Anesthetized Yorkshire pigs weighing 25–35 kg were pretreated with glycopyrrolate (0.01 mg/kg, intramuscular) 10–15 min before intubation and were pre-anesthetized with telazol (5 mg/kg, intramuscular) and xylazine (2 mg/kg, intramuscular). Sodium pentobarbital (6 mg/kg per hour) was delivered intravenously via a Harvard infusion pump (model 907; Harvard Apparatus, Holliston, MA, USA) to achieve continuous anesthesia. Animals were ventilated using a Galileo™ ventilator (Hamilton Medical, Reno, NV, USA) with baseline ventilation (Vt 12 cc/kg, PEEP 5 cmH2O, and fractional inspired oxygen 100%) at a rate of 15 breaths/minute, adjusted to maintain arterial carbon dioxide tension at 35–45 cmH2O.
A left carotid artery cutdown was performed to gain access for blood gas measurements (Model ABL 2; Radiometer Inc., Copenhagen, Denmark), blood oxygen content analysis (Model OSM 3; Radiometer Inc.), and systemic arterial blood pressure monitoring. A thermodilution pulmonary artery catheter was inserted through the right femoral vein for mixed venous blood gas and oxygen content sampling, along with cardiac output and lung function determinations (Baxter Explorer™ Baxter Healthcare Corp., Irvine, CA, USA). A triple lumen catheter was placed into the right internal jugular vein for fluid, anesthesia, and drug infusion. Pressures were measured using transducers (Argon™ Model 049-992-000A, CB Sciences Inc., Dover, NH, USA) leveled with the right atrium and recorded on a 16 channel Powerlab/16s (AD Instruments Pty Ltd, Milford, MA, USA) with a computer interface.
Surfactant deactivation was achieved by endotracheal instillation with Tween-20 surfactant detergent as previously described [22
]. Briefly, pigs were placed in the right lateral decubitus position and a 0.75 cc/kg 10% solution of Tween-20 in saline was instilled into the right, dependent lung beyond the tracheal bifurcation. Following lavage, the endotracheal tube was reconnected to the ventilator for three breaths and the lungs were then inflated with a Collins supersyringe to twice the baseline Vt for one breath in order to enhance Tween distribution. The endotracheal tube was suctioned, rendering it free from residual Tween and the previous mechanical ventilation regimen was resumed for several minutes. The animal was then rotated to the left lateral decubitus position, and the Tween lavage procedure was repeated in the left lung.
In vivo microscopy
A right thoracotomy was performed with removal of ribs five to seven to expose the lung for in vivo
microscopy. The in vivo
microscope (epiobjective, epillumination) provides real-time images of subpleural alveoli. Our technique for in vivo
microscopy is described in detail elsewhere [24
] (video footage illustrating the technique is available on the internet [25
]). Briefly, the microscope uses a coverslip suction head apparatus. The apparatus is positioned on the visceral pleural surface of the diaphragmatic lobe of the exposed right lung, and gentle suction is applied (5 cmH2
O) at end-inspiration to affix the lung in place. Suction was minimal to limit motion artifact with respiration, without altering alveolar mechanics [22
]. The microscopic images were viewed using a video camera (CCD SSC-S20; Sony), recorded using a Super VHS video recorder (SVO-9500 MD; Sony, Tokyo, Japan), and analyzed using a computerized image analysis system (Image Pro™; Media Cybernetics, Carlsbad, CA, USA). Still images of alveoli were extracted from video at peak inspiration and end-expiration, and alveolar areas were measured using computer image analysis (Figure ). Alveolar stability was expressed as the dynamic change in alveolar area between inspiration and expiration (I-EΔ), with higher values of I-EΔ representative of greater alveolar instability. I-E% was calculated by dividing I-EΔ by the alveolar area at end-expiration.
Figure 1 Photomicrographs of the same subpleural alveoli on inflation and deflation. Alveoli of interest are outlined with black dots and depict the same alveolus at expiration and inspiration. Alveolar area at end-expiration (E) was subtracted from the area of (more ...)
Phase I (conducted in three pigs)
Following surgical preparation, continuous filming of subpleural alveoli was performed before surfactant deactivation to serve as controls. Video was recorded during ventilation with all possible permutations of three experimental levels of Vt (6, 12, and 15 cc/kg) and three experimental PEEP levels (5, 10, and 20 cmH2
O), generating a total of nine experimental groups (Table ). We chose these tidal volumes because 6 and 12 cc/kg were used in the ARDSnet trial and 15 cc/kg is still used in some hospitals. We felt that the PEEP levels covered the gambit between low, medium, and high PEEP used in current clinical practice. In addition, we chose not to conduct a recruitment maneuver before applying PEEP for two reasons: although recruitment maneuvers are used by many clinicians, they are not currently the standard of care; and it is possible that the recruitment maneuver itself, with a high airway pressure for an extended period of time, could damage the lung [26
] and obscure our primary goal of determining the role of multiple ventilator strategies (combination of Vt and PEEP) on alveolar stability and VILI.
Phase I protocol: alveolar size and stability
The order of the nine combinations was randomized. Ventilation was maintained at each combination for 5 min to acquire video in order to assess alveolar mechanics before changing ventilation. After all nine Vt/PEEP combinations in healthy lung, Tween instillation was performed as described above. The in vivo microscope was again placed on the visceral pleural surface and video was recorded for all nine combinations of Vt and PEEP in the surfactant-deactivated lung in a similar manner. It is important to note that the same alveoli were filmed for each Vt/PEEP combination. In the event that alveoli moved out of our field of view for any of the Vt/PEEP combinations, they were excluded from the data analysis. Thus, our data represent the effect of each Vt/PEEP combination on the same individual alveoli in the normal and surfactant-deactivated lung.
The phase I protocol was designed to determine which combination of Vt and PEEP was most effective at stabilizing alveoli. In the subsequent phase II protocol, we tested the hypothesis that the combination of Vt and PEEP determined in the initial phase that resulted in the most stable alveoli would produce the least lung injury, and that the combination that resulted in the most unstable alveoli would result in more severe lung injury. In phase I, we found that a Vt/PEEP combination of 5 cmH2O PEEP and 15 cc/kg Vt caused the most alveolar instability (highest I-EΔ and I-E%), and a combination of 20 cmH2O PEEP with 6 cc/kg Vt caused the least alveolar instability (lowest I-EΔ and I-E%). Thus, these were the two Vt/PEEP combinations that were tested in phase II.
Phase II (conducted in six pigs)
Following surgical preparation, the in vivo microscope was placed on the visceral pleural surface of healthy swine lung and subpleural alveoli were recorded before Tween instillation to serve as controls. Lavage was then performed with Tween as described above. The in vivo microscope was again placed on the visceral pleural surface and animals were divided into two groups: animals in the high Vt/low PEEP group (least alveolar stability) were ventilated with Vt 15 cc/kg and PEEP 5 cmH2O; and those in the low Vt/high PEEP group (most alveolar stability) were ventilated with Vt 6 cc/kg and PEEP 20 cmH2O. Alveolar size at expiration, inspiration, and the number of alveoli per field were measured at each time point. Five minutes of in vivo microscopic footage was recorded every 30 min for three hours. It should be noted that the same four microscopic fields were recorded at each time point to standardize the data collected.
At necropsy the lungs were inflated to 25 cmH2O pressure and held at this pressure for 60 s to normalize lung volume history. The lungs were than allowed to deflate to atmospheric pressure and the samples were taken immediately as described below. A 3 × 3 × 3 cm cubic section of the right lung taken directly beneath the in vivo microscope viewing field and was fixed in 10% formalin. The fixed tissue contained the alveoli that were being observed with the in vivo microscope. The tissue was blocked in paraffin and serial sections were made for staining with hematoxylin and eosin.
A blinded observer evaluated lung tissue; details of this scoring methodology are published elsewhere [6
]. Briefly, the slides were reviewed at low magnification to exclude areas containing bronchi, connective tissue, large blood vessels, and areas of confluent atelectasis, such that histologic data was from parenchymal tissue. These parenchymal areas were assessed at high magnification (400×) in the following manner. Five high power fields (HPFs) were randomly sampled. Features including alveolar wall thickening, intra-alveolar edema fluid, and number of neutrophils were assessed in each of the five HPFs. Specifically, alveolar wall thickening, defined as greater than two cell layers thick, was graded as '0' (absent) or '1' (present) in each field. Intra-alveolar edema fluid, defined as homogenous or fibrillar proteinaceous staining within the alveoli, was graded as '0' (absent) or '1' (present) in each field. A total score/five HPFs for alveolar wall thickening and intra-alveolar edema fluid was recorded for each animal. The total number of neutrophils was counted in each of the five HPFs and expressed as the total number/five HPFs for each animal. All data are expressed as mean ± standard error.
Serum/bronchoalveolar lavage fluid cytokines
Serum and bronchoalveolar lavage (BAL) fluid were obtained at baseline and when the animals were killed. Serum and BAL levels (ng/ml) of IL-1, IL-6, IL-8, IL-10, and tumor necrosis factor (TNF)-α were determined by enzyme-linked immunosorbent assay (Endogen, Woburn, MA, USA).
Neutrophil elastase activity
Neutrophil elastase activity was determined in serum drawn both at baseline and at the end of the experiment, and in BAL fluid obtained at necropsy. Specifically, elastase activity was determined by incubating either 100 μl serum or BAL fluid and 400 μl of 1.25 mmol/l methoxy succinyl-ala-pro-val-p-nitroanilide (specific synthetic elastase substrate) in a 96-well enzyme-linked immunosorbent assay plate at 37°C for 18 hours. After incubation, the optical density was read at 405 nm. Data are expressed as nanomoles elastase substrate degraded per milligram of protein per 18 hours (nmol/l per 18 hours per mg).
Matrix metalloproteinase (MMP)-2 and MMP-9 activities were measured using a type I gelatin zymography technique. A volume of 20 μl BAL fluid or 2.5 μl serum was electrophoresed (30 mA) for two hours at 4°C. The slab gels were then incubated for one hour with 2.5% Triton X-100 at 22°C and the gels washed with water, then incubated at 37°C in TRIS/NaC/CaCl2 buffer overnight. The gels were stained with Coomasie blue, destained with 20% methanol/5% acetic acid (22°C), and the molecular weights of the gelatinolytic zones were compared with standard MMP-2 and MMP-9. The concentrations of MMP-2 and MMP-9 were calculated by scanning of the gels using an image densitometric system (Kodak Image Analysis System; Kodak, Rochester, NY, USA). MMP-2 and MMP-9 concentrations are expressed in densitometric units.
A 2 × 2 × 2 cm section of lung directly adjacent to each histologic section was used for wet-to-dry weight ratio determination. The samples were placed in a dish and weighed, dried in an oven at 65°C for 24 hours, and weighed again. This was repeated until there was no weight change over a 24-hour period, at which time the samples were deemed to be dry. Lung water is expressed as a wet to dry weight ratio.
The experiments described in this study were performed in adherence with the US National Institutes of Health guidelines for the use of experimental animals in research. The protocol was approved by the Committee for the Humane Use of Animals at our institution.
All values are reported as mean ± standard error. Differences between groups were determined using one-way analysis of variance, and differences within groups were determined using repeated measures analysis of variance. Whenever the F ratio indicated significance, a Newman-Keul test was used to identify individual differences. P < 0.05 was considered statistically significant.