Structural changes underlying compensatory increase of diffusing capacity after left pneumonectomy in adult dogs

Amendment history:

Correction (February 1994)

Structural changes underlying compensatory

increase of diffusing capacity after left

pneumonectomy in adult dogs.

C C Hsia, … , R C Reynolds, E R Weibel

J Clin Invest. 1993;92(2):758-764. https://doi.org/10.1172/JCI116647.

To determine if the functional compensation in diffusing capacity of the remaining lung following pneumonectomy is due to structural growth, we performed morphometric analysis of the right lung in three adult foxhounds approximately 2 yr after left pneumonectomy (removal of 42% of lung) and compared the results to those in normal adult dogs previously studied by the same techniques. Diffusing capacity was calculated by an established morphometric model and compared to physiologic estimates at peak exercise in the same dogs after pneumonectomy. The major structural changes after left pneumonectomy are hyperinflation of the right lung, alveolar enlargement, and thinning of the alveolar-capillary tissue barrier. These changes confer significant functional compensation for gas exchange by reducing the overall resistance to O2 diffusion. The magnitude of compensation in diffusing capacity estimated either morphometrically or physiologically is similar. In spite of morphometric and physiologic evidence of functional compensation, there is no evidence of significant growth of structural components. After pneumonectomy, morphometric estimates of diffusing capacity are on average 23% higher than physiologic estimates in the same dogs at peak exercise. We conclude that the previously reported large differences between morphometric and physiologic estimates […]

Research Article

Find the latest version:

Structural

Changes Underlying Compensatory

Increase of Diffusing

Capacity

after Left Pneumonectomy in Adult

Dogs

C.C.W.Hsia,* F. Fryder-Doffey,t V.Staider-Navarro,*R.L. Johnson, Jr.,* R. C.Reynolds,* andE.R.Weibel*

*Departments ofInternalMedicine and Pathology, UniversityofTexasSouthwesternMedicalCenter,Dallas, Texas 75235-9034; and

tInstitute ofAnatomy, University ofBern, CH-3000,Bern, Switzerland

Abstract

To determine if thefunctional compensationin diffusing capac-ity of theremaininglung following pneumonectomy is due to structural growth, we performed morphometric analysis of the right lung in three adultfoxhounds -2yr after left

pneumonec-tomy(removal of 42% of lung) and compared the results to those in normal adult dogs previously studied by the same tech-niques.Diffusingcapacity wascalculatedby an established

mor-phometric model and compared to physiologic estimates at peakexerciseinthe same dogs after pneumonectomy. The ma-jor structural changes after leftpneumonectomyare hyperinfla-tion of therightlung, alveolar enlargement, and thinning of the

alveolar-capillarytissue barrier. These changesconfer

signifi-cantfunctional compensationfor gas exchange by reducing the overall resistanceto

02

diffusion.Themagnitude of compensa-tion indiffusing capacityestimated eithermorphometricallyor

physiologicallyis similar.Inspite of morphometric and

physio-logic evidenceoffunctional compensation,thereisnoevidence ofsignificant growth ofstructural components. After pneumo-nectomy,morphometricestimates ofdiffusingcapacity are on average 23% higher than physiologic estimates inthe same

dogs

at peak exercise. We conclude that the previously

re-ported large differences between morphometric and

physio-logic estimates ofdiffusing capacity reflects the presence of largephysiologicreserves available for recruitment.(J. Clin.

Invest. 1993. 92:758-764.) Key words: morphometry *

re-breathing*

lung

resection*exercise* recruitment

Introduction

Leftpneumonectomyremoves 42% ofthepulmonarygas

ex-changeunits(1,2)and thuswouldbeexpectedto

significantly

reducemaximaloxygenuptake.This isparticularly important

indogs that appeartohaveasmallamountofredundancyin

theirgasexchangerformeetingoxygen demandsatheavy

exer-cise(3). However, in

previous

studies on dogs,wehave oW

servedthatexercise

performance

isnot

significantly impaired

by left pneumonectomy

(4). Diffusing

capacity

for carbon

Parts of this work have beenpublishedin abstract form(1991.Clin. Res.39:226A).

Address correspondencetoConnie C. W.Hsia,M.D., Division of Pulmonary Research, Department of InternalMedicine, Universityof Texas Southwestern Medical Center, 5323 HarryHines Boulevard, Dallas, TX 75235-9034.

Received for publication20October1992 andinrevisedform26 January 1993.

monoxide

(DLco)'

measuredphysiologicallyat peak exercise is reduced by only 23% after leftpneumonectomycompared with before,suggestingthat

_DLco

of theremaininglung must have increased by some 30% above that ofthe same lung before

pneumonectomy (5). Two general sources ofcompensation

areavailableto theremaininglung that could haveaccounted for the increase in

DLcO

after pneumonectomy: (a) greater

utilization of existing physiologicalreserves of

DLcO;

and (b)

growth ofnew gas exchange tissue.

Diffusing capacity ofthelung may be estimated

physiologi-callyby therateofdisappearance ofcarbon monoxide from the

inspiredgas by the single breath or therebreathing technique.

Large reservesof diffusingcapacity normallyexist in the lung and can be usedduringexercise aspulmonaryblood flow and lung volume increase (5). Theanatomicalbasis ofrecruitment

ofdiffusing capacity has not been fully defined, but is pre-sumedtobe due to theunfolding of alveolar-capillary mem-brane(6),openingand/or thedistentionof pulmonary capil-laries (7), and an increase inhematocrit(8). Diffusing capac-ity of the lung may also be estimated from morphometric

measurementsofthe total

alveolar-capillary

surface area, the

effective thickness ofthediffusion barrier,andthe pulmonary

capillaryred cell volume in the lung.Assumingthat these diffu-sion barriersarearrangedinseries,theseanatomicaldata have been used to give an estimate ofdiffusing capacity for

02

(DL02) and forcarbon monoxide

(DLco)

(9). The question

ariseswhetherpostpneumonectomy functional compensation

is solely due to physiological adjustments or whether it also involves structuraladaptation; the latter may include altered

characteristics of existinglung tissue or theaddition ofnew

lung tissue. Theobjectiveof our report is todeterminewhether andtowhatextentfunctionalcompensationofdiffusing capac-ity in adult dogsafterleftpneumonectomyhasbeen achieved bystructural adjustmentsin theremaining lung.If the normal reservesofdiffusing

capacity

inonelung is fully recruited after pneumonectomy,

physiological

estimates of

DLcO

atpeak

ex-erciseshould

approximate

morphometric

estimatein thesame

animal.Theextentofpostpneumonectomy

compensation

esti-matedby either technique shouldbesimilar.

Methods

Animals. Extensive physiologic studieswereperformed in three adult male foxhounds(bodywt22-24kg)before and after left pneu-monectomy. These results have beenreportedindetailpreviously(4,

1. Abbreviationsusedin this paper:

DLco,

diffusingcapacityfor

car-bonmonoxide;DLO2, diffusingcapacityfor oxygen;V,,, pulmonary

capillarybloodvolume;

V1,

tissuevolume; Vv(a, s),volumedensityof alveoliin septum;Vv(c, s),volumedensityof_capillariesin septum; Vv(cp,L),volumedensityofcoarseparenchymainlung;Vv_{(fp, cp),}

volumedensityof fineparenchymaincoarseparenchyma; Vv(s, fp),

volumedensityof alveolar septa in fineparenchyma.

J. Clin.Invest.

© The AmericanSocietyfor ClinicalInvestigation,Inc.

0021-9738/93/08/0758/07 $2.00

5, 10-13). Dogs were trained to run on a treadmill up to maximal voluntary effort. Exercise training began before pneumonectomy, re-sumed 14 dafterleftpneumonectomy, andcontinued until killing. Physiologic studies at exercise were carried out beforepneumonectomy andbeginning at 4-6 mo after surgery; these consisted of measuments of ventilation, gas exchange including ventilation-perfusion re-lations, hemodynamics, and dynamic work ofbreathing at various lev-elsofsteady state exercise.DLcowasmeasured at different intensities ofsteady stateexerciseby the rebreathing technique described previ-ously indetail( 5, 10).After the completion ofphysiologic studies ( - 2

yr afterpneumonectomy),theanimals were killed and morphologic analysis oftheremainingright lungs was carried out.

Lungfixation. The animal was deeply anesthetized with pentobar-bital (25 mg/kgintravenously), intubated via a tracheostomy, and placed in the supine position. The abdomen was opened through a midlineincision and a small rent made in the diaphragm to collapse the rightlung. An overdoseofpentobarbital was given and the lung imme-diately reinflated withinthe intact thorax byintratrachealinstillation of 2.5% glutaraldehyde buffered with potassium phosphate to pH 7.40 andosmolarity 350 mOsm at a constant hydrostatic pressure of 25 cm H20above the highest point of the sternum ( 14). After the flow of fixativesinto the lung had stopped, the endotracheal tube was tightly clamped. After 60 minoffixation in situ, the thorax was opened; the lung and heart were removed en bloc and completelysubmerged in

10%bufferedformaldehyde.

Sampling procedures. The previously established four-level strati-fied samplingscheme(15) wasused:(I) gross,(II) low power light microscopic, (III)high power light microscopic, and (IV) electron mi-croscopic analysis.Thelungs were stratifiedaccordingtolobe. Volume ofeach lobe wasmeasuredbywaterdisplacement.Each lobe was ex-haustivelysliced at 2-cmintervals.Eachslice was photographed using 35-mm Ektachrome color film toestimatethe volumedensity ofcoarse parenchymain lung,Vv(cp,L),byquantifyingallstructures

measur-ing> 1mm(levelI).Six blocksof tissueweretaken from each lobeby

a systematic, volume-weighted sampling procedure with a random start( 15). Samples fromeach block were embedded in methacrylate forthick sections(5gm)stainedwithhematoxylinandeosinto esti-mate thevolumedensityoffineparenchyma in coarse parenchyma, Vv(fp,cp),byquantifyingallstructures measuring between_20,umand 1 mm(levelII,X200). Additional samplesfrom each blockwere

em-bedded inEpon.Thesewereused forpreparationof semi-thin sections (1jIm)to estimate the volumedensityof alveolar septa in fine paren-chyma bylight microscopy, Vv(s, fp), by quantifyingstructures

mea-suring<20_{,m (level}III,X400),andtoestimate volume and surface densities ofalveolarstructuresin septum, e.g.,capillaries Vv(c,s), al-veoli Vv(a,s),as well as harmonicmeanthicknessesofplasmaand tissuediffusion barriersbyelectronmicroscopy (level IV,x- 11,000).

Detailofthe methods have beendescribedpreviously (15).

Morphometric analysis.Slidesateach level wereplacedunder

stan-dardcountinggrids.Volumedensityofalveolarstructures were deter-minedby pointcounting.Surfacedensityofalveoli andcapillarieswere

determined by intersectioncounting.Allmorphometricdatawere cal-culated foreach lobeseparately; avolume-weighted average for the entire lungwasthencalculated. Absolute volume and surfaceareaof individual alveolarstructures wereobtainedby relatingtherespective

volume and surfacedensitiesateach level backthroughthe cascade of levels to the measured volumeofthe lobe(15).

Morphometric diffusingcapacityfor 02 (DL02). DL02was esti-matedby the modelpreviouslydescribed(9).The model describes the gas diffusionpathfrom alveolar airtothebindingsitesonhemoglobin

asthreeseriallylinked conductance stepsthroughthe tissue_(DtO2),the plasma(Dp02),and theerythrocyte(DeO2):

DL02- =DtO2 1+Dp021+DeO2

where

(S+S2

S)

Dt02=Kt02*(2.rht)

(Eq. 1)

DPo2

=

KP02

SC

Thp

De02

=

_0)02.V

(Eq. 3)

(Eq.4)

SA

and

_S,

arethe measured total alveolar and

_capillary

sur-facearea,andVCthe measured total capillary blood volume.Tht and Thp are the harmonic mean thicknesses of the tissue and plasma barriers, respectively.

Kt02

and

_Kpo2

are the Krogh diffusion coefficients for 02 in tissueand plasma,respectively; these were taken from literature and were the same as used previously ( 14).002is thereaction rate of whole blood with 02 (inml-(ml*mmHg*

s)f').

For the purpose ofcomparison,we

used the same value of 002 as used by Weibel et al. ( 14) in normal dogs, calculated from the equation:

002

=

K6C(%)

f(T)

*

(0-0587

*

aO2)*(

I -

S02)

X (0.01333.[Hb]) (Eq. 5)

whereK',

%)is

the red cell reaction velocity at 60%saturation.

The value of

_{K'w%)measured}

by Holland et al. (16) is 220 mM' * s'; however, more recent data have shown that the presence of an unstirred layer of plasma surroundingthered cell in the stop-flowtechnique may cause underestimation of theinitial reaction rate by a factor of - 2 (17, 18).Therefore, we used a value of K'

60%)

= 440 mM * s-l to take into ac-count this effect. f(T) is the temperature factor derived from the Arrhenius equation that corrects Kc from the standard 370C to the core temperature measured at peak exercise work-load.

aO2

is the solubility of 02 at the core temperature during peak exercise workload.

S02

isthe initial fractionalsaturation

of 02. [Hb] is the hemoglobin concentration in g/dl of blood.

Morphometric

diffusingcapacityfor

CO

_(DLco).

_DLCo

was calculated using the same model described above (Eqs. 1-4) substituting CO for 02. Again, for the purpose of comparison with previous data, we used the same

_0co

as in the study by

Weibelet al.(19), calculatedfrom the equation given by

Hol-land (20,21)fordog blood at 40'C, the core temperature ex-pected of exercising dogs:

l =(0.929+ OoO379P~o2)

[14.4

0co

=

(0.929 +0.00379PA02) _[Hb (Eq. 6)

"COhas the units of [ml.(ml-mmHg

min)f'. PA02

isthe

meanalveolar 02 tensiontaken to be100mmHg. [Hb] is taken tobe 15 g/dl.

Dataanalysis. Since physiologic studies wereperformed before and after pneumonectomyonthesamedogs,no

simulta-neous controls were available for morphometric measure-ments. Therefore, we compared morphometric results with

publisheddataby Weibelet al. ( 14)inthreenormalmongrel

dogsusingthesame sampling and morphometric techniques by one-tailed unpairedttest.Physiologic-morphometric

com-parison of

DLco

wassimilarly relatedto thoseof Weibeletal.

( 19)indifferent species ofnormalcanids.A Pvalueof<0.05

wasconsidered significant.

Results

Morphometric

measurements areshown in Table I. Volume

compensation

was

essentially

complete; the volume ofone

TableI.ComparisonofMorphometric Data inRightLung ofDogsafterLeft Pneumonectomywith That inBothLungsofNormalDogs

Pvalues

Dogsafter left Normal Unpaired

pneumonectomy dogs ttest

Number 3 3

Body mass (kg) 23.8±0.6 28.2±0.7 0.004*

Totallung volume (ml *kg-') 53.4±4.4 56.3±1.9 0.29

Right lung volume (ml-kg-') 53.4±4.4 32.7±1.1 0.005*

Morphometrichematocrit (%) 43.5±3.3 55.0± 1.5 0.02*

Volume density

Parenchymain lung Vv(fp,L) 0.850±0.001 0.819±0.002 0.0001*

Septa in parenchyma Vv(s,fp) 0.106±0.005 0.157±0.008 0.0009*

Capillariesin lung Vv(c,L) 0.058±0.005 0.074±0.002 0.02*

Tissue in lung Vv(t,L) 0.032±0.0003 0.054±0.002 0.0001*

Surface density

Alveoli in lung Sv(a,L) cm-' 343.8±2.3 488.5±10.4 0.0001*

Capillariesinlung Sv(c,L)cm-' 292.7±22.7 385.0±6.6 0.009*

Arithmeticmeanthickness(l0-'cm)

Tissuebarrier 1.02±0.04 1.24±0.02 0.005*

Septum 2.61±0.13 2.63±0.05 0.45

Harmonicmeanthickness (10-4cm)

Tissuebarrier (Tht) 0.34 1±0.014 0.509±0.019 0.00 *

Plasma barrier(Thp) 0.155±0.004 0.117±0.001 0.005*

Volumes (ml *kg-')

Capillaries Total 3.08±0.29 4.18±0.18 0.02*

Right lung 3.08±0.29 2.43±0.10 0.05

SeptumTotal 4.791±0.425 7.238±0.351 0.005*

Right lung 4.791±0.425 4.198±0.203 0.14

Tissue Total 1.711±0.159 3.055±0.179 0.003*

Right lung 1.711±0.159 1.772±0.104 0.38

Surfacearea(M2.kg-')

Alveoli Total 1.836±0.142 2.756±0.145 0.005*

Right lung 1.836±0.142 1.598±0.084 0.11

Capillaries Total 1.564±0.174 2.167±0.069 0.02*

Right lung 1.564±0.174 1.257±0.040 0.08

Resultsaremean±SEM. *Significantat0.05 level.

pneumonectomized dogs.

Volumedensitiesof all

_septal

struc-tureswere

significantly

lowerin

pneumonectomized

animals than in

_controls,

_indicating

alveolar

_{hyperinflation,}

_stretching

and

_thinning

of the alveolarseptumwhichisevidentunder low

power

magnification (Fig.

1).

Surface densities of alveoliand

capillaries

were

significantly

lower than normal. Arithmetic meanthickness of the tissue barrierwas

significantly

lowerin

pneumonectomized

dogs

but themeanthicknessofthe septum

wassimilartothat in controls.Theharmonicmeanthickness of thetissue barrierwasalso lowerin

_{pneumonectomized}

ani-mals thanin controls. Harmonicmeanthickness ofthe

_plasma

barrierwas_greaterin

pneumonectomized

animals

owing

tothe

lower hematocrit.In

_{pneumonectomized}

_dogs,

total volumes

ofthe

_capillary

blood and the _septum were

_{significantly}

re-ducedto74and66%of

_normal,

_{respectively;}

total

_septal

tissue

volumewasreducedto 56% ofnormal. Totalsurfaceareaof

thealveoliand

_capillaries

werealso

significantly

reducedto66 and 72% of

_normal,

_{respectively.}

There were no

significant

differencesin

_morphometric

_{parametersamong lobes.}

Estimations of

DL02

and DLcoareshown inTableII.Inthe

pneumonectomizedgroup,

DtO2

and

DM02

were not

signifi-cantlydifferent(88%)from that oftwolungsin normal_dogs.

De02wassignificantlylower duetoalower Vc.DL02was lower

(81%) than normal at a borderline significance (P = 0.07). Similarchangesareseeninthecomponentsof

DLco.

Theright lungin thedogconstitutesonaverage 58% of total

lungvolume andweight,and receivesasimilarpercentage of totalventilation and blood flow( 1, 2, 22, 23). Ourresults in theright lungofdogsafter pneumonectomy maybecompared

tothat intheright lungofnormaldogs by multiplying resultsin normaldogs byafactor of 0.58. This comparison is shownin

Fig.

2.

Compared

with the expected value inthe normal right

lung, morphometricparameters (Fig. 2 a) including surface

areaof thealveoliand capillaries,andvolumes ofcapillaries andseptum,wereonaverage15-26%higher after

pneumonec-tomyandreached borderlinestatisticalsignificance(Table I). Tissue volume(Vt)was notdifferentfromthat in the normal

right lung. Similar comparisons ofcomponents of diffusing

capacityof theright lungshowed that

DeO2, DMO2,

and

DLO2

weresignificantlyhigherin postpneumonectomy dogs (Fig. 2

b);

Deco,

Dmco,

and DLcowerealsosignificantly higher in

,

a%s,' %

/~

.e

I

J.

6.ww4,...

_{e 40.}

w %

,I-.

-%.

(}

41k%

~

~

.

:',

<.

+

~

FA.,,

icp

(

)ogh

.

p'

Figure 1. Photomicrograph(x 100) of lungparenchymain

normnal

dog(left panel) and in a dog afterleft pneumonectomy(rightpanel).

Physiologic

_DLco

measuredby therebreathing technique

inthesedogs before and after leftpneumonectomyhave been previously published (5, 12). A linear relationship between

DLo

and

02

uptake and between

DLco

andpulmonaryblood

TableII. Components ofDiffusingCapacityCalculatedfrom

MorphometricMeasurements inthe RightLungofDogsafter Left Pneumonectomy andinBoth LungsofNormal Dogs

P value

Dogs after left Normal Unpaired

pneumonectomy dogs ttest

Partial conductances for02 [mI 02*(S* mmHg*kg)-']

Tissue(DtO2)* 0.280±0.037 0.267±0.021 0.39 Plasma(Dpo2)* 0.568±0.072 1.030±0.072 0.005t Membrane (DM02) 0.187±0.024 0.212±0.014 0.21 Erythrocyte(DeO2) 0.197±0.019 0.268±0.011 0.02*

Lung (DL02) 0.095±0.011 0.118±0.007 0.07 Partial conductancesfor CO

[mlCO *(s- mmHg- kg)-'

Tissue(Dtco)"I 0.225±0.030 0.217±0.017 0.41 Plasma(Dpco)II 0.454±0.058 0.837±0.058 0.005*

Membrane(Dmco) 0.150±0.019 0.172±0.011 0.20 Erythrocyte(Deco)' 0.041±0.004 0.056±0.002 0.02*

Lung(DLco) 0.032±0.003 0.042±0.002 0.03*

Results aremean±SEM. *_KtO2 andKp02 = 5.5 x lO-10

cm *s *mmHg . *_Significantat0.05 level. §002 = 0.0641

ml-(ml * mmHg-s)-'.1. KtcoandKpco=4.467 x l0-'°

cm ._s ._mmHg'.

'Oco

=0.0 133 ml *(ml-

mmHg.

s)'.

flow is evident in each dog. Thehighest

DLo

obtainedin each

dogatexercise before andafter pneumonectomyareshown in

Fig. 3 incomparisonwithmorphometricestimates from

nor-malcanids ( 14, 19). Inthepresentdogs, theratio of physio-logic DLcoafterpneumonectomy tothat in thesameright lung

beforepneumonectomyis 1.28, 1.21, and 1.30 in each of the

threedogs, withanaverageratio of 1.27. The ratio of

morpho-metricDLcoinpostpneumonectomydogstothatin theright

lungofnormaldogs is also 1.27.Fig.4 shows theDLco

esti-mated postpneumonectomybyrebreathingatexercise andby

morphometry postmortem in each of the threedogs. On the

average,morphometricestimatesare23%higherthan

physio-logic estimates inthe samedog.

Discussion

Summary

of findings.

This is the first study

correlating

morphometriccompensation withphysiologic

compensation

atheavy exercise in thesame dogsafterextensive lung

resec-tion.At two yearsafter leftpneumonectomy,

equivalent

tothe

removal of about 42% of lung, the major

morphological

changesarehyperinflationoftheremaining lung, enlargement

ofthealveolarairspaces,andthinningofthe

alveolar-capillary

tissuebarrier. Thesechangesconfersignificantfunctional

com-pensation forgasexchangebyreducingtheoverall resistance for02diffusion. Postpneumonectomycompensatoryincrease indiffusingcapacitywasthesame (27%)when estimatedby

physiologic andmorphometric techniques. Inspiteof

signifi-cant morphometric and physiologic evidence for functional

[ai

IE

£

10%

.~~~~~~~~~H

0

=₁

Sa c c s Vt_ .l-=_:

Sa SC VcC vs Vt a

[l1

60

40

o Morphometry,NormalCanids * Rebreathing, Prepnx

* Rebreathing,Postpnx

0

0 0

.

0 10 20 30 40

Body Weight (kg)

Figure 3.Comparisonofmorphometric

DLcO

of the right lung in four speciesof normal canids ( 14, 19)(open circlesandleast squares re-gression line) with physiologic

DLcO

at peak exercise in foxhounds before (prepnx, dark squares)and afterpneumonectomy (postpnx, dark circles)(5).

gives

anestimate of thetruestructural frameworkof thelung

DeO2 DpO2 DtO2 DmO2 DLO2 unaffected by any mechanical instability (6).Tracheal

instilla-tion alsofixes the red cellsin perfused capillariesand allows

direct estimation oftheharmonicmeanthickness oftheentire diffusionpath, including theplasma barrierwhich cannotbe

FC] measured after perfusion fixation. Since physiologic

measure-mentswere obtained in these foxhounds both before and after

.111I11

leftpneumonectomy, no simultaneous control animals were available formorphometric analysis. However, thepublished

control data used (14) were obtained using the same

tech-niques of fixation, sampling, and analysis asin the present report. In different species ofcanids, Weibel et al. (14, 19)

C

0

UL

0

DeCO DpCO DtCO DmCO DLCO

Figure2.Comparisonofmorphometricestimatein theright lungof

dogsafter pneumonectomyexpressedas afraction ofexpectednormal value in theright lungofnormaldogs. (A) Sa, Sc,surfaceareaof alveoliandcapillaries,respectively. Vc, Vs, Vt,volume ofcapillaries,

septum,andtissue,respectively. (B) Componentsofdiffusing

capac-ityfor

02.

(C) Componentsofdiffusingcapacityfor CO. *Signifi-cantlydifferent fromunity.

100

-E

._

E

0

C.

-1

0

pulmonary structuralcomponents.Thereisagood correlation

betweenDLcoestimated bymorphometricandbyphysiologic techniquesatpeak exercise in normaldogs and indogsafter pneumonectomy. MorphometricDLcois - 23%b_higherthan

physiologic

DLIo

atpeakexerciseinthesameanimals.

Critique of methods. Tracheal instillation of fixatives was

chosen over perfusion fixation because the former leads to

complete unfolding of the alveolar capillary membrane and

80

-60

-40

-20

-

0-After Left Pneumonectomy

El Rebreathing

M Morphometry

0

-_ , I | ROES , Cog

I

|

|

NOB

I

-

_

.1._

-22.6

_

_

_

Rub _ REAR RIMS RIMS MM _

_

OR

"

24.1

| #OR _

I #I

RIMS* ,

_

s

I _

_ _ A= _

En

_

x>

-I

24.6

Body Weight (kg)

Figure4. Comparisonofphysiologic DLcOatpeakexercise(5)and

morphometricDLcointhesamefoxhounds after pneumonectomy.

762 Hsiaetal.

21

E

0

z

0

0

a-V.

2-'RE

0

z

0

r_

0 U.

U-

1-0*

21-E

0

z

0

1

found thatdifferencesinmorphometric estimatesaredirectly relatedtobody weight. Thuscomparisonsbetween foxhounds

andmongrel dogs should be valid when resultsare normalized

bybodyweight, particularly since differences in body weight between these two speciesare small. Thetrue value of 0

re-mainscontroversial. 0for CO has been measured byseveral independent methods, yielding similar values inmanand dog (21,24-26). Thusvariabilityinmeasurementsof

%co

issmall.

On the other hand, the in vitro measurement of0I for02 iS moresusceptibleto artifact than

0co

becausereaction veloci-tiesfor 02arefaster. However, errorsin theestimation of 002 does notaffect the interpretation ofourdatasince the same

value of 002 was used in the two groups. After completion of measurementsofdiffusing capacity, these dogs underwent im-plantationofaleft atrial catheter for themeasurementof

respi-ratory muscle bloodflow 6 wkbefore

killing (

13

).

Anemia developedas aresultof chronic catheterimplantation; arterial hematocrit at rest was 43.8±1.4% before and 37.3±1.8% (mean±SE) after catheterimplantation.This anemia contrib-utedto aslightlylower plasmacomponentofdiffusingcapacity

(DPO2

and

_Dpco).

However, allphysiologicmeasurementsof

diffusing capacity were expressed at a uniform hemoglobin concentration of 14.9 g/dl, closetothat used inmorphometric

calculations.

Compensatory lung growthinadultdogs after

pneumonec-tomy. Significantcompensatorygrowthofgasexchange struc-turesshould result inhighersurfaceareasofalveoliand capillar-ies,aswell ashigher volumes of septum,tissue,andcapillaries

thanin the normalright lung.Asshown inFig.2inourdogs after pneumonectomy, theseparametersare_{in general slightly}

higher than normal,reachingborderline statisticalsignificance.

There is no increase in lung tissue volume. The modest

in-crease in surfaceareaand volume ofcapillaries is consistent with either openingordistention ofexisting capillarybed by thediversion of the entire cardiac outputthroughthe remain-ing lungafterpneumonectomy, but could alsoreflectthe for-mation ofnew capillary loops when the septa become

ex-panded. Thus thepresentstudysupportstheconclusion from previous studies ofalimitedstructural compensation in dogs after left pneumonectomy asadults (27, 28). Inaddition,we

demonstratethefunctional consequenceofpassivestructural

adaptationtolung resection. Enlargement ofexistingalveolar

spacesandthinningof the alveolarseptaresult inalower har-monicmeanthicknessof thetissuebarrier

(Tht)

and a

signifi-cantincreasein thegasconductance ofthetissue barrier

(DtO2)

andof the membrane barrier(DM02).Openingordistention ofthepulmonarycapillariesresultsinasignificant increase in

gasconductance oftheerythrocyte component

(DeO2).

Hence the apparentmorphometricdiffusing capacityoftheremaining lung isincreasedbyan averageof27%, a value similar to the averagecompensatoryincrease obtainedby physiologic

meth-odsatpeakexercise inthe sameanimals.

Comparison ofmorphometric and physiologic

DLco.

Wei-bel etal.(19), using airway instillationoffixatives, compared

DLco

estimated by morphometry and by the single breath

method at rest in various species of canids; morphometric

DLco

wasfoundtobeapproximatelytwice that ofphysiologic

DLcointhe same animal. Crapo et al. (29), also usingairway instillation of fixative,compared

_DLco

bymorphometry and

bytherebreathing techniqueatrest in the same dogs;

morpho-metric estimatesare

nearly

threetimes

higher

than

_physiologic

estimates.The

_large

difference between

_techniques

_persists

in

lungsfixedbytheperfusion technique (24),andappearstobe dueto adifferenceinthe estimate ofmembranediffusing

capac-ity

(Dmco)

which is nearly 10times larger by

morphometry

than

by

therebreathingmethod.By morphometric technique,

Dmco

hasbeenestimatedtobe 10-fold

larger

than

Dmco

estimatedbytheRoughton-Forstermethodin thedogat rest

(29, 30).Estimatesofpulmonary capillaryblood volume

(Vc)

bythe twotechniquesatrest areabout the same. Incontrast,

when we studied the present foxhoundsbefore pneumonec-tomy,physiologic DLcoincreasedbymorethanthreefoldfrom rest topeakexercise and reached the normalrangeof

morpho-metric estimates (Fig. 3). After pneumonectomy, DLco by

morphometry is on average only 23% higherthan that esti-matedphysiologicallyin the sameanimal,acloseragreement

thananypreviouscomparison.

Onepossible explanation for thediscrepancyin

compari-sonsofDLcoandDmcobythesetechniquesis the existence of

physiological reserves ofdiffusing capacitythat are not

nor-mallyused atrest,butmaybemobilizedfor02transportupon exercise and after pneumonectomy. Morphometric estimates using intratracheal instillationoffixativescausestheunfolding

ofalveolar-capillary membranes(6);allcapillaries,including

those devoid of red cells, are included in the calculation of

diffusing capacity. Hence morphometric estimates more

closelyreflect the structuralcapacitywhere all alveoli and

capil-laries areavailableforgas exchange.Onthe other hand, invivo

videomicroscopic observations by Wagner and colleagues (7)

indicate that there is aprogressive increasein thenumberof

perfused capillary segments as pulmonary pressure is in-creased.Wagner'sdatasuggestthatcapillariesarenotfully re-cruited under normal resting conditions. After pneumonec-tomy,physiologicreservesofdiffusing capacityin the

remain-inglungmustbe used since the entire cardiacoutput must now flowthroughonelungatany given work load.Thus thelimitof

structural capacityfor gasexchange in the remaining lung may

bemorereadilyapproached during heavy exercise. In previous publications (5)weshowed that physiologic diffusing capacity increasestwo- tothreefold fromrest toheavy exercise,achange

similar inmagnitudeto thedifference between morphometric

and resting physiologic measurements. Our interpretation is

thatmorphometric

DLco

reflects the maximum structural

ca-pacity forgasexchange,alimitnotphysiologicallyapproached

exceptperhaps duringvery heavy exercise (3, 5). Hence the

difference between morphometric

DLco

andphysiologic

DLco

at rest representsfunctionalreservespotentially availablefor

recruitmentuponexercise. After pneumonectomy,

DLco

con-tinues to increase beyond that achieved before

pneumonec-tomy as a resultofahigher pulmonary blood flow through the remaining lung at any given workload. Although no clear

plateau in the linear rise of

DLco

isobserved up to peak exer-ciseequivalentto acardiac output of 35 liters. min-' through

twolungs,functionalreservesin the remaininglung are insuffi-cientto meetthedemands for 02 transport, evidencedby

arte-rial 02 desaturation dueto diffusion limitationat moderate andheavy workloads after pneumonectomy(4).Furthermore, thepatternofdecline in arterial 02 saturationafter

pneumon-ectoiny could be predicted from physiologically measured

DLco

(12). The severity of arterial02desaturationis mitigated

byaconcomitant reductionin maximal cardiac output.Under theseconditions,acloseragreementbetweenphysiologicaland

morphometric

estimates of

DLcO

canbeexpected.

mayoverestimate true

_Dmco.

So calledresistance ofthe mem-brane to CO uptake (1

_/Dmco)

may be conceptualized as the

sumof multiple resistances in series, i.e., resistances of the gas phase, the surfactant layer, the tissue, and plasma. The present morphometric model includes only the last two components; physiologic measurements include all. Ifdiffusion resistance in the gas phase is significant, overexpansion of the lung after pneumonectomy should increase gas phase resistance and par-tially offset the apparentcompensatory increasein

DLo

due to thinning of the tissue barrier and to capillary recruitment.In

addition, the gas diffusion coefficients ofcompressed or homog-enized lung used to estimatemorphometric

Dmco

(31 ) may be

significantlygreaterthan thoseofinvivo alveolar membranes in which surfactant monolayers remain intact; surfactant monolayers at an air-surface interface may potentially increase theresistance to gas permeation. The present study was not

designedto clarify the contribution of these opposing influ-ences.

Inconclusion, we report that in adult dogs after left pneu-monectomy, compensation by tissue growth is quite limited. However, adjustmentsin existing structures of the remaining lung, including hyperinflation, thinning of the tissue barrier, andcapillary recruitment after pneumonectomyresult in a sig-nificant compensatory increase in diffusing capacity which

con-tributes significantly tothe overall functional compensation.

The degree offunctional compensation estimated by either

morphometricorphysiologic technique is similar.Wefound a closer agreement betweenestimates of

DLco

by morphometric andbyphysiologicmethodsthanpreviouslyreported, possibly becauseourphysiologicalmeasurements wereperformed dur-ing heavy exercise when nearly all of existdur-ing functional re-serves areexploited.Thelargediscrepancybetween these

tech-niquesreported inearlier studies is at least partly due to the

failuretotakeintoaccounttheexistence of functionalreserves indiffusingcapacity.

Acknowledgment

Wewishtodedicate thismanuscripttothe memoryofDr.RollandC. Reynolds; hiskindness,support, andfriendshiparedearlymissed. We also expressourgratitudetoAnnaSiler, DavidTreakle,andBarbara Kalley-Taylor for their technical assistance.

ThisprojectwassupportedbytheNationalHeartLungandBlood Institute(HL-40070)andbytheSwissNationalScienceFoundation (3.172.88).

References

1. Rahn, H.,P.Sadoul,L. E.Farhi,and J.Shapiro. 1956. Distributionof ventilation andperfusionin the lobes of thedog'slungin thesupineanderect

position.J.Appl.Physiol.8:417-426.

2.Rahn, H.,and B. B. Ross. 1957. Bronchialtreecasts, lobeweightsand

anatomicaldead spacemeasurementsin thedog's lung.J.Appl.Physiol. 10:154-157.

3.Karas,R.H.,C. R.Taylor,J. H.Jones,S.L.Lindstedt,R.B.Reeves,and E. R.Weibel. 1987.Adaptive variationin themammalianrespiratorysystem in

relationtoenergeticdemand. VII. Flowofoxygen acrossthepulmonarygas

exchanger.Respir.Physiol.69:101-115.

4.Hsia,C. C.W.,J. I.Carlin,P.D.Wagner,S. S.Cassidy,and R. L.Johnson,

Jr.1990.Gasexchange abnormalities afterpneumonectomy in conditioned fox-hounds. J.Appl.Physic!.68:94-104.

5.Carlin,J.I.,C.C.W.Hsia,S. S.Cassidy,M.Ramanathan,P. S.Clifford,

and R. L.Johnson,Jr.1991.Recruitmentoflungdiffusing capacitywith exercise before andafter pneumonectomy in dogs. J.App!. Physic!. 70:135-142.

6.Gil, J.,H.Bachofen,P.Gehr,and E. R. Weibel. 1979.

Alveolarvolume-sur-facearearelation in air- and saline-filledlungsfixedbyvascularperfusion. J.

Appl. Physiol.47:990-1001.

7.Wagner,W.W.,L. P.Latham,andR. L.Capen. 1979. Capillary

recruit-mentduring airway hypoxia:role ofpulmonaryartery pressure. J.Appl.Physiol.

47:383-387.

8.Malvin,G. M. and S. C. Wood. 1992. Effects ofcapillary red cell densityon

gas conductance offrogskin.J.Appl. Physiol.73:224-233.

9.Weibel,E. R. 1970.Morphometricestimation ofpulmonary diffusion

ca-pacity.I.Model and method.Respir.Physiol.11:54-75.

10.Carlin, J. I., S. S. Cassidy, U. Rajagopal, P. S. Clifford, and R. L.Johnson,

Jr. 1988. Noninvasivediffusing capacityand cardiac output inexercising dogs.J.

Appl. Physiol.65:669-674.

11.Hsia,C.C. W., J. I. Carlin, S. S. Cassidy, M. Ramanathan, and R. L.

Johnson,Jr. 1990. Hemodynamic changes after pneumonectomy in the

exercis-ingfoxhound. J.Appl. Physiol.69:51-57.

12.Hsia,C.C.W.,J. I.Carlin,M. Ramanathan, S. S. Cassidy, and R. L. Johnson, Jr. 1991. Estimation ofdiffusion limitation after pneumonectomy from carbon monoxide diffusing capacity. Respir.Physiol.83:11-21.

13.Hsia, C. C. W., M. Ramanathan, J. L. Pean, and R. L. Johnson, Jr. 1992. Respiratory muscle blood flow in exercising dogs after pneumonectomy. J.Appl.

Physiol.73:240-247.

14.Weibel, E. R., L. B. Marques, M. Constantinopol, F. Doffey, P. Gehr,and

C.R. Taylor. 1987. Adaptive variation in the mammalian respiratory system in relation to energetic demand. VI. The pulmonary gas exchanger. Respir.Physiol.

69:81-100.

15.Weibel, E. R. 1984. Morphometric and stereological methods in

respira-tory physiology, including fixation techniques. In Techniques in the Life

Sciences.Respiratory Physiology.A. B. Otis,editor. Elsevier/Scientific Pub-lishers Ireland Inc., Limerick. P401:1-35.

16.Holland, R. A. B., W. VanHezewijk,and J. Zubzanda. 1977. Velocity of oxygen uptake by partly saturated adult and fetal human red cells. Respir. Phys-iol.29:303-314.

17.Holland, R. A., H. Shibata, P. Scheid, and J. Piiper. 1985.Kineticsof 02 uptake and release by red cells in stopped-flow apparatus: effects of unstirred layer.Respir. Physio. 59:71-91.

18.Yamaguchi, K., P. D. Nguyen, P. Scheid, and J. Piiper. 1985. Kinetics of

02 uptakeand releasebyhumanerythrocytesstudied_byastopped-flow tech-nique. J.Appl. Physiol. 58:1215-1224.

19.Weibel, E. R., C. R.Taylor,J. J.O'Neil,D. E.Leith,P.Gehr, H. Hoppeler, V.Langman,and R. V.Baudinette. 1983. Maximaloxygen consumption and pulmonary diffusing capacity: a direct comparison of physiologic and

morpho-metric measurements in canids. Respir.Physiol.54:173-188.

20.Holland, R. A. B. 1969. Rate of02dissociation from 02Hb and relative combination rate of CO and02in mammals at 370C. Respir. Physiol. 7:30-42. 21.Holland, R. A. B. 1969. Rate at which CO replaces 02 from 02 Hb in red cells of different species. Respir.Physiol.7:43-63.

22.Massion, W. H., D. R.Caldwell,N. A.Early,andJ. A.Schilling.1962. The relationship of dry lung weights to pulmonary function in dogs and humans. J. Surg. Res. 2:287-292.

23.Schilling, J. A. 1965.Pulmonaryresection andsequelaeof thoracic

sur-gery. In Handbook of Physiology, W.0.Fenn and H. Rahn, editors. Section 3: Respiration. Vol. 2. The AmericanPhysiological Society, Washington, DC. 1531-1563.

24.Reeves, R. B. and H. K. Park. 1992. CO uptake kinetics of red cells and COdiffusing capacity. Respir.Physiol.88:1-21.

25.Roughton,F. J. W., R. E.Forster,and L. Cander. 1957. Rateatwhich carbon monoxidereplacesoxygen from combination with humanhemoglobinin solution and in the red cell. J.Appl.Physiol.11:269-276.

26.Crapo,R.O.,N.Bitterman,S. L.Berlin,and R. E. Forster. 1989. Rate of CO uptake by canineerythrocytes as a function ofP02. J. Appl. Physiol.

67:2265-2268.

27.Davies, P., J. McBride, G. F. Murray, B. R.Wilcox,J. A.Shallal,and L. Reid. 1982. Structuralchangesin the caninelungandpulmonaryarteries after pneumonectomy. J.Appl. Physiol.53:859-864.

28.Wilcox,B.R.,G. F.Murray,M.Friedman,and R. L. Pimmel. 1979. The effects ofearly pneumonectomyontheremaining pulmonary parenchyma. Sur-gery(St.Louis).86:294-300.

29.Crapo,J. D. and R.0. Crapo. 1983.Comparisonof totallungdiffusion

capacityand the membranecomponentof diffusioncapacityasdetermined by

physiologicandmorphometric techniques. Respir. Physiol. 51:183-194.

30.Crapo,J.D.,R.0. Crapo,R.L.Jensen,R. R.Mercer,and E. R. Weibel. 1988. Evaluation oflungdiffusing capacitybyphysiologicaland morphometric

techniques [publishederratum appears in J.Appl. Physiol. 1988,65:following

2800]. J. App!.Physiol.64:2083-2091.

31.Grote,J. 1967. Die Sauerstoffediffusionscostanten imlungengewebeund

wasser und ihre temperaturebhangikeit. Pfluegers Arch. Gesamte Physiol.