Вы находитесь на странице: 1из 13

Respiration Physiology, 74 (1988) 163-176 163

Elsevier

RSP 01464

High-frequency oscillatory ventilation may increase


airway closure
R.B. Filuk, D.J. Berezanski, P.A. Easton and N.R. Anthonisen
Section of Respiratory Diseases, Department of Medicine, University of Manitoba, Canada
(Accepted for publication 23 June 1988)

Abstract. In five seated, normal subjects, we measured closing volumes using 133Xe boluses inhaled at
residual volume. High frequency oscillatory ventilation (HFOV) (15 Hz, 2 cc/kg) was applied during either
inspiration to total lung capacity or the subsequent expiration. Closing volume was increased (P < 0.001)
when HFOV was applied during the latter half of expiration, but not when HFOV was applied during
inspiration or the first half of expiration. Subsequently, in seven subjects, we measured the regional
distributions of t33Xe boluses delivered during open-glottis breath-hold at 14~ vital capacity after equili-
bration with N20. HFOV was applied during bolus delivery for about 16 sec. These distributions were
compared with those achieved by intravenous injections of 133Xe in saline. Regional perfusion (injected
isotope) exceeded regional N20 uptake at the lung bases and this was significantly accentuated by HFOV,
compatible with increased basal closure. We conclude that in normal subjects at low lung volumes, HFOV
may enhance airway closure, though other explanations are possible.

Airway closure; High frequency oscillatory ventilation (HFOV); Human; ~33Xe

The mechanisms of gas transport in high frequency oscillatory ventilation (HFOV) are
controversial (Chang, 1984; Drazen e t a L , 1984), and it is clear that interactions
between rapidly oscillating pressures and flows and the lungs and airways may be
complex. For example, in intact animals, HFOV has been associated with hyperinflation
with alveolar pressure exceeding those measured at the airway opening (Simon et al.,
1984), perhaps due to dynamic airway compression during the expiratory phase of the
HFOV cycle (Bryan and Slutsky, 1986; Solway et al., 1986).
We postulated that HFOV might affect airway closure. If, at lung volumes normally
accompanied by airway closure, rapidly oscillating pressures were applied to the airways
which closed, it is possible that the airways would be unable to 'follow' the pressure.
That is, airways opened by the positive pressure stroke of the oscillator might not close
on the reciprocating stroke or vice versa. We attempted to study this issue by examining
the effects of HFOV on regional gas distribution in seated normal subjects at low lung
volumes.

Correspondence address: N.R. Anthonisen, Respiratory Investigation Unit, Health Science Center, 700
William Avenue, Winnipeg, Manitoba, Canada R3E 0Z3.

0034-5687/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)


164 R.B. FILUK et al.

Methods

We studied 12 seated subjects (10 male, 2 female) who were aged 27-49 years with no
history of respiratory disease. Five subjects completed the first set of experiments and
seven completed the second. Informed consent was obtained from each subject under
the guidelines and approval of the Ethics Committee of the University of Manitoba.
The oscillation pump was a piston driven by an electrical motor which produced
sinusoidal variations in pressure and flow. It was set to produce a piston displacement
of 2 cc/kg at a frequency of 15 Hz. For all experiments subjects wore nose clips and
breathed through a mouthpiece which was connected to a stopcock. The stopcock was
connected by 10 cm of rigid tubing with an internal diameter of 3 cm to a Y-shaped
connector. The oscillator was attached to one arm of the connector by 15 cm of tubing
containing another three-way stopcock so that the oscillator could be connected either
to the breathing circuit or to room air. The other arm of the Y connector was attached
by a 1.5 m segment of similar tubing to a 9 L water-filled spirometer with a metal bell.
When the oscillator was connected to the breathing circuit and activated, both the
subject and the spirometer were subjected to HFOV. During oscillation, subjects
supported their cheeks with their hands.
In the first series of experiments, the effect of HFOV on closing volume was assessed
using the methods of Dollfuss e t al. (1967). Boluses of 133Xe were inhaled at the onset
of vital capacity (VC) inspirations and expired 133Xe recorded as a function of lung
volume during the subsequent vital capacity expirations. Expired 133Xe was measured
with a scintillation counter mounted over the mouthpiece, and volume measured with
a potentiometer mounted on the spirometer. The outputs of the scintillation counter and
the potentiometer were recorded on a chart and X-Y recorder. With the mouth stopcock
turned so that the mouthpiece was connected to room air, the subject inhaled deeply
and expired to residual volume (RV). The stopcock was turned, connecting the subject
to the spirometer and oscillator, and a bolus (approximately 2 mCi 133Xe in 3 cc air)
was injected into the mouthpiece. The subject then inspired slowly ( < 0.3 L/sec) to total
lung capacity (TLC) and after a 10 sec breath hold, expired slowly ( < 0.3 L/sec) back
to RV. The subject was then connected to room air and the spirometer disconnected
and its contained 133Xe washed out. All maneuvers were conducted with the oscillator
turned on. During control measurements, the oscillator stopcock was turned so that the
oscillator was connected to room air. By manipulating this stopcock, during other
maneuvers, HFOV was applied to the subjects throughout inspiration or expiration,
during inspiration of the first 50~o of the VC, and during the first and last 50~o of the
expiratory VC. Washout curves were discarded if the VC was not reproducible within
5 ~o, and two washout curves were obtained for each condition in each subject. Closing
volumes were designated as the volume at which 133Xe concentration underwent an
abrupt increase, departing from the gradual increase of phase III (Anthonisen e t al.,
1978). In addition, the slope of phase III - change in concentration per liter of VC -
and the height of phase IV - change in concentration - were measured for each curve
and indexed according to the 133Xe concentration at 50yo VC. Data acquired under the
same conditions were averaged.
AIRWAY CLOSURE WITH HFOV 165

In the second set of experiments, regional airway closure was studied using the
method of Engel et al. (1975). This technique compares the regional distribution of
]33Xe boluses delivered intravenously with regional distributions of 133Xe boluses
delivered during open-glottis breath hold after equilibration with N20. During the latter,
boluses are injected at the mouth, carried into the lungs by N 2 0 uptake, and distributed
within the lung according to regional N 2 0 uptake which, in the presence of open
airways, is proportional to regional blood flow. Airway closure was interpreted as
present when regional concentrations of injected isotope clearly differed from those
achieved when 133Xe was carried into the lung by N20.
Seven subjects were seated in an air-conditioned body plethysmograph with two
vertical arrays of six scintillation counters positioned behind them to assess the distribu-
tion ofregional~33Xe in the lung (fig. 1). The vertical distance between each counter was
5 cm and they were shielded by lead collimators with slits 1 cm high and 7.5 cm wide
to eliminate overlap between the counting fields of adjacent counters. The subject's
position relative to the counters was maintained by a light beam aimed at the thorax.
The outputs of the scintillation counters were measured with rate meters and recorded
on a chart recorder, along with plethysmograph volume which was sensed by a Krogh
spirometer. Plethysmograph volume was displayed to the subjects on an oscilloscope.
For these experiments, in addition to the bell spirometer, the mouth stopcock was
connected to a source of 80~o N 2 0 - 2 0 yo 02, a closed rebreathing circuit and to a small
breathing valve, the expiratory side of which was connected to a fume hood, the
inspiratory side to room air (fig. 1). The subjects inhaled two VC breaths of the N 2 0
mixture and, after the second, exhaled to a pre-determined lung volume where they were
switched to the bell spirometer and held their breath, glottis open, while a bolus of 133Xe
was injected into the mouthpiece. Successful breath holds were characterized by con-

Fume
Hod/(£ P

Fig. 1. Schematic diagram of the apparatus used in the N20 study. Subjects sat in a plethysmograph(P),
the volumeof which was sensed by a spirometer (K). They breathed through a tap which could be connected
to a bag containing N20 (A), the oscillator (O) and bell spirometer (B), a closed circuit spirometer for
equilibration (E), or a valve the expiratory side of which was connected to a fume hood. Bolus injections
were given at the mouth (I) and regional counts monitored by stationary counters (C).
166 R.B. FILUK etal.

stant plethysmograph volume and declining volume of the bell spirometer as its gas
entered the lungs consequent to N20 uptake. During successful breath holds which
lasted about 25 sec, regional count rates increased and stabilized. When this occurred
the subject inhaled to TLC, where regional count rates were recorded. Control and
experimental maneuvers differed in that, during the latter, oscillations were applied
during the open glottis breath hold. Each subject performed at least two satisfactory
maneuvers with and without HFOV, and the results of each averaged.
Regional blood flow was determined by injecting 133Xe (1-2 mCi) dissolved in 2-3 cc
of saline through a large forearm vein. Injections were given at the same lung volume
as the N20 study with the same volume history and breath hold. Subjects did not
pre-breathe N20 since this does not alter regional perfusion distribution (Forkert et al.,
1979). In one subject, perfusion distribution was measured both with and without
HFOV being applied during the injection. Like the N20 studies, injection studies were
done in duplicate and averaged.
To calculate the distribution of each bolus of 133Xe in terms of regional concentration,
each subject also underwent an equilibration maneuver (Dollfuss et al., 1967). This was
done using 6-7 mCi of 133Xe in a rebreathing circuit. After equilibration, regional count
rates were measured during breath holds at TLC.
Data analysis involved averaging the right and left lung counters in the six regions
studied. The count rates were converted to regional concentrations using the following
equation to correct for differing bolus sizes as well as regional configurations:

[Xe] = --Ub X ~ U e x 100


Ue ~Ub

where [Xe] is regional Xe concentration, U b is the regional count rate at TLC resulting
from bolus administration, U e is the count rate of the same region at TLC following
equilibration, ~ U b is the sum of the regional bolus count rates, and ~ U e is the sum
of the regional equilibration count rates. If distributions were even, [Xe] would be 100
in all lung regions. To calculate airway closure we compared the distribution of 133Xe
delivered by N20 uptake ([Xe]N2o) and delivered by injection ([Xe]iv). The ratio of
[Xe]iv in apical regions to [Xe]iv in basal regions should be the same as the ratio of
apical [Xe]N2 o to basal [Xe]N2O if all airways are patent, and reductions in basal
[Xe]N2o in relation to basal [Xe]i v reflect airway closure (Engel et al., 1975). Basal
[Xe]N2o/[Xe]i v divided by apical [Xe]N2o/[Xe]i v - the latter fraction representing the
ratio which should be obtained in all lung regions if all units were open - is equal to
the fraction of patent airways in basal regions and:

[Xe]iv apex [Xe]N2O base


~o airway closed = 1 x x 100
[Xe]N2o apex [Xe]iv base
AIRWAY CLOSURE WITH HFOV 167

Using this relationship we calculated mean closure in regions 2 5 - 3 0 cm from the apex
in comparison with the mean of regions 5-10 cm from the apex.
The first set of experiments were analyzed using A N O V A with Tukey's multiple
comparison testing. Student's paired t-test was used to determine statistical significance
in the second set of experiments.

Results

Figure 2 shows individual single breath washouts from one subject. Oscillation during
all or the latter part of expiration clearly increased closing volume, and oscillations
applied either during inspiration and expiration decreased the height o f phase IV but did
not affect phase III. Mean data for closing volume and the height of phase IV are shown
in fig. 3, and support the trends shown in fig. 2. Oscillation decreased the height of

TOT,',N,,. LATE EXP. (off/on) 1

.,~.ll~llm'~ .- L,, .,t ~/,~t ~ It ~ ; I ~ ' . ' r ' ~ ,

EARLY INSP. EARLY EXP. (on/off)

Fig. 2. Singlebreath 133Xewashouts in subject N.R.A. Examples of each experimental maneuver are shown,
the titles referring to the portion of the maneuver to which oscillations were applied. Ordinates: 133Xe
concentration measured at the mouth. Abscissas: expired lung volume, which TLC to the left and RV to
the right. Arrows indicate the onset of Phase IV.
168 R.B. FILUK etaL

~+ 45
90

80 40

~ 7O
35
4 t~
o <
z 60
3o

~ so

ii
25
4O

30 20

C T E T L E C T E T L E

INSPIR. EXPIR. I N S P I R . EXPIR.

Fig. 3. Effect of HFOV on normalized phase IV height (left) and closing volume (CV) as a per cent of VC
(right). Abscissas indicate when HFOV was applied during inspiration or expiration: C is control, T is
throughout, L is late, and E is early. Brackets are SEM, * denotes significantly different from control
(P < 0.01) and + denotes significantly different from results with inspiratory oscillation (P < 0.015).

phase IV whenever it was applied, but the effect was less when oscillations occurred
during the first half of expiration than when they occurred during inspiration or the latter
half of expiration. Closing volume, on the other hand, increased when oscillations
occurred throughout expiration or during the latter half of expiration, and was un-
changed when H F O V was applied during inspiration or the first half of expiration.
H F O V had no consistent or statistically significant effects on the slope of phase III.
During the N 2 0 studies of regional distribution, boluses were administered at
approximately 14~o VC and this volume did not differ between control and H F O V
maneuvers and was similar to the lung volume at which boluses were administered
intravenously. There was no difference in the distributions of boluses injected intra-
venously with and without H F O V in the one subject studied. The regional distributions
of boluses are shown in fig. 4 and resulting closure calculations in table 1. The data of
fig. 4 are means of duplicate measurements. In our seven subjects the difference between
duplicate measurements in the two apical and two basal regions averaged 7.7~o
(SD = 1.9~o); variability in intermediate regions was less. In control studies, basal
[ X e ] i v exceeded [Xe]y2o (fig. 4), indicating that some closure was present in all but
subject I.H. (table 1). With HFOV, this trend increased in five of seven subjects
(table 1). In four subjects (D.B., M.H., L.C., Z.B.) xenon concentrations in the two
basal regions decreased with H F O V as compared to control, and in the fifth (D.J.)
apical concentrations increased, apparently because with H F O V the NaO delivered
bolus was diverted away from regions 20 cm below the apex. In general in comparison
to control, H F O V was associated with a shift in bolus distribution from basal regions
to apical regions. These shifts were interdependent in that if [Xe]N2O decreased in some
AIRWAY CLOSURE WITH HFOV 169

TABLE 1
Results of Part II studies

Subject Condition Bolus administration Fo Airway closure


(Yo v c )

I. D.B. Control I0 19
HFO 19 56

2. M.H. Control 17 1
HFO 16 50

3. D.J. Control 16 27
HFO 16 44

4. I.H. Control 17 - 1
HFO 21 1

5. Z.B. Control 16 26
HFO 13 43

6. S.H. Control 9 56
HFO 10 59

7. L.C. Control 7 35
HFO 8 51

X+SEM 13+2 23_+8


15 + 2 43 _+7*

* P < 0.015 using Student's paired t-test.

regions, they m u s t increase in others. This interdependence is taken into account in the
calculation o f airway closure and in the group as a whole, b a s a l closure increased
significantly (P < 0.015) from 23 + 8~o to 43 + 7~o with H F O V .

Discussion

Our system for applying H F O V to the lung was relatively crude. Our oscillator was set
to p r o d u c e volumes a p p r o x i m a t i n g 2 cc/kg, but the oscillatory volume actually applied
to our subjects was m u c h smaller, b e c a u s e their i m p e d a n c e was relatively high and
b e c a u s e the oscillator stroke volume was delivered b o t h to the subject a n d the bell
spirometer. The oscillatory volume applied was less than that o f the d e a d space, but was
enough to cause substantial changes in the distribution o f inspired a n d expired gas.
It might be argued that our N 2 0 results were suspect b e c a u s e we did not always
m e a s u r e perfusion distribution with injected boluses during H F O V . W e did show that
170 R.B. F I L U K et al.

~1 '~" i "
Inhalation
Perfuslon
~. - .........

! .............. HFO

20; /
25

30 D.B;

15

~ M.H.
30 '

I,o
i
25

30, R~, l /, s H;
'

10
\ D.J.

"\ =
15

2O

25

30
100 150 50 100 150
A L V E O L A R C O N C E N T R A T I O N (%)

Fig. 4. Results of studies of regional airway closure. Abscissae:~o regional 133Xe concentration. Ordinates:
lung height measured in cm from the apex. The solid lines indicate the distribution of 133Xe with intravenous
injection, the dashed line indicates regional N 2 0 uptake, and the dot-dashed line regional N 2 0 uptake with
HFOV.
AIRWAY CLOSUREWITH HFOV 171

HFOV did not change regional perfusion distribution in one subject and did not make
further measurements of this type because several other investigators have shown the
same thing (Schmid et al., 1981; Rehder and Didier, 1984; Mansel and Gillespie, 1986).
We, therefore, think it most unlikely that HFOV affected regional perfusion distribution
in a systematic way.
At low lung volumes, basal closure occurs in normal humans as initially surmised
from measurements of expired Xe concentrations (Dollfuss etal., 1967) and sub-
sequently demonstrated by studies using the N20 technique (Engel et al., 1975; Forkert
et al., 1979). The present experiments, using the same techniques, gave results compati-
ble with HFOV increasing this closure. In the initial experiments, the onset of phase IV,
thought to represent basal airway closure, occurred at higher lung volumes when HFOV
was applied late in expiration. In the subsequent series, boluses of Xe were delivered
to the lung periphery by N20 uptake. With HFOV, such boluses produced a more
uneven distribution of regional Xe concentrations; apical concentrations were higher
and basal concentrations lower than when boluses were delivered without HFOV. This
finding was also compatible with increased basal airway closure with HFOV. During
HFOV, unstable airways may have closed during the expiratory phase of the oscillation
and failed to open during the inspiratory phase. This explanation assumes that HFOV
caused oscillatory changes in the diameter of airways susceptible to closure. This seems
reasonable since oscillatory diameter changes have been demonstrated in the large
airways of dogs (Gavriely et al., 1985), which are probably less compliant than the
airways which close in normals.
The above interpretation assumes that the application of HFOV did not fundamen-
tally alter the factors governing the regional distribution of boluses delivered by N20
uptake or the factors determining regional emptying sequences. This may not have been
the case, though the fact that during our N20 experiments regional count rates stabilized
at both the apex and base (in about 7 and 15 sec respectively) more rapidly than in the
experiments of Forkert and Burks (1984) suggests that gas inflow due to the N20 uptake
present in our experiments was important in bolus delivery. It is possible that with
HFOV neither of the techniques we used were valid indicators of airway closure and
that HFOV produced the changes in gas distribution we observed without affecting
airway closure.
If, independent of airway closure, HFOV preferentially increased gas transport into
and out of apical lung regions we might have obtained similar results. Such increased
transport would have produced apical distributions of boluses during the N20 studies,
and earlier delivery of apical gas to central airways during expiration, favoring the early
onset of phase IV. Indeed, Rehder and Didier (1984) found that in supine humans
HFOV favored both washin and washout of apical regions, which could apply to erect
humans, though it should be noted that they did not find differences between non-
dependent and dependent basal regions at lung volumes at which closure was likely in
the latter. However, the hypothesis of increased apical gas transport by HFOV indepen-
dent of airway closure does not fit all our results. First, closing volume was the same
when HFOV was applied throughout expiration as it was when HFOV was applied only
172 R.B. FILUKetal.

during the latter half of expiration. There was no evidence of increased transport of
apical gas due to HFOV at high lung volumes, so that if it existed such transport must
have been peculiar to low lung volumes. Second, if HFOV increased apical gas transport
without affecting airway closure during the N20 studies, one might have expected
similar effects when boluses were inhaled at low lung volumes, i.e. HFOV should have
increased apex to base concentration differences of inhaled isotope. However, when
HFOV was applied during the inhalation of boluses from RV, the height of phase IV
was reduced (figs. 2 and 3). Since under these circumstances the height of phase IV has
been repeatedly shown to correlate with apex to base concentration differences (Engel
etal., 1975, 1976; Anthonisen etal., 1978), these differences were decreased with
HFOV during bolus inhalation at low lung volumes, the opposite of what would be
expected with increased apical transport of inhaled gas. Thus, though preferential apical
gas transport by HFOV that was independent of airway closure might explain our
results, not all our findings fit this hypothesis. The apparent discrepancy between the
evidence favoring reduced apex to base differences when HFOV was applied during
bolus inhalation and the results of the N20 studies is discussed below. HFOV has been
reported to cause two kinds of change in volume and pressure distribution in the lung
which might bear on our results: dynamic compression of airways (Solway et al. 1986),
and regional differences in alveolar pressure (Allen et al., 1987).
Dynamic airways compression has been reported as the cause of dynamic hyperin-
flation noted in intact animals with HFOV (Solway et al., 1986). It is possible that
dynamic compression of lower zone airways occurred during our studies; both closing
volume and the volume at which we conducted the N20 studies were relatively low,
where flow limiting segments develop in lobar or sublobar airways (Mink and Wood,
1980). Given the gravitational gradient in pleural pressure, the reduced recoil of basal
lung units would render them more susceptible to dynamic compression than those at
the apex. If, during the first series of experiments, dynamic compression of lower zone
airways had occurred as HFOV was applied at lower lung volumes the result would have
been a decrease in basal emptying and a reciprocal increase in the emptying of upper
lung regions: a phase IV would have been produced that was not due to airway closure
as previously postulated by Rodarte et al. (1975). However, dynamic compression of
basal airways probably could not have accounted for the more uneven distribution of
N20 delivered boluses with HFOV. The implication of such compression is that more
gas would enter the base than leave it with basal regions increasing in volume. This
would favor basal distribution of the inspirate and one would expect decreased apex
to base concentration differences with HFOV.
Recently, Allen et al. (1987) showed that in excised dog lungs, HFOV was associated
with an increase in the alveolar pressure of basal lung regions as compared to apical,
and the magnitude of the effect was roughly equal and opposite to that of the apex to
base pleural pressure gradient. The changes in alveolar pressure were stable in fractions
of a second after a change in HFOV frequency or tidal volume (Fredberg, personal
communication), and thus would have been present in our subjects before bolus delivery
if HFOV produced similar differences in alveolar pressure in intact humans. Such
AIRWAY CLOSURE WITH HFOV 173

changes in alveolar pressure would have caused inflation of basal regions, at the expense
of apical ones, favoring increased basal distribution of inspired gas, the opposite of what
we observed with HFOV. If such a mechanism applied with HFOV during expiration,
the resulting tendency toward equalization of regional volumes would mitigate against
the early appearance of phase IV that we observed.
Several other factors that might have influenced our findings should be mentioned.
It is possible that HFOV induced regional changes in airway tone, though we know of
no evidence favoring such an hypothesis. In the N20 experiments alveolar CO2 was
probably decreased and alveolar 02 increased by the application of HFOV, but
differences in CO 2 would almost certainly have been too small to influence airway tone,
and there is little evidence that differences in 0 2 do not have this effect. Preferential use
of rib cage or diaphragm has been shown to influence the regional filling and emptying,
but not at residual volume (Roussos et al., 1977), and the unlikely possibility that HFOV
affected the sequence of respiratory muscle contraction could not have affected our N20
results.
Several aspects of HFOV mitigated against finding increased uneveness of regional
filling and emptying. HFOV has been shown to induce interregional mixing of gas
(Schmid et al., 1981; Moffat et al., 1985) probably due to pendelluft. Also, during our
experiments, HFOV was applied during net convective movements of gas in and out
of the lung and, by mixing gas in major airways, probably made both regional and
expired gas concentrations more uniform. When oscillation was applied during bolus
administration it not only mixed gas in major airways but also tended to 'spread' the
indicator bolus over a greater volume of the gas entering the lung than was the case for
control maneuvers. This might have caused the bolus to enter the lung at a slightly higher
volume during oscillation than in control studies, which would decrease apex to base
differences. Oscillations applied during expiration would also have mixed gas in major
(lobar and sublobar) airways and rendered abrupt changes in expired ~33Xe concen-
tration less likely. All of these effects would have reduced our ability to detect uneveness
of gas distribution, and there is evidence that at least some of them applied. In the first
set of experiments, oscillations applied either to inspiratory or expiratory phases of the
VC resulted in a decrease in the height of phase IV (fig. 3) which as noted above reflects
apex to base concentration differences in the absence of expiratory oscillations (Engel
etal., 1975, 1976; Anthonisen etal., 1978). Oscillations applied early in inspiration
probably caused a more uniform bolus distribution by spreading the bolus and causing
mixing in large airways, and perhaps also by inter-regional pendelluft. Interestingly,
when oscillations were applied throughout inspiration, phase IV height was the same
as that when oscillations were applied during the first half of inspiration (fig. 3). This
implies that oscillations late in inspiration, after the bolus had been delivered to the lung
periphery, produced relatively little interregional mixing by pendelluft. When oscillations
were applied during expiration phase IV height was reduced. When oscillations were
applied only early in expiration, their effect upon the height of the subsequent phase IV
must have been due to interregional mixing by mechanisms such as pendelluft. When
oscillations were applied late in expiration, in addition to pendelluft, the oscillations
174 R.B. FILUK et al.

caused gas mixing in major airways as phase IV developed, so it is not surprising that
phase IV height was reduced more with late expiratory oscillations than with early. We
did not see consistent effects upon the slope of phase III which, in some subjects, was
relatively fiat (fig. 2) in these experiments. Moffat et al. (1986) have shown that HFOV,
when applied at TLC, reduces phase III slopes but there were so many differences
between their experiments and ours that they cannot be regarded as contradictory.
Factors promoting more even distribution of boluses, when oscillation was applied early
in inspiration may well have been operative during the N20 studies employing oscilla-
tion; the duration of the oscillations was roughly similar. However, the N20 experiments
were carried out during breath hold so that mixing of the bolus in major airway with
resultant 'spread' of the indicator gas did not imply that some entered peripheral lung
regions at relatively high lung volumes as was the case when HFOV was applied during
bolus inhalation. Inter-regional mixing by pendelluft would have applied in both cases,
and in view of this our finding of more uneven bolus distribution with HFOV in the N20
studies is impressive.
While in 5/7 subjects of the N20 experiments HFOV increased the uneveness of the
distribution of boluses administered with N20, and therefore airway closure as we
calculated it, this was not the case in two (Table 1). In one of these subjects (I.H.) there
was no closure demonstrated during the control maneuver, and basal airways may have
been relatively stable. The failure of HFOV to increase apex to base concentration
differences in the absence of pre-existing closure did not favor the effect of HFOV being
independent of clo sure in the other subjects. The other subject (S. H.) demonstrated the
largest amount of basal closure observed during control measurements. It is possible
that when closure is extensive, it is not increased by HFOV, but previous data (Forkert
et al., 1979) suggest that maximal closure is present at RV, so S.H. should have had
unstable airways at the lung volume studied. The failure of HFOV to influence the
distribution of boluses administered with N20 in this subject does not shed light on the
mechanism of the effect observed in the others.

Acknowledgements. This work was supported by the MRC of Canada. R.B.F. is a Fellow of the Canadian
Lung Association. P.A.E. is a fellow of the MRC.

References

Allen, J.L., I.D. Frantz III and J.J. Fredberg (1987). Heterogeneity of mean alveolar pressure during
high-frequency oscillation. J. Appl. Physiol. 62: 223-228.
Anthonisen, N.R., L.J. Peress, D. I.M. Siegler and S. Dhingra (1978). Lung volume, volume history, and
the distribution of inhaled boluses. Respir. Physiol. 33: 279-288.
Bryan, A. C. and A. S. Slutsky (1986). Lung volume during high frequency oscillation. Am. Rev. Respir. Dis.
133: 928-930.
Chang, H. K. (1984). Mechanisms of gas transport during ventilation by high-frequency oscillation. J. Appl.
Physiol. 56: 553-563.
Dollfuss, R. E., J. Milic-Emili and D.V. Bates (1967). Regional ventilation of the lung studied with boluses
of 133Xenon. Respir. Physiol. 2: 234-246.
AIRWAY CLOSURE WITH HFOV 175

Drazen, J. M., R. D. Kamm and A. S. Slutsky (1984). High-frequency ventilation. Physiol. Rev. 64: 505-543.
Engel, L.A., A. Grassino and N.R. Anthonisen (1975). Demonstration of airway closure in man. J. Appl.
Physiol. 38: 1117-1125.
Engel, L.A., L. London, L. Taussig, R.R. Martin and G. Sybrecht (1976). Influence of bronchomotor tone
on regional ventilation distribution at residual volume. J. Appl. Physiol. 40:411-416.
Forkert, L., S. Dhingra and N.R. Anthonisen (1979). Airway closure and closing volume. J. Appl. Physiol.
46: 24-30.
Forkert, L. and J. E. Burks (1984). The effect of lung volume on regional gas transport during high frequency
oscillations. Respir. Physiol. 58: 279-287.
Gavriely, N., J. Solway, S. Loring, R.H. Ingrain, Jr., R. Brown, A. Slutsky and J. Drazen (1985). Radiogra-
phic visualization of airway wall movement during oscillatory flow in dogs. J. Appl. Physiol. 58:
645-652.
Mansel, J. K. and D.J. Gillespie (1986). Oxygenation during ventilation by high-frequency oscillation in dogs
with acute lung injury. Crit. Care Med. 14: 955-959.
Mink, S. and L. D. H. Wood (1980). How does HeO 2 increase maximum expiratory flow in human lungs?
J. Clin. Invest. 66: 720-729.
Moffat, S.L., L. Byford and L. Forkert (1985). Interregional gas mixing and gas transport during high
frequency oscillations. Respir. Physiol. 62: 1-13.
Moffat, S.L., L. Byford and L. Forkert (1986). Intraregional gas mixing in humans during high frequency
oscillations. Respir. Physiol. 66: 95-107.
Rehder, K. and E.P. Didier (1984). Gas transport and pulmonary perfusion during high-frequency
ventilation in humans. J. Appl. Physiol. 57: 1231-1237.
Rodarte, J. R., R. E. Hyatt and D.A. Cortese (1975). Influence of expiratory flow on closing capacity at low
expiratory flow rates. J. Appl. Physiol. 39: 60-65.
Roussos, D.S., M. Fixley, J. Gnest, M. Cosio, S. Kelly, R.R. Martin and L.A. Engel (1977). Voluntary
factors influencing the distribution of inspired gas. Am. Rev. Respir. Dis. 116: 457-467.
Schmid, E. R., T.J. Knopp and K. Rehder (1981). Intrapulmonary gas transport and perfusion during high
frequency oscillation. J. Appl. PhysioL 51: 1507-1514.
Simon, B. A., G. G. Weinmann and W. Mitzner (1984). Mean airway pressure and alveolar pressure during
high-frequency ventilation. J. Appl. Physiol. 57: 1069-1078.
Solway, J., T.H. Rossing, A.F. S aari and J.M. Drazen (1986). Expiratory flow limitation and dynamic
pulmonary hyperinflation during high-frequency ventilation. J. Appl. Physiol. 60: 2071-2078.

Вам также может понравиться