- Email: [email protected]
Bronchial length and diameter behavior during bronchial collapse in excised dog lungs
Respiration Physiology (1981) 43, 107-116 © Elsevier/North-Holland Biomedical Press
BRONCHIAL LENGTH AND DIAMETER BEHAVIOR DURING BRONCHIAL COLLAPSE IN EXCISED DOG LUNGS
MASAO N A K A M U R A , HIDETADA SASAKI, TAMOTSU TAKISHIMA and JACK HILDEBRANDT The First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan and Institute of Respiratory Physiology, Virginia Mason Research Center, Seattle, WA, U.S.A.
Abstract. We compared changes in bronchial diameters (Dbr) and lengths (Lbr) in six excised dog lobes at 4 volume levels during deflation from 30 cm H20 under two conditions. First, intrabronchial and alveolar pressures were kept equal to zero during lung deflation; then, bronchial collapse was induced by negative intrabronchial pressure at several constant lung volumes. During the two procedures the major intrapulmonary bronchi were isolated from the rest of the lung by 10-15 beads blocking all tributary bronchi as previously described. Lbr was estimated with a linear displacement transducer connected by a thin rod to the most distal bead. Mean Dbr was calculated from Lbr and bronchial volume at each lung recoil pressure (PL) and at each intrabronchial pressure (Pbr). When PL decreased from full inflation in Condition 1, Dbr and Lbr decreased almost proportionately; however, in Condition 2 when PL (and thus lung volume) was kept constant and Pbr was reduced from 0 to -80 cm H20, Lbr was reduced by only 5 ~ (mean) whereas Dbr was reduced by 70% (mean) from initial near-homogeneous values. This suggested that, due to the interdependence in intact lungs, Lbr in intact lungs is almost constrained by changes in lung volume. By contrast, Dbr can undergo large reductions and the stiffness of intact bronchi is related mainly to PL. Thus, the intact bronchus has some of the characteristics of a thick-walled cylinder (where the wall includes surrounding parenchyma) in which longitudinal strain is small by comparison with circumferential strain. Airway mechanics Bronchial dimensions
Lung recoil pressure Pulmonary interdependence
It is recognized that airway collapsibility is an important factor in determining maximum flow (Dawson and Elliott, 1977; Pride et al., 1967; Takishima and Sasaki, 1972). Mead, Takishima and Leith (1970) predicted and Sasaki et al. (1977, 1978) and Takishima, Sasaki and Sasaki (1975) later confirmed that intrapulmonary bronchi were much less collapsible in situ than when excised, due to the radial traction effected by the peribronchial tissue; moreover, this characteristic of intact Accepted for publication 25 October 1980 107
108
M. N A K A M U R A e: al.
bronchi tended to be much more pronounced when lung recoil pressure was increased. Nakamura, Sasaki and Takishima (1979) examined this dependence of airway collapsibility on lung recoil pressure, rather than on lung volume, in greater detail by comparing the bronchial pressure-volume relationships in air-filled and saline-filled lungs. One might ask in addition whether during bronchial collapse, bronchial length shortened while bronchial diameter narrowed, and further to what extent bronchial length in intact lung is more rigid than when excised but at the same bronchial diameter. Hyatt, Rodarte and Wilson (1975) reported that there was a difference in the bronchial diameter and length behavior in situ before and after cooling and ventilation to stiffen the lobe. They showed that as the lung was inflated bronchial diameter corresponded best to lung recoil pressure, and length to lung volume. Their purpose was to attempt to make a separation between lung volume and elastic recoil as factors determining bronchial diameter and length. However, the differences between the bronchial length and diameter behavior during regional deformation within the lobe have not been studied. Our primary purpose in the present study was therefore to quantitate changes in bronchial length and diameter during regional bronchial collapse within the lobe at several constant lung volumes. For this purpose we have employed the blocked bronchus preparation previously described (Takishima, Sasaki and Sasaki, 1975). We then compared these data with dimensional data of both excised bronchi and of intact bronchi during quasi-static lung deflation.
Methods
Six mongrel dogs weighing 13-15 kg were anesthetized with sodium pentobarbital (40 mg/kg) and sacrificed by exsanguination after an intravascular injection of heparin (300 IU/kg). The right lower lobe was suspended in a Lucite box by connecting the lobar bronchus to a short metal cannula (length = 8 mm; I.D. = 11 mm) attached to the center of the box. The cannula was inserted 1-2 mm into the intrapulmonary bronchial segment and secured with 3 6 ligatures. The pulmonary vessels were not tied. Lung volume was changed by means of collateral channels using three hollow hemispherical Lucite capsules (diameter = 10 mm) glued to the pleural surface of the expanded lobe (Sasaki, Sasaki and Takishima, 1977; Takishima et al., 1971) as shown in fig. 1. The pleural surface beneath each capsule was punctured by about 30 holes, each one approximately 1 mm in diameter and 2 mm deep. Each capsule was connected to a vinyl catheter (length = 30 cm; I.D. = 5 mm). Two of them functioned to inflate the lobe. The third was connected to two transducers (Hewlett-Packard, 267 BC) for the measurement of alveolar pressure (PA) and lung recoil pressure (PL = PA - Pbox). The lobe was then inflated by a negative box pressure (Pbox) of 30 cm H~O and the lobar bronchus was made air-tight by sealing the orifices of its tributary branches with 10-15 beads (diameters 2 9 mm), attached by tissue glue (Alonalpha, Sankyo,
BRONCHIAL LENGTH AND DIAMETER BEHAVIOR
PA-%o
109
I11
I tL
Box
/ 8..a_.,
....
~ Syringe
I
~1 Linear transducer
I ~
Vbr
Fig. l. Central bronchi were isolated by glueing obstructing beads in all tributary bronchi. The lung was inflated by negative box pressure (Pbo×). Alveolar pressure (PA) was kept atmospheric through collateral routes via 2 pleural capsules glued on the pleural surface. The third capsule was used for monitoring PA. Intrabronchial pressure (Pbr) was obtained from a pressure transducer, and bronchial volume changes (Vbr) from a syringe coupled to a linear displacement transducer. The blocked lobar main bronchus, the pressure transducer and the syringe were filled with saline. Lung recoil pressure (PL) was obtained from PA-Pbox by a differential transducer. The bronchial length (Lbr) was measured with a linear displacement transducer that was connected by a thin rod to the most distal bead.
Japan) as previously described (Takishima, Sasaki and Sasaki, 1975). Thereby we could obtain the bronchial pressure volume relationships independently of the overall lung volume which was held at various levels by means of negative box pressure. Alveolar pressure was kept uniformly atmospheric through collateral pathways. The bronchial length (Lbr) was measured with a linear displacement transducer (Linear, LDMR 73S, Japan) that was connected by a thin rod to the most distal bead. The transducer was water-proofed and laid horizontally inside the plastic adapter connected to the airway opening along the axis of the main lobar bronchi. The position of the iron core (3 g in weight) was changed smoothly inside the transducer corresponding to Lbr (the displacement of the terminal bronchial bead). The bronchial volume (Vbr) was measured with a 30-ml syringe coupled to another linear displacement transducer (Hewlett-Packard, 7DCDT-1000); the intrabronchial pressure (Pbr) was measured with a pressure transducer (HewlettPackard, 267BC). Both the bronchi and the apparatus were filled with saline. The box was partially filled with a saline solution, and laid on its side, so that the lobe floated on the fluid and the lobar bronchus was held horizontally and in a straight line. The gravitational effect was thus reduced. The bronchial pressure was measured taking the central plane of the bronchi as the zero reference level. A change in the intrabronchial pressure was obtained by withdrawing a measured volume of the saline solution with the syringe. Zero Vbr was established statically
110
M. NAKAMURA ~'t a/.
in lobes at P L = 0 cm H~O by lowering the Pbr to - 1 0 0 cm H,O. First, we measured PL vs. Lbr and PL vs. Vbr curves of the intact bronchi at 5 lung volumes (PL = 30, 10, 5, 2 and 0 cm H,O) keeping both Pbr and PA at 0 cm H~O. Then, by withdrawing saline from the syringe we obtained Pbr-Lbr and Pbr-Vbr curves of the same lobes at the constant lung volumes of PL = 30, 10, 5 and 2 cm H,O. Afterwards, Vbr was returned gradually to its initial value and the lobe was inflated once again to PL = 30 cm H,O. After completing this series of measurements, the bronchi were carefully dissected from the surrounding tissues by removing the parenchyma, branches and large vessels. The zero Vbr was established again by lowering the Pbr to - 1 0 0 cm H,O. The transmural pressure-bronchial volume (Ptm-Vbr) and the Ptm-Lbr curves were then obtained in the same manner as for the intact bronchi, by applying negative Pbox of 30, 10, 5, 2 and 0 cm H,O. The alveolar pressure, bronchial volume and lung recoil pressure were recorded on a pen-writing oscillograph (Sand, 8S). The Pbox was measured with a water manometer. The Pbr-Lbr and Pbr-Vbr relationships were recorded on two X - Y recorders (Hewlett-Packard, 7045A) during the gradual deflation of Vbr (0.03 ml/s from Pbr 0 to - 8 0 cm H 2 0 ). It took about 2 3 min to record each Pbr-Vbr curve. In separate preliminary measurements of the effect of blocking bronchi on lung PV curve at low lung volume, the lobe was inflated to alveolar pressure of +30 cm H , O via two capsules, then deflated to PL = 2 cm H,O. Lobe volume was measured by water-displacement, subtracting lobe weight and assuming the density of lung tissue to be 1.0. The difference of lobe volumes before and after blocking bronchi was within 21'~ TLC. When the lobe was deflated to PL = 0 cm H eO, the residual volumes after blocking increased 2.7_+ 1.37Jo TLC, indicating that this procedure did not significantly increase gas trapping at PL = 2 cm H20, or even at RV. All experiments were carried out at 20+_2°C room temperature. The measurement of curves began between 30-60 min after exsanguination and continued for about 3 h.
Results The mean volumes and lengths of the blocked bronchial segments of the 6 dog lobes were 3.24+_0.59 ml (SD) and 3.97+_0.60 cm (SD), respectively, at full inflation (Pbr = 0 and PL = 30 cm H:O). Length was measured as the distance from the midpoint of the terminal bead to the inlet of the cannula. Figure 2 shows the in s i t u changes of Lbr v e r s u s PL and v e r s u s Pbr. Lbr at full inflation in each of the 6 dog lobes was taken as 100~,,,. When the lobe was deflated while holding Pbr and alveolar pressure at zero, bronchial lengths decreased by 24.3+_ 1.7% (mean+ SD) as the PL decreased from 30 to 2 cm H20. However, when PL was held constant at 30 cm H,O, Lbr decreased by 1.5+- 1.3~0 as Pbr was lowered from 0 to - 8 0 cm H20. At the remaining PL (10, 5 and 2 cm H20), Lbr shortened somewhat more (an average of 6.8+-0.77{i). Thus, in intact lobes Lbr
B R O N C H I A L L E N G T H A N D D I A M E T E R BEHAVIOR
111
Intact bronchi PL 30~
~
•
- •
je
-e- 100 I(*/.max)
5~ "--'--------~
80
2 ~
70
/e
J
J
/
60 50 -80
-60
-&0
Pbr
-20
02 5
( cmH20 )
10
20
30
PL
Fig. 2. The dependence of Lbr on PL of the blocked bronchi at Pbr = 0 cm H 2 0 is shown in the right panel. When Pbr was made negative at various constant lung volumes for PL = 30, 10, 5 and 2 cm H20, the curves in the left panel were obtained. Lbr is expressed as o / o f the bronchial length at full inflation (PL= 30 and Pbr = 0 cm H 2 0 ) on all curves. Bars represent the mean values +1 SD for the 6 dog lobes.
exhibited only a slight tendency to decrease at constant lung volumes, even when Vbr was reduced by up to 70~o. Figure 3 shows corresponding changes in mean Dbr v e r s u s PL and v e r s u s Pbr. The Dbr were calculated from Vbr and Lbr at each value of PL and Pbr, assuming the geometry of a uniform cylinder: Vbr = n(Dbr/2)2(Lbr). The Dbr at full inflation in each of the 6 dog lungs was taken as 100~o. During lobe deflation, Dbr decreased spontaneously by 23.0+ 3.0~o as PL decreased from 30 to 2 cm H 2 0 at Pbr = 0 cm H20. During forced bronchial collapse to Pbr = - 8 0 cm H20, Dbr decreased by 40.2 + 3.5~o at PL = 30 cm H 2O, and by an average of 78.3 + 3.6~o at the remaining PL (10, 5 and 2 cm H:O). Thus, at fixed lobe volume the changes of Dbr during forced collapse were more than 10 times larger than those of Lbr. In fig. 4 changes in Lbr are plotted v e r s u s simultaneous changes in Dbr for the intact and dissected bronchi at the 5 different PL (or Ptm) for cases where Pbr was maintained at zero. In this plot resting dimensions are taken as 100~o. Lbr of intact bronchi increased to 143.0+ 15.1 ~o, over the vital capacity range from PL (or Ptm) 0 to 30 cm H20. At the same time Dbr increased to 142.8+ I0.1~o in intact and to 149.6+ 17.3~ in dissected bronchi. Resting dimensions themselves were slightly altered when the bronchial segment was dissected from the surrounding parenchyma: Lbr decreased by 1 0 . 3 + 9 . 8 ~ and Dbr increased by 9.1+6.3~o relative to the intact dimensions at PL = 0 cm H20. In the intact bronchi not undergoing forced collapse, the changes of Lbr and Dbr lay fairly close to the line of identity over the vital capacity range, whereas isolated bronchi first shorten markedly with only minor narrowing as Ptm falls from 30 to 5 cm H20, followed by major narrowing at lower Ptm. Thus, intact bronchi contract more isotropically
112
M. N A K A M U R A
ul al.
) Intact b r o n ~
q } / / "
"/~}
U "
~'
f -
4(
,,' ,,"
10
/ 5
~
'
"
Dissected bronchi," Length fixed .''.'"
,
, -GO
. . . .
, ~ " ,
,
-z,O
2(
Length not fixed
.-''.'-''" -80
"'
---
-20
Pbr
0
0 0 2 5
( cmH20 )
10
20
30
PL
Fig.
3. The d e p e n d e n c e of D b r on PL of the b l o c k e d b r o n c h i at Pbr = 0 cm H~O is s h o w n at the r i g h t : b r o n c h i a l collapse d a t a at the c o n s t a n t l u n g v o l u m e s for PL = 30, 10, 5 a n d 2 cm H , O are illustrated on the left. D b r was c a l c u l a t e d from Vbr a n d Lbr. D b r was expressed as I~;>o f d i a m e t e r at full inflation (PL = 30 a n d Pbr = 0 cm H 2 0 ) . Bars show the m e a n values +1 S D for the 6 dog lobes.
160] ~'~
Ptm
i
"6
" '°bJ~
140]
__
_ [//~"
.....
J ~.
,
30¢mH20
L/
,2o]
"ti
1001100
|
[ ,201 [ Lbr
,,10
1;0
(% of resting)
80-
Fig. 4. The % D b r % Lbr r e l a t i o n s h i p s o f intact b r o n c h i ( H ) a n d dissected b r o n c h i (x x ) at PL (or Ptm) = 30, 10, 5, 2 a n d 0 cm H~O. D b r a n d Lbr are expressed as i',i o f resting length at Pk (or Ptm) = 0 cm H~O. Bars represent the m e a n values _+1 SD for the 6 d o g lobes. The d o t t e d line is the line o f identity (isotropic b r o n c h i a l expansion). The d a s h e d line, ~k ~--~D2, is a theoretical relation for an inflated closed-end i s o t r o p i c elastic tube. I n t e r r u p t e d steep lines show data found d u r i n g b r o n c h i a l collapse.
B R O N C H I A L L E N G T H A N D D I A M E T E R BEHAVIOR
113
than dissected bronchi, reflecting the effects of the surrounding lung parenchyma tending to induce homogeneity and isotropicity on embedded structures. Also on fig. 4 are shown the data from forced bronchial collapse at each of the 4 different fixed PL. Under these conditions interdependence factors are manifested in quite different ways: Lbr tends to remain fixed while Dbr undergoes extensive narrowing. These features are further examined below.
Discussion
In the present study we have demonstrated that intact bronchi contract nearly isotropically during lung deflation, quite unlike the behavior of dissected bronchi on the one hand, or unlike forced bronchial collapse at fixed lung volume on the other. Our findings may be explained by the existence of a quasi-uniform distribution of stress in the intact lung, in contrast to uneven distributions of stresses in the radial and longitudinal directions both in dissected bronchi and around intraparenchymal bronchi undergoing distortion. The unevenness probably arises from the nonsymmetric bronchial structure (cylindrical rather than spherical). When non-spherical cavities in elastic media are inflated or deflated, complex deformations are to be expected, as partially analyzed below. These anisotropic strains do not necessarily imply anisotropic material properties, but these could certainly further contribute to the distortion. Based on bronchograms, the percent changes of Lbr and Dbr versus PL or VL during lung inflation or deflation have been reported by many investigators (Hughes, Hoppin and Mead, 1972; Hughes, Jones and Wilson, 1975; Hughes et al., 1974; Prakash and Hyatt, 1978; Sittipong and Hyatt, 1974). Hughes et al. reported that in intact bronchi of dog lungs Lbr decreased 38~o (1972) and 30~o (1974) during deflation of the lobes from PL = 30 to 0 cm H:O. By comparison, our results showed that Lbr decreased 30.7+2.1% over the same range of PL. In our study we employed a linear displacement transducer for measuring changes in Lbr directly. In preliminary experiments in 2 lobes we compared changes of Lbr obtained from orthogonal bronchograms at the same 5 different PL as above, using one terminal bronchial bead as a radiopaque marker, with the data from the displacement transducer connected to the same wedged bead. No systematic differences were found. Hughes et al. also reported that in intact bronchi of dog lungs Dbr decreased 42~o (1972) and 36~o (1975) during deflation from PL = 30 to 0 c m H 2 0 . Our results showed that the bronchial diameter decreased 29.8+ 5.0% over the same range of PL. This figure would be expected to vary with the resting volume of the lungs, which varies depending on the preparation, volume history, etc. Several investigators (Hughes, Hoppin and Mead, 1972; Hughes, Jones and Wilson, 1975; Hyatt et al., 1970; Prakash and Hyatt, 1978; Sittipong and Hyatt, 1974) have shown that the changes of bronchial length during deflation were essentially proportional to
114
M. N A K A M U R A
et ~tl.
changes in the cube root of absolute lung volume, but that changes of bronchial diameter were different depending on the bronchial sizes. In our study we examined mainly the behavior of larger lobar bronchi and for these the changes of Dbr in situ were similar to changes of Lbr (fig. 4). When one considers the interdependence between airspaces and bronchi (Mead, Takishima and Leith, 1970), if the specific compliance of bronchi were not too different from lung specific compliance and nearly isotropic, one would expect both Dbr and kbr to vary a s VL 1 3 However, we also showed that during bronchial collapse within the lung at constant VL, Lbr and Dbr were not at all similar. During bronchial collapse Dbr within the lung parenchyma depends on bronchial transmural pressure, but Lbr depends primarily on lung volume and only slightly on transmural pressure. The grossly unequal strains may be understood by the following approximation. Consider a closed cylindrical bronchial segment of radius R and length L surrounded by parenchyma such that alveolar and bronchial pressure are zero. The force shortening the segment, when a negative Pbr is applied, is (~R:) (Pbr) on each end. This produces an average shear stress over the walls of the bronchus (force/area) of 2(rcR:)(Pbr)/2r~RL = R(Pbr)/L. On the other hand, the radial stress is Pbr. Thus, if R/L is about 1/8, as it was in our preparation, one might expect roughly 8 times as much circumferential as longitudinal stress, and proportionally different changes in Dbr and Lbr. However, during bronchial collapse R decreases by 501'~i or more, so that the ratio R/L could fall to 1/16 or less. Our measured ratio L/R (mean 16.4+ 7.6) confirms this general trend. One would further predict that if the length of the segment were reduced, e.g. to L = R, then changes in Lbr and Dbr would become nearly equal. Conversely, if the segment were made long and narrow (R/L small), almost no shortening would occur. The strains in the inflated excised bronchus may as a first approximation be treated from the analysis of inflated cylinders made of incompressible isotropic elastic material. It is well-known that longitudinal strain (~L) is zero in such a cylinder when it is undergoing small pressure expansions. However, in moderate expansion EL is not zero and may be roughly related to the tangential strain (eY) by e L - eT: (Hildebrandt, 1970). Since tangential (or circumferential) strain is just proportional to percent change in diameter, ~D, this relation can be written as eL ~ eD 2. This equation is plotted in fig. 4 (dashed line) for diameter changes up to about 5°°/~ Jo and also indicates a reasonable agreement. The general characteristics of the segment are described by the curve, although the existence of some anisotropy is suggested by the fact that length changes always exceed those predicted. Thus, the circumferential rigidity appears to be greater than the longitudinal rigidity, and this is as would be expected from length tension studies. Using dog lobes before and after their static lung recoil had been significantly increased by cooling and ventilating, Hyatt, Rodarte and Wilson (1975) showed that the diameter corresponded best to lung recoil~ and length to lung volume. Hoppin, Lee and Dawson (1975), using a grappling hook technique, demonstrated that bronchial lengths were greatly stabilized h7 situ because of their much greater
B R O N C H I A L LENGTH A N D D I A M E T E R BEHAVIOR
115
longitudinal stiffness compared to dissected bronchi. These results also showed that the mechanical interdependence for longitudinal traction was different from that for radial traction applied by surrounding lung parenchyma. We would account for these differences largely on geometrical grounds, as indicated above. Pride et al. (1967) suggested that the maximum expiratory flow was determined by the collapsibility of the tracheobronchial tree, by lung elastic recoil, and by airway resistance. Takishima and Sasaki (1972) also suggested theoretically that maximum expiratory flow might be determined by both airway resistance and airway compliance. Using the tube wave speed concept in the analysis of flow limitation in elastic tubes, Dawson and Elliott (1977) also showed that flow rate occurring at wave speed was a function of airway dimensions and specific compliance of the tube. Therefore, airway collapsibility is an important factor in determining maximum expiratory flow. Takishima, Sasaki and Sasaki (1975) reported that the differences of the intrapulmonary bronchial collapsibility between intact and dissected bronchi were large and those differences increased with higher PL. We considered three reasons why airway collapsibility might decrease with higher PL: first and most important, radial traction applied by surrounding parenchyma increases with PL (Mead, Takishima and Leith, 1972; Nakamura, Sasaki and Takishima, 1979); second, the bronchial wall is slightly stiffened when bronchial length is extended by large lung inflation (Me!issinos and Mead, 1977); third, at constant lung volume the changes of Lbr are small during bronchial collapse. The significance of the first factor is apparent from such data as shown in fig. 3 where parenchymal distending stresses of 20-40 cm HzO are typical. To evaluate the second factor, we compared the pressure-volume relations of dissected bronchi in 4 dogs before and after fixing the initial bronchial length at that found in intact lung at PL = 30 cm H20. With fixed bronchial length, the pressure-volume curves of dissected bronchi indicated slightly less collapsibility (fig. 3) by 2.1 + 0.4 cm H20 at 50~ of maximum Dbr and 3.0+ 0.7 cm H20 at 30~o maximum Dbr. Melissinos and Mead (1977) reported that maximum expiratory flow increased in some subjects during neck hyperextension and that possibly increased tracheal longitudinal tension decreased the airway collapsibility. The third factor is related to the second, insofar as shortening of Lbr during bronchial collapse was very small, further tending to stabilize diameter.
References Dawson, S.V. and E.A. Elliott (1977). Wave-speed limitation on expiratory flow - a unifying concept. J. Appl. Physiol. 43: 498-515. Hildebrandt, J. (1970). Extension of small-strain theory to finite deformation of cylindrical vessels by internal over-pressure. Angiologica 7: 25~272. Hoppin, F . G . , Jr., G.C. Lee and S.V. Dawson (1975). Properties of lung parenchyma in distortion. J. Appl. Physiol. 39: 742-751.
1t6
M. N A K A M U R A el u/.
Hughes, J. M.B., F.G. Hoppm, Jr., and J. Mead (1972). Effect of lung inflation on bronchial length and diameter in excised lungs. J. Appl. Physiol. 32:25 35. Hughes, J. M. B., H.A. Jones, A.G. Wilson, B.J.B. Grant and N.B. Pride (1974). Stability of inlrapulmonary bronchial dimensions during expiratory flow in excised lungs. J. Appl. Plo'siol. 37 : 684-694. Hughes, J. M. B., H. A. Jones and A. G. Wilson (1975}. Bronchial hysteresis in excised lungs. J. Physiol. (Lomlon) 249:435 443. Hyatt, R.E., R. Sittipong, S. Olafson and W.A. Potter (1970). Some factors determining pulmonary pressure-flow behavior at high rates of air flow. In: Airway dynamics, edited by A. Bouhuys. Springfield, IL., Thomas, pp. 43 60. Hyatt, R. E., J. R. Rodarte and T.A. Wilson (1975). Effect of increased static lung recoil on bronchial dimensions of excised lungs. J. AppI. Physiol. 39 : 429 433. Mead, J., T. Takishima and D. Leith (1970). Stress distribution in lungs : a model of pulmonary elasticity. J. Appl. Ph3siol. 28:596 608. Melissinos, C. G. and J. Mead (1977). Maximum expiratory flow changes induced by longitudinal tension on trachea in normal subjects. J. Appl. Physiol. 43:537 544. Nakamura, M., H. Sasaki and T. Takisbima (1979). Effect of lung surface tension of bronchial collapsibility in excised dog lungs. J. Appl. Physiol. 47:692 700. Prakash, U.B.S. and R.E. Hyatt (1978). Static mechanical properties of bronchi in normal excised human lungs. J. Appl. Physiol. 45:45 50. Pride, N.B., S. Permutt, R.L. Riley and B. Bromberger-Barnea (1967). Determinants of maximal expiratory flow from the lungs. J. Appl. Physiol. 23:646 662. Sasaki, H., T. Sasaki and T. Takishima (1977). Influence of lung parenchyma on dynamic bronchial collapsibility of excised dog lungs. J. Appl. Physiol. 42:699 705. Sasaki, H., F.G. Hoppin, Jr., and T. Takishima (1978). Peribronchial pressure in excised dog lungs. J. Appl. Physiol. 45:858 869. Sittipong, R. and R.E. Hyatt (1974). Static mechanical behavior of bronchi in excised dog lung. J. Appl. Physiol. 37:201 206. Takishima, T., K. lshikawa, T. Sasaki, H. Sasaki and T. Nakamura (1971). Measurement of collateral flow at quasialveolar level in excised dog lung. Tohuku J. Exp. Med. 105:405 406. Takishima, T. and H. Sasaki (1972). Two-dimensional flow-model for analysis of expiratory check valvc. Bull. Ph.vsio-Pathol. Resp. 8:361 374. Takishima, T., H. Sasaki and T. Sasaki (1975). Influence of lung parenchyma on collapsibility of dog bronchi. J. Appl. Physiol. 38:875 881.
BRONCHIAL LENGTH AND DIAMETER BEHAVIOR DURING BRONCHIAL COLLAPSE IN EXCISED DOG LUNGS
MASAO N A K A M U R A , HIDETADA SASAKI, TAMOTSU TAKISHIMA and JACK HILDEBRANDT The First Department of Internal Medicine, Tohoku University School of Medicine, Sendai, Japan and Institute of Respiratory Physiology, Virginia Mason Research Center, Seattle, WA, U.S.A.
Abstract. We compared changes in bronchial diameters (Dbr) and lengths (Lbr) in six excised dog lobes at 4 volume levels during deflation from 30 cm H20 under two conditions. First, intrabronchial and alveolar pressures were kept equal to zero during lung deflation; then, bronchial collapse was induced by negative intrabronchial pressure at several constant lung volumes. During the two procedures the major intrapulmonary bronchi were isolated from the rest of the lung by 10-15 beads blocking all tributary bronchi as previously described. Lbr was estimated with a linear displacement transducer connected by a thin rod to the most distal bead. Mean Dbr was calculated from Lbr and bronchial volume at each lung recoil pressure (PL) and at each intrabronchial pressure (Pbr). When PL decreased from full inflation in Condition 1, Dbr and Lbr decreased almost proportionately; however, in Condition 2 when PL (and thus lung volume) was kept constant and Pbr was reduced from 0 to -80 cm H20, Lbr was reduced by only 5 ~ (mean) whereas Dbr was reduced by 70% (mean) from initial near-homogeneous values. This suggested that, due to the interdependence in intact lungs, Lbr in intact lungs is almost constrained by changes in lung volume. By contrast, Dbr can undergo large reductions and the stiffness of intact bronchi is related mainly to PL. Thus, the intact bronchus has some of the characteristics of a thick-walled cylinder (where the wall includes surrounding parenchyma) in which longitudinal strain is small by comparison with circumferential strain. Airway mechanics Bronchial dimensions
Lung recoil pressure Pulmonary interdependence
It is recognized that airway collapsibility is an important factor in determining maximum flow (Dawson and Elliott, 1977; Pride et al., 1967; Takishima and Sasaki, 1972). Mead, Takishima and Leith (1970) predicted and Sasaki et al. (1977, 1978) and Takishima, Sasaki and Sasaki (1975) later confirmed that intrapulmonary bronchi were much less collapsible in situ than when excised, due to the radial traction effected by the peribronchial tissue; moreover, this characteristic of intact Accepted for publication 25 October 1980 107
108
M. N A K A M U R A e: al.
bronchi tended to be much more pronounced when lung recoil pressure was increased. Nakamura, Sasaki and Takishima (1979) examined this dependence of airway collapsibility on lung recoil pressure, rather than on lung volume, in greater detail by comparing the bronchial pressure-volume relationships in air-filled and saline-filled lungs. One might ask in addition whether during bronchial collapse, bronchial length shortened while bronchial diameter narrowed, and further to what extent bronchial length in intact lung is more rigid than when excised but at the same bronchial diameter. Hyatt, Rodarte and Wilson (1975) reported that there was a difference in the bronchial diameter and length behavior in situ before and after cooling and ventilation to stiffen the lobe. They showed that as the lung was inflated bronchial diameter corresponded best to lung recoil pressure, and length to lung volume. Their purpose was to attempt to make a separation between lung volume and elastic recoil as factors determining bronchial diameter and length. However, the differences between the bronchial length and diameter behavior during regional deformation within the lobe have not been studied. Our primary purpose in the present study was therefore to quantitate changes in bronchial length and diameter during regional bronchial collapse within the lobe at several constant lung volumes. For this purpose we have employed the blocked bronchus preparation previously described (Takishima, Sasaki and Sasaki, 1975). We then compared these data with dimensional data of both excised bronchi and of intact bronchi during quasi-static lung deflation.
Methods
Six mongrel dogs weighing 13-15 kg were anesthetized with sodium pentobarbital (40 mg/kg) and sacrificed by exsanguination after an intravascular injection of heparin (300 IU/kg). The right lower lobe was suspended in a Lucite box by connecting the lobar bronchus to a short metal cannula (length = 8 mm; I.D. = 11 mm) attached to the center of the box. The cannula was inserted 1-2 mm into the intrapulmonary bronchial segment and secured with 3 6 ligatures. The pulmonary vessels were not tied. Lung volume was changed by means of collateral channels using three hollow hemispherical Lucite capsules (diameter = 10 mm) glued to the pleural surface of the expanded lobe (Sasaki, Sasaki and Takishima, 1977; Takishima et al., 1971) as shown in fig. 1. The pleural surface beneath each capsule was punctured by about 30 holes, each one approximately 1 mm in diameter and 2 mm deep. Each capsule was connected to a vinyl catheter (length = 30 cm; I.D. = 5 mm). Two of them functioned to inflate the lobe. The third was connected to two transducers (Hewlett-Packard, 267 BC) for the measurement of alveolar pressure (PA) and lung recoil pressure (PL = PA - Pbox). The lobe was then inflated by a negative box pressure (Pbox) of 30 cm H~O and the lobar bronchus was made air-tight by sealing the orifices of its tributary branches with 10-15 beads (diameters 2 9 mm), attached by tissue glue (Alonalpha, Sankyo,
BRONCHIAL LENGTH AND DIAMETER BEHAVIOR
PA-%o
109
I11
I tL
Box
/ 8..a_.,
....
~ Syringe
I
~1 Linear transducer
I ~
Vbr
Fig. l. Central bronchi were isolated by glueing obstructing beads in all tributary bronchi. The lung was inflated by negative box pressure (Pbo×). Alveolar pressure (PA) was kept atmospheric through collateral routes via 2 pleural capsules glued on the pleural surface. The third capsule was used for monitoring PA. Intrabronchial pressure (Pbr) was obtained from a pressure transducer, and bronchial volume changes (Vbr) from a syringe coupled to a linear displacement transducer. The blocked lobar main bronchus, the pressure transducer and the syringe were filled with saline. Lung recoil pressure (PL) was obtained from PA-Pbox by a differential transducer. The bronchial length (Lbr) was measured with a linear displacement transducer that was connected by a thin rod to the most distal bead.
Japan) as previously described (Takishima, Sasaki and Sasaki, 1975). Thereby we could obtain the bronchial pressure volume relationships independently of the overall lung volume which was held at various levels by means of negative box pressure. Alveolar pressure was kept uniformly atmospheric through collateral pathways. The bronchial length (Lbr) was measured with a linear displacement transducer (Linear, LDMR 73S, Japan) that was connected by a thin rod to the most distal bead. The transducer was water-proofed and laid horizontally inside the plastic adapter connected to the airway opening along the axis of the main lobar bronchi. The position of the iron core (3 g in weight) was changed smoothly inside the transducer corresponding to Lbr (the displacement of the terminal bronchial bead). The bronchial volume (Vbr) was measured with a 30-ml syringe coupled to another linear displacement transducer (Hewlett-Packard, 7DCDT-1000); the intrabronchial pressure (Pbr) was measured with a pressure transducer (HewlettPackard, 267BC). Both the bronchi and the apparatus were filled with saline. The box was partially filled with a saline solution, and laid on its side, so that the lobe floated on the fluid and the lobar bronchus was held horizontally and in a straight line. The gravitational effect was thus reduced. The bronchial pressure was measured taking the central plane of the bronchi as the zero reference level. A change in the intrabronchial pressure was obtained by withdrawing a measured volume of the saline solution with the syringe. Zero Vbr was established statically
110
M. NAKAMURA ~'t a/.
in lobes at P L = 0 cm H~O by lowering the Pbr to - 1 0 0 cm H,O. First, we measured PL vs. Lbr and PL vs. Vbr curves of the intact bronchi at 5 lung volumes (PL = 30, 10, 5, 2 and 0 cm H,O) keeping both Pbr and PA at 0 cm H~O. Then, by withdrawing saline from the syringe we obtained Pbr-Lbr and Pbr-Vbr curves of the same lobes at the constant lung volumes of PL = 30, 10, 5 and 2 cm H,O. Afterwards, Vbr was returned gradually to its initial value and the lobe was inflated once again to PL = 30 cm H,O. After completing this series of measurements, the bronchi were carefully dissected from the surrounding tissues by removing the parenchyma, branches and large vessels. The zero Vbr was established again by lowering the Pbr to - 1 0 0 cm H,O. The transmural pressure-bronchial volume (Ptm-Vbr) and the Ptm-Lbr curves were then obtained in the same manner as for the intact bronchi, by applying negative Pbox of 30, 10, 5, 2 and 0 cm H,O. The alveolar pressure, bronchial volume and lung recoil pressure were recorded on a pen-writing oscillograph (Sand, 8S). The Pbox was measured with a water manometer. The Pbr-Lbr and Pbr-Vbr relationships were recorded on two X - Y recorders (Hewlett-Packard, 7045A) during the gradual deflation of Vbr (0.03 ml/s from Pbr 0 to - 8 0 cm H 2 0 ). It took about 2 3 min to record each Pbr-Vbr curve. In separate preliminary measurements of the effect of blocking bronchi on lung PV curve at low lung volume, the lobe was inflated to alveolar pressure of +30 cm H , O via two capsules, then deflated to PL = 2 cm H,O. Lobe volume was measured by water-displacement, subtracting lobe weight and assuming the density of lung tissue to be 1.0. The difference of lobe volumes before and after blocking bronchi was within 21'~ TLC. When the lobe was deflated to PL = 0 cm H eO, the residual volumes after blocking increased 2.7_+ 1.37Jo TLC, indicating that this procedure did not significantly increase gas trapping at PL = 2 cm H20, or even at RV. All experiments were carried out at 20+_2°C room temperature. The measurement of curves began between 30-60 min after exsanguination and continued for about 3 h.
Results The mean volumes and lengths of the blocked bronchial segments of the 6 dog lobes were 3.24+_0.59 ml (SD) and 3.97+_0.60 cm (SD), respectively, at full inflation (Pbr = 0 and PL = 30 cm H:O). Length was measured as the distance from the midpoint of the terminal bead to the inlet of the cannula. Figure 2 shows the in s i t u changes of Lbr v e r s u s PL and v e r s u s Pbr. Lbr at full inflation in each of the 6 dog lobes was taken as 100~,,,. When the lobe was deflated while holding Pbr and alveolar pressure at zero, bronchial lengths decreased by 24.3+_ 1.7% (mean+ SD) as the PL decreased from 30 to 2 cm H20. However, when PL was held constant at 30 cm H,O, Lbr decreased by 1.5+- 1.3~0 as Pbr was lowered from 0 to - 8 0 cm H20. At the remaining PL (10, 5 and 2 cm H20), Lbr shortened somewhat more (an average of 6.8+-0.77{i). Thus, in intact lobes Lbr
B R O N C H I A L L E N G T H A N D D I A M E T E R BEHAVIOR
111
Intact bronchi PL 30~
~
•
- •
je
-e- 100 I(*/.max)
5~ "--'--------~
80
2 ~
70
/e
J
J
/
60 50 -80
-60
-&0
Pbr
-20
02 5
( cmH20 )
10
20
30
PL
Fig. 2. The dependence of Lbr on PL of the blocked bronchi at Pbr = 0 cm H 2 0 is shown in the right panel. When Pbr was made negative at various constant lung volumes for PL = 30, 10, 5 and 2 cm H20, the curves in the left panel were obtained. Lbr is expressed as o / o f the bronchial length at full inflation (PL= 30 and Pbr = 0 cm H 2 0 ) on all curves. Bars represent the mean values +1 SD for the 6 dog lobes.
exhibited only a slight tendency to decrease at constant lung volumes, even when Vbr was reduced by up to 70~o. Figure 3 shows corresponding changes in mean Dbr v e r s u s PL and v e r s u s Pbr. The Dbr were calculated from Vbr and Lbr at each value of PL and Pbr, assuming the geometry of a uniform cylinder: Vbr = n(Dbr/2)2(Lbr). The Dbr at full inflation in each of the 6 dog lungs was taken as 100~o. During lobe deflation, Dbr decreased spontaneously by 23.0+ 3.0~o as PL decreased from 30 to 2 cm H 2 0 at Pbr = 0 cm H20. During forced bronchial collapse to Pbr = - 8 0 cm H20, Dbr decreased by 40.2 + 3.5~o at PL = 30 cm H 2O, and by an average of 78.3 + 3.6~o at the remaining PL (10, 5 and 2 cm H:O). Thus, at fixed lobe volume the changes of Dbr during forced collapse were more than 10 times larger than those of Lbr. In fig. 4 changes in Lbr are plotted v e r s u s simultaneous changes in Dbr for the intact and dissected bronchi at the 5 different PL (or Ptm) for cases where Pbr was maintained at zero. In this plot resting dimensions are taken as 100~o. Lbr of intact bronchi increased to 143.0+ 15.1 ~o, over the vital capacity range from PL (or Ptm) 0 to 30 cm H20. At the same time Dbr increased to 142.8+ I0.1~o in intact and to 149.6+ 17.3~ in dissected bronchi. Resting dimensions themselves were slightly altered when the bronchial segment was dissected from the surrounding parenchyma: Lbr decreased by 1 0 . 3 + 9 . 8 ~ and Dbr increased by 9.1+6.3~o relative to the intact dimensions at PL = 0 cm H20. In the intact bronchi not undergoing forced collapse, the changes of Lbr and Dbr lay fairly close to the line of identity over the vital capacity range, whereas isolated bronchi first shorten markedly with only minor narrowing as Ptm falls from 30 to 5 cm H20, followed by major narrowing at lower Ptm. Thus, intact bronchi contract more isotropically
112
M. N A K A M U R A
ul al.
) Intact b r o n ~
q } / / "
"/~}
U "
~'
f -
4(
,,' ,,"
10
/ 5
~
'
"
Dissected bronchi," Length fixed .''.'"
,
, -GO
. . . .
, ~ " ,
,
-z,O
2(
Length not fixed
.-''.'-''" -80
"'
---
-20
Pbr
0
0 0 2 5
( cmH20 )
10
20
30
PL
Fig.
3. The d e p e n d e n c e of D b r on PL of the b l o c k e d b r o n c h i at Pbr = 0 cm H~O is s h o w n at the r i g h t : b r o n c h i a l collapse d a t a at the c o n s t a n t l u n g v o l u m e s for PL = 30, 10, 5 a n d 2 cm H , O are illustrated on the left. D b r was c a l c u l a t e d from Vbr a n d Lbr. D b r was expressed as I~;>o f d i a m e t e r at full inflation (PL = 30 a n d Pbr = 0 cm H 2 0 ) . Bars show the m e a n values +1 S D for the 6 dog lobes.
160] ~'~
Ptm
i
"6
" '°bJ~
140]
__
_ [//~"
.....
J ~.
,
30¢mH20
L/
,2o]
"ti
1001100
|
[ ,201 [ Lbr
,,10
1;0
(% of resting)
80-
Fig. 4. The % D b r % Lbr r e l a t i o n s h i p s o f intact b r o n c h i ( H ) a n d dissected b r o n c h i (x x ) at PL (or Ptm) = 30, 10, 5, 2 a n d 0 cm H~O. D b r a n d Lbr are expressed as i',i o f resting length at Pk (or Ptm) = 0 cm H~O. Bars represent the m e a n values _+1 SD for the 6 d o g lobes. The d o t t e d line is the line o f identity (isotropic b r o n c h i a l expansion). The d a s h e d line, ~k ~--~D2, is a theoretical relation for an inflated closed-end i s o t r o p i c elastic tube. I n t e r r u p t e d steep lines show data found d u r i n g b r o n c h i a l collapse.
B R O N C H I A L L E N G T H A N D D I A M E T E R BEHAVIOR
113
than dissected bronchi, reflecting the effects of the surrounding lung parenchyma tending to induce homogeneity and isotropicity on embedded structures. Also on fig. 4 are shown the data from forced bronchial collapse at each of the 4 different fixed PL. Under these conditions interdependence factors are manifested in quite different ways: Lbr tends to remain fixed while Dbr undergoes extensive narrowing. These features are further examined below.
Discussion
In the present study we have demonstrated that intact bronchi contract nearly isotropically during lung deflation, quite unlike the behavior of dissected bronchi on the one hand, or unlike forced bronchial collapse at fixed lung volume on the other. Our findings may be explained by the existence of a quasi-uniform distribution of stress in the intact lung, in contrast to uneven distributions of stresses in the radial and longitudinal directions both in dissected bronchi and around intraparenchymal bronchi undergoing distortion. The unevenness probably arises from the nonsymmetric bronchial structure (cylindrical rather than spherical). When non-spherical cavities in elastic media are inflated or deflated, complex deformations are to be expected, as partially analyzed below. These anisotropic strains do not necessarily imply anisotropic material properties, but these could certainly further contribute to the distortion. Based on bronchograms, the percent changes of Lbr and Dbr versus PL or VL during lung inflation or deflation have been reported by many investigators (Hughes, Hoppin and Mead, 1972; Hughes, Jones and Wilson, 1975; Hughes et al., 1974; Prakash and Hyatt, 1978; Sittipong and Hyatt, 1974). Hughes et al. reported that in intact bronchi of dog lungs Lbr decreased 38~o (1972) and 30~o (1974) during deflation of the lobes from PL = 30 to 0 cm H:O. By comparison, our results showed that Lbr decreased 30.7+2.1% over the same range of PL. In our study we employed a linear displacement transducer for measuring changes in Lbr directly. In preliminary experiments in 2 lobes we compared changes of Lbr obtained from orthogonal bronchograms at the same 5 different PL as above, using one terminal bronchial bead as a radiopaque marker, with the data from the displacement transducer connected to the same wedged bead. No systematic differences were found. Hughes et al. also reported that in intact bronchi of dog lungs Dbr decreased 42~o (1972) and 36~o (1975) during deflation from PL = 30 to 0 c m H 2 0 . Our results showed that the bronchial diameter decreased 29.8+ 5.0% over the same range of PL. This figure would be expected to vary with the resting volume of the lungs, which varies depending on the preparation, volume history, etc. Several investigators (Hughes, Hoppin and Mead, 1972; Hughes, Jones and Wilson, 1975; Hyatt et al., 1970; Prakash and Hyatt, 1978; Sittipong and Hyatt, 1974) have shown that the changes of bronchial length during deflation were essentially proportional to
114
M. N A K A M U R A
et ~tl.
changes in the cube root of absolute lung volume, but that changes of bronchial diameter were different depending on the bronchial sizes. In our study we examined mainly the behavior of larger lobar bronchi and for these the changes of Dbr in situ were similar to changes of Lbr (fig. 4). When one considers the interdependence between airspaces and bronchi (Mead, Takishima and Leith, 1970), if the specific compliance of bronchi were not too different from lung specific compliance and nearly isotropic, one would expect both Dbr and kbr to vary a s VL 1 3 However, we also showed that during bronchial collapse within the lung at constant VL, Lbr and Dbr were not at all similar. During bronchial collapse Dbr within the lung parenchyma depends on bronchial transmural pressure, but Lbr depends primarily on lung volume and only slightly on transmural pressure. The grossly unequal strains may be understood by the following approximation. Consider a closed cylindrical bronchial segment of radius R and length L surrounded by parenchyma such that alveolar and bronchial pressure are zero. The force shortening the segment, when a negative Pbr is applied, is (~R:) (Pbr) on each end. This produces an average shear stress over the walls of the bronchus (force/area) of 2(rcR:)(Pbr)/2r~RL = R(Pbr)/L. On the other hand, the radial stress is Pbr. Thus, if R/L is about 1/8, as it was in our preparation, one might expect roughly 8 times as much circumferential as longitudinal stress, and proportionally different changes in Dbr and Lbr. However, during bronchial collapse R decreases by 501'~i or more, so that the ratio R/L could fall to 1/16 or less. Our measured ratio L/R (mean 16.4+ 7.6) confirms this general trend. One would further predict that if the length of the segment were reduced, e.g. to L = R, then changes in Lbr and Dbr would become nearly equal. Conversely, if the segment were made long and narrow (R/L small), almost no shortening would occur. The strains in the inflated excised bronchus may as a first approximation be treated from the analysis of inflated cylinders made of incompressible isotropic elastic material. It is well-known that longitudinal strain (~L) is zero in such a cylinder when it is undergoing small pressure expansions. However, in moderate expansion EL is not zero and may be roughly related to the tangential strain (eY) by e L - eT: (Hildebrandt, 1970). Since tangential (or circumferential) strain is just proportional to percent change in diameter, ~D, this relation can be written as eL ~ eD 2. This equation is plotted in fig. 4 (dashed line) for diameter changes up to about 5°°/~ Jo and also indicates a reasonable agreement. The general characteristics of the segment are described by the curve, although the existence of some anisotropy is suggested by the fact that length changes always exceed those predicted. Thus, the circumferential rigidity appears to be greater than the longitudinal rigidity, and this is as would be expected from length tension studies. Using dog lobes before and after their static lung recoil had been significantly increased by cooling and ventilating, Hyatt, Rodarte and Wilson (1975) showed that the diameter corresponded best to lung recoil~ and length to lung volume. Hoppin, Lee and Dawson (1975), using a grappling hook technique, demonstrated that bronchial lengths were greatly stabilized h7 situ because of their much greater
B R O N C H I A L LENGTH A N D D I A M E T E R BEHAVIOR
115
longitudinal stiffness compared to dissected bronchi. These results also showed that the mechanical interdependence for longitudinal traction was different from that for radial traction applied by surrounding lung parenchyma. We would account for these differences largely on geometrical grounds, as indicated above. Pride et al. (1967) suggested that the maximum expiratory flow was determined by the collapsibility of the tracheobronchial tree, by lung elastic recoil, and by airway resistance. Takishima and Sasaki (1972) also suggested theoretically that maximum expiratory flow might be determined by both airway resistance and airway compliance. Using the tube wave speed concept in the analysis of flow limitation in elastic tubes, Dawson and Elliott (1977) also showed that flow rate occurring at wave speed was a function of airway dimensions and specific compliance of the tube. Therefore, airway collapsibility is an important factor in determining maximum expiratory flow. Takishima, Sasaki and Sasaki (1975) reported that the differences of the intrapulmonary bronchial collapsibility between intact and dissected bronchi were large and those differences increased with higher PL. We considered three reasons why airway collapsibility might decrease with higher PL: first and most important, radial traction applied by surrounding parenchyma increases with PL (Mead, Takishima and Leith, 1972; Nakamura, Sasaki and Takishima, 1979); second, the bronchial wall is slightly stiffened when bronchial length is extended by large lung inflation (Me!issinos and Mead, 1977); third, at constant lung volume the changes of Lbr are small during bronchial collapse. The significance of the first factor is apparent from such data as shown in fig. 3 where parenchymal distending stresses of 20-40 cm HzO are typical. To evaluate the second factor, we compared the pressure-volume relations of dissected bronchi in 4 dogs before and after fixing the initial bronchial length at that found in intact lung at PL = 30 cm H20. With fixed bronchial length, the pressure-volume curves of dissected bronchi indicated slightly less collapsibility (fig. 3) by 2.1 + 0.4 cm H20 at 50~ of maximum Dbr and 3.0+ 0.7 cm H20 at 30~o maximum Dbr. Melissinos and Mead (1977) reported that maximum expiratory flow increased in some subjects during neck hyperextension and that possibly increased tracheal longitudinal tension decreased the airway collapsibility. The third factor is related to the second, insofar as shortening of Lbr during bronchial collapse was very small, further tending to stabilize diameter.
References Dawson, S.V. and E.A. Elliott (1977). Wave-speed limitation on expiratory flow - a unifying concept. J. Appl. Physiol. 43: 498-515. Hildebrandt, J. (1970). Extension of small-strain theory to finite deformation of cylindrical vessels by internal over-pressure. Angiologica 7: 25~272. Hoppin, F . G . , Jr., G.C. Lee and S.V. Dawson (1975). Properties of lung parenchyma in distortion. J. Appl. Physiol. 39: 742-751.
1t6
M. N A K A M U R A el u/.
Hughes, J. M.B., F.G. Hoppm, Jr., and J. Mead (1972). Effect of lung inflation on bronchial length and diameter in excised lungs. J. Appl. Physiol. 32:25 35. Hughes, J. M. B., H.A. Jones, A.G. Wilson, B.J.B. Grant and N.B. Pride (1974). Stability of inlrapulmonary bronchial dimensions during expiratory flow in excised lungs. J. Appl. Plo'siol. 37 : 684-694. Hughes, J. M. B., H. A. Jones and A. G. Wilson (1975}. Bronchial hysteresis in excised lungs. J. Physiol. (Lomlon) 249:435 443. Hyatt, R.E., R. Sittipong, S. Olafson and W.A. Potter (1970). Some factors determining pulmonary pressure-flow behavior at high rates of air flow. In: Airway dynamics, edited by A. Bouhuys. Springfield, IL., Thomas, pp. 43 60. Hyatt, R. E., J. R. Rodarte and T.A. Wilson (1975). Effect of increased static lung recoil on bronchial dimensions of excised lungs. J. AppI. Physiol. 39 : 429 433. Mead, J., T. Takishima and D. Leith (1970). Stress distribution in lungs : a model of pulmonary elasticity. J. Appl. Ph3siol. 28:596 608. Melissinos, C. G. and J. Mead (1977). Maximum expiratory flow changes induced by longitudinal tension on trachea in normal subjects. J. Appl. Physiol. 43:537 544. Nakamura, M., H. Sasaki and T. Takisbima (1979). Effect of lung surface tension of bronchial collapsibility in excised dog lungs. J. Appl. Physiol. 47:692 700. Prakash, U.B.S. and R.E. Hyatt (1978). Static mechanical properties of bronchi in normal excised human lungs. J. Appl. Physiol. 45:45 50. Pride, N.B., S. Permutt, R.L. Riley and B. Bromberger-Barnea (1967). Determinants of maximal expiratory flow from the lungs. J. Appl. Physiol. 23:646 662. Sasaki, H., T. Sasaki and T. Takishima (1977). Influence of lung parenchyma on dynamic bronchial collapsibility of excised dog lungs. J. Appl. Physiol. 42:699 705. Sasaki, H., F.G. Hoppin, Jr., and T. Takishima (1978). Peribronchial pressure in excised dog lungs. J. Appl. Physiol. 45:858 869. Sittipong, R. and R.E. Hyatt (1974). Static mechanical behavior of bronchi in excised dog lung. J. Appl. Physiol. 37:201 206. Takishima, T., K. lshikawa, T. Sasaki, H. Sasaki and T. Nakamura (1971). Measurement of collateral flow at quasialveolar level in excised dog lung. Tohuku J. Exp. Med. 105:405 406. Takishima, T. and H. Sasaki (1972). Two-dimensional flow-model for analysis of expiratory check valvc. Bull. Ph.vsio-Pathol. Resp. 8:361 374. Takishima, T., H. Sasaki and T. Sasaki (1975). Influence of lung parenchyma on collapsibility of dog bronchi. J. Appl. Physiol. 38:875 881.