Variable mechanical ventilation

Objective To review the literature on the use of variable mechanical ventilation and the main outcomes of this technique. Methods Search, selection, and analysis of all original articles on variable ventilation, without restriction on the period of publication and language, available in the electronic databases LILACS, MEDLINE®, and PubMed, by searching the terms "variable ventilation" OR "noisy ventilation" OR "biologically variable ventilation". Results A total of 36 studies were selected. Of these, 24 were original studies, including 21 experimental studies and three clinical studies. Conclusion Several experimental studies reported the beneficial effects of distinct variable ventilation strategies on lung function using different models of lung injury and healthy lungs. Variable ventilation seems to be a viable strategy for improving gas exchange and respiratory mechanics and preventing lung injury associated with mechanical ventilation. However, further clinical studies are necessary to assess the potential of variable ventilation strategies for the clinical improvement of patients undergoing mechanical ventilation.


INTRODUCTION
Healthy biological systems can quickly adapt to changing environmental conditions and present intrinsic functional fluctuations within each subsystem, including the cardiovascular (1) and respiratory systems. (2) Respiratory physiology is characterized by intrinsic variability in the respiratory components, including the respiratory rate (RR), tidal volume (TV), respiratory times, and respiratory flow. (3) Moreover, pulmonary insufflation has a non-linear opening characteristic. (4) The typical approach to mechanical ventilation (MV) involving the application of positive pressure and adjustments of fixed parameters on mechanical ventilators distinguishes MV from the physiology of the respiratory system.
However, in pathological biological systems, the intrinsic functional fluctuation (variation) is usually lower. The decrease in the variability of RR and TV in patients with chronic obstructive pulmonary disease (5) and prolonged weaning from MV (6) has been documented. In contrast with other systems, the variability of the respiratory system can be easily affected by efforts to improve its function. (7) In MV, ventilatory parameters are modulated by adjustments to the mechanical ventilator, which can be programmed to provide fluctuating ventilatory parameters to replicate some characteristics of spontaneous ventilation in healthy subjects.
Variable mechanical ventilation (VV) attempts to incorporate the physiological basis of spontaneous ventilation during MV and is defined as a ventilatory mode characterized by the oscillation of one or more respiratory parameters. It aims to mimic the variability observed in physiological ventilation and the natural breathing pattern, which changes from cycle to cycle, as well as other physiological parameters, including heart rate and blood pressure. (8) The concept of VV was proposed by Wolff et al. in 1992. (7) The authors postulated that the cycle-to-cycle variation in the relationship between the inspiratory and expiratory times and the level of positive-end expiratory pressure (PEEP) resulted in continuous lung recruitment, thus improving respiratory compliance and gas exchange compared with conventional mechanical ventilation (CV).
Considering that MV is a commonly used intervention in intensive care units, interest in strategies that can increase the variability of the respiratory pattern has grown recently. The objective of this study was to perform a descriptive analysis of the literature on VV, its clinical and experimental application, and the main outcomes of this technique.

METHODS
This literature review involved the search, selection, and analysis of all original articles on VV, without restriction on the period of publication and language, available in the electronic databases LILACS, Medical Literature Analysis and Retrieval System Online (MEDLINE ® ), and PubMed by searching for the terms "variable ventilation" OR "noisy ventilation" OR "biologically variable ventilation".
The inclusion criteria were experimental and clinical studies that evaluated the use of VV strategies. The exclusion criteria were letters to the editor, brief communications, case reports, historical articles, editorials, commentaries, study protocols, literature reviews, pilot studies, studies using artificial models, and studies not related to the use of VV strategies.
The databases were accessed by three of the four authors at different times, and the articles related to the research topic were selected based on the information contained in the title and abstract. The studies that each researcher selected were shared with the other researchers for confirmation. After that, the selected articles were read in full, and their references were searched to identify other studies that could meet the inclusion criteria and that might not have been identified in the initial search.

RESULTS
A total of 1,809 articles were found after searching the selected databases. Of these, 1,778 were excluded after reading the title and abstract because they did not address the central theme of the study. There were discrepancies in the number of articles (28, 30 (9,10) and 31 (9)(10)(11) ) selected by the three examiners. Five other articles were extracted from the references of the articles identified in the electronic search. The analysis of the 36 articles revealed that 24 were original studies; of these, 21 were experimental studies and three were clinical trials. The remaining were review studies (4), studies that used mathematical or computer models (3), letters to the editors (2), study protocols (2), and pilot studies (1) (Figure 1).
Among the experimental studies, the animal models used were pigs, sheep, and mice, the sample sizes varied between 10 and 64 animals, and the study groups were   The selected  items are shown in table 1.  The sample size of the clinical studies ranged from 13  to 162 individuals of both sexes. These studies evaluated   different diseases and respiratory conditions and different  CV and VV strategies. The selected items are shown in  table 2. The main findings of this review pertaining to the  method are summarized in table 3. VV performed better than CV in terms of gas exchange and respiratory mechanics during severe bronchospasm but without significant differences regarding inflammatory cytokines Spieth et al. (22) G1 = 9 animals G2 = 9 animals G3 = 9 animals G4 = 9 animals Pigs (23.8 -37kg) with ARDS ventilated for 6 hours To determine the impact of VV on pulmonary function and its effect on pulmonary parenchyma compared with conventional protective MV strategies G1: CV -ARDSnet (16) G2: VV -ARDSnet (16) (variable TV) G3: CV -OLA (23) G4: VV -OLA (23)

DISCUSSION
The use of VV and its main outcomes were reviewed. VV was evaluated in experimental studies, which reported beneficial effects related to improved lung function, gas exchange, and/or respiratory mechanics without injury and/or inflammation in the lung tissue compared with CV. Nevertheless, VV has been little explored in clinical settings, and only three clinical studies were found in the literature. In addition, these studies had distinct objectives and conflicting results regarding gas exchange.
VV methods are beneficial because they use a nonlinear system to mimic the physiological variability of the respiratory system. These methods may increase TV based on the nonlinear opening characteristics of collapsed alveoli (8) and normal alveoli. (36) Two main epiphenomena form the basis for improvements of lung function during VV: the recruitment and stabilization of pulmonary zones, which contribute to gas exchange, and improvement in the corresponding ventilation-perfusion.
The amplification of ventilated lung zones is primarily achieved by the recruitment of previously collapsed alveoli. Suki et al. (37) demonstrated that once the critical opening pressure of collapsed airways/alveoli has been exceeded, all subtended or daughter airways with lower critical opening pressures will be opened like an avalanche. Considering that the critical opening pressure values of the closed airways and the time required to reach these values may differ among pulmonary regions, the addition of MV patterns that produce distinct airway pressures and inspiratory times may be advantageous for maximizing Carvalho et al. (28) 12 animals Pigs (   VV versus protective CV decreased the incidence of delirium and cognitive dysfunction in the postoperative period by reducing the systemic proinflammatory response CV -conventional ventilation; VV -variable ventilation; G -group; TV -tidal volume; RR -respiratory rate; PEEP -positive end-expiratory pressure; FiO 2 -fraction of inspired oxygen; V min -volumeminute; PSV -pressure support ventilation; VCV -volume-controlled ventilation.

Knowledge gaps
Clinical studies that use randomized controlled clinical trials in different clinical settings, including patients with and without acute pulmonary impairment pulmonary recruitment and alveolar stabilization compared with conventional ventilatory patterns.
To stabilize open lung regions and prevent collapse during MV in healthy lungs, the production and release of surfactant is critical. (38) The release of surfactant increases exponentially with the stretch of alveolar type II cells. (39) Therefore, the high TV generated intermittently during VV may increase the alveolar stretch and thus stimulate the release of surfactant from type II alveolar cells. In healthy mice, random variations in TV promote the endogenous release of surfactant -as shown by the increase in the concentration of surfactant-associated phospholipids and the decrease in the concentration of membrane-associated phospholipids -and improve alveolar stability, thus reducing lung damage. (17) In contrast, in a model of acute respiratory distress syndrome (ARDS) caused by oleic acid, the controlled variable MV showed no benefits to the surface tension of the surfactant based on capillary surfactometry of the bronchoalveolar fluid. (18) During VV, increased gas exchange is usually a consequence of an improved ventilation/perfusion ratio, which results in the redistribution of ventilation to perfused areas and the redistribution of the lung blood flow to better ventilated lung zones. In an experimental model of ARDS, the redistribution of the perfusion occurred from dependent to non-dependent lung zones. (28) A study that used a pig model of ARDS (22) analyzed the lung blood flow using fluorescent microspheres and reported that the variability in TV associated with protective MV strategies redistributed the lung blood flow towards the caudal and peripheral zones. In this sense, VV, by reducing the average airway pressure in ventilated areas and recruiting previously collapsed areas, can reduce vascular impedance and hypoxic vasoconstriction, thus contributing to the adequacy of ventilation and perfusion.
It has been observed that during variable assisted MV (variable pressure support ventilation (PSV)), oxygenation increases despite the absence of improved aeration in dependent lung zones. Variable PSV had no effect on the recruitment or redistribution of aeration compared with conventional assisted MV (conventional PSV) in a saline lung lavage model, and it only affected the redistribution of perfusion from dependent to non-dependent lung zones. (28) In contrast, during variable controlled MV in different ARDS models, there was a reduction in pulmonary shunting (13,15,18,22,25,35) with no significant effect on the dead space, (15,26) suggesting that during variable controlled MV, the reduction in pulmonary shunting is more significant than the reduction in the dead space. Similarly, the venous mixture was reduced in variable PSV but not in conventional PSV. (29) Mutch et al. (19) demonstrated that the application of VV before and after lung injury induced by oleic acid increased respiratory sinus arrhythmia with the addition of variability compared with MV with controlled TV applied during the same periods. The loss of respiratory sinus arrhythmia that occurs in pathological conditions is a consequence of the decoupling of important biological variables. Therefore, measures to restore or enhance the coupling of these variables are advantageous because the increase in respiratory sinus arrhythmia is correlated with a reduction in intrapulmonary shunting and less dead space. (40) Variable controlled MV produced better blood oxygenation than conventional controlled MV in 14 of the 17 experimental studies involving ARDS models, (9,11,(13)(14)(15)18,22,25,30,31) non-ARDS models, (17) prolonged anesthesia, (12) selective ventilation, (20) and bronchospasm. (21) In three studies, including an experimental ARDS model induced by oleic acid (26) and a preterm lamb model, (10,27) the variable controlled MV did not improve arterial oxygenation compared with conventional controlled MV. The improvement in gas exchange was also evidenced during variable PSV compared with conventional PSV in ARDS models. (24,28,29) Nonetheless, in two clinical studies (32,33) that evaluated gas exchange, only the study by Boker et al. (32) in patients subjected to aneurysmectomy of the abdominal aorta showed significant improvement in this outcome during VV compared to the group subjected to CV. In contrast, in the study by Spieth et al. (33) of patients with acute hypoxic respiratory failure, gas exchange was similar for conventional and variable PSV. However, this study was a randomized crossover trial that used each ventilation mode for only 1 hour, which may explain the similar findings.
In several studies that used experimental models of ARDS, (9,11,(13)(14)(15)18,19,22,(24)(25)(26)30) respiratory mechanics were positively influenced by VV. There is considerable clinical evidence in ARDS models (16,41) and non ARDS models (42)(43)(44) that higher TV and inspiratory pressure can proportionately trigger or worsen ventilation-induced lung injury because the cyclic opening and closing may increase the shear stress and worsen the inflammatory response, triggering or aggravating lung injury. As in VV, higher TVs are generated randomly and intermittently, critical pressures for opening different airways and alveoli are reached, and lung regions are opened. Therefore, it has been demonstrated that, although high continuous pressures may be harmful, high sporadic pressures resulting from the use of a VV mode may not be harmful and may keep the alveoli open and help open collapsed alveoli. (35,45) Experimentally, Boker et al. (15) suggest that VV may be more protective than CV. They noted that the concentration of interleukin-8 (IL-8) in the tracheal aspirate after 5 hours of VV was lower than that after protective conventional MV, although the degree of pulmonary edema was similar for these two techniques. Corroborating this finding, Arold et al. (17) found that after 3 hours of VV in mice without lung injury, the concentration of IL-6 and tumor necrosis alpha factor (TNF-α) decreased in the bronchoalveolar lavage. These authors also observed that the amount of phospholipids in the bronchoalveolar lavage fluid in VV was similar to that of the control group, whereas this amount was significantly lower in CV, suggesting possible protection against lung injury with the use of VV.
In contrast, several groups found no difference in the inflammatory response between VV and CV. In animal models of ARDS, (18) severe bronchospasm (21) and prematurity, (27) the concentrations of IL-6, IL-8, and IL-10, and total protein content in the bronchoalveolar lavage were similar for both, variable and conventional controlled MV. There were no differences in lung injury in the lung tissues of an animal model of ARDS induced by oleic acid. (18) However, in ARDS induced by surfactant depletion, the variable controlled MV reduced alveolar damage, interstitial edema, hemorrhage, and epithelial dysfunction compared with CV. (22) VV improved lung function without causing structural damage to the lungs or increasing the inflammatory response in the experimental models and, in clinical settings, significantly reduced the systemic proinflammatory response compared with conventional controlled MV during the postoperative period of open abdominal surgery. (34) It is evident that even with the use of non-fixed TV and/or pressure during VV, these variables do not cause inflammatory and structural changes. Moreover, the beneficial effects observed with this method are due to this variability.
One study compared the respiratory variability in 10 normal subjects (following 1,587 breaths) with the variability randomly generated by a computer system to evaluate the variability rate related to TV and the impact of gas exchange and pulmonary mechanics. The results indicated that the nature of the chosen variability had no effect on pulmonary function. The authors concluded that the percentage of respiratory variability, but not the pattern of variability, were crucial to the success of VV. (47) The studies analyzed in this review suggest that VV is feasible and can be an effective ventilation strategy for improving lung function, particularly in injured lungs, considering that most of the preclinical studies used ARDS models. Clinical support for VV was presented in three clinical studies, (32)(33)(34) but these studies had limitations, including the lack of blinding of the investigator and health care staff, the short-term nature of the investigations, the absence of clinically relevant outcomes, and the small sample size. Furthermore, only two clinical studies provided data on hemodynamics (32,33) and sedation, (33,34) and the latter contained information on the type and prevalence of each sedative but no information on the need for sedatives or the doses used. These factors preclude the inclusion of these studies in clinical practice despite the good results found in the studies analyzed in this review.
Although preclinical studies suggest the benefits of VV in injured lungs with large collapsed and recruitable zones, there is no available data on the use of VV in patients with ARDS. Our group has investigated the role of PEEP variation in gas exchange in patients with mild or moderate ARDS (RBR-5bb65v).
Clinical studies of VV in other populations are underway. (48,49) In 2014, a study protocol was published for a randomized clinical trial (48) of patients who underwent open abdominal surgery lasting at least 3 hours. This study used a TV variation of 30%, considering an average volume of 6 mL/kg/predicted weight. The primary endpoint of the study was the forced vital capacity the first day after surgery. Secondary outcomes included new pulmonary function tests; plasma cytokine levels; spatial distribution of ventilation, assessed by electrical impedance tomography; and pulmonary complications in the postoperative period. Another multicenter controlled randomized clinical study evaluated variable PSV in patients with different pathologies in intensive care units to compare the length of weaning from MV using conventional PSV. (49) The results of these studies, which present a more appropriate design and evaluate more consistent outcomes, may provide further evidence supporting the possible inclusion of VV in clinical practice.

FINAL CONSIDERATIONS
Variable ventilation may be one of the most extensively investigated ventilation strategies in animal models of disease. Experimental studies have shown the beneficial effects of different variable ventilation strategies for improving lung function and reducing damage in mild to moderate lung injury in the short term. Variable ventilation seems to be a viable strategy for improving gas exchange and respiratory mechanics and preventing lung injury associated with mechanical ventilation. However, little evidence is available from comparative clinical studies with appropriate designs, adequate numbers of patients, and relevant clinical outcomes. Therefore, further clinical studies that use variable ventilation are necessary to assess the potential of variable ventilation strategies for improving the clinical outcomes of patients undergoing mechanical ventilation.