Journal of Clinical InvestigationVol. 44, No. 9, 1965

Exercise Limitation Following Extensive PulmonaryResection *

A. C. DEGRAFF, JR.,t H. F. TAYLORt J. W. ORD, T. H. CHUANG,ANDR. L.JOHNSON,JR., WITH THETECHNICALASSISTANCEOF LIONEL L.

HARKLEROADANDABRAHAMPRENGLER

(From the Cardiopulmonary Division, Department of Internal Medicine, University of TexasSouthwestern Medical School, Dallas, Texas; and the Lackland Air Force Base

Hospital and San Antonio State Tuberculosis Hospital, San Antonio, Texas)

Cardiac output imposes the principal limit tomaximal oxygen intake in the normal exercisingsubject (1, 2). After pulmonary resection reduc-tion in the ventilatory capacity, the diffusing ca-pacity, or the maximal cardiac output might im-pose a lower limit to maximal oxygen intake. Therelative importance of each of these factors in lim-iting oxygen intake might depend on the amount oflung resected. This study was undertaken to de-termine the primary factor limiting oxygen intakein each of a group of eight patients whose lungremaining following resection varied from 33%to 55%.

Methods

Patients. Table I describes the patients. Resectionalsurgery was performed because of far advanced tuber-culosis. Two of the eight patients had simple pneumo-nectomy; the remaining six had pneumonectomy withcontralateral segmental resection or lobectomy. In cal-culating the amount of lung remaining, each segment wasassumed equal to X of the initial total lung tissue.

Methods. The following measurements were made: 1)maximal oxygen intake for treadmill exercise; 2) lungvolumes sitting; 3) membrane diffusing capacity for car-bon monoxide, lung capillary volume, and pulmonaryblood flow at rest sitting and during treadmill exercise;4) arterial oxygen saturation, oxygen and carbon dioxidepartial pressures, and pH standing at rest and duringtreadmill exercise; 5) right ventricular pressures andpulmonary vascular resistances supine at rest and duringexercise.

* Submitted for publication February 24, 1965; acceptedMay 27, 1965.

This work supported by grants from the U. S. PublicHealth Service (HE 07744 and HE06296).

t Address requests for reprints to Dr. Arthur C. De-Graff, Jr., Dept. of Internal Medicine, University ofTexas Southwestern Medical School, Dallas, Texas75235.

t Work performed during a fellowship from the U. S.Public Health Service.

Maximal oxygen intake (MOI) was determined ac-cording to the method of Mitchell, Sproule, and Chap-man (1) on a motor driven treadmill; oxygen intakewas measured during the last minute of a 2i-minute ex-ercise period. Expired air was collected in a Douglasbag; volume of expirate was measured in a Tissot spi-rometer after a known volume had been removed for gasanalysis. Oxygen and C02 concentrations were measuredby a Beckman paramagnetic oxygen analyzer and by agas chromatograph, respectively. Each patient was stud-ied on successive days until the MOI was determined;treadmill speed was increased at i mile per hour incre-ments until further increments in work load produced nogreater than 54 ml per minute increments in the patient'soxygen intake, or until the patient could not sustain fur-ther increments of work load for the required time.

Vital capacities and timed expiratory volumes weremeasured with a Stead-Wells spirometer. Functionalresidual capacity was measured by the closed-circuithelium method (3).

Apparent CO diffusing capacity (DLco) was measuredduring breath holding, both at a high and at a low alveo-lar oxygen tension so that pulmonary membrane diffusingcapacity (DMco) and capillary blood volume (Vc) couldbe calculated by the method of Roughton and Forster(4). Measurements were made at rest and at MOT.Change in pulmonary blood flow may alter pulmonarycapillary blood volume (5, 6) and apparent CO diffusingcapacity (DLco). Hence, in order to calculate DMco andVc, it is essential that blood flow be the same when DLcois measured at high and low alveolar oxygen tensions.Pulmonary capillary blood flow and DLco were measuredsimultaneously as described by Johnson, Spicer, Bishop,and Forster (5), and measurements of DLco made athigh oxygen tensions were required to have a blood flow(Qc) within 10% of that measured under correspondingconditions at a low oxygen tension. In measurement ofDLco and Qc, test gas mixtures containing 0.5%o Ne,0.3% CO, and 0.3% C2H2 in a balance of oxygen andnitrogen were used. For measurements at rest, the testgases contained 20% and 98% oxygen; for measurementsduring exercise, the test gas mixtures contained 30%oand 98% oxygen. The subjects breathed an oxygen-ni-trogen mixture of similar oxygen concentration to thatin the test gas to be inspired for 2 to 3 minutes beforemeasurements of DLco to insure uniform alveolar oxygen

1514

EXERCISE LIMITATION FOLLOWINGEXTENSIVE PULMONARYRESECTION11

tension during breath holding. Analysis of expired airwas made with a gas chromatograph (7). Normalranges for DMCO, Vc, and Qc at rest and exercise arefrom measurements made in 12 young normal subjects inour laboratory. These data have been previously re-ported in part (8).

Blood was collected for analysis from an indwellingcatheter placed in the brachial artery during uprightexercise at loads varying from rest to MOI. Arterialoxygen saturation was measured by a Van Slyke ap-paratus; pH, arterial PoN, and Pco2 by an Instrumenta-tion Laboratories physiologic blood gas analyzer.

Right heart catheterization was performed with thesubject in the supine position. A portable bicycle genera-tor was pedaled at a constant rate for the 5 minutes ofsupine exercise. Cardiac output was measured by theFick method at rest and during the last 3 minutes ofexercise.

Results

Maximal oxygen intake (MOI). Six of theeight patients reached a maximal oxygen intakeas defined in the Methods. Two, one of whom

S.s1000-

S

(Ccc/mmi)

a02(c/nin.)

500

1000

(cc/min.)soo

1000q°2

(Cctmin.) "

Ii,)0

0 2 4Wofk load (MPH)

TABLE I

Description of patients

LungBody re- Time

surface main- afterPatient Sex Age Height area ing surgery

can m2 % monthsMS F 27 166 1.52 55 2BW F 29 163 1.57 45 3MG F 41 150 1.29 42 36CY M 32 170 1.72 39 29MC F 29 159 1.55 39 26RT M 36 172 1.63 36 24AB F 29 167 1.42 36 46RH F 47 153 1.57 33 47

was trained for 3 weeks, were unable to achieveMOI. Oxygen intakes at graded exercise loadsare illustrated in Figure 1. The highest oxygenconsumption achieved is listed in Table II.

Ventilation. Lung volumes are presented inTable III. Half-second forced expiratory volume

[ .H. 0

_ By6 0 2 4

Work load (MPH)6

FIG. 1. OXYGENINTAKE (Vo2) VS. TREADMILL SPEED. Each discrete pointrepresents a single measurement of oxygen intake. Definite plateaus in oxy-gen intake with increases in treadmill speed indicate valid estimates of maxi-mal oxygen intake (MOI) in six of the eight patients.

M.G.lh--

_S~~~~~

I

Mc.c

I

F

1515

Il

R.T

0

I

DEGRAFF, TAYLOR, ORD, CHUANG,ANDJOHNSON

TABLE II

Maximal oxygen intake

Maximal oxygen intake

Patient Work load* %Pred.t

mph mil/minMS 4.0 957 55BW 5.5 1,246 65MG 6.5 984 70CY 4.5 1,318 51MC 4.5 1,047 52RT 4.0 898 40AB 4.25 724 50RH 3.5 796 43

Average 4.6 996 53

* All patients exercised at the speed indicated, no gradefor 2.5 minutes.Wt Maximal oxygen intake in milliliters per minute perkilogram body weight was predicted from the followingregression formulas derived from Andersen (9, 10):women Vo2 = 42.7 - 0.248 X age; men Vo2 = 50.1 -0.247 X age.

(FEVo.5) was low with respect to the amount oflung remaining. This suggests some degree ofpartial airway obstruction. The maximal breath-ing capacity (MBC) can be roughly predicted fromthe FEVo.5 (13). The comparison between pre-dicted MBCand minute ventilation during maxi-mal exercise illustrated in Figure 2 suggests thatmost patients were operating at or near their ven-tilatory capacities during maximal exercise.

Diffusion. Average capillary blood volume(Vc) was normal at rest with respect to bloodflow and amount of lung remaining, but it in-creased less than would be expected from the in-crease in blood flow during exercise (Figure 3).

MBC CL/min.)

FIG. 2. COMPARISONOF VENTILATION DURING MAXIMAL

EFFORT (VE) WITH MAXIMAL BREATHING CAPACITY

(MBC). VE is plotted against MBCpredicted from thehalf-second forced expiratory volume. In all but twopatients, ventilation during maximal effort equals or ap-

proaches the diagonal line of identity, indicating that atmaximal exercise the patients were breathing at or near

their MBC.

Average membrane diffusing capacity for carbonmonoxide (DMco) at rest and exercise, was lessthan predicted for the amount of lung remaining,but the difference was not statistically significant(Table IV). DMco was reduced most with re-

spect to lung remaining in the two patients whohad the least remaining lung. MOI was reducedmuch more than would be predicted from the re-

duction in DMco in all but two patients, AB andRH (Figure 4).

TABLE III

Ventilatory function

Residualvolume X 100 0.5-sec forced

Total lung Total lung Forced vital expiratory volumecapacity capacity capacity

Patient Meas. %Pred.* Meas. Pred.* Meas. %Pred.* Meas. %Pred.t

ml ml ml

MS 2,678 58 37.7 31.6 1,667 45 997 39BW 3,088 69 38.4 32.3 1,902 53 1,109 45MG 2,761 71 30.5 36.2 1,918 65 663 33CY 3,883 64 39.2 30.1 2,360 50 896 28MC 1,920 45 29.2 32.3 1,359 40 672 29RT 2,460 40 36.2 31.8 1,570 33 750 23AB 2,142 46 53.9t 32.3 986 26 520 22RH 1,939 48 32.3 38.2 1,313 44 643 32

* Predicted values from Needham, Rogan, and McDonald (11).t Predicted from Miller, Johnson, and Wu (12).4 Difference between measured and predicted significant at .05 level.

1516

EXERCISE LIMITATION FOLLOWINGEXTENSIVE PULMONARYRESECTION

Blood pH and gas analyses and alveolar ventila-tion. The results of blood pH and gas analysesare contained in Table V.

Resting measurements of arterial pH and Poo2did not, in any instance, suggest hypoventilation;on the contrary, patients BWand RH appear tohave been hyperventilating, as indicated by alkalo-sis and low arterial Pco2. During treadmill exer-

cise at previously determined maximal loads, av-

erage alveolar partial pressure of CO2 (PAco2)rose minimally, with the highest individual maxi-mal PAOO2when breathing room air being 48 mm

Hg. However, except for MS, whose volume ofexpired air (VE) was 29 L per minute with pre-

dicted MBC53 L per minute, alveolar oxygen ten-sion did not fall with exercise, and average arterialpH fell to only 7.35 (normal 7.19 0.09) (1).

Resting arterial oxygen saturation when breath-ing room air was above 94%o in all patients exceptAB and RH. At rest, most of the difference be-tween the oxygen saturation of blood in equilib-rium with alveolar oxygen tension and the oxy-

c

oLCS 160

1280

cm

E"-I 1400

2

0

0

20

0 2 4 6 8 10 12 14 18

lood Flow (Limin. x Flung)

FIG. 3. PULMONARYCAPILLARY VOLUME VS. CARDIACOUTPUT. Pulmonary capillary volume (Vc) is expressedin milliliters per body surface area per fraction of lungremaining (Fiug), cardiac output in terms of cardiacindex per fraction of lung remaining. The normal range

of capillary volume with increasing blood flow (Qc) isindicated by the heavy lines, resting and exercise pointsfor the patients indicated by the closed and open circles,respectively. Regression line constants for patient andnormal data are indicated in the upper left panel: a=

intercept, = slope, o. = standard error of estimate atmean. The rate of increase of capillary volume with in-creasing blood flow is significantly less than normal, anddespite greater than normal blood flow per unit lungduring exercise, capillary volumes per unit lung do notexceed the normal maximum.

TABLE IV

Lung capillary volume, membrane diffusingcapacity, and blood flow

Lung mem-brane dif-

fusingLung capacitycapil- for CO

laryWork Cardiac Heart blood %

Patient load index rate volume Meas. Pred.*

mpht Limin beats ri ml/ minXBSA min mmHg

MS Rest 2.6 100 42 30 521.5 3.4 109 54 262.5 3.4 112 50 284.0 4.0 156 59 255.0 4.3 156 + 56 36 51

PVC

BW Rest 2.8 116 54 25 45Trained 5.5 6.9 170 67 31 45

MG Rest 2.4 95 54 14 29Trained 4.0 7.2 127 46 28 49

CY Rest 2.8 80 62 32 424.5 5.0 176 83 30 32

MC Rest 2.2 88 50 28 521.5 3.2 1114.0 4.9 175 71 27 44

RT Rest 1.0 124 29 24 314.0 3.3 128 55 27 28

AB Rest 4.8 102 48 18 304.0 6.1 163 56 16 22

RH Rest 2.1 72 17 12 25Trained 3.0 4.1 128 39 10

3.5 5.0 118 + 29 13 21many PVC

* Dmt = 12.13 X height' X K - 2.16 [K = 1.18 for males, 1.10for females (14) ], and DMmx= DMret + 0.2 X Dor.

t No grade.

gen saturation of arterial blood can be explainedby the venous admixture caused by anatomicalshunts (Table V), but during exercise, arterialoxygen saturation fell more than can be explainedby the measured anatomical shunt. Even in thepresence of pulmonary disease, ventilation probablybecomes more uniform with respect to perfusionduring exercise (18), and hence this discrepancybetween measured and calculated saturation prob-ably results from impaired diffusion (15). Thefall in arterial oxygen saturation with exercise wasmost pronounced in AB and RH, the patients whohad the least remaining lung.

Hemodynamics. Blood flow and heart rate dur-ing treadmill exercise are presented in Table IV.The average resting heart rate of 96 beats per min-ute rose to 156 beats per minute at maximal exer-cise (normal 187 ± 10) (1). Although cardiacoutput was low at rest, it was not reduced in pro-

oc CBS4*xkmg BSAxFkg

NO POAT RESmDr-273 11L4 e0=51.824.0 P<.001_a- 71 ± 1.2 13= 2.9t 2.3 P<.006quaM.1(I-2t.0r .895 r .537

0

* 0

- *1Ese_ / O.~~~~~~~~~~~Ezeycase

. ! I' -

1517

DEGRAFF, TAYLOR, ORD, CHUANG,ANDJOHNSON

TABLE V

Blood gas and pH, and alveolar air during rest and exercise*

Sc'o,-Patient Exercise Vo2 FIO2 O20.p Paioo pH PAOI Sc'O2t Pao2 ASc'o2t Sao2

mph!grade/minMS Rest

Rest3.5/0/3.54/0/2.54/0/2.5Cath restCath exercise

BW RestRest5.5/0/2.55.5/0/2.5Cath restCath exercise

MG RestRest7/0/2.57/0/2.5Cath restCath exerciseCath exercise

CY RestRest1.5/0/2.51.5/0/2.54.5/0/2.54.5/0/2.5Cath restCath exercise

MC RestRest4/0/2.54/0/2.5Cath restCath exerciseCath exercise

RT

AB RestRest2.5/0/2.54/0/2.54/0/2.5Cath restCath restCath restCath exerciseCath exerciseCath exercise

RH RestRestRest1.5/0/1.53/0/2.53/0/2.53.5/0/2.5Cath restCath exerciseCath exercise

192

865982

174792

197

1,246§

198590

224§

965§

171818750

267

405

1,318§

196739

198

1,021

171489872

240

530632

201203

607609530

223237

580514

673188369590

0.21 0.180 40 7.42 99 971.00 36 7.42 6570.21 0.180 43 7.34 89 95.40.21 0.190 47 7.28 94 95.31.00 45 7.28 6480.21 0.1540.21 0.171

0.21 0.170 23 7.50 117 98.51.00 24 7.51 6710.21 0.176 24 7.41 124 98.41.00 26 7.38 6690.21 0.1360.21 0.139

0.21 0.200 42 7.42 96 96.91.00 43 7.45 6520.21 48 7.29 102 96.51.00 58 7.27 6370.21 0.1510.21 0.1651.00

0.21 0.210 33 7.42 108 97.71.00 26 7.42 6670.21 7.42 113 98.01.00 36 7.40 6560.21 36 7.28 106 96.51.00 43 7.28 6490.21 0.1830.21 0.189

0.21 0.175 39 7.39 97 96.71.00 36 7.40 6620.21 0.188 41 7.33 101 96.51.00 44 7.29 6540.21 0.1590.21 0.1680.21 0.172

No blood studies

0.21 0.174 42 7.40 88 96.01.00 46 7.38 6490.21 0.171 40 7.38 94 96.40.21 0.183 44 7.40 98 96.91.00 53 7.24 6420.21 41 7.41 94 96.60.21 0.160 37 7.42 100 97.11.00 44 7.40 6510.21 0.169 47 7.30 100 96.30.21 45 7.28 101 96.11.00 51 7.28 644

0.21 36 7.48 96 97.30.21 0.203 35 7.48 97 97.41.00 38 7.48 6540.21 7.48 97 97.40.21 0.213 38 7.45 100 97.41.00 42 7.41 6500.21 36 7.48 103 97.90.21 0.1810.210.21 0.194

* VO2 = oxygen intake in milliliters per minute; FIO2 = fractional concentration of inspired oxygen; 02,ap = oxygen capacity of arterial bloodin milliliters 02 per milliliter blood; Paco2 = arterial partial pressure of CO2 in millimeters Hg; PAO2 = partial pressure of oxygen in alveolar air;Sc'o2 = end lung capillary blood saturation in per cent calculated from PAO2and arterial pH; Pao2 = partial pressure of oxygen in arterial blood;SC'02-ASC'O2 = predicted arterial saturation in per cent; Sao2 = measured arterial saturation.

. From oxyhemoglobin dissociation curve according to Dill (16), at measured alveolar Po2 and arterial pH.Theoretic saturation of mixture of blood in equilibrium with alveolar gas plus shunted blood calculated according to the method of Linderholm

(17).§ Assumed value from previous studies at similar exercise load.

1518

96.1

89.890.8

95.894.1

94.5

94.4

96.095.8

94.4

90.5

90.991.7

96.6

96.0

97.1

96.095.4

96.3

88.7

97.594.291.7

85 96.160566 94.480 94.3

585

89 96.7570

72 97.2595

90 96.5625

75 94.3560

97 95.7529

92 96.7566

86 95.9539

72 96.3637

63 95.6595

60695.3 93.9

88.682.294.3

47995.796.2

60595.295.0

580

80 95.680 95.7

5405358 95.8

53555 96.3

90.4

82.5

93.093.8

86.986.9

84.093.990.190.3

EXERCISE LIMITATION FOLLOWINGEXTENSIVE PULMONARYRESECTION

portion to the amount of lung removed, and bloodflow per unit lung was high, both at rest and exer-cise (Figure 4). [Average cardiac index by theacetylene method in 12 normal subjects was: rest3.5 L/BSA ± 2; 1 maximal exercise 12.0 L/BSA± 2.2.1 Normal blood flow per unit lung is then3.5 L/(BSA X Flung) at rest and 12.0 L/(BSAX Flung) during exercise.2] The results of car-diac catheterization are presented in Table VI.Resting cardiac outputs during cardiac catheteriza-tion are of similar magnitude to those measuredby the acetylene method. Average maximal su-pine output was somewhat lower than maximalupright output, probably as a result of the positionand the type of exercise. The average mean pul-monary arterial pressure of 19 mmHg at restrose to 39 mmHg during supine exercise. Rest-ing pulmonary vascular resistance (pressure XBSA X Flung/blood flow) in the remaining lungwas normal (Figure 5). Exercise resistance wasabnormally high because pulmonary vascular re-

1 Two X standard deviation.2 Flung = fraction of lung remaining.

'a01)

0)

0o

CL

-

0

x0c4_

x

EI,

a

80p

60F

40

20

0 20 40 60 80Membrane Diffusing Capacity for CO

(% predicted maximum)

100

FIG. 4. MEMBRANE DIFFUSING CAPACITY FOR CO(DMco) VS. MAXIMALOXYGENINTAKE. The theoretic re-

duction in maximal oxygen intake resulting from reducedalveolocapillary membrane diffusing capacity (15) isrepresented by the curved line. It is most closely ap-

proached by the two patients with least lung remaining,who also developed arterial desaturation (Sao2) withexercise.

.5

x

"- 7

,c

E

x 6

EE

5a)

-44

-0z

E

2 4 6 8 1o 12Blood Flow (L/min. x m2 x Flung)

14 16

FIG. 5. PULMONARYVASCULARRESISTANCEVS. CARDIACOUTPUT. Pulmonary vascular resistance is expressed asresistance X Flung, where resistance = pressure in milli-meters Hg/CI; CI = cardiac index; Fiung = fraction oflung remaining. Cardiac output is expressed as CI/Fiung.The normal 95% confidence limits are indicated by theheavy solid lines and were derived from 175 measurementsat various exercise levels reported in the English litera-ture (19-26). The normal mean is indicated by thelight line. Closed circles are patients' resting values,open circles measurements during exercise. See text forfurther discussion.

sistance failed to fall with exercise (Table VI,Figure 5) (19-26). However, the duration ofexercise may not have been sufficient to demon-strate a fall in pulmonary vascular resistance, sinceresistance only begins to fall after 3 to 4 minutesof exercise (24, 27). Resting wedge pressureswere normal in two patients. Right ventricularend-diastolic pressure was normal at rest and intwo patients after exercise.

Discussion

Ventilation. During maximal exercise, ventila-tion in seven of the eight patients approached pre-dicted MBC. However, patients who are limitedin exercise capacity as a result of impaired ven-tilation are unable to achieve a maximal oxygenintake (18) as defined by Mitchell, Sproule, andChapman (1), and hence the six patients who wereable to exercise to MOI were not limited in short-term exercise by their reduced ventilatory capaci-

*Rest0 Exercise

100 % 02

FAedicted relationship between/ 'Mco and i2 if t2 limited

by diffusion alone

A

A

AA A

0A

ASao2>88.5%/ Sao2<88.5%l~~~~~~~L

1519

DEGRAFF, TAYLOR, ORD, CHUANG,ANDJOHNSON

TABLE VI

Hemodynamic data from cardiac catheterization

Rightventric-

ular end-dias- Pulmonary Totaltolic Wedge arterial pulmo- Systemic Total

Heart Blood Oxygen pres- pres- pressure nary pressure peripheralPatient State rate flow intake sure sure S/D/M* resistance S/D/M resistance

Llmin dyne-sec- dyne-sec-cm6 cm5

MS Rest 85 4.4 174 22/ 9/15 253 107/69/85 1,550Exercise, air 160 8.5 792 / /38 358 / /120 1,130Exercise, 100% 02 142 8.1 734 / /27 268 155/77/113 1,120Recovery, exer- 96 2 34/15/21

cise, 100% 02

BW Rest 92 2.9 198 6 22/15/17 352 105/70/88 1,800Exercise 131 7.8 590 47/21/30 307 137/90/111 1,142

MG Rest 85 3.9 171 6 31/14/20 310 105/55/79 1,620Exercise, air 155 8.4 818 61/26/43 360 138/80/117 1,110Exercise, 100% 02 133 7.0 750 / /25 229 133/76/103 1,180

CY Rest 78 3.6 196 1 31/ 5/17 373 122/79/96 2,100Exercise, air 123 7.5 739 60/18/36 386 157/88/117 1,250

MC Rest 82 3.7 171 32/10/23 458 116/81/93 2,050Exercise, air 143 7.4 489 53/18/30 326 146/91/117 1,270Exercise, air 162 9.1 872 54/22/38 335 128/72/90 793

RT No cardiac catheterizationAB Rest 132 5.1 202 5 5 30/15/21 329 116/80/95 1,490

Exercise, air 179 6.4 607 59/38/49 608 171/95/128 1,580Exercise, 100% 02 174 6.0 504 54/23/38 511 / /120 1,610

RH Rest 83 3.3 188 4 7 28/12/19 462 108/68/94 2,260Exercise, air 104 4.2 369 36/17/28 526 126/77/96 1,810Recovery, exer- 78 3

cise, airExercise, air 130 6.0 590 51/14/37 497 154/90/122 1,640Recovery, exer- 116 7

cise, afr

* Systolic/Diastolic/Mean.

ties. While the remaining two patients were

breathing room air, alveolar oxygen tension rose

with peak exercise load while arterial CO2 tensionwas not significantly changed from rest, -therebysuggesting that, even in these patients, maximaloxygen intake was not limited significantly by thelow ventilatory capacity. Although impaired ven-

tilation per se in no instance appeared to limitmaximal oxygen intake, increased work of breath-ing could divert an increased portion of oxygenatedblood from muscles of locomotion to muscles ofventilation. This diversion of oxygen to musclesof breathing would not lower the maximal oxygenintake, but could lead to early exhaustion duringexercise.

Diffusion. The capillary blood volume per unitof remaining lung did not exceed normal maxi-mum, despite the fact that peak blood flow perunit lung during exercise was often much higher

than normal (Figure 4). This observation lendssupport to the suggestion made by Johnson, Tay-lor, and Lawson (8) that the maximal potentialvolume of the pulmonary capillary bed is ap-proached in normal subjects as they exercise tomaximal oxygen intakes. An alternative hypothe-sis is that the lung capillaries in these patients were

less compliant than normal capillaries and hencefailed to expand further as blood flow increasedwith exercise.

In a previous paper, Johnson, Taylor, and De-Graff (15) discussed the functional significanceof an impaired alveolocapillary diffusing ca-

pacity, indicating the manner in which patientswith impaired diffusing capacity might be limitedin maximal oxygen intake. A prediction curve

indicating the diffusion imposed limit to MOI ispresented in Figure 4. This limiting curve is ap-proached only by the two patients with the least

1520

EXERCISE LIMITATION FOLLOWINGEXTENSIVE PULMONARYRESECTION

lung remaining, and upon exercise, only they de-veloped arterial oxygen desaturation below thenormal minimum, 88.5%o; hence, oxygen consump-tion during exercise was probably limited in partby impaired alveolocapillary oxygen transfer inthese two patients. Theoretically, if normal ven-tilation and cardiac output are achieved during ex-ercise, diffusing capacity should become a factorlimiting maximal oxygen intake when membranediffusing capacity is reduced below 50%o of nor-mal (15). In our patients, frank alveolar capil-lary block was not manifest until membrane diffus-ing capacity was reduced to less than 30%o of thatpredicted at peak exercise. On the other hand,Cournand, Himmelstein, Riley, and Lester havestudied two patients who had simple pneumo-nectomy during childhood (28); they exhibiteda fall of arterial oxygen saturation to 85%o dur-ing exhausting exercise, and hence may havebeen limited in exercise capacities partly by theirreduced diffusing capacities. The difference be-tween Cournand's and our patients may be re-lated to the fact that our patients were unable toachieve normal maximal cardiac outputs duringexercise.

Hemodynamics. In five of the eight patients,reduced maximal cardiac output imposed the prin-cipal limit to maximal oxygen intake. Linderholm(17) has made a similar observation in patientsafter pneumonectomy. Although impaired myo-cardial function is suggested by the low maximalheart rates, the cause of the reduced maximal car-diac output, whether resulting from myocardialfailure, reduced pulmonary venous capacity, or ob-struction to venous return, was not determined.

Despite chronically increased blood flow perunit lung, pulmonary vascular resistance per unitlung was normal at rest, thereby indicating thateven the adult lung has the capacity to accommo-date itself to a chronic increase in blood flow. Ex-tended follow-up of these patients will be requiredto see whether pulmonary vascular resistance ulti-mately rises, as it does in patients with increasedpulmonary blood onow due to congenital heart dis-ease (29-30).

Summary

Factors limiting oxygen intake were studied ineight patients after recovery from resection offrom 45 to 67%o of their lungs. Maximal oxygen

intake and carbon monoxide diffusion at rest andexercise were measured in all, and arterial bloodgases measured at similar exercise loads in seven.Cardiac catheterization was also performed inseven patients with pressure and flow measure-ments made at rest and exercise. The followingconclusions were reached. 1) Maximal oxygenintake was reduced in all patients, although not inproportion to the amount of lung resected. 2)Although some degree of partial airway obstruc-tion was noted in all patients, in no instance couldthe reduction in maximal oxygen intake be as-cribed to impaired ventilation. 3) Reduced dif-fusing capacity appeared to contribute signifi-cantly to the reduction in maximal oxygen intakeonly in the two patients with least lung remaining.4) Since maximal oxygen intake in five of sevenpatients was limited neither by reduced ventila-tion nor diffusing capacity, it must have beenlimited by reduced cardiac output.

References1. Mitchell, J. H., B. J. Sproule, and C. B. Chapman.

The physiological meaning of the maximal oxy-gen intake test. J. clin. Invest. 1958, 37, 538.

2. Astrand, P., and B. Saltin. Maximal oxygen uptakeand heart rate in various types of muscular ac-tivity. J. appl. Physiol. 1961, 16, 977.

3. Meneely, G. R., and N. L. Kaltreider. The volume ofthe lung determined by helium dilution; descriptionof the method and comparison with other proce-dures. J. clin. Invest. 1949, 28, 129.

4. Roughton, F. J. W., and R. E. Forster. Relativeimportance of diffusion and chemical reaction ratesin determining rate of exchange of gases in the hu-man lung, with special reference to true diffusingcapacity of pulmonary membrane and volume ofblood in the lung capillaries. J. appl. Physiol. 1957,11, 290.

5. Johnson, R. L., Jr., W. S. Spicer, J. M. Bishop, andR. E. Forster. Pulmonary capillary blood volume,flow and diffusing capacity during exercise. J.appl. Physiol. 1960, 15, 893.

6. Daly, W. J., R. A. Krumholz, and J. C. Ross. Thevenous pump in the legs as a determinant of pul-monary capillary filling. J. clin. Invest. 1965, 44,271.

7. Lawson, W. H., Jr., and R. L. Johnson, Jr. Gaschromatography in measuring pulmonary bloodflow and diffusing capacity. J. appl. Physiol. 1962,17, 143.

8. Johnson, R. L., Jr., H. F. Taylor, and W. H. Lawson,Jr. Maximal diffusing capacity of the lung forcarbon monoxide. J. clin. Invest. 1965, 44, 349.

1521

DEGRAFF, TAYLOR, ORD, CHUANG,ANDJOHNSON

9. Andersen, K. L. Cardiopulmonary functional ca-pacity in healthy humans in relation to sex and age.T. norske Lxegeforen. 1963, 83, 227.

10. Andersen, k. L. Physical fitness in relation to ageand sex. Scientific seminar arranged by Riksi-drottsf6rbundets Poliklinik-committe, Sweden.Stockholm, August 23-24, 1962, pp. 15-26.

11. Needham, C. D., M. C. Rogan, and I. McDonald.Normal standards for lung volumes, intrapulmo-nary gas-mixing, and maximum breathing capacity.Thorax 1954, 9, 313.

12. Miller, W. F., R. L. 'Johnson, Jr., and N. Wu. Thehalf-second expiratory capacity test: a convenientmeans of evaluating the nature and extent of pul-monary ventilatory insufficiency. Dis. Chest 1956,30, 33.

13. Miller, W. F., R. L. Johnson, Jr., and N. Wu. Re-lationships between maximal breathing capacityand timed expiratory capacities. J. appl. Physiol.1959, 14, 510.

14. Hepper, N. G. G., W. S. Fowler, and H. F. Helmholz,Jr. Relationship of height to lung volume in healthymen. Dis. Chest 1960, 37, 314.

15. Johnson, R. L., Jr., H. F. Taylor, and A. C. De-Graff, Jr. Functional significance of a low pul-monary diffusing capacity for carbon monoxide.J. clin. Invest. 1965, 44, 789.

16. Dill, D. B. in Handbook of Respiratory Data inAviation. Washington, D. C., National ResearchCouncil, 1944.

17. Linderholm, H. Diffusing capacity of the lungs asa limiting factor for physical working capacity.Acta med. scand. 1959, 163, 61.

18. Pierce, A. K., H. F. Taylor, R. K. Archer, andW. F. Miller. Responses to exercise training inpatients with emphysema. Arch. intern. Med. 1964,113, 28.

19. Dexter, L., J. L. Whittenberger, F. W. Haynes,W. T. Goodale, R. Gorlin, and C. G. Sawyer.Effect of exercise on circulatory dynamics ofnormal individuals. J. appl. Physiol. 1951, 3, 439.

20. Donald, K. W., J. M. Bishop, G. Cumming, and 0.L. Wade. The effect of exercise on cardiac outputand circulatory dynamics of normal subjects.Clin. Sci. 1955, 14, 37.

21. Fishman, A. P., H. W. Fritts, Jr., and A. Cournand.Effects of acute hypoxia and exercise on the pul-monary circulation. Circulation 1960, 22, 204.

22. Freedman, M. E., G. L. Snider, P. Brostoff, S. Kim-belblot, and L. N. Katz. Effects of training onresponse of cardiac output to muscular exercise inathletes. J. appl. Physiol. 1955, 8, 37.

23. Hickam, J. B., and W. H. Cargill. Effect of exerciseon cardiac output and pulmonary arterial pressurein normal persons and in patients with cardiovas-cular disease and pulmonary emphysema. J. clin.Invest. 1948, 27, 10.

24. Holmgren, A., B. Jonsson, and T. Sjbstrand. Cir-culatory data in normal subjects at rest and duringexercise in recumbent position, with special refer-ence to the stroke volume at different work in-tensities. Acta physiol. scand. 1960, 49, 343.

25. Riley, R. L., A. Himmelstein, H. L. Motley, H. M.Weiner, and A. Cournand. Studies of the pulmo-nary circulation at rest and during exercise in nor-mal individuals and in patients with chronic pul-monary disease. Amer. J. Physiol. 1948, 152, 372.

26. Sancetta, S. M., and J. Kleinerman. Effect of mild,steady state exercise on total pulmonary resistanceof normal subjects and those with isolated aorticvalvular lesions. Amer. Heart J. 1957, 53, 404.

27. Widimsky, J., E. Berglund, and R. Malmberg. Ef-fect of repeated exercise on the lesser circulation.J. appl. Physiol. 1963, 18, 983.

28. Cournand, A., A. Himmelstein, R. L. Riley, andC. W. Lester. A follow-up study of the cardio-pulmonary function in four young individuals afterpneumonectomy. J. thorac. Surg. 1947, 16, 30.

29. Dexter, L. Atrial septal defect. Brit. Heart J. 1956,18, 209.

30. Besterman, E. Atrial septal defect with pulmonaryhypertension. Brit. Heart J. 1961, 23, 587.

1522