Mechanism of Paradoxic Pulse in Bronchial AsthmaFRAN§;OIS JARDIN, M.D., JEAN-CHRISTIAN FARCOT, M.D., Louis BOISANTE, M.D.,

JEAN-FRAN§OIS PROST, M.D., PASCAL GUERET, M.D., AND JEAN-PIERRE BOURDARIAS, M.D.

SUMMARY To elucidate the mechanism of paradoxic pulse in severe bronchial asthma, we performedhemodynamic studies and measured esophageal pressure in nine patients who had status asthmaticus andclinical paradoxic pulse. Two-dimensional echocardiography allowed simultaneous assessment of cyclicchanges in right- and left-heart size throughout the respiratory cycle. Esophageal pressure varied from a

markedly negative level during inspiration (-24.4 ± 6.5 cm H20) to a positive level during expiration (7.6+ 6.0 cm H20). Competition between right- and left-heart chambers for pericardial space during inspira-tion was suggested by the reduced left ventricular cross-sectional area at end-systole (- 24%, p < 0.01) andend-diastole (-32%, p < 0.01), the leftward septal shift, and the increased right ventricular internaldiameter at end-systole (42%, p < 0.01) and end-diastole (40%, p < 0.001). Competition for filling,however, could not entirely account for the paradoxic pulse, for systemic and pulmonary pulse pressures

were almost (within one cardiac cycle) in phase: both were minimal at inspiration and maximal at expira-tion. The increase in impedance to right ventricular ejection is another major factor reducing left ventricu-lar preload at inspiration. This reduction in preload was shown to be the predominant mechanism for thedecrease in left ventricular stroke output at inspiration.

INSPIRATORY DECLINE of the arterial pulse wasfirst described during attacks of bronchial asthma.'This inspiratory decrease in systolic arterial pressurewas later referred as "paradoxic pulse" and empha-sized as a cardinal manifestation of pericarditis.2 Pul-sus paradoxus has been recognized in many patientswith status asthmaticus,3 and is now considered anindex of the severity of airways obstruction.4 Para-doxic pulse has been noted in other clinical settings,including acute pulmonary embolism,5 chronic ob-structive pulmonary disease6 and tricuspid atresia.]Hemodynamict and echocardiographic9 studies in

patients with cardiac tamponade due to a tense peri-cardial effusion have improved our understanding ofthe mechanisms of pulsus paradoxus,'0 emphasizingthe severe competition for filling between the right andleft ventricle. In other clinical settings, however, themechanism of paradoxic pulse may differ. In bronchialasthma, pleural pressure is very negative at inspira-tion," and several mechanisms, all related to theselarge swings in pleural pressure, have been suggestedto explain paradoxic pulse. Most of our knowledge ofthe pathophysiological mechanisms derives from ob-servations of the hemodynamic alterations induced bya markedly negative pleural pressure during the Mullermaneuver in experimental animals'2 " and in man.During asthmatic attacks, however, the alveolar vol-ume is markedly increased. Moreover, only a fewclinical hemodynamic studies consisting of small num-bers of patients have been reported during statusasthmaticus.'5

In this study, we investigated the mechanisms ofparadoxic pulse in status asthmaticus using conven-tional hemodynamics combined with two-dimensional

From the Departments of Intensive Care Medicine and Cardiology,U.E.R. Paris-Ouest, Universite Ren6 Descartes, and the H6pital Am-broise Par6, Boulougne, France.

Supported in part-by grant 1981-129 from U.E.R. Paris-Ouest.Address for correspondence: Franqois Jardin, M.D., Hdpital Am-

broise Par6, 9, avenue Charles de Gaulle. 92100 Boulogne, France.Received October 13, 1981; revision accepted March 21, 1982.Circulation 66, No. 4, 1982.

echocardiography to assess right and left ventriculardimensions and configurations.

Materials and MethodsPatients

Nine adults with a clinically detectable (cuff-mea-sured decrease of 10mm Hg in systolic blood pressure)pulsus paradoxus during a severe attack of asthmawere included in the study. There were four men andfive women, ages 22-55 years (mean 34 years). He-modynamic and echocardiographic measurementswere performed during the first hour after the patient'sadmission to the respiratory intensive care unit. Stan-dard therapy for asthma was started, including nasaladministration of humidified oxygen, an i.v. bolus ofhydrocortisone, continuous i.v. infusion of theophyl-line, adequate fluid administration and mild sedationwith i.v. diazepam. In two patients who had a majortachycardia (heart rate greater than 160 beats/min),rapid blood volume expansion with 1000 ml of plasmaexpanders was performed before the study. In threepatients, standard therapy was ineffective and persis-tent respiratory failure necessitated mechanical venti-lation; in no patient was mechanical ventilation used atthe time of the study. Eight patients recovered; onepatient died 18 hours after admission despite mechani-cal ventilation. In the eight patients who recovered, theparadoxic pulse disappeared, accompanied by clinicalimprovement and progressive normalization of bloodgases. Four patients underwent two-dimensional echo-cardiography after recovering, and the findings werecompletely normal.

Hemodynamic StudiesSystemic arterial pressure was measured with a

small Teflon catheter inserted percutaneously into theright radial artery. Pulmonary capillary wedge, pulmo-nary arterial and right atrial pressures were measuredwith a triple-lumen Swan-Ganz catheter inserted per-cutaneously into the main pulmonary artery throughthe right basilic vein. Esophageal pressure was mea-sured with an esophageal balloon advanced through

887 at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

Voi, 66. No 4. OCTOIBFR 1982

the nose into the esophagus. down to 40 cm from thenares. All pressures were measured with Hewlett-Packard transducers positioned at the midaxillarv lev-el. with atmospheric pressure as a zero reference level.and recorded on a Honeywell LS 8 multichannel re-corder. Transmural pressure was calculated as intra-vascular pressure minus esophageal pressure. Heartrate was measured from a standard ECG lead. Cardiacoutput was measured by the thermodilution technique(right atrial injection with temperature recording in thepulmonary artery). Simultaneous sampling of arterial(a) and mixed venous (v) blood permitted determina-tion of oxygen and carbon dioxide tensions (Pao,Paco. Pvcco,) and pH by standard electrode tech-niques. Base excess was read on a Siggaard-Andersenalignment nomogram. Blood hemoglobin concentra-tion (Hb) was measured by spectrophotometr\ andhemoglobin saturation (Sao,. Sx70 ) was determinedusing a cooximeter. Oxygen consunmption (MO ) wNascalculated as the product of the arteriov!enous oxveencontent difference and cardiac inndex.

Echocardiographic StudiesEchocardiographic studies were performed in the

semisupine left lateral position with a phased-arravsector scanner and a di.ital scan converter (Roche RT400 or Varian V-340() R). An electrocardiographiclead was recorded during each study. Using a left later-al parasternal or a subcostal approach, a short-axiscross section of the left ventricle was selected at thehigh papillary muscle level. Two-dimensional echo-cardiographic images were recorded on a Sanvo VTC7100 videotape recorder. In three patients. standardM-mode echocardiographic measurements were alsoobtained simultaneously usine the two-dimensionalcross-sectional view to direct the ultrasound beamfrom the right ventricular free wlall to the left ventricu-lar posterolateral wall and crossina the middle inter-ventricular septum. Inspiration was signaled by inter-mittent visualization of an electronic marker on thevideo screen during two -dimensional echocardiogra-phic study! or by appropriate manipulation of the base-line of the electrocardiographic tracing during M-mode examination.

Two-dimensional echocardiographic studies weretaped and played back for subsequent single-frame.stop-motion analvsis of left ventricular cavit cross-sectional area. as previously described. The endocar-dial outlines of the left ventricle were drawn with agrease pen directly from the video screen onto a trans-parent paper bv two trained, independent observers.The end-svstolic frame coincided with the end of the Twave of the ECG. and the end-diastolic frame with theonset of the R wave on the ECG. For definition of thecavity outlines, the procedure of tracing the left ven-tricular lumen to measure short-axis areas was stan-dardized bv drawing the inner endocardial margin.Final agJreement on delineation of endocardial borderswas settled bv observin,. in slow motion, the preced-ing as well as the succeeding beats. The end-svstolicand end-diastolic left ventricular endocardial outlines

were digitized (Hewlett Packard 9871 A digitizer) andprocessed for the measurement of ventricular areasusing a Hewlett Packard 9825 A desktop computer.

Because delineation of the right ventricular outlinesin cross-sectional views was often imprecise and diffi-cult to draw accurately. only end-diastolic and end-systolic right ventricular internal diameters were mea-sured. These measurements were made with calipersdirectlv from the two-dimensional echocardiogramssynchronized with the ECG. These measurementswere validated in three patients by M-mode tracings.

During the echocardiographic analysis, special at-tention was paid to the spatial changes of the interven-tricular septal configuration. To evaluate septal shapequantitativelv. the radius of curvature of the interven-tr-icular septum was determined during inspiration andexpiration at both end-diastole and end-systole oncross-sectional views. The radius of curvature was ob-tained bv tracing endocardial borders of the interven-tricular septum and constructing arc segment. Twochords. each spanning separate parts of that arc, weredrawn and orthogonal lines bisecting each chord wereconstructed. The intersection of these two orthogonallines defined the center of the circle described by thearc seament. This technique was described by Brinkeret al.We also studied the variations in size of the inferior

vena cava during the respiratory cycle in four patients.The transducer was placed in a subxiphoid or rightsubcostal position and rotated so that the two-dimen-sional sector- was parallel to the inferior vena cava. Inthis manner. the course of the inferior vena cava be-hind the liver. extending through the diaphraTm andanastomosing with the right atrium, was imaged. Thetransducer was rocked slightlv medially and laterallyto record the maximal vena caval diameter. Inferiorvena caval end-diastolic internal diameters at expira-tion and inspiration were measured under the junctionof the hepatic veins, about 6 cm below the diaphragm.These measurements were made with calipers on stillframes. Special attention was also paid to the changesin abdominal vena caval size in its initial part, betweenthe diaphragm and the junction of the hepatic veins.

Statistical AnalysisTo compare hemodynamic and echocardiographic

data obtained at expiration with those obtained at inspi-ration. we used the t test for paired values. The resultsare presented as mean + sD unless otherwiseindicated.

ResultsAll patients exhibited an inspiratory decline (greater

than 2(0 mm Hg) of systolic blood pressure, markedtachycardia ( 142 -+- 12 beats/min) and increased cardi-ac index (4.7 0.8 I/min/m`). The stroke index wasless than 40 ml 'mn in all patients but one (patient 7) andaveraged 33.9 + 6.2 ml/m' (table 1). The bloo(d hemno-globin concentration was higher than 15 g10()() ml inseven patients and was normal in the two patients inwhom it was measured after blood volume expansion

X SX CIRCULATION

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

PARADOXIC PULSE IN BRONCHIAL ASTHMA/Jardin et al.

TABLE 1. Hemodvnamic and Blood Gas Data

Inspiratorydecline ofsystolic Po aoblood Heart Cardiac Stroke Pao2 Paco2 Base

pressure rate index index Hb kPa AMDo2 V02 excessPt (mm Hg) (beats/min) (1/min/m2) (ml/beat) (g/100 ml) (mm Hg) pH (ml/100 ml) (ml/M2) (mEq/1)

1 40 150 4.5 30 15.5 6.65 6.65 7.31 3.6 160 -3(50) (50)

2 50 160 5.2 33 16.4 4.65 10.64 7.13 3 156 -9(35) (80)

3 55 150 3.7 25 15.3 7.05 7.18 7.28 3.5 130 -2(53) (54)

4 30 145 5.6 39 16.6 8.91 6.11 7.33 3.1 173 - 3(67) (46)

5 40 130 4.4 34 17.2 11.30* 5.05 7.21 3.3 145 - 12(85) (38)

6 58 135 4.7 35 17.0 10.51* 10.91 7.17 2.6 122 -2(79) (82)

7 46 130 6.1 47t 11.7 12.90* 10.64 7.17 2.5 153 - 2(97) (80)

8 35 126 3.9 31 15.8 12.77* 7.31 7.39 4.7 182 +6(96) (55)

9 21 148 4.5 31 t 12.9 5.88 7.58 7.36 4.4 198 +4(44.5) (57)

Pulsus paradoxus (inspiratory decline of systolic blood pressure) was demonstrated by direct measurement of the radial artery pressure.*Oxygen, 6 1/min through a nasal tube.tAfter blood volume expansion.Abbreviations: Hb = hemoglobin; Pao2 = arterial oxygen tension; Paco2 = arterial carbon dioxide tension; MADo2 = arteriovenous

oxygen difference; V02 = oxygen consumption.

(mean hemoglobin concentration 15.4 + 1.9 g/100

ml). Characteristic features of severe asthma were

shown by blood gas analysis. Hypoxemia was presentin all patients except patients 7 and 8, who receivedoxygen therapy. Eight of the nine patients had hyper-capnia. Five patients had mild and two had severe

metabolic acidosis. The arteriovenous oxygen contentdifference was normal in seven patients and narrowedin patients 6 and 7. Calculated oxygen consumptionwas slightly elevated (157 24.2 ml/min/m2).

Pressure changes from expiration to inspiration are

summarized in table 2. Absolute pressures are given in

the top half of the table. At inspiration, pleural (eso-phageal) pressure decreased dramatically to a marked-ly negative value (-24.4 + 5 cm H20), and increasedat expiration, reaching an average positive value (7.6

6.0 cm H20). Systolic and diastolic radial arterypressures both decreased at inspiration; the decreasewas greater in systolic than in diastolic pressure, lead-ing to a large and significant decrease in radial arterypulse pressure. Systolic and diastolic pulmonary arterypressure both decreased at inspiration; systolic pres-

TABLE 2. Pressure Values at Expiration and InspirationPleural Radial artery pressure Pulmonary artery pressure Capillary Righlpressure (m'g mm Hg)wde ara

(esophageal) pressure pressure(mm Hg) S D PP S D PP (mm Hg) (mm Hg)

A. Absolute pressure

Exp 7.6 ± 6.0 147.2 81.3 65.9 37.0 20.4 16.6 16.0 11.9± 10.3 ±12.7 ±16.5 +9.3 ±6.4 ±7.0 +7.9 ±6.7

Insp -24.4 ± 6.5 105.5 63.2 42.3 11.1 -2.8 13.9 -12.4 3.0+ 12.2 ±10.1 ±16.2 +5.2 ±5.9 +4.2 +8.3 ±4.9

p < 0.001 < 0.001 < 0.001 <0.001 < 0.001 < 0.001 < 0.05 < 0.001 < 0.001B. Transmural pressure

Exp 0 139.6 73.7 65.9 29.4 12.8 16.6 8.4 4.3± 11.6 ± 15.1 ±16.5 ±8.4 ±5.8 ±7.0 ±4.1 +3.5

Insp 0 129.9 87.6 42.3 35.5 21.6 13.9 12.0 21.4± 14.4 ± 11.3 ±16.2 +5.9 ±3.5 ±4.2 ±5.9 ±9.8

p < 0.01 < 0.01 < 0.001 < 0.01 < 0.01 < 0.05 < 0.05 < 0.01

Values are mean ± SD.Abbreviations: S systolic; D = diastolic; PP = pulse pressure (i.e., systolic minus diastolic).

889

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

VOL 66, No 4. OCTOBER 1982

sure decreased more than diastolic pressure, whichcaused a small but significant decrease in pulmonaryartery pulse pressure. Both capillary wedge and rightatrial pressures decreased at inspiration, but the inspi-ratory decrease in capillary wedge pressure was largerthan that in right atrial pressure. An example of theselarge cyclic variations of esophageal and intravascularpressures is shown in figure 1. Both right atrial andpulmonary capillary wedge pressures (left panel) de-creased at inspiration along with the decrease in eso-phageal pressure, but pulmonary wedge pressure de-creased more than right atrial pressure, and thepulmonary wedge recording crossed the right atrialrecording. Deflation of the Swan-Ganz catheter bal-loon (right panel) allowed comparison of five succes-sive beats in the same respiratory cycle from simulta-neous recordings of pulmonary and radial pulsepressures. The minimal pulmonary pulse pressure (9mm Hg, beat 2) occurred at the end of inspiration andcoincided with the minimal radial pulse pressure (65mm Hg, beats 2 and 3); conversely, the maximal pul-monary pulse pressure (29 mm Hg, beat 3) occurred atthe onset of expiration and was followed after a one-beat delay by the maximal radial pulse pressure (79mm Hg, beat 4). Transmural pressure values (i.e.,absolute values minus esophageal pressure) are givenin the bottom half of table 2. At inspiration, systolicradial artery pressure was still significantly decreased(-7%), but diastolic pressure increased significantly(19%). Systolic (29%) and diastolic (70%) pulmonaryartery pressures also increased significantly. Both left(capillary wedge) and right (right atrial) ventricularfilling pressures increased during inspiration, but theincrease in right atrial pressure was three times greaterthan the increase in pulmonary capillary wedge pres-sure. Consequently, right atrial pressure was lowerthan pulmonary wedge pressure at expiration andmuch higher at inspiration.

Two-dimensional echocardiographic data are sum-marized in figure 2. Compared with the expiratoryvalues, the end-systolic and end-diastolic left ventricu-lar cross-sectional areas were reduced by 24% and32%, respectively, during inspiration (p < 0.01). Asmaller decrease in areas between end-diastole andend-systole (-3.3 + 1.2 cm2 at inspiration vs -6.0± 2.9 cm2 at expiration, p < 0.01) was also evi-denced, leading to the assumption that left ventricularstroke output was decreased during inspiration. Insharp contrast to the changes in the left ventricle, rightventricular end-systolic and end-diastolic internal di-ameters were increased by 42% (p < 0.01) and 40%(p < 0.001), respectively, during inspiration. Thesephasic alterations in left and right ventricular cavitydimensions were associated with septal flattening, asdemonstrated by a significant increase in the end-sys-tolic and end-diastolic radii of septal curvature (by31% and 38%, respectively p < 0.001). Figure 3shows the respiratory variations in left ventricular cav-ity cross-sectional area that were demonstrated by two-dimensional echocardiograms. In figure 4, phasicchanges in left and right ventricular internal diametersare demonstrated by M-mode examination. At inspira-

tion, the left ventricular internal diameter decreasedand the right ventricular internal diameter increased; atexpiration, opposite changes occurred.

Two-dimensional echocardiographic measurementsof abdominal vena cava internal diameter obtainedabout 6 cm under the diaphragm in four patients aregiven in table 3. The vena caval diameter decreased(-60 ± 13%) during inspiration in each patient. Thisdecrease was associated in each case with an end-inspiratory collapse of the inferior vena cava in itsinitial abdominal portion (fig. 5).

DiscussionOur patients with status asthmaticus had a high car-

diac output state presumably caused by elevated oxy-

I lFIGURE 1. Respiratory changes in intrathoracic and intravas-cular pressures. From top: ECG, systemic arterial pressure(SAP), pulmonary capillary wedge pressure (PCWP) and pul-monarv artery pressure (PAP), right atrial pressure (RAP), andesophageal pressure (EP). All pressures are expressed in mmHg. V = balloon deflation of the Swan-Ganz catheter. Tonset of inspiration; I = onset of expiration. Arabic numbersindicate five successive cardiac beats during a respiratorycycle.

890 CIRCULATION

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

PARADOXIC PUL-SE IN BRONCHIAL ASTHfIA.Jordin tcl.

CROSS SECTIONALLV AREA (cm2)

20- U Er181 T n0 EF16-

14-

12-

1018

6-

4.

2-

0*

nd -Systo

nd-Diost

L VS RADIUS OF D\I INTERNALCURVATURE(cml n v DIAMETER (cm)

)le p cO,Ol

ole *** p c 0,001

3-

2

1-

0*

gen demand due to the increased effort of breathing.Stroke volume was lower than normal, but markedtachycardia resulted in increased cardiac output. Thisfinding, and the fact that seven patients had metabolic

Ficium.- 2. Compatrison bettween ('tpitox(LXP) (c7(1 in spilratorY (1'SP) 1('ft v'e7tricidl/ar(LV ) cross- stcitinal areal.0, interventricidarseptail (IVS) rad,i of enevatore (i/iright ven11-turicnar (PVt interaJil d/iamiieters., it en1d1- vs.tocl'(/htc hed12(1 s)bandIIil n1d1- iaso/ (open bars).A// sa's(v/C 177(17 +1m /).

"I./

EXP INSPacidosis, suggested inadequate cardiovascular adaptwt-tion. 7Paradoxic pulse4 mnay result from one or sexeralof the following: impairmnent to left ventricular ejec-tion: 1 "' mechanical impairment of left ventricular fill-inr caused by increased transpulmonary pressure (i.e..alveolar pressure minus pleural pressure): and hvpo-volemia. as su2cested in our study by heminocon-centration. 18

Clinically detectable paradtoxic pulse correspondingyto an inspiratory decrease of 10( mm Hg or more insystolic blood pressure' was a prerequisite ttr includ-ing patients in our study. and was further documentedlbv invasive measurements of the radial artery pressurerelative to atmospheric pressure. However, a sianifi-cant decline was still noted when arterial pressure wasmeasured relative to pleural pressure. Therefore. di-rect transmission of intrathoracic pressure changes tothe left ventricle and subsequentlv into the arterial svs-tem does not totallv account for the observed decreasein arterial pressure. This findin(e. as well as a substan-

lINSP IEXP

FClOuR!^- 3. [Tio-dimnensional echloeardi)graphliic short-axiscr0.$ s'ectio.s ofltlte heartt in paitient 6 tit epiCrationl (t1p) (17(1 aitinspiration (bottoftt). RV }right 'entricle: IVS = inter etwrit-a lafa septiom: LV lefi ventricle.

Fj(it!Ri- 4. M-tnode traeinig illustraitin2 the p/hasie vlitiotOlsin righit an16d lefti sentricdlar internal dicaiiteter-s (RVID a17dLVII)D respectivelv ). Inspiration (INSP) mnd exvjiratioli (EXP)atTr sign led I)s mlaini1ipu7la(tionis. 0/ tlie ICG laiseline.

. I3* *K*

I~~~~~~~~~***

EXP INSP EXPIN2SP

EXP INSP EXP INSP

Xt31

__

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

Voi 66, No 4. OCT013FR 1982

TABLim 3. Diameter o;f Iifjerior Venai Cavtiat Etad-dinistole in1 FonlfPaientit

Exp lnsp InspiratoryPt (cm) (cm) decrease

6 16.7 6.7 -60t}4

7 23.3 13.0 -44(cV

8 20.0 5.0 -75 ck

9 11.4 4.3 -622

Abbreviations: Insp inspiration: Exp expiration.

tial inspiratory decrease in systemic pulse pressure,clearly suggest that the inspiratory decrease in left ven-tricular stroke volume plays an important role in themechanism of paradoxic pulse. A reduced left ventric-ular stroke output during inspiration was also indirect-ly evidenced by a smaller reduction in left ventriculararea between diastole and systole.

Theoretically, the lower the pleural pressure duringinspiration, the larger the pressure gradient betweenthe intra- and extrathoracic compartments againstwhich the left ventricle must eject a given stroke vol-ume.'3 Our data, in which an increased inspiratoryimpedance for left ventricular ejection was suggestedby an increased inspiratory transmural diastolic aorticpressure, are consistent with those of Robotham et al.'-Accordingly, it has been argued that the inspiratoryincrease in left ventricular afterload resulted in de-creased left ventricular stroke output, which in turnwas reflected by the decrease in aortic pressure.'" Re-cent studies using implanted markers suggest that leftventricular end-diastolic and end-systolic dimensionsare increased during sustained decrease in pleural pres-sure in man'9 and in experimental animals.2'

In contrast, our study tends to demonstrate an inspi-ratory decrease in left ventricular size at both end-diastole and end-systole. These seemingly paradoxicobservations can be explained by the fact that none ofthe studies mentioned above took into account the freewall to interventricular septum dimension. Our results

are consistent with those of Brinker et al.'" during theMuller maneuver and Settle et al.6 in patients withchronic obstructive lung disease, who reported a re-duction in left ventricular two-dimensional echocar-diographic cross-sectional area, and those of Sharf etal.,12 who found in dogs that both systolic and diastolicdimensions decreased in the early phase of an obstruc-tive inspiratory effort. One might object to these stud-ies and also to our work because possible compensa-tory changes in the left ventricular long axis may beundetected by two-dimensional echocardiography be-cause these measurements are based on short-axisviews; but it is unlikely that a 25% reduction in leftventricular cross-sectional areas could be totally com-pensated for by an increase in the left ventricular longaxis. In other clinical settings, important modificationsin left ventricular volumes without signiticant changesin the left ventricular long axis have been evidenced,'which suggests that acute modifications in left ven-tricular volume result from changes in short-axisdimensions.

In our patients, echocardiographic left ventriculardimensions were smaller during inspiration and leftventricular transmural filling pressure (estimated bythe pulmonary capillary wedge pressure minus esopha-geal pressure) was significantly higher. Although pul-monary capillary wedge pressure might not reflect thetrue left ventricular filling pressure in patients withincreased alveolar volume,'9 increased left atrial andleft ventricular end-diastolic pressures during deep in-spiration or during the Muller maneuver have beenreported.'2 1 Increased left ventricular filling pressurein the presence of decreased left ventricular dimensionindicates a reduction of the apparent compliance of theleft ventricle. Such a reduction is consistent with adisplacement of the interventricular septum toward theleft ventricular cavity, and has also been noted in otherclinical settings in which lung volume was increased'3or pleural pressure decreased. We attempted to quan-tify variations of septal curvature suggested by obvious

Fi(URi--. 5. Tinn-dimensional echocardiogra-phi c subcostal vie w ofthe abdlonlminal portion ofthe inferior vena cava in the long-axis view at

expiration (left pantiel) anid inspiration (rightpaniel). Legend.s are givieni oni a retouched pho-tocop oft/ihe glossy.print: IVC infe'rior veenZacava;: HV hepatic rein; RA = right atriun:D = diaphragm. The arrow indicate.s the enzd-inspiratorv collapse of the ressel in its iniitialabdominal part.

892 CIRCULA'FION

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

PARADOXIC PULSE IN BRONCHIAL ASTHMA/Jardin et al.

changes in ventricular shape noted during real-timeviewing and playback analysis. We used the techniquedescribed by Brinker et al., ' which may be sensitive tonoise and somewhat subjective because it uses twochords of septal arc segment drawn in a somewhatarbitrary fashion. However, this technique has beenhelpful in the study of pathophysiologic consequencesof right ventricular loading.15 23 In the present study,two-dimensional echocardiograms clearly demonstrat-ed an inspiratory increase in the septal radius of curva-ture associated with a reduced left ventricular cross-sectional area. Leftward displacement and flatteningof the interventricular septum during diastole is consis-tent with an inverse transseptal pressure gradient. Infact, during inspiration right-heart filling pressure de-creased less than left-heart filling pressure, althoughboth chambers were exposed to the same negative ex-ternal pressure. During expiration, opposite pressurechanges were noted and the septum returned to itsnormal circular configuration. Respiratory variationsof the transseptal pressure gradient may be explainedby respiratory changes in systemic venous return andincreased impedance to ejection of the right ventricleduring inspiration.

Facilitation of venous return to the right-heart cham-bers with subsequent increase in right ventricular vol-ume by inspiratory negative pleural pressure is wellknown20 and was further documented in our study. Theinspiratory reduction of the abdominal vena cava di-ameter in our patients (fig. 5) illustrated the boostingeffect of increased subatmospheric pressure at inspira-tion. These two-dimensional echocardiographic find-ings do not differ from those in normal subjects.24 Butin our patients, the initial part of the abdominal venacava collapsed at the end of inspiration (fig. 5). Ve-nous collapse is usually considered the result of theinability of collapsible vessels to transmit negative

25 reutopressures. However, reduction of venous diameter in

high-velocity flow secondary to a decreased lateralpressure could also play a role. However, positiveintrathoracic pressure during expiration has been notedin severe asthma1' and was present in our patients. Anexpiratory increase of the abdominal vena caval sizeand a reduction of right ventricular diameter are con-sistent with diminished right ventricular filling duringexpiration. In fact, distention of the jugular veins dur-ing expiration and venous collapse during inspirationare common findings in patients with status asthmati-cus. Accordingly, in severe asthmatic attacks, a mark-edly negative pleural pressure during inspiration accel-erates venous return, whereas a suddenly positivepleural pressure during expiration severely impedesvenous return. The contribution of respiratory vari-ations of systemic venous return to the development ofparadoxic pulse in patients with cardiac tamponade hasbeen emphasized. In an experimental study by Shabe-tai et al. ,26 maintenance of constant systemic venousreturn prevented paradoxic pulse from developing. In-deed, increased venous return to the right heart duringinspiration in patients with cardiac tamponade is asso-ciated with a substantial increase in pulmonary arterial

blood flow.26 In contrast, in status asthmaticus, despitemarkedly enhanced venous return to the right-heartchambers during inspiration, pulmonary arterial bloodflow does not increase, as suggested by the decreasedpulmonary pulse pressure. Thus, in cardiac tampon-ade, right and left ventricular pulse pressures havebeen shown to be approximatively (by one or twobeats) 1800 out of phase,26 whereas in status asthmati-cus they were almost in phase (i.e., both minimalduring inspiration and maximal during expiration).Therefore, the mechanism of pulsus paradoxus in se-vere asthma is not merely caused by a competitionbetween the two ventricles for intrapericardial space,as in cardiac tamponade; an additional contributingfactor must be considered, namely, increased impe-dance to right ventricular ejection.A marked inspiratory decrease in pleural pressure

results in an increased transpulmonary pressure and, inturn, in increased alveolar volume. Raising the alveo-lar volume increases resistance to flow through thealveolar vessels (capillaries) and therefore the after-load upon the right ventricle. In addition, this effectwould combine with any eventual increase in pulmo-nary vascular resistance due to hypoxia, hypercapniaand acidosis. During inspiration, the finding of a largerincrease in diastolic than in systolic pulmonary arterialtransmural pressure suggested increased pulmonaryvascular resistance in our patients. Hence, markedsubatmospheric intrathoracic pressure tends to impederight ventricular ejection, which is consonant with aninspiratory increase in diastolic and systolic right ven-tricular dimensions observed in the present study andin others.6' 15 This finding may also explain why pul-monary and systemic pulse pressures were approxi-mately in phase in status asthmaticus. Thus, acute rightventricular overloading appears to be the major hemo-dynamic alteration induced by markedly negative in-trathoracic pressure during inspiration. Reduced pul-monary venous return and decreased apparent leftventricular compliance, both secondary to right ven-tricular overloading, result in decreased left ventricu-lar volume. Furthermore, lung distention may com-press the heart chambers, and thereby enhance thenormal ventricular interaction.A rationale for the mechanisms of paradoxic pulse

must take into account all the above factors throughouta respiratory cycle. At the onset of inspiration, pleuralpressure abruptly drops from the supraatmosphericlevel at end-expiration to a markedly negative level.This sudden fall of pleural pressure increases lung vol-ume and vascular resistance (increased afterload) andboosts extrathoracic blood into the right ventricle (in-creased preload). The net result, however, is a de-crease in right ventricular stroke volume. Reducedright ventricular output combined with exaggeratedinput results in overdistention and a subsequentleftward shift of the interventricular septum. Concomi-tantly, left ventricular stroke volume decreases, be-cause of reduced filling due to diminished pulmonaryvenous return and increased stiffness, when left ven-tricular afterload is also increased. At the onset of

893

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

VOL 66. No 4. OcirOBER 1982

expiration, pleural pressure suddenly increases to asupraatmospheric level, which leads to a decrease inlung volume and in turn unloads the right ventricle.The unloaded and overfilled right ventricle ejects alarger stroke volume and thereby allows the septum toreturn to its normal shape and the left ventricle to bemore compliant. As soon as the next beat, systemicarterial pulse pressure increases. However, as expira-tion continues, pleural pressure remains positive andan adverse pressure gradient for filling is created be-tween the right heart and extrathoracic vessels. Conse-quently, right ventricular stroke volume decreasesgradually, as shown by the progressive decline in pul-monary pulse pressure, and systemic pulse pressuredeclines after the same lag of one beat. This shorttransit time (one beat) between right and left ventricu-lar events may be due to the hyperkinetic state andpresumably reduced capacitance of the pulmonary vas-cular bed in the presence of increased alveolar volume.

In conclusion, direct transmission of intrapleuralpressure changes to the left ventricle and subsequientlyinto the arterial system does not totally account for therespiratory variations of systemic blood pressure inpatients with severe asthma. Paradoxic pulse also ap-pears to result from the interplay of respiratory vari-ations in both systemic venous return and impedance toright ventricular ejection and of an exaggerated ven-tricular interdependence due to lung distension.

ReferencesI. Floyer J: A treatise of the asthma. 2nd ed. London. R Wilkins.

17172. Kussmaul A: Leber schwieliee MNediastino-Pericarditis unci den

patadoxen Puls. Bed Klin Wochenschr 10: 433. 18733. Rebuck AS, Pengellx LD: Developinent ot pulsus paradoxus in the

presence of airways obstruction. N En,-I J Med 288: 66. 19734. Knowles GK, Clark TJH: Pulsus paiadoxus as a valuable situn

indicatinr severitv of asthma. Lancet 2: 1356o 19735. Winer H. Kronzon I. Glassman E: Echocairdographic hindings in

severe paradoxical pulse due to pulmonary embnolization. Am JCardiol 40: 808. 1977

6. Settle HP. Engel PJ. F1o vler NO: Echocardiographic studv of theparadoxical arterial pulse in chronic obstructive lutig- disease. Cir-culation 62: 1297. 198U

7. Baum VC, Tarnoft H, Hoffman JL: Pulsus paradoxus in a patientwith tricuspid: atresia and hvpoplastic right heart. Circulation 62:63 1, 1980

8. Shabetai R. Fowler NO. Fenton JC: Pulsus paradoxus. J ClinInvest 44: 1882, 1965

9. DCruz IA, Cohen HC, Prabhu R. Glick G: Diagnosis of cardiactamponade by echocardiography: changes in mitral valve motionand ventricular dimension with special reference to paradoxicalpulse. Circulation 52: 460. 1975

10. Cosio FG. Martinez JP, Serrano CM. Calzada CC. Alcaine CC:Abnormal septal motion in cardiac tamponade with pulsus para-doxus. Chest 71: 787, 1977

11. Stalcup SA, Mellins RB: Mechanical forces producing pulmonaryedema in acute asthma. N Engl J Med 297: 592. 1977

12. Scharf SM, Brown R. Saunders N. Green LH: Effects of normaland loaded spontaneous inspiration on cardiovascular function. JAppI Physiol 47: 582, 1979

13. Robotham JL. Lixfeld W, Holland L. McGregor D. Bryan AC,Rabson J: Effects of respiration on cardiac performance. J ApplPhysiol 44: 703, 1978

14. Brinker IA. Weiss JL, Lappe DL, Rabson JL. Summer WR, Per-mutt S. Weisfeld M: Leftward septal displacement during rivhtventricular loading in man. Circulation 61: 626. 1980

15. Weitzenblum E. Pauli G, Roeslin N. Vandevenne A. Bohner C,Oudef P: Les donnees hemodynamiques pulmonaires au cours de lacrise d'asthme: incidence des conditions mecaniques endothora-ciques. J Fr Med Chir Thor 25: 487, 1979

16. McFadden ER: Respiratory mechanics in asthma. In BronchialAsthma, edited by Weiss EB. Segal MS. Boston, Little, Brown,1976. pp 259-278

17. Jardin F. BarthelemyIM: Frequence de l acidose metabolique aucours de l etat de mal asthmatique. Nouv Presse Med 6: 329. 1977

18. Straub PW. Bihlman AA. Rossier PH: Hypovolenmia in statusasthmaticus. Lancet 2: 923. 1969

19. Buda AJ. Pinsky MR. Ingels NB. Daughters GT. Stinson EB,Alderman EL: Effect ol intrathoracic pressure on left ventricularperformance. N Engl J Med 301: 453. 1979

20. McGregor M: Pulse paradoxus. N Engl J Med 301: 480. 197921. Summer W. Bromberger-Bamea B, Shoukas A. Sagawa K. Per-

mutt S: The effects ot respiration on left ventricular function(abstr). Circulation 54 (suppl 11): 11-13, 1976

22. Zwehl W. Gueret P. Meerbaum S Holt D. Corday E: Quantitativetwo-dimensional echocardiographv during bicycle exercise in normal subjects. Am J Cardiol 47: 866. 1981

23. Jardin F. Farcot JC. Boisante L. Curieii N, Mlarairaz A, Bourdar-ias JP: Influence ot positive end-expiratorv pressure on leit ventric-ular function. N Engl J Med 304: 387. 1981

24. Mintz GS. Morris NK. Parry WR. Iskandrian AS. Kanie SA: Real-time interior vena cayval ultrasonog-raphy: normal and abnormaltindings and its use in assessing rtiht-heart tunction. Circulation64: 1018. 1981

25. Guvton AG: Circulatory Physiology: Cardiac Output and its Reu-lation. Philadelphia. WvB Saunders. 1963

26. Shabetai R. Mangiardi L. BhargJava V. Ross J. Higgins CB: Thepericardium and cardiac tunction. Proig Cardiovasc Dis 22: 107.1979

894 CIRCULATION

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from

F Jardin, J C Farcot, L Boisante, J F Prost, P Gueret and J P BourdariasMechanism of paradoxic pulse in bronchial asthma.

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1982 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.66.4.887

1982;66:887-894Circulation. 

http://circ.ahajournals.org/content/66/4/887the 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:

at VA MED CTR BOISE on October 9, 2015http://circ.ahajournals.org/Downloaded from