Physiologically variable ventilation reduces regional lung inflammation in a pediatric model of acute respiratory distress syndrome

Acute respiratory distress syndrome (ARDS), characterized by the acute onset of severe hypoxic respiratory failure, remains a prevalent and often lethal condition in intensive care [1]. Although mechanical ventilation is a crucial life-saving treatment for ARDS, there is a considerable body of evidence indicating that prolonged positive-pressure ventilation can initiate, perpetuate or aggravate injury to lung tissue [2, 3]. The resulting exaggerated mechanical stress, along with the monotonous alveolar opening and closing, exerts shear stress and increased strain in the lung tissue [4], conditions that contribute to ventilator-induced lung injury (VILI).

While various modalities of mechanical ventilation have been proposed to reduce VILI [5‐8], protective ventilation with monotonous tidal volume (VT) may not be the only rational strategy. In recent years, it has been advocated that mechanical ventilation reproducing the natural variability of breathing is better than conventional modes [9, 10]. Variable ventilation has been shown to be beneficial for gas exchange and respiratory mechanics in various animal models with healthy [11‐13] or injured lungs, including ARDS [14‐17]. We have previously established a variable ventilation modality using pre-recorded breathing patterns of healthy animals [18]. This physiologically variable ventilation (PVV) is characterized by breath-to-breath variability of VT and respiratory rate, in contrast to the monotonous conventional ventilation modes.

Recent interest in variable ventilation stems from the need to reduce cyclic alveolar reopening during mechanical ventilation, especially in injured lungs, to avoid development or propagation of lung inflammation, atelectasis and subsequent hypoxemia [19]. Whereas some studies demonstrated the beneficial effect of introducing variability into lung recruitment [20, 21], and others reported improvement in global respiratory mechanical and functional parameters [11‐18], there is still a lack of detailed knowledge about the pathophysiological background related to the functional and regional behavior of the lung during variable ventilation. Moreover, the potential of PVV in the context of pediatric ARDS has not been characterized. To investigate the effect of PVV, lung functional and structural changes were compared to those obtained with conventional monotonous ventilation in normal lungs and ARDS, in a pediatric model. Global respiratory parameters were measured to characterize the overall lung condition. Regional lung aeration, pulmonary perfusion and inflammation were assessed by functional imaging using positron-emission tomography (PET) and single-photon emission computed tomography (SPECT) combined with X-ray computed tomography (CT).

A more detailed description of the methods can be found in Additional file 1.

In the present study, a combined approach consisting of lung functional and structural assessment was used to investigate differences in the global and regional effects of PVV and the conventional monotonous pressure-controlled mode in a pediatric model of normal lungs and ARDS. The use of PVV decreased pulmonary inflammation, as assessed by ¹⁸F-FDG uptake, independent of lung condition. The decreased lung inflammation observed with PVV was also detected as an improvement in respiratory tissue elastance. Neither the use of PCV nor PVV affected blood gas and lung morphology indices.

Respiratory system mechanical parameters obtained in BL conditions or following induction of lung injury exhibited excellent agreement with previous data from the same species with similar weight range [14‐16, 28]. Furthermore, the time course of the respiratory mechanical parameters over 5 h of ventilation in the control groups is in accordance with that observed previously in an experimental model using adult rabbits [18].

Since increases in tissue damping and elastance reflect lung volume loss and stiffening of the lung tissue [29, 30], the lack of an increase of elastance in the PVV-CTRL group suggests that lung derecruitment did not occur, and this conclusion is also supported by the lower inspiratory driving pressure achieved in this group. Moreover, the significant differences in elastance between the PCV-ARDS and PVV-ARDS groups observed after the 5-h ventilation suggest a protective effect of the variability on the conservation of lung volume in the presence of ARDS. Studies using models of mild-to-moderate lung injury have found similar beneficial effects on respiratory mechanics for variable ventilation [15, 31], and this protective effect was not observed in the presence of more severe ARDS [14].

Global and regional lung metabolic activity were measured by ¹⁸F-FDG uptake, a reliable biomarker of inflammation in the lung [32]. This marker is indicative of neutrophil activation in acute lung injury and ARDS [33‐35]. Previous studies have shown that voxelwise ratio of lung parenchyma and air content influences ¹⁸F-FDG uptake quantification, requiring normalization for the tissue fraction [26, 36], which was performed in the current study. After 5 h of ventilation, we observed significantly lower indices of global and regional lung inflammation in the animals ventilated with PVV. Specifically, a significantly higher inflammatory activity characterized the well aerated and non-dependent lung zones, both in control and injured groups. This finding is consistent with results from previous experiments studying injured lungs, in which lung inflammation assessed by ¹⁸F-FDG uptake was correlated with regional strain [37, 38]. The significantly lower inflammation associated with PVV may be explained by the fact that the variability of the delivered VT contributes to tidal recruitment [12, 15], therefore reducing strain in the open, aerated zones. It is worth noting that PVV exerts the most beneficial effect in the well and poorly aerated zones under both control and ARDS conditions (Fig. 5). Conversely, the collapsed non-aerated zones were obviously unaffected by ventilation modes since these units were not subjected to strain. These findings further confirm the importance of focusing on regional ventilation when assessing the benefit of ventilation strategies. SPECT imaging confirms differences in regional distribution of lung perfusion when it is related to aeration zones. However, the lower blood perfusion in the ventral lung regions as compared to the dorsal zones can be attributed to the gravity effect and/or to the blood shift to the dorsal zones as a consequence of positive pressure and lung overdistension.

The beneficial effects of PVV on respiratory mechanics and lung inflammation were not reflected in changes in blood gas parameters. The lack of improvement in oxygenation may be related in part to the more severe hypoxemia in this group, which required a higher FiO₂ (65% vs 55% in groups PVV-ARDS and PCV-ARDS, respectively). Moreover, the increase in lactate levels suggest the development of metabolic acidosis in both groups of ARDS animals, which may be the consequence of inadequate tissue oxygen delivery. Moreover, the timespan of the experiment (5 h) may be too short to detect effects on gas exchange. We may hypothesize that the more prominent inflammation observed in the PCV groups would build up and potentially cause gas exchange problems over the course of days.

The presence of ARDS was evident in the elevated lung injury score compared to control groups. In agreement with previous studies, lung injury score did not differ between the ventilation modes [14, 39]. The discrepancy between the functional and structural findings may be explained by the faster onset of functional changes, compared to the relatively longer time needed for morphological changes to become apparent. Lung inflammation quantified using BALF cell counts and pro-inflammatory cytokines, unlike in vivo imaging, did not reveal differences between the ventilation modes. In vivo imaging gives a more comprehensive measure of pulmonary inflammation at the early phase of ARDS, as it demonstrates the alveolar as well as the interstitial compartments of the lung. Additionally, ¹⁸F-FDG uptake reflects the acute metabolic activation of neutrophils and captures lung inflammation without barrier disruption, opposite to BALF neutrophils and cytokines, providing a more rapid assessment of inflammatory processes. In this context, it is worth noting that the control groups also showed increased inflammation and lung injury indices (BALF cytokines and histological injury score). These findings suggest that, despite the use of protective ventilation in the control groups, prolonged mechanical ventilation triggered the development of lung inflammation. This could potentially explain the lack of significant difference in normalized ¹⁸F-FDG uptake between control and ARDS lungs.

The similarity in the values of systemic hemodynamic parameters observed for the experimental groups is expected from the similarity in the overall lung perfusion as assessed by SPECT imaging. However, the significantly higher regional perfusion measured in the dependent zones can be attributed to the physiological distribution of lung perfusion that occurs in supine position [40] and is enhanced under positive pressure ventilation [41]. Considering the regional aeration of lung tissue, the significantly lower perfusion observed in the poorly and non-aerated zones can be explained by the hypoxic pulmonary vasoconstriction mechanism [42].

There are some methodological aspects of the present study that warrant consideration. In this study we used a Cone Beam CT [43]. This device uses less radiation and creates higher resolution images than the regular fan beam CT; however, it produces more scatter artefacts, which can alter the measured values [44, 45]. Due to technical limitations, breath gating was not performed in any of the acquisitions; therefore, basal lung areas had artefacts due to motion of the abdominal organs during breathing. The lung volume containing these artefacts was similar however, among rabbits.

The animal model to induce ARDS calls for some considerations as well. The components of the model were chosen to mimic the various pathophysiological aspects of ARDS observed in humans. Namely, intravenous LPS contributes to the inflammatory component of the disease and it has also been described to induce surfactant dysfunction [46]. Injurious ventilation using high VT combined with no PEEP contributes to development of volume- and barotrauma due to the supraphysiologic tidal volumes and respiratory pressures, whereas the absence of PEEP promotes tidal closures and exerts shear stress on the lung tissues [47]. The use of an FiO₂ of 1.0 during this injurious ventilation period facilitates lung volume loss and development of ventilation heterogeneities [48]. While the surfactant dysfunction can restore to some extent during the 5-h timeframe of the experimental protocol, the functional and morphological damage is still present in the lungs, supported by the marked and highly significant changes observed between the control and ARDS groups regardless of the ventilation mode applied.

Measurements of respiratory mechanical parameters also warrant some considerations. While Raw is mainly specific to the flow resistance of the conducting airways [49], the tissue parameters damping and elastance include not only pulmonary components but are also influenced by other structures of the total respiratory system, mainly the chest wall [49]. Previous literature attributed a chest wall contribution of approximately 30–50% to these parameters [50] and since the chest wall contribution is not expected to change after lung injury and mechanical ventilation [51], the observed changes are interpreted as being mainly of pulmonary origin. Therefore, the corresponding changes registered in tissue damping and elastance are predictably underestimating the real pulmonary changes.

Our data demonstrate the beneficial effect of variable ventilation based on a physiological breathing pattern in healthy lungs and in mild to moderate ARDS, in an experimental pediatric model. This positive effect was detected in the absence of deterioration in respiratory tissue elastance and in decreased regional lung inflammation measured by PET imaging. Ventilation for five hours with physiologically variable ventilation provided better protection on aerated lung zones than with monotonous pressure-controlled ventilation. While further studies in humans might be needed, our results suggest that the application of a physiological breathing pattern as the driving signal of mechanical ventilation may have a better lung protective ability than conventional modes in scenarios where prolonged mechanical ventilation is required.