|
|
||||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Infant Pulmonary Research Center (Drs. Manaligod, Bendel-Stenzel, and Mammel, Ms. Meyers, and Mr. Bing) Childrens Hospital and ClinicsSt. Paul, MN; and the Departments of Pediatrics (Drs. Manaligod, Bendel-Stenzel, and Mammel) and Biostatistics (Dr. Connett), University of Minnesota, Minneapolis, MN.
Correspondence to: Mark C. Mammel, MD, Department of Neonatal Medicine, Childrens Health CareSt. Paul, 345 N Smith Ave, Room 2100, St. Paul, MN 55102;
| Abstract |
|---|
|
|
|---|
Design: Prospective animal study.
Setting: Animal physiology research laboratory.
Subjects: Nine piglets.
Interventions: Animals underwent saline solution lavage to produce lung injury. Perflubron was instilled via the endotracheal tube in a volume estimated to represent functional residual capacity. The initial PEEP setting was 4 cm H2O, and stepwise changes in PEEP were made. At 30-min intervals, the PEEP was increased to 8, then 12, then decreased back down to 8, then 4 cm H2O.
Measurements and results: After 30 min at each level of PEEP, arterial blood gases, aortic and central venous pressures, heart rates, dynamic lung compliance, and changes in EELV were recorded. Paired t tests with Bonferroni correction were used to evaluate the data. There were no differences in heart rate or mean BP at the different PEEP levels. CO2 elimination and oxygenation improved directly with the PEEP level and mean airway pressure (Paw). Compliance did not change with increasing PEEP, but did increase when PEEP was lowered. EELV changes correlated directly with the level of PEEP.
Conclusions: As previously reported during gas ventilation, oxygenation and CO2 elimination vary directly with PEEP and proximal Paw during PLV. EELV also varies directly with PEEP. Dynamic lung compliance, however, improved only when PEEP was lowered, suggesting an alteration in the distribution of perflubron due to changes in pressure-volume relationships.
Key Words: lung compliance lung injury perfluorocarbon positive end-expiratory pressure ventilation
| Introduction |
|---|
|
|
|---|
Liquid ventilation involves the instillation of a perfluorocarbon endotracheally to provide respiratory support. Perfluorocarbon is a chemically and biologically inert compound that has a low surface tension and a high solubility for oxygen and carbon dioxide.6 Partial liquid ventilation (PLV) involves using perfluorocarbon in conjunction with either a conventional mechanical ventilator7 or with high frequency ventilation.8 The instillation of perfluorocarbon into the lungs opens collapsed alveoli and increases EELV.9 Because it is a noncompressible liquid, PFC acts as a kind of "liquid PEEP" and prevents collapse of liquid-filled alveoli. Thus, the role of PEEP during PLV is not clear. Previously, a certain PEEP level was set to clear the proximal airways of perfluorocarbon and to prevent reflux.10 This study was designed to investigate the effects of PEEP during PLV on gas exchange and lung volume. We hypothesized that because perfluorocarbon mechanically maintains EELV, the application of PEEP would not further improve lung compliance, oxygenation, or EELV.
| Materials and Methods |
|---|
|
|
|---|
Physiologic Measurements
Before lung injury, after lung injury, and after 30 min at each
level of PEEP, the following were recorded: arterial blood gases,
intravascular pressures, and heart rates (model 521 neonatal monitor;
Spacelabs, Inc; Redmond, WA); dynamic respiratory system compliance
(VenTrak 1550 Respiratory Mechanics Monitor; Novametrix Medical
Systems; Wallingford, CT); and ventilator settings.
Inductance coils were placed around the chest and abdomen of each animal. As previously described, changes in EELV were obtained using the sum signal of respiratory inductive plethysmography (Respitrace Systems; Ambulatory Distributing Co; Ardsley, NY) in the directcurrent-coupled mode.12 13 14 The Respitrace system is a respiratory inductive plethysmograph that can be used to provide calibrated outputs corresponding to rib cage compartment volume changes, abdominal compartment changes, and the total volume changes associated with respiration. Changes in the impedance of the coils are proportional to changes in the volume of the chest and abdomen, and the sum of the two impedances is proportional to lung volume.15 16 Measurements in the changes in EELV were made after 30 min at each PEEP level.
Lung Injury Model
Pulmonary compromise was induced by repeated normal saline
solution lavage. This was accomplished by filling the lungs with normal
saline solution until a meniscus was seen in the endotracheal tube. The
saline dwelled in the lung for 3 to 5 min while the peak inspiratory
pressure (PIP) was automatically adjusted to maintain a VT
of 15 mL/kg. The animals were rotated after each lavage to ensure
uniform lung injury. The animals qualified when
PaO2 was < 100 mm Hg in an
FIO2 of 1.0, and if there was a
30% reduction in compliance.
Experimental Protocol
After initial instrumentation and stabilization, we measured
physiologic parameters, arterial blood gases, and lung compliance.
After lung injury was induced using the above surfactant washout
technique, measurements were again recorded. Each piglet received
perflubron (LiquiVent; Alliance Pharmaceutical Corp; San Diego, CA).
The perflubron was warmed to room temperature and preoxygenated by
bubbling oxygen at 3 to 4 L/min through the liquid for 1 min prior to
instillation. Perflubron was incrementally instilled into the trachea
using the side port of the endotracheal tube until a meniscus was
visible at end expiration and zero PEEP. This technique approximates
the animals functional residual capacity (FRC).7
During
instillation, the animal was disconnected from the ventilator, and a
volume of 5 mL was given as a bolus down the endotracheal tube. The
animals were rotated bidirectionally in the lateral decubitus and prone
positions to provide appropriate distribution. Repeated boluses were
given until a meniscus was visible at the end of the endotracheal tube.
During and immediately following the administration of perflubron, the
PIP was automatically adjusted to maintain a VT of 15
mL/kg, PEEP was maintained at 4 cm H2O, and the
rate was adjusted to keep PaCO2
within 35 to 55 mm Hg. Animals were stabilized for 1 h after the
instillation of perflubron. Animals were then paralyzed with
pancuronium (0.2 mg/kg/dose), and stabilized for an additional 1
h. Replacement doses for evaporative losses were not given. Animals
received VTtargeted, time-cycled positive pressure
ventilation with the Dr
ger Babylog. After instilling perflubron,
the initial settings were as follows: PEEP, 4 cm
H2O; VT, 15 mL/kg; inspiratory time
adequate for gas flow to return to zero; and an
FIO2 of 1.0. Stepwise changes in
PEEP, at increments of 4 cm H2O, were made at
30-min intervals. The PEEP was increased to 8 and then 12 cm
H2O, then decreased back down to 8 and 4 cm
H2O. PIP was adjusted to maintain VT
at 15 mL/kg. Ventilator rate was adjusted to maintain similar minute
ventilation per kilogram (
E/kg) throughout the
different PEEP levels. Sodium bicarbonate was given when the base
deficit was > -8 Meq/L.
Statistical Analysis
The primary physiologic variables were heart rate, mean BP,
central venous pressure (CVP), pH,
PaCO2, oxygenation index (OI),
arterial/alveolar (a/A) gradient, mean airway pressure (Paw),
respiratory system compliance, and changes in FRC. OI is calculated
with the following equation:
OI = Paw x FIO2 x 100/PaO2 The a/A gradient is calculated as follows:
a/A gradient = PaO2/[(PB - 47) x FIO2)] - (PaCO2/R)
where PB is barometric pressure and R is the respiratory quotient.
Data were analyzed with paired t test comparisons using Bonferroni correction. Values were compared with the preceding values. Values at the highest PEEP level (12 cm H2O) were compared with the two lowest PEEP settings (initial and final settings, 4 cm H2O). Values at the end of the study were compared with the original baseline values at the beginning of the study. A p value of < 0.05 was accepted as statistically significant.
| Results |
|---|
|
|
|---|
EELV after manipulating PEEP are shown
in Table 2
.
|
|
Gas Exchange
pH varied directly and PaCO2
varied inversely with the PEEP level (Fig 1
). There were no significant differences in ventilator rate or
VT because VT was kept constant at 15 mL/kg.
E/kg was similar as PEEP was increased, but was
slightly lower when PEEP was decreased from 12 cm
H2O to 8b (p < 0.04).
E/kg
at 4b was lower than the initial value at 4a (p < 0.03).
PaCO2 significantly decreased when
the PEEP was raised from 4 to 8 cm H2O
(p < 0.04), and the value at 12 cm H2O was
significantly lower than the baseline value at 4a (p < 0.003). There
were also significant increases in
PaCO2 as PEEP was lowered. The value
at 8b was different from that measured at 12 cm
H2O (p < 0.002), 4b was different from 8b
(p < 0.0001), and 4b was different from 12 cm
H2O (p < 0.0001). pH also varied with PEEP,
and was a reflection of the changes in
PaCO2. The a/A gradient varied
directly with the PEEP level (Fig 2
). There were significant increases in the a/A gradient when PEEP was
increased from 4 cm H2O to 8a (p < 0.03) and
to 12 cm H2O (p < 0.03). There were also
significant decreases in a/A gradient when PEEP was lowered from 12 cm
H2O to 8b (p < 0.003), and from 8b to 4b
(p < 0.002). The value at 4b was also different from that measured
at 12 cm H2O (p = < 0.0001). The a/A gradient
when PEEP was lowered back down to 4 cm H2O at 4b
was significantly lower than the original baseline value at 4a
(p = 0.0012. Because PEEP is a major determinant of Paw, there were,
as expected, significant changes in Paw as well (Fig 2)
. There were
significant differences in Paw at all PEEP levels and for all
comparisons. As was previously seen with a/A gradient, when PEEP was
lowered back to 4 cm H2O (4b), the Paw was
significantly lower than the baseline value at 4a (p < 0.05). There
were no significant changes in OI until the PEEP was lowered back down
to 4 cm H2O (Fig 3
). This was significantly higher than the baseline value at 4a
(p < 0.04).
|
|
|
|
|
|
| Discussion |
|---|
|
|
|---|
In this study, we compared gas exchange, dynamic lung compliance, hemodynamics, and changes in EELV during PLV at different levels of PEEP in an animal lung injury model. In contrast to our hypothesis, we have shown that the application of PEEP, in the range we studied, has similar effects during PLV to those seen during conventional gas ventilation. As PEEP increased, gas exchange improved and EELV increased; as PEEP decreased, the opposite effects occurred.
Our gas exchange results are consistent with findings from large animal studies.18 19 Kirmse and colleagues18 demonstrated that applying a PEEP level 1 cm H2O above the value needed to define the lower inflection point on a subjects volume-pressure curve improves gas exchange when using FRC amounts of perflubron. Kaisers et al19 used CT to show that the administration of PEEP during PLV resulted in a spreading of perfluorocarbon from the tracheobronchial system into the lung periphery. We speculate that the air spaces that are only partially filled with perflubron would become more completely filled with higher PEEP, resulting in further recruitment of partially opened alveoli, an increase in EELV, and an improvement in ventilation/perfusion matching. Decreasing PEEP, however, would shift perflubron back to the tracheobronchial system, which would result in lower EELV and less effective gas exchange.
Dynamic lung compliance did not change when PEEP was increased, consistent with the findings of Kaisers et al.19 In our study, dynamic lung compliance did significantly increase when PEEP was lowered. One explanation for this result relates to the hysteresis of the lung. As PEEP levels increase in a stepwise fashion, the effect is to define the inspiratory limb of the pressure-volume relationship of the lung. As PEEP is then similarly reduced, dynamic lung compliance measurements then reflect the deflation limb of the curve, with its increased volume for a given pressure. As respiratory efforts occur on this portion of the ventilation/perfusion relationship, the C/kg measured for such breaths will produce higher values. Oxygenation, on the other hand, decreased as would be expected at the lower PEEP and Paw. This effect likely reflects a redistribution of perflubron into the nondependent regions of the lung and the larger proximal airways. Further studies are needed to better resolve this issue.
The effects of PEEP on oxygenation that we observed during PLV are in contrast to the findings of Wolfson et al,20 who investigated the effects of PEEP during total liquid ventilation (TLV). During TLV, the entire alveolar-gas interface is replaced with a liquid-liquid interface. The application of PEEP during TLV did not improve oxygenation. This suggests that in the partially filled lung, where alveoli-gas interfaces existespecially in the nondependent regions of the lungrecruitment of alveoli and improvements in gas exchange can be achieved with PEEP.
At the end of our study, when PEEP was lowered back down to the original value of 4 cm H2O, a/A gradient, OI, and Paw were all significantly different from the baseline values at the beginning of the study. These observations may be a result of perflubron loss due to evaporation, evolution of the lung injury itself, and the improved compliance noted at the end of the study. We chose not to replace hourly losses because any small movement of the animals would affect the recordings from the inductive plethysmograph. The evaporative loss of perflubron, along with improved compliance and lower Paw delivered by the ventilator, may account for the decreased oxygenation seen at the end of the study.
There were other potential problems that could have affected the
results of our study. As previously mentioned, we chose not to replace
evaporative losses of perflubron. Although this was a short-term study,
an evaporative loss of approximately 2 mL/kg/h was
likely.8
This might be sufficient to affect gas exchange
and
EELV data as liquid content of the lung also affects the
baseline of the inductive plethysmograph. However, given that the study
was only 2.5-h long, we felt that the effects of evaporation on the
baseline would be minimal. We also chose to measure only dynamic
compliance and not include static compliance measurements. Static
compliance might have been preferable, as this more accurately reflects
the elastic properties of the lung, while dynamic compliance also
depends on inspiratory flow, airway resistance, inertance of the
medium, and equilibration between proximal and distal airways.
Manipulations required for measurement of static compliance would have
disrupted our measurement of respiratory inductive plethysmography.
The design of the study raises possible limitations. A true control group (no PLV or zero PEEP) was not included in the design. The initial PEEP setting of 4 cm H2O was chosen instead of zero PEEP to prevent reflux of perflubron back up the endotracheal tube, which might interfere with the flow sensor. Stepwise increases were made in PEEP, as in the work by Suter et al2 with conventional gas ventilation, and stepwise decreases in PEEP were made to determine if perflubron had any stabilizing effect on EELV once recruitment was achieved. We also did not include a separate group of animals that did not receive perflubron because previous studies have investigated the effects of PEEP during conventional gas ventilation.2 3 4 5
Cox et al21 observed that air leak and pneumothoraces occurred more frequently in animals receiving PLV and large VT compared with animals receiving PLV and normal VT. Although we did not perform autopsies or obtain chest radiographs, we did not observe any clinical evidence of air leak in any of the animals in the short-term study. We chose appropriate VT for piglets, and did not see any overdistention based on pressure-volume curves. High airway pressures during PLV should be used cautiously. Further studies are needed to determine the association of air leaks with PLV.
In conclusion, in a surfactant-deficient lung injury model, our study demonstrates that the application of PEEP during PLV has physiologic effects on gas exchange similar to those seen during gas ventilation, without obvious adverse cardiovascular effects. This study further suggests that the concept of "liquid PEEP" during PLV is not entirely valid, because improvements in both gas exchange and lung volume occur with the application of PEEP in the partially liquid-filled lung.
| Footnotes |
|---|
E/kg = minute ventilation per kilogram;
VT = tidal volume Supported in part by a grant from the Childrens Hospitals and Clinics Foundation. Liquivent was supplied by Alliance Pharmaceutical Corp., Dräger Babylog supplied by Dräger Critical Care Systems Inc. Dr. Mammel and Ms. Meyers own stock in Alliance Pharmaceutical Corp.
Received for publication April 6, 1999. Accepted for publication August 18, 1999.
| References |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
S.A. Loer, D. Kindgen-Milles, and J. Tarnow Partial liquid ventilation: effects of liquid volume and ventilatory settings on perfluorocarbon evaporation Eur. Respir. J., December 1, 2002; 20(6): 1499 - 1504. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |