| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Guest Access | Sign In via User Name/Password
|
|
|
|||||||
Guest Access | Sign In via User Name/Password |
|||||||||
* From the Department of Clinical Physiology (Drs. Maurenbrecher and Hedenstierna), University Hospital, Uppsala, Sweden; Department of Anesthesia and Intensive Care (Drs. Lamy and Deby-Dupont), University Hospital, Liège, Belgium; and Department of Anesthesiology (Dr. Frascarolo), University Hospital, Lausanne, Switzerland.
Correspondence to: Göran Hedenstierna, MD, Department of Clinical Physiology, University Hospital, S-75185 Uppsala, Sweden
 |
Abstract |
|---|
 TOP Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Design: Animal study.
Setting: Experimental laboratory in a university hospital.
Interventions and measurements:
We studied eight pigs under general anesthesia (mean weight, 26.2 kg)
receiving mechanical ventilation adjusted to normocapnia, with a
fraction of inspired oxygen (FIO2) of 0.5 to
1.0. Blood gases, endotoxin concentration, and central hemodynamics
were measured hourly, and ventilation-perfusion (
/
)
relationships were assessed by multiple inert gas elimination technique
before and after inhalation of NO. NO was delivered at 40 ppm for 10
min at 4 h and 6 h of endotoxin exposure.
Results: Seven of eight pigs were responders to NO at
4 h, defined as a 
20% increase in oxygenation index
(PaO2/FIO2)
[223 ± 43 to 330 ± 56 mm Hg; p = 0.001]. The same pigs
exhibited a 20% fall in mean pulmonary artery pressure
(39.4 ± 2.2 to 30.0 ± 2.1 mm Hg; p < 0.001). The response
correlated to the perfusion to "normal /" regions
(r = - 0.82) and negatively to shunt and dead space
ventilation (r = 0.76 and r = 0.87,
respectively). At 6 h, seven of eight pigs were nonresponders,
despite unaltered hemodynamics and gas exchange. Correlations at 4
h between physiologic variables and response to NO were abolished. The
logarithmic SDs of the perfusion distribution, a measure of the degree
of / mismatch, increased significantly from 4 to 6 h
(p = 0.04).
Conclusion: Response to inhaled NO is
abolished over time in endotoxin-induced ARDS pig lungs. The response
seems to be related to the degree of / mismatch, which
may indicate an important role of hypoxic pulmonary
vasoconstriction.
Key Words: endotoxin • lung injury • nitric oxide • pig
|
Introduction |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
IV vasodilators such as nitric oxide (NO) donors and prostaglandins reduce pulmonary hypertension but might worsen arterial oxygenation by increasing the blood flow to nonventilated areas and have undesirable hypotensive systemic effects.2 Exogenous and endogenous NO acts by stimulating soluble guanylate cyclase, which increases intracellular concentration of guanosine 3',5'-cyclic monophosphate in the smooth-muscle cell and causes relaxation.3 Endogenous NO, produced by the endothelium by a constitutive NO synthetase, contributes to the basic vascular tone. Inducible NO synthase (iNOS) is stimulated by endotoxin and cytokines, producing large amounts of NO and inhibiting hypoxic pulmonary vasoconstriction (HPV).4 5
Inhaled NO, by its unique route of administration and extremely short half-life, selectively dilates vessels in ventilated lung units. This results in redistribution of lung blood flow, reduction of shunt with improvement of arterial oxygenation, and a decrease in mean pulmonary artery pressure (MPAP).6 However, the initial enthusiasm for its clinical application in patients with ARDS has been tempered by its inconsistent efficacy. Clinical studies7 8 9 10 have shown an effectiveness of 40 to 60% in terms of improvement in arterial oxygenation. Attempts to identify predictive factors for the gas exchange response to inhaled NO in patients with ARDS, such as the level of shunt, oxygenation index (PaO2i) [PaO2/fraction of inspired oxygen (FIO2)], pulmonary vascular resistance (PVR), and cardiac output prior to NO inhalation, have yielded inconclusive results.10 However, septic shock is a condition associated with less response to inhaled NO.8 11
The knowledge about the mechanisms of nonresponse is limited. Most animal studies12 13 testing the effect of inhaled NO have had an observation period of 3 to 4 h, and the nonresponse phenomenon has not been reported except in two studies13 14 of oleic acid-induced ARDS, where the histologic analysis of the nonresponder animals revealed more severe and diffuse lung damage. Hyporesponsiveness in terms of a decrease in MPAP has been reported in isolated lung.15
Based on these observations and of the clinical experience that NO
inhalation is often introduced in a later phase of ARDS, we tested the
degree of response to inhaled NO delivered at two different times (4 h
and 6 h) after the induction of lung injury by endotoxin. In
previous experimental studies,12
13
animals were found to
be responders in an early phase (3 to 4 h). We hypothesized that
the progression of the injury will lead to nonresponse at 6 h. We
therefore investigated the question of whether the response to NO
inhalation in an endotoxin porcine model changed over time with regard
to hemodynamic and blood gas parameters and to ventilation-perfusion
(/) distribution.
|
Materials and Methods |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
Anesthesia
The animals were brought to the laboratory in a heated car the
morning of the experiment. General anesthesia was induced with IM
atropine, 0.04 mg/kg; tiletamine-zolazepam, 6 mg/kg; and xylazine, 2.2
mg/kg. After loss of the eyelid reflex, the pig was placed in a supine
position on the operation table, a venous line was introduced into an
ear vein, and the trachea was intubated with a 7-mm inner-diameter
tube. IV fentanyl, 200 µg, and pancuronium, 0.2 mg/kg, were
administered, and mechanical ventilation was started in a volume-cycled
mode (Servo 900C, Siemens-Elema AB; Lund, Sweden). At baseline, the
ventilator settings were as follows: tidal volume (VT) of
12 mL/kg at 20 breaths/min, inspiration to expiration ratio of 1:2, and
positive end-expiratory pressure of 5 cm
H2O. The FIO2
was 0.5 in five animals and 1.0 in three animals throughout the
experiment. The VT was adapted during the experiment to
attempt normoventilation, as guided by the end-tidal
CO2 tension (35.9 ± 5.2 mm Hg) and by
intermittent PaCO2
measurements (45.9 ± 0.9 mm Hg). Anesthesia was maintained by
continuous infusion of clomethiazole, 400 mg/h; fentanyl, 150
µg/h; and pancuronium, 2.5 mg/h. Additional fentanyl and pancuronium
were administered if needed. Ringer’s lactate at body temperature
(1,000 mL) was infused prior to baseline measurements, and the infusion
was thereafter adjusted to yield a stable systemic arterial pressure
and hemoglobin concentration (8.43 ± 0.11 g/dL).
Catheterization
An arterial line was inserted into the left carotid artery
together with a thermistor-tipped catheter, which was advanced into the
descending aorta for measurement of cardiac output. A Swan-Ganz
catheter and an 18-gauge catheter were introduced into the right
jugular vein, and their positions were confirmed by the corresponding
pressure curves. A silicone-coated balloon catheter (Ch 20; Rüsch
AG; Kernen, Germany) was inserted into the bladder, and peritoneal
fluid was drained by a multihole catheter.
Measurements
Mean systemic arterial pressure (MAP), MPAP, central venous
pressure (CVP), and pulmonary capillary wedge pressure (PCWP) with zero
reference at the level of the anterior third of the thoracic wall, and
oxyhemoglobin saturation were displayed continuously (series 7010 Tram;
Marquette Electronics; Milwaukee, WI). CVP and PCWP values were
recorded at the end of expiration.
Cardiac output was measured by injection of 8 mL of cold glucose 5%. The bolus was injected into the right atrium at the end of expiration. The dilution curves for temperature were recorded in the aorta with the thermistor-tipped catheter connected to a cardiac output computer. Two measurements were made and averaged; three measurements were made and averaged if the first two measurements differed by > 10%.
Arterial and mixed venous blood samples were analyzed for pH, PO2, PCO2, O2 saturation, hemoglobin, and methemoglobin (models ABL 300 and OSM 3; Radiometer; Copenhagen, Denmark). PaO2i was used for statistical analysis.
Arterial blood samples were obtained for cell counts and measurement of
albumin and plasma endotoxin concentrations; the blood samples at
4 h and 6 h were drawn just before NO inhalation. For
endotoxin measurements, blood was drawn into pyrogen-free Falcon tubes
(Becton Dickinson; Franklin Lakes, NJ), and the assay was
performed with a quantitative end point chromogenic method (Coatest;
Chromogenix AB; Mölndal, Sweden), with Escherichia
coli 0111:B4 reference endotoxin for the standard curve obtained
with pig serum. The value was expressed in endotoxin units (EU) per
milliliter, and the lowest limit of detection was 0.06 EU/mL. The
/ distribution was determined at 4 h and 6 h
after induction of lung injury by the multiple inert gas elimination
technique (MIGET). This technique is based on the elimination of six
inert gases according to their different blood-gas solubility
coefficients (sulfur hexafluoride, ethane, cyclopropane, enflurane,
diethyl ether, acetone). At steady state, after 40 min of continuous IV
infusion (2 to 3 mL/min, depending on minute ventilation) of these
gases diluted in glucose 5%, arterial and mixed venous blood samples
were collected together with an expired gas sample and analyzed by gas
chromatography (5890, series II; Hewlett-Packard; Wilmington,
DE). These data allow the construction of a virtually continuous
distribution of blood flow and ventilation against different
/ ratios: shunt (/ < 0.005), low
/ (0.005 < / < 0.1), normal
/ (0.1 < / < 10), and high
/ areas (10 < / < 100), and dead
space (/ > 100). The dispersion of /
ratios is expressed as the logarithmic SDs of perfusion distribution
(logSDQ) and describes the degree of / mismatch (not
including shunt).
End-tidal CO2 and FIO2 were monitored continuously (Capnomac Ultima; Datex Instrumentarium; Helsinki, Finland). Compliance and resistance of the total respiratory system were determined during an inspiratory-hold maneuver (approximately 4 s). Resistance was calculated as the difference between peak airway pressure and the pressure at 2 s of the end-inspiratory pause, divided by the inspiratory flow. Compliance was calculated as VT divided by end-inspiratory pressure minus end-expiratory pressure. Airway pressures and flow were measured in the ventilator on the inspiratory side and recorded with a personal computer for on-line signal processing, taking into account the gas compression within the ventilatory circuit. The mean of two inspiratory-hold maneuvers was used for statistical analysis (LabVIEW 3.1; C-O Sjöberg Engineering; National Instruments). The same monitoring was used for VT, minute ventilation, respiratory rate, and airway pressures.
Protocol
After a stabilization period of 1 h and baseline
measurements, lung injury was induced by endotoxin infusion (30
µg/kg/h; Escherichia coli lipopolysaccharide O111:B4;
Sigma-Aldrich; Stockholm, Sweden) administered via a peripheral venous
line over 2 h. The dose of 30 µg/kg/h over 2 h that was
used in the study is based on prior experiments in pigs.16
The animals were submitted to an observation period of 6 h with
two challenges of inhaled NO of 10 min each: the first challenge at
4 h and the second challenge at 6 h after the start of
endotoxin infusion.
NO was delivered from a 1,000 ppm gas tank (AGA Gas AB; Lidingö, Sweden) together with air/oxygen from a low-flow air-oxygen blender (AGA AB; Sundbyberg, Sweden) into the low-pressure gas-flow inlet of the ventilator. The NO delivery was adjusted by means of a calibrated flowmeter yielding an inspiratory concentration of 40 ppm, which was controlled by an NO chemiluminescence analyzer (9841 NOx; Lear Siegler Measurement Controls Corporation, Englewood, CO) connected to the inspiratory limb of the respiratory circuit. After 10 min of NO inhalation, circulatory and ventilatory measurements were made as described above. At the end of the experiment, the pig was killed with an IV injection of potassium chloride (20 mg).
Hemodynamic parameters including MAP, MPAP, cardiac output, and blood
gases were measured at baseline, at 30 min, and thereafter every hour,
as well as after 10 min of NO inhalation. CVP, PCWP, compliance, and
resistance were measured every hour. Blood sampling for cell counts and
measurements of albumin and plasma endotoxin levels was performed
hourly. MIGET evaluations of / distribution were
performed before and after 10 min of NO inhalation under fixed
FIO2 and ventilatory settings. A
response was defined as a 20% increase in
PaO2i compared to the pretreatment
level, in accordance with the report of a recent multicenter study as
well as other studies.8
11
17
Statistical Analysis
Data are given as mean ± SEM if not otherwise indicated. For
each parameter, the normality was tested by the Wilcoxon-Shapiro test.
One-way analysis of variance for repeated measurements was performed to
assess the effect of endotoxin over time. A paired Student’s
t test, or Wilcoxon’s signed rank test if appropriate, was
applied to detect differences due to inhaled NO and to compare the
effects of endotoxin at 4 and 6 h. Bonferroni correction for
multiple comparison was then applied. Simple regression analysis was
carried out with the Pearson coefficient to explain the difference in
response to inhaled NO at 4 h and 6 h. All p values are given
as raw data, and p < 0.05 was considered the level of significance.
|
Results |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
/ results are shown in Table 2
. At baseline, all variables were within normal limits as compared to
values from previous studies in healthy pig lungs.12
|
|
/ regions was minor at 4 h; dead space ventilation
was approximately 50% of the total ventilation, and there was no
ventilation to high / areas. logSDQ, a measure of
/ ratio dispersion, was increased compared to
measurements in normal pigs in previous studies.12
Compliance fell and resistance increased after induction of the lung
injury (Table 1)
.
|
The endotoxin concentration, leukocyte count, and fraction of neutrophils are shown in Figure 2 . The endotoxin concentration increased rapidly after the start of the endotoxin infusion, reaching a plateau between 1 h and 2 h in parallel with a major initial fall in neutrophil count. The endotoxin concentration then decreased and regained almost baseline values at 6 h, whereas the neutrophil count increased (64 ± 5% at 4 h and 75 ± 4% at 6 h; p < 0.001). The thrombocyte count and albumin concentration decreased significantly between baseline and 6 h (389 ± 30 x 103 to 164 ± 10 x 103, and 21.7 ± 0.6 to 11.2 ± 1.1 g/L, respectively), with a significant difference between 4 h and 6 h (p = 0.024 and p = 0.035, respectively).
|
/ ratios (logSDQ). Inhalation of NO also resulted in a
significant decrease in MPAP (Fig 1
, Table 1
). Effect of NO at 6 h: One pig died 6 h after the induction of lung damage, just before the second NO trial, which was thus conducted in seven pigs. Only one pig responded to the second NO trial, according to the definition given above. Thus, at 6 h, NO inhalation had no significant effect either on PaO2i or on shunt. However, MPAP fell, though only to a minor extent, from 36 to 33 mm Hg. The MAP fell from 79 to 75 mm Hg (p = 0.06; Tables 1 , 2 ).
Pooled data from the two NO trials showed that the increase in PaO2i on inhalation of NO was correlated to the logarithmic endotoxin concentration (p = 0.006), and that the decrease of MPAP on NO inhalation was negatively correlated to the logarithmic endotoxin concentration (p = 0.029; Fig 3 ).
|
/"; Fig 4
). The fall in MPAP was negatively correlated to the magnitude of shunt
and to the dead space ventilation (Table 3
). Thus, when taking these correlations together, the results suggest
that inhalation of NO had a better effect on MPAP when /
mismatch was less marked.
|
|
|
/ areas, and dead space ventilation were the same at
4 h and 6 h (Table 2) . |
Discussion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
20%) to
a nonresponder at 6 h (absence or negative response to NO). The
functional lung injury was characterized by shunt, /
mismatch, and increased dead space ventilation. Respiratory mechanics
were impaired, with decreased compliance and increased resistance. A
capillary leak syndrome (decreased albumin concentration) with
generalized edema was present. Endotoxins, lipopolysaccharide components of Gram-negative bacteria, cause an inflammatory injury mediated by activation of alveolar macrophages and neutrophils in the lungs18 19 20 and other organs. This stimulates iNOS to produce endogenous NO.4 Sepsis is one of the situations related to nonresponse, and we therefore chose an endotoxin infusion for inducing an acute lung injury. The differential cell count in the present study reflected a Gram-negative sepsis with leukopenia and thrombocytopenia, similar to that seen in humans. The blood endotoxin concentrations, measured to validate the model, confirm the efficacy of the infusion as well as of the hepatic clearing of the toxin.
The 2 h of endotoxin infusion established a hemodynamically stable model at 4 h and 6 h, and seven of the eight pigs survived the 6-h study period. The animals exhibited lung damage of variable severity (shunt between 8% and 39%, PaO2i between 63 mm Hg and 435 mm Hg) despite the fact that we chose young and pathogen-free animals to avoid endotoxin-primed lungs. It is well-known, however, that the extent of the injury in the lungs as well as in other organs is subject to interindividual and interspecies differences.20 The pigs that received ventilation with FIO2 of 1.0 (n = 3) seemed to be more severely affected; thus, an additional insult by O2 toxicity was possible in these animals.
Response to Inhaled NO
The aim of using a high NO concentration (40 ppm) was to elicit a
response (and avoid false-negative responses) during a short period of
inhalation (10 min) in order not to create an impact on the second NO
trial 2 h later, although this cannot be fully excluded. The
correlation 4 h after the start of endotoxin infusion between the
fall in MPAP induced by NO inhalation and the amount of perfusion and
ventilation going to normal / areas may suggest that
inhaled NO must reach a certain amount of lung tissue to create a
response. logSDQ had increased further from 4 h to 6 h,
indicating a larger dispersion of / ratios. This finding,
without a change in other variables, such as cardiac output, PVR,
PaCO2, and mixed venous
O2 tension between 4 h and 6 h,
may indicate a decline in the strength of HPV.21
Moreover,
at 6 h, there was no longer a response to inhaled NO, and the
close relationship between a decrease in MPAP and an increase in
PaO2i induced by inhaled NO that was
observed at 4 h and that has been reported from clinical
studies6
8
9
11
was also abolished. Thus, it is possible
that blunted HPV is a cause of nonresponse to inhaled NO; loss of
vascular tone should reasonably reduce the effectiveness of a
vasodilator.22
Another observation in the present study was that the response to inhaled NO correlated to the endotoxin concentration in the blood. This finding may have been coincidental and should be interpreted with caution. However, Gust and coworkers23 reported that a low dose of endotoxin added to an oleic acid-induced acute lung injury model created a responder to inhaled NO, whereas oleic acid alone resulted in a nonresponse. This was attributed to increased release of prostacyclin, an effect of endotoxin that has been demonstrated earlier.24
There are also other potential mechanisms underlying the nonresponse to inhaled NO. The endogenous production of NO, stimulated by activation of iNOS via endotoxin and cytokines3 as well as by hypoxia itself, as shown for bovine endothelial cells in culture,25 may counteract the hypoxic vasoconstrictor stimulus26 and contribute to the decreased response to exogenous NO. Free radicals produced by inflammatory cells, especially activated neutrophils and alveolar macrophages, may scavenge the inhaled NO before it has reached the target cell.27 The presence of generalized edema with albumin loss might have thickened the alveolar-capillary membrane and prevented NO from reaching the vascular smooth muscle.
The loss of effectiveness of inhaled NO might also be due to obstruction of blood vessels by microthrombi or by architectural changes in the alveolar-capillary membrane, reducing the area of the vascular bed available for gas exchange, as observed in other studies.28 29 These observations are in concordance with our findings that the response to inhaled NO was related to the amount of normally perfused and ventilated lung.
|
Conclusion |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
/ ratios in the nonresponders
suggests that attenuation of HPV may be another determinant of the
nonresponse phenomenon. Extended studies with comparisons of response
and nonresponders concerning endotoxin levels, vasoactive inflammatory
mediators, and histologic changes in responders and nonresponders are
required to elucidate this problem further.
|
Footnotes |
|---|
/ = ventilation/perfusion; VT = tidal
volume This study was supported by grants from the Swedish Medical Research council (grant 5315), the Swedish Heart and Lung Fund, the Laerdal Foundation, and the AGA Medical Fund.
Received for publication November 30, 1999. Accepted for publication February 21, 2001.
|
References |
|---|
|
TOPAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Gerlach, D. Keh, A. Semmerow, T. Busch, K. Lewandowski, D. M. Pappert, R. Rossaint, and K. J. Falke Dose-Response Characteristics during Long-Term Inhalation of Nitric Oxide in Patients with Severe Acute Respiratory Distress Syndrome: A Prospective, Randomized, Controlled Study Am. J. Respir. Crit. Care Med., April 1, 2003; 167(7): 1008 - 1015. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
