
==== Front
BJA Open
BJA Open
BJA Open
2772-6096
Elsevier

S2772-6096(24)00046-7
10.1016/j.bjao.2024.100302
100302
Original Research Article
Variable positive end-expiratory pressure in an experimental model of acute respiratory distress syndrome: an advanced ventilation modality☆
Südy Roberta 1
Diaper John 1
Bizzotto Davide 2
Dellàca Rafaelle 2
Petak Ferenc 3
Habre Walid 1
Dos Santos Rocha Andre andrerocha@medicina.ulisboa.pt
14⁎
1 Unit for Anaesthesiological Investigations, Department of Anaesthesiology, Pharmacology, Intensive Care and Emergency Medicine, University of Geneva, Geneva, Switzerland
2 Dipartimento di Elettronica, Informazione e Bioingegneria, Politecnico di Milano, Milan, Italy
3 Department of Medical Physics and Informatics, University of Szeged, Szeged, Hungary
4 Department of Anaesthesia, North Lisbon University Hospital Centre, Lisbon, Portugal
⁎ Corresponding author. Unit for Anaesthesiological Investigations, Department of Anaesthesiology, Pharmacology, Intensive Care and Emergency Medicine, University of Geneva, Geneva, Switzerland. andrerocha@medicina.ulisboa.pt
14 8 2024
9 2024
14 8 2024
11 10030214 3 2024
1 7 2024
© 2024 The Authors
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Background

Introducing variability in tidal volume, ventilatory frequency, or both is beneficial during mechanical ventilation in acute respiratory distress syndrome (ARDS). We investigated whether applying cycle-by-cycle variability in the positive end-expiratory pressure (PEEP) exerts beneficial effect on lung function in a model of ARDS.

Methods

Rabbits with lung injury were randomly allocated to receive mechanical ventilation for 6 h by applying a pressure-controlled mode with constant PEEP of 7 cm H2O (PC group: n=6) or variable PEEP (VEEP) with a coefficient of variation of 21.4%, range 4–10 cm H2O (PC-VEEP group; n=6). Lung oxygenation index (Pao2/FiO2) after 6 h of ventilation (H6) was the primary outcome and respiratory mechanics, lung volume, intrapulmonary shunt, and lung inflammatory markers were secondary outcomes.

Results

After lung injury, both groups presented moderate-to-severe ARDS (Pao2/FiO2 <27 kPa). The Pao2/FiO2 was significantly higher in the PC-VEEP group than in the PC group at H6 (12.3 [sd 3.5] vs 19.2 [7.2] kPa, P=0.013) and a lower arterial partial pressure of CO2 at 1–3 h (P<0.02). The ventilation-induced increases in airway resistance and tissue elastance were prevented by PC-VEEP. There was no evidence for a difference in minute volume, driving pressure, end-tidal CO2, lung volumes, intrapulmonary shunt fraction, and cytokines between the ventilation modes.

Conclusions

Prolonged mechanical ventilation with cycle-by-cycle VEEP prevents deterioration in gas exchange and respiratory mechanics in a model of ARDS, suggesting the benefit of this novel ventilation strategy to optimise gas exchange without increasing driving pressure and lung overdistension.

Keywords

gas exchange
lung function
lung oxygenation index
variable ventilation
ventilator-induced lung injury
Handling editor: Susan M. Goobie
==== Body
pmcThe physiological benefits of variable ventilation (VV), a mode that applies breath-by-breath variability to reproduce the natural variation of spontaneous breathing, have been demonstrated in healthy and injured lungs.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 The deterioration of lung structure and function that is observed during conventional ventilation (pressure- or volume-controlled) was prevented by the application of VV in animal models with healthy lungs1,10,12 and diseased lungs with acute respiratory distress syndrome (ARDS),2,4,9 asthma,7 chronic obstructive pulmonary disease,6 and pulmonary fibrosis.5 These experimental findings were corroborated by two clinical trials using VV in humans with healthy lungs and ARDS, demonstrating improved oxygenation, respiratory compliance, and lower ventilation deadspace.3,8 The advantages of VV stem from the fact that breath-by-breath variations in tidal volume (VT), ventilatory frequency (VF), or both improve gas exchange, respiratory mechanics, and alveolar recruitment while reducing shear stress and inflammation1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 as a result of the non-linear properties of the respiratory system.13 Furthermore, VV is also beneficial for the ventilation–perfusion matching and surfactant production.14

Previous studies of VV were based on variations of only VT, VF, or both while maintaining the positive end-expiratory pressure (PEEP) constant. However, it has been well established that PEEP is of paramount to optimise gas exchange in protective ventilation strategies, particularly in the presence of ARDS.15,16 Therefore, we hypothesised that an advanced ventilation strategy characterised by cycle-by-cycle variability in PEEP can offer advantages in gas exchange and respiratory function compared with conventional, constant PEEP. To test this hypothesis, we compared the effects of variable PEEP (PC-VEEP) with those of constant PEEP (PC) over a 6-h period of conventional pressure-controlled ventilation in a rabbit model of ARDS.

Methods

Ethical statement

The study was approved by the Animal Welfare Committee of the Canton of Geneva and the Experimental Ethics Committee of the University of Geneva, Switzerland (no. GE 144/20, approved 12/08/2020). All procedures were performed according to the current animal protection laws of Switzerland (LPA, RS455). The current work follows the ARRIVE guidelines for reporting animal studies.

Experimental animals

Twenty adult New Zealand White rabbits (male n=9, female n=11, aged 20 weeks, weighing 2.7 kg on average, with a range of 2.3–3.35 kg) were involved in the current study. Animals were purchased from the farm of University of Geneva (Arare, Geneva, Switzerland) and were delivered 1 week before the experiments to allow acclimatisation. Food and water were provided ad libitum before the experiments.

Study design and protocol

The scheme of the prospective randomised study protocol is depicted in Figure 1. After completing surgical preparation, a 15- to 30-min period was allowed for the stabilisation of the haemodynamic and respiratory conditions. A recruitment manoeuvre was performed to standardise lung volume history, followed by recording the baseline (BL) measurements that included ventilation (VT, VF, and inspiratory pressure [Ppeak]) and blood gas parameters, respiratory system mechanics using oscillometry, and the multiple breath washout technique to characterise lung volume. Lung injury was then induced using a triple-hit injury, as described below. After this intervention, another recruitment manoeuvre was performed and the ventilation parameters were set: PEEP set to 7 cm H2O, Ppeak set to achieve VT of 7 ml kg−1, and VF set to achieve normocapnia. After a period of stabilisation, these measurements were performed again to establish the respiratory outcomes before beginning mechanical ventilation (H0). Animals were randomised to receive 6 h of conventional PC with constant PEEP or variable PEEP (PC-VEEP). Ventilation parameters, blood gas, and respiratory mechanics were characterised hourly during the 6-h ventilation period. After 6 h (H6), the final set of measurements was performed, which was the same as the initial measurements at H0 but also included lung volume and intrapulmonary shunt fraction (Qs/Qt) assessments.Fig. 1 Study protocol. Acute respiratory distress syndrome (ARDS) induction was achieved with injurious ventilation with positive end-expiratory pressure (PEEP) 0 and 100% fraction of inspired oxygen, total lung lavage, and i.v. endotoxin. BG, blood gas; BL, baseline; FOT, forced oscillometry; H0 to H6, hourly measurements from hour 0 to hour 6; MBW, multiple breath washout; PC, pressure-controlled ventilation; VEEP, variable PEEP.

Fig. 1

Primary and secondary outcomes

The primary outcome of the present study was the oxygenation index (partial pressure of oxygen in arterial blood divided by fraction of inspired oxygen [Pao2/FiO2]) after 6 h of mechanical ventilation. Secondary outcomes included airway pressure (Paw), VT, VF, airway resistance (Raw), respiratory tissue elastance (H) and damping (G), Qs/Qt, end-expiratory lung volume (EELV), lung clearance index (LCI), and lung inflammatory markers.

Experimental procedures

Anaesthesia and surgical preparation

Animals in all groups were anaesthetised with an i.m. injection of ketamine 25 mg kg−1 and xylazine 3 mg kg−1. After infiltration of the anterior cervical region with lidocaine 1%, a surgical tracheostomy with a 3.5-mm uncuffed tube was performed. The ear vein was cannulated with a 24-G catheter for the administration of i.v. anaesthesia and analgesia (propofol 10 mg kg−1 h−1, fentanyl 5 μg kg−1 h−1). After confirming adequate depth of anaesthesia by the absence of movement in response to stimuli and haemodynamic monitoring (stable heart rate and arterial blood pressure), neuromuscular block was performed with atracurium 0.6 mg kg−1 h−1. Body temperature was maintained at 38 (sd 1) °C with a heating pad (Harvard Apparatus, South Natick, MA, USA). I.V. fluid replacement was administered with Ringer's acetate 2 ml kg−1 h−1. Then, the femoral artery and internal jugular vein were cannulated for arterial and venous blood sampling and invasive pressure monitoring. At the end of the study protocol, still under general anaesthesia, the animals were euthanised by injecting a single i.v. dose of pentobarbital (50 mg kg−1).

ARDS model

Lung injury was induced using a triple-hit injury involving surfactant depletion with lung lavage, i.v. lipopolysaccharide, and injurious ventilation with PEEP 0 cm H2O, VT 10 ml kg−1 and FiO2 100% for 20–30 min, as described previously.17

Mechanical ventilation modalities

Mechanical ventilation was applied by a computer-controlled, custom-made, blower-based ventilator with software that continuously recorded tracheal airflow (V′), Paw, and VT, as described previously.6 For BL measurements, all ventilation parameters were set identically between groups: Ppeak to deliver an average VT of 7 ml kg−1, PEEP of 3 cm H2O, FiO2 of 40%, and inspiratory-to-expiratory ratio (I:E) of 1:2.

After induction of lung injury (H0), the Ppeak was set to deliver VT of 7 ml kg−1, PEEP was set to 7 cm H2O and FiO2 was titrated between 60% and 90% to achieve oxygen peripheral saturation above 92%. At this time point, animals were randomised to receive either PC with the aforementioned parameters or PC-VEEP. The PC-VEEP group received the same average PEEP (7 cm H2O) with a cycle-by-cycle variability following a Gaussian distribution and a coefficient of variation of 21.4% (PEEP range 4–10 cm H2O). A representative period of PC-VEEP application is presented in Figure 2.Fig. 2 Representative tracing of the airway pressure from one animal during the application of pressure-controlled ventilation with variable positive end-expiratory pressure (PC-VEEP). The mean PEEP was set to 7 cm H2O, and the driving pressure was 12 cm H2O.

Fig. 2

The Ppeak and driving pressure were kept constant throughout the 6 h of ventilation (H0–H6), and VF was adapted to achieve constant minute volume and normocapnia (end-tidal CO2 of 5.5–6%).

Measurements

Blood pressure, electrocardiogram, and Paw were continuously recorded using PowerLab model 8/35 and LabChart 7 (ADInstruments, Dunedin, New Zealand). Peripheral oxygen saturation was continuously measured with LifeVet P Pulse Oximeter (Eickemeyer Veterinary Equipment, Appenzell, Switzerland). Ventilation parameters were continuously recorded by the ventilator and averaged over periods of 60 min to measure mean VT, mean VF, and mean Ppeak. Blood analyses were performed from arterial and central venous samples using a point-of-care blood gas analyser (i-Stat; Abbott Laboratories, Chicago, IL, USA) to measure the partial pressure of oxygen, carbon dioxide, and lactate concentrations in the arterial and venous blood. The Qs/Qt was calculated using the Berggren equation.18

The oscillatory impedance spectra of the respiratory system (Zrs) were analysed using a method described previously.19 Briefly, 2 cm H2O peak-to-peak amplitude pseudorandom oscillations were applied for 10 s during end-expiratory pauses (15 non-integer multiples between 0.5 and 21 Hz) by the computer-controlled ventilator turbine. The flow (V′) was measured using a pneumotachograph (PNT 3700 Hans Rudolph Inc., Shawnee, KS, USA) connected to a differential pressure transducer (Honeywell model 24PCEFA6D, Charlotte, NC, USA). The Paw was measured by a second pressure transducer connected to a side port of the tracheal cannula. Zrs (Zrs=Paw/V′) was calculated using Fast Fourier Transformation from the 10-s-long recordings with 4-s time windows and 95% overlap. The measured impedance was fitted to a model to obtain Raw and tissue damping and elastance. The Raw represents the flow resistance of the airways, whereas the tissue viscoelastic parameters reflect the energy loss (damping) and the energy storage capacity (elastance) of the respiratory tissues.

An ultrasonic flowmeter (Spiroson Scientific; ECO Medics AG, Dürnten, Switzerland) using helium as tracer gas was used to apply a multiple breath washout technique described in previous studies.5,20 The washout curve was analysed to obtain the EELV and LCI before and after 6 h of ventilation. The LCI, an index which reflects the inhomogeneity of alveolar ventilation, was determined as the number of turnovers required to decrease the end-tidal helium concentration to 1/40th of the initial value before washout, as described previously.21

After euthanasia, the bronchoalveolar lavage fluid (BALF) and the frozen lung tissue homogenate were analysed with enzyme-linked immunosorbent assay (ELISA) to quantify interleukin (IL)-6 and IL-8 and tumour necrosis factor (TNF-α), as described in Supplementary material.

Sample size

Based on previous work using an animal model of ARDS2 demonstrating that VV improved oxygenation after 5 h of mechanical ventilation (Pao2 23.1 [4.0] vs 15.9 [3.1] kPa for VV and conventional ventilation, respectively), we hypothesised that VEEP increases oxygenation to a similar extent in comparison with constant PEEP. Accordingly, and assuming an alpha level of 0.05 and a power of 80%, the sample size estimation resulted in six rabbits per group. However, our previous experiments revealed a large intersubject variability (25%) and premature animal death (10–20%) when inducing lung injury, and thus, we added four supplementary animals per group to account for this.

Statistical methods

The statistical analysis plan and the outcome variables were approved by the authors before analyses began. Data are presented as mean (sd). Two-way repeated measure analysis of variance (anova) with Holm–Šídák post hoc tests were used to assess differences between ventilation modes (PC vs PC-VEEP) and time points (BL and H0 to H6). Relative changes between H0 and H6 were compared using an unpaired Student's t-test or a Mann–Whitney test for data with normal and non-normal distribution, respectively. The statistical analysis was performed using SigmaPlot (Version 14.5; Systat Software, Inc., Chicago, IL, USA). Results were considered statistically significant if P<0.05, and all P-values are two-sided.

Results

Twenty rabbits were involved in the present study. Four animals did not survive the 6-h experimental protocol as a result of severe ARDS, pneumothorax, and technical problems (two randomised to PC and two to PC-VEEP). Four other animals were excluded from the final analysis for not meeting ARDS criteria (Pao2/FiO2 >40 kPa, two randomised to PC and two to PC-VEEP). Accordingly, six rabbits were included in the final analysis in both PC and PC-VEEP groups.

Changes in arterial blood gas parameters

After inducing lung injury, marked and significant decreases in Pao2/FiO2 and increases in Paco2 were observed (BL vs H0, P<0.001 for both; Fig 3), demonstrating the presence of moderate-to-severe ARDS according to the Berlin criteria.22 The significant interactions between the ventilation mode and time for Pao2/FiO2 and Paco2 (P<0.05 for both) obtained by anova indicate that the ventilation mode significantly affected the temporal changes in both blood gas parameters. The Pao2/FiO2 values after PC-VEEP ventilation became significantly higher when compared with PC after 5-h ventilation (P=0.027 and P=0.013 at H5 and H6, respectively). This finding was associated with lower Paco2 in the ventilation period of 1–3 h (P=0.002, 0.01, and 0.014 for H1, H2, and H3, respectively). Application of PC-VEEP prevented the deterioration in Pao2 during the 6-h ventilation period, as demonstrated by significant differences in the percentage change of Pao2/FiO2 between H0 and H6 (P=0.046; Fig 3, top right panel).Fig. 3 Arterial blood gas parameters. Shown are the ratio of partial pressure of oxygen (Pao2, kPa) to the fraction of inspired oxygen (FiO2) and the partial pressure of carbon dioxide (Paco2) before (BL) and after (H0) the induction of lung injury and during the 6 h of ventilation (H1–H6) with pressure-controlled ventilation and constant PEEP (PC, black) or variable PEEP (PC-VEEP, red). The relative change in percentage (▵) between H0 and H6 is represented on the right for Pao2/FiO2 and Paco2 for groups PC and PC-VEEP. Values are mean (standard deviation) or median (inter-quartile range). ∗P<0.05 for PC vs PC-VEEP. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Ventilation parameters

The ventilation parameters are summarised in Figure 4. The induction of ARDS resulted in a significant increase in VF (P=0.02) and Ppeak (P<0.001) irrespective of the ventilation mode. As the Ppeak was set to achieve a VT of 7 ml kg−1 at H0, VT remained stable between BL and H0 in both groups. There was a significant increase in VF in group PC from H3 (P<0.05) and a significant decrease in VT from H2 of mechanical ventilation (P<0.05), whereas VF and VT did not differ from H0 to H6 in the PC-VEEP group (P=0.531 and 0.112, respectively). Accordingly, the relative change in VF in the PC group during the 6-h ventilation protocol was higher than in the PC-VEEP group (P=0.002; Fig 4, top right panel). No significant difference was observed between the protocol groups in Ppeak or its relative change (P=0.262 for anova groups, and P=0.110 for relative change). For the PC and PC-VEEP groups, there was no evidence for significant differences regarding minute volume (688 [99] vs 735 [147] ml min−1, respectively, at H6; P=0.546) and end-tidal CO2 (6.7 [0.8] vs 6.7 [0.9]%, respectively, at H6; P=0.962) during the 6 h of mechanical ventilation.Fig. 4 Changes in ventilatory frequency (VF, min−1), tidal volume (VT, ml kg−1), and airway inspiratory pressure (Ppeak, cm H2O) before (BL) and after (H0) the induction of lung injury, and during the 6 h of ventilation (H1–H6) with pressure-controlled ventilation and constant PEEP (PC, black) or variable PEEP (PC-VEEP, red). The relative change in percentage (▵) between H0 and H6 is represented on the right panels for groups PC (grey) and PC-VEEP (red). Values are mean (standard deviation) or median (inter-quartile range). ∗P<0.05 for PC vs PC-VEEP. †P<0.05 vs H0. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Respiratory function

There was no evidence for a statistical difference between the protocol groups in the mechanical parameters of the respiratory system under BL condition: Raw (12.6 [1.8] vs 11.5 [2.6] cm H2O s L−1, P>0.05), damping (96.7 [14.9] vs 87.4 [9.4] cm H2O L−1, P>0.05), and elastance (348 [16] vs 337 [43] cm H2O L−1, P=0.978). Figure 5 shows the relative changes of these outcomes during the 6-h ventilation period. After 6 h of ventilation, Raw and elastance elevated significantly in group PC (P=0.017 and P<0.001, respectively), whereas these elevations were prevented by group PC-VEEP (P>0.05 and P=0.618, respectively).Fig. 5 Relative changes in percentage (▵) between H0 and H6 for the respiratory mechanical parameters. Airway resistance (Raw), tissue damping (G), and tissue elastance (H) in the group with pressure-controlled ventilation and constant PEEP (PC, grey) or variable PEEP (PC-VEEP, red). ∗P<0.05 for PC vs PC-VEEP. Values are median (inter-quartile range). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

The ventilation mode had no significant effect in the relative change in EELV after 6 h of ventilation (–13.3 [10.6]% vs +3.0 [30.3]% for the PC and PC-VEEP groups, respectively, P=0.312), and in LCI (16.5 [18.3] vs 3.7 [49.0], P=0.638).

The induction of lung injury resulted in significant increases in Qs/Qt (7.36 [1.2] vs 40.6 [13.7]% BL and 8.7 [0.3]% vs 32.1 [8.5]% H0 for the PC and PC-VEEP groups, respectively, P<0.001 for both) without affecting the lactate concentration. No difference in Qs/Qt between the protocol groups was detectable after 6 h of mechanical ventilation (52.5 [3.4]% vs 46.5 [14.1]% for the PC and PC-VEEP groups, respectively, P=0.146). Conversely, lactate concentrations elevated significantly after 6 h ventilation in both groups (1.8 [0.7] vs 5.7 [1.8] mM and 1.9 [1.5] vs 3.7 [1.3] mM at H0 and H6 in the PC and PC-VEEP groups, respectively, P<0.001 for both) despite stable haemodynamic conditions (data not shown).

Inflammatory markers

The total protein in the BALF and in the lung frozen tissue homogenates, and the inflammatory cytokines, TNF-α, IL-1β, and IL-8 did not differ between the groups after 6 h mechanical ventilation (see Supplementary material).

Discussion

In the present study, we evaluated the benefits of an advanced ventilation strategy by introducing cycle-by-cycle variability to PEEP and compared gas exchange and respiratory function parameters with those obtained during 6 h of conventional ventilation in a rabbit model of moderate-to-severe ARDS. We demonstrated that the long-term deterioration in blood gas parameters observed during conventional ventilation was mitigated by the application of VEEP. This improvement in lung oxygenation was associated with the protective potential of VEEP against ventilation-induced deteriorations in Raw and respiratory tissue elastance. VEEP allowed the maintenance of minute volume and driving pressure without the need for adjustments in VT and VF. The inflammatory profile of the lungs was not affected by the advanced ventilation modality.

The injury model used in this work was the previously described triple-hit model of ARDS,17 which combines surfactant depletion, i.v. lipopolysaccharide, and injurious ventilation. The rationale for this experimental injury is to mimic the pathophysiological aspects of ARDS in humans, including tissue inflammation, alveolar and interstitial oedema, capillary leak, lung stiffness, and ventilation–perfusion mismatch. Previous imaging and histological analysis of this triple-hit ARDS model used with rabbits demonstrated considerable inflammation, lung condensation, neutrophilic infiltration, alveolar septal thickening, hyaline membranes, and increased protein and cell count in the bronchial lavage fluid.4,17

The application of pressure-controlled ventilation for 6 h with constant or variable PEEP demonstrated that VEEP provided a significant benefit on the primary outcome variables—lung oxygenation ratio and CO2 clearance. This finding is in line with previous data obtained after introducing variability in VT under constant PEEP in healthy lungs and in models of lung diseases.2,5, 6, 7,23, 24, 25 Interestingly, the magnitude of the benefit for gas exchange observed previously using VV waveforms is in concordance with the present results demonstrating Pao2 improvements after 5–6 h of ventilation.

To our knowledge, this is the first study to systematically characterise the potential benefit of varying PEEP levels over the conventional, monotonous ventilation. The benefits of VV have been attributed to the non-linear properties of the respiratory system.13,14 Although the principle of applying VEEP may also be related to this phenomenon, the rationale may be somewhat different: by varying the PEEP in a cycle-by-cycle manner (keeping a constant driving pressure for each cycle), the non-linear nature of the pressure–volume curve will likely generate improved VT delivery for the same level of average Ppeak. This hypothesis is supported by our data demonstrating that despite a constant level of driving pressure during the 6 h of ventilation, VT significantly decreased and VF was increased to maintain the minute volume with the conventional ventilation using constant PEEP, whereas VT and VF did not significantly differ during the 6 h under PC-VEEP. A progressive decrease in VT also indicates that derecruitment was more important with constant PEEP than with VEEP, although the same average PEEP level of 7 cm H2O was maintained in both groups.

The level of PEEP was chosen considering the respiratory system properties of this species and for this disease model based on previous research, indicating a synergistic effect of VV and PEEP between 6 and 9 cm H2O.17 Although recruitment manoeuvres (sighs) could have equally prevented derecruitment, previous research with this model has indicated that lung function deterioration was better prevented by VV than a pressure-controlled mode with regular sighs.17

As the respiratory mechanical properties are non-linear, it seems reasonable to assume that under the same average driving pressure, the ‘gain’ during higher-than-average PEEP cycles exceeds the ‘loss’ during lower-than-average PEEP cycles. Thus, varying PEEP levels seems to prevent alveolar derecruitment. It is difficult to determine whether this beneficial effect resulted solely from the higher PEEP breaths or from the variability of PEEP itself, as the varying levels in consecutive breaths could benefit from different time constants and optime regional ventilation. This mechanism is reflected in the ability of PC-VEEP to protect ventilation-induced deleterious changes in the Raw and respiratory tissue elastance (Fig 5), whereas both elastance and resistance significantly increased with PC. In addition, redistribution of pulmonary blood flow might have also played a role in the gas exchange benefit of VEEP, as introducing variability in VT or pressure support was shown to redistribute pulmonary blood flow to the well-ventilated alveolar compartments.26,27

In addition to the benefits in ventilation parameters and gas exchange, VV has been shown to prevent derecruitment and improve ventilation–perfusion matching, respiratory mechanics, and inflammation.2,4, 5, 6, 7,9,23, 24, 25, 26, 27 In the present work, no difference could be evidenced in EELV, LCI and Qs/Qt, and inflammatory markers. In line with our findings, several studies on VV did not reveal differences in the cytokine quantification despite significant improvements in gas exchange and respiratory mechanics.17,28, 29, 30, 31, 32 Considering that our study was not powered to detect a difference in these secondary outcomes with large variance, it is likely that improvements in these parameters would require more targeted investigations, in particular using lung functional imaging,4 as previously reported.

PEEP titration and the determination of the ‘optimal PEEP’ are subjects of extensive research and clinical interest, especially in the fields of ARDS and intensive care medicine.33 Maintaining an optimal PEEP balances the deleterious lung overdistension contributing to haemodynamic impairment and ventilator-induced lung injury, and underventilation causing derecruitment and atelectrauma. Another approach could have been the individualisation of PEEP for each animal. Although this approach might have identified the optimal PEEP, individualising and titrating PEEP would have introduced confounding effects into this experimental setting, which was designed to isolate a sole study variable: the coefficient of variation in PEEP. This approach represents both a limitation and a strength of the study design, as all animals in both groups received the same average PEEP. Using an advanced ventilation modality incorporating cycle-by-cycle variation in PEEP has the potential advantage of utilising the non-linear properties of lung tissue, thereby preventing derecruitment during mechanical ventilation, avoiding the need for recruitment manoeuvres and increased driving pressure.

The study has certain methodological considerations warranting discussion. In the present experiments, the mean PEEP level was set to 7 cm H2O for both groups. Applying a high PEEP level (9 cm H2O) was shown to abrogate most of the benefits of VV in a rabbit model of ARDS17 that can be attributed to an extreme right-shift in the lung pressure–volume curve. Thus, it can be anticipated that even greater benefits are incurred by using a lower mean PEEP level during the application of VEEP. Another limitation of the study design is the ventilation time (6 h), which precludes drawing conclusions about the long-term effects of VEEP. Accordingly, whether the observed benefits in gas exchange and respiratory mechanics would extend over the course of days or weeks in ventilated ARDS subjects is a subject of further investigations.

In conclusion, the cycle-by-cycle variability in PEEP during pressure-controlled ventilation has beneficial effect on gas exchange and prevents ventilation-induced respiratory mechanical impairment in an experimental model of ARDS. Introducing variability in PEEP has the promise to offer a different modality to improve mechanical ventilation in the presence of lung injury without the need to increase the driving pressure to the potentially deleterious range. Further studies on larger cohorts are warranted for a more detailed investigation of the effects of VEEP and its optimisation under various lung conditions.

Authors’ contributions

Conceived the study protocol: RS, WH, ADSR.

Data collection: RS, JD, ADSR.

Performed animal experiments: RS, JD, ADSR.

Created software and hardware for the experiments: DB, RD.

Data analysis: FP, HW, RS, ADSR.

Drafted the manuscript: FP, HW, RS, ADSR.

All authors read and approved the final manuscript.

Funding

This work was performed with internal funding from the Unit for Anaesthesiological Investigations and a scientific grant atributed to ADSR from Ligue Pulmonaire Genevoise, Geneva, Switzerland.

Data availability statement

The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.

Declaration of interest

The authors declare that they have no conflicts of interest.

Appendix A Supplementary data

The following are the Supplementary data to this article:Multimedia component 1

Multimedia component 1

Multimedia component 2

Multimedia component 2

Acknowledgements

The authors sincerely thank Xavier Belin and Sylvie Roulet for animal handling and preparation as well as technical help with the animal experiments.

☆ A preliminary account of the results has been given during the EuroAnanesthesia conference, Milan, Italy, 4–6 June 2022 and during the SwissAnaesthesia Congress, Interlaken, Switzerland, 3–5 November 2022.

Appendix A Supplementary data to this article can be found online at https://doi.org/10.1016/j.bjao.2024.100302.
==== Refs
References

1 Arold S.P. Suki B. Alencar A.M. Lutchen K.R. Ingenito E.P. Variable ventilation induces endogenous surfactant release in normal Guinea pigs Am J Physiol Lung Cell Mol Physiol 285 2003 L370 L375 12851212
2 Boker A. Graham M.R. Walley K.R. Improved arterial oxygenation with biologically variable or fractal ventilation using low tidal volumes in a porcine model of acute respiratory distress syndrome Am J Respir Crit Care Med 165 2002 456 462 11850336
3 Boker A. Haberman C.J. Girling L. Variable ventilation improves perioperative lung function in patients undergoing abdominal aortic aneurysmectomy Anesthesiology 100 2004 608 616 15108976
4 Dos Santos Rocha A. Fodor G.H. Kassai M. Physiologically variable ventilation reduces regional lung inflammation in a pediatric model of acute respiratory distress syndrome Respir Res 21 2020 288 33129315
5 Dos Santos Rocha A, Peták F, Carvalho T, Habre W, Balogh AL. Physiologically variable ventilation prevents lung function deterioration in a model of pulmonary fibrosis. J Appl Physiol (1985) 2022; 132: 915–924
6 Dos Santos Rocha A. Südy R. Bizzotto D. Benefit of physiologically variable over pressure-controlled ventilation in a model of chronic obstructive pulmonary disease: a randomized study Front Physiol 11 2020 625777
7 Dos Santos Rocha A. Sudy R. Petak F. Habre W. Physiologically variable ventilation in a rabbit model of asthma exacerbation Br J Anaesth 125 2020 1107 1116 33070949
8 Kowalski S. McMullen M.C. Girling L.G. McCarthy B.G. Biologically variable ventilation in patients with acute lung injury: a pilot study Can J Anaesth 60 2013 502 503 23564337
9 Lefevre G.R. Kowalski S.E. Girling L.G. Thiessen D.B. Mutch W.A. Improved arterial oxygenation after oleic acid lung injury in the pig using a computer-controlled mechanical ventilator Am J Respir Crit Care Med 154 1996 1567 1572 8912782
10 Mutch W.A. Eschun G.M. Kowalski S.E. Graham M.R. Girling L.G. Lefevre G.R. Biologically variable ventilation prevents deterioration of gas exchange during prolonged anaesthesia Br J Anaesth 84 2000 197 203 10743453
11 Samary C.S. Moraes L. Santos C.L. Lung functional and biologic responses to variable ventilation in experimental pulmonary and extrapulmonary acute respiratory distress syndrome Crit Care Med 44 2016 e553 e562 26963321
12 Walesa M. Bayat S. Albu G. Baudat A. Petak F. Habre W. Comparison between neurally-assisted, controlled, and physiologically variable ventilation in healthy rabbits Br J Anaesth 121 2018 918 927 30236254
13 Suki B. Alencar A.M. Sujeer M.K. Life-support system benefits from noise Nature 393 1998 127 128 9603516
14 Huhle R. Pelosi P. de Abreu M.G. Variable ventilation from bench to bedside Crit Care 20 2016 62 26979175
15 Brower R.G. Matthay M.A. Morris A. Schoenfeld D. Thompson B.T. Wheeler A. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome N Engl J Med 342 2000 1301 1308 10793162
16 Santa Cruz R. Rojas J.I. Nervi R. Heredia R. Ciapponi A. High versus low positive end-expiratory pressure (PEEP) levels for mechanically ventilated adult patients with acute lung injury and acute respiratory distress syndrome Cochrane Database Syst Rev 2013 2013 Cd009098 23740697
17 Fodor G.H. Bayat S. Albu G. Variable ventilation is equally effective as conventional pressure control ventilation for optimizing lung function in a rabbit model of ARDS Front Physiol 10 2019 803 31297064
18 Berggren S.M. The oxygen deficit of arterial blood caused by non-ventilating parts of the lung Acta Physiol Scand II 1942 1 92
19 Hantos Z, Daroczy B, Suki B, Nagy S, Fredberg JJ. Input impedance and peripheral inhomogeneity of dog lungs. J Appl Physiol (1985) 1992; 72: 168–178
20 Albu G. Wallin M. Hallbäck M. Comparison of static end-expiratory and effective lung volumes for gas exchange in healthy and surfactant-depleted lungs Anesthesiology 119 2013 101 110 23571638
21 Becklake M.R. A new index of the intrapulmonary mixture of inspired air Thorax 7 1952 111 116 14913505
22 Ranieri V.M. Rubenfeld G.D. Thompson B.T. Acute respiratory distress syndrome: the Berlin Definition JAMA 307 2012 2526 2533 22797452
23 McMullen M.C. Girling L.G. Graham M.R. Mutch W.A. Biologically variable ventilation improves oxygenation and respiratory mechanics during one-lung ventilation Anesthesiology 105 2006 91 97 16809999
24 Mutch W.A. Buchman T.G. Girling L.G. Walker E.K. McManus B.M. Graham M.R. Biologically variable ventilation improves gas exchange and respiratory mechanics in a model of severe bronchospasm Crit Care Med 35 2007 1749 1755 17522581
25 Mutch W.A. Harms S. Lefevre G.R. Graham M.R. Girling L.G. Kowalski S.E. Biologically variable ventilation increases arterial oxygenation over that seen with positive end-expiratory pressure alone in a porcine model of acute respiratory distress syndrome Crit Care Med 28 2000 2457 2464 10921579
26 Carvalho A.R. Spieth P.M. Guldner A. Distribution of regional lung aeration and perfusion during conventional and noisy pressure support ventilation in experimental lung injury J Appl Physiol (1985) 110 2011 1083 1092 21270348
27 Spieth P.M. Carvalho A.R. Pelosi P. Variable tidal volumes improve lung protective ventilation strategies in experimental lung injury Am J Respir Crit Care Med 179 2009 684 693 19151194
28 Froehlich K.F. Graham M.R. Buchman T.G. Physiological noise versus white noise to drive a variable ventilator in a porcine model of lung injury Can J Anaesth 55 2008 577 586 18840587
29 Funk D.J. Graham M.R. Girling L.G. A comparison of biologically variable ventilation to recruitment manoeuvres in a porcine model of acute lung injury Respir Res 5 2004 22 15563376
30 Kiss T. Silva P.L. Huhle R. Comparison of different degrees of variability in tidal volume to prevent deterioration of respiratory system elastance in experimental acute lung inflammation Br J Anaesth 116 2016 708 715 27106975
31 Spieth P.M. Carvalho A.R. Guldner A. Pressure support improves oxygenation and lung protection compared to pressure-controlled ventilation and is further improved by random variation of pressure support Crit Care Med 39 2011 746 755 21263322
32 Spieth P.M. Guldner A. Beda A. Comparative effects of proportional assist and variable pressure support ventilation on lung function and damage in experimental lung injury Crit Care Med 40 2012 2654 2661 22743778
33 Millington S.J. Cardinal P. Brochard L. Setting and titrating positive end-expiratory pressure Chest 161 2022 1566 1575 35131298
