In a previous study (Allen G, Lundblad LK, Parsons P, and Bates JH. J Appl Physiol 93: 1709-1715, 2002), our laboratory used deep inflations (DI) in mice to show that recruitment of closed lung units can be a very transient phenomenon in lung injury. The purpose of this study was to investigate how this transience of lung recruitment depends on the nature and degree of acute lung injury. Mice were administered 50 μl of either saline (n = 8), 0.01 M (n = 9) or 0.025 M (n = 8) hydrochloric acid, or 50 μg (n = 10) or 150 μg (n = 6) of LPS and were mechanically ventilated 24-48 h later. At various levels of positive end-expiratory pressure, two DIs were delivered, and forced oscillations were used to obtain a measure of lung stiffness (H) periodically over 7 min. After LPS exposure, pressure-volume curve hysteresis and recovery in H after DI were no different from saline-exposed controls despite 500 times more neutrophils in bronchoalveolar lavage fluid. Pressure-volume hysteresis and recovery in H were increased in acid-exposed mice (P < 0.001) and were correlated with bronchoalveolar lavage fluid protein content (R = 0.81). Positive end-expiratory pressure reduced recovery in H in all groups (P < 0.01) but reduced pressure-volume hysteresis in the acid-injured groups only (P < 0.001). We conclude that the effects of DIs in acute lung injury depend on the degree of lung injury but only to the extent that this injury reflects a disruption of the air-liquid interface.
- acute lung injury
- lung stiffness
acute lung injury (ALI) continues to be a major cause of morbidity and mortality in the intensive care unit (37, 38, 48). Although ALI may have many different etiologies (10, 13), its physiological derangements invariably manifest as decreased lung compliance and expanded pressure-volume (PV) hysteresis (6, 35). To some degree, these mechanical abnormalities in ALI can be reversed by a recruitment maneuver (RM) designed to reopen collapsed regions of the lung, because when collapsed regions of the lung are recruited, the increased surface area should result in a decrease in overall lung elastance. However, the effect is often transient. In a previous study using mice, we demonstrated that even normal lungs progressively regain elastance (H) after recruitment with a deep inflation (DI) (1). When the lungs were lavaged with saline, the recovery of H to plateau was so rapid that RMs would have to be given several times per minute to achieve a meaningful increase in the degree of mean open lung (1). Previous studies using different animal models of ALI, however, have demonstrated a wide range in the effect of RMs on lung mechanics and gas exchange (27, 29). Furthermore, among human subjects, the response to RMs has been shown to be more robust in patients with ALI of a nonpulmonary vs. primary pulmonary source (14, 41) and at an earlier stage of development (17). These findings suggest that varying types and degrees of ALI may exhibit different dynamics of progressive alveolar closure after a RM.
We speculated that closure should occur most rapidly in lung injury associated with alveolar flooding because free alveolar fluid would be expected to reobstruct airspaces almost immediately after being displaced by a DI. Occasional DIs in this situation would thus be essentially futile. Conversely, in lung injury that primarily manifests as neutrophilic infiltrate, closure might be less rapid after a DI. We would expect some degree of efficacy for occasional DIs in this type of ALI. In any case, there is clearly a need to understand how the dynamics of derecruitment after a RM differ in various types and degrees of ALI. The purpose of this study was, therefore, to investigate how the dynamics of recruitment and derecruitment after a DI depend on the nature and degree of ALI.
Animal preparation. We studied 8- to 9-wk-old BALB/c female mice (Jackson Laboratories, Bar Harbor, ME). Before lung mechanics measurements, mice were assigned to one of five groups. Mice were anesthetized with an intraperitoneal injection of avertin [tribromoethyl alcohol (Aldrich, Milwaukee, WI) in tert-amyl alcohol, 20 mg/ml] at a dose of 400 mg/kg. We used two accepted methods for inducing ALI in animals. The first method involved nasal instillation of LPS. In one group of 10 mice, we administered a low dose of LPS (Escherichia coli 0111:B4, Sigma Laboratories) in two separate 25-μl (1 mg/ml) aliquots delivered intranasally 48 and 24 h before measurement of lung mechanics (total dose = 50 μg). In a separate group of six mice, equal volumes of more highly concentrated LPS (3 mg/ml) were given at the same time points for a total dose of 150 μg. Phosphate-buffered normal saline was delivered in equal volumes at the same time points in eight anesthetized control mice. For the second method of inducing ALI, 50 μl of either 0.01 M (n = 9) or 0.025 M (n = 8) hydrochloric acid (HCl) was instilled by transilluminating the neck and intubating the trachea with a 22-gauge bulb-tipped steel feeding tube. All mice were observed and kept warm until they recovered from anesthesia.
On initiation of the lung mechanics measurements, mice in all experimental groups were anesthetized with an intraperitoneal injection of pentobarbital at 90 mg/kg. They then underwent tracheostomy with a secured 18-gauge metal cannula, were connected to a flexiVent (SCIREC, Montreal, Canada) computer-controlled small animal ventilator, and were ventilated in a quasisinusoidal fashion at a rate of 200 breaths/min. The cylinder piston displacement was set at 0.25 ml, which resulted in tidal volumes of 0.20 ml (∼10 ml/kg) when gas compression was accounted for. Positive end-expiratory pressure (PEEP) was controlled by submerging the expiratory limb from the ventilator into a water trap. The animals were paralyzed with an intraperitoneal injection of pancuronium bromide (0.5 ml/kg) and allowed 5 min to adjust to the ventilator at a PEEP of 3 cmH2O. To ensure adequate anesthesia, heart rate was monitored by continuous electrocardiogram, which was measured via transcutaneous needle electrodes. Halfway through the protocol, an additional dose of pentobarbital (30 mg/kg) was administered for maintenance of deep anesthesia. These studies were approved by the Institutional Animal Care and Use Committee of the University of Vermont and were carried out in accordance with the Guiding Principles in the Care and Use of Animals of the American Physiological Society.
Experimental protocol. After the initial stabilization period, the level of PEEP was set at 1 cmH2O, and two DIs were delivered at constant flow with a pressure limit of 25 cmH2O. Each DI lasted 2 s. The mice were then returned to quasisinusoidal ventilation at 200 breaths/min with a tidal volume of 0.20 ml. Respiratory system input impedance (Zrs) was measured via a forced-oscillation technique (described in Data analysis) 4 s after the two DIs, then subsequently every 15 s for 5 min, and then every 30 s for an additional 2 min. The entire protocol was timed by a computer and repeated at a PEEP of 3 and 6 cmH2O. The order of PEEP was not randomized to minimize the possible damaging effects of the higher PEEP levels on measurements at lower PEEP.
Bronchoalveolar lavage and tissue collection. After completion of the protocol, the mice were killed with a 200 mg/kg ip injection of pentobarbital sodium. At the onset of asystole, the mice were disconnected from the ventilator, and bronchoalveolar lavage fluid (BALF) was obtained by instilling 0.8 ml of phosphate-buffered saline into the tracheal cannula and slowly suctioning back for a return of ∼0.6 ml in every animal. After bronchoalveolar lavage, the lungs were surgically removed en bloc, instilled with 10% buffered formalin to a pressure of 30 cmH2O, then embedded in paraffin, cut, mounted, and stained with hematoxylin and eosin. In some mice, bronchoalveolar lavage was omitted in favor of obtaining less adulterated histological specimens.
BALF analysis. Cell counts were determined manually with a hemacytometer, and differentials were calculated from fixed, hematoxylin and eosin-stained cytospun slides. Protein content was calculated by using a colorimetric assay (Bio-Rad Laboratories, Hercules, CA), standardized to graded concentrations of bovine serum albumin.
Data analysis. Zrs was determined by measuring piston volume displacement and pressure in the ventilator cylinder while 2-s oscillatory volume perturbations were delivered to the airway opening. These perturbations were composed of 13 superimposed sine waves of varying amplitude and frequency, ranging from 1 to 20.5 Hz. The frequencies were set at mutually primed values to reduce harmonic distortion that can occur in nonlinear systems (18). Before beginning the protocol, we obtained dynamic calibration signals necessary to correct for the physical characteristics of the flexiVent in subsequent measurements of Zrs (20, 43). Zrs itself was determined via Fourier transform from the signals of ventilator piston volume and cylinder pressure as described previously (16, 20). Zrs was interpreted by being fit with the model 1 where 2 where Raw is airway resistance, i is √-1, f is frequency (Hz), Iaw is airway inertance, and G and H characterize the dissipative and elastic properties of the lung tissues (18), respectively. In the present study, we focused our attention on H, which is essentially the conventional elastance of the respiratory system (indeed, it is precisely equal to respiratory elastance at an oscillation frequency of 1/2π Hz). We followed H as a traditional surrogate for the amount of open lung, expecting it to decrease with alveolar recruitment and increase with alveolar derecruitment. We thus obtained a set of H values vs. time for 7 min after the DIs in each mouse at each level of PEEP. Because each mouse had varying surgical times and depths of anesthesia, we gave two DIs immediately before data collection to standardize volume history. Thus H assumed its lowest value after DI and increased as time progressed. In separate pilot experiments, this protocol was repeated immediately at the same level of PEEP (data not shown) and gave the same time course in H, demonstrating the effects of the DI to be reproducible.
In a subset of animals from each experimental group, PV curves were obtained (saline, n = 8; 50 μg of LPS, n = 6; 150 μg of LPS, n = 6; 0.01 M HCL, n = 8; 0.025 M HCl, n = 8). Starting at functional residual capacity, seven steps of inspiratory volume were delivered to a total volume of 0.8 ml, followed by seven equal expiratory steps, pausing at each step for 1 s. Plateau cylinder pressure was measured during each pause and plotted against piston displacement from the flexiVent to obtain a quasistatic PV curve (corrected for gas compression and slow leak). This was performed after 7 min of ventilation at a PEEP of 1, 3, and 6 cmH2O.
All graphing and statistical analyses were performed by using Origin (version 7.03, Origin Laboratory, Northampton, MA) software. Time constants were fit to the data from acid-exposed mice by using a multiple-regression subprogram within Origin software. When grouped data (at all PEEP levels) were compared between experimental groups, repeated-measures ANOVA was used to examine the overall effects of PEEP, LPS, or acid exposure on baseline and final values for H, total recovery in H after DI, and PV hysteresis. Bonferroni means comparison tests were used when the means between individual groups were compared after ANOVA. Differences were considered significant when P < 0.05. When differences between individual levels of PEEP within the same group were compared, paired t-tests were used. When between-group differences at a specific level of PEEP were compared, unpaired t-tests were used.
Endotoxin ALI. Cell counts from BALF were significantly greater in the LPS groups (2.5 × 106 ± 4.3 × 105 and 3.8 × 106 ± 6.2 × 105 cells/ml for the 50- and 150-μg groups, respectively) compared with the control group (2.9 × 105 ± 3.0 × 104 cells/ml; P < 0.001). Differential analysis revealed neutrophil predominance (80%) in the LPS group vs. macrophage predominance (98%) in the controls, yielding a 500- to 900-times greater mean absolute BALF neutrophil count in the LPS-exposed groups (2.0 × 106 ± 4.0 × 105 and 3.5 × 106 ± 5.9 × 105 cells/ml for the 50- and 150-μg groups, respectively, vs. 3.8 × 103 ± 9.8 × 102 cells/ml; P < 0.001) (see Fig. 1). Light microscopic examination of hematoxylin and eosin sections from LPS-exposed mice (compared with saline controls) revealed a patchy distribution of injury that consisted of abundant neutrophil accumulation within the interstitium, surrounding the airways, and within the alveolar spaces. The alveoli, however, had relatively well spared architecture when compared with those of saline-exposed controls. In areas of injury, the alveoli exhibited slight expansion of the interstitium, but no alveolar fluid or hyaline membranes could be identified, suggesting mild interstitial but negligible alveolar edema (Fig. 2, A and B).
With respect to lung mechanics measurements, the total recovery in H over the 7 min after DI was lower with added PEEP in both the LPS and saline-exposed groups (P < 0.001). Initial and final values for H were not significantly different between the groups at any level of PEEP (Fig. 3). The total recovery in H after DI was no different between the 150-μg LPS-exposed and saline-exposed controls. Interestingly, both groups exhibited a greater recovery in H than the naive controls of our previous study (1), indicating that simple nasal instillation of fluid into the lungs increases the propensity for closure, presumably because of addition of fluid to the airway lining. In the 50-μg LPS-exposed mice, however, the total recovery in H after DI was slightly yet significantly increased compared with controls (P < 0.01), but when each level of PEEP was examined separately, recovery in H was higher only at a PEEP of 3 cmH2O (P < 0.01). Additionally, the ratio of G/H (i.e., hysteresivity), which is traditionally used as an index of airways and tissue heterogeneity (25, 32), did not change to any greater degree throughout the protocol in either LPS-exposed group, at any level of PEEP, when compared with controls (Fig. 4, data from 150-μg group not shown but overlapped that of 50-μg group). Newtonian Raw was no different between the saline and LPS groups at the beginning or end of the protocol, at any level of PEEP (data not shown). Analysis of the PV curves from both the 50- and 150-μg LPS-exposed groups failed to demonstrate any greater degree of hysteresis relative to those from control mice, at any level of PEEP (Fig. 5, curves from 150-μg group are not shown but overlapped those of 50-μg LPS and control groups).
Acid-induced ALI. Overall, both the 0.025 and 0.01 M concentrations of acid exposure significantly increased the total recovery in H after DI relative to saline-exposed controls (P < 0.001 and P < 0.01, respectively; Fig. 6). When both groups were analyzed together, incremental increases in PEEP significantly reduced the total recovery in H (P < 0.01). When each level of PEEP was examined separately, the total recovery in H in the 0.025 M HCl-injured mice was greater than that of controls at a PEEP of 1 and 3 cmH2O (unpaired t-tests, P < 0.05) but not at a PEEP of 6 cmH2O(P = 0.09). In contrast, the
0.01 M HCl-injured mice exhibited a greater recovery in H than did controls only at a PEEP of 1 cmH2O (P < 0.05). The rates of rise in H also decreased with increasing PEEP (Fig. 6); time constants of recovery were 37, 50, and 163 s, respectively, for PEEP of 1, 3, and 6 cmH2O in the 0.25 M acid group. G/H did not change to any greater degree throughout the protocol (Fig. 4) at any level of PEEP in the 0.1 M HCl-injured mice and only dropped significantly in the 0.025 M HCl-injured mice at a PEEP of 1 cmH2O when compared with the other groups (P < 0.05). Raw was only significantly elevated in the 0.025 M HCl-injured group when compared with the other groups at the beginning and end of the protocol at a PEEP of 1 cmH2O only (P < 0.05) (data not shown). PV curves from acid-injured mice demonstrated expanded hysteresis when compared with those obtained from LPS- and saline-exposed mice (P < 0.0001) (Fig. 5). In the 0.01 M HCl-exposed mice, the deflation limb of the PV curve followed that of the controls at all levels of PEEP, and the expansion in hysteresis (rightward shift of the inflation limb) was partially mitigated at a PEEP of 3 cmH2O and completely mitigated at a PEEP of 6 cmH2O (Fig. 5). In contrast, the deflation limbs of the PV curves obtained from mice exposed to the more concentrated acid (0.025 M) were shifted to the right, and the expansion in hysteresis (compared to controls) persisted at all levels of PEEP.
The BALF from the acid-injured mice demonstrated slightly higher total cell counts in the 0.025 M HCL group (Fig. 1), when compared with saline-exposed controls (P < 0.05), and demonstrated ∼60- and 90-fold greater neutrophil counts (0.01 and 0.025 M HCl, respectively) (P < 0.05). On histological examination (Fig. 2, C and D), hemorrhagic intra-alveolar edema and hyaline membranes (blue arrow) were noted, and significant distortion in alveolar duct size and shape could be appreciated, suggesting disruption of the air-liquid interface. Analysis of BALF from acid-injured mice revealed a dose-dependent increase in protein (Fig. 7), with almost an entire order of magnitude greater concentration of protein in the BALF from 0.025 M acid-injured mice (1,376 ± 223 μg/ml) when compared with controls (190 ± 10 μg/ml) (P < 0.001). Protein in BALF from the LPS-treated groups (368 ± 28 and 368 ± 72 μg/ml in the 50- and 150-μg LPS-exposed groups, respectively) was only about twice that of the controls (P < 0.001; Fig. 7). With the use of mice from which all the data were available, BALF protein levels were plotted against both total recovery in H (Fig. 8A) and the area of the PV curves (Fig. 8B) after ventilation at a PEEP of 1 cmH2O. This data was not always available from all mice, particularly in the 50-μg LPS-exposed group, but controls and LPS-exposed mice tended to cluster toward the bottom, with acid-injured mice driving the majority of the regression. When plotted individually, protein levels from BALF were found to correlate with the total recovery in H (Fig. 8A) and the area of the PV curves (Fig. 8B) after ventilation at a PEEP of 1 cmH2O. This was also the case between integrated PV area and BALF protein at a PEEP of 3 cmH2O (R = 0.85) and to a lesser extent at a PEEP of 6 cmH2O (R = 0.80) (data not shown).
In this study, we compared the lung mechanical manifestations of two important and widely used models of ALI. The endotoxin injury was utilized for its well-known ability to invoke a profound inflammatory response, characterized by accumulation of neutrophils, protein, and proinflammatory cytokines within the airways (5, 8, 9, 47, 49). The acid injury was used for its potential to cause early and direct injury to the alveolar epithelium with alveolar edema, and a delayed second phase involving neutrophil accumulation (26, 28). This study clearly demonstrates that the dynamics of lung mechanical function after DI recruitment are strongly dependent on the type and degree of ALI sustained (Figs. 3 and 6). However, an unexpected result was the lack of any mechanical effect in the LPS mice (Fig. 3), despite their high accumulation of lung neutrophils (Figs. 1 and 2). The ubiquity of decreased compliance in human patients with ALI would suggest that a derangement of lung mechanics should be required of any animal model of this condition. Does this mean that our LPS animals did not have ALI even though their lungs were clearly inflamed? We are not sure how to answer this question at present, so for the time being we will continue to treat the LPS mice as just another model of ALI. Nevertheless, this is an issue that will no doubt continue to surface as the pathophysiology of ALI becomes more precisely defined.
The degree of recovery in H over the 7-min protocol period after DI was inversely related to PEEP in both the acid and endotoxin models of ALI (Figs. 3 and 6). Certainly, a higher degree of PEEP could increase the amount of recruited lung before and during DI and thus lower the initial value for H after DI. However, the slower rise in H after DI at higher PEEP supports the notion that PEEP is also able to retard alveolar derecruitment in the injured lung and is consistent with the findings of other investigators who observed a PEEP dependence in compliance over time by using less frequently sampled data points (7, 50). This phenomenon is also consistent with a theoretical model recently published by Bates and Irvin (3). Admittedly, we have thus far equated the rise in H with alveolar derecruitment when other forces such as stress adaptation within the tissues (22), surfactant dispersal, and increasing airway and tissue heterogeneities after DI may be contributing to such mechanical changes. Interestingly, hysteresivity (Fig. 4) was somewhat increased for the injured mice, which could mean either increased heterogeneity or changes in intrinsic tissue properties, we cannot say which. However, the effects were rather small, and hysteresivity actually decreased as H increased, the opposite of what would be expected if heterogeneity were playing a major role in the mechanical changes we observed (32). Raw was also unchanged between groups (except in the 0.025 M HCl-injured group at a PEEP of 1 cmH2O), discounting airway heterogeneity as a significant contributing factor. Furthermore, because stretch is known to stimulate surfactant secretion (51) and may lead to its more effective dispersion, the gradual loss of surfactant from the air-liquid interface (21) might also explain these changes. Nevertheless, the scientific community has focused heavily on elastance as an index of recruitment (24, 33, 42), and it is difficult to imagine that derecruitment is not the primary mechanism responsible for these findings.
Interestingly, the nature of the post-DI time course in H differed between the two types of ALI. H increased approximately linearly in the saline and LPS-exposed groups (Fig. 3), whereas in the acid-exposed mice the time course of H was more exponential in character (Fig. 6), as our laboratory found previously in naïve and saline-lavaged mice (1). These differences suggest that the underlying mechanisms governing derecruitment may vary depending on how ALI is produced. For example, in the inflammatory LPS model, there was marked neutrophil accumulation within the distal airways but no evidence of alveolar flooding (Fig. 2B), whereas in the HCl group there was an accumulation of protein-rich exudates in the airways. In saline-lavaged animals, the airways contained a substantial volume of extra fluid. Thus it could be that derecruitment follows a rapid quasiexponential time course in situations involving alveolar flooding, whereas when lungs simply have an enhanced level of airway inflammation, they derecruit at a much slower rate that does not have time to approach its plateau over only 7 min. Admittedly, the time courses in human patients could be on a different scale than those demonstrated in our mouse models, possibly due to such factors as increased gravitational forces, a greater number of airway generations, and differences in respiratory rate. However, the central point regarding the differing behavior of derecruitment in different types of ALI seems less likely to be influenced by these factors than the actual time signatures.
Another important finding of our study was that the effect of PEEP on derecruitment dynamics depended heavily on the degree of injury. This was demonstrated by the response of both the PV hysteresis (Fig. 5) and the recovery in H (Fig. 6) to PEEP in the two groups of acid-induced ALI. PV curve hysteresis and total recovery in H in the 0.01 M acid group approximated that of controls after PEEP was changed from 1 to 3 cmH2O. Conversely, although PEEP reduced both PV hysteresis and recovery in H in the 0.025 M group, they remained elevated compared with saline controls, even at the highest PEEP tested (6 cmH2O). These results suggest that the “ideal PEEP” after DI in the 0.01 M acid-exposed mice was between 1 and 3 cmH2O but was >6 cmH2O in the 0.025 M acid group. Indeed, the lower inflection point (LIP) on the inflation limb of the PV curve obtained in the 0.025 M acid group lay between 10 and 15 cmH2O (see Fig. 5), suggesting that PEEP should be set at least at this level. Admittedly, the pressure limit of 25 cmH2O may have prevented sufficient DI in the most injured mice (0.025 M HCl), perhaps explaining the higher baseline H after DI in this group. However, this should not have changed the effect of PEEP on the overall change in H over 7 min. Additionally, this limit was designed to help prevent overinflation when set volumes were applied on top of different baseline resting PEEP levels, allowing for less interference of total inflation when different levels of PEEP were compared.
We estimated the position of the LIP in the 0.025 M HCl animals at each level of PEEP (Fig. 5) by using the traditional technique of continuing, by eye, the linear portions of the upper and lower segments of the inflation limb of the PV curve. The point where these lines intersect defines the LIP (2). The LIP obtained in this way shifted leftward with increases in the level of PEEP at which the mouse was ventilated just before the PV curves were obtained, despite all the curves being started at a PEEP of 0 cmH2O. If one subscribes to the theory that the inspiratory limb of the PV curve reflects a spectrum of critical opening pressures (19), this suggests that these pressures were affected by ventilation history. One potential explanation for this is that PEEP influences the dispersal of surfactant at the air-liquid interface (50), thereby affecting surface tension. However, an alternative explanation is that since the PV curves were obtained within seconds of removing the preceding PEEP, there was simply more open lung at the initiation of the PV curve when following ventilation at higher PEEP, in support of the model of Bates and Irvin (3). A prior study in our laboratory (31) used body plethysmography to demonstrate higher thoracic gas volume at higher PEEP in support of this explanation. Regardless of the mechanism responsible, however, our findings illustrate the importance of ventilation history on measurements of lung mechanics in ALI and may explain why it is often difficult to define the ideal PEEP simply as the LIP from a quasistatic PV curve (33, 36, 41).
Unexpectedly, LPS had no effect on respiratory mechanics, either with regard to initial and final values of H or total change in H, except to a minimal but statistically significant degree at a PEEP of 3 cmH2O in the 50-μg group. Furthermore, the PV curves from the LPS-exposed mice failed to exhibit any greater degree of hysteresis than the control mice at any level of PEEP. Despite these similarities in mechanics, the BALF of the LPS-exposed mice had 10-fold greater cell counts, 500- to 900-fold greater neutrophil counts, and 2-fold greater protein content than controls. These results demonstrate that neutrophil recruitment alone is not enough to alter lung mechanics or increase the rate of alveolar derecruitment. This raises an interesting question regarding how to best define the presence of lung injury in animal models.
Clinically, ALI is defined by the presence of poor oxygenation and bilateral airspace filling disease of a noncardiogenic origin (4), and neutrophils are felt to play a key role in its pathogenesis (48). Neutrophil recruitment is also thought to play a crucial role in both acid (12, 15, 28) and endotoxin-induced (5, 8) animal models of ALI. Pathology in these models is often attenuated, but not entirely circumvented, by neutrophil depletion. However, the recruitment of neutrophils has been shown in other experiments to occur without any significant change in epithelial or endothelial permeability (34, 40, 49). Also, other studies have shown that protein leakage and edema does occur after neutrophil recruitment, but not until the later stages of injury (8, 23), so it could be that it is not the neutrophils per se but rather the eventual effects they induce on the air-liquid interface that cause the mechanical consequences of ALI. On the other hand, other investigators using nebulized LPS have shown derangements in lung mechanics in mice at identical time points as in our study (11, 23). Possible explanations for this discrepancy include differences in response to LPS between different strains of mice (30), as well as differences in the volume of LPS actually reaching the lungs from intranasal instillation vs. the aerosol delivery employed by these other investigators. There is no question that intranasal delivery of LPS in mice results in a significant dose reaching the lungs, because it has been successfully employed by others to elicit increases in BALF protein and neutrophils (8) and was recently validated by using a 99mtechnicium-labeled sulfide colloid (45). Also, the BALF and histology specimens in the present study confirm that we successfully delivered LPS to the lungs via intranasal instillation. It might be, therefore, that the aerosol delivery method results in a different dose or distribution of LPS.
Our data thus suggest that the absence of changes in lung mechanics after neutrophil recruitment by LPS in our model was due to a lack of interference with surfactant function (23, 39). Evidence supporting this suspicion can be found on closer examination of the mean BALF protein levels from each of the experimental groups. Whereas mean BALF protein concentration in both LPS-exposed groups was significantly higher than that of saline controls, it was significantly less than levels observed in the acid-injury groups (Fig. 7). Furthermore, the wide variation in the degree of lung injury among the acid-exposed mice revealed a correlation between the level of BALF protein and both the degree of PV hysteresis and the total recovery in H at a PEEP of 1 cmH2O (Fig. 8), which strongly supports the role of plasma proteins within the airway-lining fluid interfering with surfactant function (44). Admittedly, it was the acid-injured mice that drove most of this correlation (Fig. 8), and although a linear function was fit to the data, the data distribution in Fig. 8A suggests a possible threshold effect for BALF protein content on recovery in H with an exponential rise in total recovery beyond 750 μg/ml. It may be that a critical amount of protein seepage into the alveolar space is required before surfactant function is suffi-ciently disrupted so as to decrease alveolar stability and pulmonary compliance.
We conclude that the mechanical consequences of both a DI and PEEP depend on the type and degree of lung injury. We also conclude that the BALF protein content is predictive of the response in lung function to an RM, whereas the degree of BALF neutrophilia is not. These conclusions have important implications for the utility of RMs in managing patients with different types of ALI (14, 41) or at varying stages of injury (17). For example, in severe ALI, the increases in amount of open lung after a DI may only persist for a useful period of time if some critical level of PEEP is employed. Determining the ideal PEEP needed to keep the lung open, however, has proven to be a formidable task (33, 41, 46). The present study demonstrates that the position of the traditionally used LIP can depend on the antecedent ventilation settings, which may explain some of the controversy surrounding its use. Finally, our study demonstrates how the utility of RMs and PEEP can be better appreciated though the use of frequent measurements of lung function during mechanical ventilation and entreats the use of this approach for further investigation of recruitment strategies in the clinical setting.
This work was supported by National Institutes of Health Grants HL-62746, R01 HL-67273, and NCRR-COBRE P20 RR-15557.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society