| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Division of Pediatric Critical Care and 5Division of Pulmonary and Critical Care Medicine, 4University of Virginia School of Medicine, Charlottesville, Virginia 22908-0386; 3Department of Pathology and 2Critical Care Research Laboratories and Department of Anesthesia, Children's Hospital and Harvard Medical School, Boston, Massachusetts 02115
Submitted 8 July 2002 ; accepted in final form 6 May 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
acute respiratory distress syndrome; high-frequency oscillatory ventilation; perfluorochemical; perflubron; myeloperoxidase activity
Both high-frequency oscillatory ventilation (HFOV) and partial liquid breathing techniques have been advanced as offering lung protection by improving lung mechanical properties as well as limiting ongoing lung inflammation. HFOV, when utilized with an optimal lung volume strategy, has been shown to improve histopathological outcomes in animal models as well as limit the elaboration of proinflammatory cytokines (14, 18, 20, 38).
The clinical experience with high-frequency oscillation (HFO) has suggested a decreased incidence of chronic lung disease in premature newborn infants (6, 11, 28) and also in pediatric patients with acute respiratory distress syndrome (3) when an optimal lung volume strategy is utilized. Furthermore, despite early disappointment in the use of high-frequency techniques in adults (5), there is now renewed interest in the application of high-frequency ventilation in adults with lung injury (9, 10, 12, 21).
Partial liquid breathing techniques have also been shown to improve the histopathological outcomes in animal models of acute lung injury (13, 25, 31, 40). Recent investigation has suggested that high-frequency techniques may act synergistically with liquid ventilation to provide enhanced lung protection in the management of the acutely injured lung (8, 35, 37). We sought to extend these observations by examination of gas exchange, histopathological outcomes, and vapor loss characteristics in an animal model of acute lung injury managed with HFO-partial liquid ventilation (PLV).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Lung injury. After intubation and instrumentation, repetitive
saline lavage was performed with the use of isotonic saline (30 ml/kg) warmed
to body temperature until a uniform degree of gas-exchange impairment was
achieved: arterial oxygen saturation from pulse oximetry
(SpO2) 
90% on an
FIO2 of 1.0. Ventilation was provided during
this period and in the immediate postlavage period by using volume-cycled
mechanical ventilation at preinjury settings with rate adjusted to maintain
PaCO2 of 40 ± 5 Torr (5.3 ± 0.7 kPa).
After surfactant depletion via lavage had been performed, the lung injury was
amplified over a 4-h period by subjecting animals to cyclic alveolar
distension and derecruitment, as follows: Synchronized Intermittant Mechanical
Ventilation/Volume Control; VT 15 ml/kg, PEEP 2 cmH2O,
I:E 1:2, FIO2 1.0, RR adjusted to target
PaCO2 40-50 Torr (5.3-6.6 kPa). Lung recruitment
using HFOV. After the induction of lung injury, all animals were
converted to HFOV (Sensor Medics 3100A, Sensor Medics Critical Care, Yorba
Linda, CA). During this period, an attempt was made to optimize lung volume by
using an aggressive strategy designed to increase mean airway pressure until
oxygenation targets were achieved: SpO2 
90% on an
FIO2 of 0.60. HFO was instituted by using
a mean airway pressure 10 cmH2O greater than that used during
conventional ventilation and increased until oxygenation targets were
achieved. The power setting of the oscillator was adjusted to maintain a
peak-trough pressure difference (
P) that achieved a
PaCO2 between 45 and 55 Torr (6.0-7.3 kPa) at a
frequency of 7 Hz. The inspired oxygen concentration remained 0.6 for the
remainder of the experiment. After lung recruitment and stabilization on HFO
for 60 min, animals were then randomly assigned to one of three treatment
groups: HFO-gas control, HFO-PLV compartmented perflubron (PFB) [PLV(C)], and
HFO-PLV uniform PFB [PLV(U)].
Initial PFB administration. HFO-PLV was initiated with the
instillation of PFB via an endotracheal tube sideport adapter at a rate of 0.5
ml · kg-1 · min-1 to
achieve a dose of 10 ml/kg. The administration rate was determined in pilot
work and is the fastest administration rate possible without incurring reflux
of PFB into the HFOV circuit. At the onset of the dosing interval, frequency
was reduced in all groups from 7 to 4 Hz; no other changes were made to
ventilator settings during the dosing period. In the animals assigned to the
PLV(U) group, 5 ml/kg PFB were administered in the supine position over 10
min; animals were maneuvered into the prone position over 10 min, dosed with a
second 5 ml/kg PFB over 10 min, and returned to the supine position after 10
min. In the animals assigned to the compartmented-PFB group, all PFB was
administered in the supine position in two 5 ml/kg aliquots, as in the PLV(U)
group; however, a 10-min interval was observed before and after the second
aliquot. The animals in the PLV(C) group were not repositioned. Once dosing
was completed, the mean airway pressure was adjusted to achieve a
SpO2 of 90 ± 2% with a
FIO2 of 0.6. The five control animals that
were not dosed with PFB underwent exactly the same frequency change from 7 to
4 Hz at the onset of the dosing interval and were not repositioned. As in the
PFB-treated animals, the mean airway pressure during this period was adjusted
to achieve a SpO2 of 90 ± 2% with an
FIO2 of 0.6, and no manipulation of the
peak-trough P was made.
Quantifying PFB loss. After the initiation of PLV, intrapulmonary
PFB losses were estimated by continuous quantification of both the PFB vapor
content and the exhaust flow rate from the HFOV circuit. A thermal mass
flowmeter (model 4040, TSI, St. Paul, MN) was interposed at the exhaust port
of the HFOV circuit, yielding continuous data, and the expiratory limb of the
ventilator circuit was heated to prevent any PFB vapor condensation. PFB vapor
content of exhaust gas was quantified via infrared vapor (IR) analysis by
using a device custom designed for this purpose (Alliance Pharmaceutical, San
Diego, CA) and is described in detail elsewhere
(19). Briefly, the specificity
of this device for the IR detection of PFB in the presence of oxygen, carbon
dioxide, nitrogen, and water vapor has been confirmed, and, additionally, IR
quantification and gravimetric quantification (by weight loss) of PFB
evaporative loss correlate well both in a simulator device and in swine over a
large range of vapor loss rates
(19). A custom couple captured
all exhaust gas from the HFOV circuit and provided access for continuous
sidestream sampling via a heated umbilical tube through which gas was pumped
to the optical cell of the IR instrument, yielding IR absorbance data and thus
vapor pressure for PFB (PPFB). PPFB was converted to an
absolute PFB concentration (CPFB) (mg/l) after generation of a
calibration curve for PFB, according to the idealized form of the Beer-Lambert
law, where absorbance (A) is defined
 | (1) |
 | (2) |
 | (3) |
 | (4) |
E is minute ventilation and
t is time. The exhaust gas stream was sampled every 30 s, generating a series of instantaneous PFB vapor loss rates and also yielding cumulative PFB vapor loss, which was tracked as a running total.
Maintenance PFB dosing. After the initial PFB dosing, vapor loss was monitored, and no PFB redosing was undertaken in the animals assigned to the uniform-PFB group. In the animals assigned to the compartmented-PFB group, however, intrapulmonary PFB volume was maintained by adherence to a redosing schedule: every 2 h the evaporative PFB loss quantified over the preceding 2-h interval was delivered in the supine position at a rate of 0.5 ml · kg-1 · min-1. We chose to deliver the replacement dose in a bolus, rather than continuously, to limit time that PFB was present in the large airways. This minimized vapor loss of the replacement PFB itself and permitted selected monitoring of vapor loss from the alveolar PFB pool in the intervals between PFB redosing.
HFO management during HFO-PLV. HFO management was identical in all
groups over the 15-h HFO-PLV study period. All animals were maintained with
frequency set at 4 Hz and an I:E of 1:2. The power setting of the oscillator
was adjusted to maintain a peak-trough P that achieved a
PaCO2 between 45 and 55 Torr (6.0-7.3 kPa). During this
period, airway pressure (Paw) was manipulated to maintain the oxygenation
targets: SpO2 = 90-95% on an
FIO2 of 0.60.
Data collection. Baseline measurements were made 15 min after the induction of anesthesia and were repeated every 30 min after induction of lung injury. After the completion of PFB dosing, data were recorded every 15 min for the first hour and then hourly until the termination of the experiment.
Data collection was performed in all groups at the same time points and included arterial blood-gas tensions, hemodynamic parameters, and ventilator settings. The directly measured hemodynamic parameters included heart rate, femoral artery pressure, pulmonary artery pressure, pulmonary capillary wedge pressure, and cardiac output by thermodilution. Pulmonary capillary wedge pressure was recorded during its nadir at end expiration. Cardiac output was calculated with a Baxter COM-2 cardiac output computer (American Edwards Laboratories, Irvine, CA) after serial injections of 5 ml of iced saline (0°C) into the right atrium. At each data point, iced saline injection was repeated until three cardiac output computations grouped within 10% were obtained, and the group mean was recorded. Both cardiac output and pulmonary vascular resistance were indexed to weight.
Data collection also included the following ventilator and pulmonary mechanics information during conventional mechanical ventilation: RR, VT, peak inspiratory pressure, PEEP, and dynamic compliance.
During HFOV, mean Paw, the power setting, and the peak-trough P were
recorded. In addition, at each time point, arterial blood-gas tensions and pH
were measured by use of a standard laboratory blood-gas analyzer (Ciba
Corning, 278 blood-gas system, Ciba Corning Diagnostics, Medfield, MA), and
the oxygenation index was calculated.
In addition, an estimation of intrapulmonary shunt
(Qps/QT) was calculated by using the standard formula
 | (5) |
 | (6) |
 | (7) |
 | (8) |
Lung preparation. After termination of the 20-h experimental protocol, the anesthetic was deepened and the lungs were exposed via median sternotomy; the right lung was prepared for histopathological analysis of lung injury, and the left lung was prepared for tissue banking and myeloperoxidase assay. The main pulmonary artery trunk was exposed and isolated. A 10-Fr catheter was placed in the main pulmonary artery trunk and secured. To protect the right lung from saline perfusion, the right hilum was cross-clamped. The superior vena cava and inferior vena cava were isolated and ligated at their entry to the right atrium. Simultaneously, an incision was made in the left atrium, and the left lung was perfused with normal saline via the catheter in the main pulmonary artery trunk until the efflux from the left atrium was clear. The left hilum was then divided, the lung was removed expeditiously from the thoracic cavity, and multiple 2 x 2-cm tissue blocks from all lung regions were fast frozen in liquid nitrogen before storage at -80°C. The animal was then euthanized with a lethal dose of 15 mg/kg pentobarbital intravenously.
The trachea was then divided at its entry to the thoracic cavity, the right main pulmonary artery and veins were ligated and divided, and the lung was dissected free and removed from the thoracic cavity. The lung was placed in a 10% formalin bath and inflated with 10% formalin to a pressure of 25 cmH2O via a catheter secured in the right mainstem bronchus and fixed over a 48-h period. Tissue blocks for sectioning were then prepared from dependent and nondependent regions of the lower, middle, and upper lobes of the right lung (6 tissue blocks per animal, 72 total blocks). The tissue was sliced, and slides were prepared from each block.
Lung injury scoring system. Hematoxylin and eosin-stained samples underwent review by a blinded pathologist using a semiquantitative scoring system to quantify the degree of lung injury. The following variables were scored: airway plugs (mucus and cell debris within airway lumens), atelectasis (partially to completely collapsed alveoli), alveolar and interstitial inflammation (neutrophils in alveolar spaces and interstitium), edema (alveoli filled with homogeneous eosinophilic material), alveolar distension (expansion of alveoli to 1.5-2 times normal size), hemorrhage (free red blood cells within alveoli), necrosis, and hyaline membrane formation. For each component, the percentage of injured lung was recorded. Severity of injury was then reduced to an integer from 0 (no injury) to 10 (100% injury).
Lung myeloperoxidase activity. Myeloperoxidase activity, indexed by protein content of lung tissue samples (16), was determined as a measure of lung neutrophil sequestration. Lung tissue was minced, homogenized, and placed in an ultrasonic dismembranator (Branson Ultrasonic, Danbury, CT). To further disrupt cellular membranes, three cycles of freezing (-80°C) and thawing were performed, after which the ultrasonic treatment was repeated. After further high-speed centrifugation for 30 min, supernatant was added to chilled 0.1% O-dianisidine dihydrochloride (Sigma Chemical, St. Louis, MO) in 0.05 M sodium phosphate buffer (pH 6.0) activated by 0.005% hydrogen peroxide. Myeloperoxidase activity in each sample was determined by measuring the change in absorbance over 2 min at 460 nm by use of a spectrophotometer (Shimadzu UV 1601, Wood Dale, IL). The protein content of sample supernatant was determined via the bicinchoninic acid method (BCA protein assay kit, Sigma Chemical). Myeloperoxidase activity was expressed as change in optical density per minute per microgram protein.
Statistical analysis. Gas exchange, hemodynamic parameters and pulmonary mechanics data were compared by using repeated-measures ANOVA with PFB dose strategy as the main effect and time as the within-subject factor. The effects of time and the interaction between PFB dose strategy and time were evaluated by using two-way repeated-measures ANOVA; post hoc analysis of between-groups differences was made by use of Fisher's paired least-significant difference test. At the time points immediately before HFO lung recruitment and PFB dosing, all groups were compared by using a two-sample Student's t-test. Lung injury scores were compared with Kruskall-Wallis ANOVA; post hoc analysis of between-groups differences of individual lung injury scores components was performed by using Dunn's test for multiple comparisons. A two-tailed P value <0.05 was considered statistically significant throughout. Both linear and logarithmic curve-estimation regression models were utilized to plot and generate regression lines and the coefficient of determination (R2) for vapor loss data. Analysis was performed by using SPSS version 9.0 (SPSS, Chicago, IL) and StatView version 4.51 (Abacus Concepts, Berkeley, CA).
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
For all groups, saline lavage produced a significant drop in both arterial PO2 (PaO2) and dynamic lung compliance (P < 0.001) and an increase in the oxygenation index (P < 0.001) (Table 1). Lavage also led to a significant increase in cardiac index (P = 0.005) and pulmonary vascular resistance index (P = 0.002). There were no differences between any of the experimental groups after induction of lung injury (Table 1).
|
At the end of the period of lung recruitment using HFOV, we found a significant increase in PaO2 for all groups, although there was a concomitant increase in oxygenation index after lung volume recruitment (P = 0.045) (Table 1). Although there was a significant increase in the pulmonary vascular resistance index (P = 0.006), there was not a significant change in the cardiac index (P = 0.117) after lung volume recruitment (Table 1). At the end of lung volume recruitment with HFOV, there were no significant differences between any of the experimental groups before dosing in blood-gas tensions, hemodynamic parameters, or lung mechanics (Table 1).
Gas exchange during PFC dosing. PFC dosing at 0.5 ml · kg-1 · min-1 was initiated after lung recruitment with HFOV; data were gathered every 10 min after each of four dosing intervals. In the gas control group (n = 5), we noted a modest increase in PaO2, enabling a small reduction in the Paw over the period of sham dosing; oxygenation index was thus unchanged in the control animals by the end of the sham dosing interval (Table 2). PaCO2 was stable throughout the dosing interval. The initiation of HFO-PLV in the PLV(U) group led to somewhat improved oxygenation. Specifically, oxygenation improved on prone positioning and on return to the supine position after each 5 ml/kg aliquot of PFB (P = 0.07) (Table 2), permitting a reduction in Paw and leading to a nonsignificant reduction in oxygenation index (P = 0.18). PaCO2 was stable throughout the dosing interval. The initiation of HFO-PLV in the PLV(C) group produced no notable changes in gas exchange (Table 2). Gas exchange between groups, however, was significantly different over the dosing interval for PaO2 (P = 0.02). There were no significant differences in oxygenating efficiency as assessed by the oxygenation index between groups during the 40-min dosing interval.
|
Gas exchange during HFO-PLV. After PFC dosing, the only
experimental manipulation consisted of adjustment of Paw to achieve a
SpO2 of 90 ± 2% with
FIO2 of 0.6. No manipulation of the
peak-trough P or change in oscillator frequency was made. In the gas
control group, we noted a steady increase in PaO2
(Fig. 1), permitting
progressive reduction in Paw (Fig.
1) over the course of the 15-h experimental period, which produced
a significant decrease in the oxygenation index within this group over time
(Fig. 1). Similar changes were
noted in the PLV(C) group (Fig.
1). In the PLV(U) group, PaO2 improved
dramatically after the dosing period (P < 0.01)
(Fig. 1), permitting a
progressive reduction in Paw. These changes were reflected in a substantially
lower oxygenation index in the PLV(U) group, although it was not sustained
beyond 5 h. Comparison with repeated-measures ANOVA revealed significant
differences between the groups over time in PaO2
(P < 0.001) and oxygenation index (P = 0.02).
|
Hemodynamic changes. The cardiac index did not change significantly over the course of the experimental protocol, and there were no significant differences among the three treatment groups (Fig. 2). Administration of PFB did not result in significant changes in pulmonary vascular resistance index in any treatment group, and there were no significant changes in pulmonary vascular resistance over time in any group. Between-group comparison did not reveal differences between the treatment groups (Fig. 2).
|
PFB vapor loss. In the PLV(U) group, hourly PFB vapor loss normalized to body weight was observed to decay in a logarithmic fashion from 2.8 ml · kg-1 · h-1 to 0.6 ml · kg-1 · h-1 (Fig. 3); as losses were not replaced in this group, 50% of the initial fill was lost 3.8 h into the experiment (Fig. 3). A curve-fitting procedure generated a logarithmic decay curve with R2 of 0.87 (Fig. 3 inset). In the PLV(C) group, hourly PFB vapor loss was fairly steady, fluctuating about 1.25 ml · kg-1 · h-1 (Fig. 3). In the PLV(C) group, a total of 16 ml/kg was replaced over the course of the experiment to maintain the initial fill volume of 10 ml/kg (Fig. 3). Of note, in the PLV(U) group, PaO2 exhibited an inverse correlation with weight-normalized cumulative PFB vapor loss with R2 of 0.703 (Fig. 3).
|
Histopathology. The eight-component lung injury scoring system was
used to create an injury score for the dependent and nondependent lung. The
regional scores were combined to yield a cumulative lung injury score for the
entire lung. Differences in the individual components of the lung injury score
were also examined for between-group differences. The cumulative injury score
in the gas control group was 
25% higher than in either of the PLV groups;
this trend was not significant (Fig.
4). In dependent lung, there was a 30% reduction in injury score
in the PLV(U) group and a nearly 50% reduction in injury score in the PLV(C)
group (P = 0.045) (Fig.
4). In nondependent lung, injury scores in the gas control and the
PLV(C) group were essentially the same
(Fig. 4). A reduction of
30% was noted in the PLV(U); this trend was not significant
(Fig. 4). In examining
components of the cumulative scores, a significant reduction in alveolar
distension and atelectasis was noted in both PLV groups
(Fig. 4). When individual score
components were examined separately by lung region, a significant reduction in
atelectasis was noted in both PLV groups in dependent lung
(Fig. 4). In nondependent lung,
a significant reduction in alveolar distension was noted in both PLV groups
(Fig. 4).
|
Myeloperoxidase assay. Myeloperoxidase activity indexed by weight to tissue protein content for both PLV groups was compared both to gas control for the lung as a whole and in a separate analysis for dependent and nondependent lung individually. For whole lung, there was a similar 65% reduction in myeloperoxidase activity in both PLV groups (P < 0.01) (Fig. 5). In dependent lung, myeloperoxidase activity was attenuated by 69% in the PLV(C) group (P = 0.02) and by 64% in the PLV(U) group (P = 0.002) (Fig. 5). In nondependent lung, myeloperoxidase activity was attenuated by 45% in the PLV(C) group (P = 0.02) and 52% in the PLV(U) group (P = 0.03) (Fig. 5).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Oxygenation, as assessed by the arterial oxygen tension and the oxygenation index, was transiently improved in the group treated with perfluorocarbon by using rotational positioning. However, 3 h after the animals were returned to the supine position, these improvements in oxygenation were not evident compared with the other two study groups. The absence of a persistent improvement in oxygenation in the rotationally dosed animals may be due to the beneficial effects of the prone position (24), which were no longer present after 3 h in the supine position. It is also possible that evaporative losses, which were not replaced in this group, interrupted the positive effects on oxygenation that were produced by the combination of rotational dosing, perfluorocarbon, and HFOV. Such a relationship was also noted by Jeng et al. (15), who, while quantifying PFB evaporative loss during conventional mechanical ventilation, demonstrated a positive correlation between PaO2 and residual intrapulmonary PFB volume (15). Indeed, there appeared to be a drop in oxygenation below a critical volume of PFB remaining in the lungs, suggesting a role for monitoring PFB vapor loss during PLV. Furthermore, we suspect that the optimum residual PFB volume that should be maintained in the lungs (as does the need for PEEP) may both vary with the severity of lung injury between patients as well as vary over time during the course of illness; precise titration of intrapulmonary PFB volume in the clinical setting would necessitate monitoring of PFB vapor loss. There were no other differences noted in PaCO2, arterial pH, or hemodynamic variable between groups.
We did note a significant improvement in the histopathological outcomes in the animals treated with the combination of perfluorocarbon and HFO. In the dependent lung, the degree of atelectasis was significantly lower in both perfluorocarbon-treated groups. Atelectasis and "atelectrauma," produced by the repetitive reopening of lung units that close at end-expiration in dependent lung, are important considerations in the design of lung-protective ventilatory strategies (33). The combination of HFOV and perfluorocarbon appears to reverse atelectasis more efficiently than high-frequency ventilation alone. In the nondependent lung, the rotationally dosed animals had a significantly lower incidence of alveolar necrosis compared with the other two groups. When the lung injury data for the whole lung were analyzed, both perfluorocarbon-treated groups had less alveolar distension and less atelectasis compared with the HFO-treated controls. This is an important finding because one of the concerns when using aggressive increases in mean airway pressure to achieve an optimal lung volume during HFO is the overdistension of more compliant lung regions. This phenomenon appears to be mitigated by the combination of perfluorocarbon and HFOV. When the components of the lung injury score were combined, the improved histopathological outcomes were most evident in the dependent lung, which likely relates to the gravity-dependent distribution of the perfluorocarbons in the lung (4, 26).
Neutrophil-mediated lung injury is an important and early step in the sequence of acute lung injury (17, 18). Both HFOV and PLV have been associated with decreased pulmonary migration of neutrophils (18, 30, 36). In our study, myeloperoxidase activity, which represents the presence of the neutrophil enzyme, was also shown to be reduced in the perfluorocarbon-treated groups, and the reductions in myeloperoxidase activity were evident in both dependent and nondependent lung in both perfluorocarbon-treated groups. We also indexed myeloperoxidase activity to protein content in each sample, which may have reduced the variability of the assay (31) and enhanced our ability to detect lung protection during the combination of HFO and PLV.
We also examined vapor loss rates in both groups of perfluorocarbon-treated animals using an IR device that sampled gas in the expiratory limb of the ventilator circuit. Not surprisingly, the animals managed in the supine position had a higher cumulative vapor loss rate than the rotationally dosed animals. This likely relates directly to the fact that the supine animals were managed with regular replacement of evaporative perfluorocarbon losses and thus received a higher cumulative perfluorocarbon dose. Replacement dosing of perfluorocarbon compounds has been noted by others to produce increased perfluorochemical saturation of the expired gas and increased hourly loss rates, which may be cumulative (22). In our study, the animals managed with rotational dosing were noted to have a higher hourly vapor loss rate during the first 3 h of measurement. In fact, the vapor loss rate measured in these animals (peak 2.8 ml · kg-1 · h-1) is significantly higher than has been reported in animals managed with conventional mechanical ventilation in the supine position (0.8-1.2 ml · kg-1 · h-1) (23). This is most likely related to the high degree of alveolarization achieved in the rotationally dosed animals, which resulted in a large air-liquid interface and greatly increased vapor loss rate (32). The vapor loss rate measured in the animals treated with perfluorocarbon in the supine position ranged between 1.0 and 1.5 ml · kg-1 · h-1, with the higher vapor loss rates noted in the later phases of the protocol. In these animals, the vapor loss rate may have been influenced by regular replacement of perfluorocarbon on the basis of measured losses that resulted in a stable rather than diminishing gas-liquid surface area. Because hourly vapor loss rate appears to be related to the size of the gas-liquid interface, quantification of perfluorocarbon in expired gas may provide useful information regarding the degree of alveolarization of the drug.
A number of investigators have examined the combination of high-frequency ventilation and perfluorochemical administration in animal models of lung injury. Smith and colleagues (34) described the effects on gas exchange of PFB administration in combination with a variety of high-frequency devices in neonatal piglets treated with saline lavage. In an important follow-up study, the same group of investigators examined the effect of combined high-frequency ventilation and PFB administration on lung pathology in the same model (35). There was evidence of enhanced lung protection in both dependent and nondependent lung regions.
Sukumar and coworkers (37) described the changes in hemodynamics and gas exchange during high-frequency ventilation of PFB-treated preterm lambs compared with a control group ventilated with HFO. Histopathological analysis in a small subgroup of the subjects suggested increased alveolar air spaces and preservation of alveolar morphometry in the animals managed with a combination of PFB and HFO (37). The interpretation of these data is complicated by the absence of attempts to optimize lung volume in each animal and the use of the same mean airway pressures in PFB-treated and non-PFB-treated animals.
In a recent study in larger animals, we described the effects of PFB administration during HFOV on gas exchange, hemodynamic function, and lung histopathology (8). Healthy swine underwent repetitive saline lavage and were then randomized to HFO or combined HFO-PLV. Lung volume was not optimized in either treatment group, and the HFO-PLV animals received a dose of 30 ml/kg of PFB. There were no differences in the two groups of animals regarding gas exchange or hemodynamic function. However, the animals treated with the combination of HFO and PFB manifested significantly less severe atelectasis after 2 h (8).
In the present study, we attempted to extend our preliminary findings and included a rotationally dosed group to elucidate potential benefits of maximal alveolarization of PFB (4). We ensured a uniform and profound degree of lung injury by repetitive saline lavage to a physiological end point, followed by 4 h of injury amplification using large VT values and low-end expiratory pressures. We acknowledge that all animal models of lung injury have their shortcomings when attempting to elucidate the complex sequence of events that occur during mechanical ventilation of the acutely injured lung (29). We also chose to emulate the use of HFOV in the clinical setting by manipulation of the mean airway pressure to achieve an optimal PaO2-to-FIO2 ratio, the so-called "optimal lung volume strategy" (3). Our study protocol included 15 h of mechanical ventilation after the induction of lung injury and perfluorocarbon dosing to maximize our ability to detect differences in histopathological outcomes between the modes of ventilatory support examined in this study. Finally, we used a robust, eight-variable histopathological scoring system with the variable of interest scored for the entire lung section sampled to assess regional differences in lung injury by evaluation of both dependent and nondependent lung regions.
HFO of the perfluorocarbon-treated lung is an evolving story; in particular, there are insufficient data regarding histopathological findings during the combination of HFO and partial liquid breathing at varying doses of PFB. We found salutary effects in gas exchange, although unsustained, after uniform distribution of PFB and reduced indexes of lung inflammation in both PLV groups during HFO-PLV in our extended protocol; similar benefit in gas exchange, mechanics, and indexes of inflammation were also found for PLV during conventional ventilation that did not translate into benefit during clinical trials. Our present data provide preliminary information that require further study in guiding optimal lung-protective strategies in the management of acute lung injury as well as providing insight regarding the utility of quantifying vapor loss as an estimate of the distribution of perfluorocarbon in the injured lung.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. L. Bull, S. Tredici, E. Komori, D. O. Brant, J. B. Grotberg, and R. B. Hirschl Distribution dynamics of perfluorocarbon delivery to the lungs: an intact rabbit model J Appl Physiol, May 1, 2004; 96(5): 1633 - 1642. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
