanged in controls ( n = 6). Doppler echocardiographic flow measurements in the left pulmonary artery (LPA) and ductus arteriosus (DA) were correlated with flow probe measurements. Compared with controls, high PEEP reduced LPA flow from baseline (10-cmH2O PEEP: 43 ± 8% vs. control: 83 ± 21%; P = 0.029). High PEEP increased the proportion of right-to-left (R-L) shunting through the DA, with a trend to an increased oxygenation index compared with controls (oxygenation index: 44.5 ± 13.5 at 10-cmH2O PEEP vs. 19.4 ± 4.5 in controls; P = 0.07). Increasing PEEP decreased heart rate (17 beats/min; P = 0.03) and tended to lower systolic arterial pressure (5.0 mmHg; P = 0.052) compared with controls. Doppler echocardiography measurement of LPA flows correlated strongly with indwelling flow probe ( r2 = 0.73, P < 0.001), except during highly turbulent flows. Increases in PEEP have significant cardiopulmonary consequences in preterm lambs, including reduced LPA flow and increased R-L shunt through the DA. These changes are likely due to the concomitant increase in downstream pulmonary vascular resistance and increased cardiovascular constraint induced by PEEP.