pathophysiology of cardiac tamponade* · 2017. 6. 21. · cardiac tamponade is defined as...

Pathophysiology of CardiacTamponade*David H. Spodick, MD, DSc, FCCP

(CHEST 1998; 113:1372-78)

Key words: cardiac compression; pericardium; tamponadephysiology

Cardiac tamponade is always life threatening andnearly always requires urgent and precise ther¬

apeutic intervention. It is perhaps unique in thatappreciation of its pathophysiologic state is essentialto precise diagnosis and rational treatment. Since the19th century, investigations in experimental animalsprovided a basic understanding that has been con¬

tinually refined by recent investigators to the pointwhere an integrated picture of its macrophysiologiccondition emerges. This review covers the resultantconcepts. Procedural details of investigations can befound in the references (thus, for example, the textomits the relative merits of catheters vs flat balloonsfor intrapericardial pressure determination).

General Considerations

Cardiac tamponade is defined as significant com¬

pression of the heart by accumulating pericardialcontents, including effusion fluids, blood, clots, pus,and gas, singly or in combinations. "Significant com¬

pression" depends on whether tamponade is ap¬proached from a purely physiologic or clinical stand¬point. Since tamponade is a pathophysiologiccontinuum, hearts can be said to be mildly to floridlytamponaded, the latter being a life-threateningemergency and the former a stage that can progressin that direction. "Life threatening" is a result ofprogressive, ultimately critical, reduction in cardiacoutput that can occur from as little as 150 mL ofblood flooding the pericardium after a cardiacwound to much more than 1 L of fluid in slowlyaccumulating pericardial effusions. The "continuum"concept, formulated by Reddy et al,1 involves stagesseen in intact, unanesthetized animals, fully recov¬

ered from operations to implant instrumentation(with responses quantitatively different from openchest, anesthetized animals undergoing short-term

*From the Cardiology Division, Saint Vincent Hospital, and theDepartment of Cardiology, University of Massachusetts MedicalSchool, Worcester.Manuscript received August 20, 1997; revision acceptedOctober 13.

experiments23) and also in human tamponade asmeasured by cardiac catheterization.1 However, thecontinuum begins before catheterization is clinicallyjustified when small, medium, and large asymptom¬atic, "nontamponading" effusions couple the parietalpericardium to the heart, thereby exaggerating ven¬

tricular interaction and the consequent respiratoiyresponses, as measured by sensitive noninvasive tech¬niques in patients without pulsus paradoxus, the classicrespiratory response of critical cardiac tamponade.4Interaction Between the Pericardium and

Its Contents

Clinically significant cardiac compression by peri¬cardial fluids depends on three interrelated condi¬tions. The pericardial contents must do the follow¬ing: (1) fill the relatively small pericardial reservevolume (Fig 1).the volume which, added to thenormal 15 to 35 mL of pericardial fluid, will justdistend the parietal pericardium by filling its numer¬ous recesses and sinuses;5 (2) thereafter increase at arate exceeding the stretch rate of the parietal peri¬cardium;6 and (3) exceed the rate at which venous

blood volume expands to support the small normalpressure gradient for right heart filling.1'6 Because at

any instant the pericardium is relatively inextensible,the heart and the pericardial contents competecontinuously for the relatively fixed intrapericardialvolume. Therefore, unless the abnormal pericardialcontents increase veiy slowly, permitting progressive"give" ofthe parietal pericardium, their increase is atthe expense of cardiac chamber volume. The keyoperational defect is reduced cardiac inflow due tocompression of all cardiac chambers, progressivelyreducing their diastolic compliance and ultimatelyequalizing the mean diastolic pressures in eachof them.

Pericardial stiffness determines the fluid incre¬ments causing tamponade. Despite elastic tissue andinitially wavy collagen, the initial elements of peri¬cardial relaxation,7 even the normal parietal pericar¬dium is relatively noncompliant8 so that after theinitial give, the pericardial pressure-volume curve

angulates and becomes vertical5 (Fig 1). Thus, pa¬tients with critical tamponade function on the steepvertical portion of the pericardial pressure-volumecurve with progressively smaller fluid increments

1372 Reviews

Downloaded From: http://journal.publications.chestnet.org/pdfaccess.ashx?url=/data/journals/chest/21765/ on 06/21/2017

Pressureor

Strain

Pericardial ReserveVolume

Volume, or

Stress

Figure 1. Dimensionless schematic pericardial pressure-volume(and also stress-strain) curve also indicating the pericardialreserve volume (see text). Dimensions omitted because of wideindividual variations in principal determinants of the curve: rateof fluid increase and parietal pericardial compliance. FromSpodick DH. In: Spodick DH, ed. The pericardium: a compre¬hensive textbook. New York: Dekker, 1997; 19 (with permission).

provoking progressively large pressure increments. Ifunchecked by compensatory mechanisms (Fig 2) or

effective treatment, the heart becomes criticallycompressed, because cardiac filling, ie, maximumdiastolic volume, competes unsuccessfully withpathologic pericardial filling for the relatively fixedspace within the parietal pericardium. This results in

progressive reduction in cardiac chamber volumes.

Cardiac Responses to SignificantPericardial Compression

Figure 2 schematizes the major hemodynamicevents and compensatory mechanisms in uncompli¬cated cardiac tamponade.When increasing pericardial contents put the in¬

trapericardial pressure on the steep portion of itsJ-shaped pressure-volume curve (Fig 1), the cardiacchambers must operate on parallel steep pressure-volume curves, a form of diastolic dysfunction where,at any diastolic volume, there is necessarily excessiveintracardiac pressure.129 Thus, progressive com¬

pression of the atria and ventricles progressivelyresists cardiac filling, progressively reducing ventric¬ular stroke volume. In intact, unanesthetized, long-term instrumented animals2 compensatory mecha¬nisms maintain arterial pressure until relativelysudden decompensation, as in comparably intact

human patients.9 This is a "last straw".literally "lastdrop".phenomenon due to the final increment ofpericardial contents. This situation is demonstratedby the reciprocal effect of therapeutic pericardialfluid drainage.the initial decrement usually pro¬duces the largest relative hemodynamic improve¬ment by rapidly shifting the stretched pericardiumback toward the "flat" portion of its pressure-volumecurve with parallel improvement in intracardiacpressure-volume relations and arterial pressures.6Although the key chambers maintaining cardiac

output are the right and left ventricles, within thepathophysiologic continuum of tamponade, the ear¬

liest targets of significant compression are the thin¬ner chambers: the right atrium and ventricle, dia¬stolic pressure in which equilibrates with risingpericardial pressure before the left atrial and ven¬

tricular pressures do.1-2-6 That right atrial and ven¬

tricular pressures are normally somewhat lower thanin the corresponding left chambers contributes totheir earlier entrainment by inexorably rising peri¬cardial pressure.

Significance of Transmural Pressures andRespiratory ReciprocationFor each cardiac chamber, its transmural

pressure.intracardiac pressure minus pericardialpressure.is a principal determinant of its filling.(Transmural pressure is a true filling [distending]pressure that contributes to ventricular preload.610)Normal pericardial pressure is lower than the rightatrial mean and right ventricular diastolic pressuresso that right atrial transmural pressure (right atrialpressure minus pericardial pressure) is normallyhigher than its cavitary pressure. In tamponade,rising pericardial pressure progressively reduces.and ultimately can make phasically negative.theaverage transmural pressure of first the right andsubsequently the left cardiac chambers.19 Survivalnecessitates the ensuing parallel rise in diastolicpressures, first in the right side ofthe heart and laterthe left side of the heart, critically reducing allchamber transmural pressures, and with them allfilling. Like most tamponade-induced abnormalitiesof pressure and flow, transmural pressures are recip¬rocally reduced and increased during the respiratoryphases for the left vs the right heart. Thus, inspira¬tion increases right heart filling at the expense of theleft heart with reversal in expiration. In criticaltamponade, when cardiac output usually has fallenby at least 30%,1_5 transmural pressures are, on

average, zero (typically between 15 and 30 mm Hgwithin the pericardium and between 15 and 30 mmHg within the heart in euvolemic patients) so thatrespiratory reciprocation becomes a principal physi-

CHEST/113/5/MAY, 1998 1373


Tamponade: \COMPENSATION

t INTRAPERICARDIAL PRESSURE

? Ventricular and Atrial Volumes

,t StimulatesY Opposes

t BLOOD VOLUME

f SYSTEMIC &PULMONARY

VENOUS PRESSURES

If Diastolic Ventrjcuiar Pressure-<t Systolic

TJ Stroke Volume

t Atrial Pressure

TACHYCARDIA

\ VENTRICULAR END-SYSTOLIC VOLUME

t EJECTION FRACTION

INOTROPIC EFFECT

\ Cardiac Output

t Arterial Pressure

Yf PERIPHERAL RESISTANCE

ADRENERGIC STIMULATION

Figure 2. Cardiac tamponade: relationships among major hemodynamic events and major compen¬satory mechanisms (see text). Simple arrows= tamponade sequences; pointed arrowheads= stimulatorycompensatory7 actions; blunt arrowheads= oppositions! compensatory actions. Modified from SpodickDH. In: Spodick DH, ed. The pericardium: a comprehensive textbook. New York: Dekker, 1997; 182(with permission).

ologic mechanism contributing at some level tocardiac input and output. A significant component ofrespiratoiy reciprocation is the marked shift of theventricular septum into the left ventricle when inspi¬ration fills the right heart at the expense of the leftwith reversal on expiration. Clinically, respiratoryreciprocation is expressed as pulsus paradoxus.

Vascular and Cardiac Pressure Responses

Since the systemic and pulmonary venous bedsmust generate sufficient pressure to fill both sides ofthe heart, cardiac filling is supported by a parallelrise in systemic and pulmonary venous pressureswhich, in tamponade, are determined primarily bypericardial rather than myocardial compliance.1 Sec¬ondarily, of course, cardiac chamber compliance isreduced by pericardial compression and this pro¬duces the immediate force progressively resistingfilling. In response, blood volume expands (Fig 2).Indeed, part of the ability to tolerate increasingcardiac compression is the rate of venous volumeexpansion. This, like pericardial stretch, requirestime since it occurs by fluid transfer from the tissuesto the venous side of the cardiovascular system. Incontrast, in rapid, hemorrhagic tamponade, eg, dueto cardiac wounds, venous volume expansion is

virtually inoperative. Finally, venous filling may beaugmented at very low ventricular volumes by dia¬stolic suction, the relative effectiveness of which isundetermined.6

Eventually, diastolic pressure in both ventriclesand the pulmonary artery equilibrate with mean

right and left atrial pressures at approximately intra¬pericardial pressure. Conventionally, "equilibrate"represents diastolic pressures differing by no more

than 5 mm Hg. Clinically, this corresponds to "florid"tamponade, when most patients will have frankexaggeration of the respiratory fluctuation in arterialBP.pulsus paradoxus.conventionally a 10 mm Hgor greater inspiratory fall that reflects the exagger¬ated respiratoiy pressure and volume reciprocationbetween the right and left heart chambers. Forexample, during peak inspiration, left heart filling isminimal and its pressure differences from pericardialpressure are least, nil or even negative, with re¬

versed, ie, negative, transmural pressure.36 In suchsevere tamponade, the mitral valve may only openwhen atrial systole occurs during expiration.11 Withinspiration, pulmonary wedge pressure falls belowpericardial pressure. In contrast, right atrial pres¬sure, which tends to "track" pericardial pressure12(Fig 3) also falls, but not below pericardial pressure,enhancing inspiratoiy filling of the right ventricle.

1374 Reviews


Pre-Pe ricardicentesis

Figure 3. Cardiac tamponade. Simultaneous right atrial (RA) and pericardial (Pm) pressure curves;Insp=beginning of inspiration. Patient with atrial flutter. Top: prepericardicentesis: high atrial pressure(mean, 25 mm Hg) tracks high pericardial pressure; in early inspiration, pericardial pressure transientlyfalls below RA pressure. Note QRS electric alternation and atrial flutter. Bottom: postpericardicentesis:mean RA pressure has fallen to 15 mm Hg with little respiratory variation (pericardial restraintremoved). Pericardial pressure has fallen below RA pressure with marked respiratory fluctuations:expiratory level, 8 mm Hg; inspiratory level, approximately zero. Note disappearance of electricalternation; flutter waves on atrial pressure tracing and ECG.

(Tracking is perhaps best demonstrated in experi¬mental animals by intrapericardial pressure mea¬

surement with a flat balloon.)13Tamponading pericardial fluids compress the

heart throughout systole and diastole. Although theatria fill continuously, blood mainly enters the heartwhen blood is leaving it during the right and leftventricular ejection periods, since ventricular ejec¬tion expels blood, reducing ventricular volumes.Ejection thus transiently reduces pericardial pres¬

sure, transiently increasing transmural pressure.Ejection simultaneously aids atrial filling throughenlarging the atria by pulling their "floors" (valvelevels) toward the ventricular apices; this producesthe normal % descent in atrial pressure curves.

Pericardial volume and pressure thus vary continu¬

ously during the cardiac cycle reflecting the varia¬tions in cardiac chamber volumes. Ventricular pres¬sure is high throughout diastole resisting filling. Inearly diastole, the normally rapid peak ventricular

CHEST/113/5/MAY, 1998 1375


filling rate becomes radically reduced, along with thefilling fraction; this increases the relative contribu¬tion to ventricular filling of end-diastolic atrial con¬

traction. At end-diastole, the compressed ventriclesare thus maximally expanded by filling, raising intra¬pericardial pressure to its maximum. Although con¬

tinuous atrial filling tends to expand the atria, whichalso tends to raise pericardial pressure, at end-systole, ventricular ejection (emptying) is completeso that the ventricles are at minimal volume permit¬ting intrapericardial pressure to fall. In tamponade,when ventricular filling is "maximal," the ventriclesremain critically underfilled relative to their normalcapacity (' underpreloaded"). They therefore operateat the low end of their Frank-Starling curves whileejecting the reduced stroke volume.

Intracardiac Pressure Curves

In full-blown tamponade, ventricular pressurecurves remain flat throughout diastole at their highequilibrated level. During systole and ventricularejection, the atrial cavitary pressure drop preservesthe normal x descent in atrial pressure curves.

However, continuously high ventricular diastolicpressures resist rapid ventricular filling so that atrialemptying in early diastole is resisted and ultimatelyaborted. Consequently the y descent of atrial pres¬sure following atrioventricular valve opening is pro¬gressively amputated and ultimately eliminated.Moreover, high ventricular diastolic pressure favorspremature mitral and tricuspid valve closure, furtherresisting filling with further reduction of ventricularpreload reflected in the shorter fiber length of thecompressed, underfilled chambers. This reduces theventricular ejection rate which, with the progressiveunderfilling, progressively reduces stroke volume. Ininspiration, because the right atrium and ventricleexpand at the expense of the left (due to both septalshift and elevation of pericardial pressure; Fig 2),inspiratory left ventricular chamber compliance isminimal. Ultimately the mitral and aortic valves mayopen during expiration only with atrial systole.11Influence of State of HydrationThe progressively increasing diastolic chamber

pressures ultimately equilibrate (averaged over therespiratory cycle) throughout the heart. The typicalrange of 15 to 30 mm Hg applies mainly to euv-

olemic patients and animals. All physiologic phe¬nomena in tamponade occur earlier and at lowerpressures in hypovolemic patients and later and at

higher pressures in hypervolemic than in euv-

olemic patients. Indeed, significantly hypovolemicpatients can have tamponade at average diastolicpressures as low as 6 mm Hg, a form of low-

pressure tamponade.2414 (This can be difficult to

recognize clinically15 particularly in patientstreated with diuretics.)

Coronary Blood Flow

With normal coronary arteries, coronary bloodflow is reduced but remains adequate to supportaerobic metabolism.16 This is due to a proportionatereduction in cardiac work, because the ventricles are

underloaded."underafterloaded" as well as "under-preloaded."615 Even coronary vasodilatory reserve,capacitance, and resistance are not reduced suffi¬ciently to add a measurable ischemic burden.16

Myocardial Status

In otherwise normal hearts, gross systolic functionremains intact17 and the ventricles support strokevolume by an increased ejection fraction (oftenvisibly striking during echocardiography). In con¬

trast, previously diseased or injured hearts may notmaintain cardiac output and arterial pressure as

well,18 and tamponade decompensates earlier.

Compensatory Responses: DecompensatedTamponade

In addition to time-dependent blood volumeexpansion, pericardial stretch, and increased ejec¬tion fraction, compensatory mechanisms for tam¬

ponade include tachycardia and peripheral vaso¬

constriction due to intense adrenergic stimulationevoked by the falling cardiac output.19 Rising rightatrial pressure reflexly contributes to the tachycar¬dia that enhances the minute cardiac output(stroke volumeXheart rate) in the face of thefalling stroke volume (Fig 2).Adrenergic stimulation, both alpha and beta,

and including increased serum catecholamines,19is a principal compensatory response to the re¬

duced cardiac output (Fig 2) with four majoreffects: (1) P-adrenergic contribution to heart rateincreases; (2) (3-adrenergic-dependent augmenta¬tion of diastolic relaxation; (3) a-adrenergicallyincreased peripheral resistance to maintain centralblood pressure and support the gradient for coro¬

nary flow; and (4) increased inotropy to minimizeventricular end-systolic volume via increased ejec¬tion fraction which, as already noted, is normal to

high in tamponade in the absence of cardiacdisease.3'6-12-18

Despite progressively falling cardiac output, in¬creased peripheral resistance supports arterial BPuntil relatively late, partly by an a-adrenergic mech¬anism. Thus, increased peripheral resistance is not

1376 Reviews


affected by |3-blockade, but is opposed by a-block-ade.20 Ultimately BP tends to decline precipitous¬ly.the "last drop" phenomenon. (In experimentalanimals that are anesthetized or not recovered fromsurgery, BP falls relatively slowly in contrast to intactanimals that more closely resemble human respons¬es.23) The critical BP fall is also influenced by an

opioid-dependent mechanism, demonstrated bynaloxone-induced BP increase during tamponade.20This increase occurs without increasing cardiac out¬put, so that this mechanism must act through in¬creased systemic vascular resistance. (Further evi¬dence that systolic function does not limit cardiacoutput in uncomplicated tamponade.9) Naloxone-induced BP increase further suggests a "peripheralresistance reserve," despite the already high periph¬eral resistance that is a principal compensatory re¬

sponse (Fig 2).

Further Neurohormonal Activation

BP is augmented by further neurohormonal acti¬vation associated with arterial and atrial baroreceptorunloading,21 resembling neurohormonal activation incardiac failure (but much less investigated). Therenin-angiotensin-aldosterone system contributes re-

nin, angiotensin II, arginine vasopressin, and aldo-sterone, with consequent sympathetic nervous sys¬tem stimulation, vasoconstriction, decreased urineflow, decreased renal sodium and potassium excre¬

tion, and water retention. These effects are mainlyrelatively late, during decompensating tamponade,after aortic BP decreases by about 30%2-3 and are

followed by increased production of adrenocortico¬tropic hormone. They are preceded by neurogenicreduction in renal sodium output.22 However, unlikecardiac failure at comparably high central pressures,serum atrial natriuretic factor does not increase,because tamponade prevents myocardial stretch.23-25This is associated with the external compression ofthe heart and the critically reduced transmural fillingpressures of tamponade that prevent atrial disten¬tion. (In contrast, transmural pressures are preservedin cardiac failure, permitting myocardial stretch withconsequent atrial natriuretic factor production).Thus, tamponade prevents this mechanism for in¬creasing renal natriuresis. This, with the neurogenicrenal sodium retention,23 contributes to increasedblood volume, tending to support the compensatorilyincreased venous pressures (Fig 2).

Tamponade in Summary

Figure 2 summarizes the principal elements oftamponade dynamics and the main compensatory

mechanisms. The discussion emphasizes the com¬

plexity of each of these, adding the important neu¬

rohormonal elements that resemble those of conges¬tive heart failure. The pathophysiologic continuumof the tamponade state."not all or none"1.is fun¬damental to its appreciation.

References1 Reddy PS, Curtiss El, Uretsky BF. Spectrum of hemody¬namic changes in tamponade. Am J Cardiol 1990; 66:1487-91

2 Bernath GA, Cogswell TL, Hoffman RG, et al. Influences on

distribution of blood flow during cardiac tamponade in theconscious dog. Circ Res 1987; 60:72-81

3 Cogswell TL, Bernath GA, Keelan MH, et al. Laboratoryinvestigation.cardiac tamponade: the shift in the relation¬ship between intrapericardial fluid pressure and volumeinduced by acute left ventricular pressure overload duringcardiac tamponade. Circulation 1986; 74:173-80

4 Spodick DH, Paladino D, Flessas AP. Respiratory effects on

systolic t me intervals during pericardial effusion. Am JCardiol 1983; 51:1033-37

5 Spodick DH. Threshold of pericardial constraint: the pericar¬dial reserve volume and auxiliary7 pericardial functions. J AmColl Cardiol 1985; 6:296-99

6 Spodick DH. The normal and diseased pericardium: current

concepts of pericardial physiology, diagnosis and treatment.J Am Coll Cardiol 1983; 1:240-51

7 Spodick DH. Macro and microphysiology and anatomy ofthepericardium. Am Heart J 1992; 124:1046-51

8 Shabetai R. Function ofthe pericardium. In: Fowler NO, ed.The pericardium in health and disease. Mt. Kisco, NY:Futura, 1985; 22-26

9 Shabetai R. Changing concepts of cardiac tamponade. J AmColl Cardiol 1988; 12:194-95

10 Spodick DH. Inferior vena cava plethoric with bluntedrespiratory response [letter]. J Am Coll Cardiol 1989; 13:1217-18

11 Sakamofo T. Some new or poorly recognized physical andgraphic findings in various forms of pericardial disease. JpnCirc J 1979; 12:149-55

12 Boltwood CM Jr. Ventricular performance related to trans¬mural filling pressure in clinical cardiac tamponade. Circula¬tion 1987; 75:941-55

13 Smiseth OA, Refsom H, Tyberg JV. Pericardial pressureassessed by right atrial pressure: a basis for calculating leftventricular transmural pressure. Am Heart J 1984; 188:603-09

14 Antman EM, Cargill V, Grossman W. Low-pressure cardiactamponade. Ann Intern Med 1979; 91:403-09

15 Reddy PS, Curtiss El, O'Toole JD, et al. Cardiac tamponade:hemodynamic observations in man. Circulation 1978; 58:265-72

16 Lader AS, Mihailescu LS, Abel FL. Effect of pericardialpressure on determinants of coronary flow. FASEB J 1993;7:A319

17 Shabetai R, Fowler NO, Guntheroth WG. The hemodynam¬ics of cardiac tamponade and constriction. Am J Cardiol 1970;26:480-89

18 Hoit BD, Gabel M, Fowler NO. Cardiac tamponade in leftventricular dysfunction. Circulation 1990; 82:1370-76

19 Cogswell TL, Bernath GA, Raff H, et al. Total peripheralresistance during cardiac tamponade; adrenergic and angio¬tensin roles. Am J Physiol 1986; 251 (Regulatory IntegrativeComp Physiol 20):R916-22

20 Klopfenstein HS, Mathias DW. Influence of naloxone on

CHEST/113/5/MAY, 1998 1377


response to acute cardiac tamponade in conscious dogs. Am JPhysiol 1990; 259(Heart Circ Physiol 28):H512-17

21 Hirsch AT, Dzau VT, Creager MA. Baroreceptor function in

congestive heart failure: effect on neuro humoral activationand regional vascular resistance. Circulation 1987; 75(supplIV):36-46

22 Osborn JL, Lawton MT. Neurogenic antinatoresis duringdevelopment of acute cardiac tamponade. Am J Physiol 1986;250(Heart Circ Physiol 19):H195-201

23 Koller PT, Grekin RJ, Nicklas JM. Paradoxical response ofatrial natriuretic hormone to pericardiocentesis in cardiactamponade. Am J Cardiol 1987; 59:491-92

24 Spodick DH. Low atrial natriuretic factor levels and absentpulmonary edema in pericardial compression of the heart.Am J Cardiol 1989; 63:1271-72

25 Wolozin MW, Ortola FV, Spodick DH, et al. Release of atrialnatriuretic factor after pericardiectomy for chronic constric¬tive pericarditis. Am J Cardiol 1988; 62:1323-25

1378 Reviews


pathophysiology of cardiac tamponade* · 2017. 6. 21. · cardiac tamponade is defined as...

Documents