Lung Injury Is Induced by Abrupt Increase in Respiratory Rate but Prevented by Recruitment Maneuver in Mild Acute Respiratory Distress Syndrome in Rats

This study was approved by the Ethics Committee of the Health Sciences Center, Federal University of Rio de Janeiro (CEUA: 015/19). All animals received care in compliance with the “Principles of Laboratory Animal Care” formulated by the National Society for Medical Research (Bethesda, Maryland) and the U.S. National Academy of Sciences (Bethesda, Maryland) Guide for the Care and Use of Laboratory Animals. The current study followed the Animal Research: Reporting of In Vivo Experiments guidelines for reporting of animal research.¹²  Animals were housed at a controlled temperature (23°C) and in a controlled light–dark cycle (12–12 h), with free access to water and food.

In the early morning (8:00 am), 42 consecutive male Wistar rats (body weight, 352 ± 16 g, 10 to 12 weeks old) were anesthetized with isoflurane (1.5 to 2.0%; Isoforine, Cristália, Brazil). Animals were breathing spontaneously and then Escherichia coli lipopolysaccharide (Merck Millipore, Burlington, Massachusetts, 9.6 × 10⁶ endotoxin units/ml suspended in 0.9% saline solution [total volume 200 μl]) was instilled intratracheally to induce mild ARDS.¹  Animals were then allowed to recover from anesthesia and monitored for 24 h. Twenty-four hours after endotoxin instillation (at 8:00 am), animals were premedicated intraperitoneally with 10 mg/kg diazepam (Compaz, Cristália, Itapira, São Paulo, Brazil), followed by 100 mg/kg ketamine (Ketamin-S, Cristália) and 2 mg/kg midazolam (Dormicum, União Química, São Paulo, Brazil). After local anesthesia with 1% lidocaine (0.4 ml), an intravenous catheter (Jelco 24G, Becton, Dickinson and Company, Franklin Lakes, New Jersey) was inserted into the tail vein, and anesthesia was maintained with midazolam (2 mg · kg^–1 · h^–1) and ketamine (50 mg · kg^–1 · h^–1). A midline neck incision and tracheostomy were performed, and the trachea was cannulated with a polyethylene tube (inner diameter 1.8 mm). An 18-gauge catheter (Arrow International, USA) was then placed in the right internal carotid artery to enable blood gas analysis (Radiometer ABL80 FLEX, Copenhagen, Denmark) and continuous monitoring of mean arterial pressure as a surrogate marker of hemodynamic status. Temperature was measured with a rectal catheter (Networked Multiparameter Veterinary Monitor LifeWindow 6000V; Digicare Animal Health, USA). Body temperature was maintained at 37°C ± 1°C using a heating bed. Once anesthetized, animals were placed in dorsal recumbency, paralyzed with pancuronium bromide (2 mg/kg IV) and their lungs were mechanically ventilated (V500; Dräger Medical, Lübeck, Germany) for 5 min (BASELINE), in pressure-controlled ventilation mode. During that period, airway pressure was adjusted to reach a V_T of 6 ml/kg; positive end-expiratory pressure (PEEP), 3 cm H₂O; RR, 70 breaths/min; and inspired fraction of oxygen (Fio₂), 0.4. After arterial blood gases were evaluated, each animal was randomly assigned, by sealed envelopes, to one of six groups (n = 7 animals per group): (1) RR, 70 breaths/min for 2 h (control); (2) RR, 70 breaths/min during the first hour followed by an abrupt increase in RR to 130 breaths/min, keeping that higher RR constant during the second hour (abrupt increase of RR for 1 h); (3) RR, 130 breaths/min during both first and second hours (abrupt increase of RR for 2 h); (4) RR, 70 breaths/min during the first 30 min, increasing gradually from 70 breaths/min to 130 breaths/min (increments of 2 breaths/min) until minute 60, followed by 130 breaths/min during the second hour (shorter RR adaptation); (5) RR, 70 breaths/min, immediately followed by a gradual increase from 70 breaths/min to 130 breaths/min until minute 60 (1 breath increment per minute), followed by constant 130 breaths/min during the second hour (longer RR adaptation) (fig. 1). An RR of 130 breaths/min was chosen based on a previous study,¹³  which reported signs of ventilator-induced lung injury while allowing animals to survive 2 h without severe hemodynamic impairment. The gradual increases in RR were made manually by an independent researcher (P.H.X.). To avoid V_T reduction and to keep it in a protective range, the delta pressure in pressure-controlled ventilation was constantly manually adjusted by a dedicated researcher (P.H.X.). Seven nonventilated animals were used for comparisons of molecular biology analyses. Respiratory mechanics were analyzed every 30 min, while arterial blood gases were analyzed at INITIAL and FINAL. Increased RR may raise the partial pressure of carbon dioxide (Paco₂); however, to keep Paco₂ within a range from 30 to 45 mmHg, circuit dead space (polyethylene tubing) between the ventilatory circuit and pneumotachograph was added and a new arterial blood gas analysis performed (Supplemental Digital Content 1, fig. 1, https://links.lww.com/ALN/D8). Thus, the distance between the tracheal tube and pneumotachograph was constant, which minimizes the compressible volume at high RR. At the end of 2 h, heparin was injected intravenously (1,000 IU), animals were euthanized by IV overdose of sodium thiopental (60 mg/kg, Cristália), and their lungs were extracted at PEEP, 3 cm H₂O, for histologic and molecular biologic analyses.

In response to peer review, two additional control groups were included (n = 7/each): (1) control 2, abrupt increase of RR maintained for 1 h; and (2) control 3, in which the control-2 protocol was performed after recruitment maneuver (continuous positive airway pressure, 30 cm H₂O for 30 s). We hypothesized that the application of the recruitment maneuver could open alveolar units otherwise collapsed at end-expiration, thereby reducing inhomogeneity among so-called slow and fast alveolar units.

A pneumotachograph (internal diameter, 1.5 mm; length, 4.2 cm; distance between side ports, 2.1 cm) was connected to the tracheal cannula for airflow (V′) measurements.¹⁴  The pressure gradient across the pneumotachograph was determined using a SCIREQ differential pressure transducer (UT-PDP-02; SCIREQ, Montreal, Quebec, Canada). Airflow and airway pressures were recorded continuously throughout the experiments with a computer running custom-made software written in LabVIEW (National Instruments, USA).¹⁴  In brief, V_T was calculated by digital integration of the airflow signal obtained from a custom-made pneumotachograph that was connected to the Y-piece of the ventilator tubing.

Both inspiratory and expiratory pauses for 5 s were done to accurately measure the respiratory system peak pressure, respiratory system plateau pressure, the respiratory system driving pressure, intrinsic PEEP, and extrinsic PEEP. The respiratory system plateau pressure was computed after a 5-s end-inspiratory pause according to the time point of evaluation. Extrinsic PEEP was measured at the pneumotachograph. Inspiratory airflow, RR, and V_T were also measured. Mechanical power was also calculated by the following equation:

Cumulative power exposure was calculated in the 1-h and 2-h groups.¹  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). Mechanical data were computed offline by a routine written in MATLAB (version R2007a; MathWorks, USA).¹ 

Lungs were extracted during an expiratory pause at PEEP, 3 cm H₂O, and slices (4 μm thick) were cut and stained with hematoxylin and eosin. The diffuse alveolar damage score was quantified.^15,16  Photomicrographs at magnifications of ×100, ×200, and ×400 were obtained from 10 nonoverlapping fields of view per section using a light microscope (Olympus BX51, Olympus Latin America, São Paulo, Brazil). Diffuse alveolar damage was quantified using a weighted scoring system by two investigators (A.C.F.F. and V.L.C.) blinded to group assignment and independently, as described elsewhere.¹⁷  The scores of each expert were combined to yield a final score by arithmetic averaging. In brief, scores of 0 to 4 were used to represent ductal overdistension, alveolar collapse, inflammation, and edema or hemorrhage, with 0 standing for no effect and 4 for maximum severity. In addition, the extent of each scored characteristic per field of view was determined on a scale of 0 to 4, with 0 denoting no visible evidence and 4 denoting widespread involvement. Scores were calculated as the product of the 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 characteristic and ranged from 0 to 64.^1,16 

Airspace enlargement was assessed by measuring the mean linear intercept between alveolar walls at a magnification of ×400, as described elsewhere.¹⁸  To characterize the heterogeneity of airspace enlargement, the central moment of the mean linear intercept (D₂ of mean linear intercept between alveolar walls) was computed from 20 airspace measurements¹⁹  according to equation (2)

where μ is the mean, σ is the variance of airspace diameters, and γ is the skewness of the diameter distribution. After D₂ calculation, the heterogeneity index (β) was derived from D₂ and mean linear intercept between alveolar wall values by their ratio.²⁰  Quantification of heterogeneous airspace enlargement was performed by one expert investigator (A.C.F.F.) blinded to group assignment.

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 (Catalog No. 610181; BD Transduction Laboratories, 1:250, USA). 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 graphical interface (version 2.0 software. 5, Quantitative Imaging, Surrey, British Columbia, Canada). High-quality images (2048 × 1536 pixel buffer) were captured far from the airways and analyzed using ImagePro Plus software (version 4.5.1, Media Cybernetics). Quantification of immunohistochemistry was performed by two expert investigators (A.C.F.F. and L.A.) blinded to group assignment.

Gene expression of biomarkers associated with inflammation (interleukin-6 [IL-6]),²¹  epithelial cell damage (club cell secretory protein [CC-16]),²²  and endothelial cell damage (vascular cell adhesion molecule [VCAM-1])²³  in lung tissue was measured by reverse transcriptase-polymerase chain reactions. Primer sequences are listed in Supplemental Digital Content 2, table 1 (https://links.lww.com/ALN/D9). 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 ReliaPrep RNA Tissue Miniprep System (Promega Corporation, USA) according to the manufacturer’s recommendations. RNA concentrations were measured by spectrophotometry in a Nanodrop ND-1000 system (Thermo Scientific, USA). First-strand cDNA was synthesized from total RNA using a Quantitec reverse transcription kit (QIAGEN, Hilden, Germany). Relative mRNA levels were measured with a SYBR green detection system in a Mastercycler ep realplex real-time polymerase chain reaction system (Eppendorf, Applied Biosystems, USA). Samples were run in triplicate. Expression of each gene was normalized to the housekeeping gene acidic ribosomal phosphoprotein P0 (36B4)²⁴  and expressed as fold change relative to nonventilated animals, using the 2^−∆∆Ct method, where ΔCt = Ct (target gene) − Ct (reference gene).²⁵  Molecular biology data were analyzed by one investigator (M.M.) blinded to group assignment.

The number of animals was calculated according to a previous study.¹³  Taking into account that differences in diffuse alveolar damage score between low power or low V_Tversus high power or low V_T achieved an effect size of d = 1.68, assuming a sample size ratio of 1 and a statistical power (1 − β = 0.8) to identify significant differences (α = 0.05), 7 animals per group were necessary. The sample size calculation was done in G*Power software (G* Power 3.1.9.2, University of Düsseldorf, Germany).

The primary outcome was the diffuse alveolar damage score, and secondary outcomes were heterogeneous airspace enlargement, immunohistochemistry, and molecular biology. To compare functional parameters over time, a fixed and mixed linear model based on random intercept was used for each animal, followed by Holm-Šidák comparison tests. For the linear mixed model, we used the terms group, time, and group × time interaction. To compare diffuse alveolar damage score, heterogeneous airspace enlargement, immunohistochemistry, and molecular biology, the Kruskal-Wallis test was performed, followed by Dunn’s comparison tests. Spearman correlation was done between cumulative mechanical power and cumulative diffuse alveolar damage score. Data that satisfied parametric assumptions were expressed as the mean and SD, and data that did not satisfy parametric assumptions as the median (interquartile range). Statistical significance was established at P < 0.05 (two-tailed). Statistical analysis was performed using GraphPad Prism for Windows (version 8.1.1, GraphPad, USA).

All animals survived to the end of the experiment (FINAL); thus, there were no missing data. Cumulative fluid administration was similar among groups (control, 11.4 ± 0.6 ml; abrupt increase of RR for 1 h, 11.7 ± 0.5 ml; abrupt increase of RR for 2 h, 11.8 ± 0.6 ml; shorter RR adaptation, 11.8 ± 0.9 ml, longer RR adaptation, 11.9 ± 0.7 ml). Oxygenation improved through time (Supplemental Digital Content 3, table 2, https://links.lww.com/ALN/D10). No differences in Paco₂, arterial pH blood, or bicarbonate levels were observed among the groups. Mean arterial pressure remained higher than 90 mmHg, with no significant differences among groups throughout the experiments (Supplemental Digital Content 3, table 2, https://links.lww.com/ALN/D10).

Table 1 depicts the respiratory parameters. V_T did not differ among groups through time. At FINAL, respiratory system peak pressure was higher in both abrupt increase of RR groups and shorter RR adaptation, but not in longer RR adaptation, compared with control (abrupt increase of RR for 1 h vs. control, mean difference 3.9 [95% CI 0.8 to 6.9]; abrupt increase of RR for 2 h vs. control, mean difference 4.8 [95% CI 1.7 to 7.9]; shorter RR adaptation vs. control, mean difference 3.7 [95% CI 0.5 to 6.9]). Peak inspiratory flow was higher in the abrupt increase of RR groups, but not in the adaptation groups, compared with control (abrupt increase of RR for 1 h vs. control, mean difference 10.3 [95% CI 3.1 to 17.5]; abrupt increase of RR for 2 h vs. control, mean difference 10.9 [95% CI 3.7 to 18.0]). The respiratory system plateau pressure and respiratory system driving pressure were higher in the abrupt increase of RR for 2 h and shorter RR adaptation groups, but not in the abrupt increase of RR for 1 h and longer RR adaptation groups compared with the control group (respiratory system plateau pressure: abrupt increase of RR for 2 h vs. control, mean difference 3.6 [95% CI 0.9 to 6.4]; shorter RR adaptation vs. control, mean difference 2.9 [95% CI 0.1 to 5.6]; respiratory system driving pressure: abrupt increase of RR for 2 h vs. control, mean difference 3.8 [95% CI 1.1 to 6.5]; shorter RR adaptation vs. control, mean difference 3.0 [95% CI 0.2 to 5.8]). Mechanical power was lower in the longer RR adaptation group than both abrupt increase of RR groups (longer RR adaptation vs. abrupt increase of RR for 1 h, mean difference −47.1 [95% CI −93.9 to −0.2]; longer RR adaptation vs. abrupt increase of RR for 2 h, mean difference −60.7 [95% CI −107.6 to −13.9]). Both abrupt and adaptation groups showed increased cumulative mechanical power compared with the control group (abrupt increase of RR for 1 h vs. control, mean difference 6,487 [95% CI 3,735 to 9,239]; abrupt increase of RR for 2 h vs. control, mean difference 15,776 [95% CI 13,131 to 18,420]; shorter RR adaptation vs. control, mean difference 8,956 [95% CI 6,204 to 11,708]; longer RR adaptation vs. control, mean difference 10,335 [95% CI 7,691 to 12,979]). Abrupt increase of RR for 2 h and longer RR adaptation were associated with increased cumulative mechanical power compared with abrupt increase of RR for 1 h (abrupt increase of RR for 2 h vs. abrupt increase of RR for 1 h group, mean difference 9,288 [95% CI 6,536 to 12,040]; longer RR adaptation vs. abrupt increase of RR for 1 h group, mean difference 3,848 [95% CI 1,096 to 6,600]). Both the shorter and longer adaptation groups showed reduced cumulative mechanical power compared with abrupt increase of RR or 2 h (shorter RR adaptation vs. abrupt increase of RR or 2 h, mean difference –6,820 [95% CI –9,572 to –4,068]; longer RR adaptation vs. abrupt increase of RR or 2 h, mean difference –5,441 [95% CI –8,085 to –2,796]). There was a positive correlation between cumulative mechanical power and cumulative diffuse alveolar damage score (r = 0.62, P < 0.001) (Supplemental Digital Content 4, fig. 2, https://links.lww.com/ALN/D11).

Both abrupt increase of RR groups presented higher cumulative diffuse alveolar damage scores than the control group, due to increases in ductal overdistension, alveolar collapse, inflammation, and edema/hemorrhage (table 2, fig. 2). The cumulative diffuse alveolar damage score was reduced in the shorter and longer RR adaptation groups in comparison with both abrupt increases of RR groups. In addition, alveolar collapse and inflammation were lower in the shorter RR adaptation group than both abrupt increases of RR groups. Edema or hemorrhage was lower in the longer RR adaptation group than in both abrupt increases of RR groups (table 2, fig. 2). The diffuse alveolar damage score was lower in both adaptation groups (shorter and longer) compared with abrupt increase of RR for 1 h during 1 h (Supplemental Digital Content 5, table 3, https://links.lww.com/ALN/D12, Supplemental Digital Content 6, fig. 3, https://links.lww.com/ALN/D13). In addition, E-cadherin was fragmented, which denotes loss of epithelial integrity, in the abrupt increase of RR for 1 h during 1 h group (Supplemental Digital Content 7, fig. 4, https://links.lww.com/ALN/D14).

The Supplemental Digital Content 8, fig. 5 (https://links.lww.com/ALN/D15), shows data on cumulative mechanical power in the control, abrupt increase of RR during 1 h, and longer and shorter adaptation groups. The cumulative power of abrupt increase of RR during 1 h was similar to that imparted to the control group ventilated for 2 h (Supplemental Digital Content 8, fig. 5, https://links.lww.com/ALN/D15 and Supplemental Digital Content 9, table 4, https://links.lww.com/ALN/D16). However, the diffuse alveolar damage score was higher in the abrupt increase of RR for 1 h group than in the control groups (Supplemental Digital Content 5, table 3, https://links.lww.com/ALN/D12). Furthermore, the cumulative power of longer and shorter RR adaptation groups ventilated for 2 h was higher than in the abrupt increase of RR for 1 h group. However, the diffuse alveolar damage score was lower in longer and shorter RR adaptation groups ventilated for 2 h than in animals exposed to abrupt increase of RR for 1 h.

The heterogeneity index (β) of airspace enlargement was higher in the abrupt increase of RR for 2 h group compared with the control group (P = 0.010) (table 2). The lower RR adaptation group showed lower β compared with both abrupt increase of RR groups (1 h and 2 h; P = 0.048 and P = 0.005, respectively) (table 2). Both abrupt increase of RR groups (1 h and 2 h) showed reduced alveolar integrity, as measured by lung tissue E-cadherin expression, compared with the control group (P = 0.026 and P = 0.017, respectively) (fig. 3). Similarly, the lower RR adaptation group showed better alveolar integrity compared with both abrupt increase of RR groups (1 h and 2 h; P = 0.041 and P = 0.026, respectively) (fig. 3).

Abrupt increase of RR for 2 h was associated with a higher level of IL-6 (marker of inflammation) than in the control group (P = 0.038). IL-6 was reduced in the longer RR adaptation group compared with the abrupt increase of RR for 2 h group (P = 0.045). CC-16 (a marker of alveolar epithelial cell damage) and VCAM-1 (a marker of endothelial cell damage) were higher in both abrupt increases of RR groups (1 h [P = 0.025 and P = 0.017, respectively] and 2 h [P < 0.001 and P < 0.001, respectively]) compared with the control group. In addition, lung tissues from animals in the shorter and longer RR adaptation groups showed reduced VCAM-1 gene expression compared with the abrupt increase of RR for 2 h group (P = 0.013 and P = 0.045, respectively) (fig. 4).

VCAM-1 gene expression was lower in animals subjected to recruitment maneuver before abrupt increase of RR (control 3) for 1 h than in the nonventilated group in and animals not subjected to recruitment maneuver before the abrupt increase in RR (control 2, P = 0.040 and P = 0.016, respectively). CC-16 gene expression increased in animals not subjected to recruitment maneuver before the abrupt increase in RR (P = 0.014), but not in those animals subjected to recruitment maneuver, compared with the nonventilated group (fig. 5).

In mild experimental ARDS, we found that (1) the cumulative diffuse alveolar damage score decreased in both the shorter and longer adaptation groups compared with the abrupt increase of RR for 1 h and 2 h groups; (2) mechanical power was lower in the longer adaptation group than in the abrupt increase of RR for 1 h and 2 h groups; (3) the longer RR adaptation group showed less lung heterogeneity compared with the abrupt increase of RR for 1 h and 2 h groups; (4) alveolar integrity, measured by the amount of E-cadherin expression in lung tissue, was better preserved in the longer RR adaptation group compared with both abrupt increase of RR groups; (5) markers of lung inflammation (IL-6) and of epithelial and endothelial cell damage (CC-16 and VCAM-1) were higher in the abrupt increase of RR groups than in both RR adaptation strategies compared with the control group; (6) recruitment maneuver before the abrupt increase of RR prevented the rise in VCAM-1 and CC-16 gene expressions. Compared with an abrupt increase to a given RR, more gradual increments of RR to the same final level and time predictably decreases mechanical power, attenuating ventilator-induced lung injury, as measured by lung histology and molecular markers of damage. Cumulative power in the abrupt increase of RR for 1 h group was similar to that observed in the control 1 group, but diffuse alveolar damage score was higher in animals exposed to abrupt increase of RR for 1 h than in controls (Supplemental Digital Content 4, table 3, https://links.lww.com/ALN/D11). This result suggests an injurious effect of abruptly increasing RR for 1 h. Furthermore, the cumulative power of longer and shorter RR adaptation groups ventilated for 2 h was higher than in animals exposed to abrupt increase of RR for 1 h, but the diffuse alveolar damage score was lower in the former than in the latter, suggesting a protective effect of gradually increasing RR, even if more mechanical power accumulates over time; in other words, high levels of RR applied in an abrupt manner are associated with more lung-tissue injury.

We used a model of acute lung injury induced by intratracheal instillation of E. coli endotoxin,^3,13,26,27  a well-established “first hit” for lung inflammation. After 24 h, different ventilatory RR strategies can superimpose a “second hit” to cause lung damage. The E. coli lipopolysaccharide model reproduces changes in lung function and histology comparable to human ARDS.²⁸  Accordingly, the oxygenation index at BASELINE showed values consistent with mild ARDS (Supplemental Digital Content 2, table 2, https://links.lww.com/ALN/D9). In addition, the diffuse alveolar damage score, as well as markers of decreased epithelial integrity, were pronounced in the current study. Animal models have been used to advance the field of ventilator-induced lung injury pathophysiology because direct and invasive measurements cannot be performed in humans. For example, the pioneering study of Webb and Tierney²⁹  on rats was revisited,³⁰  and the results were confirmed. In preclinical and clinical studies, when V_T and pressures are increased, higher respiratory rates predispose to ventilator-induced lung injury.^7,8,10,31,32  However, to date, no study has compared whether gradual, compared with abrupt, increments in RR mitigate ventilator-induced lung injury in mild experimental ARDS.

We tested the hypothesis that applying an “adaptive” strategy that gradually approached a higher RR would produce less injury than an abrupt increase to that same RR; the latter is regularly applied in clinical settings.³³  Our rationale was that RR represents an important component in the mechanical power formula, because it is not directly related to cumulative energy load over the span of multiple cycles.^2,5,34  Although RR has received less attention, it has been associated experimentally with ventilator-induced lung injury.^5,35  During protective ventilatory strategies in ARDS patients (e.g., V_T ~6 ml/kg), it is usual practice to increase RR abruptly to control hypercapnia.^33,36  However, higher levels of RR augment effective mechanical power, not only by increasing the number of cycles per unit time, but also by increasing microstresses and strains per cycle due to inadequate time to distribute parenchymal forces within the viscoelastic and mechanically heterogeneous injured lung.³⁷  Dynamic compliance dropped for about 30 to 60 s after the abrupt RR increase. One likely explanation is that, during pressure-controlled ventilation, alveolar collapse in slow and/or poorly aerated alveolar units can increase strain heterogeneity and may cause lung injury.^38,39  Alveolar collapse increases the tidal strain for the remaining fraction of the aerated lungs and acts together with short inspiratory time (due to high RR) to affect gas volume distribution.

An injured lung may be characterized by both slow and fast alveolar units. By promoting a gradual increase in RR, both alveolar units remain open and better accommodate stress (reduced airway pressures) for the same strain (V_T). On the other hand, by promoting an abrupt increase in RR and shortening of inspiratory time, only fast alveolar units remain open, which may favor alveolar overdistension, heterogeneity, and lung damage. Thus, fast alveolar units that better accommodate strain tend to overdistend.^2,7,34,40  After recruitment maneuver application, the fraction of slow alveolar units tends to decrease,⁴¹  as does the propensity of alveolar units to become atelectatic, which may decrease regional tidal strains and heterogeneity. Although the injurious biologic impact of abruptly increasing RR was observed in a heterogeneous lung, it was not detected when the lungs were previously subjected to recruitment maneuver. The injurious effects of abruptly increasing the RR may thus depend on the baseline condition of the lung, whether heterogeneous or homogeneous. In addition, the protective effect of the adaptation groups, whether shorter or longer, may rely on the gentle and continuous recruitment of slow alveolar units, while helping to prevent their becoming atelectatic. Although we have shown differences between shorter and longer V_T adaptation in a previous study,¹  no major differences were observed in shorter and longer RR adaptation groups. These differences may depend on the participation (and likely weights) of each variable in the mechanical power formula. For instance, it has been shown that increasing V_T while keeping a stable and low mechanical power by proportionally reducing RR led to signs of ventilator-induced lung injury.⁴²  This means that the potential of V_T to cause lung damage was not fully annulled by RR modulation. We may emphasize that each variable has its own mechanism of injury and adaptations in heterogeneous ARDS lungs. However, because we did not perform experiments dealing with recruitment maneuver application followed by adaptation groups, this hypothesis should be tested in future preclinical studies.

Different components may contribute differently with mechanical power, such as respiratory system peak pressure, V_T, respiratory system driving pressure, peak inspiratory flow, PEEP, and RR.^1–4  During patient enrollment in the ARDS Network trial of V_T (1996 to 1999),³³  mechanical power was not a primary focus, and the potential contribution of an increase of RR to potential harm to the respiratory system was not considered. However, a recent observational retrospective study of 4,549 patients with ARDS showed that RR was associated with mortality,⁵  and should be considered when estimating the potential to inflict lung damage. Here, we expand this debate; not only is a given high RR injurious in mild experimental ARDS, but so, too, may be the rate of increase in RR toward that higher level. In addition, recruitment maneuver may play a relevant role in minimizing the injurious effects of abruptly increasing respiratory rate, by decreasing alveolar unit heterogeneity.

This study has limitations that must be noted. As a set of experiments conducted in a small-animal model of mild ARDS, these results cannot be directly translated to patient care. On the other hand, as far as we know, there are no specific recommendations on the best way to increase RR at the bedside. In addition, we acknowledge that our findings apply to the specific experimental circumstances and time intervals used herein, and that the duration of experiments was not sufficient to determine whether injury was attenuated or simply delayed by the RR adaptation strategies. To address that question directly, however, the observation time would need to be longer (4 to 6 h), which might interfere with gene expression, because additional fluids, vasopressors, and/or inotropes would be required for life support,⁴³  introducing confounding variables. The abrupt increase of RR for 2 h group was chosen as a positive control for ventilator-induced lung injury, even though the abrupt increase of RR for 1 h group is the most translational to clinical practice when an increase in RR is suggested. Measurements of respiratory system mechanics, E-cadherin by immunohistochemistry, diffuse alveolar damage score, and molecular biology are quantitative and, thus, objective indexes, although they may appear to be subjective. Although cumulative diffuse alveolar damage score indicates a significant difference between adaptation and abrupt groups, some biologic markers did not reflect these differences. Indeed, this may be a limitation of the model (type and severity) and/or of the characteristics of the biologic marker (level of lung damage that may lead to an increase of lung injury biomarkers) and the timing of analysis (early or late in the course of lung injury). Finally, morphologic changes are not always associated with early or late modifications in the biologic markers, and morphologic changes might be a more sensitive and early indicators of ventilator-induced lung injury. In addition, one researcher was fully dedicated to changes in RR through time and the maintenance of protective V_T (6 ml/kg) by changing the delta pressure during pressure-controlled ventilation. Although RR is one factor in the formula for mechanical power, changing each factor has its own contribution in the mechanical power calculation.²  For example, in table 1, the differences in RR among groups at FINAL are not the same differences observed in mechanical power values among the groups. This means that respiratory system peak pressure and respiratory system driving pressure contribute to the unifying mechanical power as the causal factor of ventilator-induced lung injury. Dynamic compliance decreased after abruptly changing the RR, as the result of narrowing inspiratory time and likely promoting ventilation to fast alveolar units (Supplemental Digital Content 10, fig. 6, https://links.lww.com/ALN/D17). There is an intrinsic variance in respiratory system driving pressure at BASELINE. However, this is inherent to animal models using intratracheal E. coli lipopolysaccharide, as previously shown by our group.^1,13,42,44  This study was intended as a proof of concept, not to determine the best ventilatory strategy. Future studies are needed to confirm whether a threshold of transition to injury exists, to precisely determine the pace of gradually increasing RR, and to refine by imaging studies that show in real time what happens to the number of fast alveolar units. Future studies should also focus on the association with different biomarker inhibition and ventilatory strategies. Recruitment maneuver prevented the increase in VCAM-1 and CC-16 gene expressions in the abruptly increasing RR groups. We did not perform controlled experiments dealing with recruitment maneuver application followed by adaptation. Nevertheless, we may infer that the protective effect of the adaptation groups, whether shorter or longer, may rely on the gentle and continuous recruitment of slow alveolar units while attenuating their tendency to become atelectatic. This hypothesis should be tested by future studies with adequate power. It has been shown that there are no differences in susceptibility to ventilator-induced lung injury determined by sex.⁴⁵  Rather than justify the use of animals of a single sex in experimental studies, these findings support the inclusion of both.⁴⁶  Nonetheless, apart from any mechanistic insights, these experimental data do have potential clinical relevance; low V_T values and restrained driving airway pressure are sometimes impossible to apply in clinical settings without increases of RR to assure adequate carbon dioxide elimination and pH. Along with our previous work regarding recruitment maneuvers,⁴⁷  V_T,¹  and PEEP,⁴⁸  the present results should be considered further evidence in support of the novel “adaptation” principle of lung protection.

In mild experimental ARDS in rats, we found that gradually increasing RR, compared with abruptly doing so, can mitigate the development of ventilator-induced lung injury. In addition, recruitment maneuver prevented the injurious biologic impact of abrupt increases in RR.

The authors express their gratitude to the following people from the Laboratory of Pulmonary Investigation, Carlos Chagas Filho Biophysics Institute, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil: Andre Benedito da Silva, B.Sc., for animal care; Arlete Fernandes, B.Sc., for her help with microscopy; Camila Machado, Ph.D. student, for her help with microscopy; Maíra Rezende Lima, M.Sc., for her assistance with molecular biology analysis; and Moira Elizabeth Schöttler, M.Sc., Rio de Janeiro, Brazil, and Filippe Vasconcellos, M.Sc., São Paulo, Brazil, for editing assistance.

This study was supported by the Brazilian Council for Scientific and Technological Development (CNPq), the Rio de Janeiro State Research Foundation (FAPERJ E-26/202.766/2018, E-26/010.001488/2019), the São Paulo State Research Foundation (FAPESP 2018/20403-6), the Coordination for the Improvement of Higher Education Personnel (CAPES, 88881.371450/2019-01), and the Department of Science and Technology - Brazilian Ministry of Health (DECIT/MS).

The authors declare no competing interests.

Supplemental Digital Content 1, Figure 1. Experimental configuration, https://links.lww.com/ALN/D8

Supplemental Digital Content 2, Table 1. Oligonucleotide sequences of target gene primers, https://links.lww.com/ALN/D9

Supplemental Digital Content 3, Table 2. Arterial blood gases and mean arterial pressure, https://links.lww.com/ALN/D10

Supplemental Digital Content 4, Figure 2. Spearman correlation, https://links.lww.com/ALN/D11

Supplemental Digital Content 5, Table 3. Diffuse aveolar damage score, https://links.lww.com/ALN/D12

Supplemental Digital Content 6, Figure 3. Histology in abrupt increase of RR during 1 h, https://links.lww.com/ALN/D13

Supplemental Digital Content 7, Figure 4. E-cadherin in abrupt increase of RR during 1 h, https://links.lww.com/ALN/D14

Supplemental Digital Content 8, Figure 5. Cumulative power during ventilatory strategies, https://links.lww.com/ALN/D15

Supplemental Digital Content 9, Table 4. Respiratory parameters, https://links.lww.com/ALN/D16

Supplemental Digital Content 10, Figure 6. Cdyn,_RS during abrupt increase in respiratory rate, https://links.lww.com/ALN/D17