Published ahead of print on September 4, 2003, doi:10.1164/rccm.200206-527OC
© 2003 American Thoracic Society Dose–Response Effect of Perfluorocarbon Administration on Lung Microvascular Permeability in RatsEA 3512, IFR02 Claude Bernard, Institut National de la Santé et de la Recherche Médicale, Faculté de Médecine Xavier Bichat, Paris; Service de Réanimation Médicale, Hôpital Louis Mourier (Assistance Publique—Hôpitaux de Paris), Colombes; and Service de Radiologie, Hôpital Bichat (Assistance Publique—Hôpitaux de Paris), Paris, France Correspondence and requests for reprints should be addressed to Georges Saumon, M.D., EA 3512, Faculté Xavier Bichat, BP 416, 75870 Paris Cedex 18, France. E-mail: saumon{at}bichat.inserm.fr
The effect of various perflubron doses on overdistension lung injury was evaluated. Rats were given perflubron at 0 ml/kg (control) to 20 ml/kg and ventilated with a VT of 33 ml/kg without or with 5 cm H2O of positive end-expiratory pressure (PEEP). High (20 ml/kg), but not lower, perflubron doses aggravated lung capillary leak in the absence of PEEP. PEEP application aggravated capillary leak in controls, had no effect in those given a low (10 ml/kg) dose, but decreased the leak in rats ventilated with a large dose compared with zero end-expiratory pressure. In the presence of PEEP, this low dose decreased capillary leak compared with controls or with rats given the large dose. Lung computerized tomography scans showed that the large dose increased functional residual capacity by 68% and produced gas trapping that was reduced by PEEP. Thus, large doses predispose to overdistension injury whereas low doses do not and may even have a protective effect in the presence of PEEP. The paradoxical beneficial effect of PEEP when large doses are given may be due to gas trapping reduction. These findings confirm that liquid ventilation does not aggravate volutrauma provided perflubron doses are adjusted. They provide a lead to further investigate partial liquid ventilation in the clinical setting.
Key Words: partial liquid ventilation • acute respiratory distress syndrome • tomography, X-ray computed • functional residual capacity • ventilator-induced lung injury A study with high-dose perfluorocarbon (PFC) did not show reduction of mortality with partial liquid ventilation (PLV) (1). These data highlight the fact that, contrary to drugs, the dose–efficacy–safety relationship of PFC has not yet been assessed. Numerous experimental studies have stressed the substantial improvement in oxygenation and lung mechanics obtained with PFC ventilation in animal models of hypoxemic respiratory failure. Because of its density and tensioactive properties and its high solubility coefficients for O2 and CO2, PFC distributes preferentially in dependent regions, recruits zones excluded from the ventilation, and improves gas exchange (2). During early clinical studies (3, 4) a dose of PFC of about 30 ml/kg was used, the lungs being considered satisfactorily filled when a PFC reflux was seen at end expiration in the endotracheal tube. Pneumothoraces have been reported during PLV of premature infants (3) and of adults (1) with this dosing. An experimental study suggests that such high doses of PFC are associated with increased pneumothorax incidence (5). This suggests that PLV may in some instances result in lung overdistension. Instillation of a small amount of PFC may reduce the severity of ventilator-induced lung injury in the setting of alveolar flooding (6). However, such a PFC instillation sometimes failed to reduce lung injury (6), perhaps because flooded lungs were not fully recruited. Positive end-expiratory pressure (PEEP) is a widely used strategy to recruit the lungs (the "open lung approach") and regularly ameliorates oxygenation during PLV (2, 7). The pressure–volume curve of the respiratory system has been used to adjust PEEP during PLV (8, 9). Although these studies provide strong evidence for a beneficial effect of high PEEP level in combination with high doses of PFC, their authors acknowledge that "possible volutrauma to lung units at perflubron and PEEP levels resulting in maximal PaO2 cannot be ruled out" (8). Indeed, the effect of PFC filling on lung recoil pressure and thus on the resting volume of the respiratory system (i.e., FRC) has not been specifically addressed. PFC reflux in the trachea may correspond to instillation of a volume of liquid well above the actual FRC if the lung pressure–volume curve is shifted to the left by PFC instillation. In addition, PEEP application may augment the risk of overinflation in the absence of appropriate reduction of tidal volume (10). This study was designed to evaluate the influence of PLV with various doses of perflubron (LiquiVent; Alliance Pharmaceutical, San Diego, CA) together with the effect of PEEP on overinflation lung microvascular permeability. We found that perflubron dose and PEEP interact in a complex manner. Low perflubron doses do not affect microvascular permeability alterations observed when ventilating with a large tidal volume. However, they decrease those produced by a large tidal volume and PEEP. Large perflubron doses increase lung capillary protein leak; however, this increase was less but still important in the presence of PEEP. Preliminary results of this study have already been presented in abstract form (11).
All experiments were conducted on male Wistar rats (weighing 280–320 g; Charles River, St-Aubin-lès-Elbeuf, France) in accordance with regulations of the French Ministry of Agriculture. Rats were anesthetized by intraperitoneal injection of thiopental (50 mg/kg; Sigma, Saint-Quentin Fallavier, France). Rodents metabolize thiopental slowly and remain deeply anesthetized for at least 4 hours with this dose. They were tracheostomized, paralyzed with succinylcholine (Sigma), and ventilated with a Harvard volume respirator (Ealing, Courtaboeuf, France) to achieve a VT of 33 ml/kg body weight with 25 breaths per minute for 10 minutes. At the onset of ventilation, animals received either no PFC or PFC at 6.7, 10, 13.3, 16.7, or 20 ml/kg body weight and zero end-expiratory pressure (ZEEP) (nine animals per group). Additional experiments using 5 cm H2O PEEP were performed in rats receiving PFC at 0, 10, or 20 ml/kg body weight (nine animals per group). Animals ventilated with a VT of 7 ml/kg and ZEEP served as controls (n = 5). End-inspiratory pressure (plateau pressure, Pei) was measured with a piezoelectric transducer during mechanical ventilation. The extravascular distribution space of albumin in lungs (Alb-sp) was used to assess microvascular injury. It was measured as previously described (12). Briefly, lung blood content was measured using 99mTc-labeled erythrocytes. Radioiodinated human serum albumin was intravenously injected 30 minutes before the rat was killed. 125I activity was measured in a blood sample and in the lungs. Lung blood activity was calculated, and subtracted from total activity. The extravascular distribution space of albumin in lungs was defined as the space (volume) in which albumin would distribute in blood-free lungs, assuming that extravascular concentration was the same as plasma concentration. Five animals underwent computed tomography (CT) scan evaluation (see online supplement for CT scan technique) of lung volume before and after PLV; two of these animals were also evaluated after 15 minutes of PLV with PEEP. Animals anesthetized and tracheostomized as described above were placed in the supine position and mechanically ventilated, after paralysis with succinylcholine, to achieve a VT of 6.7 ml/kg with 60 breaths per minute under 100% FIO2. A CT scan was done before perflubron instillation to measure FRC. Each animal was disconnected from the ventilator during helical acquisition that lasted 17 seconds. After this the animal was immediately returned to mechanical ventilation. Another acquisition was obtained after instillation of perflubron at 20 ml/kg. PLV was performed for 15 minutes either with or without PEEP, before the rat was disconnected, and the next CT scan was taken. Effect of PEEP on PFC distribution in the lungs was assessed by measuring Hounsfield unit (HU) density in distinct slices of each lung (from apices to diaphragm) before and 15 minutes after PEEP was applied.
Statistical Analysis
Effect of PFC Dosing on Lung Microvascular Permeability Alb-sp in rats not given perflubron and ventilated with a high VT was about threefold that of controls (p < 0.05). Alb-sp did decrease, but not significantly, after instillation of perflubron at 6.7 or 10 ml/kg in comparison with animals not given perflubron. Alb-sp had a tendency to increase with further (13.3 and 16.7 ml/kg) doses of perflubron, but this increase did not reach statistical significance. This increase was, however, significant with perflubron at 20 ml/kg (p < 0.01) (Figure 1) .
Effect of PEEP on Lung Microvascular Permeability Compared with ZEEP, PEEP significantly increased Alb-sp when no perflubron was administered, did not affect Alb-sp in the presence of perflubron at 10 ml/kg, but significantly decreased it when perflubron was administered at 20 ml/kg (Figure 2) .
End-inspiratory Pressure End-inspiratory pressure (Pei) increased significantly with incremental doses of perflubron (Figure 3) . Immediately after perflubron instillation, Pei was significantly greater when perflubron was given at 16.7 and 20 ml/kg than at lower doses (p < 0.05). After 5 and 10 minutes of mechanical ventilation, Pei was greater only in rats given perflubron at 20 ml/kg (p < 0.05) (Figure 3).
Adding 5 cm H2O PEEP led to a significant increase in Pei in rats not given perflubron, but had no effect in those given perflubron at 10 ml/kg. The Pei had a tendency to decrease in rats given perflubron at 20 ml/kg, but without reaching statistical significance (p = 0.11; Figure 4) .
Lung Volume FRC measured before perflubron instillation was 4.4 ± 0.1 ml. After instillation of perflubron at 20 ml/kg, FRC increased to 7.2 ± 0.34 ml (p < 0.01). After ventilation with 5 cm H2O PEEP, FRC was 6.8 ± 0.5 ml (not significant in comparison with PFC at 20 ml/kg and ZEEP). Three-dimensional reconstruction of the lungs of one of these rats is shown in Figure 5 .
Gas Trapping HU density analysis was performed on 40 slices at the same location with ZEEP and then with PEEP. There was a modest increase in slice HU density from ZEEP to PEEP (2,065.8 ± 41.8 versus 2,174 ± 41.8 HU, t = 3.37, p < 0.01). Slice surface area (arbitrary units: SU) was slightly lower with PEEP (101.0 ± 9.86 versus 93.1 ± 9.11 SU, t = 3.22, p < 0.01). Total HU (mean HU x surface area) was identical in ZEEP and in PEEP (206,541 ± 19,322 versus 199,761 ± 19,317 HU x SU, t = 1.36, not significant), reflecting the fact that perflubron quantity in lungs was the same under the two conditions. However, the mean standard deviation of HU density was significantly greater with ZEEP than with PEEP (375.6 ± 24.3 versus 283.5 ± 13.5 HU, t = 5.67, p < 0.001) reflecting more heterogeneity in HU density in slices with ZEEP than with PEEP. One of the 20 paired slices is shown in Figure 6 . In these medium thoracic slices, an obvious trapping of gas in perflubron-filled lungs in ZEEP is considerably reduced with PEEP.
In the present study, we show the following: 1. Mechanical ventilation with a high VT and ZEEP produces microvascular permeability alterations that are aggravated by the administration of a high dose, but not a low dose, of perflubron. 2. PEEP worsens microvascular protein leak in rats not given perflubron because it increases end-inspiratory pressure (and hence lung volume [10]); this deleterious effect is prevented by administration of a low dose of perflubron. 3. PEEP also has a protective effect on the greater microvascular protein leak observed in rats given a high dose of perflubron. 4. However, even in the presence of PEEP, high doses of perflubron increase microvascular permeability because of high volume ventilation. 5. In the absence of PEEP, PFC administration is associated with a significant amount of gas trapping. Adding PEEP significantly reduces this gas trapping. These results suggest that in the setting of overinflation lung injury, the interaction of PEEP and perflubron is complex. The contrasted effects of PEEP and perflubron, depending on the context, seem to be the consequence of the balance between improved respiratory mechanics and increased end-expiratory volume. Administration of perflubron at 20 ml/kg in rats did not result in perflubron overflowing, a meniscus of liquid being seen in the tracheal cannula at end expiration as in clinical trials (1). We were surprised at that, because careful measurements of FRC in rats of similar weight gave values ranging from 2.51 ml (13) to 4.89 ml (14). We thus verified whether perflubron administration affected rat end-expiratory volume by measuring FRC by CT scan and three-dimensional lung reconstruction. FRC values of animals ventilated with air were similar to those previously published (13, 14). Instillation of perflubron at 20 ml/kg resulted in a marked increase in FRC. This increase may be explained by two mechanisms (15): perflubron weight and gas trapping. Even if the size of rat lungs is small (i.e., 2 cm maximum along the dorsoventral axis), the weight of perflubron may play a nonnegligible role because of the high respiratory system compliance of rats (typically 0.6 ml/cm H2O). Thus, the maximum increase in pressure in dorsal segments of rat lungs can be expected to be 3.8 cm H2O (perflubron density, 1.9), which would correspond to a maximum increase in volume of 2.3 ml, or two-thirds the FRC. The real effect of perflubron weight indeed stays between 0 and this extreme, but cannot be considered inconsequential. This effect is apparent on CT scans that show that the dependent zones are preferentially enlarged after perflubron instillation. CT scans also clearly show that some gas is trapped in these lungs. This trapping may be the consequence of the asymmetrical branching of the tracheobronchial tree (16). This hypothetical explanation is depicted in Figure 7 . The effect of high VT gas ventilation on a liquid-filled lung would necessarily result in gas trapping if short and long pathways to alveolar ducts coexist. When the dose administered is small (i.e., 10 ml/kg), perflubron distribution is more distal and no significant trapping occurs because the liquid does not flow back in main bronchi. This increase in FRC explains why a high dose of perflubron increases end-inspiratory pressure during high VT ventilation, increases tissue stress, and results in more severe microvascular permeability alterations.
We have previously shown that application of PEEP may worsen overinflation and aggravate edema when cardiac output is not markedly depressed by too high intrapulmonary pressure (10). This previous observation is confirmed in rats not given perflubron. The beneficial effect of perflubron (10 ml/kg) instillation on permeability alterations may be explained by improved mechanical properties at high lung volume, as attested by the much lower Pei values seen in these rats. It has been shown that surface tension at the air–liquid interface increases during inflation in normal, air-filled lungs (17) and that this increase contributes to the decrease in compliance above the "upper inflection point" and to the high transpulmonary pressures seen approaching the TLC. It is conceivable that perflubron administration decreases surface forces at high lung volume by providing an interface with constant and low (18 dyn/cm) surface tension, resulting in less negative tissue pressure and, thus, reducing filtration and protein leak (18). End-inspiratory pressure was only slightly, but not significantly, lower with PEEP in rats given perflubron at 20 ml/kg. However, tissue pressure at end inspiration may be more affected than indicated by the changes in Pei measured at the tracheal aperture if trapping and overinflation are inhomogeneous. Indeed, the only clear explanation for the lessening of permeability alterations with PEEP is that PEEP paradoxically decreased end-expiratory volume in distal lung spaces during partial liquid ventilation (see, e.g., Figure 6), the exact opposite to what occurred in rats ventilated with air. This decrease is obviously the consequence of less distal gas trapping because the amount of perflubron in lungs being unchanged, the "liquid PEEP" effect is not expected to change. This manifests as less heterogeneity of HU density in lung slices after PEEP in comparison with that before PEEP. This is also obvious by eye (Figure 6). A possible mechanism is that PEEP displaces the perflubron front to the periphery, reducing the risk of trapping during the next breath (Figure 7). This mechanism is supported by the disappearance with PEEP of the "pressure surge" seen at low lung volume with ZEEP as previously reported by others (19, 20). Experimental studies have investigated the use of the pressure–volume curve to adjust PEEP during partial liquid ventilation (8, 9). Efficacy of perflubron administration in these studies on animals depleted in surfactant by saline lavage was based on PaO2 increase. Authors found that the highest PaO2 was obtained with high PEEP (1 cm H2O above the lower inflection point) combined with a high (30 ml/kg) perflubron dose (8). However, potential negative effects of such regimens were not assessed in this study and, as stated by the authors, "possible volutrauma to lung units at perflubron and PEEP levels resulting in maximal PaO2 cannot be ruled out" (8). Cox and coworkers (5) investigated these adverse effects of PLV in surfactant-depleted rabbits. The risk of barotrauma was greater in animals given high doses of PFC (5). Our results showing that rats given perflubron at 20 ml/kg had greater ventilator-induced lung injury (even in the presence of PEEP) are in agreement with this study. Taken together, these results may provide some kind of explanation for the absence of beneficial effects of PLV on mortality in patients with acute respiratory distress syndrome described by Hirschl and colleagues (1). In the clinical setting, an international multicenter randomized controlled trial compared two doses of perflubron with conventional mechanical ventilation after a preliminary feasibility pilot study (21). Although the study is not yet published, its preliminary results were given during the American Thoracic Society meeting in May 2001 (H. Wiedemann, public announcement, ATS meeting, San Francisco, May 2001) and published at the Alliance Pharmaceutical Web site (http://www.allp.com/press/press.exe?@B0521). There was no reduction of mortality with PLV and a trend to an increased risk for barotrauma with the higher doses of PFC. One possible explanation for these results is that in some instances high-dose PFC administration reduced ventilatable lung volume (baby lung effect), thus increasing ventilator-induced lung injury (5, 22). These observations stress the difficulty in extrapolating experimental findings to the clinical setting. Our results clearly illustrate the occurrence of gas trapping during partial liquid ventilation with high PFC doses and suggest that use of PEEP may lessen overinflation lung injury when applied with these high doses by reducing gas trapping. Our results also indicate that this can be better achieved by using lower doses of perflubron. Further clinical studies using low doses may be interesting.
J-D.R. received $10,000 from Alliance Pharmaceuticals (that sells Liquivent) as a grant for performing post-doctoral research in Dr. Saumon's laboratory and Alliance Pharmaceuticals did not interfere with design or interpretation or the writing of this paper; D.D. has no declared conflict of interest; J.P-L. has no declared conflict of interest; G.S. received $10,000 from Alliance Pharmaceuticals (that sells Liquivent) as a grant for functioning of my laboratory and Alliance Pharmaceuticals did not interfere with design or interpretation or the writing of this paper. The authors thank David Blaysius for helpful technical assistance.
Supported in part by Alliance Pharmaceutical Corporation. J-D.R. is a recipient of a Research Fellowship from the Fonds de Recherche et d'Etudes du Corps Médical des Hôpitaux. This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form June 7, 2002; accepted in final form September 3, 2003
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
