Mechanical properties in excised pony lungs

of small airways

L. E. OLSON Department of Veterinary Physiology and Pharmacology, The Ohio State University, Columbus, Ohio 4321 O-l 092 OLSON, L. E. Mechanical properties of small airways in excised pony lungs. J. Appl. Physiol. 73(2): 522-529, 1992.-We

evaluated the pressure-flow relationships in collaterally ventilating segmentsof excisedpony lungsby infusing N2, He, Ne, or SF, at known flows (v) through a catheter wedgedin a peripheral airway. Measurementswere madeat segment-(Ps) to-airway opening (Pao) pressuredifferentials of 345 cmH,O when the lungs were held at transpulmonary pressuresof 5, 10, and 15 cmH,O. The data were analyzed both by calculating collateral resistance(Ps - Paolv) and by constructing Moody-type plots of normalized pressuredrop [(Ps - Pao) /( 1/2pU2, where p is density and U is velocity)] against Reynolds number to assessthe pattern of flow through the segmentand the change in dimensionof the flow channelsas Ps and Pao were changed. The interpretations from these analyseswere compared with radiographic measurementsof the diameters of small airways within the collaterally ventilating lung segmentat similar pressures.Collateral resistanceincreasedas Ps - Pao increasedat high Reynolds numbers,i.e., high flows or densegas(SF,). Analysis of the Moody-type plots revealed that flow was density dependent at Reynolds number >lOO, which frequently occurred when N, was the inflow gas.The radiographic data revealed that small airway diameter increased as Ps - Pao increasedat all lung volumes.In addition, at 5 cmH,O Pao, smallairway diameter was smaller for a given Ps in the nonhomogeneouscase (Ps > Pao) than small-airway diameter for the samePs in the homogeneouscase(Ps = Pao). We interpret thesedata to suggestthat the surrounding lung prevented the segment from expanding in the nonhomogeneouscase. Taken together, these data suggestthat collateral resistance measuresthe properties of structures that behave like small airways in pony lungs. collateral ventilation; small airway diameter

COLLATERAL RESISTANCE is calculated by infusing gas through a catheter wedged in a small airway and measuring the pressure difference between the distal lung segment and the trachea relative to the rate of airflow (V). Collateral resistance is therefore the pressure cost associated with the bulk transfer of gas between adjacent lung units through anatomic structures collectively referred to as “collateral pathways.” Controversy exists regarding the effect of increasing airflow on collateral resistance. Although most studies report that collateral resistance increases as the pressure difference between the segment and trachea increases (2, 3, 15), there are reports in intact lungs to the contrary (6), which may be attributable to the concentration of CO, in the inflow gas (5). 522

The mechanism for the increase in collateral resistance with increasing segment pressure has been investigated in excised and intact dog lungs by assessing whether collateral resistance changed as the density of the inflow gas was changed (3,9,13). These studies demonstrated that, at a constant pressure differential, collateral resistance increased with the density of the inflow gas. These results strongly suggest that changes in collateral resistance cannot be simply interpreted to reflect changes in airway geometry. Collateral resistance is an order of magnitude higher in excised horse lungs than in excised dog lungs, a finding that correlates with the frequency and severity of obstructive lung diseases in these two species (15). However, whereas increasing the rate of airflow into the collaterally ventilating segment increased collateral resistance in dog lungs held at volumes corresponding to airway pressures of 5, 10, or 20 cmH,O, collateral resistance in horse lungs increased only at airway pressures of 10 or 20 cmH,O. These results suggest that either the flow characteristics or compliance of small airways or parenchymal interdependence differs between horse and dog lungs. The present study was therefore undertaken to evaluate the mechanical properties of small airways in excised pony lungs. Two different techniques were used, and the interpretations from each were compared. We reasoned that the pressure differential necessary to generate a given collateral airflow would be a function of both the geometry of the air-conducting pathways distal to the wedged catheter and the pattern of flow through those pathways. Furthermore, we reasoned that the geometry would be a function of the degree of expansion of the segment and the compliance of the pathways relative to the compliance of the parenchyma and would be fixed by the difference between segment alveolar pressure and alveolar pressure in the surrounding parenchyma. We considered that the degree of expansion of the segment for a given segment pressure would be a function of the effective compliance of the segment, which would depend on the mechanical relationship between the segment and the surrounding parenchyma. The relationship between driving pressure and flow through a collaterally ventilating segment was assessed by using gases of different densities and viscosities to determine whether the observed increase in collateral resistance could be attributed to a change in flow pattern rather than a decrease in the effective cross-sectional

0161-7567/92 $2.00 Copyright 0 1992 the American Physiological Society

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

AIRWAY

MECHANICS

area for flow. If the relationship between pressure and flow was linear, resistance could be calculated by the Hagan-Poiseuille equation. Calculated resistance would be directly proportional to the viscosity of the inflow gas and independent of the density of the inflow gas. If the magnitude of the calculated resistance was influenced by the density of the gas, the calculation of resistance as the simple ratio of pressure to flow would be inappropriate. The diameters of small airways within the collaterally ventilating segment were measured under similar pressure conditions to provide an estimate of the change in small airway diameter with increases in driving pressure.

IN

PONY

TABLE

Viscosity, Density, Data

523

LUNGS

1. Gas physical

pP g/l from

Braker

properties

He

Ne

N?

SF,

201 0.18

318 0.82

179 1.14

156 6.16

et al. (1).

(p; Table 1), and the flowmeter was calibrated by timed collections of each gas. Flow through the collaterally ventilating lung segment was measured for each gas when Pao was maintained at 5, 10, or 15 cmH,O and Ps - Pao was increased to 15 cmH,O and then decreased to 3 cmH,O in 3-cmH,O inMETHODS crements. When gases were changed the new test gas was infused through the segment at a rate estimated to wash Lungs were carefully excised from ponies that were the resident gas from the segment before measurements killed by an intravenous overdose of pentobarbital so- were made. The order in which Pao’s were studied was dium. The right and left lungs were separated and varied. usually vacuum degassed before study. A total of 19 right Collateral resistance was calculated as the ratio of or left lungs were studied from 17 ponies. For studies Ps - Pao to flow (V) through the segment. In addition, involving isoproterenol, 100 ml of 0.5% isoproterenol Moody-type plots were generated for each lung. The kiwere poured into the airways and then drained after 10 netic energy loss was calculated to be insignificant; theremin. During the studies, the lungs were covered with fore the frictional pressure losses were assumed to equal moistened gauze and periodically sprayed with normal the static pressure difference (Ps - Pao). The normalsaline to prevent drying. All transducers were calibrated ized pressure (Pn) drop across the collaterally ventilatdaily before use. ing lung segment was computed as Ps - Pao/( l/ZpU2), Pressure-lung v&me studies. The lungs were rapidly where Uis velocity at the catheter tip (U = V/n?, where cycled through several large volume inflation-deflation r is radius of wedged catheter = 0.25 cm). Reynolds numcycles with use of a compressed air source through a ber (Re) was calculated as pdU/p, where d is the diameter water-bottle pop-off system. The lungs were then in- of the wedged catheter. Log Pn was plotted as a function flated to an airway opening pressure (Pao) of 25 cmH,O, of log Re to permit inferences regarding the pattern of and Pao was recorded while known volumes of gas were flow through the segment and the dependence of airway removed from the lungs with a 1.5-liter syringe. At least diameter on pressure (3, 9, 12, 13, 16). two curves were recorded for each lung, and the results Pressure-airway diameter studies. The lungs were prewere averaged. Pressure was measured with a Validyne pared in the same manner as the pressure-flow studies, transducer (DP 45) and recorded on a chart recorder except humidified N, was used as the inflow gas. Con(Gould 2400 S). Lung volume was expressed as a percenttrast was enhanced in airways distal to the catheter (“inage of the volume of gas that could be removed from the trasegmental airways”) by blowing tantalum dust lung as it was deflated from 25 to 0 cmH,O Pao (vital through a small cannula placed in the wedged catheter. capacity). Specific compliance was calculated as the Airways in a region of lung some distance from the slope of the curve between 5 and 10 cmH,O Pao normalwedged catheter were also dusted with tantalum to serve ized to lung volume at 5 cmH,O Pao. as control airways (“extrasegmental airways”). Radiographs were obtained when Ps and Pao were equal (Ps = Pressure-flow studies. The pressure-flow characterisPa0 = 0, 5, 10, 15, 20, 25, and 30 cmH,O), the homogetics of the air-conducting pathways in the collaterally neous case, and when Pao was maintained constant at 5, ventilating lung segment distal to the wedged catheter were evaluated as previously described (9, 12, 13). A Y 10, or 15 cmH,O and Ps was greater than Pao (Ps = connector was attached to the main stem bronchus, and a Pao + 5 cmH20, Ps = Pao + 10 cmH20, and Ps = Pao + 15 cmH,O), the nonhomogeneous case. The lungs were polyethylene catheter (0.5 cm OD) was advanced into the lung until it became wedged in a peripheral airway. A inflated to 25 cmH,O before each measurement, and the small side-hole cannula within the wedged catheter was order in which the airway pressures were tested was used to measure pressure at the catheter tip to provide an varied. The method for obtaining radiographs has been deestimate of segment pressure (Ps). Pao was measured, scribed previously (10). Radiographs were taken with a and lung volume was adjusted by inflating the lung with compressed air channeled through a water-bottle pop-off fixed Machlett 50-D X-ray tube (focal spot l-2 mm) ussystem. Ps - Pao was adjusted by regulating the flow of ing Kodak OMl film in Kodak X-omat cassettes with single Lanex fine screens. The technique was adjusted to gas through the wedged catheter and was measured with a Validyne transducer (DP 45). A manifold was used to give good contrast and had the following ranges: 80-100 deliver humidified N,, Ne, He, or SF, through a flow- kV, 4-8 mA, and 0.03-0.05 ms. Films were developed in a meter (Fleisch no. 0000 pneumotachograph) into the Kodak X-omat M35 processor. The diameters of airways wedged catheter and distal lung segment. Gases were that were clearly visible in each film were measured with chosen to provide a range of densities (p) and viscosities a back-lighted digitizing tablet and hand-held cursor Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

524

AIRWAY

MECHANICS

IN

PONY

LUNGS

2. Specific compliance of excised pony lungs 5 cmH,O Pao

Pao and Ps - Pao of 3 cmH,O. These points are therefore not shown in the composite data presented in Fig. 2 and were not included in the statistical analysis. At 15 Isoproterenol Rinsed Untreated cmH,O Pao, collateral resistance increased as Ps - Pa0 increased, irrespective of the inflow gas. At 5 cmH,O 0.11 0.09 Pao, collateral resistance increased as Ps - Pao in0.12 0.11 0.14 0.11 creased only when SF, was infused. Collateral resistance 0.19 0.16 increased as Ps - Pao increased when He, N,, or SF, was 0.17 infused at 10 cmH,O Pao. If the increased resistance was 0.13 because of a decrease in the effective cross-sectional area Values are in cmH,O-‘. for flow caused by the increase in Ps - Pao, the effect should have been observed with each gas. These data (GTCO). Between 7 and 10 intrasegmental airways and 6 suggest that at Pa0 = 5 and 10 cmH,O the increase in and 10 extrasegmental airways were measured in each resistance was likely due to a change in the pattern of lung. The distances between the cassette, X-ray source, flow. and stand on which the lung was placed were constant; Moody-type plots of log Pn as a function of log Re were therefore, the diameters were not corrected for magnificonstructed for each lung studied to assess the nature of cation. the relationship between pressure and flow. Although The diameter of each airway was normalized to its dithere was variability in the data, particularly at 5 cmH,O Ps = 30 cmH,O. For each combination ameter at Pa0 = of pressures studied, normalized intrasegmental and ex- Pao, there were also some clear trends that became aptrasegmental airway diameters were averaged for each parent when individual curves for each lung were visually analyzed. Cu rves for the least and most (1 and 6) variable lung. cases are shown for comparison in Fig. 3. Visual analysis Statistical analyses. Specific compliances at 5 cmH,O of the slopes of the curves suggested that, at Re < 100, Pao in freshly excised and isoproterenol-rinsed lungs viscous losses predominated (slope = -1) and, at Re > were compared by t test for unequal sample sizes (17) to 1,000, dynamic pressure losses predominated (slope = 0). assess the effect of isoproterenol on the pressure-volume A transitional pattern was observed between these two characteristics of the lung. The effect of isoproterenol on extremes (slope between 0 and -1). the pressure-diameter behavior of small airways was deAt equivalent Ps - Pao, increasing lung volume by termined by a repeated-measures two-way analysis of increasing Pao shifted the curve downward and rightvariance (19). The effects of pressures on airway diameward (lower Pn at constant Re). This result is consistent ter and on collateral resistance were analyzed with Friedman’s test, the nonparametric alternative to the with an increase in the dimensions of the flow pathways two-way analysis of variance (17). Airway diameters at as Pao increased. Because the magnitude of the increase was unknown, the dimension used in calculating velocity equal Ps in the homogeneous and nonhomogeneous and the linear dimension in the Re was the same for each cases were compared by Wilcoxon’s rank sum test (17). Pao. If the magnitude of the increase was known, recalResults were considered significant at P < 0.05. culating Pn and Re to account for the change in dimension would collapse all curves to a single line (12). In RESULTS addition, increasing lung volume decreased the slope of Pressure-uohme studies. The pressure-volume characthe curve at low Re. This is consistent with either a trend teristics of the lung were not affected by isoproterenol, as shown in Table 2; therefore the results were combined. -^ ‘” J Specific lung compliance was 0.13 t 0.03 (SD) cmH,O-l (n = 10 lungs), which compared well with values reported for intact horse lungs of between 0.08 and 0.18 cmH,O-l + + (8, 1% Pressure-flow studies. Pressure-flow studies were performed on six lungs from four ponies. Results using N, as = l the inflow gas are shown in Fig. 1. At a constant Ps Pao, collateral resistance decreased as Pao increased. At a constant Pao > 5 cmH,O, collateral resistance increased as Ps - Pao increased. There was considerable variability in collateral resistance. In addition, for purl Pa0 = 5 cm H,O poses of clarity, one measurement was deleted from Fig. v Pa0 = 10 cm H,O n Pa0 = 15 cm H,O 1 (Pao = 5 cmH,O, Ps - Pao = 3 cmH,O) because the calculated resistance was an order of magnitude greater than other measurements and was >Z SDS above the mean. This value, however, was used in the (nonparametFIG. 1. Collateral resistance measured with N2 as a function of difric) statistical analysis. between pressure in collaterally ventilating segment (Ps) and Figure 2 shows the effect of increasing Ps - Pao on ference airway opening (Pao) pressures in excised pony lungs (n = 6 except collateral resistance for each gas tested. In two lungs, it where noted). * Resistance increased as Ps - Pao increased. + Resiswas not possible to measure flow for all gases at 5 cmH,O tance significantly affected by Pao at constant Ps - Pao. TABLE

at

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

AIRWAY Pa0

0

=

I

]

Pa0

z 30 E < 25 a = 20

MECHANICS

=

10

I

cm

I

I

I 1

1

H,O

-I Pao=l5 ‘2‘ .- 30 E + 25

0

l

a * 20 E s 15 ! ifi = 10 3 .-03 t : 5 r-Y t 0

cm

H,O

N2

He

D SF, V Ne

* f * i

0

I

I

I

I

I

3

6

9

12

15

Ps-Pao

(cm

IN PONY LUNGS

525

and N, followed approximately the same curve irrespective of Ps - Pao and the curve had a slope that was more positive than -1, which indicated that flow was not independent of density. In contrast, at 5 cmH,O Pao points for He, Ne, and N, at identical driving pressures tended to define different curves and have slopes that were equal to or more negative than -1, consistent with an increase in pathway dimension as Ps - Pao increased. If the magnitude of the change in dimension with Ps - Pao was known, recalculating Pn and Re would collapse all points on a single curve. Radiographic studies. Radiographic studies were performed on 14 lungs from 12 ponies. Lung weights averaged 923.6 -t 241.4 (SD) g (n = 9 lungs). Seven of the lungs were rinsed with isoproterenol before study. Although 100 ml of isoproterenol were poured into the airways, on average, 17 ml (range 2-43 ml) were poured out before study. The pressure-diameter relationship in the homogeneous case (Ps = Pao) was not affected by isoproterenol; therefore the results were combined. A total of 119 intrasegmental airways and 127 extrasegmental airways were studied. Average airway diameter at 25 cmH,O Pao was 0.162 t 0.019 (SD) cm for intrasegmental airways and 0.145 t 0.017 cm for extrasegmental airways (n = 14 lungs). The pressure-diameter relationships of the intrasegmental and extrasegmental airways are represented in Fig. 4. Normalized extrasegmental airway diameter increased as pressure increased in the homogeneous case (Ps = Pao) and was unaffected by pressure in the nonhomogeneous case (Ps > Pao). Normalized intrasegmental #6 Pa0

= 5 cm

H,O

i

18

H,O)

FIG. 2. Collateral resistance measured with Na, SF,, He, and Ne as a function of difference between Ps and Pao in excised pony lungs (n. = 6). * Resistance increased as Ps - Pao increased. Pa0

=

10

cm

H,O

toward transitional flow at higher Re or a reduction in the dimensions of the flow pathways as Ps - Pao was increased. We assumed that the geometry of the flow pathways was fixed by a combination of lung volume and driving pressure. Therefore, for each Pao, changes in the dimensions of the flow pathways as Ps - Pao changed could be Pa0 = 15 cm inferred from a comparison of the slopes of the curves for 7 the individual gases at constant Ps - Pao (3,9,13,16). At l He 6 each combination of Pao and Ps - Pao and, therefore, F ‘, \ VSF, each fixed geometry, there were four estimates of flow, v Ne \. one for each gas. Because the flow rate and viscosity and density of the gases differed, these estimates described four distinct points on the Moody-type plot. The line I I I 1 l--l--. -_ 1 I..--J 2’ connecting the four points therefore indicated the flow 0 1 2 3 4 0 1 2 3 4 pattern for the geometry fixed by that lung volume and Log Re Log Re driving pressure. At 10 or 15 cmH,O Pao, and at Re < 100 FIG. 3. Log-normalized pressure (Pn) as a function of log Reynolds there was no evidence to suggest that the dimensions of number (Re) at Pao = 5, 10, and 15 cmH,O for 2 individual excised the flow pathways changed as Ps - Pao increased be- pony lungs. Lung 1 showed least variability and lung 6 showed most tween 3 and 15 cmH,O. In other words, points for He, Ne, variability. Solid lines, slope of -1 for comparison.

ON2

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

AIRWAY

526

MECHANICS

IN PONY LUNGS

when transpulmonary pressure was held constant at 10 or 15 cmH,O and Ps - Pao was increased. Collateral resistance was unaffected by Ps when the lungs were held 1.0 at a transpulmonary pressure of 5 cmH,O. The simplest interpretation of these results is that the effective crosssectional area for flow was directly related to lung vol0.9 ume and inversely related to Ps at higher lung volumes. In other words, the cross-sectional area for flow in3 q 0.8 creased when Pao increased and decreased when Ps in13 creased at constant Pao of 10 or 15 cmH,O. The effective cross-sectional area for flow could change by recruitment 0.7 or derecruitment of flow pathways or by a change in the diameter of a fixed number of patent pathways. This in/ (n-l 1) terpretation assumes that the relationship between pres0.6 sure and flow was independent of the density of the gas. However, at each lung volume studied, collateral resis0.5 tance depended on the density of the inflow gas and increased as Ps was raised when SF, was the inflow gas. At 1.1 the largest lung volume studied, collateral resistance increased as Ps was increased, irrespective of the density of the inflow gas. These findings indicate that flow through the collaterally ventilating lung segment was not purely viscosity dependent. The simple interpretation that a 0.9 change in collateral resistance represents a change in effective flow area is therefore clearly suspect. z Moody-type plots were constructed to estimate the q 098 cl flow pattern and to indicate the appropriateness of the Homogeneous assumption that the characteristic dimension chosen to J . (n-l 2) 0 Ps=Pao 0.7 calculate both Re and Pn was fixed and therefore indeNonhomogeneous pendent of changes in either Pao or Ps. Visual inspection 0 Pao=5 cm H,O of the slopes of the plots revealed that purely viscosity. 0.6 v Pao=l 0 cm H,O dependent flow occurred only at Re < 100, which for the D Pao=l5 cm H,O pressures and flows studied required the use of He or Ne 0.5 as the inflow gas. Re frequently exceeded 100 when N, -5 0 5 10 15 20 25 30 35 was the inflow gas. Furthermore, these plots indicated Pressure (cm H,O) that the characteristic dimension chosen to calculate Re FIG. 4. Normalized airway diameter (DID,,) of intrasegmental (A) and Pn increased as transpulmonary pressure increased, and extrasegmental (B) airways as a function of pressure in homogeconsistent with the simple interpretation of the resisneous and nonhomogeneous cases (Ps > Pao, open symbols; n = 14 tance data. Although variability in the data made addilungs except where otherwise noted). * Diameter increased as Ps increased. tional interpretations difficult, analysis of individual curves offered some suggestion that the characteristic airway diameter increased as Ps increased in both the dimension increased as Ps - Pao increased when the lungs were held at a transpulmonary pressure of 5 homogeneous and nonhomogeneous cases. At 5 cmH,O cmH,O, which was not consistent with the simple interPao and 15 and 20 cmH,O Ps, intrasegmental airway pretation of the resistance data. There was no suggestion diameter was smaller in the nonhomogeneous case than of a similar increase in dimension as Ps increased at in the homogeneous case. transpulmonary pressures of 10 or 15 cmH,O. However, this result must be interpreted cautiously, because an DISCUSSION increase in the characteristic dimension and a change These studies were undertaken to evaluate the me- from a flow pattern representing purely viscous pressure chanical properties of small airways in pony lungs. Ro- losses to a flow pattern in which dynamic pressure losses binson and Sorenson (15) showed that collateral resis- were significant can have opposing effects. tance measured with 0, as the inflow gas was high in Radiographic studies were therefore undertaken to diequine lungs compared with measurements with a simirectly determine the relationship between intrasegmenlarly sized catheter (0.25 cm OD) in dog lungs, (68.3 vs. tal airway diameter and alveolar pressures (PA). Compar9.5 cmH,O 1-l min, respectively, at 5 cmH,O Pao and isons between pressure-diameter data and pressure-flow airway diameter can Ps - Pao of 5 cmH,O). At 5 cmH,O Pao and Ps - Pao of 6 data presume that intrasegmental cmH,O, using a larger catheter (0.5 cm OD) and N, as be related to the characteristic dimension of the collatthe inflow gas, we measured a resistance of 25.9 erally ventilating lung segment that was chosen to calcucmH,O l min in excised pony lungs. In addition, we late Re and Pn. The diameter of the catheter tip was used confirmed their results showing that collateral resistance in the calculation of Re and Pn for two reasons. First, the airwav in which the catheter was wedged was essentiallv decreased as lung volume increased and then increased 1.1

.A

l

1-l

l

l

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

AIRWAY

MECHANICS

an extension of the catheter and therefore likely to be the same diameter as the external diameter of the catheter. Second, it was necessary to estimate velocity, and we knew that the volume flow through the airway just distal to the catheter equaled measured volume flow through the catheter. We know that the resistance to collateral airflow includes the resistances of the intrasegmental airways and the collateral pathways and that the precise anatomic structures conducting collateral airflow in equine lungs are unknown. We reasoned that if the interpretation of the pressure-diameter studies was consistent with the interpretation of the flow studies, we could hypothesize that collateral resistance assessed the properties of pathways that behaved like small airways. The radiographic studies in the homogeneous case demonstrated that the diameter of small airways in pony lungs increased as the lungs inflated, as has been shown previously (4, 7, 10, 11, 14). This finding was consistent with the interpretation from the pressure-flow data that collateral resistance decreased and the resulting line on the Moody-type plot was moved downward and rightward. In addition, intrasegmental airway diameter increased when Ps was increased and transpulmonary pressure was held constant at 510, or 15 cmH,O (nonhomogeneous case), i.e., during conditions similar to those established when the pressure-flow data were collected. This result was not consistent with the simple interpretation of the finding that collateral resistance increased as Ps increased. Although analysis of the Moody-type plots could be interpreted to be consistent with this finding, the finding of a slope between -1 and 0 coupled with variability in the data made a conclusive interpretation difficult. A conclusion from these data could be that intrasegmental airways and collateral pathways respond differently to increased Ps and that collateral resistance increased even as intrasegmental airway diameter increased because of a decrease in the effective cross-sectional area of the collateral pathways. Although it is possible that the intrasegmental airways do not represent the major resistance to collateral airflow, experimental evidence does not support the hypothesis that airways within the segment and air-conducting pathways connecting the segment with the surrounding lung respond differently to increases in Ps (3). Fuller and Robinson (3) used an alveolar capsule to measure PA in a collaterally ventilating segment of excised dog lung, in addition to measuring Ps at the tip of the wedged catheter. Collateral resistance was then partitioned into an intrasegmental airway component (Ps PA/~) and an intersegmental component (PA - Paolv). Both resistances increased as Ps increased, although intersegmental resistance exceeded intrasegmental resistance when flow was viscosity dependent. We conclude that, as a first approximation, the results from the pressure-diameter and pressure-flow studies should be directly comparable. We tested the validity of this approximation by using the radiographically determined estimates of the increase in intrasegmental airway diameter with increases in either Pao or Ps to recalculate Re and Pn. Plots resulting from these recalculations are shown in Figs. 5 and 6. Figure 5A shows the Moody-type plot generated from the

527

IN PONY LUNGS

A76 -’

0” J4

-.

3 -.

2

!

If

I

II

I

II 1

II 2

II 3

I 4

87 T bt -

25 m

I

t

0

‘4

7 i

2

1 0

Log Re

FIG. 5. Pn as a function of Re at Pao = 510, and 15 cmH,O calculated using average flows at Ps - Pa0 = 6 cmH,O from 6 pony lungs. A: Re and Pn calculated assumin .g that d was independent of Pao. B: Re and Pn at Pao = 10 and 15 cmH,O calculated assuming that d increased in proportion to radiographically determined increase in small-airway diameter from Ps = Pao = 5 cmH,O to Ps = Pao = 10 cmH,O and from Ps = Pao = 5, cmH,O to Ps = Pao = 15 cmH,O.

average flow for all six lungs when Ps - Pao was 6 cmH,O and Pao was 5, 10, or 15 cmH,O, assuming that the cha racteristic dimension used to calculate Re and Pn did not change as Pao increased; i..e., d = 0.5 cm for each. Figure 5B shows the same data, where d was 0.5 for 5 cmH,O Pao only. From the normalized diameters in the homogeneous case, we estimated that d decreased by 12% at 10 cmH,O Pao and by 21% at 15 cmH,O Pao, and we recalculated Pn and Re accordingly. Both the horizontal and vertical separation between the three 1ines is reduced, consis tent with the interpretati .on that increasing lung volume at a constant Ps - Pao increased the dimensions of the flow-conducting pathways. A similar recalculation is shown in Fig. 6. Figure 6A represents calculations of Re and Pn from average flows of-Ne, He, and N, at 5 cmH,O Pao, for Ps - Pao of 6,9, and 15 cmH,O with the assumption that d = 0.5 cm. Figure 6B represents recalculations of Re and Pn made by estimating th e increase in d from the radiographically determined in .crease in n.orma lized diameters at 5 cmH,O Pao, Ps - Pao of 5, 10, and 15 cmH,O, which were 6.4, 9.3, and l4.1%, respectively. Note the reduction in the horizontal separation between the three lines, consistent with the interpretation from the Moody-type plot that, at 5 cmH,O Pao, the dimensions of the flow pathways increased as Ps - Pao increased. Although the recalculated data shown in Figs. 5 and 6 are by no means proof, they clearly show that the. interpretati .ons from the pressure -flow and pressure -dia .meter

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

528

AIRWAY

P-s-Pao= 0 15cm 9 cm 6 cm

2

I

I

I

MECHANICS

H-0 Hi0 H,O

I

R7-F

FIG. 6. calculated Re and Pn Pn calcula tally deter cmH,O to

data are consistent and lead to the conclusion that measurements of collateral resistance reflect the properties of structures that appear to behave like small airways. We emphasize, however, that the precise anatomic structures of collateral pathways in equine lungs have not been determined. A similar conclusion was reached in studies of dog lungs (10). Collateral resistance therefore seems to measure the properties of structures that behave like small airways in both dog and pony lungs. Yet the pressureflow data in pony lungs confirm the findings of Robinson and Sorenson (15) that the response to increases in Ps at 5 cmH,O Pao differs between canine and equin e lungs. We interpret these findings to suggest species differences in the mechanical properties of the airways or species differences in parenchymal interdependence. The shape of the pressure-diameter curve for pony airways resembled the curve for small airways in freshly excised dog lungs (7, l4), indicating the airways to be very compliant at low transpulmonary pressures. The observed difference between horse lungs and dog lungs in the response of collateral resistance to increased Ps Pao is therefore not likely to be due to differences in small airway compliance. The response of intrasegmental airways to increased Ps in the nonhomogeneous case has been postulated to depend on parenchymal interdependence as well as airway compliance (10). We assume that the diameter of the small airways reflects the degree of expansion of the segment, provided that the airways are operating on the

IN PONY LUNGS

steep part of the pressure-diameter curve, which limits the analysis to 5 cmH,O Pao (11). The finding that airway diameter is smaller for a given Ps in the nonhomogeneous case (Ps > Pao) than in the homogeneous case (Ps = Pao) can be interpreted to indicate that the segment was less expanded in the nonhomogeneous case than in the homogeneous case. Figure 4 shows that increasing Ps increased airway diameter in excised pony lungs at each lung volume studied. Furthermore, at Ps - Pao of 5, 10, or 15 cmH,O at both 10 and 15 cmH,O Pao, the normalized diameter of the small airways in the nonhomogeneous case was identical to the normalized diameter of the small airways in the homogeneous case, indicating that diameter was purely a function of Ps. However, in contrast to excised pig (11) and dog (10) lungs held at a lung volume corresponding to 5 cmH,O Pao, small airway diameters in pony lungs were smaller in the nonhomogeneous case (Ps > Pao) than in the homogeneous case (Ps = Pao) at equivalent Ps. We know that airway diameter is a function of the airway transmural pressure, which is the difference between airway lumen pressure and interstitial pressure. Furthermore, there is good evidence that interstitial pressure around small airways approximates pleural pressure in the homogeneous case (11). The finding of differences in diameter at equivalent Ps at 5 cmH,O Pao, i.e., diameter at 10 cmH,O Ps and 5 cmH,O Pao, and diameter at Ps = Pao = 10 cmH,O, could be interpreted to indicate that interstitial pressure was greater in the nonhomogeneous case than in the homogeneous case in pony lungs but not pig or dog lungs. A reasonable explanation for the increase in interstitial pressure would be that the surrounding lung prevented the segment from expanding to the same extent in the nonhomogeneous case as in the homogeneous case. The difference in small-airway pressure-diameter behavior between horse and dog lungs could therefore be explained if interdependence was greater in horse lungs. However, this would not explain the different response-of collateral resistance to increased Ps at 5 cmH,O Pao. In fact, if parenchymal interdependence was greater in pony lungs than in dog lungs, the opposite result would be predicted. If the segment in pony lungs failed to expand to the same extent as in dog lungs, a larger increase in collateral resistance as Ps - Pao increased would be predicted in pony lungs. A logical extension of this argument is that even though collateral resistance measures the properties of structures that behave like small airways in pony and dog lungs, the pathways are not identical. In conclusion, this study has confirmed that collateral resistance increases as collateral airflow is increased at high lung volumes in equine lungs (15). Additional results showed that the increase in collateral resistance with increasing Ps at a constant lung volume was likely due to a change in flow pattern rather than a decrease in the effective area for flow. This inference was confirmed by documenting that, at all lung volumes, small-airway diameter increased as Ps increased. Although these results suggest that collateral resistance measures the property of structures that behave like small airways, they

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.

AIRWAY

MECHANICS

IN

also emphasize the limitations of interpreting results on the basis of measurements of collateral resistance alone. We thank V. P. Wright and L. M. Tobin for expert technical assistance. This work was supported in part by National Heart, Lung, and Blood Institute Grant HL-37246 and a grant from the American Lung Association of Ohio. L. E. Olson is the recipient of a Research Career Development Award from the National Institutes of Health. Address for reprint requests: L. E. Olson, The Ohio State University, 309 Sisson Hall, 1900 Coffey Rd., Columbus, OH 43210-1092. Received

4 October

1991; accepted

in final

form

11 February

1992.

8.

10.

11.

12.

REFERENCES 1. BRAKER, W., AND A. L. MOSSMAN. Matheson Gas Data Book (6th ed.). Lyndhurst, NJ: Matheson, 1980, p. 344, 504, 522, and 649. 2. FULLER, S. D., AND N. E. ROBINSON. Effect of regional inhomogeneity on collateral airway resistance. J. Appl. Physiol. 57: 254-261, 1984. 3. FULLER, S. D., AND N. E. ROBINSON. Mechanism of increased collateral airway resistance during inhomogeneous inflation of excised dog lungs. Respir. Physiol. 74: 254-264, 1988. 4. HYATT, R. E., K. P. OFFORD, AND S. J. LAI-FOOK. Effect of length changes on bronchial diameters. J. Appl. PhysioZ. 50: 1168-1172, 1981. 5. KAISE, A., A. N. FREED, AND W. MITZNER. Interaction between CO2 concentration and flow rate on peripheral airway resistance. J. Appt. Physiol. 70: 2514-2521, 1991. 6. KAPLAN, J., R. C. KOEHLER, P. B. TERRY, H. A. MENKES, AND R. J. TRAYSTMAN. Effect of lung volume on collateral ventilation in the dog. J. AppZ. Physiol. 49: 9-15, 1980. 7. LAI-FOOK, S. J., R. E. HYATT, AND J. R. RODARTE. Effect of parenchymal shear modulus and lung volume on bronchial pressure-diameter behavior. J. AppZ. Physiol. 44: 859-868, 1978.

13 ’ l4 . l5

l6

17. 18. 19.

.

.

PONY

LUNGS

529

LEITH, D. E., AND J. R. GILLESPIE. Respiratory mechanics of normal horses and one with chronic obstructive lung disease (Abstract). Federation Proc. 30: A556, 1971. OLSON, L. E. Effect of transpulmonary and driving pressures on collateral gas flow in dog lungs. J. AppZ. Physiol. 59: 1757-1765, 1985. OLSON, L. E. Small airway pressure-diameter relationships during nonhomogeneous lung lobe inflation. J. AppZ. Physiol. 62: 23772382,1987. OLSON, L. E. Regional compliance and bronchial pressure-diameter relationships in excised pig lungs. Respir. Physiol. 86: 25-39, 1991. OLSON, L. E., J. R. RODARTE, AND N. E. ROBINSON. Pressure-flow relationships in a collaterally ventilating dog lung segment. J. AppZ. Physiol. 54: 956-960, 1983. OLSON, L. E., AND P. A. SOCHA. Effect of flow direction on collateral ventilation in excised dog lung lobes. J. AppZ. Physiol. 60: 770776,1986. OLSON, L. E., L. M. TOBIN, AND V. P. WRIGHT. Effect of bronchomotor tone on the response of small airways to unequal alveolar pressure (Abstract). FASEB J. 3: Al300, 1989. ROBINSON, N. E., AND P. R. SORENSON. Collateral flow resistance and time constants in dog and horse lungs. J. AppZ. Physiol. 44: 63-68,1978. SNEDDON, S. L., AND J. D. BRAIN. Steady expiratory flow in dog lungs: an isovolume preparation. J. AppZ. Physiol. 51: 1331-1337, 1981. SOKOL, R. R., AND F. J. ROHLF. Biometry. San Francisco, CA: Freeman, 1969, p. 222-399. SORENSON, P. R., AND N. E. ROBINSON. Postural effects on lung volumes and asynchronous ventilation in anesthetized horses. J. AppZ. Physiol. 48: 97-103, 1980. WINER, B. J. StatisticaL PrincipZes in Experimental Design. New York: McGraw-Hill, 1971, p. 262-300.

Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on January 11, 2019.