Biologic Impact of Mechanical Power at High and Low Tidal Volumes in Experimental Mild Acute Respiratory Distress Syndrome

MECHANICAL ventilation potentially worsens or even initiates lung injury.^1,2  High distending pressures (barotrauma), as well as repetitive collapse and reopening of lung units (atelectrauma), may trigger an inflammatory response (biotrauma) that can result in ventilator-induced lung injury,³  organ dysfunction,^4,5  and death.⁶ 

The role of low tidal volume (V_T) in lung protection has been well established in both experimental⁷  and clinical studies.⁸  In addition, the importance of choosing adequate levels of positive end-expiratory pressure (PEEP) has been well documented, especially in patients with severe acute respiratory distress syndrome (ARDS).^2,9  More recently, the importance of limiting the so-called driving pressure (ΔP = inspiratory plateau pressure—PEEP) has been emphasized even when combinations of low V_T and adequate PEEP are used.¹⁰  Furthermore, experimental studies have shown that the respiratory rate (RR),¹¹  as well as the inspiratory^12,13  and expiratory¹⁴  airflow, modulate the severity of lung injury. When combined, all of these variables determine the mechanical energy, or power, that is transferred from the mechanical ventilator to the lungs.^15 ,16

Mechanical power (energy over time) is associated with deterioration of lung function and formation of lung edema in pigs without previous lung injury and ventilated with increased V_T.¹⁶  While the contribution of RR to mechanical power and its impact on lung injury has been described,¹⁷  it is still unclear whether low and high V_T result in comparable lung injury when similar levels of mechanical power are achieved. Based on this aforementioned, we hypothesized that low V_T would minimize ventilator-induced lung injury regardless of the degree of mechanical power.

In the present study, we investigated the impact of mechanical power, as obtained by combinations of different V_T and respiratory rate (RR), on lung function and surrogates of ventilator-induced lung injury (histologic lung damage, inflammation, and mechanical alveolar stretch) in a rat model of mild ARDS.

This study was approved by the Ethics Committee of the Carlos Chagas Filho Institute of Biophysics, Health Sciences Center, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil. All animals received humane care in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research and the Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences, USA.

Forty Wistar rats (weight 408 ± 63g) were anesthetized by inhalation of sevoflurane 2% (Sevorane; Cristália, Brazil) and underwent intratracheal instillation of Escherichia coli lipopolysaccharide (O55:B5, LPS Ultrapure; Invivogen, France), 200 µg suspended in saline solution to a total volume of 200 μl,¹⁸  to induce mild-to-moderate ARDS. After 24 h, animals were premedicated intraperitoneally with 10 mg/kg diazepam (Compaz; Cristália, Brazil), followed by 100 mg/kg ketamine (Ketamin-S+; Cristália, Brazil) and 2 mg/kg midazolam (Dormicum; União Química, Brazil). After local anesthesia with 2% lidocaine (0.4 ml), a midline neck incision and tracheostomy were performed.

An intravenous (iv) catheter (Jelco 24G, Becton, Dickinson and Company, USA) was inserted into the tail vein, and anesthesia induced and maintained with midazolam (2 mg · kg^–1 · h^–1) and ketamine (50 mg · kg^–1 · h^–1). Additionally, 10 ml · kg^–1 · h^–1 Ringer’s lactate (B. Braun, Crissier, Switzerland) was administered iv; Gelafundin (B. Braun, São Gonçalo, Brazil) was administered (iv, in 0.5 ml boluses) to maintain mean arterial pressure (MAP) greater than 60 mmHg. A second catheter (18G; Arrow International, USA) was then placed in the right internal carotid artery for blood sampling and arterial blood gas analysis (ABL80 FLEX; Radiometer Medical, Denmark), as well as monitoring of MAP (Networked Multiparameter Veterinary Monitor LifeWindow 6,000 V; Digicare Animal Health, USA). A 30–cm-long water-filled catheter (PE-205, Becton, Dickinson and Company) with side holes at the tip, connected to a differential pressure transducer (UT-PL-400, SCIREQ, Canada), was used to measure the esophageal pressure. The catheter was passed into the stomach and then slowly returned into the esophagus; its proper positioning was assessed with the “occlusion test,” as described elsewhere.¹⁹  Heart rate (HR), MAP, and rectal temperature were continuously monitored (Networked Multiparameter Veterinary Monitor LifeWindow 6000V, Digicare Animal Health, USA). Body temperature was maintained at 37.5 ± 1°C using a heating bed.

Animals were paralyzed with pancuronium bromide (2 mg/kg, iv), and their lungs mechanically ventilated (Servo-i, MAQUET, Sweden) in volume-controlled mode (VCV) with constant inspiratory flow, V_T = 6 ml/kg, RR to maintain normocapnia (Paco₂ = 35 to 45 mmHg), PEEP = 3 cm H₂O, Fio₂ = 1.0, and inspiratory-expiratory ratio of 1:2 (fig. 1). After 5 min (BASELINE), arterial blood gases (300 µl) were determined (Radiometer ABL80 FLEX, Denmark). Fio₂ was reduced to 0.4 to prevent possible iatrogenic effects, and lung mechanics were assessed. Thirty-two rats were then randomly assigned to be mechanically ventilated with a combination of different V_T (6 ml/kg and 11 ml/kg) and RR settings that resulted in low and high mechanical power (N = 8/group): (1) low power/low V_T: V_T = 6 ml/kg and RR adjusted to maintain normocapnia; (2) low power/high V_T: V_T = 11 ml/kg and RR to maintain normocapnia; (3) high power/low V_T: V_T = 6 ml/kg and RR to increase mechanical power by threefold compared to the first two groups; and (4) high power/high V_T: V_T = 11 ml/kg and RR also adjusted to increase mechanical power by threefold compared to the first two groups. The “low” mechanical power levels were derived from a previous publication,⁷  whereas the “high” mechanical power levels were defined according to pilot studies (data not shown). The different levels of V_T corresponded approximately to the lower limits of protective and nonprotective ventilation (6 and 11 ml/kg, respectively). Settings were kept for 2 h. Arterial blood gases were determined after randomization (INITIAL) and at the end of the experiment (FINAL). Lung mechanics were assessed after randomization (INITIAL), at 60 min, and at the end of the experiment (FINAL). Eight rats were used for molecular biology analysis and were not mechanically ventilated (nonventilated). At BASELINE and FINAL, the endotracheal tube was clamped, and animals were connected to a flexiVent mechanical ventilator (SCIREQ); standardized pressure-volume curves at the same Fio₂ and PEEP as implemented by the manufacturer were then obtained. Heparin (1000 IU) was injected into the tail vein and animals were euthanized by overdose of sodium thiopental (60 mg/kg). The lungs were extracted at PEEP = 3 cm H₂O for histology and molecular biology analysis.

To rule out the potential effects of a mismatch in Paco₂ between the high and low mechanical power groups on diffuse alveolar damage score and the expression of biomarkers of inflammation, 10 additional animals were ventilated with high mechanical power and low as well as high V_T, but the dead space adapted to yield normocapnia (Supplemental Digital Content 1, https://links.lww.com/ALN/B640).

Airflow, airway pressure (Paw), and esophageal pressure were continuously recorded throughout the experiments with a computer running custom-made software written in LabVIEW (National Instruments, USA).^20,21  Inspiratory flow was measured from the recorded curve. Briefly, V_T was calculated by digital integration of the airflow signal obtained from a custom-made pneumotachograph²²  that was connected to the Y-piece of the ventilator tubing, while RR was calculated from the esophageal pressure swings as the frequency per minute of each type of breathing cycle. All signals were amplified in a four-channel signal conditioner (SC-24, SCIREQ), and sampled at 200 Hz with a 12-bit analog-to-digital converter (National Instruments, USA). Mechanical data were computed offline by a routine written in MATLAB (Version R2007a; The Mathworks Inc., USA).

Respiratory system plateau pressure was measured in occluding airways at end-inspiration during 5 s.²⁰  Respiratory system driving pressure was calculated as the difference between respiratory system plateau pressure and PEEP. Transpulmonary pressure was calculated as the difference between the pressure in the alveoli and the pressure in the pleural cavity (transpulmonary pressure at end-inspiration), and transpulmonary driving pressure as the difference between transpulmonary pressure at end-inspiration and at end-expiration. Intrinsic PEEP was measured by occluding the airway at end-expiration.²³ 

The mechanical energy was calculated based on: (1) the equation described by Guerin et al.²⁴  and Marini and Jaber²⁵  (simplified formula); (2) a quasi-static pressure-volume curve (flexiVent); and (3) the equation proposed by Gattinoni et al.,²⁶  based on the equation of motion.

The first approach to compute energy is presented in equation 1:

where ΔP,_L is the transpulmonary driving pressure and E,_L is the lung elastance.

In order to perform the quasi-static pressure-volume curve, a stepwise increase in airway pressure up to 30 cm H₂O was done. The area of the pressure-volume curve was computed (Supplemental Digital Content 2, https://links.lww.com/ALN/B641). The third method to compute mechanical energy was based on the formula presented in equation 2:

where ΔV is the variation of tidal volume, E,_RS is the respiratory system elastance, I:E is the inspiratory to expiratory ratio, and Raw is the airway resistance. All values were converted to mJ. Mechanical power was calculated as the product of mechanical energy multiplied by RR.

The lungs and heart were removed en bloc. The left lung was frozen in liquid nitrogen and immersed in formaldehyde solution (4%), embedded in paraffin, cut longitudinally in the central zone by means of a microtome into slices 4 μm thick, and stained with hematoxylin-eosin for histologic analysis.²⁶  Photomicrographs at magnifications of ×100, ×200, and ×400 were obtained from eight non-overlapping fields of view per section using a light microscope (Olympus BX51; Olympus Latin America Inc., Brazil). Diffuse alveolar damage was quantified by two investigators (V.M. and V.L.C.) blinded to group assignment and independently, using a weighted scoring system, as described elsewhere.²⁷  Briefly, scores of 0 to 4 were used to represent edema, atelectasis and overdistension, with 0 standing for no effect and 4 for maximum severity. Additionally, the extent of each scored characteristic per field of view was determined on a scale of 0 to 4, with 0 standing for no visible evidence and 4 for complete involvement. Scores were calculated as the product of severity and extent of each feature, on a range of 0 to 16. The cumulative diffuse alveolar damage score was calculated as the sum of each score, and thus ranged from 0 to 48.

To analyze the adherens junction protein E-cadherin, immunohistochemical procedures were performed on 4-μm-thick, paraffin-embedded lung sections using a mouse polyclonal antibody against E-cadherin (1:250; cat. number, 610181; BD Transduction Laboratories, Becton, Dickinson and Company).¹⁸  After dewaxing and rehydrating, endogenous peroxidase was blocked with 3% H₂O₂ in methanol for 15 min. Heat-mediated antigen retrieval and enzymatic techniques were performed according to the specific antibody. After blocking the nonspecific binding of immunoglobulins to the tissue, primary antibodies were then incubated overnight at 4°C in a humidified chamber for approximately 16 h. The sections were then washed in 0.25% Tween/phosphate-buffered saline solution for 5 min, and the secondary antibodies were incubated (Nichirei-Histofine Simple Stain Rat MAX-PO-Mouse; Nichirei Bioscience Inc., Germany). The chromogen substrate was diaminobenzidine (cat. number, K3468; Liquid DAB, Dako Denmark A/S, Denmark). Negative control slides were incubated with mouse isotype immunoglobulins or with antibody diluent solution. Visualization and image capture were performed using a light microscope (Eclipse E800; Nikon, Japan) coupled to a digital camera (Evolution; Media Cybernetics, USA) with the Q-Capture 2.95.0 graphic interface software (version 2.0.5; Quantitative Imaging, Canada). High-quality images (2,048 × 1,536 pixel buffer) were captured away from airways. After calibration of program settings, images were analyzed using the Image Pro Plus software (version 4.5.1; Media Cybernetics, USA).¹⁸ 

Three slices (2 × 2×2 mm) were cut from three different segments of the right lung and fixed for electron microscopy. On each electron micrograph (20 fields per animal), type II epithelial cell damage, basement membrane thickness, and endothelial cell damage were graded on a five-point, semiquantitative, severity-based scoring system as follows: 0 = normal lung parenchyma, 1 to 4 = changes in 1 to 25%, 26 to 50%, 51 to 75%, and 76 to 100% of examined tissue, respectively.^20,28  Electron microscopy analyses were performed by two investigators (V.M. and V.L.C.) blinded to group assignment.

Quantitative real-time reverse transcription polymerase chain reaction was performed to measure biomarkers associated with alveolar pulmonary stretch (amphiregulin), inflammation [interleukin (IL)-6], epithelial cell damage [club cell protein 16 (CC16)], and endothelial cell damage [intercellular adhesion molecular (ICAM)-1]. The primers used are shown in the Supplemental Digital Content 3 (https://links.lww.com/ALN/B642). Central slices of the right lung were cut, collected in cryotubes, flash-frozen by immersion in liquid nitrogen, and stored at −80°C. Total RNA was extracted from frozen tissues using the RNeasy Plus Mini Kit (Qiagen, Germany), following the manufacturer’s recommendations. RNA concentrations were measured by spectrophotometry in a Nanodrop ND-1000 system (ThermoScientific, USA). First-strand cDNA was synthesized from total RNA using a Quantitec reverse transcription kit (Qiagen). Messenger RNA (mRNA) levels were measured with a SYBR green detection system in an ABI 7500 real-time polymerase chain reaction system (Applied Biosystems, USA). Samples were run in triplicate. For each sample, the expression of each gene was normalized to the acidic ribosomal phosphoprotein P0 (36B4) housekeeping gene²⁹  and expressed as fold change relative to respective nonventilated animals, using the 2^–ΔΔCt method, where ΔCt = Ct_{(target gene)} – Ct_{(reference gene).}³⁰ 

Sample size was based on pilot studies and on our past experience with ventilator strategies in small animals.²⁰  We tested the hypothesis that low V_T would decrease IL-6 gene expression regardless of the degree of mechanical power. Accordingly, a sample size of eight animals per group (providing for one animal as dropout) would provide the appropriate power (1-β = 0.8) to identify significant (α = 0.05) differences in the symmetry score, taking into account an effect size d = 1.9, a two-sided test, and a sample size ratio = 1 (G*Power 3.1.9.2, University of Düsseldorf, Germany). Data were tested for normality using the Kolmogorov-Smirnov test with Lilliefors’ correction, while the Levene median test was used to evaluate homogeneity of variances. If both conditions were satisfied, Mauchly’s test of sphericity with repeated-measures ANOVA (P < 0.05) was used.³¹  If epsilon was higher than 0.75, the Huynh-Feldt P value was shown; otherwise, the Greenhouse-Geisser P value was provided. Additionally, to compare all parameters among groups at each time point, a mixed linear model based on a random intercept for each animal followed by Bonferroni’s test was used.

Molecular biology, diffuse alveolar damage, and electron microscopy variables were assessed with the Mann–Whitney U test and Bonferroni correction for 4 comparisons (P = 0.0125) (low power/low V_Tvs. low power/high V_T and high power/low V_T, as well as high power/low V_Tvs. high power/high V_T). Association and agreement between energy calculated from the pressure-volume curve (flexiVent) and mechanical power equation were determined by Pearson correlation and Bland-Altman analysis, respectively. Associations of diffuse alveolar damage score, IL-6, amphiregulin, and club cell protein 16 with energy and mechanical power were assessed with Spearman coefficients. Multiple linear regression analysis was performed to predict associations with the dependent variables (diffuse alveolar damage, IL-6, amphiregulin, and club cell protein 16) and independent variables (V_T, RR, and transpulmonary driving pressure). Parametric data were expressed as mean ± SD, while nonparametric data were expressed as median (interquartile range). The impact of hypocapnia on histologic damage and biologic markers was evaluated with the Student’s t test and Mann–Whitney U tests, as appropriate. The mixed linear models, Mauchly’s sphericity test, and multiple linear regression analyses were performed using IBM SPSS Statistics for Windows, Version 19.0 (IBM Corp., USA). All the other tests were performed in GraphPad Prism version 6.00 (GraphPad Software, USA). Significance was established at P < 0.05.

The survival rate was 100% in all groups.

A comparison of mechanical energy computed by equation 1 and according to the quasi-static volume-pressure curve (flexiVent) is presented as Supplemental Digital Content 4 (https://links.lww.com/ALN/B643). Even though the energy level calculated by different methods differed due to the final pressure achieved in each method, the variation of energy from FINAL to INITIAL was not different between the two methods (69% with flexiVent, 72% with the formula).

At BASELINE, no significant differences were observed in MAP and gas exchange variables among groups (Supplemental Digital Content 5, https://links.lww.com/ALN/B644). MAP did not differ over time or among groups (table 1). Cumulative fluid volumes administered were comparable among groups: low power/low V_T = 10.0 ± 1.7 ml, low power/high V_T = 13.1 ± 3.1 ml, high power/low V_T = 15.3 ± 5.2 ml, and high power/high V_T = 14.5 ± 6.8 ml. Pao₂/Fio₂ did not differ significantly either among groups or between INITIAL and FINAL. In the high mechanical power groups, Paco₂ was lower and arterial pH was higher compared to low mechanical power (table 1).

At BASELINE, no significant differences were observed in respiratory parameters among groups (Supplemental Digital Content 6, https://links.lww.com/ALN/B645). As depicted in table 2, high V_T was accompanied by lower RR in high mechanical power and low mechanical power groups alike. V’_E and flow increased in high mechanical power compared to low mechanical power groups, regardless of V_T. Transpulmonary plateau pressure, respiratory system plateau pressure, transpulmonary driving pressure, and respiratory system driving pressure were increased in high power/low V_T compared to low power/low V_T, as well as high power/high V_T compared to low power/high V_T. Additionally, in the high mechanical power groups, mechanical energy increased with high compared to low V_T. As desired, mechanical power was approximately three times higher in high mechanical power compared to low mechanical power groups. Intrinsic PEEP was higher in high power/low V_T compared to low power/low V_T, as well as in high power/high V_T compared to low power/high V_T and high power/low V_T (table 2).

Figure 2 depicts light microscopy images of representative animals. As shown in table 3, high V_T resulted in increased diffuse alveolar damage scores in both low and high mechanical power groups. Furthermore, despite low V_T, diffuse alveolar damage scores were increased in high versus low power groups. E-cadherin expression in lung tissue was lower in low power/high V_T compared to low power/low V_T group (table 3, Supplemental Digital Content 7, https://links.lww.com/ALN/B646).

Electron microscopy images of one representative animal per group are shown in figure 3. In both low and high mechanical power groups, high V_T increased type II epithelial cell damage and basement membrane thickness compared to low V_T (table 4). In the high mechanical power group, high V_T also led to greater endothelial cell damage.

Levels of IL-6, amphiregulin, and club cell protein 16 mRNA expressions, but not ICAM-1 expression, increased in all groups compared to nonventilated, as shown in figure 4. However, IL-6 and amphiregulin mRNA expressions were higher in high power/high V_T than in low power/high V_T. Moreover, high power/high V_T animals exhibited increased amphiregulin and club cell protein 16 expression compared to high power/low V_T animals.

Correlation analyses revealed significant associations of diffuse alveolar damage score, IL-6, amphiregulin, and club cell protein 16 mRNA expressions with mechanical energy (calculated by a simplified formula) and power (fig. 5). Similar associations of diffuse alveolar damage score, IL-6, amphiregulin, and club cell protein 16 mRNA expressions with mechanical energy computed by flexiVent and the formula based on the equation of motion (Supplemental Digital Content 8, https://links.lww.com/ALN/B647) were observed.

On multiple linear regression analyses, variances in diffuse alveolar damage score and E-cadherin, Il-6 and amphiregulin gene expressions were better predicted by V_T (Supplemental Digital Content 9, https://links.lww.com/ALN/B648).

Mean arterial pressure (Supplemental Digital Content 10, https://links.lww.com/ALN/B649), gas exchange (Supplemental Digital Content 11, https://links.lww.com/ALN/B650), and diffuse alveolar damage scores (Supplemental Digital Content 12, https://links.lww.com/ALN/B651; and Supplemental Digital Content 13, https://links.lww.com/ALN/B652) in experiments performed with matched Paco₂ at high mechanical power are reported as Supplemental Digital Content. Diffuse alveolar damage score variables (edema, atelectasis, and overdistension) and cumulative diffuse alveolar damage score did not differ between high-power groups with low levels of Paco₂ and high-power groups with normal Paco₂. On comparison of molecular biology analysis of high-power groups at low Paco₂ with those which presented similar levels of Paco₂ at low mechanical power, expression of IL-6, amphiregulin, club cell protein 16, and ICAM-1 behaved as in high- power groups at low Paco₂ levels (Supplemental Digital Content 14, https://links.lww.com/ALN/B653).

In the model of mild ARDS used herein, we found that, at low mechanical power, low V_T reduced diffuse alveolar damage score, without changing expression of biomarkers associated with inflammation, alveolar pulmonary stretch, or epithelial cell damage. At high mechanical power, high V_T increased diffuse alveolar damage, promoted ultrastructural impairment in alveolar epithelial and endothelial cells and alveolar–capillary membrane, as well as loss of cell-cell adhesion. Diffuse alveolar damage score and expression of IL-6, amphiregulin, and club cell protein 16 were associated with energy and power, and these findings were independent of Paco₂ levels.

To the best of our knowledge, this was the first experimental study to investigate the impact of low and high mechanical power on ventilator-induced lung injury. Different methods have been proposed to calculate mechanical power. We chose a simplified equation in order to facilitate its routine use in the clinical setting.²⁹  This equation computes the most important component (driving mechanical power), without taking into account resistive properties and PEEP, unlike the equation proposed by Gattinoni et al.¹⁵  The resistive properties depend on flow magnitude and profile, and may contribute to lung damage. However, it is difficult to directly link the mechanical power dissipated in the proximal airways to alveolar injury. Additionally, the contribution of PEEP to mechanical power was comparable among groups. We calculated mechanical energy using a quasi-static pressure-volume curve and observed that, even though the degree of energy differs between the two methods, due to the maximum pressure achieved in the pressure-volume curve (30 cm H₂O), the variation from INITIAL to FINAL was similar. We also calculated mechanical power based on the aforementioned Gattinoni equation, which includes resistive properties and PEEP. Regardless of the method used to calculate energy, its correlation with diffuse alveolar damage, IL-6, club cell protein 16, and amphiregulin did not differ. Cressoni et al. calculated energy as the area between the inspiratory limb of the delta-transpulmonary pressure–volume curve. Direct comparison between our data and those of Cressoni et al. was not feasible due to differences in animal size (piglets vs. rats), use of healthy versus ARDS animals, positioning (supine vs. prone-Trendelenburg position), duration of mechanical ventilation (2 vs. 54 h), and the combination of V_T and RR. In our study, not only RR but also V_T was modulated to ensure tight control of the two levels of mechanical power and maintained within a safe range to keep animals hemodynamically stable and alive during the experiments. Even the highest V_T values used were within ranges reported in clinical practice.³²  Additional groups were subjected to ventilation with high power and low as well as high V_T, but the dead space was adapted to yield normocapnia, in order to exclude the potential effects of a mismatch in Paco₂ between high and low power groups on diffuse alveolar damage score and expression of biomarkers.

It is worth noting that, despite an approximately threefold increase in mechanical power, transpulmonary plateau pressure, transpulmonary driving pressure, and respiratory system driving pressure differed between low V_T, but not high V_T groups. This finding might be explained by the fact that, in high power groups, RR was increased substantially under low V_T, resulting in a small level of intrinsic PEEP.

Both RR¹⁷  and inspiratory flow¹³  have been recognized as potential determinants of ventilator-induced lung injury. We observed that, at the same power (low or high), V_T increased and RR reduced but inspiratory flow did not differ significantly between low and high V_T. Inspiratory flow was higher at high compared to low power, suggesting the participation of airflow as a factor promoting ventilator-induced lung injury, since lung damage was observed even at low V_T. However, the damage induced by high V_T at high power was greater than that induced by low V_T. Thus, at high power, the role of airflow as a determinant of ventilator-induced lung injury was minor compared to that of V_T. Even though the high RR observed at low V_T and high power resulted in intrinsic PEEP, it was not high enough to affect transpulmonary plateau pressure, transpulmonary driving pressure.

Diffuse alveolar damage score was lower in low than high V_T, regardless of the power level. In fact, at low V_T, energy did not differ, while RR increased in the high, compared to low, power group. In the first proof-of-concept study¹⁶  investigating the association between mechanical power and ventilator-induced lung injury, lung edema increased while oxygenation decreased proportionally to RR. In contrast, in our study, the histologic damage resulting from high V_T was greater than that associated with low V_T at comparable levels of low and high power, suggesting that the proportional role of V_T in promoting lung damage is greater than predicted by the mechanical power model. In line with these results, E-cadherin expression was further reduced with high V_T compared to low V_T at low power, suggesting increased mechanical stress/strain at the alveolar epithelial layer, leading to loss of cell–cell adhesion.^18,33  High V_T might also have exceeded the plasticity limits of the lungs, resulting in disproportionally high parenchymal strain.^34,35  In rats, respiratory system plateau pressure exceeding 14 cm H₂O can trigger expression of biomarkers of inflammation,²⁰  which supports this hypothesis. We were also unable to rule out the possibility that inhomogeneity of injury might have amplified the effects of V_T on lung damage. Intratracheal instillation of LPS results in heterogeneous distribution of diffuse alveolar damage features, as also reflected by altered viscoelastic properties.³⁶  In a physical model of ARDS, inhomogeneities in lung structure can increase the local forces generated during lung expansion, amplifying regional stress.³⁷ 

Oxygenation did not differ among groups at FINAL. One might expect worse oxygenation in the presence of high V_T and high power due to greater lung damage. However, oxygenation depends not only on the pulmonary structure, but also on regional ventilation and perfusion, including effects on pulmonary vascular resistance and venous return yielding differences in and dead space.³⁸  We hypothesize that, in our study, oxygenation probably did not decline in the presence of high V_T and high power due to increased dead space.

Although all ventilated groups showed increased gene expression of biomarkers of inflammation, alveolar stretch, and alveolar epithelial and endothelial cell damage as compared to nonventilated, the combination of high power with high V_T led to more pronounced responses. As observed for lung damage, the contribution of high V_T to triggering the pro-inflammatory response was higher than predicted by mechanical power. Moreover, the mechanical energy and power showed association with variables of ventilator-induced lung injury. As mechanical energy, unlike mechanical power, is not determined by RR, the intratidal phenomenon of extracellular matrix deformation and anchored endothelial and epithelial cells seem to play a central role in ventilator-induced lung injury. Multiple linear regression analyses revealed that V_T predicted most of the impact of mechanical ventilation on diffuse alveolar damage score, E-cadherin, and gene expressions of amphiregulin and IL-6.

Our findings reinforce the concept that mechanical power should be measured in ARDS patients under protective mechanical ventilation. However, the hypothesis that mechanical power is the main determinant of ventilator-induced lung injury could inadvertently lead to acceptance of higher V_T, even when RR and/or transpulmonary driving pressure are within a “safe” range. Although this range has yet to be defined clearly, our data suggest that mechanical ventilation at low power may promote lung injury in the presence of inappropriately high V_T. Therefore, despite use of low RR and transpulmonary driving pressure, V_T should always be kept in the low protective range. This claim certainly remains to be confirmed in clinical studies.

Several limitations of the present study must be acknowledged. First, experimental ARDS was induced by endotoxin, and our results cannot be extrapolated to other models or to human ARDS. Second, the observation time was relatively short, precluding extrapolation of our findings to longer periods of ventilation. To keep small animals with ARDS alive for 6 h would require larger fluid volumes, perhaps vasopressors to maintain MAP greater than 70 mmHg, and bicarbonate to counteract intense metabolic acidosis. Therefore, as a primary study design, even though a 2-h duration represents a short period of observation, it allowed us to evaluate the gene activation induced by different V_T and power without any interference from therapies that would be needed to keep the animals alive in a longer experiment. Third, PEEP was constant during the experiments and we cannot exclude different results with different PEEP levels. Fourth, Paco₂ and arterial pH were not matched between low and high-power groups. There is a controversy on the potential harmful³⁹  or beneficial^40,41  effects of these variables in terms of lung protection. Nevertheless, in additional experiments performed with high power, Paco₂ and arterial pH were kept in normal ranges, but no differences were observed in lung damage nor in gene expression of biomarkers of inflammation, alveolar mechanical stretch, epithelial and endothelial cell damage.

In this model of mild ARDS, high mechanical power was associated with ventilator-induced lung injury, even at low V_T. The impact of high V_T on ventilator-induced lung injury was greater than predicted by mechanical power. To minimize ventilator-induced lung injury, low V_T should be combined with low mechanical power.

The authors express our gratitude to Andre Benedito da Silva, B.Sc. (Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil), for animal care; Arlete Fernandes, B.Sc. (Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil), for her help with microscopy; Moira Elizabeth Schottler, B.A., and Filippe Vasconcellos, B.A., for their assistance in editing the manuscript, and Ronir Raggio Luiz, Ph.D. (Institute of Public Health Studies, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil), for his help with statistics.

This study is supported by the Brazilian Council for Scientific and Technological Development (CNPq; Brasília, Brazil), the Rio de Janeiro State Research Foundation (FAPERJ; Rio de Janeiro, Brazil), the São Paulo State Research Foundation (FAPESP; São Paolo, Brazil), the National Institute of Science and Technology for Regenerative Medicine (INCT-REGENERA; Rio de Janeiro, Brazil), and the Coordination for the Improvement of Higher Education Personnel (CAPES; Brasília, Brazil).

The authors declare no competing interests.