In the face of widespread use of lung-protective, low-volume ventilation in patients with acute lung injury, interest in the recruitment maneuver (RM) is growing. Little is known about lung-morphometric effects of the RM as compared with positive end-expiratory pressure (PEEP) titration (PT) without the RM.
RM was defined as a stepwise change in PEEP from baseline to 10, 20, 30, and 20 cm H(2)O every 30 s, after which PEEP was reset at the lower inflection point + 2 cm H(2)O. For PT, PEEP was simply increased from baseline to the lower inflection point + 2 cm H(2)O. Both maneuvers were performed in 10 lung-lavaged dogs. Computed tomography of the lung was performed before and 30 s and 30 min after the maneuver.
Thirty seconds after the maneuver, the decrease in the amount of nonaerated plus poorly aerated lung was greater and decreases in Hounsfield units in the caudal and dorsal lung regions were greater with the RM than with the PT. The hyper-aerated lung volume after the RM tended to be greater than that after the PT. At 30 s and 30 min after the maneuver, gas plus tissue volume, gas-only volume, and gas-tissue ratio of the lung were greater with the RM than with the PT. At both time points after the maneuver, the coefficient of variation of regional Hounsfield units, an index of regional heterogeneity of aeration, was lower with the RM than with the PT.
Compared with PT, the RM resulted in a greater lung volume, better aeration of the most dependent lung, and less regional heterogeneity of aeration. However, the RM tended to induce a greater increase in hyperaerated lung volume than did the PT.
ALVEOLAR recruitment is an important goal in the treatment of acute respiratory distress syndrome (ARDS) or acute lung injury (ALI). Alveolar recruitment is used not only to reverse hypoxia but also to circumvent ventilator-associated lung injury associated with repetitive shearing of bronchioloalveolar units. 1–3To achieve alveolar recruitment in the ARDS or ALI lung, titrating positive end-expiratory pressure (PEEP) to a value near the lower inflection point (LIP) of the inflation pressure–volume (P–V) curve has long been in vogue. 4–7Accumulating evidence, however, indicates that alveolar recruitment continues well above the LIP, 8–11and thus titrating PEEP to the LIP on the inflational P–V curve is not adequate for maximal recruitment of an ARDS lung. 12,13
This article is accompanied by an Editorial View. Please see: Rouby J-J: Lung overinflation: The hidden face of alveolar recruitment. Anesthesiology 2003;99:2–4.
Opening of collapsed alveoli in ARDS is known to require a distending pressure higher than the inspiratory pressure usually achieved during conventional ventilation, due to the heightened collapsing forces. 8,9,14Furthermore, reopening collapsed alveoli is time-dependent, requiring 20–60 s (which is presumably inversely related to the level of applied pressure). 15–17The concept of augmented (pressure × time) for alveolar opening is now incorporated into a variety of maneuvers referred to as “recruitment maneuvers” (RMs). 1,9,14,18,19
A few clinical and laboratory studies have investigated the physiologic effects of the RM with regard to gas exchange and respiratory mechanics. 9,18–23To our knowledge, however, little is known on how lung morphometry is affected by the RM in both quantitative and qualitative terms of recruitment. In addition, the RM could induce generalized and/or regional overdistension of the lung as was shown with the use of PEEP. 7,24,25Because the RM generally exploits higher airway pressures than a simple PEEP titration (PT), 1,9,14,18,19such untoward effects may occur to a greater degree with the RM than with the PT. In this study, we investigated the volumetric and topographic changes in the lung induced by the RM in surfactant-depleted canine lungs and compared them with those induced by a traditional method of recruitment, i.e. , PT without the RM.
Materials and Methods
Preparation of Animals
Ten male mongrel dogs (mean weight ± SD, 21.0 ± 2.8 kg) were used for this study. Care and handling of the animals complied with the guidelines of the United States National Institutes of Health. Our institution's animal care committee approved the experimental protocol. Ketamine (5 mg/kg) and atropine sulfate (0.1 mg/kg) were administered intramuscularly to prepare for general anesthesia, which was induced by an intravenous bolus injection of sodium pentobarbital (25 mg/kg) in two divided doses via a peripheral vein. Anesthesia was maintained by a continuous intravenous infusion of sodium pentobarbital (3 mg·kg−1·h−1), and muscle paralysis was maintained with a continuous intravenous infusion of pancuronium (0.1 mg·kg−1·h−1).
After orotracheal intubation with an 8.0-mm (internal diameter) endotracheal tube, the dogs were mechanically ventilated using the intermittent positive pressure ventilation mode with an Evita 4 ventilator (Draeger, Lubeck, Germany) at a pressure limit of 40 cm H2O, tidal volume of 10 ml/kg, rate of 26/min, inspiratory-to-expiratory ratio of 1:1, inspiratory flow rate of 40 l/min, fractional concentration of inspired oxygen of 1.0, and PEEP of 0 cm H2O. An 18-gauge cannula was introduced into the femoral artery. Three electrodes for electrocardiography were placed in appropriate positions on the dog's trunk. The arterial catheter and the electrodes were connected to an Escort II pressure monitor (Medical Data Electronics, Arleta, CA) to monitor blood pressure and heart rate. The left external jugular vein was cut down for the insertion of a 7.0-French Swan-Ganz catheter (Baxter Healthcare Corporation, Irvine, CA), which was advanced to the pulmonary artery under the guidance of the pressure profile. The pulmonary artery catheter was connected to a COM-1 cardiac output computer (Baxter Healthcare Corporation).
Saline Lavage, ALI, and Determination of the LIP
After measurements of normal state had been made, lung lavage was performed three to four times using warmed (37°± 1°C) normal saline at a dose of 40 ml/kg, which was administered via the endotracheal tube and removed by gravity. ALI was deemed established if Pao2taken 90 min after the last lavage was 300 mmHg or less. At the establishment of ALI, the LIP of the inflational P–V curve was determined by the interruption method, as described previously. 26The pressure corresponding to the intersection of the starting compliance and the inflational compliance was defined as the LIP. The mean LIP ± SD in seven experimental animals was 10.7 ± 1.4 cm H2O. In the other three dogs in which the LIP was not easily discernible, 11 cm H2O was arbitrarily chosen for the LIP.
RM and PT
Our method of performing the RM 16was designed to gradually apply and withdraw a high distending pressure over a prolonged period (fig. 1): from the baseline level of 3 cm H2O, PEEP was changed in stepwise manner to 10, 20, 30, and 20 cm H2O, each step being 30 s (2 min in total duration). Because ventilation by the intermittent positive pressure ventilation mode was pressure limited at 40 cm H2O, the tidal volume at the phase of the highest PEEP (30 cm H2O) was accordingly reduced. After the RM, PEEP was reset at the determined LIP + 2 cm H2O. The PT was performed by simply increasing the level of PEEP from 3 cm H2O to the LIP + 2 cm H2O.
The two maneuvers were performed sequentially in random order with a baseline measurement in between. Five animals were treated with the RM first, whereas the others were treated with the PT first. Thirty minutes after the first maneuver, the dogs were disconnected from the ventilator circuit for 5 s so that the recruited lung resulting from the first maneuver could undergo passive collapse. The baseline ventilation with PEEP at 3 cm H2O was resumed, and baseline data were collected in a second round 10 min later.
Multidetector Row Computed Tomography and Morphometric Analysis
Computed tomography (CT) studies of the whole lung were performed with a multidetector Light Speed QX/i scanner (GE Medical System, Milwaukee, WI). Scanning parameters were 5-mm collimation with a high-speed mode (pitch = 6), 120 kV(peak), and 100 mA. Scanning of the whole lung was accomplished during both end-expiratory and end-inspiratory breath holdings, which ranged from 7.5 to 8.5 s depending on the size of the dog's lungs. As the number of pixels was 512 per field of view (25–30 cm), the area of a pixel ranged from 0.49 × 0.49 to 0.59 × 0.59 mm (0.24–0.35 mm2), and the volume of a voxel ranged from 0.49 × 0.49 × 5 to 0.59 × 0.59 × 5 mm (1.20–1.74 mm3).
Using an Advantage Workstation (GE Medical System), lung volumes were determined with a three-dimensional rendering method for the tissues corresponding to the attenuation from +100 Hounsfield units (HU) to −1,000 HU. Depending on the value of the HU, the lung was divided into atelectatic lung, poorly aerated lung, normal aerated lung, or hyperaerated lung, which are defined below. 27In assessing volume and regional attenuation, only the lung parenchyma was included by excluding the chest wall, mediastinum, large vessels, and airways using a roller ball of the computer. 24,28This process was done by two investigators (S.S.L. and J.S.L.). The concordance of their lung volume measurements was 97.6 ± 2.2%, and an average of their measurements was taken as data.
In addition to the whole-lung volume and compartment volumes, regional aeration of the lung was evaluated using the regional HU. Three levels of the lung were chosen: the cephalad top of the aortic arch (apical level), 1 cm above the dome of the diaphragm (caudal level), and the midpoint between the two levels (hilar level). For each level, the average of the pixels in a region of interest (1 cm2in area and centered 2 cm from the inner wall of the thorax) was determined in three planes: the anterior (ventral), middle, and posterior (dorsal) planes, which were equally divided in the vertical distance of the lung. In this way, 18 topographic regions from both lungs (nine per lung) were assessed in each animal. Because the amount of tissue would not change over this short period, the decrease in the HU after the maneuver was interpreted as representing an increase in aeration of the region due to a given maneuver. The following are the definitions used for the morphometric analysis of the lung:
end-expiratory lung volume (EELV), whole-lung (gas plus tissue) volume at end expiration
aerated EELV, EELV–nonaerated lung volume (VNA)
end-inspiratory lung volume, whole-lung volume at end inspiration
mean lung volume, (EELV + end-inspiratory lung volume)/2 taking the inspiratory-to-expiratory ratio of 1:1 into consideration
volume of gas, volume of gas was calculated based on the assumption that a lung area exhibiting −500 HU is made up of 50% of gas and 50% of tissue
volume of tissue, (total volume of lung − volume of gas)
VNA, volume of voxels exhibiting +100 to −100 HU
poorly aerated lung volume (VP), volume of voxels exhibiting −100 to −500 HU
normal aerated lung volume, volume of voxels exhibiting −500 to −900 HU
hyperaerated lung volume (VH), volume of voxels exhibiting −900 to −1,000 HU
lung recruitment, decrease in the amount of collapse-prone lung with maneuver, as calculated by (VNA+ VPof baseline at end expiration) − (VNA+ VPafter maneuver at end expiration)
change in regional aeration, (regional HU of before maneuver) − (regional HU of after maneuver)
heterogeneity of lung aeration, degree of variation of regional aeration, assessed by coefficient of variation (CV) of the HUs of 18 topographic lung regions
percent lung volume undergoing tidal recruitment, decrease in the amount of collapse-prone lung with tidal inspiration as calculated by ([VNA+ VPat end expiration]−[VNA+ VPat end inspiration])/mean lung volume
percent lung volume undergoing tidal hyperinflation, increase in the amount of VHwith tidal inspiration, as calculated by ([VHat end inspiration]−[VHat end expiration])/mean lung volume
CT scans, hemodynamics, and respiratory data for the dogs were obtained immediately before each maneuver (before), every 30 s during the RM or the first 2 min of the PT, and 30 s after each maneuver (after). Another set of data was obtained 30 min after each maneuver (delayed) to evaluate the effect of time on the lung morphometry.
Cardiac output was the average of three thermodilution measurements, determined with 10 ml normal saline at a temperature of 23°–25°C (computation constant, 0.595). Respiratory parameters (peak pressure, inspiratory pause pressure, and mean airway pressure) were taken directly from the digital display of the ventilator. Dynamic and static compliance of the respiratory system was calculated by tidal volume/(peak pressure − PEEP), and tidal volume/(inspiratory pause pressure − PEEP), respectively. In calculating compliances, the set PEEP was chosen instead of total PEEP because intrinsic PEEP was not observed in our model of ALI. Blood gas analysis was performed for simultaneous pairs of arterial blood and mixed venous blood. Shunt was calculated by (capillary oxygen content − arterial oxygen content)/(capillary oxygen content − mixed vein oxygen content) × 100.
All data are expressed as mean ± SD. Friedman analysis was performed to test the difference between different conditions, followed by the Wilcoxon signed rank test if significant differences were detected. The CV was calculated as SD/mean. The correlation between two variables was tested by linear regression analysis. Normality of distribution of a continuous variable was examined by the Kolmogorov–Smirnov test. P < 0.05 was considered statistically significant.
P–V Curve and Position of Tidal Ventilation Loop
Figure 2, above —constructed with the sequential points of EELV or gas-only volume and corresponding PEEP before, during the maneuver at 30-s intervals, and after the maneuver—shows both inflation and deflation curves with the RM, whereas only a smaller inflation curve is shown with the PT. The tidal ventilation loop, plotted with EELV, end-inspiratory lung volume, PEEP, and inspiratory pause pressure, was relocated higher after the RM than after the PT (fig. 2, below ). The tidal excursion of airway pressure was smaller after the RM than after the PT for the same tidal volume.
Figure 3show representative scans of a dog's lung with the RM and with the PT, respectively. Increased densities in the most dependent lung regions were noted after the PT. In contrast, such lesions were much less after the RM. Furthermore, the lung and thorax assumed a bigger and more rounded shape after the RM than after the PT.
Lung Recruitment, Tidal Recruitment–Hyperinflation, Total and Compartment Lung Volumes, and Gas-to-tissue Ratios of the Lung
VNA, VP, normal aerated lung volume, and VHin absolute amount and percent of total lung volume at baseline were similar between the maneuvers (table 1). Lung recruitment was greater with the RM than with the PT (190.3 ± 70.1 vs. 139.8 ± 72.7 ml, respectively;P = 0.035). The increase in the normal aerated lung volume was greater with the RM than with the PT (783.3 ± 191.3 vs. 549.7 ± 115.7 ml, respectively;P = 0.028). VHwas increased with both maneuvers (both, P = 0.018), and the increase tended to be greater with the RM than with the PT (138.8 ± 134.5 vs. 110.3 ± 104.6 ml, respectively;P = 0.075). As assessed by percent of total lung volume, percent VPafter the RM was smaller than that after the PT, whereas other compartments were similar between the two maneuvers.
Percent lung volume undergoing tidal recruitment decreased after the RM (6.8 ± 5.3% at baseline and 1.5 ± 2.1% after maneuver;P = 0.036), whereas percent lung volume undergoing tidal recruitment did not change after the PT (4.3 ± 5.2% and 2.6 ± 1.9%, respectively;P = 0.575). Percent lung volume undergoing tidal hyperinflation did not change with both maneuvers: RM, 2.9 ± 2.2% at baseline and 2.9 ± 1.3% after maneuver (P = 0.612); PT, 3.4 ± 2.0% at baseline and 3.0 ± 1.8% after maneuver (P = 0.128).
The EELV, which did not differ at baseline (P = 0.1), was greater after the RM than after the PT (P = 0.018) (fig. 2, above ). Gas-only volume of the lung, which did not differ at baseline (P = 0.483), was greater after the RM than after the PT (P = 0.012). Tissue volumes of the lung at baseline (524.3 ± 96.2 ml for RM and 516.0 ± 87.0 ml for PT;P = 0.499) and after the maneuver (574.8 ± 119.1 and 550.7 ± 108.2 ml, respectively;P = 0.136) did not differ between the RM and the PT. Tissue volume of the lung was not significantly changed by either the RM (P = 0.084) or the PT (P = 0.311). The gas-to-tissue ratio of the lung, which did not differ at baseline (0.77 ± 0.31 for RM and 0.76 ± 0.29 for PT;P = 0.865), was higher after the RM than after the PT (2.02 ± 0.36 and 1.64 ± 0.27, respectively;P = 0.011).
Regional Aeration of the Lung
Whereas the increase in regional aeration was similar at the apical and hilar levels with the two maneuvers, the increase in regional aeration at the caudal level was greater with the RM than with the PT (P = 0.026) (fig. 4, above ). Similarly, whereas the increase in regional aeration was similar in the ventral and middle planes with the two maneuvers, the increase in regional aeration in the dorsal plane was greater with the RM than with the PT (P = 0.006) (fig. 4, below ).
Heterogeneity of Overall Lung Aeration
The CV of the HU at end expiration was smaller after the RM than after the PT (0.13 ± 0.06 vs. 0.27 ± 0.11, respectively;P = 0.026). The CV of the HU at end inspiration was also smaller after the RM than after the PT (0.19 ± 0.19 vs. 0.29 ± 0.18, respectively;P = 0.028).
Physiologic Outcomes and Correlation between Lung Morphometry and Hemodynamics
Dynamic compliance (P = 0.021) and static compliance of the respiratory system (P = 0.011) were higher after the RM than after the PT (table 2). Shunt was lower after the RM than after the PT (P = 0.047). Paco2was increased with the RM, whereas not so with the PT. Cardiac output was lower (P = 0.017) and pulmonary artery occlusion pressure was higher (P = 0.038) after the RM than after the PT.
An increase in the EELV (fig. 5, above ) and an increase in aerated EELV (fig. 5, below ) induced by the RM or the PT both correlated with the maximum decrease in mean blood pressure recorded during both maneuvers.
Time Effects on Lung Morphometry and Physiology with Both Maneuvers
With the RM, the delayed EELV did not differ from the EELV after the maneuver (P = 0.128) (fig. 2, above ). In contrast, with the PT, the delayed EELV was greater than the EELV after the maneuver (P = 0.018). With both maneuvers, the delayed values for the gas-only volume of the lung were greater than the values after the maneuver (P = 0.035 for RM;P = 0.012 for PT). With the RM, the delayed CVs of the HU at end expiration (P = 0.1) or at end inspiration (P = 0.715) did not differ from the values after the maneuver. In contrast, with the PT, the delayed CVs of the HU both at end expiration (0.22 ± 0.10) and at end inspiration (0.22 ± 0.16) improved relative to the values after maneuver (P = 0.027 and P = 0.028, respectively).
Nevertheless, the delayed values for the EELV (P = 0.018) and gas-only volume with the RM (P = 0.012) were greater than those with the PT (fig. 2, above ). The delayed CVs of the HU with the RM at end expiration (0.13 ± 0.06) and at end inspiration (0.17 ± 0.16) were both less than those with the PT (P = 0.042 and P = 0.027, respectively). The delayed values for dynamic compliance (P = 0.013) and static compliance of the respiratory system (P = 0.011) with the RM were greater than those with the PT (table 2). The delayed value for shunt with the RM was lower than that with the PT (P = 0.037).
In this study, we demonstrated a difference between the RM and the PT (without the RM) on the radiologic morphometry of the lung in an animal model of ALI. In quantitative terms, the RM brought about a greater recruitment of the lung. In addition, lung volume (both EELV and gas-only volume) after the RM was greater than that after the PT, and this difference persisted for 30 min. However, our results also demonstrated the risk of lung overinflation with the RM, although the proportion of hyperaerated lung after the maneuver was similar between the two methods of recruitment.
A few important qualitative differences between the two methods of recruitment were also noted. The complex curve plotted for the RM demonstrated that the end-expiratory P–V of the lung was transferred onto a deflation curve within the short time frame of the maneuver (2 min). With the PT, however, the end-expiratory P–V of the lung was shifted along a simple shorter line, and the lung inflation appeared to cease prematurely. Consequently, more lung units remained open after the RM than after the PT at equal pressures in the range of (LIP + 2 cm H2O) to inspiratory pause pressure. Our findings agree with those of Rimensberger et al. 20who showed that a ventilatory strategy without an RM fails to relocate the tidal ventilation cycle onto the deflation limb of the P–V envelope. The increase in lung volume after the PT, in fact, represents the alveoli recruited by virtue of the augmented inflation pressure built upon an elevated baseline pressure. 8,29In view of the growing recognition that recruitment and the prevention of derecruitment are two distinct phenomena, 12,13,20the indiscriminate use of PEEP as equivalent to “recruitment” no longer seems valid. Rather, because PEEP is an end-expiratory pressure, it should be regarded as an “antiderecruitment pressure.”
The change in regional aeration was more revealing of the effects of the RM relative to the PT. The increase in regional aeration (decrease in the HU) induced by the two methods did not differ at the apical or hilar levels of the lung or in the ventral or middle planes of the lung. However, the increase in regional aeration in the dorsocaudal region (the most dependent portion of the lung in our animal) was greater with the RM than with the PT. The clinical implications of these findings are twofold: an augmented inflation pressure built upon an increase in PEEP could work for certain lung regions with relatively lower opening pressure, 8,10and an RM that employs a higher distending pressure than a PT may be necessary for the conversion of the most dependent lung regions into functioning units.
The CVs for the regional HU in our study suggest that the lung aeration, at a static condition, was more homogeneous along the gravitational (sternovertebral) and nongravitational (cephalocaudal) axes after the RM than after the PT. Improved homogeneity of aeration in a lung with ALI should translate into at least two physiologic advantages. First, oxygenation may improve through a better ventilation–perfusion match. Second, evenly distributed aeration will alleviate the elastic interdependence force impinging on collapsed lung units. 30The latter aspect, not assessed in the present study, might be worth investigating to determine whether an RM can reduce ventilator-associated lung injury more than a PEEP adjustment without an RM.
After the RM, the proportion of the lung subjected to tidal recruitment decreased from the baseline, suggesting that repeated phasic alveolar collapse could be less after the RM. However, such a decrease in tidal recruitment was not observed with the PT. In view of the importance of shearing damage in the mechanism of ventilator-associated lung injury, 31this result supports the RM as a necessary component of lung-protective ventilation.
Interestingly, the delayed (30 min later) lung volume after the PT increased significantly relative to the lung volume immediately after the PT (30 s), which contrasts with the stability of lung volume after the RM over the same period. This phenomenon may be explained by the “avalanche effect,”i.e. , the subsequent opening of collapsed lung units may have been facilitated by the elastic interdependence of adjacent recruited units. 30,32Nevertheless, both EELV and gas-only volume after the PT did not reach the values observed after the RM at least during our study period. Consistent with this, the delayed values of physiologic parameters, such as compliance and shunt, were still more favorable with the RM.
Considering the qualitative and quantitative superiority of the RM over the PT in our study, it might be insufficient to use the inflation limb of the P–V curve to optimize recruitment in the ARDS lung. The LIP is inaccurate in predicting optimal recruitability. 9,11,12First, once tidal ventilation is resumed after P–V plotting, the tidal P–V loop will no longer remain on the same previous inflation limb. 20Second, the phenomenon of recruitment may continue above the LIP. 8–11Third, the LIP may represent the opening pressure of the small airways, whereas true recruitment of the collapsed alveoli requires a much higher pressure than required for opening small airways. 5,8,10In this study, the difference between the two maneuvers also obviously involved the inflation pressure imposed on the lung over the first 2 min of the maneuver: the inflation pressure of the RM was 40 cm H2O, whereas that of the PT to the LIP peaked around 25 cm H2O. In view of recent data, a recruiting pressure as high as 60–70 cm H2O may be necessary for alveolar opening. 9,22,23
Unlike the pressure factor, the time factor for the RM has not received much attention until recently. In a pioneering study reported 51 yr ago, Day et al. 15showed that a certain time as well as a minimum threshold pressure was necessary to reverse atelectasis of the lung. Interestingly, they found that the (pressure × time) function for satisfactory inflation was similar across a variety of animal species. The time adopted for the RM has varied from 10 s to 2 min depending on the investigator. 1,9,15,18,19Further studies are required to define the optimum duration for the application of high pressure in patients with ARDS to avoid jeopardizing hemodynamic or respiratory stability. The lack of change in lung volume over the 30 min after the RM suggests that the 2-min period used in our protocol may have been sufficient, at least for our dogs and our pressure strategy.
It is noteworthy that the increase in the EELV correlated with the maximal decrease in blood pressure seen during the 2-min period of intervention. This phenomenon might reflect the greater transmission of airway pressure to the pleura as the lung becomes more compliant. 33The stronger correlation of blood pressure with the “aerated EELV” (lung volume exclusive of nonaerated volume) than with the total EELV in our study suggests that transmission of airway pressure to the pleura occurred through “functioning” lung units. In view of the lower cardiac output, the higher pulmonary artery pressure after the RM than after the PT is more likely to be a spurious result and lends support to the concept described above. Therefore, the larger the recruitment, the greater the decrease in blood pressure during the maneuver that should be anticipated. This could be a drawback of the RM at the bedside although transient in duration, because systemic oxygen delivery (arterial oxygen content × cardiac output) could be compromised despite improved oxygenation of the blood in the lung.
CT has contributed significantly to the advances of the concept and treatment of ALI or ARDS. Previous CT studies on ALI or ARDS, however, mostly involved a single (usually juxtadiaphragmatic) region or a few slices of the lung because of the technical difficulties associated with whole-lung scanning. 7,8,28The validity of lung-morphometric data determined on a single or even three CT sections has been seriously questioned, because such limited data have been shown to be biased and correlated poorly with data from the whole lung. 34The main reason for the discrepancy arises from the inhomogeneity of lung injury along the cephalocaudal axis of the lung. The present study exploited the CT data from the whole lung and analyzed both the cephalocaudal axis and the sternovertebral axis. In previous studies, 10,27in which the whole lung was scanned, the scanning necessitated an apnea of 15–20 s in the subjects, which may be too long to be acceptable in a clinical setting. The present study used an advanced CT technique, i.e. , a multidetector row scanner, which allowed whole-lung scanning in less than one half of the respiration hold of the study by these investigators. Setting aside the concern of radiation hazard, this CT technique with a shorter period of apnea might be more useful in the clinical situation for ARDS than previous CT techniques. In our study, we considered lung recruitment as the conversion of collapse-prone lung (the composite volume of nonaerated lung and poorly aerated lung) into collapse-resistant lung (normal aerated lung and hyperaerated lung), not merely as the increase in gas volume into these regions as defined in previous studies. 8,24Because “tissue” in these lung regions becomes the “substance” for shearing damage in a collapse-prone lung, 35–37salvaging this composite volume inclusive of tissue may be a clinically relevant concept in ARDS or ALI.
Obviously, lung injury induced by saline lavage does not represent all the possible pathophysiologic diversity of human ARDS. Moreover, our results may not be reproducible in ARDS at a more advanced stage, in which compressive atelectasis is considered to be less extensive than in the early stage of the disease. 38Even in the present study with an acute form of lung injury, the volume of hyperaerated lung was significantly increased with both RM and PT. This finding is in line with the recent CT report on ARDS patients 7,24,25and indicates that alveolar recruitment employing a high proximal airway pressure may occur at the expense of hyperaeration in some regions of the lung. Therefore, the aforementioned effects of the RM as compared with the PT must be carefully weighed against this complication in clinical application. The optimal strategies for determining recruiting pressure and its duration are as yet unclear. It has been suggested that the end-expiratory pressure should be higher than the prerecruitment level to preserve the immediate result of recruitment. 18,21,39Ideally, a point at which the critical closing pressure is most concentrated needs to be determined for the application of holding pressure (PEEP) that best preserves recruitment. 12,13Our data on the relationship between lung volume and hemodynamics are incomplete in that blood pressure is influenced by factors other than left ventricular preload. A continuous measurement of stroke volume and systemic and pulmonary vascular resistance during the RM are necessary to better understand the interaction between the changes in lung volume and hemodynamics that occur with the maneuver. The increase in Paco2after the RM could have been related to a compromised tidal volume during the RM associated with the ventilation mode (pressure-limited mode) and increased wasted ventilation from the increased volume of hyperaerated lung and/or decreased cardiac output compared with the baseline. However, because mixed expired Pco2or ventilation–perfusion distribution was not determined in the animals, the mechanisms of this phenomenon cannot be ascertained for the present study. The lower cardiac output after the RM or PT, especially cardiac output at 30 s, compared with baseline could have contributed to the decrease of shunt. 40In view of the stability of shunt over time with the RM and similarity of cardiac output between both maneuvers at 30 min, however, the difference of shunt between maneuvers was more attributable to the quantitative or qualitative difference in morphologic change of the lung.
In conclusion, in a canine saline-lavaged lung, the lung-morphometric effects of a RM were different from those of a PT to the LIP, in both qualitative and quantitative terms. Compared with the PT, the RM resulted in a greater functioning lung volume, better aeration of the most dependent lung, and less regional heterogeneity of lung aeration. However, the RM tended to induce a greater increase in hyperaerated lung volume than did the PT.
The authors thank Sung-Joon Yang (a medical student at the University of Ulsan College of Medicine, Seoul, Korea) for his help with volumetric analysis of the dogs’ lungs and Robb W. Glenny, M.D. (Associate Professor at the Division of Pulmonary and Critical Care, University of Washington, Seattle, Washington), for his constructive comments on the manuscript.