| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vol. 83, Issue 6, 1907-1916, December 1997
1 Academisch Ziekenhuis, ABSTRACT Verbanck, S., D. Schuermans, A. Van Muylem, M. Paiva, M. Noppen, and W. Vincken. Ventilation distribution during histamine provocation. J. Appl. Physiol. 83(6):
1907-1916, 1997.
multiple-breath washout experiments; nonspecific agent; bronchodilatation; salbutamol
INTRODUCTION FOR THE STUDY of ventilation distribution in the normal
or the diseased lung, various techniques have been used to evaluate the
sites and mechanisms of ventilation inhomogeneity. In the area of
bronchoconstriction, radioisotope techniques have been applied to
identify central and peripheral deposition of a radiolabeled bronchoconstrictor agent and to assess its effect on lung function and
gas exchange (19, 21) or, alternatively, to evaluate modifications in
distribution of a radiolabeled aerosol or radioactive xenon when
inhaled immediately before and after bronchoprovocation (5, 27). In
general, the resolution of the gamma camera used for this type of study
is limited and only allows differences in ventilation between
relatively large lung zones, i.e., comprising several hundreds of
acini, to be evaluated. A second group of tests to differentiate
between peripheral and central action of bronchoconstrictive agents
makes use of their different effect on lung function parameters such as
airway resistance vs. anatomic dead space (22); forced end-expiratory
flow rates vs. forced expired volume in 1 s (FEV1) (19); or
forced expiratory flows of air vs. those of a
He-O2 mixture (3, 15). In the case
of forced expiratory flow rates of air vs.
He-O2, for instance, the dividing
line between proximal and peripheral zones of the bronchial tree is
located at the transition between turbulent and laminar flow, i.e.,
typically at the level of the so-called "small airways."
Finally, a third group of tests used to evaluate the site of airway
narrowing is the N2 washout
technique, either by measuring the phase III alveolar slope of the
single-breath washout (SBW) (14, 20) or by computing an index derived
from the washout curve that represents the expired concentration of
each subsequent breath of a multiple-breath washout (MBW) (10, 12). In
several of these washout studies, ventilation inhomogeneity after
bronchoprovocation has been attributed to peripheral units. Although
these units may potentially be involved in decreasing overall
ventilation efficiency, the reported analyses of the
N2 washout test, SBW or MBW, as
such cannot clarify whether concentration differences were generated in
large units or rather in peripheral ones. In this study we use a more
detailed analysis of the MBW test, previously applied in normal
subjects by Crawford et al. (6-9), which provides the ability to
distinguish between possible gas concentration differences generated in
very small units, i.e., at the level of acini, and those generated in
much larger lung units, which would also become apparent with the
above-mentioned tests.
MATERIAL AND METHODS
We investigated ventilation inhomogeneity during
provocation with inhaled histamine in 20 asymptomatic nonsmoking
subjects. We used N2
multiple-breath washout (MBW) to derive
parameters Scond
and Sacin as a
measurement of ventilation inhomogeneity in conductive and acinar zones
of the lungs, respectively. A 20% decrease of forced expiratory volume in 1 s (FEV1) was used to
distinguish responders from nonresponders. In the responder group,
average FEV1 decreased by 26%,
whereas Scond
increased by 390% with no significant change in
Sacin. In the
nonresponder group, FEV1 decreased
by 11%, whereas
Scond increased by 198% with no significant
Sacin change.
Despite the absence of change in
Sacin during
provocation, baseline
Sacin was
significantly larger in the responder vs. the nonresponder group. The
main findings of our study are that during provocation large
ventilation inhomogeneities occur, that the small airways affected by
the provocation process are situated proximal to the acinar zone where
the diffusion front stands, and that, in addition to overall decrease
in airway caliber, there is inhomogeneous narrowing of parallel
airways.
Fig. 1.
Schematic representation of experimental setup used for multiple-breath
washout (MBW) experiments. A nonrebreathing valve (3) separates inhaled
(IN) from exhaled air (EX) in bags. Air breathing occurs via
valves 1 and
4 and
O2 breathing via
valves 1 and
2; in both cases, pressure changes in
box or bags are recorded by pneumotachograph (5). In front of
mouthpiece sits needle valve probe (6) from
N2 analyzer. PC, personal
computer.
[View Larger Version of this Image (24K GIF file)]
1 liter and FRC 3 liters,
as was the case for the subject with the
N2 washout curve in Fig.
3A, it takes about three breaths to
reach one lung TO. The reason for using lung TO instead of breath
number on the abscissa in Fig. 3A is that it allows for better comparison of subjects with different lung
volumes and dilution (6).
The MBW tests were also analyzed according to a method first proposed in a theoretical work by Paiva (16) and subsequently applied experimentally by Crawford et al. (9). Basically, this consists of treating each expiration as a single-breath N2 washout and determining breath by breath the alveolar slope, by linear regression of N2 concentration vs. expired volume in the alveolar phase III. We used a linear regression between 0.65 liter and the end of expiration (nominally, 1 liter), with a possibility for readjustment of slope limits to avoid possible disturbance of, e.g., cardiogenic oscillations, especially in the baseline phase MBW tests. For each breath, alveolar slope is then divided by mean expired N2 concentration of that breath, to give a normalized alveolar slope (S). The inset of Fig. 2 illustrates a large increase in the normalized phase III slope between breaths 1 and 20 in the case of a provocation MBW test performed by a hyperresponsive subject. Figure 3B is a graphical representation of all S values as a function of TO, obtained in the same subject, where closed and open symbols represent the average of three baseline MBW tests and two provocation MBW tests, respectively. In Fig. 3, A and B, the provocation curves (open symbols) are used to illustrate how the MBW indexes (solid symbols) are derived. The mathematical description of these indexes, without physiological background at this point, is as follows. Derived from the N2 washout curve in Fig. 3A are its curvilinearity (Curv) and its value for TO = 6 (log[N2]6TO). Curv equals RS1/RS2, i.e., the ratio of two regression slopes in the log[N2] vs. TO plot: RS1 is the regression slope between TO = 3 and TO = 6, and RS2 is the regression slope between TO = 0 and TO = 3. In this way, Curv is always smaller than or equal to one, and a more curvilinear N2 washout curve leads to a smaller value for Curv. The other measurement of the mixing efficiency of the lung, as derived from the classic N2 washout curve in Fig. 3A, is simply the value of log[N2] for TO = 6 (log[N2]6TO). Scond and Sacin represent the contributions of the conductive airways and acinar airways, respectively, to the ventilation inhomogeneity reflected in the alveolar slopes of the MBW (see Theoretical background). The magnitude of Sacin and Scond is determined by use of the entire S curve in Fig. 3B as follows. Scond is the normalized slope difference per unit TO, which is determined by linear regression in that part of the MBW where only conductive airways are known to contribute to the rate of rise of S, i.e., between TO = 1.5 and TO = 6 (see Theoretical background). Sacin is determined by subtracting that part attributable to the conductive airways from the slope of the first breath, i.e., Scond multiplied by the TO value of the first breath (~0.3 in the case of Fig. 3B). In the example in Fig. 3B, the baseline MBW leads to Scond = 0.02 liter
1 and
Sacin = 0.15 liter1, and the provocation
MBW leads to
Scond = 0.12 liter1 and
Sacin = 0.15 liter1. In fact, a sixfold
rate of rise of the provocation S
curve (open symbols) with respect to the baseline
S curve (solid symbols) is translated
into a sixfold increase in
Scond, whereas
Sacin is
unaffected.
Essentially, characterization of the
N2 washout curve in terms of Curv
or
log[N2]6TO
(Fig. 3A) is reported here to relate to indexes that have been used in the clinical context before. For this
same reason we also computed anatomic and physiological dead space
volume of the first breath
(VDanat and
VDphys,
respectively). In contrast, the
Scond and
Sacin values
derived from the plot of normalized slope vs. lung TO (Fig.
3B) are new and also most relevant
with respect to the present study. They necessitate some degree of
background information given below, although extensive reference of
modeling (16, 17, 26) and experimental work (6-9) can be found
elsewhere.
Theoretical background.
Basically, the particular advantage of the normalized alveolar slope
S is that, as the washout progresses,
the behavior of S reveals the
mechanisms by which it is generated. In general terms, the normalized
alveolar slope is a measure of 1)
N2 concentration differences that
are generated after each O2
inspiration relative to the mean alveolar
N2 inspired concentration, and
2) the emptying pattern during each
exhalation. The larger the ventilation inhomogeneity between lung
units, the larger the normalized alveolar slope. Two major mechanisms
are held responsible for the ventilation inhomogeneities resulting in
an alveolar slope.
The first mechanism, also referred to as convection-dependent
ventilation inhomogeneity, originates from convective flow differences to and from different lung units because of differing
pressure-volume characteristics of these units. When the
least-ventilated unit (with largest
N2 concentration) empties
predominantly late in the expiration, this results in a positive
N2 slope. One of the factors that
has been hypothesized to contribute to the alveolar N2 slope is gravity-dependent flow
sequencing between upper and lower lung units. Although the lung units
involved need a priori not be as large as, e.g., entire lung regions,
they need to be subtended from airways proximal to the diffusion
front to be solely convection dependent. The second mechanism, also
referred to as diffusion-convection-dependent inhomogeneity, reflects a
far more complex diffusion-convection interaction process without
necessity for convective flow sequencing during expiration to
produce a positive N2 slope. For
this mechanism to apply, two conditions need to be fulfilled:
1) comparable magnitude of
convective and diffusive transport and
2) asymmetry of the lung structure
where diffusion and convection interaction can develop. Asymmetry may be due to unequal narrowing of parallel airways or differences in
volume subtended by two daughter branches. Even in normal healthy subjects, these two conditions are met in the lung periphery, more
specifically at the acinar level of the bronchial tree where the
diffusion front stands. In the case of abnormal lung behavior such as
airway inflammation or emphysematous lesions, asymmetry may be
increased, leading to an increased
N2 slope.
The inset of Fig.
3B shows theoretical predictions of
S generated by the two mechanisms
described above (solid lines), the sum of which typically corresponds
to a smoothed version of the experimental provocation
S curve (open symbols). The
diffusion-convection interaction produces an initial
S value that only slightly increases and very rapidly reaches a horizontal asymptote (x). This
S asymptote corresponds to an
equilibrium state of convection and diffusion, in which relative
concentration differences remain constant throughout most of the MBW.
The convective sequential emptying produces a steady increase of
S (asterisk), reflecting the fact that
concentration differences relative to the mean alveolar concentration
increase progressively because the best-ventilated lung units get
better ventilated at every subsequent inspiration. Moreover, distances between these relatively large units are too large to be covered by
diffusive transport, i.e., diffusive homogenization of the concentration differences is negligible. In fact, with this mechanism, S can only eventually reach a
horizontal asymptote if one of the large lung units gets washed out
completely.
With respect to the experimental S
curves, we can summarize here that the
S value for the first breath of the
MBW is predominantly generated by diffusion-convection-dependent
ventilation inhomogeneity in the peripheral acinar lung units. The
actual acinar contribution to ventilation inhomogeneity can be
characterized by
Sacin by subtracting the estimated convection-dependent contribution from the
slope of the first breath. The convection-dependent
ventilation inhomogeneity, which is generated by unequal inspired
concentration and flow sequencing between larger lung units, becomes
more apparent as the MBW progresses. Because these large units roughly
correspond to lung units subtended by branch points in the conductive
airway zone, we refer to
Scond for this
large-scale ventilation inhomogeneity.
Given these definitions of
Sacin and
Scond, their
baseline values should be considered as two independent indexes of
ventilation inhomogeneity in the lungs:
Sacin reflects
ventilation inhomogeneity resulting from a normal peripheral lung
structure with a given asymmetry, and
Scond results
from a given difference in ventilation between any two
diffusion-independent lung units. Whenever
Sacin undergoes
important changes with respect to baseline, this is due to an important
alteration in the peripheral lung structure. Whenever
Scond is
increased, there has been a change in the conductive airways or the
pressure-volume characteristics of the lung units subtended by these conductive airways.
Statistical analysis.
All values are given as means ± SE. Using a software package
(Primer of Biostatistics, McGraw-Hill), we performed a
repeated-measures analysis of variance followed by a post hoc
Bonferroni test for pairwise comparisons. Paired and unpaired
comparisons were made with a t-test.
For all statistical analyses, P < 0.05 was considered significant.
RESULTS
The groups of 10 nonhyperresponsive and 10 hyperresponsive subjects participating in this study will be subsequently referred to as the nonbronchohyperresponsive (NBHR) and bronchohyperresponsive (BHR) groups, respectively. Despite the fact that these subjects were neither trained nor locked into a 1-liter breathing pattern by end-inspiratory valve switching, actual tidal volume obtained from the baseline MBW tests in the NBHR and BHR groups were 1,108 ± 64 and 1,070 ± 37 (SD) ml, respectively. In addition, coefficients of variation of tidal volume within each MBW test averaged 9 ± 2 (NBHR) and 9 ± 1 (SD) % (BHR). Breathing frequency in the baseline phase was 11 ± 3 and 9 ± 2 (SD) breaths/min in the NBHR and BHR groups, respectively. No significant differences in tidal volume, its coefficient of variation, or breathing frequency were found among different phases (baseline, provocation, dilatation) nor between NBHR and BHR groups in any given phase of the study.
Table 1 lists the baseline values (mean ± SE) of all lung function and MBW parameters obtained in NBHR and BHR groups. All spirometry data reported in this section were those obtained by averaging, for each subject, the values recorded before and after the two or three MBW tests performed at a given stage (at baseline or after bronchoprovocation and bronchodilatation). FEV1 and FEF75 values recorded before and after the MBW tests showed no significant difference (P > 0.05). The MBW parameters such as Sacin and Scond are derived from an S curve obtained by averaging, breath by breath, two or three alveolar slope curves, each one computed from one MBW test. The same procedure was followed to derive Curv and log[N2]6TO from an average of two or three washout curves. VDanat and VDphys were average values of two or three values of anatomic and physiological dead space, respectively, determined in the first breath of each MBW. Of the lung function parameters, FEV1, FEV1/FVC, and FEF75, and of the MBW parameters, Sacin, were significantly different between the NBHR and the BHR groups (Table 1).
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Table 2 shows how bronchoprovocation and bronchodilatation affect lung function parameters (FEV1, FEF75), dead space volumes (VDanat, VDphys), N2 washout characteristics (Curv, log[N2]6TO), and proximal and peripheral MBW components of ventilation inhomogeneity (Scond, Sacin). We performed a repeated-measures analysis of variance on each of the parameters in Table 2, with a Bonferroni t-test for pairwise comparisons to check the significance of the following: changes from baseline after provocation (between baseline and provocation), reversal of changes after bronchodilatation (between provocation and bronchodilatation), and return to baseline after bronchodilatation (between bronchodilatation and baseline). The latter was not represented in Table 2, but the result was that, only for FEV1 (in NBHR and BHR groups) and for FEF75 (in NBHR group), values were not entirely back to baseline after dilatation. We also checked with an unpaired t-test whether an increase or decrease of a parameter from baseline to provocation was significantly larger in the BHR group than in the NBHR group and found that this was the case only for FEV1, PEF, FEF75, Curv, and log[N2]6TO (Table 2).
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Figure 4 shows a graphical representation of the FEV1 and FEF75 decreases (A), and Sacin and Scond increases (B). Closed and open symbols are averages in the NBHR group and the BHR group, respectively. For changes in FEV1 with respect to baseline, which is taken as the criterion for hyperresponsiveness, we show the distribution of 20 individual data points (some of which are superimposed), indicating a continuous distribution of hyperresponsive and nonhyperresponsive subjects in terms of FEV1. After bronchoprovocation, average reduction in FEF75 (31% in the NBHR group; 49% in the BHR group) was significantly greater than reduction in FEV1 (11% in NBHR; 26% in the BHR group). Scond showed an average 198% increase in the NBHR group and an average 390% increase in the BHR group on bronchoprovocation, with large standard error bars and with no significant difference between the Scond increase in the BHR and NBHR groups (P = 0.06). Sacin showed only a 13-20% increase in both groups, which did not reach statistical significance.
, NBHR group; 
, BNR group.
Table 2 shows that
VDanat
decreased to the same extent, namely, 20 ml, in the NBHR and BHR
groups. However, because on average VDanat was
10 ml lower in the BHR group, the relative change was slightly more
important in the BHR group. For
VDphys, no
significant changes could be demonstrated in any group. Nevertheless,
the so-called "alveolar dead space," defined as
VDphys 
VDanat, did increase significantly.
VDanat and
VDphys VDanat were
back to baseline control values after bronchodilatation.
Bronchoprovocation leads to significantly increased curvilinearity of
the N2 washout curve (i.e.,
decreased value of Curv) only in the BHR group. The change in Curv did
not reach statistical significance in the NBHR group. Also, the
logarithmic value of the mean expired
N2 concentration for TO = 6 (i.e.,
log[N2]6TO)
increased significantly on bronchoprovocation only in the BHR group.
Both these parameters derived from the N2 washout curve were back to
baseline control values after bronchodilatation.
DISCUSSION
To evaluate the role of conductive and acinar lung zones in the
histamine bronchoprovocation process, an analysis of the normalized slope in the N2
MBW (9) was applied, from which two indexes of
ventilation inhomogeneity, i.e.,
Scond and
Sacin, were
derived. In the process of extracting a small set of parameters from
the MBW normalized slopes for quantitative analysis, we based our choice of Sacin
and Scond on
elements from the MBW papers of Crawford et al. (6-9) and
subsequent model analysis by our laboratory (26).
Sacin was based
on the extrapolation method decribed by Crawford et al. (9) but used
linear instead of exponential fitting on the latter part of the MBW and
used lung TO instead of breath number as the abscissa, as suggested in
a subsequent paper (6). The computation of the conductive component
Scond only
differed from the method described by Crawford et al. (8), where linear
regression was also used, by the choice of lung TO as the abscissa.
Because the subjects in the series of papers by Crawford et al. (8) are
not documented in terms of parameters that are pertinent to our study
(e.g., hyperresponsive or nonhyperresponsive, FEF75 values, diffusing capacity),
direct comparison with our baseline MBW data needs to be handled with
caution. However, using the lung volumes reported in Crawford et al.
(6) from subjects who also participated in another study by Crawford et
al. (7), we used the average normalized slope curve in their Fig.
3 (Ref. 9) to evaluate
Sacin and
Scond according
to our method, yielding Sacin = 0.070 liter1 and
Scond = 0.027 liter1. When possible
differences in subjects, flow rates, equipment, and slope-determination
limits are taken into account, these values are in general agreement
with our baseline
Sacin and
Scond data (Table
1).
Besides the small methodological differences in the computation of
convective and diffusion-convective inhomogeneity indexes, the main
issue remains whether the exponential extrapolation rather than the
linear extrapolation to the slope of the first breath (to obtain
Sacin) could
lead to different results. Model analysis actually predicts that the
normalized slope curve vs. lung TO has two components, one fast and one
slow (insets, Fig.
3B). That is why we also submitted
all our data (including those from provocation and dilatation MBW
tests) to a Levenberg-Marquardt routine, which fitted a sum of two
exponentials (with boundary conditions on the curvature and asymptotes)
to our normalized slope curves as a function of lung TO. Using the
statistical method of Bland and Altman (2) to compare
Sacin values
obtained with two-exponential vs. linear extrapolation, we obtained a
mean difference of 0.004 ± 0.005 (SD)
liter1. Placed against the
baseline Sacin
values in Table 1, this result led us to conclude that, even in extreme
conditions such as provocation, our linear method for
Sacin
determination is as valid as the more cumbersome exponential method.
In this study, it was expected that 1) Scond would increase if large ventilation differences and asynchronous emptying occur, e.g., as a result of inhomogeneous narrowing of parallel conductive airways; and 2) Sacin would increase if any significant alteration occurs at the level of the acinar structure, even in the absence of flow asynchrony. We observed large Scond increases during histamine-induced airway narrowing in both BHR and NBHR groups, and no significant changes in Sacin in either group. However, prehistamine Sacin was significantly larger in the BHR group. Together with measurements of dead space (VDanat, VDphys) and lung function parameters (FEV1, FEF75), our MBW study suggests that 1) the airways involved during the histamine bronchoprovocation process, part of which are the small airways, are situated proximal to the acinar entrance; 2) between relatively large lung units, i.e., those containing several groups of acini, large differences in inspired gas concentration develop; and 3) the baseline acinar structure is not affected by the provocation process itself but may be related to hyperresponsiveness.
Large-scale inhomogeneities (Scond ). The fact that histamine bronchoprovocation generates average Scond increases on the order of 200 and 400% in the NBHR and BHR groups, respectively (Fig. 4B), indicates not only an average decrease in airway lumen of parallel airways down to a given level or at a given level of the bronchial tree (which reduces FEV1 and FEF75) but also an important inhomogeneity in constriction between parallel airways. Indeed, to generate an alveolar slope, parallel differences in ventilation distribution must exist, associated with sequential emptying. Therefore, the present results suggest that the inhomogeneity of airway narrowing that is known to exist in the case of acute asthmatic attack is also present, although to a lesser degree, during bronchoprovocation with a nonspecific agent in asymptomatic subjects. The inequality in response of parallel airways could reflect density differences in muscarinic receptors and/or cholinergic innervation between airways located at a given lung depth (i.e., airways of more or less the same lung generation), in addition to proximal vs. peripheral density differences observed along the bronchial tree (13). Another category of MBW indexes that can reflect convection-dependent inhomogeneities is that derived from the classic washout curve, Curv and log[N2]6TO (Fig. 3A). Their modifications after bronchoprovocation did not reach statistical significance in the NBHR group (Table 2). From a theoretical viewpoint, this is surprising because these two parameters should reflect all convective, i.e., large-scale, concentration differences. In particular, the specific ventilation differences between lung units that empty asynchronously during expiration and therefore increase Scond should also tend to decrease Curv and increase log[N2]6TO. In addition, possible specific ventilation differences generated between lung units that empty synchronously, and therefore do not contribute to Scond, would nevertheless tend to decrease Curv and increase log[N2]6TO even more. Therefore, the absence of significant change in Curv or log[N2]6TO and the twofold increase in Scond (200% baseline) in the NBHR group could indicate that specific ventilation differences are small, whereas flow asynchrony is more apparent during mild bronchoprovocation. In contrast, in the BHR group, both specific ventilation and flow asynchrony become important enough to affect all large-scale MBW parameters (Table 2). Alternatively, it could be argued that any index derived from the classic N2 washout curve, whether it be related to its curvilinearity (such as Curv) or to its value after a number of breaths or lung TO (such as log[N2]6TO), is not sensitive enough to detect the mild bronchoprovocation in the NBHR group. Small-scale inhomogeneities (Sacin ). Our results show that, in contrast to Scond, Sacin is not significantly affected by the bronchoprovocation process itself (Fig. 4B). Nevertheless, Sacin is significantly larger in the hyperresponsive subjects (Table 1). Using the alveolar N2 slope of the vital capacity SBW test, Hudgel and Roe (11) found a larger baseline slope for 9 hyperresponsive subjects in a group of 22 coal miners. Inasmuch as the slope of a SBW maneuver and Sacin (i.e., a major part of the slope of the first breath in a MBW) can reflect, at least in part, the same ventilation inhomogeneity, the findings in the group of miners are compatible with our baseline Sacin data. However, Hudgel and Roe could not demonstrate such a baseline N2 slope difference between 8 responders and 33 nonresponders in a group of 41 nonminers. Taylor and Clarke (25) also failed to observe a different baseline SBW alveolar N2 slope in the responder vs. nonresponders to histamine in a group of 21 nonsmoking subjects. One reason could be that, in the case of small changes and for small groups of subjects, the computation of Sacin, without the confounding effects of large-scale inhomogeneities as would be the case with a SBW alveolar slope, is indeed crucial. The larger baseline Sacin value in the BHR group should be interpreted with caution. It merely adds a piece of information to the controversy about the relationship between baseline lung function and hyperresponsiveness (24). The significantly smaller FEV1 and FEF75 values in the BHR group (Table 1) are in support of such a dependence in a group of 20 otherwise asymptomatic subjects. Nevertheless, FEV1 and FEF75 averages are supranormal or normal in both groups. Because baseline DLCO and KCO values are normal and not different between BHR and NBHR groups (Table 1), it is unlikely that intra-acinar alterations reflected in the larger Sacin in the BHR group took place at the level of the alveolar structure. Rather, the larger Sacin points to some degree of intra-acinar airway narrowing, maybe due to inflammation (23). This issue surely needs further investigation in a larger group of subjects with different degrees of hyperresponsiveness, possibly a group that also includes symptomatic subjects, in whom Sacin can, for instance, be related to PD20, the provocative dose necessary to reach a 20% fall in FEV1. VDanat and VDphys. In contrast to the other MBW parameters, VDanat derived from the first expiration showed a very similar decrease in the BHR and NBHR groups, indicating a similar degree of volumetric reduction of the conductive airways. VDanat is expected to be less sensitive to airway narrowing than any resistance-related parameter simply because VDanat is related to the second power of the airway radius, whereas resistance is related to approximately the fourth power of the airway radius. Alternatively, one could argue that a possible increase in lung volume after provocation (20, 27) would tend to oppose a VDanat decrease. We did not find a significant change in FRC after bronchoprovocation in any of the two groups, in line with FRC measurements by Langley et al. (12) using the same technique. Despite these arguments for a lack of sensitivity of VDanat to evaluate bronchospasm, the fact that it does not decrease more in the BHR group at all remains surprising. Perhaps it is an indication of upper airway constriction with a limit that is already reached in the NBHR group, where the average FEV1 decrease was 12%. We did not find a correlation between VDanat and FEV1 decrease in the NBHR group, a correlation that could have confirmed the hypothesis of a progressive VDanat decrease with FEV1 below the 20% FEV1 threshold. However, the range of changes in VDanat is probably too small to verify this. Model analysis predicted that, for the study of ventilation distribution in normal human subjects, VDphys, or the difference VDphys VDanat, is
not very sensitive to evaluate changes in ventilation inhomogeneity
(26). Our experimental
VDphys and VDphys VDanat data
confirm that the same is true in the case of bronchoprovocation, during
which important inhomogeneities are known to occur. In general, our
dead space data coincide with the findings of Burke et al. (4), who
also found a 20-ml decrease of
VDanat and
virtually no effect on
VDphys.
Implication of bronchoprovocation in gas-exchanging units.
Despite the small effect of histamine provocation on
VDphys and
VDanat,
Burke et al. (4) found a large degree of ventilation-perfusion mismatch, and our data provide an explanation for this gas-exchange impairment. The fact that
Scond increases
so dramatically points to large differences in gas concentration
between relatively large units, comprising several acini or clusters of
acini. The size of these units remains somewhat speculative. In
combination with a similar
VDanat
decrease in the NBHR and BHR groups, the larger Scond increase in
the BHR with respect to the NBHR group is an indication of the fact
that the ventilation differences during bronchoprovocation are
generated between units subtended by the more peripheral of the
conductive airways. The large decreases in
FEF75, which, despite its poor
reproducibility, is used in clinical practice as a marker of the small
airways, provide further support for this (Fig.
4A). The implication of the smaller
conductive airways in the bronchoprovocation process is also not
surprising in view of the conclusions of a MBW study in normal subjects
(7) suggesting that bronchomotor tone of relatively small airways is
responsible for a relatively uniform distribution of ventilation. When
normal bronchomotor tone is disturbed and large inspired concentration
differences occur as a result, gas-exchange impairment is likely to
ensue.
Our suggestion that the small conductive airways are the major
determinant of gas-exchange impairment during bronchoprovocation is
also compatible with bronchoprovocation data in the literature. In the
case of the bronchoprovocation study by Olgiati et al. (15), one needs
to assume that the peripheral airways that were held responsible for
impairment of gas exchange were indeed small airways but nevertheless
were situated proximal to the acini, i.e., conductive airways. This is
probably also the reason why Schmekel et al. (21) found no difference
in impairment of gas exchange, whether methacholine was deposited
centrally or peripherally in the lungs. With central vs. peripheral
deposition of histamine, Ruffin et al. (19) even found that a lower
dose was necessary to decrease
FEF75 in the case of central
deposition. Although these authors concluded that action on central
airways was the main determinant of histamine provocation, doubt
remained about the dispersion of the histamine dose over the more
numerous peripheral airways, leading to submaximal reaction of the very
peripheral airways. The same reasoning could lead us to believe that
this is why Sacin
did not change significantly (Fig.
4B). It is possible that, to elicit
the hypothesized peripheral action of histamine, also at the acinar
level, intravenous injection of histamine would be more appropriate.
Potential of the MBW method.
Provided one corrects for the convective component, as was done here to
obtain Sacin, the
alveolar slope of the first breath of a MBW may be considered as
reflecting intra-acinar alterations. The fact that the
convection-dependent part, which turns out to be the most important
effect during bronchoprovocation, is only poorly reflected in the first
breath probably explains the relatively moderate increases in the SBW
phase III slope increases seen after provocation (14, 20).
Nevertheless, the first breaths of a MBW and the SBW are not strictly
comparable because, for the SBW vital capacity maneuver, the conductive
and acinar contribution to ventilation inhomogeneity may be quite
different (17). In the study by Scano et al. (20), this relative
contribution is further complicated by the end-inspiratory breath hold
of 5-10 s, which tends to reduce the acinar contribution of the
alveolar slope.
The MBW test has been used in association with bronchoprovocation tests
before. Langley et al. (12) quantified ventilation distribution in
terms of a mixing-efficiency index, derived from the classic
N2 washout curve. A MBW was
performed before and after methacoline provocation and
81mKr ventilation lung scans were
obtained in both phases. From the inspection of the patches on the lung
scans, Langley et al. (12) concluded that convection-dependent
inhomogeneity alone could not account for the marked decrease in MBW
mixing efficiency, and that some diffusion-related mixing inefficiency
must be involved in the provocation process. Our interpretation of
those data is that important ventilation inhomogeneities exist between
the smaller convection-dependent units, which also contribute to
decrease the MBW mixing efficiency, but cannot be distinguished on the ventilation scans because of the poor resolution. In fact, the MBW data
published by Harris et al. (10) showed that bronchoprovocation had the
same effect on He- and sulfur hexafluroide mixing-efficiency curves.
Inasmuch as these curves (obtained in only 3 asthmatic subjects) are
sensitive enough to make the diffusion-dependent part of mixing
efficiency appear, these data also contradict the hypothesis of a
diffusion-dependent component in the provocation process. The fact that
in our study
Sacin did not
increase significantly after bronchoprovocation confirms the findings
of Harris et al. in a larger group of hyperresponsive, but otherwise
asymptomatic, subjects.
Of interest is that bronchoprovocation has been reported to generate a
reversal of the apex-to-base ventilation gradient (5, 27). With the
gravity-dependent vertical pressure gradient around the lung remaining
as it is, the reversal of the apex-to-base ventilation gradient after
bronchoprovocation would lead to a negative alveolar slope, or at least
the gravity-dependent part of it. This opposing gravity-dependent
effect could have contributed to counteract the alveolar slope increase
resulting from much smaller lung units. However, human physiological
experiments recently performed onboard the Spacelab Life Sciences 1 mission show that the MBW maneuver, involving near-tidal breathing from
FRC, and the normalized slopes it generates are not significantly
affected by gravity (18). This means that, in the absence of the
blurring effect of gravity, the alveolar slope generated in a MBW test can be entirely attributable to intrinsic structural and elastic properties of the lung, a fact that renders this test particularly useful in the clinical context of the lung function laboratory.
In conclusion, our MBW results suggest that, in otherwise asymptomatic
subjects with airway hyperresponsiveness to inhaled histamine, airway
narrowing occurs predominantly in airways proximal to the acini. These
airways are at the point of origin of important inspired gas
concentration differences between relatively large units and of a
sequential emptying pattern between them. By contrast, the
acinar component of ventilation inhomogeneity is not affected by the
bronchoprovocation itself, but its baseline value is significantly increased in the BHR group of subjects.
ACKNOWLEDGEMENTS
We thank Johan Goris from the Biotechnology Department of Academisch Ziekenhuis, Vrije Universiteit Brussel, for technical support.
FOOTNOTES
Address for reprint requests: S. Verbanck, AZ-VUB, Dienst Pneumologie (CPNE), Laarbeeklaan 101, 1090 Brussels, Belgium (E-mail: pnevks{at}az.vub.ac.be).
Received 26 December 1996; accepted in final form 25 July 1997.
REFERENCES
| 1. | American Thoracic Society Statement. Standardization of spirometry. Am. Rev. Respir. Dis. 136: 1285-1298, 1987[Medline]. |
| 2. | Bland, J. M., and D. G. Altman. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet 1: 307-310, 1986[Medline]. |
| 3. |
Brown, N.,
E. McFadden,
and
R. H. Ingram, Jr.
Airway responses to inhaled histamine in asymptomatic smokers and non-smokers.
J. Appl. Physiol.
42:
508-513,
1977 |
| 4. |
Burke, T. V.,
M. Kung,
and
N. K. Burki.
Pulmonary gas exchange during histamine-induced bronchoconstriction in asthmatic subjects.
Chest
96:
752-756,
1989 |
| 5. | Clague, H., D. Ahmad, M. J. Chamberlain, W. K. C. Morgan, and S. Vinitski. Histamine bronchial challenge: effect on regional ventilation and aerosol deposition. Thorax 38: 668-675, 1983[Abstract]. |
| 6. | Crawford, A. B. H., D. J. Cotton, M. Paiva, and L. A. Engel. Effect of lung volume on ventilation distribution. J. Appl. Physiol. 66: 2505-2510, 1989. |
| 7. |
Crawford, A. B. H.,
M. Makowska,
and
L. A. Engel.
Effect of bronchomotor tone on static mechanical properties of the lung and ventilation distribution.
J. Appl. Physiol.
63:
2278-2285,
1987 |
| 8. |
Crawford, A. B. H.,
M. Makowska,
S. Kelly,
and
L. A. Engel.
Effect of breath holding on ventilation maldistribution during tidal breathing in normal subjects.
J. Appl. Physiol.
61:
2108-2115,
1986 |
| 9. |
Crawford, A. B. H.,
M. Makowska,
M. Paiva,
and
L. E. Engel.
Convection- and diffusion-dependent ventilation maldistribution in normal subjects.
J. Appl. Physiol.
59:
838-846,
1985 |
| 10. | Harris, E. A., P. R. Buchanan, and R. M. L. Whitlock. Human alveolar mixing efficiency for gases of differing diffusivity in health and airflow limitation. Clin. Sci. (Lond.) 73: 351-359, 1987[Medline]. |
| 11. | Hudgel, D. W., and R. Roe. Nonspecific airway hyperreactivity in nonsmoking bituminous coal miners demonstrated by quantitative methacholine inhalation challenge. J. Lab. Clin. Med. 111: 684-691, 1988[Medline]. |
| 12. | Langley, F., K. Horsfield, G. Burton, W. A. Seed, S. Parker, and G. Cumming. Effect of inhaled methacholine on gas mixing efficiency. Clin. Sci. (Lond.) 74: 187-192, 1988[Medline]. |
| 13. | Mak, J. C., and P. J. Barnes. Autonomic receptors in the upper and lower airways. In: Neuropeptides in Respiratory Medicine, edited by M. A. Kaliner, P. J. Barnes, G. H. Kunkell, and J. N. Baranicek. New York: Dekker, 1994, vol. 4, p. 251-274. (Clin. Allergy Immunol. Ser.) |
| 14. | Nielsen, J., M. Dahlqvist, H. Welinder, Y. Thomassen, R. Alexandersson, and S. Skerfving. Small airways function in aluminium and stainless steel welders. Int. Arch. Occup. Environ. Health 65: 101-105, 1993[Medline]. |
| 15. | Olgiati, R., S. Birch, A. Rao, and A. Wanner. Differential effects of methacholine and antigen challenge on gas exchange in allergic subjects. J. Allergy Clin. Immunol. 67: 325-329, 1981[Medline]. |
| 16. | Paiva, M. Two pulmonary functional indexes suggested by a simple mathematical model. Respiration 32: 389-403, 1975[Medline]. |
| 17. | Paiva, M., and L. A. Engel. Gas mixing in the lung periphery. In: Respiratory Physiology: An Analytical Approach, edited by H. K. Chang, and M. Paiva. New York: Dekker, 1989, vol. 40, p. 245-276. (Lung Biol. Health Dis. Ser.) |
| 18. |
Prisk, G. K.,
H. J. B. Guy,
A. R. Elliott,
M. Paiva,
and
J. B. West.
Ventilatory inhomgeneity determined from multiple-breath washouts during sustained microgravity on Spacelab SLS-1.
J. Appl. Physiol.
78:
597-607,
1995 |
| 19. | Ruffin, R. E., M. B. Dolovich, R. K. Wolff, and M. T. Newhouse. The effects of preferential deposition of histamine in the human airway. Am. Rev. Respir. Dis. 117: 485-492, 1978[Medline]. |
| 20. | Scano, G., L. Stendardi, M. Bracamonte, A. Decoster, and R. Sergysels. Site of action of inhaled histamine in asymptomatic asthmatic patients. Clin. Allergy 12: 281-288, 1982[Medline]. |
| 21. |
Schmekel, B.,
H. Hedenström,
M. Kämpe,
L. Lagerstrand,
G. Stalenheim,
P. Wollmer,
and
G. Hedenstierna.
The bronchial response, but not the pulmonary response to inhaled metacholine is dependent on the aerosol deposition pattern.
Chest
106:
1781-1787,
1994 |
| 22. | Sekizawa, K., M. Yanai, Y. Shimizu, H. Sasaki, and T. Takishima. Serial distribution of bronchoconstriction in normal subjects. Am. Rev. Respir. Dis. 137: 1312-1316, 1988[Medline]. |
| 23. | Smith, L., and E. R. McFadden, Jr. Bronchial reactivity revisited. Ann. Allergy Asthma Immunol. 74: 454-469, 1995[Medline]. |
| 24. | Suppli, C. Bronchial responsiveness to inhaled histamine in both adults with intrinsic and extrinsic asthma: the importance of prechallenge forced expiratory volume in 1 second. J. Allergy Clin. Immunol. 91: 120-126, 1993[Medline]. |
| 25. | Taylor, R. G., and S. W. Clarke. Bronchial reactivity to histamine in young male smokers. Eur. J. Respir. Dis. 66: 320-326, 1985[Medline]. |
| 26. |
Verbanck, S.,
and
M. Paiva.
Model simulations of gas mixing and ventilation distribution in the human lung.
J. Appl. Physiol.
69:
2269-2279,
1990 |
| 27. | Whyte, K. F., M. Ip, T. Kirby, C. G. Wathen, and D. C. Flenley. Changes in regional ventilation during histamine bronchial challenge in stable asthma. Respiration 61: 68-73, 1994[Medline]. |
This article has been cited by other articles:
|
|
 |
|
  V. Suresh, D. A. Shelley, H.-W. Shin, and S. C. George Effect of heterogeneous ventilation and nitric oxide production on exhaled nitric oxide profiles J Appl Physiol, June 1, 2008; 104(6): 1743 - 1752. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Verbanck, Y. Kerckx, D. Schuermans, W. Vincken, M. Paiva, and A. Van Muylem Effect of airways constriction on exhaled nitric oxide J Appl Physiol, April 1, 2008; 104(4): 925 - 930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Verbanck, D. Schuermans, S. Van Malderen, W. Vincken, and B. Thompson The effect of conductive ventilation heterogeneity on diffusing capacity measurement J Appl Physiol, April 1, 2008; 104(4): 1094 - 1100. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Verbanck, D. Schuermans, M. Paiva, M. Meysman, and W. Vincken Small Airway Function Improvement after Smoking Cessation in Smokers without Airway Obstruction Am. J. Respir. Crit. Care Med., October 15, 2006; 174(8): 853 - 857. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. M. Olfert and G. K. Prisk Effect of 60{degrees} head-down tilt on peripheral gas mixing in the human lung J Appl Physiol, September 1, 2004; 97(3): 827 - 834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Verbanck, D. Schuermans, M. Meysman, M. Paiva, and W. Vincken Noninvasive Assessment of Airway Alterations in Smokers: The Small Airways Revisited Am. J. Respir. Crit. Care Med., August 15, 2004; 170(4): 414 - 419. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P.M. Gustafsson, P. Aurora, and A. Lindblad Evaluation of ventilation maldistribution as an early indicator of lung disease in children with cystic fibrosis Eur. Respir. J., December 1, 2003; 22(6): 972 - 979. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Gronkvist, E. Bergsten, O. Eiken, and P. M. Gustafsson Inter- and intraregional ventilation inhomogeneity in hypergravity and after pressurization of an anti-G suit J Appl Physiol, April 1, 2003; 94(4): 1353 - 1364. [Abstract] |
ABSTRACT
