Real-time effects of PEEP and tidal volume on regional ventilation and perfusion in experimental lung injury.

Background Real-time bedside information on regional ventilation and perfusion during mechanical ventilation (MV) may help to elucidate the physiological and pathophysiological effects of MV settings in healthy and injured lungs. We aimed to study the effects of positive end-expiratory pressure (PEEP) and tidal volume (VT) on the distributions of regional ventilation and perfusion by electrical impedance tomography (EIT) in healthy and injured lungs. Methods One-hit acute lung injury model was established in 6 piglets by repeated lung lavages (injured group). Four ventilated piglets served as the control group. A randomized sequence of any possible combination of three VT (7, 10, and 15 ml/kg) and four levels of PEEP (5, 8, 10, and 12 cmH2O) was performed in all animals. Ventilation and perfusion distributions were computed by EIT within three regions-of-interest (ROIs): nondependent, middle, dependent. A mixed design with one between-subjects factor (group: intervention or control), and two within-subjects factors (PEEP and VT) was used, with a three-way mixed analysis of variance (ANOVA). Results Two-way interactions between PEEP and group, and VT and group, were observed for the dependent ROI (p = 0.035 and 0.012, respectively), indicating that the increase in the dependent ROI ventilation was greater at higher PEEP and VT in the injured group than in the control group. A two-way interaction between PEEP and VT was observed for perfusion distribution in each ROI: nondependent (p = 0.030), middle (p = 0.006), and dependent (p = 0.001); no interaction was observed between injured and control groups. Conclusions Large PEEP and VT levels were associated with greater pulmonary ventilation of the dependent lung region in experimental lung injury, whereas they affected pulmonary perfusion of all lung regions both in the control and in the experimental lung injury groups.


Introduction
Real-time information on regional ventilation and perfusion, and their changes, during mechanical ventilation (MV) may help elucidate the physiological and pathophysiological effects of MV settings in healthy and injured lungs.
A single-compartment model of the healthy lung can describe aeration and its changes associated with a homogeneous distribution of airway pressures, expansion, and stretching across the lung parenchyma [1]. The positive airway pressure applied during MV may affect pulmonary ventilation and perfusion distributions, and their tidal changes, especially in the presence of acute lung injury.
The effects of positive end-expiratory pressure (PEEP) on the regional distribution of tidal volume (V T ) and recruitment have been investigated by chest computed tomography in sedated-paralyzed patients with the acute respiratory distress syndrome (ARDS) [2]. It was evidenced that PEEP made the gas distribution more homogeneous, stretching the upper levels and recruiting the lower ones [2].
Electrical impedance tomography (EIT) has emerged as a new functional-imaging method potentially meeting many clinical needs. It is a non-invasive, radiation-free tool to monitor, in real-time and at the bedside, the distribution of pulmonary ventilation [3][4][5][6][7][8][9][10][11]. Subjecting the patient's chest to minute electrical currents, EIT measures the electric potentials at the chest wall surface to produce two-dimensional (2D) images that reflect the impedance distribution within the thorax. Cyclic variations in pulmonary air and blood content are the major determinants for the changes in thoracic impedance. Because cyclic changes in local impedance mainly correspond to changes in lung aeration, EIT can reliably assess imbalances in the distribution of regional ventilation [3,4,11,12]. Besides other features like portability and the possibility of aroundthe-clock monitoring, the high temporal resolution (modern EIT devices generate up to 50 images per second) is another important aspect of this imaging method, which also allows the study of rapid physiological phenomena, such as the regional perfusion [13].
We aimed to measure pulmonary regional ventilation and perfusion at 12 combinations of PEEP and V T levels in piglets with saline lavage-induced lung injury and in mechanically ventilated control piglets. We sought to determine (1) if the PEEP and V T effects on the regional distribution of tidal volume in injured lungs are also detectable by EIT; (2) whether there are also effects on perfusion; and (3) does the presence or absence of injury affect PEEP-or V T -related changes in ventilation and perfusion distributions.

Materials and methods
The study was performed at the Hedenstierna Laboratory, Uppsala University. The Regional Animal Ethics Committee approved the study.
Eleven piglets (2-3 months old, weight 30.7 ± 1.5 kg, mean ± SD) of mixed Hampshire, Yorkshire, and Swedish country breeds were included in the study. All animals were studied lying in the supine position under general anesthesia with mandatorymode mechanical ventilation provided via tracheostomy. They were pre-medicated with intramuscular xylazine 2 mg/kg, ketamine 20 mg/kg, and midazolam 0.5 mg/kg. An ear vein was cannulated and intravenous ketamine 32 mg/kg/h, fentanyl 4 mcg/kg/h, and midazolam 0.16 mg/kg/h were administered. Adequacy of anesthesia was confirmed by the absence of reaction to painful stimulation between the front hooves and the absence of any signs of sympathetic stimulation after paralysis. Muscle relaxation was achieved using continuous infusion of rocuronium titrated against the spontaneous respiratory effort. Normovolemia was maintained with intravenous infusion of Ringer's lactate solution at 20 ml/kg/h for the first hour followed by 10 ml/kg/h. MV was performed using a Servo-I ventilator (Maquet, Rastatt, Germany). During the instrumentation phase, ventilation was provided in volume-controlled mode, with V T of 10 ml/kg, respiratory rate (RR) 25 breaths/min, PEEP 5 cm H 2 O, inspiratory-toexpiratory (I:E) ratio 1:2, and fraction of inspired oxygen (F I O 2 ) 0.4. ECG, invasive systemic, central venous, and pulmonary artery blood pressures were transduced using a standard clinical monitor (IntelliVue M8004A, Philips Healthcare, Best, Netherlands). Digital outputs from the ventilator and clinical monitor were continuously recorded using the acqIS software (EPiQ Life Science AB, Kista, Sweden). A femoral artery was cannulated for pulse contour cardiac output monitoring (PiCCO, Pulsion Medical Systems, Munich, Germany). Pulmonary artery flotation catheter thermodilution cardiac output measurements and arterial blood gas analyses were performed at the beginning and end of the scanning series for each animal.

Acute lung injury model
Following baseline measurements, a one-hit acute lung injury model was established in 7 randomly chosen animals (injured group), with repeated lung lavages (30 ml/kg) of isotonic saline applied until an arterial partial pressure of O 2 and fraction of inspired oxygen ratio (PaO 2 /F I O 2 ) of 200 mmHg was reached. During the lavages, mechanical ventilation was set in pressure-controlled mode, with F I O 2 1.0, RR 30, PEEP 5, and a driving pressure resulting in a V T of 6 ml/kg. If required, an infusion of noradrenaline (0.01 to 0.1 mcg/kg/min) was commenced following an injury to maintain adequate mean arterial blood pressure.

Investigational protocol
We studied the effects of combinations of PEEP and V T on regional ventilation and perfusion by EIT in volume-controlled mode. A randomized sequence of any possible combination of three V T (7, 10, and 15 ml/kg) and four levels of PEEP (5, 8, 10, and 12 cmH 2 O) was performed. In total, 12 conditions were studied.

Electrical impedance tomography
Pulmonary EIT data were recorded at 50 Hz with 32 electrodes equidistantly placed around the circumference of the thorax just below the level of the axilla (Enlight, TIM-PEL SA, São Paulo, Brazil) [13,14]. The following functional images were generated by EIT: 1. Ventilation maps derived from relative impedance changes, which reliably track local, pixel-by-pixel changes in the content of air within the lung [12,15]. It is expressed as the percentage of total pulmonary ventilation through each of the three ROIs (total 100%). 2. Perfusion maps obtained by injecting a bolus of 10 ml of a hypertonic solution (NaCl 10%) into a central venous catheter during an expiratory breath hold for 20 s. Due to its high conductivity, NaCl 10% acts as an EIT contrast agent [16], which after injection into the right atrium during apnea passes through the pulmonary circulation, thereby producing a dilution curve that follows typical first-pass kinetics. The resulting regional impedance curves are then analyzed to quantitatively assess regional perfusion [13,17,18], expressed as the percentage of total pulmonary blood flow through each of the three ROIs (total 100%).
For the quantitative analysis of the ventilation and perfusion distributions by EIT, the lungs were sub-segmented into three isogravitational regions-of-interest (ROIs): nondependent, middle, and dependent regions (Fig. 1).

Control group
The same EIT imaging protocol and data analysis were performed in four animals, which did not receive saline lavages.

Statistics
The Shapiro-Wilk test was used to test data for normality. A mixed design with one between-subjects factor (group: intervention or control), and two within-subjects factors (PEEP and V T ) was used, with a three-way mixed analysis of variance (ANOVA). The Bonferroni adjustment for multiple tests was applied for post hoc comparisons. The statistical analyses were conducted with SPSS (version 20; IBM Corp, IBM SPSS Statistics for Windows, Armonk, NY). Individual p values to indicate statistical tests' significance are reported were relevant. Values presented are mean and SEM unless otherwise stated.

Results
The fatality rate was 1/11 piglets, due to a cardiovascular event (this was in the injured group, thus 4 controls and 6 injured completed the study). The 10 piglets that survived the whole experiment were included in the analysis.
Just before starting the application of the twelve combinations of PEEP and V T , cardiopulmonary parameters were measured (Table 1). During the measurements of these cardiopulmonary parameters, mechanical ventilation was set in volume-controlled mode, with F I O 2 0.3-0.4 for the control animals and 0.7-0.8 for those with lung injury, RR 25, PEEP 5, and a V T of 10 ml/kg.  Figure 3 shows representative EIT functional images illustrating the differences in the distribution of pulmonary ventilation and perfusion at two levels of PEEP and V T . Figure 4 shows the distribution of regional pulmonary ventilation at different PEEP and V T levels. Higher PEEP and V T levels were associated with greater percent ventilation in the dependent ROI in the injured group than in the control group.

Regional ventilation
A two-way interaction between PEEP and group (injured/control) was observed for the dependent ROI (p = 0.035), with PEEP-related increase in dependent ROI ventilation being greater in the injured than in the control group.
A two-way interaction between V T and group (injured/control) was also observed for the dependent ROI (p = 0.012), indicating that the increase in the dependent ROI ventilation was greater at higher V T in the injured group than in the control group.
For instance, at PEEP 12 cmH 2 O with V T 15 ml/kg the dependent ROI ventilation in the injured group was 34.5 ± 5.6% versus 22.0 ± 2.4% at PEEP 5 cmH 2 O with V T 7 ml/kg.

Regional perfusion
No interactions between PEEP and group (injured/control) and/or between V T and group (injured/control) were observed for regional perfusion. Figure 5 shows the distribution of regional pulmonary perfusion at different PEEP and V T levels. A two-way interaction between PEEP and V T was observed for each ROI: nondependent (p = 0.030), middle (p = 0.006), and dependent (p = 0.001).

Discussion
Regional ventilation of the most dependent lung region was greater with larger PEEP and tidal volume in piglets with saline lavage-induced lung injury. In contrast, pulmonary blood flow of all lung regions was altered by larger PEEP and tidal volume both in control piglets and in piglets with saline lavage-induced lung injury. Our study demonstrates that EIT can measure physiological and supra-physiological levels of pulmonary ventilation and perfusion, and their changes associated with different MV settings such as PEEP and V T , continuously, in real-time, and at the bedside. This information may be critical for a timely evaluation of the precision and efficacy of attempts to minimize some of the main ventilator-induced lung injury (VILI) mechanisms, particularly in the presence of alveolar collapse, common in the lungs of patients with the ARDS [19]. Our data highlight that airspaces collapse in the injured group altered the pulmonary regional ventilation response to different PEEP and V T levels. Notwithstanding, when analyzing the effects of PEEP and V T on regional perfusion, airspaces collapse did not seem to play a major role. Regarding regional perfusion, the mutual interaction between PEEP and tidal volume seemed to be more relevant. Our findings on regional ventilation may be interpreted as predominantly gravity-related, since the ventilation of the more dependent region exhibited a greater increase in its percentage with increasing V T and PEEP in the injured group than in the control one. On the other hand, our findings on regional perfusion, where all regions were altered and the presence or not of lung injury did not have an effect, suggest that V T and PEEP interacted all over the lung parenchyma on the redistribution of regional perfusion to the dependent zones [20,21], also in agreement with pulmonary blood volume tidal redistribution [22].
Pulmonary perfusion can be affected by anesthesia and by mechanical ventilation, affecting arterial blood oxygenation [23][24][25][26][27]. The greatest determinant of pulmonary blood perfusion distribution is the difference between alveolar and pulmonary capillary pressures [28]. Perfusion increases down the gravitational gradient in the lung, due to hydrostatic forces and structural factors. Atelectasis is pronounced in the dependent lung regions, where perfusion of non-ventilated lung parenchyma produces a shunt effect of around 8-10% of cardiac output in healthy lungs [24], and a greater shunt in injured lungs. In addition, nongravitational inhomogeneity of perfusion can reduce arterial blood oxygenation. Elevated airway pressures applied to the lung during mechanical ventilation can compromise venous return and redirect blood flow to dependent lung regions, with the effect being more marked at higher levels of PEEP and/or V T [29]. Our data demonstrate that these effects can be regionally monitored, measured, and tracked in real-time and at the bedside by EIT. Our distributions of regional pulmonary perfusion at different PEEP and V T levels may have been also partly influenced through hypoxic pulmonary vasoconstriction (HPV) [30]. In conditions of atelectasis, as in the injured group here, HPV could have been greater at low levels of PEEP and V T . Noteworthy, we have shown before that perfusion is similar, whether a region is collapsed is aerated but non-ventilated [13], findings that fit with observations by Benumof [31]-that HPV is the main mechanism of reduced blood flow in atelectatic regions, not mechanical obstruction.
The main limitation of our study may be in the animal model that is not capable of reproducing all of the key characteristics of lung injury in humans, for example in ARDS patients, where EIT titration of MV may be most helpful. Any animal model is relevant for only limited aspects of ARDS pathophysiology. However, if the specific characteristics of an animal model are carefully taken in account, and its findings are interpreted in the context of the study limitations, animal investigations can provide valid assessments of relevant elements of ARDS in human patients. The surfactant depletion by saline lavage model was developed based on the observation that ARDS is associated with depletion of surfactant from the air spaces and reduced concentrations of surfactant-associated proteins in bronchoalveolar lavage fluid [32]. However, determining precisely the extent to which the lung injury is caused by the saline lavage, by mechanical ventilation, or both remains challenging. Depletion of surfactant may be associated with lung injury via two mechanisms: greater lung collapse and increased likelihood of mechanical injury during repeated cycles of airspaces opening/closure, and impaired alveolar host defenses. Characteristically, saline lavage leads to almost immediate hypoxemia, which may be rapidly reversed by recruitment maneuvers, suggesting that the gas exchange abnormalities reflect collapsed alveoli with otherwise intact alveolar walls. The saline lavage by itself has little consequence in terms of permeability changes or inflammation [33], although TNF-alpha is detectable in lavage fluid. Despite surfactant depletion being an important feature of ARDS in humans, it usually appears as a consequence rather than the primary cause of lung injury [34]. In ARDS, surfactant abnormalities occur because of injury to the alveolar epithelium and exudation of protein-rich edema fluid into the alveolar spaces. Saline lavage of the lungs results in surfactant depletion in the absence of major alveolar epithelial damage. Epithelial damage occurs only when the saline lavage is followed by an injurious ventilatory strategy. Therefore, surfactant depletion followed by mechanical ventilation simulates established ARDS and provides information about the consequences of surfactant depletion, but it is less useful in modeling the initial pathophysiological mechanisms that lead to ARDS. The major advantage of the saline lavage model is that it provides an ideal way to test the effects of different ventilatory strategies on the development of lung injury [33].
The approach applied in this study for the determination of regional pulmonary perfusion by EIT presents some limitations. The regional time-impedance curves resulting from 10% NaCl injection [16,35,36] were fitted on a pixel-by-pixel basis by a gammavariate function to quantitatively assess regional perfusion [17,18]. The subtraction of the cardiac component of perfusion from mixed pixels while maintaining the net lung component by using an EIT gamma variate algorithm is an essential feature of this method. Such was made possible by the fitting also of the early cardiac component, also on a pixel-by-pixel basis by a corresponding gamma-variate function. However, at the edges of the heart, some challenging overlap between the behavior of lung and heart tissue may have remained, sometimes causing uncertainties in the double-function fitting process applied. Although the location of the cardiac chambers in humans could make this differentiation easier in the clinical setting, more studies on this subject are needed.
The maximal slope method applied is based on the assumption that no tracer leaves the ROI before the peak artery concentration is reached. Such assumption could be violated in the presence of the combination of high blood flow and low blood volume, leading to small estimation errors, provided that blood volume decreases by no more than one order of magnitude.
It is possible that some sodium chloride diffused outside the pulmonary blood vessels. In this situation, the solute that remains in the vessel will leave the lungs through the venous drainage while the diffused solute will tend to stay in the lungs, violating the conservation of mass principle (unless an extravascular compartment is accounted for). For the calculations of the maximal slope; however, the conservation of mass could still be applied correctly, since one of the assumptions is that there is no outflow of hypertonic saline before the peak of the pulmonary artery input function. In this case, the conservation of mass implies that all the solute that reaches the ROI (feeding vessel and extravascular compartment together), irrespective of whether it remains inside the vessels or not, came through the feeding artery.
EIT imaging has 2D and 3D features (the electrodes are placed within a single plane, but the finite mesh is 3D). Although reasonably large, the thickness of the EIT crosssectional slice (~15 cm) may represent different proportions of the lungs, depending on the size and shape of the animal [4], without guarantee that most of the lung is represented in all animals, unless other imaging techniques are employed in parallel [37]. Finally, due to the low spatial resolution of EIT, the maximum slope time point might be slightly displaced (in time) among sub-regions within the ROI, likely causing inaccurate estimates of spatially averaged maximum slopes. The extent of this potential limitation deserves future studies.

Conclusions
Our findings suggest that ventilation of the most dependent lung region is greater with larger levels of PEEP and tidal volume in piglets with saline lavage-induced lung injury. Moreover, pulmonary blood flow of all lung regions is affected by larger levels of PEEP and tidal volume, both in control animals and in piglets with saline lavage-induced lung injury. They also suggest that EIT is a promising bedside and noninvasive clinical tool for continuous and real-time monitoring of pulmonary ventilation that can be especially useful in severe mechanically ventilated patients such as those with ARDS. EIT can help to optimize mechanical ventilation settings, detect complications such as derecruitment, and provide estimates of perfusion distribution. More clinical validation studies are awaited to explore the full potential of the technology.