How does HeO2 increase maximum expiratory flow in human lungs?







Full text


How does HeO2 increase maximum expiratory

flow in human lungs?

S N Mink, L D Wood

J Clin Invest. 1980;66(4):720-729.

We used the retrograde catheter technique to investigate the effect of HeO2 on resistance to maximum expiratory flow (Vmax) in airways subsegments between alveoli and the equal pressure point (EPP), and between EPP and the flow-limiting segment (FLS). FLS were found at the same airway site in sublobar bronchi (i.d., 0.54 +/- 0.13 cm) on both air and HeO2 in the six human excised lungs studied. Static elastic recoil pressure (5 +/- 1 cm H2O) and the lateral pressure at FLS (critical transmural airway pressure -6 +/- 3 cm H2O) were not different on the two gases. delta Vmax averaged 37 +/- 8.9% and was similar to the value found in healthy subjects of similar age (66 +/- 10 yr). EPP were located on HeO2 in peripheral airways (i.d., 0.33 +/- 0.03 cm), and EPP on air were located more downstream. Resistance between EPP and FLS was highly density dependent. Resistance between alveoli and EPP behaved as if it were density independent, due in part to Poiseuille flow in the peripheral airways and in part to the consequent narrowing of peripheral airways on HeO2. This density-independent behavior in peripheral airways reduced delta Vmax on HeO2 from its predicted maximal amount of 62%. Assuming that FLS is the "choke point" these findings are consistent with wave-speed theory of flow limitation modified […]

Research Article

Find the latest version:


How Does HeO2 Increase Maximum

Expiratory Flow




STEVENN. MIINKandL. D. H.WOOD, Sectioni ofRespiratoryDiseases, Departmenit of

Medicine, University of Manitoba, Winnipeg, Manitoba, Canada R3E OZ3

A BS T R A C T We used the retrograde catheter







maximum expiratory flow (Vmax) in airways stubseg-ments between alveoli and the equtal pressuire point (EPP),and between EPP and the flow-limiting segment (FLS). FLS were found at the same airway site in sub-lobar bronchi(i.d.,0.54+0.13 cm) onboth air and HeO2 in the six human excised lungs studied. Static elastic recoil pressure (5±1 cm H20) and the lateral pressuire at FLS (critical transmural airway pressuire -6+3 cm H20) were not different on the two gases. AVmax aver-aged 37±8.9% and was similar to the valuie fouind in healthy subjects ofsimilar age (66+10 yr). EPP were

located on HeO2 inperipheral airways (i.d.,0.33±0.03

cm), and EPP on air were located more downstream. Resistance between EPP and FLS was highly density dependent. Resistance between alveoli and EPP

behavedas ifitweredensityindependent, dueinpart

toPoisecuille flowintheperipheralairways andinpart totheconsequent narrowing of peripheral airways on HeO2. This density-independent behavior in periph-eral airways reduced AVmax on HeO2 from its predicted maximal amount of 62%. Asstuming that FLS is the "choke point"thesefindingsareconsistent with

wave-speed theory of flow limitationi modified to include


density-independent pressure losses in

peripheral airways. These restults and concltusions are

similar to those fouind in living dogs. They question previous interpretation of


as an index of

periph-eral airway obstruction, anddemonstrate the utility of the wave-speed theory in explaining complicated mechanisms of expiratory flow limitation.



a helium-oxygen (HeO2) gas mixtuire has a

lower gas density than air, the normal response to Dr.WoodisaScholar of the Canadian Life Insurance Asso-ciation.Dr.Minkisarecipientofan AmericanLung

Associa-tioin Traininlg Fellowsship. Dr. Mink's presenit address is

Division of Pulnionary Diseases, University of Texas Health

Scienice Center-atSaniAntonio,Tex.

Receivedforpublication 18July 1979andint revisedform 29April1980.


breathing this mixtture is an increasein maximum ex-piratory flow (Vmax) (1, 2). The magnittude of this in-crease


isoften less inpersons with obstructive airway disease and the decrease in response is pre-sumed to be dependent upon the site of air-flow ob-struction.Those persons whohavelittle or no response tobreathingHeO2 are thought tohave increased airflow resistance in the peripheral airways where the flow regime is laminar and unresponsive to gas density. Those who are responsive to HeO2breathing are said to haveobstructioncentrally when the flowregimesare responsive to the effect of gas density.

Yet,eveninnormals, the responsetobreathing


is variable and AVmax may vary from 20 to 80% (1, 2). This nonuiniform response to HeO2 may indicate that the density dependence ofresistance to airflow


also show wide variability, and the resistance to dif-ferent flow regimes may contribute to this varied re-sponse. To determine the cause of this wide variation, we previously examined and described the mecha-nisms of


in dogs (3). The latter study was de-signed within the framework of

ouir uniderstanding

the mechanism of flow limitation, which included the equal


point (EPP) theory ofMead et al. (4), and the Starlingresistorconcept ofPrideetal. (5). The pertinent concepts of these approaches havebeen

re-cently reviewed (3). In the


canine experi-ment, to partition the airway segment relevantto flow limitation,weplaced retrograde cathetersattheEPP on HeO2 and jtust uipstream from intrathoracic loci

where flow became limited (i.e., theflow-limiting seg-ment [FLS] of Pride et al.). Accordingly, we

meas-ured the pressure drop and flow resistance along the twoairwaysegments

dturing breathing


at maximumexpiration. We therefore determined the airwaysitethatallowed


toincrease.Theresistance to airflow upstream from the HeO2 EPP was inde-pendent of thegas uised to ventilate the lun-g and the pressure drop was

explained by

viscous losses


Resistance betweentheHeREPPandFLSwashighly density dependent and the


loss was due to

convective acceleration (Pca). Moreover, the magni-tude of




could be

explained by



tive proportion of the total presstire drop (Ptot) from the alveoltus to the FLS that wasdependentongas density. We fturther ainalyzed ourrestults in terms of the


theory (6), aind




derived from this theorv, we were al)le to predict A17max from the ratio of Pfr/Ptot.

To determine whether the above mechanisms and concepts were pertinent to the response to breathing HeO2 inthe human lung as well, weperformedasimilar experiment in excised human lungs.


A* Criticall cross-sectionial area ofthe choke point


lps P*


pressure point


segment Litersper second

Pressure headat thechoke point

Pbr- Ppl Lateral bronchial pressure minus pleural surface pressure (transmural airway pres-sure)

Pc Transmural airway pressure atthe central catheter positioin

Pca Pressureloss duetoconvectiveacceleration Pel Elastic recoil pressure ofthe lung Pfr Pressure losses due tofriction

Pp Transmural


pressure atthe periph-eral catheter

Ptot Total pressure drop

Rc Airwayresistanceupstream from thecentral catheter

Rc-p Airway resistance between the two


Rp Airway resistance upstream from the peripheral catheter

Rus Upstream resistance

VEPP Volume atthe HeO2 EPP Vmax

AVmax Maximum expiratory flow

Increase in


on HeO.,


The six humilcan luinigs (three right, three left) in this

experi-ment kverestuidiedwithin 6h of postmortemexatminationl and

vere normiial


gross inspectioni.The vesselsatthe hiluim

ofthelungweretie(dand theIlungplacedin avoluime

displace-mentplethysmograplh,wvhereit wasventilatedthrouigha

maini-stemilbronchus.Asingleretrogradle catheter wasplacedinthe lowerlobetomeasuirelateral bronchialpressuire (Pbr)

accord-ingtothetechni(quieofMacklein anidMead(7),anidan identi-cal catheterwasplacedwithin theplethysmographtorecord pleural surflce pressuire (Ppl). The experimiient protocol and the



imieasurinig lulng

flow, volume, anid

tranis-multral airway


(Pbr- Ppl) at two dlifferent catheter

sites (HeO2 EPP andcl FLS) were described inldetail else-,,here(3). Abrief description isgivenheretoindicateminior


With theretrogradecatheteriniaperipheralairway (He(2),

(lutasistatic auidforced expiratory(Pbr- Ppl)-lungvolumeanid

lung flow-volume culrves were recorded on a duatl beam

oscilloscope dturing successive deflations from total Ilung

capacitytominimalgasvolume.Aftertwosetsof curveswere

obtainedonleither airor an80:20mixtureonHeO2, theywere

repeated three times on the othergas and then twice more on the first gas. Then the catheter was moved mouithward

tothe site offlow limitation atthe volume ofEPP on HeO2 (VEPP)detected and described(3).As inthecanine stuidies,

this central catheter positionwas0.25-0.75cmtupstreamfrom the FLS where Pbr-Ppldecreasedabruptlyand varied with

reservoirpressure.Oncethis central catheter positionwas so

identified, the se(quienice of measurements were repeated as

described above for the peripheral catheter position. By stuper-imposing thephotographicrecords of all curveson each gas,

composite flow-voluime, static presstire (Pel)-voltume, and dynamic (Pbr -Ppl)-lung volume curves for the peripheral

(Pp) and cenitral (P(.) catheterpositionwere obtained. These

were analyzedatVEPP forVmax, Pel, Pc,and Pp.Resistance

tothe total airway segment (Rc = [Pel-



airwaysegmenit (Rp =[Pel - Pp]/Vmax),and interveninig seg-ment (Re-p =(Pe -


were caleculated as described (3). Statistical comp)arison between air and HeO2 measure-mentswereperformed uising the Wilcoxon matched pairtest.

The adcqtiate frequency responise ofthe apparatus anid the

smnallerror inlateral pressure measuremeintwasdescribed(3).

Atthe end of each experiment, the lungwasplaced inliq(uid formaliin at atranspulmonarv pressture of 20 cm H20 for 4-5 d. Theni the positioin of each catheter was determined in the fixedluinginterms ofdistanicefrom the carina, the upper lobe

bronchus, airway diameter, and airway generation. Trachea

wasconsidered generation 1, whereas lobar bronchi to the left lower lobe and right lower lobe were generations 4 and 5,

respectively. In two of the exiced lungs (5 and6),

morpho-metries wereperformed. This included grading the extent of

any emphysemra present, and measiurenments of meani in-ternall bronchiolar diameter and the ratio of internal bronlchiolar dianmeter to the external adventitial arterial diameter.Mloreover,todetermine if therewasaniyartifact pro-duced byuisingonlyonelung, oine additioInal experimentwas



meassurementsweremiade with


luings intact andthe retrograde cathleter placed in its ustual position

in the lower lobe. After all measturements were made, the leftluingwas removedaindthe proceduire repeatedas inthe othersixexperimenits.


Table I shows theclharacteristies ofthe excised luings usedinthisstudy. Allwerembalewithanaverage ageof

66 yr. Ingeneral, deathwasduetoarterioscleroticheart

disease aind autopsy weights of four of ouir lungs were

slightly high when compared tothe normal valuies of

this instituition. Becautse ofthe natuire of the autopsy

series, ac more detailedpulmonary historycouldnotbe obtained. However,althouigh some lutng weights were

sliglhtly heavy,Fig. 1 showsthat theirindividualvaltues of


on air were not different from


predicted fromanormal poptulation ofsimilarage(8). Moreover,

their values of


on HeO2 were similar to those

fouind b)yDosmnan etal.(2)fornormal elderly sul)jects, anud the mean value of theiruipstreamii resistance (see below) was similar to that caleulated by Mead et al. (4). Additionally, in the two lungs in which morpho-metrics were performed, bronchiolar measturemenits



Clinical Summary of Excised Lunigs

Ltung Age Sex Lungwt* Causeof death

Jr g

1 60 M 380(R) Arteriosclerotic heart disease 2 84 M 770 (L) Acute myocardial infarct

3 70 M 910(R) Acuite myocardial inf:arct

4 59 M 410 (L) Arrythmia

5 60 M 720(R) Arteriosclerotic heart disease 6 63 M 480(L) Bleeding ulcer

66±10 nl R (360-570)

ni L(325-480)

Abbreviatiotns used in thistable: M, male; L, leftluing; R, right lunilg; nI,normal.

semawasjudged mild and focal.' However, artificially ventilatingthe mainstembronchuis may have produced physiological changes in the parameters tested, and its possible consequences are ouitlined in the discus-sion below.

Fig. 2shows the composite flowvolume andpressure volume curves at the two retrograde catheter positions foreach experiment. There were nosystemic changes

observed in the curves on onegas measuredbefore and

after the second gas, or changes in the flow volume curveafter the catheterwasmoved from itsperipheral

to downstream position. Flow was -37% higher on

HeO2 over the upper range of thevitalcapacity. There was no difference observed between the quasi-static pressure volume curves on air or


so a single cturve is displayed (Fig. 2, dotted line). At most lung

IEmphysemawasfocal and grade20 orlessinbothlungs.

Internal bronchialdiameter for excised lungs 5 and 6 was 0.88

mmn aindl 0.82 mm, respectively, anil externial advenititial

arterial Diam was0.86and 0.70,respectively.









50 100

Vmax5o%Predicted on Air

FIGURE 1 Percent predictedVmaxon air for the sixexcised

lungs (closed circles) at50% vital capacity isplotted onthe abscissa against their response to breathing the HeO2

mix-ture onthe ordinate. Mean (+SD) of the presentstudyshown

by closed triangleand the closed square represents themean

value (+SD) found by Dosman et al (2).

volumes, the dynamic pressure attheperipheral cathe-terwas less on HeO2 (solid line) than on air (broken line) indicating thatthe flow-relatedpressure drop be-tween alveoli and the peripheralcatheter was greater

on HeO2. Thiswas not true inthe downstream catheter where at some lung voltumes dynamic Pbr - Ppl was greater on air and at other voltumes it was greater

on HeO2.

The mean lung volumeatVEPP was56±8.2% vital

capacity. Table II presentsthe analysis ofthe curvesat

VEPP,indicated by the arrow in Fig. 2. At that lung

vol-ume,Vmaxincreased inallexperiments andmeanVmax increased from 2.1±1.1 to 2.8±1.5 liters per second (Ips). This 37% mean increase in Vmax was not

asso-ciated with achange in Pel, which averaged 5.4±1.3

cm H20 on airand5.1±1.4 cm H20 onHeO2.The pres-suremeasuredatthedownstream position tendedtobe lower on HeO2 than on air but was not significantly

dif-ferent.Themeanvalueswere -5.3±3.9cm H20on air

and -6.4±2.5 cmH20 on


These values

approxi-mate critical transmural airway pressure on the two

gases. Pel - Pcaveraged 10.7±3.2 cm H20on airand 11.5±2.8 cm H20 on


However, because of

in-creasedflow,Rcdecreasedon HeO2inall experiments from6.6±4.08 to 5.09±2.63cmH20/lps. This 30%

de-crease inRccorrespondedtothe37% mean increase in

AVmaxon HeO2.

The retrograde catheter in the peripheral position partitionedthe resistance between alveoli and FLS.At

VEPP on HeO2, Pbr - Ppl wasgreater on airthan on

HeO2. Pp wasabout zero on


(0.1±0.4 cm H20). Mean Pp on air (1.6±1.3 cm H20) was significantly greater than that on HeO2 (P <0.05). Therefore, the pressure drop between alveoli and the peripheral catheterwas smalleron air(3.8cm H20)thanon HeO2 (5.0cmH20). Moreover,sincePpon air washigherthan

2pponHeO2wasslightlygreater thanzerobecauselung6

was examined at a slightly higher lung volume than VEPP, where theperipheralcatheter plugged.








20- 1

















1 0


FIGURE 2 Flow-volume (A) andl pressure-volumnie (B) curves in each of the sixexperimentts on air (hroken lines) and HeO2 (solid linesi). The static (Pbr- Ppl)-lung voltume cutrves were not

different between gases(dotted lines).Thedynamic (Pbr- Ppl)-lungvolume curves at the

periph-eral and downstream catheter position are representedlby the upper aincllowercurves, respec-tively.ArrowindicatesVEPP.

on HeO2, the peripheral catheter locationi did not

represenit the position of the air EPP. An additional

1.6 cm H20 pressure drop would be necessary before Pbr- Ppl was zeroduring airbreathing. Accordingly,

the EPP oIn air were located downstream from the peripheral catheterwherethelateralintraluminal pres-suiredecreased byanaverage of1.6 cmH20. Resistance

between alveoli and the periplheral caltheter was

greateron air intwoexperiments and greateron HeO2

in fouir. Mean Rpwas2.04±+0.91 on airand2.02+(0.61 cm H20/lps on HeO2 and these values were not

dif-ferent on the two gases indicatinig that resistaince inthis airwaysegmiienitwasindependent ofthe

venitilat-ing gas.

Alechanisin of Inicrease in Maximum Expiratorit Flowv oilHeO2 7




Effectof HeO2onPressure-VmaxRelationshipsinAirtvay Subsegmenits. Antalysisof VEPP oni HeO2*

V... Pel Pc Pp Rc Rp Rc-p

Air 2.1+1.1 5.4±1.3 -5.3±3.9 1.6±1.3 6.61±4.08 2.04±0.91 4.57±3.66

HeO2 2.8±1.5 5.1±1.4 -6.4±2.5 0.1±0.4 5.09±2.63 2.02±0.61 3.07±2.07

P < 0.03 NS NS 0.05 0.03 NS 0.05

* Mean±SDVL at VEPP was


vitalcapacity. All values aremean±SD.

The pressure drop between the catheters (Pc - Pp) averaged6.9+±3.5on airand6.5±+2.7cm H20 onHeO2. Resistancebetween the catheters was greater on air in all experiments and the increase from 3.07+2.07 to 4.57+3.66 cm H2O/lps was 49% indicating greater density dependence betweenthe catheters than in the wholeRc segment.

The average position of the HeO2 EPP was in a peripheral bronchus(generation11±1.3)havingan air-way Diam of0.33±0.03 cm, and situated 14.6±2 cm from the carina. This corresponded to order 23 in model 2 in reference 9. Using the values in the latter reference, this catheter subtended 2.8+9% of the total flow. Thelocation of the HeO2 FLS wasin a sublobar bronchus between the seventh and eighth generation having an airway Diam of0.54 cm and 10.4 +2.7 cm from the carina. The FLS were usually positioned at the junctionofposteriorand lateral basal segments or just slightly upstream and approximately correspond to order 25 in the same reference. At this location, the catheter subtendedalargerpercentageof the totalflow which averaged10.4±3%.Intheoneexperimentwhere twolungswere used and theentireintacttrachea con-nected to the gas source, the FLS was also located peripherally and did not change when one lung was



As in the previous


this experiment was de-signed within the


framework of the two

prevailing theories offlow limitation (4,



selecting a mid-vital capacity



we pro-spectively placed our intraluminal catheters to lie in

mid-vital capacityon HeO2attheequalpressure point (4, 10, 11) and just upstream from the FLS (5, 12). In

this way, wedetermined inwhich airway segment the low density gas was acting to lower flow resistance

andso increase Vmax.The FLS representsa site in the airway where alarge pressure




Up-stream from this location the geometry of the airwayis

fixed, whereas just downstream fromthispointevents

are determined by changes in reservoir pressure. Be-cause slightly downstream from our central catheter (<1 cm) intrabronchialpressure became very negative (<-50 cm H20) and varied with reservoir pressure, 724 S. N. Mink andL. D. H. Wood

we conclude that flow became limited in the airway segment just downstream from ouir central catheter. We will interpret our measurements partitioning the pressure drops and resistances upstream from the site of flow limitation to describe the mechanisms setting Vmax on each gas. We will try to integrate our results with the intuitive approaches ofMead et al. (4) and Pride et al. (5) as well as the mncathemiiatical approaclh described by the wave-speed theorv (6, 13). Fturther-more, we will use equations derived from the previouis caninestudy (3) andapply themtothe data foundinthe excised human lungs.

Mllechatnisms of AV,naX. In each of the six lungs, the position ofthe FLS was consistently located in an intra-parenchymal airway 2-3 cm upstream from the takeoff ofthe superior segment of the lower lobe where intra-bronchial pressure averaged about -6 cm H20 (Table II). Therefore, for the similar driving pressures on the two gases, AVmax increased by 37% on HeO2 due to a reduction in flow resistance between alveolus and FLS. Whereas the increase in Vmax was the result of adecreaseinflow resistance between alveoli and FLS, HeO2 had no effect on the resistance between alveoli and the peripheral catheter position (Table II). This was because both Vmax and the peripheral pressture dropwerelargeron


Accordingly, the resistance between alveoli and HeO2 EPP was independent ofthe gas used to ventilate the lung. Fturthermore, sinlce the transmural pressture at the peripheral catheter washigheronairthan on HeO2 (Fig. 2), the EPP on air muist have been positioned further downstream from the peripheral catheter a distance long enough to

ac-countforthe additional 1.6cm H20pressuredrop.We conclude therefore that one effect of ventilating ex-cised human lutngs with HeO2 was to shorten the air-way segment upstream from the EPPbymoving EPP peripherally.

In the airway segment between the HeO2 EPP and the FLS, Rc-p was 50% greater on airthan on HeO2, and the influence of the ventilating gas on the

resis-tancetoairflowtookplaceentirely within this segment. Accordingly, intheselungs,the totalairwayresistance to the FLS was composed of Rp between alveoli and HeO2 EPP, which was independent of the ventilating gas and Re-p between the two catheters, which was


Although Rp behaved as if it were density inde-pendent, this wasnotnecessarilydueto alaminar flow regime in this airway segment. Because (Table II) Pp on air was 1.6 cm H20 and less on HeO2 (0.1 cm H20),there mayhave beenachangeinthe dimensions of the airway geometry when thelungwas ventilated onthe two gases.Consequently,thediameter ofthe

air-waywouldbeexpectedtobelargeronairthan on HeO2. We found a similar result in the canine study (3) and argued that since the airways were subjected to dif-ferent transmural pressureonthetwogases, the airway wouldbenarroweronHeO2(14)(Fig. 3B),therefore

in-creasing the resistancetoairflowand


in-creasing thepressurecost. Because resistancewouldbe equal onthe two gases,itwouldsuggestalaminar flow


AIR Vmax= 2

+4 +2

+4 HeO2 t' = 2


AIR Vmax = 2 +4 +2


0 -3 -6

+2 0 -1



0 -3 -4


EPP(air) FLS(air)


+3 0 -4 -6

HeO2 Vmax= 3 Pbr-PpI

FIGURE 3 Schematicmodel comparing airway eventsduring

Vmaxon air(2lps) withthose on HeO2. (A) When HeO2flow

isalso2Ips, transmural airwaypressure (Phr- Ppl)and air-way area decrease more on air (solid lines) than HeO2(broken lines) because pressuire losses are density dependent. Flow limitsat Pbr- Ppl = -6 cm H20 onair, fixing EPP air, but does not limitat increased Pbr-Ppl (-4cm H20) on HeO2.

(B) Fturther lowering of reservoir pressure increases HeO2

flowuntil it becomes limited at 3 lps at the same airway site and Pbr - Ppl as FLS air. EPP HeO2 are fixed further up-stream, and Pbr-Ppl and airway geometry are reduced on

HeO2 (broken lines).

regime when in reality, the flow regime could be tur-bulent with different airway geometries. The initial mechanism for the differences in airway geometry wouildbe the directrestultofatnielaminar flow regime

in aperipheral airway site near the alveoluis.

The results of the partitionedresistances were

simi-lartothosenotedinmostofthe dogs (group 1) thatwe

studied (3) and suggest a similar model as to how breathing


increased Vmax, Fig. 3 depicts a

schematic diagram of the lung at mid-vital capacity which was stubjected to progressively decreasing

reservoir pressure during breathing air until flow limitation occurred. The transmural airway presstire from alveoli toairway openingare depictedatthe on-set offlow limitation on air (Vmax =2 lps). At similar flows on HeO2, e.g., 2lps which was notVmax,the EPP would be identically located or even fturther down-streamon HeO2.The latter may occur becauise at

simi-larflows there would be no narrowing of the airway on


and resistance between alveoli andEPPmaybe less thanonair. Between EPPand FLS resistancealong the airway segment onHeO2would also be less thanon air.Therefore,the transmural pressureonHeO2 wouild be larger and the cross-sectional area of the airway would be greater than on air. Accordingly, expiratory flow limitation would notoccur.

When the reservoir was lowered frirtherto increase flow onHeO2,asshown in Fig. 3B, transmural pressure decreased at each airway site because the flowresistive

pressure drop increased. When flow reached 3 lps, Pbr - Ppl fell to the critical value of-6atthe site of the air FLS, converting the airway to a Starling resistor. As wasthe case with air, the EPP on


werefixed at the point where Pbr-Ppl was zero. Because resistanice upstream to this pointappearedtobehaveasifitwere independent of the ventilating gas and becaause Vmax



wasgreater than thatonair, the resistive pres-sure drop to this point was greater on


than air.

Therefore, the


EPP were further upstreamii. According to the EPP theory (4), the pressure drop from the alveoli to EPP is equal to Pel, and the ratio of this pressuredroptoVmax isdefinedasupstream

re-sistance (Rus). For a given Pel, a reduction in Vmax would imply an increase in Rus. Therefore, although Rusdecreased during breathingof HeO2, this occurred because of a geometry changewhich was unrelated to peripheral airwayfunction.Wetherefore concludethat

Rus is a very indirect indicator of peripheral airway resistance.

In summary, ventilating the lung with HeO2 de-creased resistance from alveolito FLSby30%and was associated with a tVmax of 37%. Whereas resistance


letween alveoli and


comparedto air. Inspite of


flowactuallyincreased and the resistance between alveoli and FLS was less on HeO2 than on air. Thus density dependence of resistance outweighed the geometric changes. This density dependence of re-sistancemtusthaveoccurred throughoutall segments of the upstream airways where there was any signifi-cant geometric difference in order to allow for the increase in flow through the segment. Our data indi-catethatdensitydependence was greatest between the catheters where resistance decreased by 50%.

Although the above explanation and analysis appear todescribethe mechanism ofAVm5xon HeO2 at


dlifferentmechanisms may be responsible at other lung volumes.In someof thelungs shown in Fig.2,it can be seen that at high lung volumes, 1)oth Rc and Rc - p appear to be density independent.


this wouldinvokeexplanations of Vmax on HeO2 other than that previously described. Similar seemingly density independent observations of both partitioned seg-ments have been shown to occur in group II dogs (3). This finding has been associated with different loca-tions ofthe FLS on the two gases. Since ourcatheter

was notplaced at the FLS at these high lungvolumes, it is notclear whether similar conclusions hold true for human lungs.

Effect of peripheral pressure drop on AVmax. Our results show that mean AVmaxincreased by 37%. If re-sistance to airflow were completely inertial,


would vary as density -0'5, and



have in-creased by 62%(1, 2, 6, 15-17). Such an inertial pres-sure loss due to Pca was probablythe majority of the pressure









EPP and FLS. The finding that the resistance in the peripheral segment behaved as ifit were


in-dependent may explain whyAVmax waslessthan maxi-mal. Since the Pfrwas due to resistance thatbehaved

asifitweredensity independent, an appreciable

por-tionof thePtotbetweenairspacesandFLSvariedwith Vmax independent of the ventilatinggas.Consequently, the intuitive reason why flow on


increased by only 37% was thatmeasured Pfrreduced the density-dependentcomponentofPtotthatwas




In effect, the density-independent peripheral flow regime and the consequent airway geometry change induced on HeO2reducedthe influence of gas density on


SincePfron HeO2 accounted for0.45

Ptot, AVmax lies about half way between the maximal value of62%predicted for the frictionlessextreme(Pfr/

Ptot =0) and the minimal value ofzeropredicted for the completely viscous extreme (Pfr/Ptot= 1).

A more rigorous explanation of how Pfr/Ptot would reduceVmaxcanbe derived from thewave-speed theory of flow limitation (3, 6, 13).


to the


Vmax occurs when at a point in the airway, the choke point, linearvelocityequalsthe


of propagation of

the pressure pulse waves. We previously used equa-tions derived from the wave-speed theory to describe how increasing amounts of Pfr affected AVmax on HeO2 (3). In the present experiment, where the FLSwas lo-cated at the same airway site on both gases, similar equations canbe used. Additionally,from these equa-tions, the ratio of the area at the choke point as well asthe ratio ofthe pressure drop to the choke point could bedetermined on the two gases.

Fig. 4shows the graphic solutions of these equations for these three relationships (3, appendix). In Fig. 4A


is plotted on the ordinate against increasing valuesof Pfr/Ptot. When Pfr/Ptot is zero, AVmax is maxi-mal at 62%. With increasing values of Pfr/Ptot AVmax

de-creases in a curvilinear manner. As illustrated in the Fig. 4 whenmean (±SD) Pfr/Ptot (0.45±0.15) found in the presentexperiment is plotted against mean (±SD)





predicted by the wave-speed theory. In effect,




A* HeO2









1.2-0.2 0.4 0.6 0.8 1.0


r I

0.2 0.4 0.6 0.8




0 0.2 0.4 0.6 0.8 1.0


FIGURE 4 Effects of viscous resistance on predictions of HeO2effects ofwave-speedtheoryof flow limitation. Abscissa: ratiooffunctionallydensity-independentPfronHeO2 to Ptot between alveoli and choke point. Ordinates: (A) AVmax

de-creases in acurvilinearmannerfrom0.62atPfr/Ptot=0to0 at Pfr/Ptot= 1.0. Note thatmean (-) (+SD) values arejust below the predicted relationship. (B)The ratio ofA*HeO2/ A*airdecreases as Pfr/Ptot increases,explaining why AVmax decreases. (C) TheratioPel-P*HeO2/Pel -P*airincreases initially as Pfr/Ptot increases and then declines as Pfr/Ptot


726 S. N. MinkandL. D. H. Wood



?0-c I I


4 2


tionally density-independent losses caused wave-speed theory topredict a lowerthan maximal AVmax.

What is the mechanism responsible for this effect? Dawson and Elliott (6) point out that such viscous losses reduce thelocaldistending pressureheadatthe choke point (P*) below the static recoil value. This re-duced head implies a rere-duced critical cross-sectional areaofthechokepoint (A*)resultinginreducedcritical flow. Becausethe viscous losseswere greateron


P*wasreduced by a greater amount than on air so that A* was reduced more on


than air.

These latter two relationships are illustrated in the second and thirdpanelsof Fig. 4. Inthe second panel, the ratio A* HeO2/A* air is plotted against Pfr/Ptot. WhenPfr/Ptot is zero, A* HeO2/A* air is one andAVmax is 62%. With a value ofPfr/Ptotfound in the present ex-perimentA* HeO2/A* airwould be about 10% less on


thanonair.The third panel shows that the ratio of the pressure drops Pel-critical transmural airway pressure for HeO2/Pel-critical transmural airway pres-sure for air would also change at different values of Pfr/Ptot. When


is62%,Pfr/Ptot would be zero and the ratio of the pressure drop would be one. As Pfr/ Ptot increased, this ratiowould becomegreater and at a Pfr/Ptot= 0.45 the pressure drop on HeO2 would be -10% greater than on air. This prediction would be consistentwiththeslightly increasedpressure


on HeO2 found in the present study, although our methodology was not sensitive enough to detect this change systematically.Notethatwith increasing values of




yet the ratio again ap-proaches one. Thephysiologic mechanism that would allow theratioofthe pressuredropstodecrease while the ratio of the areas increased is unclear to tis at the present time.

Insummary, we were able to predictAVmaxon HeO2 from equations derived from the wave-speed theory. The lower


than predicted for the frictionless state (62%) occurred because functionally density-independent losseson HeO2exceededthoseon air, re-ducing theP*andA* on HeO2. Greater Pfr occurred on HeO2 because


was greater, and visctous resis-tancebehavedas ifitwas independent ofgasdensity. Comparison ofthepresentstudytoprevious studies. Macklem and Wilson (18) measured intrabronchial pressure in manand showed that EPP develop in the trachea, move upstream as lung flow increases, and become fixed atthe level of segmental bronchi when


isattained. Over the middle half of vital capacity, it isestimated that EPP remained fixed in this airway location,while below25% vital capacity, it was thought thatEPPmovefurther upstream. Our results show that

EPPwerelocated more peripherally than found in the above study. There are many reasons for this dis-crepancy considering the varying techniques and methods used in each experiment including the fact

that this stucdy measured the EPPon HeO2. However, the most important reason may be relate(d to the dif-ferent age group used in the two stuidies. The meain age in Macklem's study was 32 yr whereas in our study, the mean age was 66 yr. Previous work (4) has shown that age is accompanicd by a dcecrease in

elastic recoil of lutngs resuilting in decreased static

re-coil pressuire at a given lung volume. This would tendto decrease the drivingpressture andl move EPP peripherally.

In contrastto the canine lung in which FLS was in

the trachea (3), FLSin theagedhumacn excised lung is locatedinasublobar bronchus. We wondlered whether

our studies ofonelung changed


by eliminating

alength of trachea ancd reducing the airway compliance

asdescribed by Jones et al. (19). To the extent that the tule in the mainstem bronchus prevented flow limita-tion inthat airway at reducedVmax,flow increased until transmural pressures of upstream intraparenchymal air-ways reached their critical transmuiral airway pressure. This se(quence predicts that stucdy of both lungs expir-ing throuigh a common trachea segment is associated with lower


on each gas and a more downstream locus of FLS. Note that even if this occurred, AVmax will again be determined by Pfr/Ptot. Although we

can-notpredict the effect of a changed FLSlocus oneither Pfr or Ptot, we know from our canine stuidies (3) that trachealFLS are associated with AVmax(Illite similarto

that observed in aged human lungs. Furthermore, the one pair of human lungs stuidied with tracheal

cannula-tion demonstrated the same sublobar loctus ofFLS as when it was studied with cannulation of the mainstem bronchus. Finally, ouir findings of intraparenchymal FLS supports the prediction ofMelissiinos et a]. (20), that flow limitation shifts into intraparenchymal air-ways as lung volume decreases during forced expira-tion. We conclude that the determinant ofAVmaxinthe aged human lung as in the canine lung is the ratio of the apparent viscous pressuredroptothe Ptot between

air spaces and thechoke point.

Defining this mechanism does not simplify our un-derstanding of AVmax as a test of early and peripheral airwaysobstruction. Despas et al. (1) reasoned that be-cause HeO2 is a lessdensegasthan air, it could be used

to determine the site of airway obstruction in asthma. They thereforeassumed that in those asthmatics when AVmax increases <20%, airway obstruction was present peripherally in the small airway (<2 mm) where the flow regime was unresponsiveto gas density and con-sidered laminar. On the other hand, those who had a

>20% increase were considered to have obstruction more centrally where turbulent and convective flow regimes would be responsive to the lower density of HeO2. More recently, Dosman et al. (2) used similar concepts to study airway function in smokers. A

de-crease in AVmax was fouind in smokers whose airway


function wasotherwise normal on air,andthis finding was considered to represent early evidence of small airway obstruction. Therefore, the response to

breath-ing HeO2 was thought a useful clinical test to detect persons with early and presumably peripheral airway


Our study introduces some complexities to the

in-terpretations given by the latter studies because from

ourfin(linlgs, anabnormal response to breathingHeO2

isnot necessarily indicative of eitherairway or

paren-chymal disease. Consider, for example, howinthe

nor-mal population different magnitudes of central airway

caliber and therefore A* will alter AVmaxon HeO2

in-dependently ofany additional factors. In the instance in which A* is large and Ptot the same, since Pca is in-versely related to A*, Vmax will also be large, and Pfr (Vmax x Rfr) will tend to represent a large proportion

ofthe Ptot fromairspaces tothe choke point. This will resultin anI increase in Pfr/Ptot and will reduce AVmax

oIn HeO2 eveninalungfree of airway obstruction.

Ac-cordingly, normal subjects with large central airways

and so largeVmax will have low


and viceversa.

Indeed, Usinlg similar analysis anid assumptionis, stucha

resuiltlhas been f'oundl by Castile etal. (21). Mioreover, extendlingthe influenceofcentral airway sizeon /Vmax

topersons withperipheral airways disease complicates the interpretationi of breathing HeO2 even further. In at given lung, peripheral airways ol)struction tends to

reduce Vmax by reducing P*andA*atthechokepoint.

Ifchoke point loci donotchange with peripheral

air-ways obstruction, Ptot will not change acndc Pfr/ Ptot will inereasebecause Peaismore flow dependent thani Pfr. Accordingly, Vmax anid LVmax decrease with




butwhichvariable falls outside the rangeclassifiedlas normaiiil firstdepends on

the caliber of the central airwatys.Asperipheralairways obstruct, the subject having small


airways and large AVmaxwillbe considered tohave aclinicallylow


before his AVmax noticeably decreases. On the other hand, the subjectwith a large trachea and

there-fore small AVmax will develop appreciably low AVmax before Vmax isreducedbeyondthe normal range.


the ability of HeO2 breathing to

dis-criminate between normal aInd obstructed peripheral airways is affected by large variationi in normal Vmax attributable to normal variations in cenitral airways caliber (22). Thisconfounds theinterpretationofAVmax oIn HeO2 as ain indicatorofperipheralairways obstruc-tion, an(l (letracts from the tuse ofHeO2 as avaltuable cliniical tool.


The authors gratefully acknowledge the contrilbution ofDr. Ted Wilson who developed the mlathematical approach

uisedl hereini toadljuistwavespeed theory for viscouspressure

losses. We are gratefulto Dr. Norbert Berrand, Ms. Connie

Skoog,and Dr. W. M. Thurlbeck for providing the lungs for

studyandforperformingthe morphometric analyses. Weare

gratefulto Dr.N. R.Anthonisenfor his helpful review of this manuscript.

Thisstudy wassupportedbytheMedical Research Coun-cilofCanada.


1. Despas, P. J.,M.Leroux,andP. T.Macklein. 1972. Site of airway ol)struction in asthma as determined by

measur-ing maximal expiratory flow breathing air and helium-oxygen mixture.J.Clin. Invest. 51: 3235-3242.

2. Dosman, J., F. Bode,J. Urbanetti, R. Martin, and P. T.

Macklem. 1975. The use of a helium-oxygen mixture

duringmaximum expiratory flow todemonstrate obstruc-tion in small airways in smokers. J. Clin. Invest. 55:


3. Mink, S., M.Ziesman, andL. D. H.Wood. 1979.

Mecha-nismsof increased maximum expiratory flow during HeO2

breathingindogs.j.Appl. Physiol.Respir. Environ. Exer-cisePhysiol. 47(3):490-502.

4. Meacd, J., J. M. Turner, P. T. Macklein, aindlJ. B. Little. 1967. Significancee of the relationship between lung and recoiland maximum expiratoryflow.J.Appl. Physiol. 22: 95-108.

5. Pride,N.B., S. Permutt, R. L. Riley, and B. Bromberger-Barneia. 1967. Determinants of maximal expiratory flow from the lungs.J.Appl. Physiol. 23: 646-662.

6. Dawson,S.V., andE.A.Elliott.1977. Wave-speecl

limita-tion on expiratory flow-a uinifying concept. J. Appl. Physiol. Respir. Envirotn. Exercise Physiol. 43: 498-515.

7. Macklem, P.T., and J. Meaid. 1967. Resistanceofcentral andperipherall atirwaysmeassured byaretrogratde cattheter. J.Appl. Physiol. 22: 395-401.

8. Protte,D.J.,N.Craven,A.Naimalrk, aindR.M.Cherniack. 1973.Computer assistanieeinpulmonary fuinctionistudies.

illethio(ds lI Af.'Je(I. 12: 102-107.

9. Horsefield, K.,G. Dart, D. F. Olsen, G. F. Filley, andG.

Ctmminng. 1971. Models of the human bronchial tree. J.Appl.Physiol. 31: 207-217.

10. Mackleni, P. T., aind J. Mead. 1968. Fatetors dletermininig mnaximumn expiratory flow in clogs.J. Appl. Physiol. 25:


11. Gardiner,A.J., L.Wood,P.Gayrard,H.Menkes, aind P.T. Macklem. 1974. Influence of constrictioni in central and peripherall airways onmaximutm expiratory flowratesin

dogs.J.Appi.Physiol. 36: 554-560.

12. Smaldone, G. C., and E. H. Bergofsky. 1976.Delineationi offlow-limitingseggmentacnd predicted airway resistance

bymovablecatheter.J.Appl.Physiol. 40: 943-952. 13. Elliott, E. A., and S. V. Dawson. 1977. Test of waive-speed

theoryofflow limitationiinelastictulbes.J.Appi. Physiol. Respir. Environt. Exercise Physiol. 43: 516-522. 14. Huighes, J. M. B., H. A. Jones, A. B. Wilson, B. J. B. Granit,

and N. B. Pride. 1974. Stability ofintraptulmonary

bron-chial dimensionis dturing expiratory flow in excise(d tlungs.J. Appl. Physiol. 37: 684-694.

15. Wood,L. D. H.,ancd A.C.Bryani. 1969. Eflect of increasedl

gas density on flow-voltume cuirve of the luing.J. Appl.

Physiol. 26: 4-8.

16. Wood,L. D. H.,aindA.C. Bryan. 1978. Exericise

venltilai-tory mechanics at inereased ambient pressuire.J. Appl.

Physiol.Respir.Environ. Exercise Physiol. 44: 231-237.

17. Wood. L. D. H., L. A. Engel,P.Griffin,P.Despas, andP.T. Mackleni. 1976. Effect of gas physicall properties and


flow on lower pulmonary resistance. J. Appl. Physiol. 41: 234-244.

18. Macklem, P. T.,andN. J.Wilson. 1965. Measurement of intrabronchial pressure in man. J. Appl. Physiol. 20: 653-663.

19. Jones, J. G., R. B. Fraser, and J. A. Nadel. 1975. Effectof changing airway mechanics on maximum expiratory flow. J.Appl. Physiol. 38: 1012-1021.

20. Melissinos, C. G., and J. Mead. 1977. Maximum expira-tory flow changes induced by longitudinal tension on

trachea in normal subjects. J. Appl. Physiol. Respir.

Environ.Exercise Physiol. 43: 537-544.

21. Castile,R.G.,R. E.Hyatt, andJ. R. Rodarte. 1980. Deter-minants of maximal expiratory flow and density

de-pendence in normal man. J. Appl. Physiol. Respir. En-viron. ExercisePhysiol. In press.

22. Meadows, J. A., III,J. R. Rodarte,and R. E. Hyatt. 1980. Density dependence of maximal expiratory flow in

chronic obstructive pulmonary disease. Am. Rev. Respir.

Dis.121(1): 47-54.