special article the main the pulmonary...

SPECIAL ARTICLE

The Main Functions ofthe Pulmonary Circulation

By JULIUS H. COMROE, JR., M.D.

I HAVE chosen the title of this lecture tocorrect the belief of some that the pulmo-

nary circulation has only one function; somelater Conner Lecturer, however, will surelycorrect the assumption in my title that in 1965we know all of the main functions of thepulmonary circulation. *We do know several important functions:

The first-participating in pulmonary gas ex-change-has long been known. The second-serving as a blood reservoir for the left ven-tricle-has been known for some years butcould be demonstrated quantitatively only re-cently. The third-providing the nutrition ofthe alveoli and alveolar ducts-has been un-covered in the past 2 years. The fourth-filter-ing particles from the mixed venous blood-

From the Cardiovascular Research Institute, Uni-versity of California San Francisco Medical Center,San Francisco, California.The original investigations reported in this lecture

were supported in part by Grant HE-06285 ofU. S. Public Health Service.The Lewis A. Conner Memorial Lecture, presented

at the meeting of the American Heart Association,Miami, Florida, October 15, 1965. A slightly modi-fied version was presented as the Annual Lecture tothe Harveian Society, London, England, April 5,1965.

*There is now a vast literature on the pulmonarycirculation, which includes several recent books, re-ports of symposia, and scholarly review articles. Fish-man's "Dynamics of the Pulmonary Circulation" (inthe Handbook of Physiology, Section 2, Circulation,Volume 2, pp. 1667-1743) contains 446 references forthose readers who wish further details and documenta-tion. This Conner Lecture is not intended to beanother review, but more of a personal account ofrecent and current work by my colleagues in theCardiovascular Research Institute.

sometimes results in dysfunction, but I pro-pose to consider it instead a function that issometimes abused. The fifth-removing excessfluid from the alveoli-has been known since1873, when Colin poured 25 liters of fluiddown the trachea of a horse in a 6-hourperiod.

Gas ExchangeThe alveoli and the alveolar capillaries are

magnificently designed to exchange gases; noartificial lung ever made approaches them incapacity, speed, efficiency, compactness, or

M/IXE VENOIISBLOOD

ARTERIALBLOOD

Figure 1

Schema of the lungs and pulmonary circulation. Therounded areas represent alveoli; the shaded tubesleading to them represent all of the conducting air-ways. Mixed venous blood (dark) flows throughcapillaries in intimate contact with ventilated alveoliand becomes arterial blood (light). The large arrowin the upper airway represents the tidal volume; thetwo smaller arrows in the lower airways and alveoliillustrate uniform distribution of the tidal volume.The fine arrows represent the transfer of °2 andCO2 between gas and blood.

Circulation, Volume XXXIIl, January 1966146

by guest on July 6, 2018http://circ.ahajournals.org/

Dow

nloaded from

http://circ.ahajournals.org/

PULMONARY CIRCULATION

durability. For teaching purposes, I have oftenpictured the lung by a greatly simplifiedschema (fig. 1). Actually it is a vast, com-plicated and marvelous system of branchingtubes that is in itself a superb engineeringjob. The air tubes begin at the nose andmouth and then join to become the trachea.The trachea divides into bronchi and thebronchi divide and subdivide into bronchioles.After 18 branchings, there are several hun-dred thousand respiratory bronchioles andafter 24 branchings the air tubes end in about300,000,000 alveoli,1 where most of the gasexchange occurs. The blood tubes begin asthe pulmonary trunk, divide into the rightand left pulmonary arteries, which thendivide into arterioles, and these in turn intohundreds of millions of alveolar capillaries.The volume of blood in these capillaries atany one time is only about 75 ml, but thetotal surface area of the capillaries is 70 M2;the alveoli have a similar area for gas ex-change.

If the gas exchanger is an ideal one, all ofthe inspired gas and mixed venous bloodshould be distributed uniformly to these mil-lions of gas-blood interfaces. In the systemiccirculation there is need for both regionalregulation (to shift blood from nonvital organs

to vital organs during stress, or from in-active to active organs) and for over-allregulation (to maintain a proper pressure headto drive blood through the cerebral andcoronary arteries). Is there any need forregional or over-all regulation of blood flowthrough the pulmonary circulation? Becauseall of the right ventricular output must flowthrough the pulmonary blood tubes, thesetubes, theoretically, should always offer mini-mal resistance to blood flow so that the workof the right ventricle is minimal.* And yetwe now know that there is both regionaland over-all regulation of pulmonary vascu-lar resistance.2

Regional (local) regulation seems to servethe purpose of matching gas and blood flow.If gas distribution is not uniform but blooddistribution is uniform, the gas exchanger isnot an ideal one; if gas distribution is uni-form and the blood distribution is not, againthe gas exchanger is not an ideal one. Whenboth are nonuniform, the gas exchanger will

*This is not true of the fetal circulation. In thefetus, gas exchange occurs in the placenta and it isto the advantage of the fetus to reduce pulmonaryblood flow to the minimnum required for maturationof the lungs; this can be achieved by general pul-monary vasoconstriction.

Figure 2

Sections of cat lung frozen by the ultra-rapid freeze technique of Staub and Storey.4a Thelung on the left received a mixture of gas low in oxygen and high in CQ2; the lung on theright pure oxygen. Compare the narrowed arteriole (arrow) in the "hypoventialted" lung onthe left with the wide arteriole (arrow) in the oxygenated lung on the right.

Circulation, Volume XXXIJI, January 1966

147


Dow

nloaded from


COMROE

still function well if the alveoli that havelittle ventilation also have little blood flow,and if the alveoli with more ventilation alsohave more blood flow. The matching of bloodand gas is, therefore, crucial to good gas ex-change.Two mechanisms are known which help to

achieve matching:1. Regional alveolar hypoventilation results

in pulmonary arteriolar constriction in thesame region. This was demonstrated first byLiljestrand.3 Recently, Staub4 and associatesin our Institute have been able to see andidentify the vessels that constrict. They pro-vided a low oxygen-high carbon dioxide mix-ture to one lung of an anesthetized cat andpure oxygen to the other lung; in this waythey mimicked severe regional hypoventila-tion without significantly changing the oxygencontent or tension of arterial blood and (be-cause of the local vasoconstriction) withoutchanging the arterial PCo2 or pH. They theninundated both lungs simultaneously withliquid propane, freezing the lungs almost asthey were in life4a: Sections of the still-frozenlungs clearly demonstrated constriction of ar-terioles in the "hypoventilated" lung and notin the other (fig. 2). This constriction reducespulmonary blood flow to regions with re-duced air flow and helps to match air andblood flow throughout both lungs. The mech-anism is a local one and does not depend onnerves or reflexes.

2. Regional ischemia results in airwayconstriction in the same region. Severinghaus,Swenson, and associates"' 6 inflated a balloonat the tip of a catheter to occlude one pul-monary artery in both dog and man. Withina few breaths, air flow decreased to the lungwith its pulmonary artery occluded and in-creased to the other lung. Such a regionalincrease in airway resistance serves todirect more of the inspired air to alveoli withnormal or increased blood flow and so helpsto match air and blood flow to the alveoli.It is partly for this reason that some patientswho have complete obstruction of some pul-monary arteries by emboli have little or nounsaturation of their arterial blood.

Because alveoli with no pulmonary capil-lary blood flow receive no carbon dioxidefrom mixed venous blood, it seemed reasonablethat the bronchoconstriction might be relatedto low alveolar Pco2. This was the case, be-cause addition of 5% carbon dioxide to theinspired gas at the moment of inflating theballoon and occluding the pulmonary arteryprevented the increased airway resistance, theshift of alveolar ventilation, and the matchingof gas and blood. Several years ago Sever-inghaus and Stupfel,7 and later Robin andassociates,8 believed that a low P.02 inmixed alveolar gas-relative to the PCO2 ofsimultaneously measured arterial blood-might serve as a sensitive diagnostic test of

40

-I

NKY 40 i;

20.5I46 1

/I--1 40

40

40

22

46

1D40

43

40

Figure 3Efects of unilateral pulmonary artery obstruction.The schema is similar to that in figure 1; the numbersrefer to the CO2 tension in mixed venous blood, al-veolar gas, mixed expired alveolar gas, and arterialblood. (A) Normal lung with uniform distribution ofgas and blood; there is no difference between theC02 tension of arterial blood and mixed expiredalveolar gas. (B) Immediate effects of occlusion of onepulmonary artery; the lung which receives the totalpulmonary blood flow,now has inadequate ventilationto eliminate CO2 and alveolar and arterial CO2 ten-sions increase. (C) Ventilation has increased to bothlungs and maintains a normal arterial blood CO2tension but cannot reduce the difference betweenthe CO2 tension of arterial blood and mixed expiredalveolar gas. (D) Bronchoconstriction has occurred inthe nonperfutsed lung and shifts air from it to theperfused lung. Total ventilation returns toward normaland the difference between the CO2 tension of arterialblood and mixed expired alveolar gas decreases greatly.

Circulation, Volume XXXIII, January 1966

148

40-A,


Dow

nloaded from



pulmonary embolism and might even providea quantitative measure of the proportion ofthe pulmonary circulation that was obstructedby emboli. Unfortunately, it is not so use-ful as predicted, probably because of thiscompensatory increase in airway resistance,which leads to a partial correction of mismatch-ing of blood and gas that would otherwiseinstantaneously follow pulmonary embolism(fig. 3). The decrease in wasted venti-lation helps the patient but hinders the phy-sician in diagnosis. However, if the broncho-constriction is severe enough to be measured,it, instead of the difference in carbon dioxidetensions, can become the clue to pulmonaryembolism.9

Blood ReservoirThe pulmonary vessels normally contain

about 600 ml of blood; most of this is inreadily distensible vessels. This blood andthat in the left atrium together serve as areservoir that supplies blood to fill the leftventricle and maintain its output, even whenthe right ventricular output falls behind fora few beats. Guz and associates'0 have demon-strated this nicely; they placed one electro-magnetic flowmeter on a pulmonary trunk andanother on the ascending aorta so that theycould simultaneously measure pulmonary ar-

AORTICFLOWml/sec

P.A.FLOWml/sec

300 [1

100

0

tery flow (the inflow to the pulmonary circula-tion) and aortic flow (which in a steadystate is the output from the pulmonary circu-lation). In this way they could calculate beat-by-beat the change in pulmonary blood vol-ume. Figure 4 shows the response to completelyblocking the inflow to the pulmonary circula-tion. The stroke volume of the left ventriclecontinued unchanged for two beats and thengradually declined; when the pulmonary cir-culation was congested and its blood volumeconsiderably increased, left ventricular outputwas maintained even longer without any bloodadded to the pulmonary circulation.

NutritionMany physiologists think of the pulmonary

circulation as a set of pipes, delivering bloodto the alveolar capillaries, collecting it, andsending it on to the left atrium and leftventricle, but having nothing to do withnourishing the tissue through which it passes.This view was strengthened by clinical obser-vations: Surgeons had sometimes found itnecessary to tie off one pulmonary artery butstill leave the lung in place. A year or twolater the tissues in such lungs appeared to behealthy. There was, of course, little uptakeof oxygen because there was no pulmonary

il W~A~ ; ftA aft

_.~AJXWWOCcL.AON

Figure 4

Response to occlusion of the main pulmonary artery by inflation of a balloon; the left ventricularoutput remains unchanged for two beats and then gradually decreases.


I

149


Dow

nloaded from


COMROE

circulation but the ventilation of the ischemiclung appeared to be normal. The obvious con-

clusion was that the pulmonary circulation isnot essential for the nutrition of the pul-monary tissues, and, since it is not, the bron-chial circulation must be. We assume thatthe anatomists knew better, but they were

silent. Several years ago we got a rude shock.Finley, Tooley, and associates1" in our Insti-tute wanted to see how long compensatorybronchoconstriction lasted after one pulmo-nary artery was occluded. To the surprise ofno one, occlusion of one pulmonary artery for24, 48, or 72 hours led to a decrease in venti-lation to that lung, but to the surprise ofeveryone, this lung was atelectatic and mar-

kedly congested. Grossly and microscopical-ly, it resembled the atelectatic lung of thenewborn. But still the lung did not die and,with continued occlusion of the pulmonaryartery, the ventilation of that lung increased,its appearance improved, and in 6 to 8 monthsit was reasonably normal in both ventilationand appearance, although somewhat scarred.The time required for its recovery seemed

to be similar to the time required for theestablishment of connections between thebronchial artery and the alveolar capillariesand it is a reasonable assumption that thebronchial circulation, by enlargement of ex-

isting channels or growth of new vessels, comes

to the rescue of the pulmonary tissues.The anatomists are correct that the bronchial

arteries supply the airways down to andincluding the terminal bronchioles and thatthe pulmonary artery supplies the alveoli, al-veolar ducts, and respiratory bronchioles.Nadel and his associates12 confirmed this intheir experiments which showed that drugsthat constrict smooth muscle affected only thealveolar ducts when injected into a pulmonaryartery, and affected only the bronchioles wheninjected into the bronchial arteries.To function properly, the alveoli and al-

veolar ducts require a minimal amount ofblood flow per minute whether it comes fromthe pulmonary artery or the bronchial circu-lation. In the dog with unilateral pulmonaryartery ligation, Tooley and associates (person-

al communication) have estimated that thisminimal flow is about 10 ml per kg of bodyweight per minute. If these estimates are trans-ferred to adult man, the minimal flow wouldbe 700 ml per minute or only about one

seventh of the normal pulmonary blood flow.It is unlikely that total pulmonary blood flowcould ever be low enough during adult life toproduce bilateral pulmonary ischemia and col-lapse, but regional blood flow can be too littleto sustain cell function.When an organ becomes ischemic, those

cells with a high metabolic rate suffer first.The alveolar cells of the lung have many

mitochondria (fig. 5) and a high metabolicrate. They are unusually active in the syn-

thesis of phospholipids and have a high con-

tent of phospholipids, particularly of the sat-urated phospholipids and of dipalmitoyllecithin.'3 The alveolar cells contain intracellu-lar structures unique in human tissues-thelamellar inclusion bodies. It appears at pres-

ent that these contain dipalmitoyl lecithin or

a chemical closely related to it; in some way,

not yet understood, this substance reaches thealveoli and lines them with a monomolecularlayer. In this form it has the unique propertyof decreasing surface tension and preventingcollapse of alveoli during expiration.*

*Why does the surface active material, whichlowers surface tension, prevent collapse of alveoli?We are accustomed to think that the elastic recoilof the lungs is due solely to recoil of stretchedelastic fibers. Actually, it is also due to the surfaceproperties of liquid-air interfaces in the 300,000,000alveoli. Von Neergaard14 showed in 1929 that fluid-filled lungs, which have a fluid-fluid interface, en-large more for a unit increase in transpulmonarypressure than air-filled lungs, which have a fluid-airinterface. Since elastic fibers recoil the same amountwhether alveoli are filled with fluid or filled withair, the additional recoil of air-filled lungs must bedue to the surface tension effects-the tendency oftiny bubbles to become smaller. The surfactant whichnormally lines the alveoli is remarkable in this respect:-As the alveoli get smaller and their surface film iscompressed, the surface tension decreases almost tozero and the recoil due to surface forces almostvanishes.15 The surfactant produced by alveolar cellsis, therefore, an anti-atelectatic factor.

Circglation, Volume XXXIII, January 1966

1,50


Dow

nloaded from



Figure 5Mitochondria and lamellar inclusion bodies of alveolar cell. Epithelial (alveolar) cell of adultmouse lung. A - alveolus; M = mitochondrion; LIB - lamellar inclusion bodies; Nuc -nucleus. (Courtesy of Dr. Michel Campiche.)

When alveolar cells receive less than theminimal blood flow necessary for nutrition,two things can happen: (1) the quality orquantity of surfactant formed is not proper,or (2) pulmonary capillaries become leaky,plasma enters the air spaces, and plasmafibrinogen interacts with pulmonary surfac-tant to cause both an inactivation of surfactantand an inhibition of fibrinolysis.16 The resultis that alveoli collapse and are filled withfibrin threads, the so-called hyaline mem-branes.Although a total pulmonary blood flow less

than the minimal required for nutrition isvirtually impossible during adult life, it is

Circelation, Voluze XXXIII, January 1966

possible in the newborn because there is alarge bypass, the ductus arteriosus, which canaccept all of the right ventricular output ifit is not able to go through the pulmonarycirculation. Ischemia of the newborn lungmight, therefore, be a cause of atelectasis orrespiratory distress syndrome.

In 1964, Drs. Clements, Tooley and Klauswere convinced that the pulmonary surfac-tant was dipalmitoyl lecithin or a chemicalvery similar to it. Because of mounting evi-dence that loss or inactivation of dipalmitoyllecithin was a major factor in respiratorydistress syndrome of the newborn, they andfive associates spent six months studying th-is

151


Dow

nloaded from


COMROE

disease at the Kandang Kerbau Hospital atthe University of Singapore. This hospitalwas chosen because it is by far the largestmaternity hospital in the world and becauseit is affiliated with the University of Californiathrough its International Center for MedicalResearch and Training. At first the group

tried to propel additional dipalmitoyl leci-thin into the lungs of the newborn with re-

spiratory distress syndrome after aerosolizingit. This replacement-type of therapy helpedsome infants but not all. The failures were

probably due to inability to propel enoughmaterial into alveoli that were already col-lapsed, to the inability to obtain a suitablefreon propellent that was free from mildanesthetic properties, and to the presence ofmaterial in the alveoli (probably plasmafibrinogen) which inactivated some or all ofthe dipalmitoyl lecithin that did reach thealveolar surfaces. When they studied babieswho died of respiratory distress syndrome,they found that the pulmonary circulationoffered unusually great resistance to the flow

of fluids through it. They then measured pul-monary blood flow in living babies and foundthat in those with respiratory distress syn-

drome it was only one third of that in healthybabies.'7 18 The increased resistance to pul-monary blood flow was not due to thrombibecause it returned toward normal valueswhen pulmonary vasodilators were infusedintravenously. Intravenous acetylcholine oftenled to an outpouring of previously retainedC02, a decrease in arterial PCO2, an increasein arterial blood pH and P02, and a markeddecrease in wasted ventilation. But in some

babies, sustained pulmonary vasodilatationwas obtained only by intravenous adminis-tration of sodium bicarbonate, thus confirm-ing and findings of Liliestrand,3 and Ensonand associates'9 that acidosis is a potent pul-monary vasoconstrictor. The same combinationof pulmonary ischemia, atelectasis, and hyalinemembranes occurs in babies with respiratorydistress syndrome and in dogs following ex-

perimental pulmonary artery occlusion.No one knows yet whether the pulmonary

ischemia in babies comes first and is causative

or whether it comes later as a result ofhypoventilation, hypoxia, and acidosis. And noone knows whether, if it comes first, it is dueto episodes of fetal asphyxia or fetal hypo-tension or whether it is due to a failure offetal pulmonary vessels to relax normally atbirth. We still have to learn much about themechanisms which, at the birth of a baby,simultaneously cause the pulmonary arteriolesto dilate but the ductus arteriosus and theumbilical arteries to constrict. High oxygenconcentrations are known to dilate pulmonaryarterioles in the fetus and to constrict theductus arteriosus and umbilical arteries evenin vitro, but reflexes may also be involved.

Injection of nicotine (which stimulates boththe aortic and carotid bodies) into the ascend-ing aorta of man causes pulmonary vasocon-striction (Burgess, personal communication).In the dog, stimulation of only the aortic bodyleads to reflex pulmonary vasoconstriction20;presumably lack of stimulation leads to pul-monary vasodilatation. It is of interest that al-though stimulation of either the carotid or theaortic bodies increases respiration, stimulationof the carotid body in the dog produces brady-cardia and hypotension while similar stimula-tion of the aortic body produces tachycardiaand hypertension (fig. 6).21 The aortic bodiesof adult animals receive their blood supplyfrom a branch of a coronary artery or ofthe aortic arch. In the fetus this vesselcommunicates with a small branch of thepulmonary artery; this vascular system issimilar to the ductus arteriosus in that it con-nects the pulmonary artery and the aorta(though the diameter of the channel is farsmaller) and it closes shortly after birth. Wedo not know whether this unusual blood sup-ply has some significance in the reflex controlof pulmonary vascular resistance or resistanceto flow through the ductus arteriosus in thefetus.We do not believe that there is a separate

pulmonary body in the adult which receivesblood from a branch of the pulmonary arterybecause we have never been able to find,beyond the newborn period, a patent arteryto chemoreceptor tissue that arises from the

Circulation, Vorlme XXXIII, January 1966

152


Dow

nloaded from



.4 -Resp. 2i "Flow ,1^.OhA

ec. ./ ½

1 2 3 4

10,ug/kg A.A.

Figure 6Temporal separation of aortic and carotid body stimulation. Above: Nicotine injected througha catheter placed in the aorta just beyond the aortic valves reaches the aortic bodies within1 sec, but because it must pass through long delay paths (coil made of plastic tubing) insertedin the common carotids, does not reach the carotid bodies until 75 sec later. Below: Nicotineinjected at 2 stimulates aortic bodies and causes tachycardia and hypertension; later (at 4)it reaches carotid bodies and causes bradycardia and hypotension. A neuromuscular blockingagent (succinylcholine) was injected at 1 to produce apnea and eliminate any effects ofhyperventilation. Top tracing is respiratory air flow; bottom tracing is blood pressureon the cardiac side of the delay coils. A portion of the record was deleted at 3 to savespace. (From Comroe, J. H., Jr., and Mortimer, L.: J. Pharmacol Exp Therap 146: 33, 1964.)

pulmonary trunk or the right or left pulmonaryarteries.Hughes22 has presented anatomic evidence

that the blood supply of the adult aortic bodycomes from the aorta in a roundabout way.Figure 7 shows that a branch of the left coro-

nary artery first becomes the vasa vasorum

of the wall of the pulmonary artery; theserecombine to form a venous portal systemwhich then breaks up into a second set ofcapillaries which supply the chemoreceptorcells that we call the aortic bodies. Can thecomposition of blood flowing through the pul-Circulation, Volume XXXIII, January 1966

monary artery affect significantly the Po.,and pH of blood flowing through these vasa

vasorum in the wall of the pulmonary artery?Theoretically, it is possible if the diffusionpath is very short and if the blood flowthrough the vasa vasorum is very slow; prac-

tically, it has not yet been possible to stimulatethese chemoreceptors by changing the chemi-cal composition of pulmonary arterial bloodin intact animal.

Filtration

The pulmonary circulation, located as it

1/S4

BPmmHg

153


Dow

nloaded from


COMROE

CAUSES OF DECREASED DIFFUSING CAPACITY

Normal

AlveolarWoll\WI 02

OpenCopillaries Copillary

Thick Interstitial Tissue

Figure 7

Diagram of the vascular connections of the aortico-pulmonary glomus tissue of the cat. (From Hughes,T.: Nature (London) 205: 149, 1965.)

is between the mixed venous blood and thesystemic circulation, serves the importantfunction of retaining fine particles presentin mixed venous blood and preventing thesefrom entering the systemic circulation andlodging in the end-arteries in other, potentiallymore troublesome, sites such as the coronaryor cerebral circulation. There are many morepulmonary capillaries than are needed foreffective gas exchange in resting man andsome of these can be sacrificed to protectother vascular beds; one lung and some ofthe other lung can be removed without pro-ducing anoxemia. It is now possible to esti-mate the size of the functioning pulmonarycapillary bed by special tests that measureuptake of carbon monoxide at normal, high,and low oxygen tensions in inspired gas23; ifthere is no primary alveolar disease, low car-bon monoxide uptake means that there arefewer blood-filled alveolar capillaries in con-tact with ventilated alveoli (fig. 8). Recently,Gold and McCormack (personal communica-tion) have measured pulmonary capillary bloodvolume in 12 patients before and 1 hour afterinjecting radioactive iodinated macro-aggre-gated serum albumin into an antecubital veinfor the purpose of diagnostic lung scanning,as proposed by Wagner and associates.24 Thistest depends upon the plugging of some of thepreviously open pulmonary vessels by the

Thick Alveolar Wall

A A

Decreased CapillaryBlood Volume

0~~~~~~~

Figure 8Causes of decreased diffusing capacity of the lung.Diffusion may be impaired because of a longer pathbetween 02 in the alveolar gas and pulmonarycapillary blood (thick alveolar wail or thick inter-stitial tissue) or becatuse of fewer blood-filled ca-

pillaries.

macro-aggregates of albumin. The maximalamount of albumin injected was mg; assum-

ing that the average diameter of aggregateswas 30 microns, the maximal number of em-

boli would be 70,000, enough to occlude about5% of the precapillary vessels in the lung, ifeach entered a different vessel. As one mightpredict, the injection of these aggregates didnot lead to measurable change in the uptakeof carbon monoxide in any of the patients.However, in another group of patients, Goldand associates25 injected 20 ml of iodized oil(Ethiodol) into the dorsal pedal lymphaticvessels for diagnostic lymphangiography. Thismaterial eventually enters the venous bloodand goes to the pulmonary circulation wherethe oil micro-emboli lodge in the pulmonarycapillaries. Assuming that the average diam-eter of the Ethiodol droplets is 10 microns,20 ml would produce 40 billion emboli-more than enough to block every pul-monary capillary. Gold and his associates


154


Dow

nloaded from



found that the uptake of carbon monoxidediminished after the injection of Ethiodol; onthe average, the pulmonary diffusing capacitydecreased by 32% of control values and pul-monary capillary blood volume by 42% of theinitial figure. Yet at this time these patientshad no symptoms of breathlessness* and nochanges from control values in lung volumesor arterial blood gas tensions.Of great interest is the fact that the dif-

fusing capacity for carbon monoxide and pul-monary capillary blood volume returned tocontrol values within 4 days. Obviously thepulmonary circulation possesses mechanismsfor passing or disposing of macro- or micro-emboli and reestablishing the full number ofopen capillaries. These mechanisms deservemore study because it is likely that mixedvenous blood often contains micro-emboli oftissue cells, fat globules, agglutinated redblood cells, sickle cells, white blood cells,platelets, or parasites that occur normally,follow minor trauma, or occur in disease.tNadel and associates (personal communica-

tion) have recently studied a group of patientswhose main pulmonary abnormalities were adecrease in carbon monoxide (presumablyowing to obstruction of small pulmonary ves-sels) and an increased pulmonary complianceat all lung volumes. Because such a change incompliance has been reported previously onlyin patients with emphysema, they wonderedwhether prolonged block of capillaries (in sucha way that they can be supplied neither bypulmonary nor by bronchial arterioles) pre-cedes and causes loss of alveolar septal tissue.If so, the mechanism for opening plugged capil-laries must be inoperative in these patients.

*Lymphangiography is a potentially dangerous testin patients who have pre-existing disease of thepulmonary circulation and it can be fatal; it is wise,when possible, to measure pulmonary diffusing capac-ity or pulmonary capillary blood volume before inject-ing Ethiodol.

fIn 61 consecutive autopsies at the Beth IsraelHospital in Boston, Freiman and bis associates26found evidence of antemortem thrombi in 64 %; itseems that large or small thrombi or emboli can befound in most older patients if a proper search ismade for them.

Circulation, Volurme XXXIII, January 1966

This means that the factors which inhibit thenormal scavenging of emboli also deservestudy.

I have labeled the filtering of particlesfrom the mixed venous blood as a function ofthe pulmonary circulation-made possible bythe large reserve in pulmonary vascular bedand the apparent harmlessness of some em-boli. We have noted that block of a smallpercentage of precapillary vessels with macro-aggregates of albumin or of a large percentageof the capillaries with Ethiodol causes few ifany measurable changes in respiration or sys-temic circulation. Yet sometimes, pulmonaryemboli cause dysfunction and indeed can befatal. The events occurring after pulmonaryemboli are not necessarily related to the frac-tion of the vascular bed that is occluded.Block of one pulmonary artery by inflating aballoon at the tip of a catheter causes nosymptoms and no important disturbances inventilation or in the systemic circulation if theother lung is normal. But certain types ofpulmonary emboli can cause reflex rapid andshallow breathing,27 constriction of alveolarducts owing to release of histamine frommast cells,28 constriction of bronchiolesowing to liberation of serotonin from bloodclots,29 or systemic effects. Further, there is awell-defined pulmonary chemoreflex in someanimals which, when stimulated by body con-stituents such as serotonin or adenosine tri-phosphate, can cause marked systemic hypo-tension, bradycardia and apnea.30 Andstrangely enough, some mechanoreceptors inthe lungs can be stimulated by certain chemi-cal substances when these are injected intothe pulmonary artery.Much more remains to be learned about

why some micro-emboli do and some do notcause effects in addition to local mechanicalcirculatory obstruction. Is the clue in theirphysical or chemical nature, in the point oftheir impaction, or in secondary chemicalchanges that occur locally after impaction?Even with more animal experimentation, therestill will be the problem of species difference.Figure 9 shows that serotonin injected into thepulmonary artery of a cat produces apnea-

155


Dow

nloaded from


COMROE

Resp.

BPmmHg200-

1BP

mmHg4200-2

Ilo# 100-

Figure 9Effects of serotonin on respiration and blood pressure in the cat (left) and dog (right). Serotonininjected at arrows.

hypotension-bradycardia (from activation ofthe pulmonary chemoreflex) and into adog produces hyperpnea-hypertension-tachy-cardia (from stimulation of the aortic andcarotid bodies); in man, it produces variablechanges.31 32The variability of the clinical and laboratory

manifestations of pulmonary embolism inman have long been puzzling and have led tocontroversy. Some radiologists believe that thetypical roentgenogram of the chest after pul-monary embolism shows an ischemic lung.Others have stated that the typical film showssharply demarcated or wedge-shaped densi-ties (infarcts), often associated with pleuraleffusion. Some radiologists who have reportedthese densities have seen them in the first 24hours after the onset of symptoms, but othersnot until several days have passed. Clinicianshave been puzzled by the fact that one pa-tient can have occlusion of the right or leftpulmonary artery with no effects but anothermay have occlusion of finer vessels and beseriously ill with dyspnea, substernal oppres-sion, cough, hemoptysis, or shock. If we knewthe precise moment of impaction of an em-bolus in man (rather than the moment atwhich symptoms occur), we might see thesequence of bronchoconstriction, ventilatorychanges, ischemia, congestive atelectasis, andthen recovery that would be predicted fromstudies of experimental pulmonary vascularocclusion. But we would not expect this se-quence to occur if the patient's lung containedpre-existing bronchopulmonary anastomosesof any magnitude.

Removal of Alveolar FluidThe pulmonary circulation is designed so

that it can rapidly remove water from thealveoli. The pulmonary capillary pressure ofa healthy man (8 to 10 mm Hg), which tendsto filter fluid from the blood to the alveoli, isnormally far less than the colloidal osmoticpressure of the plasma proteins (25 to 30 mmHg), which tends to pull alveolar fluid intothe blood. This imbalance prevents transu-dation of fluid from blood to alveoli andhastens the reabsorption of alveolar fluid intoblood. When tagged small molecules are dis-solved in water and introduced into the al-veoli, either by injection through a catheteror by inhalation as an aerosol, they enter thecirculation almost as rapidly and completelyas would an intravenous injection. This iswhy administration of procaine or isopro-terenol into the lower airways can be hazard-ous.

It is widely believed that large moleculessuch as the plasma proteins leave alveolislowly and only through lymphatic channels.Recently, it has been shown by Schultz andhis co-workers33 that labeled proteins placedin alveoli enter the pulmonary circulation ofperfused, isolated lungs. The particles couldnot have reached there through the usuallymphatic channels draining into the jugularveins and must have entered the pulmonarycapillaries directly. This has been confirmedby Lee in our Institute (personal commu-nication). The mechanism of the uptake oflarge molecules by pulmonary capillaries re-quires further study.


156


Dow

nloaded from



CommentIn concluding, I have discussed the major

known functions of the pulmonary circulation.I predict that important new functions willbe uncovered in the future. The pancreas hasan exocrine function (long known) and anendocrine function, discovered many yearslater; the kidney has a filtration-reabsorptionfunction (long known) and an endocrine func-tion (formation of renin and hematopoietin)discovered much later. We may find unsus-pected functions of the lung and the pulmo-nary circulation. Why is the lung rich in hista-mine, heparin, adenosine deaminase, and incertain proteolytic enzymes? Why are leuko-cytes, separated briefly from blood and thenreinjected, trapped preferentially in the pul-monary circulation? Why do certain bacteria,given intravenously, stay in fine pulmonaryvessels although they are small enough topass through any capillary bed? Why do cer-tain diseases of the lung lead to inappropriatesecretion of antidiuretic hormone? Why isbronchial carcinoma often associated withclubbing of the fingers although there is noarterial hypoxemia? The answers to thesequestions will be interesting; I hope that youwZill find them.

References1. WEIBEL, E. R., AND GOMEZ, D. M.: Architecture

of the human lung. Science 137: 577, 1962.2. FISHMAN, A. P.: Respiratory gases in the regula-

tion of the pulmonary circulation. Physiol Rev41: 214, 1961.

3. LILJESTRAND, G.: Chemical control of thedistribution of the pulmonary blood flow. ActaPhysiol Scand 44: 216, 1958.

4. STAUB, N. C.: Site of action of hypoxia onpulmonary vasculature. Fed Proc 22: 453,1963.

4a. STAUB, N. C., AND STOREY, W. F.: Relationbetween morphological and physiologicalevents in lung using rapid freezing, J ApplPhysiol 17: 381, 1962.

5. SEVERINGHAUS, J. W., SWENSON, E. W., FINLEY,T. N., LATEGOLA, M. T., AND WILLIAMS, J:Unilateral hypoventilation produced in dogsby occluding one pulmonary artery. J App]Physiol 16: 53, 1961.

6. SWENSON, E. W., FINLEY, T. N., AND GUZMAN,S. V.: Unilateral hypoventilation in man dur-ing temporary occlusion of one pulmonaryartery. J Clin Invest 40: 828, 1961.


7. SEVERINGHAUS, J. W., AND STUPFEL, M. A.: Al-veolar dead space as an index of distributionof blood flow in pulmonary capillaries. J ApplPhysiol 10: 335, 1957.

8. ROBIN, E. D., FORKNER, C. E., JR., BROMBERG,P. A., CROTEAU, J. R, AND TRAVIS, D. M.:Alveolar gas exchange in clinical pulmonaryembolism. New Eng J Med 262: 283, 1960.

9. SASAHARA, A. A., AND STEIN, M. (editors): Pul-monary Embolic Disease. New York, Grune &Stratton, 1965.

10. Guz, A., HOFFMAN, J. I. E., CHARLIER, A.,WEIRICH, W. R., ZANGER, L., AND BENKE, J:Simultaneous observations of right and leftventricular stroke volume in the conscious dog.Fed Proc 20: 133, 1961.

11. FINLEY, T., TOOLEY, W., SWENSON, E., GARDNER,R., AND CLEMENTS, J.: Pulmonary surface ten-sion in experimental atelectasis. Amer RevResp Dis 89: 372, 1964.

12. NADEL, J. A., COLEBATCH, H. J. H., AND OLSEN,C. R.: Location and mechanism of airwayconstriction after barium sulfate microembolism.J Appl Physiol 19: 387, 1964.

13. FELTS, J. M.: Carbohydrate and lipid metabolismof lung tissue in vitro. Med Thorac 22: 89,1965.

14. VON NEERGAARD, K.: Neue Auffassungen fibereinen Grundebegriff der Atemmechanik. Z GesExp Med 66: 373, 1929.

15. CLEMENTS, J. A.: Surface phenomena in relationto pulmonary function: Sixth Bowditch Lec-ture. Physiologist 5: 11, 1962.

16. TAYLOR, F. B., JR., AND ABRAMS, M. E.: Theeffect of surface active lipoprotein on clottingand fibrinolysis and the effect of fibrinogenon surface tension of surface active lipopro-tein. Amer J Med. In press.

17. CHU, J., CLEMENTS, J. A., COTTON, E., KLAUS,M. H., SWEET, A. Y., THOMAS, M. A., ANDTOOLEY, W H.: The pulmonary hypoperfusionsyndrome. Preliminary report. Pediatrics 35:733, 1965.

18. CLEMENTS, J. A.: Surfactant in pulmonary dis-ease. New Eng J Med 272: 1336, 1965.

19. ENSON, Y., GIUNTINI, C., LEWIS, M., MORRIS,T., FERRER, M., AND HARVEY, R.: The influ-ence of hydrogen ion concentration and hy-poxia on the pulmonary circulation. J ClinInvest 43: 1146, 1964.

20. STERN, S., FERGUSON, R. E., AND RAPAPORT,E.: Reflex pulmonary vasoconstriction dueto stimulation of the aortic body by nicotine.Amer J Physiol 206: 1189 1964.

21. COMROE, J. H., JR., AND MORTIMER, L.: Therespiratory and cardiovascular responses oftemporally-separated aortic and carotid bodiesto cyanide, nicotine, phenyldiguanide andserotonin. J Pharmacol Exp Ther 146: 33, 1964.

157


Dow

nloaded from


COMROE

22. HUGHES, T.: Portal blood supply to glomus tis-sue and its significance. Nature (London)205: 149, 1965.

23. ROUGHTON, F. J. W., AND FORSTER, R. E.:Relative importance of diffusion and chemicalreaction rates in determining the rate of ex-change of gases in the human lung. J ApplPhysiol 11: 290, 1957.

24. WAGNER, H. N., JR., SABISTON, D. C., JR.,MCAFEE, J. G., Tow, G., AND STERN, H. S.:Diagnosis of massive pulmonary embolism inman by radioisotope scanning. New Eng J Med271: 377, 1964.

25. GoLD, W. M. YOUKER, J., ANDERSON, S., ANDNADEL, J. A.: Pulmonary function abnormali-ties following lymphangiography. New EngJ Med 273: 519, 1965.

26. FREIMAN, D. G., SUYEMOTO, J., AND WESSLER,S: Frequency of pulmonary thromboembolismin man. New Eng J Med 272: 1278, 1965.

27. WH1TTERDGE, D.: Multiple embolism of thelung and rapid shallow breathing. Physiol Rev30: 475, 1950.

28. COLEBATCH, H. J. H., OLSEN, C. R., ANDNADEL, J. A.: Effect of histamine, serotoninand acetylcholine on the peripheral airways.J Appl Physiol. In press.

29. COMROE, J. H., JR., VAN LINGEN, B., STRouD,R. C., AND RONCORONI, A.: Reflex and directcardiopulmonary effects of 5-OH-tryptamine(serotonin). Amer J Physiol 173: 379, 1953.

30. DAWES, G. S., AND COMROE, J. H., JR.: Chemo-reflexes from the heart and lungs. PhysiolRev 34: 167, 1954.

31. STONE, G. H., AND NEMIR, R., JR.: Study of therole of 5-hydroxytryptamine (serotonin) andhistamine in the pathogenesis of pulmonaryembolism in man. Ann Surg 152: 890, 1960.

32. SKINNER, S. L., AND WHELAN, R. F.: Carotidbody stimulation by 5-hydroxytryptamine inman. J Physiol (London) 162: 35, 1962.

33. SCHULTZ, A. L., GRISMER, J. T., WADA, S.,AND GRANDE, F.: Absorption of albumin fromalveoli of perfused dog lung. Amer J Physiol207: 1300, 1964.

A Thought Twister

. . . In rational problem solving, goals will change not only in detail but in a morefundamental sense through experience with a succession of means-ends and ends-meansadjustments.... In an important sense a rational problem solver wants what he can getand does not try to get what he wants except after identifying what he wants byexamining what he can get.-ALBERT 0. HIRSCHMAN AND CHARLES E. LINDBLOM. Eco-nomic Development, Research and Development, Policy Making: Some ConvergingViews. In: Behavioral Science 7: 218, 1962.


158


Dow

nloaded from


JULIUS H. COMROE, JR.The Main Functions of the Pulmonary Circulation

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1966 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.33.1.146

1966;33:146-158Circulation.

http://circ.ahajournals.org/content/33/1/146.citationlocated on the World Wide Web at:

The online version of this article, along with updated information and services, is

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. and Rights Question and Answer

Permissionsthe Web page under Services. Further information about this process is available in thewhich permission is being requested is located, click Request Permissions in the middle column ofClearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of the CopyrightCirculationoriginally published in Requests for permissions to reproduce figures, tables, or portions of articlesPermissions:


Dow

nloaded from

http://circ.ahajournals.org/content/33/1/146.citation

http://www.ahajournals.org/site/rights/

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


special article the main the pulmonary...

Documents