10
863 Noninvasive Evaluation of Pulmonary Artery Pressure During Exercise by Saline-Enhanced Doppler Echocardiography in Chronic Pulmonary Disease Ronald B. Himelman, MD, Michael Stulbarg, MD, Barbara Kircher, MD, Edmond Lee, MD, Laura Kee, RN, MS, Nathan C. Dean, MD, Jeff Golden, MD, Christopher L. Wolfe, MD, and Nelson B. Schiller, MD With the technical assistance of Salvatore Geraci, RDMS To determine the feasibility of noninvasive determination of right ventricular systolic pressure (RVSP) during a graded-exercise protocol, saline contrast-enhanced Doppler echocardiography of tricuspid insufficiency was performed in 36 patients with chronic lung disease and 12 normal controls. In the patients with chronic pulmonary disease, symptom-limited, incremental supine bicycle exercise and pulse oximetry were performed on and of high-flow oxygen. Technically adequate Doppler studies were initially obtained in 20 patients (56%) at rest and 14 (39%) on exercise; these numbers increased to 33 (92%) and 32 (89%o), respectively, after enhancement with agitated saline (both p<0.001). In 10 patients with chronic lung disease who had simultaneous hemodynamic monitoring during exercise, the correlation between Doppler and catheter measurements of pulmonary artery systolic pressure was close (r=0.98). Among controls, RVSP increased from 22±4 at rest (mean± SD) to 31±7 mm Hg at peak exercise. In patients with chronic lung disease, RVSP increased from 46±20 to 83±30 mm Hg (both p < 0.001 vs. controls). Despite normal resting values for RVSP in 28% of study patients, nearly all showed abnormal increases in RVSP during supine bicycle exercise. Increases in RVSP during exercise were greatest in patients who showed oxyhemoglobin desaturation. The short-term administration of oxygen significantly blunted the increase in RVSP during exercise. Saline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems a potentially valuable noninvasive method of determining the exercise response of RVSP in patients with chronic pulmonary disease. (Circulation 1989;79:863-871) Pulmonary hypertension is a complication of chronic lung disease that is associated with augmented morbidity and mortality.1-4There- fore, identifying pulmonary hypertension at rest or in response to exercise is important for diagnostic, prognostic, and therapeutic purposes. The measure- ment of pulmonary artery pressure during exercise From the Department of Medicine, Cardiology and Pulmonary Divisions, the Cardiovascular Research Institute, and the John Henry Mills Echocardiography Laboratory, University of Cali- fornia, San Francisco. Supported by an Institutional National Research Service Award, HL-07192, Training Program in Heart and Vascular Diseases, National Heart, Lung, and Blood Institute, Bethesda, Maryland. Address for correspondence: Ronald B. Himelman, MD, 1670 Tenth Avenue, San Francisco, CA 94122. Received July 6, 1988; revision accepted November 22, 1988. in patients with chronic lung disease has been accomplished only with invasive techniques involv- ing right heart catheterization.5-8 In patients with chronic obstructive pulmonary disease, exercise often provokes elevations of pulmonary artery pres- sure, pulmonary capillary wedge pressure, and right ventricular end-diastolic pressure.5-8 Doppler echocardiography offers an accurate non- invasive approach to determining right ventricular systolic pressure (RVSP) at rest.9 In patients with chronic pulmonary disease, we used saline contrast- enhanced Doppler of tricuspid insufficiency to deter- mine the exercise response of RVSP while the patients were on and off oxygen. The study was performed to determine 1) the feasibility and accu- racy of noninvasive determination of RVSP during exercise in patients with chronic lung disease, 2) the by guest on June 17, 2018 http://circ.ahajournals.org/ Downloaded from

Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

Embed Size (px)

Citation preview

Page 1: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

863

Noninvasive Evaluation of Pulmonary ArteryPressure During Exercise by Saline-Enhanced

Doppler Echocardiography in ChronicPulmonary Disease

Ronald B. Himelman, MD, Michael Stulbarg, MD, Barbara Kircher, MD,

Edmond Lee, MD, Laura Kee, RN, MS, Nathan C. Dean, MD, Jeff Golden, MD,

Christopher L. Wolfe, MD, and Nelson B. Schiller, MD

With the technical assistance of Salvatore Geraci, RDMS

To determine the feasibility of noninvasive determination of right ventricular systolic pressure

(RVSP) during a graded-exercise protocol, saline contrast-enhanced Doppler echocardiographyof tricuspid insufficiency was performed in 36 patients with chronic lung disease and 12 normalcontrols. In the patients with chronic pulmonary disease, symptom-limited, incremental supinebicycle exercise and pulse oximetry were performed on and of high-flow oxygen. Technicallyadequate Doppler studies were initially obtained in 20 patients (56%) at rest and 14 (39%) onexercise; these numbers increased to 33 (92%) and 32 (89%o), respectively, after enhancementwith agitated saline (both p<0.001). In 10 patients with chronic lung disease who hadsimultaneous hemodynamic monitoring during exercise, the correlation between Doppler andcatheter measurements of pulmonary artery systolic pressure was close (r=0.98). Amongcontrols, RVSP increased from 22±4 at rest (mean± SD) to 31±7 mm Hg at peak exercise. Inpatients with chronic lung disease, RVSP increased from 46±20 to 83±30 mm Hg (bothp < 0.001 vs. controls). Despite normal resting values for RVSP in 28% of study patients, nearlyall showed abnormal increases in RVSP during supine bicycle exercise. Increases in RVSPduring exercise were greatest in patients who showed oxyhemoglobin desaturation. Theshort-term administration of oxygen significantly blunted the increase in RVSP during exercise.Saline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems a potentiallyvaluable noninvasive method of determining the exercise response of RVSP in patients withchronic pulmonary disease. (Circulation 1989;79:863-871)

Pulmonary hypertension is a complication ofchronic lung disease that is associated withaugmented morbidity and mortality.1-4There-

fore, identifying pulmonary hypertension at rest orin response to exercise is important for diagnostic,prognostic, and therapeutic purposes. The measure-ment of pulmonary artery pressure during exercise

From the Department of Medicine, Cardiology and PulmonaryDivisions, the Cardiovascular Research Institute, and the JohnHenry Mills Echocardiography Laboratory, University of Cali-fornia, San Francisco.

Supported by an Institutional National Research ServiceAward, HL-07192, Training Program in Heart and VascularDiseases, National Heart, Lung, and Blood Institute, Bethesda,Maryland.Address for correspondence: Ronald B. Himelman, MD, 1670

Tenth Avenue, San Francisco, CA 94122.Received July 6, 1988; revision accepted November 22, 1988.

in patients with chronic lung disease has beenaccomplished only with invasive techniques involv-ing right heart catheterization.5-8 In patients withchronic obstructive pulmonary disease, exerciseoften provokes elevations of pulmonary artery pres-sure, pulmonary capillary wedge pressure, and rightventricular end-diastolic pressure.5-8

Doppler echocardiography offers an accurate non-invasive approach to determining right ventricularsystolic pressure (RVSP) at rest.9 In patients withchronic pulmonary disease, we used saline contrast-enhanced Doppler of tricuspid insufficiency to deter-mine the exercise response of RVSP while thepatients were on and off oxygen. The study wasperformed to determine 1) the feasibility and accu-racy of noninvasive determination of RVSP duringexercise in patients with chronic lung disease, 2) the

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 2: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

864 Circulation Vol 79, No 4, April 1989

relation between RVSP and arterial hypoxemia dur-ing exercise, and 3) the effect of short-term supple-mental oxygen therapy on these indexes.

MethodsPatient PopulationThe protocol was approved by the Committee for

Human Research at the University of California, SanFrancisco. The patient population consisted of 12control patients (aged 27-68 years, 11 male) with nohistory of cardiopulmonary disease and 36 patientswith moderate-to-severe chronic pulmonary disease(aged 32-80 years, 21 male) from the Fort MileyVeterans Hospital or the University of California,San Francisco, who consented to participate in thestudy. Pulmonary diagnoses were based on the resultsof clinical evaluation, chest radiography, and pulmo-nary function testing. Etiologies of pulmonary dis-ease included chronic obstructive pulmonary diseasein 23 patients, interstitial lung disease in eight (sar-coidosis in four, idiopathic in three, radiation-induced in one), primary pulmonary hypertension intwo, thromboembolic disease in two, and sclero-derma with pulmonary vascular disease in one.Pulmonary function tests, resting arterial blood

gas, and hemoglobin were performed on all patientswithin 1 week of the study. Total lung capacity wasmeasured by a nitrogen washout technique10 at theFort Miley Veterans Hospital and by a heliumdilution method at the University of California;diffusing capacity was measured by a single-breathcarbon monoxide method and corrected for hemo-globin.'1 We excluded patients with known coro-nary artery disease, congestive heart failure, andvascular, orthopedic or neurologic problems of thelower extremities that precluded bicycle exercise.In addition, two patients were excluded after base-line echocardiography showed evidence of a previ-ous inferior wall myocardial infarction in one patientand moderate left ventricular dysfunction with dif-fuse hypokinesis in another. Evaluations were sched-uled when patients were clinically stable; testingwas postponed when clinical exacerbations of symp-toms occurred within the previous 4 weeks.

OximetryAn N-10 or N-200 (R wave-gated) pulse oximeter

(Nellcor, Hayward, California) with continuous read-out was placed on the index finger of the right hand.The right hand was gently immobilized to preventmotion artifact during bicycle exercise. The lowestvalue for arterial saturation during each stage ofexercise was recorded. Arterial desaturation duringexercise was defined as a 4% or greater decreasefrom baseline levels of oxyhemoglobin saturation toless than 90% saturation.12,13

Exercise ProtocolExercise studies were performed at 9:00 AM.

Patients were instructed to eat a light breakfast on

the morning of testing and to take their usualprescribed medications. Of the 36 study patients, 33exercised on room air and oxygen (23 patients on 101/min by nasal cannula and 10 patients on 40%oxygen). The order of testing on room air or oxygenwas randomized. Before each exercise test, patientsbreathed the chosen gas for at least 15 minutes.Between tests, patients rested for 40 minutes. Thefirst 10 patients were aware of which gas they werebreathing; for the final 23 patients, testing wasconducted in a double-blinded fashion.A variable-load supine bicycle ergometer (Engi-

neering Dynamics Corporation Model 8420, Low-ell, Massachusetts) was used for all testing with theergometer table tilted to 200 (patient's left sidedown) to obtain the echocardiographic apical view.Patients pedalled at a constant speed, beginning at aworkload of 10 W and increasing by 10 W every2-minute stage until reaching dyspnea or symptom-limiting fatigue.14 Heart rate, cuff blood pressure,minimum oxyhemoglobin saturation, and maximumtricuspid insufficiency velocity (see below) wererecorded at rest, at each stage of exercise, and after4 minutes of recovery. Continuous lead 2 electro-cardiogram was monitored throughout exercise.The duration of exercise on and off oxygen and thelimiting symptoms were recorded in each patient.

Control patients exercised on room air only, begin-ning at 25W and increasing by 25W every 2 minutes.Control patients exercised to symptom-limiting fatigueor to a maximum workload of 200 W. The sameindexes as noted above were recorded at rest, ateach stage of exercise, and during recovery.

EchocardiographyWith a 3.5-MHz transducer for imaging and a

2.5-MHz transducer for Doppler (Irex Meridian,Ramsey, New Jersey), a complete resting two-dimensional echocardiogram, including imaging andDoppler echocardiograms in all standard views, andsubcostal imaging of the inferior vena cava wasperformed just before exercise.

Ratios of right to left ventricular end-diastolicsizes in the apical four-chamber view (RV/LV size)were determined with a phantom-calibrated com-puterized light-pen (Diasonics, Milpitas, California)by a single-plane area-length method.15 Tracingswere made during held inspiration, when right ven-tricular sizes were the largest. From previous work,the mean value for the ratio of RV/LV size innormal adults is 0.6±0.1.15

Right atrial pressure was estimated at rest by theresponse of the inferior vena cava to deep inspira-tion and was assumed to be constant throughoutexercise. With the trailing-edge to leading-edge tech-nique, maximum inferior vena cava diameters beforeinspiration and minimum diameters after inspirationwere measured in the subcostal view within the first2 cm of the entrance to the right atrium.16 When thediameter of the inferior vena cava (measured at restwith the legs raised onto the bicycle pedals)

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 3: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

Himelman et al Doppler Evaluation of Pulmonary Artery Pressure With Exercise 865

STAGE BGRADIENT

(mm Hg)

1 2 3 4

74 92 100 108

pnI FIGURE 1. Representative single-beat71 contrast-enhanced Doppler recordings

oftricuspid insufficiency at baseline (B),at the end ofeach stage ofexercise (1 to4), and at 4 minutes of recovery (R).Duning exercise, the width of the Dopp-ler signal progressively decreases as

heart rate increases and the maximumtricuspid insufficiency velocity increases.Maximum transtricuspidgradients by themodified Bemoulli equation (in mm Hg)are indicated at the top of the figure.

decreased by less than 50% after deep inspiration,right atrial pressure was defined to be 15 mm Hg;when the diameter decreased by more than 50%,right atrial pressure was defined to be 5 mm Hg.16,17All patients except one showed greater than a 50%decrease in vena cava diameter at rest.

Intravenous injections of 2 to 3 ml agitated iso-tonic saline (contrast) were performed at rest, ateach stage of exercise, and at recovery. Saline wasagitated with two syringes mounted on a three-waystop-cock and injected directly into an antecubitalvein through a 20-gauge catheter. Contrast mediawas injected to delineate the right ventricular cav-ity, to enhance the tricuspid insufficiency Dopplersignal, and to detect right-to-left shunting through apatent foramen ovale.Doppler evaluation of the tricuspid valve was

performed in the apical four-chamber view. Theregurgitant jet was localized in the pulsed mode,then maximized and measured in the continuouswave mode at a sweep speed of 75 mm/sec. By amodification of the Bernoulli formula,18 the maxi-mum transtricuspid gradient (in mm Hg) was esti-mated as the product of 4 and the square of themaximum tricuspid insufficiency velocity (AP = 4V2).RVSP was computed as the sum of the transtri-cuspid gradient and right atrial pressure (RVSP4V2+right atrial pressure).9Throughout exercise, continuous full-screen mon-

itoring of tricuspid insufficiency at a sweep speed of75 mm/sec was performed. During each 2-minutestage of incremental exercise, agitated saline wasinjected intravenously to enhance the Doppler sig-nal. We considered tricuspid insufficiency enve-lopes to be technically adequate when signals werepansystolic and showed well-defined borders. Aftervideotape review, the maximum tricuspid insuffi-ciency velocity was assigned to the highest coher-ent boundary on the spectral wave form.9 Maximumvelocities recorded at rest, at each stage of exer-cise, and during recovery permitted the calculationof the transtricuspid gradient and RVSP throughoutthe test (Figure 1).To ensure accurate transducer position through-

out exercise, the location of the continuous waveDoppler cursor was checked intermittently by two-dimensional imaging. Tricuspid insufficiency was

distinguished from aortic outflow or mitral regurgi-tation by 1) the location of the Doppler cursorwithin the boundaries of the right heart, 2) thecharacteristic high-pitched audio signal, 3) the char-acteristic sound and appearance of contrast enhance-ment of the tricuspid insufficiency profile (left heartDoppler signals will not show contrast enhance-ment in the absence of a right-to-left intracardiacshunt), and 4) the comparison to the contours of theresting Doppler envelopes.

Invasive TestingOf the 36 patients with chronic pulmonary dis-

ease in the study, 10 had flow-directed pulmonaryartery catheters, and 10 had radial or brachialarterial catheters in place at the time of exercisetesting; these catheters had been inserted for rea-sons unrelated to this study.

Etiologies of chronic pulmonary disease in the 10patients who exercised with flow-directed pulmo-nary artery catheters in place were obstructivedisease in five, primary pulmonary hypertension intwo, thromboembolic disease in two, and interstitialdisease in one. In these 10 patients, Doppler esti-mation of RVSP during exercise was correlatedwith simultaneous determination of pulmonary arterysystolic pressure by catheter. Correct placement ofthe pulmonary artery catheter was confirmed bypressure tracings and chest radiograph. Zero refer-ence was midthorax at 200 leftward tilt on theexercise ergometer. Measurements of RVSP andpulmonary artery systolic pressure by two experi-enced, independent observers were compared forthe average of five consecutive, simultaneous heartbeats that showed technically superior Dopplersignals at rest, at the end of each stage of exercise,and at 2-4 minutes of recovery.To correlate the values for calculated and mea-

sured oxyhemoglobin saturation, arterial blood gaseswere drawn at rest and at peak exercise. Calculatedoxyhemoglobin saturation was derived from arterialoxygen tension with Severinghaus equations19; mea-sured oxyhemoglobin saturation was determined bypulse oximetry.

Data AnalysisFor all variables, the mean rest value was com-

pared with the mean exercise value by the paired t

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 4: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

866 Circulation Vol 79, No 4, April 1989

TABLE 1. Pulmonary Function Data in Patients With ChronicObstructive and Interstitial Pulmonary Disease

Obstructive disease Interstitial disease

Patients (n) 23 8FEVI (1) 1.0±0.5 (35) 1.7±0.8 (73)FVC (1) 2.8± 1.1 (73) 2.0±0.8 (64)FEV1/FVC 0.36+0.15 0.85±0.31TLC (l) 6.9±1.0 (107) 4.1±0.1 (74)DLCO (ml/min/mm Hg) 21±13 (71) 16+6 (63)Pao2 (mm Hg) 67+8 64+21PaCO2 (mm Hg) 42± 10 44 ±8Hemoglobin (g) 16.5±1.5 16.4±2.3

Data are mean ± SD. Numbers in parentheses are averagepercentages of predicted values.FEV1, forced expiratory volume in 1 second; FVC, forced

vital capacity; FEV1/FVC, ratio of forced expiratory volume in1 second to forced vital capacity; TLC, total lung capacity;DLCO, diffusing capacity for carbon monoxide; Pao2, arterialoxygen tension on room air; Paco2, arterial carbon dioxidetension on room air.

test. Data on patients with chronic lung diseasewere compared with data on controls by analysis ofvariance. In patients with chronic lung disease,values for RVSP at rest and at peak exercise thatexceeded control means by more than 2 SD wereconsidered abnormally elevated.

Exercise data on room air were compared withdata on oxygen by the paired t test. Because exerciseduration was usually greater on oxygen than on roomair, comparisons were made at identical workloads:at rest (0%), 20%, 40%, 60%, 80%, and 100% of peakworkload on room air, and at 4 minutes of recovery.

Patients were subdivided into subgroups basedon 1) the presence or absence of pulmonary hyper-tension at rest, or 2) the presence or absence ofoxyhemoglobin desaturation with exercise. Analy-sis of variance was performed separately on thedata from these subgroups. Correlations amongpulmonary function and exercise data were deter-mined with least-squares linear regression analysis.Data are listed as mean±SD. The null hypothesiswas rejected at the 5% level.

ResultsPulmonary Function Tests

Results of pulmonary function testing in the 23patients with chronic obstructive pulmonary diseaseand in eight with interstitial lung disease are shown inTable 1. At the time of testing, no patient had aresting arterial oxygen tension below 55 mm Hg.

Technical Adequacy of Two-Dimensional andDoppler EchocardiographyOf the 36 two-dimensional echocardiograms in

study patients at rest, 33 (92%) had apical four-chamber views of sufficient quality to permit themeasurement of RV/LV size.Of the Doppler echocardiograms, technically ade-

quate tricuspid insufficiency Doppler signals wereinitially obtained in only 20 patients (56%) at rest

and 14 (39%) while exercising. After contrastenhancement of the Doppler signals (Figures 2Aand 2B), these numbers increased to 33 (92%) and32 (89%), respectively (bothp < 0.001). In the remain-ing patients, inability to clearly detect tricuspidinsufficiency precluded calculation of RVSP at restor exercise, despite multiple attempts at contrastenhancement in various echocardiographic win-dows. Technically adequate Doppler signals wereinitially obtained in seven of the 12 control patientsat rest and 5 of 12 on exercise; after contrastenhancement, these numbers increased to 11(p=NS) and 11 (p<0.05), respectively.

Contrast studies at rest after Valsalva maneuverrevealed right-to-left shunting through a patent fora-men ovale in two patients; both also showed shunt-ing at peak exercise. No other patient showedshunting either at rest or at peak exercise.

Limiting SymptomsOf the 36 study patients, 26 stopped exercise

because of dyspnea and 10 because of leg fatigue.No patient had chest pain, ischemic electrocardio-graphic changes, or significant arrhythmias.

Correlation of Noninvasive and Invasive DataThe correlation between RVSP by contrast-

enhanced Doppler and pulmonary artery systolicpressure by pulmonary artery flotation catheter atrest and exercise in 10 patients (30 simultaneousmeasurement points) was close (r= 0.98, p <0.0001,Figure 3). Doppler tended to slightly underestimatepulmonary artery systolic pressure as measured bycatheter, especially at peak exercise (RVSPDoppler=0.88 x pulmonary artery systolic pressurecatete. + 4, Fig-ure 3). Mean right atrial pressure by Swan-Ganzcatheter rose from 7 ± 4 at rest to 14 ± 7 mm Hg atpeak exercise (p< 0.01).The correlation between calculated and measured

values for oxyhemoglobin saturation at rest andpeak exercise in 10 patients (20 simultaneous mea-surement points) was also close (r= 0.97, p <0.0001,Saturationmeasured = 1.00 x Saturationcalculated + 1.4).

Exercise Performance of ControlsExercise data on room air for control patients are

presented in Table 2. Mean RVSP at rest was 22 ± 4mm Hg (range, 15-29 mm Hg); mean RVSP at peakexercise was 31±7 mm Hg (range, 25-44 mm Hg).The upper limits of normal values for RVSP at restand exercise were defined as two standard devia-tions greater than control means; these were 30 and45 mm Hg, respectively.

Exercise Performance of Patients on Room AirFor the 32 patients with technically adequate Dopp-

ler studies of tricuspid insufficiency at rest andexercise, mean room air RVSP at rest was 46±20and increased to 83 ± 30 mm Hg at peak exercise(p <0.001). Patients exercised 9.2 ± 5.8 minutes to aworkload of 44 ± 27 W. Mean heart rate increased by

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 5: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

Himelman et al Doppler Evaluation of Pulmonary Artery Pressure With Exercise 867

38 beats/min, systolic blood pressure increased by 47mm Hg, and oxyhemoglobin saturation decreased by13% to a mean nadir of 84±11% (Table 2).

Exercise Performance of Patients on Oxygen

Although the duration of exercise was greater onoxygen than on room air in all but three patients(p<0.01), differences in exercise time were oftensmall. Increases in workload, heart rate, and sys-tolic blood pressure were not significantly differenton room air and oxygen (Table 2).Although RVSP at peak exercise on oxygen was

not significantly lower than RVSP at peak exerciseon room air, RVSP on oxygen was significantlylower when matched for equivalent workloads. Thedifferences between mean values for RVSP onoxygen and room air increased as workload pro-gressed and became significant at 80% and 100% ofpeak exercise (Figure 4).

Comparison of Patients With and Without RestingPulmonary HypertensionOf the 32 patients with technically adequate tri-

cuspid insufficiency Doppler signals at rest and

FIGURE 2. Single-beatcontrast-enhanced Doppler recording of

} tricuspid insufficiency in apatient.with chronic obstructive pulmo-

- ..nary disease. A technically ade-quate tricuspid insufficiency jet

= could not be obtained by routineresting Doppler evaluation (JA),

......but /sec jJet as clearly:::::::::<:et:Q00:0;>Ei>:i tS;; .g^ dS:obtainedafter the injection of.........agitated saline (lB arrow).

exerc:i;ise: nine (28%) had normal resting values for

RVSP.on..room ai (<3 mmHg,grop A, ad2

groups butgretethan......otincontrol p tie intse(pi<0.05)

At pea exrcs on roo air RVSP in see of

levels.(>45 mm Hg, Figred 5) BCmaredwitgroup. A patients those in gr p B hd s an:.. ....5:-:-- ttW

heoloi (14E . vs.s 16.9± 1.9 g) an R /L

s (05. 0 2vs. 1.2.1.0. allp<0...0

Tw nt t'RSE'iwo patient (61% showed oxhe ogobindesaturaEE seeetio on roo ai exrcs (dces in sSees K

u i by. more...........a 4 t.o l.. fZE ::... ... .... 'B#'8'i..E Ei '' E

... .. ............. ... ...

.... ..... ............%..............::E........

....iE" ..'; ......'.f'." :'' '('''. '"'. 't .'.g- ': '.> ..''5 .''SRsiR.,R-*i,#: g:E......EL

\. ..: R .....R .............:RERS SSPEf-E-:i -- :fiER ER...,* .8 :,#%o .....,itR--,f-Z E -E R EER) #EtRRE ER EESE * S

....

......... ..............R:R:...........

0~~

.....................

RVSP~~~~~~~~~~~~~~. ......................ou ) an 2...in ....ni....enio..(group

RVS wit ....is wa....r ewenth wgrup but ....te ..a incnro..iet p o

........nine patients ln group A increased to abnormal.......r

exerise, nine+ v.64+8 had normandrestir values forRVSPoglbnroom>air 16.9+1.9Hggrou a), and/LV

Compadio resin pulmonaryWihyperWtensonut ru

RVSPr withDexerctisewssmlrbewe h w

grousebty-two tanicntopatients(61%shwe oxyhe oglobin pesaktuxratiseon room air,RSinrcs(derens ofnineupatient in grouptan inresetolestabnormal

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 6: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

868 Circulation Vol 79, No 4, April 1989

Exercise Doppler on Room Air vs Oxygen

120

8040

C' 40 /

0 40 80 120 160

PASP Catheter (mm Hq)

FIGURE 3. Plot ofcorrelation betweenpulmonary arterysystolic pressure (PASP) by pulmonary artery flotationcatheter and tight ventricular systolic pressure (RVSP)by contrast-enhanced Doppler of tricuspid insufficiencyat rest and during exercise. In 10 patients, 30 simulta-neous measurements were obtained over a wide range ofpulmonary pressures (r = 0. 98).

saturation). Compared with the patients who haddesaturation with exercise on room air, those whoshowed desaturation had lower values for diffusingcapacity and arterial oxygen tension and had highervalues for RV/LV size ratio, RVSP at rest, RVSP atpeak exercise, and increase in RVSP with exercise(Table 3). Differences in duration of exercise, work-load, heart rate, and systolic blood pressure onroom air were not significant (Table 3).

Correlations Among PulmonaryFunction and Exercise Data

Significant correlations did not occur betweenpulmonary function indexes and RV/LV size ratio,RVSP at rest, RVSP at peak exercise, or increase inRVSP with exercise. However, modest correlationsoccurred between RV/LV and RVSP at rest (r= 0.59)or exercise (r=0.69, both p<0.001) and betweenRVSP at rest and peak exercise (r= 0.77,p < 0.0001).

DiscussionMeasurement ofRight Ventricular SystolicPressure (RVSP) During ExerciseWe have shown the potential value of the nonin-

vasive measurement of RVSP as an estimate of

c

4-

;0

00

0E

CD

be

E

E0LQ

Li.0

o

U,

100 -* *

9 Rest Recovery

90-

851 Sat on Air/Set on 02

-.- Peck Ex. on 0280 . .

0 20 40 60 80 100 130

_ RVSP on AirRVSP on 02

-*- Peek Ex. on 02

Recovery

0 20 40 60 80 100 130of Room Air Peak Workload

FIGURE 4. Plots of comparisons of levels of oxyhemo-globin saturation (Sat, top graph) and right ventricularsystolic pressure (RVSP, bottom graph) during exercisein chronic lung disease patients on room air (blacksquares) and oxygen (open squares). Data are comparedat identical workloads, expressed as the percentage ofworkload performed on room air (0-100%). Large open

diamonds indicate the values atpeak exercise on oxygen(at an average of 130% of workload on room air). Errorbars indicate standard errors; *p<0.05 vs. saturation atsame level of workload on room air; - p < 0. 05 vs. RVSPat same level of workload on room air.

pulmonary artery pressure at rest and during exer-cise in patients with chronic lung disease. Despitethe technical limitations of standard echocardiogra-phy and Doppler in these patients, saline contrastenhancement of tricuspid insufficiency permits the

TABLE 2. Results of Exercise Testing on Room Air and Oxygen

Chronic pulmonary disease

Room air Oxygen Oxygen equivalent ControlPeak workload (W) 44±27 56±30 . . . 158±49Change in heart rate (beats/min) 38±20 40±21 34± 14 75±28Change in SBP (mm Hg) 47± 29 43 ±22 39±21 66±29RVSP rest (mm Hg) 48± 19 44± 19 . .. 22±4RVSP peak exercise (mm Hg) 86±29 78±28 66±18* 31±7Change in RVSP (mm Hg) 39±20 35±16 26±13* 11±6Nadir oxyhemoglobin saturation (%) 84±11 93±6* 92±4* 96±3

Data are mean± SD.Oxygen equivalent, data on oxygen at equivalent workload to peak exercise on room air; SBP, systolic blood

pressure; RVSP, right ventricular systolic pressure by Doppler.*p<0.05 vs. chronic pulmonary disease on room air. All entries in first three columns differ significantly from

controls except nadir of oxyhemoglobin saturation on oxygen or oxygen equiv.

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 7: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

Himelman et al Doppler Evaluation of Pulmonary Artery Pressure With Exercise 869

100-

RVSP (mm Hg) 75 -

50-

25-

150

125

100-

RVSP (mm Hg) 75 -

50 -

25 -

Exercise Doppler on Room Air150-

Group A Patients

0

Group A Baseline Group A Peak Exerase

150-

125'

100-

RVSP (mm Hg) 75'

50 -

25 -

0'

Group B Patients

Group B Baseline Group B Peak Exercise

FIGURE 5. Plots of increase in right ventricular systolicpressure (RVSP) fiom rest topeak exercise on room air in11 normal controlpatients (top graph), nine patients withnormal (<30 mm Hg) resting values for RVSP (group Apatients, middle graph), and 23 patients with pulmonaryhypertension at rest (group B patients, bottom graph).The extent ofincrease in RVSP with exercise is similar ingroup A and B patients and is significantly greater thancontrol pattents. Superimposition of data points hasobscured some individual values.

determination of RVSP in a high percentage ofpatients. Intravenous injection of agitated isotonicsaline has been previously used to enhance right-heart Doppler signals at rest with no adverseconsequences.15,20-22 During exercise, we chose toperform continuous Doppler evaluation of the tri-cuspid valve alone, because we found that attemptsto record signals from more than one valve impairedthe ability to gather high-quality data.

In this study, RVSP determined by Doppler atrest and exercise compared favorably with simulta-neous hemodynamic measurements. In patients withchronic lung disease in whom were placed indwell-ing flow-directed pulmonary artery catheters, pul-monary arterial systolic pressures were used as a

close approximation of right ventricular systolicpressures. No patient had known pulmonary valvedisease. Because the catheter may, itself, causetricuspid insufficiency, we cannot exclude the pos-sibility that the catheter improved the quality of the

TABLE 3. Comparison of Patients With and Without Oxyhemo-globin Desaturation at Rest and During Exercise on Room Air

DesaturatioiNondesaturation

Patients (n)Pulmonary functionDLCO (ml/min/mm Hg)Pao2 (mm Hg)Paco2 (mm Hg)

Hemoglobin (g)Echo/Doppler at restRVILV size ratioRest RVSP (mm Hg)

Echo/Doppler at peak exerciseExercise duration (min)Workload (W)Increase heart rate (beats/min)Increase SBP (mm Hg)Nadir oxyhemoglobin

sat (%)Peak RVSP (mm Hg)Increase in RVSP (mm Hg)

22 14

10.0± 4.7 (47)17.3 ± 6.4* (75)63±8 76±8*40±11 39±4

16.9±1.9 15.8±1.5

1.26±1.11 0.51 ±0.22*56±1 32± 12*

8.1±4.4 10.9±7.8

38±18 54±35

38±16

43±30

79±11

39±25

53±27

91±4*

98±30 61±10*

44±23 31±12*

Data are mean ± SD. Numbers in parentheses are averagepercentage of predicted values.DLCO, diffusing capacity, Pao2, arterial oxygen tension; Paco2,

arterial carbon dioxide tension; RV/LV, ratio of right to leftventricular size in apical four-chamber view by single-planearea-length method; RVSP, right ventricular systolic pressure byDoppler; SBP, systolic blood pressure.

*p<0.05 vs. patients with desaturation.

Doppler envelope in these patients. Although RVSPby contrast-enhanced Doppler showed an excellentcorrelation with pulmonary artery systolic pressureby catheter, Doppler underestimated catheter val-ues by an average of 12%. This underestimationtended to be greater at peak exercise than at restand may reflect 1) technical difficulty in aligning theultrasound beam parallel to the tricuspid insuffi-ciency jet near peak exercise because of tachypneaand tachycardia, and 2) the error encountered inassigning a constant value to right atrial pressurethroughout exercise (in the patients with indwellingcatheters, mean right atrial pressure increased from7 to 14 mm Hg during exercise). The latter problemcould be improved in future studies by allowing foran increase in right atrial pressure with exercise.

In this study, we found that RVSP during exerciseincreased to a significantly greater extent in patientswith chronic lung disease than in normal controlpatients. The pulmonary pressures estimated byDoppler in this study at rest and exercise are similarto hemodynamic data previously reported in patientswith chronic lung disease.5 823 We also noted amodest correlation between right ventricular sizeand RVSP at rest or exercise. Although patients withnormal resting pulmonary pressures had significantlysmaller right ventricles than those with elevatedpressures, both groups showed abnormal rises inpulmonary pressure during exercise. Therefore, Dopp-ler exercise testing may help identify a group of

125-I Normal Control Patients

O . ,Normal Baseline Normal Peak Exercise

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 8: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

870 Circulation Vol 79, No 4, April 1989

patients with chronic lung disease who have a nor-mal resting RVSP and right ventricular size, yet whohave abnormal increases in RVSP with exercise.

Effect of Short-term Oxygen Administrationon Right Ventricular Systolic PressureDuring ExerciseThe graded exercise protocol used in this study

has been advocated as a relatively safe and reliablemeans of ascertaining the maximum workload inpatients limited by respiratory symptoms. Patientswere exercised on and off high-flow oxygen. Toavoid either a consistent training effect or a pulmo-nary vasodilating effect of a first exercise test,24,25 theorder of exercise testing on room air or oxygen inthis study was randomized. In addition, testing wasconducted in a double-blinded fashion regarding thegas breathed in most studies. Oxygen was chosenover other pulmonary vasodilators because of itssafety and possible therapeutic utility; high flowrates were administered in an attempt to prevent thedevelopment of arterial hypoxemia during exercise.Although the short-term administration of high-

flow oxygen increased exercise duration in mostpatients, mean heart rate and cuff systolic bloodpressure were not significantly different during exer-cise on room air and oxygen. At identical peakworkloads, the increase in RVSP during exercisewas significantly blunted on oxygen compared withroom air. Although the patients in this study did nothave resting hypoxemia, our findings are similar tothose of a recent hemodynamic report of hypox-emic patients with chronic obstructive pulmonarydisease, in whom short-term oxygen administrationresulted in a small but significant decrease in meanpulmonary artery pressure at peak exercise.23

In patients with chronic pulmonary disease, theabnormal increase in pulmonary artery pressure dur-ing exercise has been attributed to destruction of thepulmonary capillary bed, resulting in limitation ofpulmonary vascular reserve.25 In addition to fixedpulmonary vascular disease, reactive pulmonary vaso-constriction due to exercise-induced hypoxia maycontribute to elevations in pulmonary pressure. Theshort-term administration of oxygen corrects hypox-emia and may also improve ventilatory muscle func-tion, alleviate right ventricular ischemia, and decreaseright ventricular afterload.26-30 Because cardiac out-put and pulmonary vascular resistance were notmeasured in most study patients, the relative contri-butions of each of these factors to the decrease inexercise RVSP on oxygen cannot be estimated.However, data from the Nocturnal Oxygen TherapyTrial Group23 showed that exercise on oxygen com-pared with room air was associated with a decreasein mean pulmonary vascular resistance at peak exer-

cise with little change in cardiac output.

Oxyhemoglobin SaturationSimilar to data in previously reported studies,31'32

the patients who had desaturation had significantly

lower values for diffusing capacity and resting oxy-gen tension than those who did not have desatura-tion. In addition, patients who had desaturation hadlarger right ventricles and higher values for RVSP atrest and peak exercise than patients who did nothave desaturation. Right-to-left shunting through apatent foramen ovale at rest or exercise was presentin only two patients (one desaturated) and thus wasnot an important factor in the development ofarterial hypoxemia during exercise.The development of pulmonary hypertension in

chronic lung disease may be related to repetitiveepisodes of transient hypoxic vasoconstriction.A2,13We have previously shown that only 21% of patientswith severe chronic obstructive pulmonary diseasewho are not hypoxemic at rest undergo desaturationduring sleep and that the degree of right ventricularenlargement and pulmonary hypertension in thesepatients is not greater than in those who do notundergo desaturation.15 In comparison, 61% ofpatients in the present study showed oxyhemoglo-bin desaturation during exercise. Unlike the resultsof the sleep study, the patients in this study whohad desaturation during exercise had significantlylarger right ventricles and higher pulmonary pres-sures at rest and peak exercise than those who didnot have desaturation.

SummaryWe have shown that saline contrast-enhanced

Doppler is potentially valuable for the noninvasiveestimation of pulmonary artery pressure duringexercise in patients with chronic lung disease.Patients who showed arterial oxyhemoglobin desat-uration during exercise had larger right ventriclesand higher pulmonary pressures at rest and exercisethan patients who did not show desaturation. Despitenormal resting values for RVSP in 28% of patientsin the study, nearly all showed abnormal increasesin RVSP during supine bicycle exercise. In addi-tion, the short-term administration of oxygen bluntedthe increase in RVSP during exercise.

AcknowledgmentWe are indebted to Nellcor (Hayward, California)

for providing the pulse oximeters used in the studyand to the technicians in the echocardiographylaboratory for their invaluable assistance.

References1. Stevens PM, Terplan M, Knowles JH: Prognosis of cor

pulmonale. N EnglJMed 1963;269:1289-12912. Fishman AP: Chronic cor pulmonale. Am Rev Resp Dis

1976;114:775 -7943. Mitchell RS, Webb WC, Filley GF: Chronic obstructive

bronchopulmonary disease: III. Factors influencing progno-sis. Am Rev Resp Dis 1964;89:878-896

4. Renzetti AD, McClement JH, Litt BD: The Veterans Coop-erative Study of Pulmonary Function: III. Mortality inrelation to respiratory function in chronic obstructive pul-monary disease. Am JMed 1966;41:115-129

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 9: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

Himelman et al Doppler Evaluation of Pulmonary Artery Pressure With Exercise 871

5. Burrows B, Kettel LJ, Niden AH, Rabinowitz M, DienerCF: Patterns of cardiovascular dysfunction in chronic obstruc-tive lung disease. N Engl J Med 1972;286:912-917

6. Brent B, Berger H, Matthay RA, Mahler R, Pytlik L, ZaretBL: Pathophysiologic correlates of right ventricular ejectionfraction in chronic obstructive pulmonary disease: A combinedradionuclide hemodynamnic study.AmJCardiol 1982;50:255-262

7. Horsfield K, Segel N, Bishop JM: The pulmonary circulationin chronic bronchitis at rest and during exercise breathing airand 80% oxygen. Clin Sci 1968;43:473-483

8. Morrison D, Goldman S, Wright AL, Henry R, Sorenson S,Caldwell J, Ritchie J: The effect of pulmonary hypertension onsystolic function of the right ventricle. Chest 1983;84:250-257

9. Yock PG, Popp RL: Noninvasive estimation of right ven-

tricular systolic pressure by Doppler ultrasound in patientswith tricuspid regurgitation. Circulation 1984;70:657-662

10. Goldman HI, Becklake MR: Respiratory function tests:Normal values at median altitudes and the prediction ofnormal results. Am Rev Tuberc 1959;79:457-467

11. Burrows B, Kaski Je, Niden AH: Clinical usefulness of thesingle breath diffusing capacity test. Am Rev Resp Dis 1961;84:789-806

12. Trask CH, Cree EM: Oximeter studies on patients withchronic obstructive emphysema, awake and during sleep. NEnglJMed 1962;266:639-642

13. Flick MR, Block AJ: Continuous in-vivo monitoring ofarterial oxygenation in chronic obstructive lung disease. AnnInten Med 1977;86:725-730

14. Jones NL: Pulmonary gas exchange during exercise inpatients with chronic airway obstruction. Clin Sci 1966;31:39-50

15. Himelman RB, Struve SN, Brown JK, Namnum P, SchillerNB: Improved recognition of cor pulmonale in patients withsevere chronic obstructive pulmonary disease. Am J Med1988;84:891-898

16. Simonson JS, Schiller NB: Sonospirometry: A non-invasivemethod for estimation of mean right atrial pressure based ontwo-dimensional echographic measurements of the inferiorvena cava during measured inspiration. J Am Coll Card1988;11:557-564

17. Kircher B, Himelman RB, Schiller NB: Estimation of rightatrial pressure by two-dimensional echocardiography of therespiratory behavior of the inferior vena cava (abstract).Circulation 1988;78(suppl II):II-550

18. Hatle L, Angelsen B: Doppler ultrasound in cardiology:Physical principles and clinical applications, 2nd ed. Phila-delphia, Lea and Febiger, 1985, p 23

19. Severinghaus JE: Blood gas calculator. JAppl Physiol 1966;21:1108-1116

20. Yock PG, Naasz C, Schnittger I, Popp RL: Doppler tricus-pid and pulmonic regurgitation in normals: Is it real (abstract)?Circulation 1984;70(suppl II):II-40

21. Hagler DJ, Currie PJ, Seward JB, Tajik AJ, Mair DD, RitterDG: Echocardiographic contrast enhancement of poor orweak continuous-wave Doppler signals. Echocardiography1987;4:63-67

22. Beard JT II, Byrd BF III: Saline contrast enhancement oftrivial Doppler tricuspid regurgitation signals for estimatingpulmonary artery pressure. Am J Cardiol 1988;62:486-488

23. Timms RM, Khaja FV, Williams GW, and the NocturnalOxygen Therapy Trial Group. Hemodynamic response tooxygen therapy in chronic obstructive pulmonary disease.Ann Intemn Med 1985;102:29-36

24. Widimsky J, Berglund E, Malmberg R: Effect of repeatedexercise on the lesser circulation. J Appl Physiol 1963;18:983-986

25. Weber KT, Janicki JS: Cardiopulmonary Exercise Testing:Physiologic Principles and Clinical Applications. Philadel-phia, W.B. Saunders, 1986, pp 319-320

26. Bradley BL, Garner AE, Billiu D, Mestas JM, Forma J:Oxygen-assisted exercise in chronic obstructive lung dis-ease: The effect on exercise capacity and arterial blood gastensions. Am Rev Respir Dis 1978;118:239-243

27. Bye PTP, Esau SA, Levy RD, Shiner RJ, Macklen PT,Martin JG, Pardy RL: Ventilatory muscle function duringexercise in air and oxygen in patients with chronic air-flowlimitation. Am Rev Respir Dis 1985;132:236-240

28. Matthay RA: Cardiovascular performance in chronic obstruc-tive pulmonary disease. Med Clin NAm 1981;65:489-524

29. Morrison DA, Adcock K, Collins M, Goldman S, Caldwell J,Schwarz MI: Right ventricular dysfunction and the exerciselimitation of chronic obstructive pulmonary disease. JAmColl Cardiol 1987;9:1219-1229

30. Morrison DA, Henry R, Goldman S: Preliminary study ofthe effects of low flow oxygen on oxygen delivery and rightventricular function in chronic lung disease. Am Rev RespirDis 1986;133:390-395

31. Owens GR, Rogers RM, Pennock BE, Levin D: The dif-fusing capacity as a predictor of arterial oxygen desaturationduring exercise in patients with chronic obstructive pulmo-nary disease. N EnglJ Med 1984;310:1218-1221

32. Sue DY, Oren A, Hansen JE, Wasserman K: Diffusingcapacity for carbon monoxide as a predictor of gas exchangeduring exercise. N EnglJ Med 1987;316:1301-1306

KEY WORDS * Doppler * echocardiography * emphysema* exercise * pulmonary hypertension * tricuspid insufficiency* contrast

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from

Page 10: Noninvasive PulmonaryArtery Pressure During …circ.ahajournals.org/content/circulationaha/79/4/863.full.pdfSaline contrast-enhanced Doppler evaluation of tricuspid insufficiency seems

B SchillerR B Himelman, M Stulbarg, B Kircher, E Lee, L Kee, N C Dean, J Golden, C L Wolfe and N

Doppler echocardiography in chronic pulmonary disease.Noninvasive evaluation of pulmonary artery pressure during exercise by saline-enhanced

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1989 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.79.4.863

1989;79:863-871Circulation. 

http://circ.ahajournals.org/content/79/4/863the World Wide Web at:

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

  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. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on June 17, 2018http://circ.ahajournals.org/

Dow

nloaded from