Vol. 85, Issue 2, 642-652, August 1998
Single-exhalation profiles of NO and
CO2 in humans: effect of
dynamically changing flow rate
Nikolaos M.
Tsoukias1,
Ziad
Tannous2,
Archie F.
Wilson2, and
Steven C.
George1
1 Department of Chemical and Biochemical
Engineering and Materials Science and
2 Division of Pulmonary and
Critical Care, Department of Medicine, University of California at
Irvine, Irvine, California 92697-2575
 |
ABSTRACT |
Endogenous
production of nitric oxide (NO) in the human lungs has many important
pathophysiological roles and can be detected in the exhaled breath. An
understanding of the factors that dictate the shape of the NO
exhalation profile is fundamental to our understanding of normal and
diseased lung function. We collected single-exhalation profiles of NO
and CO2 from normal human subjects
after inhalation of ambient air (~15 parts/billion) and examined the
effect of a 15-s breath hold and exhalation flow rate
(
E) on the
following features of the NO profile:
1) series dead space,
2) average concentration in phase
III with respect to time and volume,
3) normalized slope of phase III
with respect to time and volume, and
4) elimination rate at end
exhalation. The dead space is ~50% smaller for NO than for
CO2 and is substantially reduced
after a breath hold. The concentration of exhaled NO is inversely
related to E,
but the average NO concentration with respect to time has a stronger inverse relationship than that with respect to volume. The normalized slope of phase III NO with respect to time and that with respect to
volume are negative at a constant
E but can be
made to change signs if the flow rate continuously decreases during the
exhalation. In addition, NO elimination at end exhalation vs.
E produces a
nonzero intercept and slope that are subject dependent and can be used
to quantitate the relative contribution of the airways and the alveoli
to exhaled NO. We conclude that exhaled NO has an airway and an
alveolar source.
endogenous; exhalation flow rate; phase III slope; elimination
rate
| |
INTRODUCTION |
NITRIC OXIDE (NO) is a highly reactive and
abundant molecule in the body that has many important
physiological roles, including neurotransmission, host defense
response, and smooth muscle relaxation (5). Recently, NO has been
detected in the exhaled breath of humans (2, 21). The concentration of
NO in the exhaled breath increases in inflammatory lung diseases
such as bronchial asthma (15, 22). This finding generated excitement
about a possible use of NO as a biomarker of pulmonary inflammation
(14). In addition, exogenous NO has been shown to selectively
vasodilate the pulmonary vasculature and is capable of modulating
bronchial smooth muscle tone in animals and humans (8-10). The
above findings suggest the possible use of NO as a therapeutic
intervention for a number of pulmonary diseases. However, there is only
limited information on the fundamental gas exchange dynamics of NO in the lungs.
Early studies of exhaled NO demonstrated an increase in NO elimination
from the lungs with hyperventilation and exercise, providing evidence
that alveolar gas contained a nonzero concentration (21). Subsequent
studies demonstrated NO production from the conducting airway space
below the larynx on the basis of direct measurement (24) as well as
after breath hold (16, 21). The most recent studies have reported
significant levels of NO in the nasal cavity (3, 16, 24). In fact, the
production is approximately an order of magnitude larger in the nasal
cavity than in the lower respiratory tract. These studies have
documented that endogenous NO production in the airways and lungs is
heterogeneous and likely plays a variety of roles in normal lung
function.
The heterogeneity in the source of exhaled NO complicates the
interpretation of the exhalation profile. Historically, the exhalation
profile of the respiratory gases,
CO2 and
O2, and other inert gases, such as
He and N2, has provided great
insight into the gas exchange dynamics of the specific gas and the
lungs, in general. The shape of the NO exhalation profile has not been
fully characterized. Recently, NO concentration was shown to be
inversely related to exhalation flow rate
(E) (28). This
result was also consistent with conducting airway production of NO and
highlighted the importance of further study into the dynamics of NO gas
exchange. The objective of our present study is threefold:
1) to quantitatively characterize
the features of the NO oral exhalation profile,
2) to further explore the dependence
of the NO exhalation profile on
E, and
3) to develop an experimental
technique that can provide an estimate of the relative contribution of
the airways and the alveoli to exhaled NO.
| |
METHODS |
Subjects.
We collected oral exhalation profiles of NO and
CO2 in seven normal men
[28.4 ± 3.8 (SD) yr] with no history of smoking or lung
diseases. The protocol was approved by the Institutional Review Board
at the University of California, Irvine. Subjects were categorized as
normal on the basis of standard spirometry that included forced vital
capacity, forced expiratory volume in 1 s, and forced expiratory flow
between 25 and 75% of the exhaled volume. All subjects included in the
study had spirometry values >75% of predicted on the basis of their
race, age, weight, and height (Table 1).
Isolation of nasal cavity.
Several researchers have reported the nasal cavity as a significant
source of NO in the respiratory tract (3, 16, 24) that is capable of
contaminating the oral exhalation profile. Hence, to collect a true
oral exhalation profile to describe the exchange dynamics of NO in the
lower respiratory tract, the nasal cavity must be isolated from the
lower respiratory tract. To achieve this, we applied a small negative
pressure (
20 cmH2O) to the nasal cavity. Each subject was then instructed to swallow, which elevated the soft palate and sealed the nasal cavity. Closure of the
soft palate was confirmed by monitoring the nasal cavity pressure and
maintaining a nasal concentration of
CO2 of <0.5% during expiration.
Any maneuvers where movement of the soft palate was detected (change in
nasal cavity pressure) were discarded and repeated. All exhalation
profiles had a constant negative nasal pressure between 10 and
20 cmH2O. Under these
conditions, all subjects maintained a nasal
CO2 concentration <0.5% during the course of the exhalation maneuver.
The negative nasal cavity pressure provides several advantages over
other reported methods. 1) Elevation
of the soft palate requires no special effort from the subject. In
contrast, balloon occlusion is invasive, and voluntary closure of the
soft palate requires subject training.
2) Minor openings of the soft palate will result in a convective flow into the nasal cavity and will not
contaminate the oral exhalation. 3)
By recording the pressure in the nasal cavity, even minor openings of
the soft palate can be detected immediately.
4) The nasal cavity is isolated
during inspiration and expiration.
Protocol.
The design of the experimental protocol focuses on the effect of
E on the NO
exhalation profile. After isolation of the nasal cavity, each subject
was asked to breathe comfortably (3-5 tidal breaths) before
performing a single-exhalation maneuver. In our single exhalation a
subject inspires from functional residual capacity to total lung
capacity, then slowly exhales to approximately functional residual
capacity. The exhalate passed through a mouthpiece connected to a
"j" valve (two 1-way valves), allowing the subjects to inspire
ambient air [10-15 parts/billion (ppb) NO] through one
port and exhale through the second port. Gas (NO and
CO2) was measured from a
sampling port at the side of the mouthpiece, ~1 cm from the mouth.
Each subject performed four different
E patterns:
1) constant slow
E of ~250
ml/s (control maneuver), 2) constant
faster E of
~500 ml/s, 3) linearly (in time)
increasing E,
and 4) linearly (in time) decreasing
E. Each
subject performed the maneuvers in triplicate, with or without a 15-s
breath hold before the exhalation. Hence, each subject performed 24 technically acceptable exhalation maneuvers (4 E × 3 replicates × 2 presence or absence of breath hold). Exhalation
with constant flow rate was facilitated by use of a Starling resistor.
The resistance to flow through a Starling resistor increases with
increasing flow; thus the resistor serves to maintain a constant flow
rate, independent of expiratory effort of the subject. On-line
representation of the flow signal on the screen of the computer
assisted the subjects during performance of the increasing and
decreasing E
maneuvers.
Airstream analysis.
NO concentration was measured using a chemiluminescence NO analyzer
(model NOA280, Sievers Instruments, Boulder, CO). The instrument is
capable of providing highly accurate (repeatability ±1 ppb vol) gas
phase measurements with a very small detection threshold [<5
ppb vol-500 parts/million (ppm) vol] and a fast response
time (0-90% response time of 200 ms). We used a needle valve to
restrict flow and maintain an operating reaction cell pressure of 7.5 mmHg and a sampling rate of 250 ml/min. The factory calibration values
of the instrument were tested periodically using certified NO gas (27.8 ppm in N2).
CO2 concentration was measured by
using a fast-response (0-90% response time of 100 ms) mass
spectrometer (Perkin-Elmer Medical Instruments, Pomona, CA). The
convective transport times for NO and
CO2 from the mouth to the
analyzers were 350 and 450 ms, respectively. These times were
subtracted from the profiles before analysis.
Flow rate was measured with a pneumotachometer (model 50-MC2-M2,
Mariom). Analog signals from the pneumotachometer, the mass spectrometer, and the chemiluminescent analyzer were digitized through
a 12-bit analog-to-digital converter. The analog-to-digital card was
programmed to scan three channels simultaneously every 50 ms. The
resulting measurements (20 Hz) were stored in a computer for later
analysis. Figure 1 is a schematic of the
experimental setup.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 1.
Schematic of experimental setup. Soft palate was elevated with
negative nasal cavity pressure (20
cmH2O). Exhalation flow rate was
controlled with a Starling resistor in constant flow rate maneuvers.
Nitric oxide (NO) concentration was measured using chemiluminescence,
and CO2 concentration was measured
with mass spectroscopy. PC, personal computer.
|
|
Data analysis.
We used three parameters to characterize the exhalation profile:
1) series dead space,
2) average concentration of phase
III, and 3) slope of phase III. In
addition, we determined the elimination rate of each gas from the
lungs.
The dead space for CO2
(VDCO2)
was calculated as described by Meyer et al. (18) using
backextrapolation of the phase III slope with respect to exhaled
volume
[SIII,CO2(V)] (18, 19, 25, 26). Because the shape of the exhalation profile of NO is
quite different from that of CO2
(see RESULTS), we defined the dead
space of NO
(VDNO) in
a different manner. The exhalation profile of NO is characterized by a
rapid rise to an initial peak in phase I. The slope of phase III can be
negative or positive depending on
E (see
RESULTS). The size of the initial peak depends on the length of breath hold. We defined
VDNO as the
volume corresponding to the midpoint NO concentration between the
ambient level and the initial peak value in the breath-hold maneuver.
An average value for concentration and flow rate that characterizes the
exhalation profile was calculated using the data in phase III (alveolar
gas) of the exhalate. Average concentrations were calculated with
respect to time and exhaled volume. The average concentrations and flow
rates with respect to time
[
E,NO(t), E,CO2(t),
and
E(t)]
were calculated over a fixed time interval of 6 s beginning 2 s after
the start of exhalation and ending 8 s after the start of exhalation
|
(1)
|
The
average concentrations and flow rates with respect to volume
[E,NO(V),
E,CO2(V),
and
E(V)] were calculated using the exhalate between 20%
(V20) and 60%
(V60) of each subject's vital
capacity (VC)
|
(2)
|
The
time (2-8 s) and volume (20-60% of VC) intervals were
chosen for two reasons: 1) to
provide a maximum interval common to all subjects and
2) to guarantee that, even under the
extreme conditions (maneuver with the lowest flow rate or
subject with the smallest VC), the regression would exclude phase I and
phase II of the exhalation profile. The technique for determining the average concentrations and flow rates is shown schematically in Fig. 2.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 2.
Schematic detailing analysis of exhalation profiles for concentration
and flow rate. Average concentrations and flow rates were estimated for
exhaled volume and exhaled time by use of interval of 20-60% of
vital capacity (VC) or 2-8 s, respectively. Phase III slope was
calculated using linear least squares over this same interval and then
normalized by average value over interval.
E, exhalation
flow rate;
CCO2 and
CNO,
CO2 and NO concentration,
respectively;
E,
CO2,
and NO,
average
E,
CCO2, and
CNO, respectively;
 ,
SIII,CO2,
and SIII,NO,
phase III slopes for NO, CO2, and
E, respectively.
|
|
Phase III slopes for NO
(SIII,NO),
CO2
(SIII,CO2),
and E
()
were determined using linear least squares regression with respect to
time and exhaled volume. The slopes with respect to time
[SIII,NO(t), SIII,CO2(t),
and
(t)] were calculated over the same time interval used in determining the average concentrations (2-8 s; Fig. 2). Phase III slopes with respect to volume
[SIII,NO(V),
SIII,CO2(V),
and
(V)] were also calculated using the same volume interval used to determine the average concentrations and flow rates with respect to volume (20-60% of VC; Fig. 2). The slopes were then normalized with the appropriate average concentrations or flow rates as described above
(e.g., slope with time normalized by the average value with respect to
time)
|
(3)
|
|
(4)
|
|
(5)
|
|
(6)
|
|
(7)
|
|
(8)
|
Hence,
the resulting normalized slopes represent the fractional change in gas
concentration or flow rate with respect to the average per unit time or
per unit volume exhaled.
The elimination rate of NO from the lungs
(ENO) is defined as the product
of concentration and flow rate. Because concentration and flow rate are
functions of time or exhaled volume, we calculated the elimination rate
at end exhalation (ee). Hence, ENO
was determined as follows
|
(9)
|
The
end-exhalation point was defined at
V60, and
E,ee
and CNO,ee represent average
values over a period of 1 s.
ENO can be simply viewed as the
sum of two components: 1)
elimination from a well-mixed expansile alveolar region (product of
E and alveolar
concentration) and 2) net airstream
absorption of the gas from the tissue during transport through the
airway tree
|
(10)
|
where
Calv,ee is alveolar concentration
at end exhalation,
t:g,air is
the average (over axial position) flux
(nl · s
1 · cm2)
of NO from the airway tissue into the gas phase, and
As,air is the
total surface area of the airways
(cm2). Recently, Silkoff et al.
(28) reported a linear relationship between
ENO and
E over a wide
range of flow rates, which suggests that
1)
Calv,ee approaches a steady-state
value (Calv,ss) and 2)t:g,airAs,air
is essentially constant over the examined flow rate range. The above
observations were verified with the theoretical model (30). In this
fashion the slope of ENO vs. E provides an
estimate of Calv,ss, and the
intercept provides an estimate of the contribution from the conducting
airway space to total lung elimination.
Statistics.
A two-tailed paired t-test was used to
detect differences in the dead space between gases and between
different maneuvers. The same statistical method was used to examine
the effect of breath hold on the average exhaled concentrations and
phase III slope. The flow rate dependence of the parameters was
examined with linear least squares regression. The null hypothesis of
no dependence (i.e., zero slope) was tested with a single-sample two-tailed t-test with use of the
average slope from all seven subjects.
P < 0.05 was considered
statistically significant in all analyses.
| |
RESULTS |
Figure
3A
represents a typical exhalation profile of NO and
CO2 under the control conditions.
The E signal is
also shown. The CO2 profile
demonstrates the well-documented dead space (phase I) followed by a
rapid rise (phase II) to a sloping alveolar plateau (phase III). The
exhalation profile of NO is substantially different. The initial
concentration is equal to the ambient concentration (~14 ppb). Phase
I is present and is visually smaller than that of
CO2 (see Series
dead space). Phase I is characterized by an initial
peak and then a slow decrease during exhalation of the alveolar gas
(phase III). Hence, under the control conditions of our experiment,
SIII,NO is
negative (see Phase III slope).
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 3.
Representative exhalation profiles describing qualitatively effect of
E and breath
hold. A: control maneuver of a
constant E of
~250 ml/s. B: exhalation profiles in
same subject after 15-s breath hold (note change in
y-axis scale).
C: exhalation profiles when
E is decreasing
linearly in time during exhalation (scale as in
A). ppb, Parts per billion.
|
|
Figure 3B represents the exhalation
profiles in the same subject after a 15-s breath hold. NO and
CO2 scales have been reduced to
allow depiction of the initial NO peak. The results are typical of
those presented by previous investigators. For
CO2, phase I is reduced, phase II
rises to a larger concentration before the start of phase III, and
phase III has a flatter slope. For NO the initial peak in phase I is
much larger in magnitude (70 ppb vs. 17 ppb in this particular
subject), and phase III is similar in shape.
Figure 3C depicts the effect of a
dynamically changing flow rate on the exhalation profiles (in this case
a linearly decreasing E), again, in
the same subject. The CO2 profile
does not change appreciably. However, the NO profile demonstrates
marked changes. After the initial peak the concentration decreases
gradually to a minimum and then begins to steadily increase. This
phenomenon generates a positive phase III slope (see
Phase III slope).
Series dead space.
Mean values of
VDCO2
and VDNO
for all seven subjects, under different exhalation maneuvers, are
summarized in Table 2.
VDNO is
~50-60% smaller than
VDCO2
under all conditions. VDCO2
and VDNO
decrease significantly (~40-60%,
P < 0.05) after a 15-s breath hold
but are independent of a twofold increase in the flow rate.
Average concentrations.
Table 3 summarizes the average
concentrations for NO and CO2 for
the control maneuver. After a 15-s breath hold,
E,NO(t) and
E,NO(V)
were slightly decreased for most of the subjects; however, this
decrease was not statistically significant. In contrast, E,CO2(t)
and
E,CO2(V)
were significantly (P < 0.05)
increased after a 15-s breath hold.
Figure 4 displays the data from the
constant-flow exhalation profiles. Average concentrations for NO (Fig.
4, A and
B) and CO2 (Fig. 4,
C and
D) with respect to time and
volume are plotted as a function of
E.
The slopes for
E,NO(t)
and
E,NO(V)
were 0.0086 ± 0.0044 and 0.0048 ± 0.0039 (SD)
ppb · ml1 · s1,
respectively. Both are statistically different from 0, indicating that
E,NO(t)
and
E,NO(V)
are inverse functions of
E. In contrast, the slopes for
E,CO2(t)
and
E,CO2(V)
were +0.0006 ± 0.0009 and 0.0018 ± 0.0011% · ml1 · s1,
respectively. Only the slope for
E,CO2(V)
is statistically an inverse function of
E.
View larger version (17K):
[in this window]
[in a new window]
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 4.
NO and
CO2
as a function of
E in maneuvers
where E was held
constant throughout exhalation maneuver.
A:
NO with
respect to exhalation volume
[NO(V)].
B:
NO with
respect to exhalation time
[NO(t)].
C:
CO2
with respect to exhalation volume
[CO2(V)].
D:
CO2
with respect to exhalation time
[CO2(t)].
Each subject is represented with a different symbol, and lines of
linear regression are shown.
|
|
Phase III slope.
Relative slopes with respect to time or volume from the
constant flow rate maneuvers are shown in Fig.
5.
III,NO(t) and
III,NO(V)
did not reveal any dependence on
E. In contrast,
III,CO2(t)
under control conditions is a positive function of
E, whereas
under control conditions
III,CO2(V) is an inverse function of
E.
View larger version (20K):
[in this window]
[in a new window]
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 5.
Average SIII,NO
and
SIII,CO2
(III,NO and
III,CO2,
respectively) as a function of
E in maneuvers
where E
was held constant throughout
exhalation maneuver.
A:
III,NO with
respect to exhalation volume
[III,NO(V)].
B:
III,NO
with respect to exhalation time
[III,NO(t)].
C:
III,CO2
with respect to exhalation volume
[III,CO2(V)].
D:
III,CO2
with respect to exhalation time
[III,CO2(t)].
Statistically different from 0:
* P < 0.05 and
# P < 0.1.
|
|
The dependence of the phase III slope on the dynamics of
E can be
illustrated by plotting
III,NO(t)
or
III,CO2(t) vs.
(t).
Figure 6 summarizes the data with and
without breath hold for NO and CO2. Data with significant
negative or positive
(t) were generated during the gradually decreasing or increasing flow rate
maneuvers, respectively. Analysis of these slopes indicates that
III,NO(t),
but not
III,CO2(t),
depends on
(t)
in an inverse fashion.
III,NO(t)
can be positive or negative depending on the slope of the flow. In
contrast,
III,CO2(t)
is always positive. In addition, the mean intercept on the ordinate represents the mean phase III slope at a constant flow rate
[i.e., (t) = 0]. The intercept is positive for
CO2 and negative for NO. In
addition, a 15-s breath hold reduces the absolute value of the
intercept for CO2 and NO; i.e.,
the phase III slopes become flatter (Fig.
3B).
View larger version (19K):
[in this window]
[in a new window]
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
III,NO(t)
and
III,CO2(t)
as a function of
(t).
A: NO without breath hold;
B: NO with breath hold;
C:
CO2 without breath hold;
D:
CO2 with breath hold. Solid lines
represent least squares linear regression. * Statistically
different from 0 (P < 0.05).
|
|
Elimination rate.
ENO for the constant exhalation
maneuvers is displayed as a function of
E,ee in Fig.
7. A 15-s breath hold had no statistical effect on the elimination rates at end exhalation; hence, all the data
have been lumped together. The mean slope (an index of Calv,ss) and intercept (an index
of airway flux or production) for each subject are presented in Table
4. ENO
is generally a positive function of
E,ee; however,
the standard deviation of ENO is
large. In fact, the slope of NO for subject
1 was zero. The slope for all subjects, and hence the
average steady-state alveolar concentration, was 5.6 ± 3.1 (SD) ppb. The intercept, and hence the average net total airway flux of
NO
(t:g,airAs,air), for the seven subjects was 0.71 ± 0.31 × 106 ml/s and is
statistically different from zero.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 7.
Elimination rates for NO (ENO)
as a function of
E at end
exhalation
(E,ee). Only
constant E
maneuvers are shown, with and without breath hold. Lines represent
least squares linear regression through all data points for each
subject.
|
|
| |
DISCUSSION |
Our experimental data suggest two different but not contradictory
hypotheses: 1) NO may originate from
the airway region of the lung, which is fairly nonexpansile;
2) NO may originate from the
expansile alveolar region of the lung. The combination of hypotheses 1 and
2 can provide a possible explanation
for the unique behavior of the NO exhalation profile and the
differences with CO2 that have
been observed by us and others. However, the existence of alternative
explanations cannot be excluded. Some are discussed below. The relative
importance or contribution of each of these mechanisms is examined in
greater detail with the development of a simplified two-compartment
mathematical model (30).
Series dead space.
There is a significant difference between
VDCO2
and VDNO.
Although
VDCO2
(180 ± 24 ml) is probably overestimated [a phenomenon
previously reported (18) and attributed to an artifact arising from
backextrapolation of the nonlinear
CO2 profile], VDNO (87 ± 25 ml) is significantly smaller. Hence, even with the nasal
passages isolated, NO appears in the exhaled breath before CO2. This result is consistent
with considerable production of NO in an area of the lung before the
alveolar region, the source of CO2
(hypothesis 1). The considerable
decrease in the dead space volume for both gases after the short
breath-hold time has been previously reported (4, 18) and is attributed
primarily to cardiogenically induced convective mixing during the
breath hold. Although the techniques used to determine
VDNO and
VDCO2 were different, they are quite similar and cannot account for the large
difference between
VDNO and
VDCO2.
Average phase III concentrations.
The decrease in the average exhaled concentration over the specified
volume interval (20-60% of VC) in the constant flow rate maneuvers with higher
E (Fig. 4)
cannot be attributed to the same mechanism for both gases. For
CO2 this decrease can be explained through the mechanism of continuous gas exchange in the alveolar region
of the lung (6, 18). Over a constant-volume interval, a slower
E will
result in an increased expiration time (similar to breath holding);
hence, the result is an increased
CO2 concentration due to
continuous production and evolution into the alveolar gas. In contrast,
when the average CO2 concentration
is calculated over a specified time interval (Fig.
4D), the effective time for gas
exchange is the same under the different flow rate conditions. As a
result, there is no statistical flow rate dependence on
CO2(t). Small changes in
CO2(t)
with flow rate can be attributed to other mechanisms such as
stratified or parallel inhomogeneities of the lung during exhalation
(19).
In contrast,
NO(t)
and
NO(V)
are inverse functions of flow, suggesting inherently different gas
exchange dynamics. The exact mechanism behind this dependence is not
known. However, a probable explanation, which is in agreement with our
findings for the series dead space, is that a significant part of the
exhaled NO is derived from the relatively nonexpansile airway
region of the lung proximal to the alveolar region
(hypothesis 1). As
E increases, the
rate of washout in the airways (fixed volume) increases (or mean
residence time decreases), thus decreasing the concentration. This
phenomenon does not occur for CO2,
the source of which is the expansile alveoli. As the alveoli shrink
during exhalation, CO2
concentration in the alveolar gas is maintained relatively constant (in
the absence of production or consumption) and is independent of flow
rate. This concept is developed more fully in the form of a
mathematical model (30). In addition, consistent with the theory of
airway NO production is the fact that several cell types present in the airways, including the bronchial epithelium, fibroblast, smooth muscle
cell, macrophage, and neutrophil, have been shown in vitro to produce
NO and/or to contain the enzyme necessary for NO production, NO
synthase (12, 13, 17, 23, 27).
Interestingly, the slope of
NO(t)
vs. E is
statistically steeper than the slope of
NO(V)
vs. E. If one
assumes a net consumption during exhalation of NO in the alveolar
region because of fast reaction with the hemoglobin of the blood in the
pulmonary capillary bed, then the relationship of NO concentration to
flow rate can be viewed as a result of two different antagonistic
mechanisms: 1) concentration is an
inverse function of the flow due to NO production in the nonexpansile
part of the lungs, and 2)
concentration is a positive function of flow as a result of a decrease
in the expiration time and hence less NO consumption by the pulmonary blood. Mechanism 2 presumes a nonzero
alveolar concentration (hypothesis 2), which is consistent with experimentally observed
in vitro production of NO by the alveolar epithelium (7) as well as theoretically by Hyde et al. (11) and in our companion article (30).
These two mechanisms are consistent with a steeper slope of
NO(t)
vs. E (same
exhalation time and hence only mechanism 1) compared with
NO(V)
vs. E (different
exhalation times and hence mechanism 2 contributes).
Mechanisms other than continuing gas exchange, such as parallel or
stratified inhomogeneity, may also impact the observed results. For
example, parallel inhomogeneities in the lungs exist because of
different alveolar ventilation-to-volume ratios. During exhalation,
regions of higher ventilation will empty first from the lung and will
have a lower (or higher) concentration of gas because of alveolar
dilution (or enrichment) during inspiration (19). Thus, over a
constant time interval, a higher
E will increase
the contribution of alveolar air in poorly ventilated regions, which
have a higher (or lower) concentration. Whether the poorly ventilated
regions have a higher or lower NO concentration will depend on whether
there is a net positive or negative flux of NO in the specific alveolar
region, respectively. A positive or negative flux depends on several
factors, including the alveolar diffusing capacity, the inhaled
concentration, and the volume of the specific alveolar region (11, 30).
This mechanism would predict a positive or negative relationship of
NO(t)
to E. Because
NO(t)
is a stronger inverse function of
E than
NO(V) is, parallel inhomogeneities would be an important mechanism underlying the phase III slope if a net consumption of NO occurs in poorly ventilated regions.
Recently, Silkoff et al. (28) performed an extensive study on the
effect of E on
the exhaled NO concentration during oral exhalation (nasal cavity
isolated during expiration). Although they applied significant positive
end-expiratory pressure, which has been reported by others to affect
exhaled NO concentration (20), they also report an inverse relationship
of concentration to
E.
Phase III slope.
Phase III slopes have been defined in several ways by previous
investigators (18, 19, 25, 26). When phase III slope data are reported,
several factors must be considered:
1) absolute vs. relative values,
2) slope with respect to time or
exhaled volume, 3) range of the
exhalation profile over which to calculate slope, and
4) method of normalization.
We showed earlier that
E can affect the
average CO2 concentration by
changing the expiration time (continuous alveolar production) or the
exhaled volume (parallel, stratified inhomogeneities). The positive
dependence of
III,CO2(t)
and the small negative relationship of
III,CO2(V)
to E can be
explained using the same rationale. The change in
CO2 concentration per unit time
[III,CO2(t)]
will increase in the constant flow rate maneuvers with a higher
E due to
parallel inhomogeneity, since the exhaled volume per unit time will
increase. This change is small compared with the effect of continuing
gas exchange (18). The change in
CO2 concentration per unit volume
[III,CO2(V)] will slightly decrease with higher flow because of continuing alveolar
gas exchange (continuous production), since the expiration time per
unit volume will decrease.
If it is assumed that the average flow during the examined part of
phase III does not change significantly between maneuvers, III,CO2(t)
is not expected to change with the dynamically changing flow rate
maneuvers (Fig. 6, C and
D). On the other hand, SIII,NO behaves
completely differently.
III,NO(t)
and III,NO(V)
are essentially independent of the magnitude of a constant
E but depend
significantly on a changing
E throughout the
course of the single exhalation (Fig. 5,
A and
B, and Fig. 6,
A and
B).
Figure 6 displays this unique feature of the NO exhalation profile.
Depending on the slope of the exhalation flow, the alveolar plateau can
range from positive to negative values. The mechanisms underlying the
direct relationship of the exhaled NO concentration to the flow rate
presented earlier can also be invoked to explain this phenomenon. At
any given exhaled volume or time after the beginning of exhalation, an
increase or decrease in
E causes a
decrease or increase in NO concentration, respectively, because of
dilutional effects in the nonexpansile airways
(hypothesis 1). For example, if
E is
continuously decreasing [i.e., negative (t)],
NO concentration will continuously increase [i.e., positive
III,NO(t)].
An interesting finding related to the dynamically changing flow rate
maneuvers is the nonzero intercept for NO in Fig. 6, A and
B. The existence of a negative
SIII,NO when
E is constant indicates that, although the phase III slope is predominantly influenced by flow rate changes, other mechanisms also contribute. We
have already suggested a nonzero alveolar concentration
(hypothesis 2) coupled with
continuing gas exchange (in the case of continuing NO consumption) as a
mechanism for this negative slope. However, for continuous consumption
of NO to occur in the alveolar region, alveolar concentration at end
inspiration would have to be larger than the steady-state value. A
steady-state alveolar NO concentration is expected when the net rate of
production from the cells of the alveolar membrane is balanced by
diffusion into the pulmonary blood (11, 30). Hence, it is possible that
inspired conditions such as concentration and flow rate may affect the
alveolar concentration at end inspiration. For example, if the inspired
NO concentration were zero, then the alveolar concentration at end
inspiration may be less than the steady-state value. In this case,
continuing gas exchange would cause an increase in the alveolar
concentration during exhalation and a positive phase III slope. This
finding would be modulated by inspiration flow rate, which impacts the time of inspiration as well as absorption of NO from the airways. Our
finding of a negative phase III slope may be attributed to inhaling
ambient concentrations of NO that were on the order of 15 ppb. The
effect of inspired conditions on the NO exhalation profile is explored
further elsewhere (30) with use of a mathematical model and should be
explored further experimentally.
Breath hold.
The effect of breath hold on the NO and
CO2 exhalation profiles has been
reported previously and can be explained in a manner consistent with
the mechanisms related to average concentration and the phase III
slope. During breath hold, NO and
CO2 concentrations in the airway
and alveolar regions approach their steady-state values. NO accumulates
in the airway space [but could also accumulate in alveoli
(11)], and CO2 accumulates
in the alveolar region. Hence, during exhalation, a large initial peak
in NO concentration is observed in phase I and a larger concentration
of CO2 is observed at the start of
phase III.
The dynamics of NO exchange in the alveolar region are more complicated
because of the somewhat unexpected result of the effect of breath hold
on the average NO concentration in phase III. If continuing NO exchange
in the alveolar region is the predominant mechanism underlying the
negative phase III slope under control conditions (constant flow rate),
then the average NO concentration in phase III should decrease after
the 15-s breath hold (e.g., the alveolar concentration is expected to
decrease because of consumption by the pulmonary blood).
CO2 and NO have similar relative rates of increase or decrease, respectively, during a single exhalation with constant flow rate (~5%/s over the 2- to 8-s time period without breath hold and 1.7%/s after the breath hold). However, only
CO2 concentrations were
statistically affected by a breath hold. Although
E,NO(t)
and
E,NO(V)
were slightly decreased (consistent with continuous uptake by the
pulmonary blood) during breath hold, this was not statistically
significant. There are several possible explanations.
1) NO consumption by the pulmonary
blood may be limited and the negative
SIII,NO, in
contrast to
SIII,CO2, is generated predominantly by parallel or stratified inhomogeneities and not as a result of continuous gas exchange in the alveoli. 2) NO concentrations in the airway
space become quite elevated during the breath hold (Fig.
3B). The subsequent steep axial
gradients in NO concentration may lead to substantial axial diffusion
from NO in the airways to the alveolar region, leading to an alveolar concentration, at the end of breath hold, above its normal steady-state value. The distance of diffusion
(LD) over a
15-s breath hold can be estimated by
LD = 
from
an unsteady-state analysis (1).
Dg is the
molecular diffusivity of NO in the gas phase and is ~0.23
cm2/s; hence,
LD is ~7.4 cm.
This length spans approximately generations 5-23 (6.5 cm) in the Weibel model (31) and easily
spans the length of the respiratory bronchioles (0.62 cm in
generations 17-23).
Elimination rate.
ENO did not remain constant as
E,ee
changed for six of the seven subjects. Although the exhaled
concentration of NO decreased with increasing
E (in the
constant and in the dynamically changing maneuvers), this decrease was
not sufficient to preserve a constant ENO (Eq. 9). There are at least three explanations.
1) As discussed earlier
(hypothesis 2), there may be a
nonzero alveolar concentration of NO. Because of the expansile nature
of the alveoli, the alveolar concentration is independent of
E (in the
absence of production or consumption should alveolar concentration
remain constant during exhalation). Hence, the elimination rate of a
gas is proportional to
E
(Eq. 10). This explanation is quite
probable in light of the evidence that the alveolar epithelium is
capable of producing NO (7, 29). 2)
Equation 10 suggests that a positive
relationship between
E and
ENO can be generated by a
flow-dependent flux of NO from the airway wall
(Jt:g,air). If
transport of NO into the airstream within the airways is limited by
diffusion in the gas phase, then increasing
E may increase
the net uptake of NO within the airways (28). As
E increases, the
concentration of NO decreases by dilution. Hence, the driving force for
radial diffusion (difference between tissue concentration and airstream concentration or radial concentration gradient) is
increased. The very small solubility of NO in water and tissue creates
a very small relative gas-phase resistance to diffusion, making this
explanation unlikely (30). The consumption of NO by chemical reaction,
at least first order in NO concentration, within the airway tissue may
also generate a flow-dependent flux. Increasing E and thus
decreasing NO concentration in the airstream and within the adjacent
tissue may decrease the rate of consumption by chemical reaction while
the rate of endogenous production remains unaffected. It is difficult
to speculate on the importance of this mechanism; however, the
mathematical model provides valuable insight and suggests that reduced
consumption in the airway tissue with increasing
E is not
important in explaining the observed flow rate dependence of the NO
elimination.
There is a quite significant variability in the exhaled levels of NO
compared with CO2, even among
normal human subjects. The large standard deviations of the intercept
and slope in Fig. 6 suggest that such variability originates from the
airway as well as the alveolar region of the lung. This variability can be attributed to a number of parameters, including the endogenous production of NO in the airway or alveolar tissue as well as the diffusing capacity of NO, and they are examined in detail elsewhere (30).
Conclusions.
We have utilized a technique to establish satisfactory isolation of the
nasal cavity while collecting the experimental oral exhalation profile
of NO and CO2 from normal human
subjects. An experimental protocol was designed that focuses on the
effect of E on
the exhaled concentration of the endogenously produced gases. There are
five major findings of the study related to NO exchange dynamics:
1) the average phase III exhaled
concentration of NO is an inverse function of
E;
2)
SIII,NO is highly
dependent on the dynamic changes of flow rate during exhalation and can vary from positive to negative values;
3) for constant
E,
SIII,NO is
slightly negative when ambient levels of NO are inhaled (~15 ppb);
4) the elimination rate of NO is a
positive linear function of
E with a nonzero
intercept and can be used to quantitate the relative contribution of
the alveolar and airway regions to exhaled NO; and
5) there is a large intersubject
variability in exhaled NO concentration compared with
CO2. All these findings are
consistent with the alveoli and the airways as sources of exhaled
endogenous NO. These findings can provide directions for collecting
reproducible exhalation NO profiles. For example, one must consider the
following factors when assessing NO concentration in the exhaled
breath: E,
exhalation volume, and region of the exhalation profile where the
measurement is recorded. In addition, inspired conditions such as
volume, flow rate, and concentration may impact the exhalation profile.
These factors are considered by a mathematical model (30) but must also
be addressed in future experimental designs.
| |
ACKNOWLEDGEMENTS |
The authors thank David S. Mukai for technical support in the
collection of the exhalation profiles.
| |
FOOTNOTES |
This work was supported in part by National Science Foundation Grant
BES-9619340 and by generous start-up funds to S. C. George from the
Department of Chemical and Biochemical Engineering and Materials
Science at the University of California, Irvine.
Address for reprint requests: S. C. George, Dept. of Chemical and
Biochemical Engineering and Materials Science, 916 Engineering Tower,
University of California at Irvine, Irvine, CA 92697-2575.
Received 26 November 1997; accepted in final form 16 April 1998.
| |
REFERENCES |
1.
Bird, R. B.,
W. E. Stewart,
and
E. N. Lightfoot.
Transport Phenomena. New York: Wiley, 1960.
2.
Borland, C.,
Y. Cox,
and
T. Higenbottam.
Measurement of exhaled nitric oxide in man.
Thorax
48:
1160-1162,
1993[Abstract].
3.
Dillon, W. C.,
V. Hampl,
P. J. Schultz,
J. B. Rubins,
and
S. L. Archer.
Origins of breath nitric oxide in humans.
Chest
110:
930-938,
1996[Abstract/Free Full Text].
4.
Engel, L. A.,
H. Menkes,
L. D. H. Wood,
G. Utz,
J. Joubert,
and
P. T. Macklem.
Gas mixing during breath holding studied by intrapulmonary gas sampling.
J. Appl. Physiol.
35:
9-17,
1973[Free Full Text].
5.
Gaston, B.,
J. M. Drazen,
J. Loscalzo,
and
J. S. Stamler.
The biology of nitrogen oxides in the airways.
Am. J. Respir. Crit. Care Med.
149:
538-551,
1994[Abstract].
6.
Gronlund, J.,
E. R. Swenson,
J. Ohlsson,
and
M. P. Hlastala.
Contribution of continuing gas exchange to phase III exhaled PCO2 and PO2 profiles.
J. Appl. Physiol.
62:
2467-2476,
1987[Abstract/Free Full Text].
7.
Gutierrez, H. H.,
B. R. Pitt,
M. Schwarz,
S. C. Watkins,
C. Lowenstein,
I. Caniggia,
P. Chumley,
and
B. A. Freeman.
Pulmonary alveolar epithelial inducible NO synthase gene expression: regulation by inflammatory mediators.
Am. J. Physiol.
268 (Lung Cell. Mol. Physiol. 12):
L501-L508,
1995[Abstract/Free Full Text].
8.
Hogman, M.,
C. Frostell,
H. Arnberg,
and
G. Hedenstierna.
Inhalation of nitric oxide modulates methacholine-induced bronchoconstriction in the rabbit.
Eur. Respir. J.
6:
177-180,
1993[Abstract].
9.
Hogman, M.,
C. G. Frostell,
H. Hedenstrom,
and
G. Hedenstierna.
Inhalation of nitric oxide modulates adult human bronchial tone.
Am. Rev. Respir. Dis.
148:
1474-1478,
1993[Medline].
10.
Hogman, M.,
S.-Z. Wei,
C. Frostell,
H. Arnberg,
and
G. Hedenstierna.
Effects of inhaled nitric oxide on methacholine-induced bronchoconstriction: a concentration response study in rabbits.
Eur. Respir. J.
7:
698-702,
1994[Abstract].
11.
Hyde, R. W.,
E. J. Geigel,
A. J. Olszowka,
J. A. Krasney,
R. E. Forster,
M. J. Utell,
and
M. W. Frampton.
Determination of production of nitric oxide by lower airways
theory.
J. Appl. Physiol.
82:
1290-1296,
1997[Abstract/Free Full Text].
12.
Jorens, P. G.,
F. J. VanOverveld,
H. Bult,
P. A. Vermeire,
and
A. G. Herman.
Synergism between interleukin-1
and interferon-
, an inducer of nitric oxide synthase, in rat lung fibroblasts.
Eur. J. Pharmacol.
224:
7-12,
1992[Medline].
13.
Jorens, P. G.,
F. J. VanOverveld,
H. Bult,
P. A. Vermiere,
and
A. G. Herman.
L-Arginine-dependent production of nitrogen oxides by rat pulmonary macrophages.
Eur. J. Pharmacol.
200:
205-209,
1991[Medline].
14.
Kharitonov, S. A.,
K. Alving,
and
P. J. Barnes.
Exhaled and nasal nitric oxide measurements: recommendations.
Eur. Respir. J.
10:
1683-1693,
1997[Medline].
15.
Kharitonov, S. A.,
D. Yates,
R. A. Robbins,
R. Logan-Sinclair,
E. A. Shinebourne,
and
P. J. Barnes.
Increased nitric oxide in exhaled air of asthmatic patients.
Lancet
343:
133-135,
1994[Medline].
16.
Kimberly, B.,
B. Nejadnik,
G. D. Giraud,
and
W. E. Holden.
Nasal contribution to exhaled nitric oxide at rest and during breathholding in humans.
Am. J. Respir. Crit. Care Med.
153:
829-836,
1996[Abstract].
17.
Kobzik, L.,
D. Bredt,
C. Lowenstein,
S. H. Snyder,
J. M. Drazen,
D. Sugarbaker,
and
J. S. Stamler.
Nitric oxide synthase in human and rat lung.
Am. J. Respir. Cell Mol. Biol.
9:
371-377,
1993.
18.
Meyer, M.,
M. Mohr,
H. Schulz,
and
J. Piiper.
Sloping alveolar plateaus of CO2, O2, and intravenously infused C2H2 and CHClF2 in the dog.
Respir. Physiol.
81:
137-152,
1990[Medline].
19.
Paiva, M.,
and
L. A. Engel.
The anatomical basis for the sloping N2 plateau.
Respir. Physiol.
44:
325-347,
1981[Medline].
20.
Persson, M. G.,
P. A. Lonnqvist,
and
L. E. Gustafsson.
Positive end-expiratory pressure ventilation elicits increases in endogenously formed nitric oxide as detected in air exhaled by rabbits.
Anesthesiology
82:
969-974,
1995[Medline].
21.
Persson, M. G.,
N. P. Wiklund,
and
L. E. Gustafsson.
Endogenous nitric oxide in single exhalations and the change during exercise.
Am. Rev. Respir. Dis.
148:
1210-
Top
Abstract
Introduction
