High-frequency oscillatory ventilation (HFOV) at higher frequencies minimizes the tidal volume. However, whether increased frequencies during HFOV can reduce ventilator-induced lung injury remains unknown.
After the induction of acute respiratory distress syndrome in the model by repeated lavages, 24 adult sheep were randomly divided into four groups (n = 6): three HFOV groups (3, 6, and 9 Hz) and one conventional mechanical ventilation (CMV) group. Standard lung recruitments were performed in all groups until optimal alveolar recruitment was reached. After lung recruitment, the optimal mean airway pressure or positive end-expiratory pressure was determined with decremental pressure titration, 2 cm H2O every 10 min. Animals were ventilated for 4 h.
After lung recruitment, sustained improvements in gas exchange and compliance were observed in all groups. Compared with the HFOV-3 Hz and CMV groups, the transpulmonary pressure and tidal volumes were statistically significantly lower in the HFOV-9 Hz group. The lung injury scores and wet/dry weight ratios were significantly reduced in the HFOV-9 Hz group compared with the HFOV-3 Hz and CMV groups. Expression of interleukin-1β and interleukin-6 in the lung tissue, decreased significantly in the HFOV-9 Hz group compared with the HFOV-3 Hz and CMV groups. Malondialdehyde expression and myeloperoxidase activity in lung tissues in the HFOV-9 Hz group decreased significantly, compared with the HFOV-3 Hz and CMV groups.
The use of HFOV at 9 Hz minimizes lung stress and tidal volumes, resulting in less lung injury and reduced levels of inflammatory mediators compared with the HFOV-3 Hz and CMV conditions.
High-frequency oscillatory ventilation provides efficient gas exchange using very high frequencies and low tidal volumes, while maintaining a constant mean airway pressure
The effects of frequency in high-frequency oscillatory ventilation on ventilation-induced lung injury are yet to be evaluated
This study suggests that high-frequency oscillatory ventilation at higher frequencies minimizes lung stress and tidal volume, resulting in less lung injury and reduced local lung inflammation
ACUTE respiratory distress syndrome (ARDS) is the most severe manifestation of acute lung injury caused by various direct and indirect factors. Inflammatory pulmonary edema, severe hypoxemia, and diffuse endothelial and epithelial injury are key characteristics of ARDS, which can often lead to multiple organ failure.1,2 Despite being the most widely used approach to treat ARDS, conventional mechanical ventilation (CMV) can induce ventilator-induced lung injury (VILI) due to alveolar overdistension and the cyclic collapse/reopening of lung units, eliciting inflammatory responses, and worsening the damage.3,4
Recently, protective CMV has been shown to decrease lung and systemic inflammatory responses and improve the outcome of ARDS patients by minimizing alveolar distention (lower tidal volumes and plateau pressure limitation).5–7 Mechanical ventilation-induced alveolar instability can injure lung tissue.8,9 Adjustment of the positive end-expiratory pressure (PEEP) level has been used to maintain adequate end-expiratory lung volume at the end of expiration, which protects injury tissue from cyclic tidal opening and closing. Additionally, meta-analyses of clinical studies have shown that higher PEEP values significantly contribute to improved survival in patients with severely hypoxemic ARDS.10 Higher PEEP levels should be recommended for those who are highly likely to have a diffuse ARDS pattern and high lung recruitability.11 However, despite considerable progress in recent years, mortality rates in ARDS patients are still notably high (from 35–40%).12,13
High-frequency oscillatory ventilation (HFOV), a theoretically optimal lung-protective strategy,14 allows higher mean airway pressure (mPaw) and extremely small tidal volumes, which are often smaller than the anatomical dead space.15 HFOV application enables a consistently sufficient end-expiratory volume without inducing the overdistension or collapse of alveoli due to the much smaller tidal volumes.16 Rapid piston oscillations drive gas transport and active inspiration and expiration. Higher mPaw values and lower tidal volumes prevent alveolar derecruitment and overdistension respectively, thereby maintaining alveolar stability.17 Overall, HFOV improves oxygenation, reduces inflammatory processes and histopathological damages, and attenuates oxidative lung injury compared with protective mechanical ventilation.18 Conceptually, HFOV constitutes an attractive lung-protective ventilatory modality.
During HFOV, the tidal volume is primarily determined by frequency. Lower ventilating frequencies in HFOV generate airflow similar to that of CMV, potentially causing increased lung distension, alveolar instability, and even VILI.19 In contrast, higher frequencies in HFOV result in less distension in the small airways and air spaces. In pediatrics and neonatology, higher frequencies of 10–15 Hz are routinely used. One small animal study showed that in saline lavage-injured rabbits, HFOV at 15 Hz ventilation yielded lower tissue neutrophil infiltration than HFOV at 5 Hz after 4 h.20 In adults, the recommended HFOV frequency is 5 to 6 Hz.21 One clinical survey showed that 62.5% of adult patients with ARDS exhibited significantly improved gas exchange under HFOV.22 Another study demonstrated that higher frequencies between 6 and 10 Hz were within the feasible range of ventilation frequency for adults.23 However, whether higher frequencies in HFOV can reduce VILI in adults remains unknown. We hypothesize that HFOV at higher frequencies can prevent VILI in a sheep model of ARDS.
Our aim was to investigate the effects of ventilation at various frequencies on gas exchange, respiratory mechanics, hemodynamics, histological signs of lung injury, and expression of inflammatory cytokines and compare them between HFOV and CMV in large animal models of ARDS.
Materials and Methods
This study was approved by the Science and Technological Committee and the Animal Use and Care Committee of the Southeast University Medical College. All experiments were performed according to the Guidance for the Care and Use of Laboratory Animals.24 Twenty-six adult male sheep were studied (Southeast University Medical College Laboratory Animal Center, Nanjing, China).
Animals were kept in fasting for 24 h before the experiments but with free access to water. Before the experiment, sheep were placed in a supine position and anesthetized by intravenous injection of ketamine (3 mg/kg) and midazolam (0.2 mg/kg). To maintain anesthesia and paralysis, a continuous intravenous infusion of ketamine (2 mg·kg−1·h−1), midazolam (0.1 mg·kg−1·h−1), and atracurium (0.4 mg·kg−1·h−1) was applied. The airways were secured by tracheotomy (Tracheal tube, ID 8.0, Evac/Larz; Tyco Healthcare, Dublin, Ireland) with 2% lidocaine being provided as local anesthesia. The animals were mechanically ventilated with an Evita XL ventilator (Dräger, Lubeck, Germany) using the volume control ventilation mode. A fraction of inspired oxygen (FiO2) of 1.0, a PEEP of 5 cm H2O (1 cm H2O = 0.098 kPa), a tidal volume (VT) of 6 ml/kg, a respiratory rate of 30 breaths/min, and an inspiration/expiration (I:E) ratio of 1:2 were used as the initial settings of the mechanical ventilator. Ventilation was also adjusted to maintain the arterial carbon dioxide pressure (PaCO2) between 30 and 40 mmHg (1 mmHg = 0.133 kPa). A Swan–Ganz catheter (Arrow International, Reading, PA) was inserted through the internal jugular vein to measure central venous pressure (CVP), pulmonary arterial wedge pressure (PAWP), and also to collect mixed venous blood. A thermistor-tipped PiCCO catheter (Pulsion Medical System, Munich, Germany) was advanced through the right femoral artery to monitor the mean arterial pressure (MAP), and arterial blood samples were collected from the arterial catheter. Pneumatic tubing of VENTRAK™ Model 1550 respiratory mechanics monitoring system (Novametrix Medical Systems Inc., Wallingford, CT) was placed between the ventilator Y-piece and the inlet to the sheep and was secured in an upward direction. An esophageal balloon was inserted into the lower third of the esophagus and was attached to VENTRAK™ Model 1550 system to monitor the esophageal pressure. The correct positioning of the esophageal balloon is assessed by an occlusion test with spontaneous breaths, specifically before atracurium was given; when the airways are closed at the end of expiration, and an active inspiration occurs, a drop in esophageal pressure takes place.25 In this scenario, there are no changes in lung volume and the decrease in esophageal pressure equals the decrease in airway pressure.
Induction of Lung Injury
After the animal preparation, the animals were stabilized for 30 min and baseline measurements (TBaseline) were taken. ARDS was induced by performing bilateral lung lavages with 30 ml/kg isotonic saline (38°C). The saline was infused through the funnel while the chest was gently massaged. After the massage maneuver, the fluid was allowed to drain by gravity and excessive fluid was removed by negative pressure suction through the proximal portion of the endotracheal tube. The alveolar lavages were repeated every 10 min until the PaO2/FiO2 ratio decreased to less than 60 mmHg and remained stable for 30 min (TARDS) with unchanged ventilatory parameters.
The study design is presented in figure 1. After saline lavage-induced ARDS, animals were stabilized for 30 min, postinjury measurements (TARDS) were obtained, and the animals were randomly assigned to one of the following four treatment groups (n = 6) according to a random digital table. The following initial settings were used for each group:
Group 1, HFOV-3 Hz: frequency = 3 Hz; Fio2 = 1.0; mPaw = 20 cm H2O; bias flow = 30 l/min; amplitude = 70 cm H2O; Ti = 33% (HFOV Ventilator: 3100 B, SensorMedics Corporation, Yorba Linda, CA).
Group 2, HFOV-6 Hz: frequency = 6 Hz; Fio2 = 1.0; mPaw = 20 cm H2O; bias flow = 30 l/min; amplitude = 70 cm H2O; Ti = 33%.
Group 3, HFOV-9 Hz: frequency = 9 Hz; Fio2 = 1.0; mPaw = 20 cm H2O; bias flow = 30 l/min; amplitude = 70 cm H2O; Ti = 33%.
Group 4, CMV: volume control ventilation mode with Fio2 = 1.0; PEEP = 5 cm H2O; VT = 6 ml/kg; respiratory rate = 30 breaths/min; I:E ratio = 1:2.
After initiating the ventilation, continuous positive airway pressure was used to provide a sustained inflation at 40 cm H2O for 40 s in the CMV group, after which the PEEP was adjusted back to 20 cm H2O, and the mode was adjusted back to volume control ventilation. In the three HFOV groups, the mean airway pressure (mPaw) was increased to 40 cm H2O without oscillation for 40 s, and the mPaw was then adjusted to 36 cm H2O. The lung recruitment maneuver was performed repeatedly every 5 min until the Pao2/Fio2 was greater than 400 mmHg, or the increase in Pao2/Fio2 was less than 10%. This condition was regarded as full lung inflation or the beginning of lung overdistension,26,27 and we marked the time point as TRM.
After full recruitment, PEEP or mPaw levels were reduced at a decrement of 2 cm H2O every 10 min until PaO2/FiO2 was less than 400 mmHg, or the decrease in PaO2/FiO2 was greater than 10%. A total of +2 cm H2O was added to the lowest PEEP or mPaw, which maintained oxygenation to obtain the optimal PEEP or mPaw. The animals were then ventilated with those settings for 4 h.
Hemodynamics and gas exchange indices were determined hourly and recorded at the following time points: TBaseline, TARDS (after the lung injury period), and TPRM (postrecruitment maneuver), as well as 1, 2, 3, and 4 h during the ventilation period (fig. 1). Electrocardiography (1500; Spacelab Medical Inc., Issaquah, WA) was monitored continuously, and the heart rate was recorded. The CVP, mean pulmonary arterial pressure (PAP), and PAWP were monitored, using calibrated pressure transducers. The MAP was continuously monitored by PiCCO, and the cardiac output (CO) was calculated. The VT, airway pressure (PAir), mean airway pressure (Pmean), plateau pressure (Pplat), esophageal pressure (PEso), mPaw, airway resistance (R), and static compliance of the respiratory system (Crs) were monitored, using a VENTRAK™ Model 1550 respiratory mechanics monitoring system.
All blood gas measurements were performed using an automated blood gas analyzer (Nova M; Nova Biomedical, Waltham, MA). Paired arterial and mixed venous blood samples were drawn and analyzed at each measurement point.
The lung alveolar pressure (Palveo) in the HFOV groups was measured at each time point using the inspiratory occlusion maneuver (clamping the endotracheal tube, connected to the end of the Y-piece of the breathing circuit, for 10 s to allow measurements of the resulting intratracheal pressure during HFOV, which mimics the manual “Insp hold” function of the Evita XL ventilator during CMV (see fig., Supplemental Digital Content 1, http://links.lww.com/ALN/A930, which illustrates the measurement of alveolar pressure in HFOV groups). In the HFOV groups, the transpulmonary pressure (Ptp) was computed as: Ptp= average PAlveo − average PEso (see fig., Supplemental Digital Content 2, http://links.lww.com/ALN/A931, which illustrates the airway pressure, esophageal pressure, and transpulmonary pressure in HFOV group). The airway plateau pressure (PPlat) and esophageal pressure in CMV group were recorded as described by Chiumello et al.28 Ptp was determined as the difference between the PPlat value and the PEso value in CMV groups (see fig., Supplemental Digital Content 3, http://links.lww.com/ALN/A932, which illustrates the airway pressure, esophageal pressure, and transpulmonary pressure in CMV group).
During the experimental period, the central body temperature was monitored with a temperature probe on PICCO machine and maintained at 36.5°–37.5°C by a thermostatically controlled heating pad. A continuous infusion of a 5 ml·kg−1·h−1 balanced electrolyte solution was administered during the experiment to maintain adequate hydration, and the MAP was maintained above 60 mmHg with rapid infusions of 0.9% saline solution of up to 20 ml/kg, if required.
Tissue Removal and Lung Processing
After 4 h of ventilation, all animals were euthanized by a bolus injection of saturated potassium chloride solution. After midline sternotomy, the right lung was immediately removed, inflated with buffered 10% formalin at a pressure of 30 cm H2O, and fixed in a buffered 10% formalin bath for 24 h, to prepare the tissue for histological examination. The upper, ventral, and dorsal medial and lower lobes of the left lung were immediately removed and stored at −80°C for cytokine determination and the measurement of myeloperoxidase expression and malondialdehyde activity.
Wet/Dry Weight Ratios
Three blocks (1 × 1 × 1 cm3) were cut from the upper, ventral, and dorsal lower lobes of the left lung. They were weighed and then dried to a constant weight at 50°C on consecutive days (5–7 days) in an oven. The lung wet/dry weights ratio was calculated to estimate the severity of lung tissue edema.
Slides from the upper, ventral and dorsal medial, and lower lobes were stained with hematoxylin and eosin, and examined blindly by two lung pathologists. Lung pathology was assessed on the basis of five histological criteria: the severity of alveolar exudates, alveolar hemorrhage, polymorphonuclear neutrophil infiltrates in the air space and/or in the alveolar wall, interstitial edema, and hyaline membrane formation. The following scale was used for grading: 0 = no or minimal damage; 1+ = mild damage; 2+ = moderate damage; and 3+ = severe damage.29
Two pathologists individually scored the sample slides (10 visual fields per slide), and an average was taken between the two scores to obtain the final score. For each sample, a composite lung injury score was calculated, and then corrected according to the number of visual fields. An average of the scores of the five lobes was taken to obtain the total lung injury score.
A total of 100 mg of the medial ventral, lower ventral, and dorsal lobes of the left lung tissue was separately added to 1.0 ml of normal saline and ground. The expressions of interleukin (IL)-1β and IL-6 were measured with a goat anti–sheep-specific ELISA kit (Adlitteram Diagnostic Laboratories Inc., San Diego, CA).
Measurement of Malondialdehyde Content and Myeloperoxidase Activity
Malondialdehyde content and myeloperoxidase activity in the lung tissue homogenates were measured by spectrophotometry. Briefly, 100 mg of each of the medial ventral, lower ventral, and dorsal lobes of the left lung tissue was added to 1.0 ml of normal saline and ground separately. The lung homogenates were then used to measure myeloperoxidase activity or malondialdehyde content. Myeloperoxidase activity was measured by the change in absorbance at 460 nm and 37°C, with a 755B spectrophotometer (Analytical Apparatus Company, Shanghai, China). Malondialdehyde levels were measured as an index of lipid peroxidation using colorimetric methods (Nanjing Jiancheng Co., Nanjing, China). Malondialdehyde and thiobarbituric acid were oxidized to red products, which exhibited an absorption maximum at 532 nm.
Data were expressed as the means ± SD. All analyses were performed using the SPSS 16.0 statistical package (SPSS Inc., Chicago, IL). Before performing the analysis procedure, the distribution of the data was assessed using the Kolmogorov–Smirnov test. The results showed that all data were normally distributed. A two-way repeated measures ANOVA was applied to evaluate the effects of time and group differences on hemodynamics, gas exchange, and respiratory variables. A two-way ANOVA was used to compare lung injury scores, wet/dry weight ratio, and mediators of inflammation. In the post hoc analysis to separate differences between the means, we used Tukey pairwise multiple comparison test for a factor or for the interactions of factors, when a significant F ratio was obtained. All tests were two-tailed, and P values less than 0.05 were considered statistically significant.
ARDS was successfully induced by repeated lung lavages in all 26 sheep. A total of 4–18 lavages were used to reach the same criterion of a Pao2/Fio2 ratio below 60 mmHg. Two animals were excluded due to hypoxemia and refractory shock (one in the CMV group and one in the HFOV-3 Hz group). Twenty-four sheep were analyzed (38.3 ± 2.3 kg). As shown in table 1 and table 2, there was no significant difference in gas exchange and hemodynamic parameters among animals at baseline and TARDS. No cases of pneumothorax were observed during the experiment.
There was no significant difference in heart rate, MAP, CO, CVP, and PAWP among the four groups. CO showed no significant difference throughout the entire experiment. After ARDS induction, CVP and PAWP, in the HFOV-3 Hz group, were significantly higher with reference to the baseline. After recruitment and over the study period, MAP in the HFOV-3 Hz group showed significant decreases, whereas CVP in the HFOV-3 Hz, HFOV-6 Hz, and CMV groups were significantly higher during the 4-h ventilation period than baseline and TARDS. Furthermore, PAWP in the HFOV-3 Hz and HFOV-6 Hz groups also had significant increases than baseline (table 1).
Gas exchange and Respiratory Parameters
Relative to TARDS, the PaO2/FiO2 ratios were significantly improved in all four groups after recruitment and during the 4-h ventilation period (see fig., Supplemental Digital Content 4, http://links.lww.com/ALN/A933, which illustrates the PaO2/FiO2 ratio over the study course). The Paco2 was significantly decreased during HFOV compared with CMV after recruitment. The tidal volume values observed in the HFOV-6 Hz and HFOV-9 Hz groups were lower than those in the HFOV-3 Hz and CMV groups (P < 0.001). During the 4-h ventilation period, the Crs was significantly improved in the HFOV-9 Hz group when compared with HFOV-3 Hz and CMV groups (p = 0.013). The total resistance of the respiratory system was increased significantly after lung injury in all groups. In the HFOV-6 Hz and HFOV-9 Hz groups, the total resistance of the respiratory system was lower than other two groups (P < 0.001; table 2).
Once the optimal oxygenation was achieved after recruitment, there was no significant difference in esophageal pressure among the three HFOV groups. In addition, transpulmonary pressure in the HFOV-9 Hz group was significantly lower than the HFOV-3 Hz and CMV groups during 4-h period (fig. 2). There were no significant differences in mean alveolar pressure among the four groups (see table, Supplemental Digital Content 5, http://links.lww.com/ALN/A934, which shows the alveolar pressure among the four groups). In contrast, the pressure drop from the ventilator to the lung alveolar space was significantly higher in the HFOV-9 Hz group than in all of the other groups (see table, Supplemental Digital Content 6, http://links.lww.com/ALN/A935, which shows the difference of airway pressure and alveolar pressure among the four groups).
General Appearance and Lung Histopathology
In general, alveolar consolidation and hemorrhage mainly appeared in the dorsal areas of the lower lobes rather than in the ventral side. Lung histopathology revealed that animals ventilated under HFOV-3 Hz and CMV exhibited more hemorrhage and neutrophil infiltration in the alveoli and interstitium, alveolar atelectasis, interstitial lymphocyte infiltration, and hyaline membrane formation compared with the other two groups (fig. 3).
The semiquantitative total lung injury scores in the HFOV-9 Hz group (3.4 ± 0.9, HFOV-9 Hz) were significantly lower than those of other three groups (HFOV-3 Hz, 5.5 ± 1.5; HFOV-6 Hz, 4.45 ± 1.2; and CMV, 5.9 ± 1.4) after the 4-h mechanical ventilation period (P < 0.001; fig. 4). In addition, the lung injury scores of the dorsal lung tissues were significantly higher than those of other parts of the lung tissue (table 3).
Wet/Dry Weight Ratio
The HFOV-9 Hz group exhibited statistically significantly decreased wet/dry ratio (7.5 ± 2.5, HFOV-9 Hz) in the total lung, compared with the other group (HFOV-3 Hz, 9.5 ± 3.2; HFOV-6 Hz, 8.1.5 ± 2.4; and CMV, 9.9 ± 3.2; P < 0.001; fig. 5). Additionally, the wet/dry ratio of the left upper lobe tissue was significantly lower than that of other parts of the lung tissue (left lower ventral and left lower dorsal lobes) in all four treatment groups (table 4).
Expression Levels of IL-1β and IL-6 in Lung Tissue
Expression levels of IL-1β and IL-6 in the lung tissue were markedly lower in the HFOV-9 Hz group (224 ± 50 and 290 ± 53 pg/ml, HFOV-9 Hz) after the 4-h ventilation compared with the HFOV-3 Hz and CMV groups (HFOV-3 Hz, 287 ± 54 and 341 ± 48 pg/ml; HFOV-6 Hz, 252 ± 50 and 320 ± 44 pg/ml; CMV, 279 ± 58 and 331 ± 52 pg/ml; P < 0.001). In the four treatment groups, IL-1β and IL-6 expression levels in the left medial ventral lobe of the lung tissue were lower than those in the left lower ventral lobe and left lower dorsal lobe (fig. 6).
Malondialdehyde Content and Myeloperoxidase Activity in Lung Tissue
Malondialdehyde content and myeloperoxidase activity in the lung tissues after the 4-h ventilation period in the HFOV-9 Hz group (HFOV-9 Hz, 3.93 ± 0.58 units/g and 2.07 ± 0.34 nmol/mg) were significantly lower than those in the HFOV-3 Hz and CMV groups (HFOV-3 Hz, 4.62 ± 0.58 units/g and 2.48 ± 0.39 nmol/mg; HFOV-6 Hz, 4.23 ± 0.55 units/g and 2.27 ± 0.36 nmol/mg; and CMV, 4.65 ± 0.55 units/g and 2.47 ± 0.33 nmol/mg; P < 0.001). In the CMV, HFOV-9 Hz, and HFOV-6 Hz groups, the malondialdehyde content and myeloperoxidase activity in the lung tissue of the left lower dorsal lobe were significantly higher than those in the left medial ventral lobe and left lower ventral lobe (fig. 7).
The major findings of this study are as follows. (1) In the HFOV-9 Hz group, the histological markers, wet/dry ratio of lung injury, expression levels of IL-1β and IL-6, and malondialdehyde and myeloperoxidase activity were all statistically significantly decreased in the sheep model of ARDS after 4-h ventilation. (2) Compared with the HFOV-3 Hz and CMV groups, the transpulmonary pressure and tidal volumes were statistically significantly reduced in the HFOV-9 Hz group during the 4-h ventilation. (3) Oxygenation, carbon dioxide elimination, and respiratory system compliance were improved in all the three HFOV groups (3, 6, and 9 Hz) and the CMV group after lung recruitment.
Compared with lower frequencies, HFOV at higher frequencies resulted in milder injury. The total injury scores and wet/dry weight ratios in animals receiving HFOV at 9 Hz were considerably less than those in animals receiving HFOV at 3 Hz or CMV. Our results are consistent with a previous study, which found that HFOV at higher frequencies yielded lower scores of lung injury compared with HFOV at lower frequencies in small animals.20 The inflammatory response and the overexpression of proinflammatory mediators are contributing factors to ARDS and VILI pathogenesis.30 We found that the levels of IL-1β and IL-6 were markedly lower in the HFOV-9 Hz group after ventilation, and the same results were obtained in different areas of the lung tissue. Additionally, a previous study in large animal models of ARDS, suggested that HFOV-6 Hz may reduce inflammation more effectively than conventional lung-protective ventilation.31 Moreover, in our study, myeloperoxidase activity and malondialdehyde expression after ventilation were markedly lower in the HFOV-9 Hz group compared with the HFOV-3 Hz and CMV groups. Myeloperoxidase activity quantifies neutrophil infiltration in the lung tissue, and malondialdehyde expression measures the rate of lipid peroxidation and reflects oxidative damage; therefore, reduced myeloperoxidase activity and malondialdehyde expression potentially account for the alleviation of lung injury.
With the use of small tidal volumes within a range of safe lung volumes, HFOV minimizes the risks of both overdistension during inspiration, and derecruitment during expiration.19,32 Rather than plateau pressure and tidal volume, the primary determinants of VILI are pulmonary stress and strain, the clinical equivalents of which are transpulmonary pressure (alveolar pressure minus pleural pressure), and the ratio of volume change to the functional residual capacity, respectively.28,33 The pressure amplitude (∆P) of oscillation was markedly attenuated in the distal airway, and the frequency strongly influenced the drop in ∆P in the airway, especially at higher frequency. Additionally, throughout the entire respiratory cycle during HFOV, a higher mean airway pressure was maintained to avoid alveolar collapse. This effect occurred in synchrony with the decrease in pressure amplitude to decrease alveolar volume expansion and pressure, thereby minimizing lung injury. The transpulmonary pressure was statistically significantly lower in the HFOV-9 Hz group than in the other groups, which minimized the stress experienced by the lung parenchyma and lung injury.
HFOV at higher frequencies utilizes lower tidal volumes, which confers protection against VILI. As shown by Hager et al.,15 the VTs delivered to patients during HFOV were determined by the frequency and the diameter of the internal endotracheal tube. Gattinoni et al.34 noted that ARDS patients have both decreased lung compliance and a smaller volume for air exchange due to a large number of collapsed alveoli (the “baby-lung” concept). A previous study demonstrated that normal tidal volumes caused alveolar overdistension and exacerbated alveolar instability in injured porcine lungs.35 Additionally, in ARDS patients receiving CMV, even low tidal volumes were still found to cause tidal overdistension.36 Reducing the intratidal alveolar opening and closing was found to be imperative for keeping the lungs open in ARDS patients, especially in patients with higher lung recruitability.37 In our experiment, the tidal volume decreased from 4.8 to 1.8 ml/kg as the frequency increased from 3 to 9 Hz, evidently reducing the lung strain.
Our saline lavage-induced ARDS model demonstrated that this type of lung injury is area-dependent. The lung injury score, myeloperoxidase activity, malondialdehyde content, and expression of the proinflammatory mediators IL-1β and IL-6 were reduced in the medial ventral lobe when compared with the dorsal lobe in all four treatment groups. In a study on ARDS patients, CT scans showed heterogeneity in the location of lung injury in both the craniocaudal, and the sternovertebral gradients. Our results demonstrated heterogeneity in the distribution of inflammatory mediators across various lung regions, and this finding was in accordance with a study by Gattinoni et al.38 These results aid in the understanding of the heterogeneity of lung injury in patients with pulmonary ARDS. The transpulmonary pressure and tidal volume were statistically significantly lower in the HFOV-9 Hz group compared with the other groups, which could prevent overdistension in normal alveoli, thus attenuating lung injury.
Despite the small tidal volume used in HFOV, it is highly effective and efficient at carbon dioxide elimination. In our study, the PaO2 was significantly reduced in all ventilated groups using HFOV compared with the CMV group, and there were no differences among the three HFOV groups. Unlike gas transport by bulk delivery in CMV, gas transport in HFOV takes place via a number of convective and diffusion mechanisms. Such mechanisms include the local bulk flow of gas to alveolar units close to the proximal airways, asymmetric velocity profiles, Taylor dispersion, asynchronous filling of adjacent alveolar spaces (termed pendelluft), and cardiogenic mixing.32,39 The effective carbon dioxide elimination in HFOV may be attributed to the aforementioned gas exchange mechanisms, high-frequency, high-oscillatory pressure, and large endotracheal tube with an internal diameter of 8 mm.
Conventionally, positive pressure in mechanical ventilation affects hemodynamics. HFOV with higher mPaw values increased the intrathoracic pressure, reduced the pressure gradient and the amount of venous return to the heart, and subsequently, decreased cardiac performance. Additionally, HFOV reduced the biventricular preload, leading to decreased stroke volume, and CO without fluid resuscitation.40 Furthermore, when switching from CMV to HFOV, the hemodynamic responses were dependent on the predefined setting of PEEP during CMV, and on the applied mean airway pressure during HFOV.41 In our study, CVP and PAWP were higher after recruitment and during the ventilation period compared with baseline in all four groups. The heart rate, MAP, CO, CVP, and PAWP were not different among the four study groups. During the course of the experiment, continuous fluid infusion prevented hypovolemia, which in turn prevented hemodynamic instability.
Our study provides insights into a feasible strategy of mechanical ventilation in adult patients. In our large animal model, HFOV at higher frequencies minimized the stress and strain on the lung, resulting in reduced VILI. However, some limitations must be acknowledged. First, we did not directly measure the pressure of oscillatory ∆P changes in the alveoli. From our results, we were unable to determine the stability of the alveoli in different areas of the lung. Second, the short duration of this study may not be able to account for the discrepancies between various ventilator strategies. Finally, surfactant-depleted collapsed lungs induced by bilateral pulmonary lavages responded better to PEEP application and showed improved lung recruitment; nevertheless, this animal model lacks the ability to address ARDS induced by other factors.
In summary, our large animal model of ARDS showed that HFOV at different frequencies maintained gas exchange after recruitment. Compared with CMV and HFOV at lower frequencies, HFOV-9 Hz delivered smaller tidal volumes and minimized stress on the lung, all of which resulted in lower degrees of lung injury and reduced the expression of inflammatory mediators. Therefore, HFOV at higher frequencies may constitute a promising lung-protective approach to the treatment of ARDS.