Pericardial Pressure Research Articles

The relationship between pericardial pressure and right atrial pressure: an intraoperative study.

The objective of this study was to determine the constraining effect of the normal human pericardium. Accordingly, immediately after thoracotomy in nine patients undergoing elective cardiac surgery, we measured mean pericardial surface pressure over the lateral free wall of the left ventricle with a flat balloon as well as mean right atrial pressure while incrementally infusing up to 2.1 liters of Ringer's solution to increase right atrial pressure. In each case, the slope of the relationship between right atrial (range -4 to 20 mm Hg, overall) and pericardial pressures was near unity (1.16 +/- 0.20 mean +/- SD) and the intercept was approximately zero (0.71 +/- 2.48 mm Hg). Correlation coefficients ranged from .86 to .97. These observations suggest that right atrial pressure can be used as an estimate of pericardial surface pressure. If this is the case, true left ventricular preload (i.e., effective distending pressure or transmural diastolic pressure) might be estimated from the difference between left ventricular filling pressure and right atrial pressure, both conveniently measurable clinically by means of a triple-lumen, flow-directed catheter.

Assessment of pericardial constraint: The relation between right ventricular filling pressure and pericardial pressure measured after pericardiocentesis

Experimental studies have shown that right ventricular filling pressure (that is, intracavitary diastolic pressure) approximates pericardial surface pressure but, in many patients after removal of pericardial effusion, right ventricular filling pressure has been found to markedly exceed pericardial pressure recorded by an open catheter. The aim of this study was to determine whether this apparent contradiction was related to the technique of pericardial pressure measurement. Nine patients with chronic pericardial effusion were studied and, although these pressures diverged to varying degrees in individual patients, the previous observation was confirmed in that, although initially similar, right ventricular filling pressure and pericardial pressure (measured by means of an open catheter) tended to diverge during removal of the effusate; when the evacuation was as complete as possible pericardial pressure was 2.1 +/- 1.0 (mean +/- SE), while right ventricular filling pressure was 8.7 +/- 1.7 mm Hg (p less than 0.01). In six open chest, anesthetized, volume-loaded dogs with pericardial effusion (50 ml), right ventricular filling pressure and pericardial pressures measured with both open catheter and flat balloon were all equal. With decreasing volume of pericardial fluid, right ventricular filling pressure and pericardial pressure (by catheter) diverged as had been observed in patients. However, pericardial pressure (balloon) continued to be equal to right ventricular filling pressure. (With 0 ml in the pericardium, right ventricular filling pressure = 12.9 +/- 0.9 mm Hg, pericardial pressure [catheter] = 1.4 +/- 1.9 mm Hg and pericardial pressure [balloon] = 12.4 +/- 1.5 mm Hg.) Thus, these observations support the use of right ventricular filling pressure as an estimate of pericardial constraint in patients.

Neurogenic antinatriuresis during development of acute cardiac tamponade.

Renal function was measured during the development of cardiac tamponade in anesthetized dogs. Tamponade was induced by infusion of isotonic saline (37 degrees C) into the pericardial space. Step increases in pericardial fluid volume in 20-ml increments from 0 to 160 ml increased renal efferent nerve activity by a total of 140 +/- 58% of control (from 205 +/- 70 microV/s). Since renal nerve traffic was elevated by pericardial fluid volume infusion prior to any changes in arterial pressure, renal function was determined before and after increasing pericardial pressure (PCP) by approximately 5 and 10 mmHg. Increasing PCP by 5 mmHg decreased urinary sodium excretion and increased renin secretion rate without changing mean arterial pressure, renal blood flow, or glomerular filtration rate (GFR). Further elevation of PCP to approximately 10 mmHg also decreased urinary sodium excretion as well as arterial pressure and GFR. Renal denervation prevented both of these antinatriuretic responses to elevated PCP. In a third group of dogs, similar antinatriuresis was induced by increasing PCP by approximately 5 and 10 mmHg. Bilateral cervical vagotomy abolished the antinatriuretic response to increased PCP by 5 mmHg but was ineffective when PCP was increased by 10 mmHg, which was accompanied by decreased arterial pressure. These results demonstrate that elevation of pericardial fluid volume, and consequently PCP, reflexly decreases urinary sodium excretion via activation of renal sympathetic outflow. Early reflex antinatriuresis results primarily from activation of cardiopulmonary or splanchnic receptors with vagal afferents, whereas, at larger PCP, reflex antinatriuresis may result from hypotension and unloading of high pressure baroreceptors.(ABSTRACT TRUNCATED AT 250 WORDS)

The pericardial hypothesis: a mechanism of acute shifts of the left ventricular diastolic pressure‐volume relation

Changes in LV diastolic P-V relations may be caused by changes in myocardial distensibility and by changes in extraventricular constraint. Experimental studies suggest that the upward shift of the LV diastolic P-V relation associated with pacing tachycardia, in patients with angina pectoris, is due to decreased myocardial distensibility which possibly represents incomplete relaxation. However, shifts in the LV diastolic P-V relation with vasodilator and vasoconstrictor agents seem to be caused by changes in extraventricular constraint. Experimental and clinical data show that such interventions do not significantly change the LV transmural P-V relation. This supports the hypothesis that these shifts are due to changes in pericardial pressure. Our data suggest that such vasoactive agents act by shifting blood between the (splanchnic) venous compartment and the heart, thereby changing heart size and in turn pericardial pressure. These concepts have significantly improved our understanding of the mechanisms of action of vasoactive agents. It seemed a paradox that vasodilators (e.g., nitroglycerine) could substantially lower filling pressure of the failing left ventricle without reducing cardiac output. Because of the downward-shift in the P-V relation with nitroglycerine, preload is virtually unchanged and therefore stroke volume is maintained. Appreciation of these phenomena has considerable impact on how haemodynamic measurements are interpreted. It is obvious that the use of LV end-diastolic pressure as an index of end-diastolic volume may lead to serious misinterpretations of ventricular function. Our demonstration that right atrial pressure might be used to assess pericardial pressure provides a potentially useful way to estimate LV transmural pressure, and therefore an accurate measure of preload.

Ventilation, Circulation and Their Interactions in the Land Crab, Cardisoma Guanhumi

ABSTRACT Physiological variables for ventilation (scaphognathite frequency, branchial chamber pressure and branchial air flow) and for circulation (heart rate, intracardiac and pericardial haemolymph pressure) were measured in the land crab Cardisoma guanhumi (Latreille). Crabs were studied both in air alone and in air with access to a shallow layer of fresh, brackish or sea water. During complete air exposure, forward scaphognathite beating predominated and reversed scaphognathite beating was very infrequent. Periods of apnoea were rare. When crabs were able to immerse the Milne-Edwards openings to the branchial chambers in water, scaphognathite reversal occurred much more frequently, and most air flow through the branchial chambers was generated by this mode of ventilation. Changes in water salinity had no effect on respiratory patterns. The cyclic variation between forward and reversed scaphognathite beating appears to serve not only to ventilate the branchial chambers with air, but also to flush water through the branchial chambers for non-respiratory purposes such as ion, water and acid-base regulation. Haemolymph pressures were comparatively low (14 mmHg systolic, 6mmHg diastolic). During diastole a pressure gradient of approximately 0·6 mmHg existed between the pericardial space and the heart lumen. Pauses in heart beat were never observed. Circulatory events were closely coordinated with adjustments in ventilation. Reversed scaphognathite beating produced a transient increase in systolic and diastolic haemolymph pressure due to the rise in branchial air pressure acting directly upon the large, haemolymph-filled spaces lining the branchial chambers. A transient bradycardia accompanied this brief rise in central haemolymph pressures. Possible mechanisms for the regulation of haemolymph pressure are discussed.

Myocardial blood flow during cardiac tamponade in dogs with coronary occlusion: Effects of isoproterenol

To determine the effects of cardiac tamponade in dogs with ligation of the left anterior descending coronary artery, fluid was introduced into the pericardial space to raise right and left atrial and pericardial pressures, first to 7 to 9 mm Hg and then to 11 to 12 mm Hg. Normal and ischemic myocardial blood flow fell approximately 20% to 25% during mild tamponade (1.27 ± 0.16 to 1.00 ± 0.06 ml/min/gm and 0.52 ± 0.12 to 0.39 ± 0.08 ml/min/gm, respectively) and by 50% during moderate tamponade (0.66 ± 0.08 and 0.23 ± 0.05 ml/min/gm, respectively). The inner/outer left ventricular wall blood flow ratio decreased modestly from 1.16 to 1.08 ( p < 0.025) in normal areas but increased from 0.53 to 0.61 ( p < 0.05) in the ischemic regions, suggesting possible epicardial vessel compression. Isoproterenol resulted in prompt decreases in pericardial and filling pressures, 16% increase in aortic pressure, and 200% rise in cardiac output. Normal myocardial blood flow more than doubled (1.55 ± 0.12 ml/min/gm, p < 0.001). Although average ischemic blood flow rose slightly to 0.42 ± 0.10 ml/min/gm, the increase was not significant. Furthermore, changes in ischemic blood flow were heterogeneous with frank decrease in one dog. Therefore, although isoproterenol has salutary hemodynamic effects, its unpredictable action on myocardial blood flow should cause one to use it cautiously in those with tamponade who are believed to have coronary occlusive disease.

Function of the pericardium and pericardioperitoneal canal in elasmobranch fishes.

Previous studies of cardiac function in elasmobranch fishes have not included the influence of the pericardioperitoneal canal on pericardial pressure and volume and thus on cardiac function. Accordingly, we studied the function of the pericardium and pericardioperitoneal canal in sharks and rays. We found negative pericardial pressure that rose to a plateau of approximately 0 mmHg when fluid was infused into the pericardium with the canal undisturbed. However, this pericardial pressure elevation caused severe cardiac tamponade. After the canal was occluded, the pressure plateau was substituted with an exponential rise. We injected radioisotopes into the pericardial cavity and obtained scintigrams several hours later. The scans and counts of body fluids and tissues indicated absorption, disputing the suggestion that the primary function of the canal may be inadequate absorption of pericardial fluid. We conclude that the pericardioperitoneal canal maintains negative pericardial pressure, which is a prerequisite in elasmobranch fishes and may serve to regulate pericardial pressure level to optimize cardiac function in relation to changes in cardiac size.

Assessment of pericardial constraint in dogs.

To determine the better method of measuring pericardial constraint, pericardial pressure was recorded by a liquid-filled open-ended catheter and a liquid-containing flat balloon in six open-chest anesthetized dogs. Left ventricular pressure was measured by a micromanometer-tipped catheter and left ventricular anteroposterior diameter was measured by sonomicrometry. Left ventricular end-diastolic pressure was raised to 20 +/- 1.7 (mean +/- SD) mm Hg by intravenous saline. Left ventricular diastolic pressure-diameter loops were constructed (1) with incremental amounts of saline (0 to 50 ml) in the resealed pericardium, (2) with several small holes in the pericardium, and (3) with the pericardium widely open. Measured pericardial pressures were compared with what was assumed to be the correct pericardial pressure, i.e., the calculated difference between left ventricular diastolic pressure (at a given left ventricular diameter) before and after opening the pericardium. Pressure recorded by the flat balloon was similar to the calculated pericardial pressure at all pericardial liquid volumes. Pressure recorded by the open-ended catheter, however, was significantly lower (p less than .05) than the calculated pressure unless there was at least 30 ml of liquid in the pericardium. After several holes had been made in the pericardium it still exerted a constraining effect, as shown by a marked rightward or downward shift of the left ventricular diastolic pressure-diameter relationships after completely opening the pericardium. After holes were made in the pericardium pressure recorded by the flat balloon was still similar to the calculated pericardial pressure. However, pressure recorded by the open-ended catheter was significantly (p less than .02) lower than the calculated pressure.(ABSTRACT TRUNCATED AT 250 WORDS)

Distribution of stresses in the left ventricular wall of the intact heart

The left ventricle is modelled as a prolate spheroid of viscoelastic material with an aim to demonstrate the qualitative effects of anistropy and nonhomogeneity in the calculation of intact ventricular wall stresses. The pericardial pressure is accounted for in the analysis and an attempt is made to examine to what extent this influences the ventricular stresses. Numerical results are also obtained by computing the analytical expressions derived through the analysis.

Cardiac tamponade in dogs with normal coronary arteries. II. Myocardial flow and metabolism with moderate and severe hemodynamic impairment.

To determine the effects of cardiac tamponade on myocardial blood flow and its distribution, dogs were prepared with indwelling pericardial catheters. Hemodynamic, myocardial blood flow, and myocardial metabolic data were collected in 5 closed-chest, spontaneously breathing animals with normal blood volumes and hemoglobin concentrations and 6 with acute anemia. Instillation of an average of 89.0 +/- 14.9 ml of modified Normosol into the pericardial space in dogs with normal hemoglobin levels produced mild tamponade with a modest decline in aortic pressure (119.5 +/- 14.3 to 96.8 +/- 12.1 mm Hg) and significant rises in left and right atrial and pericardial pressures to 7-8 mm Hg. Increasing the pericardial volume to 124.0 +/- 13.6 ml produced hypotension (mean aortic pressure 86.2 +/- 10.5 mm Hg) and rises in the left and right ventricular filling pressures and pericardial pressure to 10-11 mm Hg. Total myocardial blood flow fell from 1.19 +/- 0.18 to 0.73 +/- 0.17 ml/min/g (p less than 0.02) during mild tamponade, and fell further to 0.56 +/- 0.17 ml/min/g (p less than 0.05) with more severe tamponade. Despite these declines, the left ventricular wall inner/outer flow ratio and left ventricular flow as a proportion of total cardiac output were unchanged. In dogs with anemia more severe tamponade was created, with consequently more marked hemodynamic abnormalities. However, the relative changes in myocardial blood flow and inner/outer flow ratio were similar. Myocardial metabolic parameters could be evaluated only in the dogs with less severe tamponade.(ABSTRACT TRUNCATED AT 250 WORDS)

Pericardial pressure assessed by right atrial pressure: A basis for calculation of left ventricular transmural pressure

Pericardial pressure assessed by right atrial pressure: A basis for calculation of left ventricular transmural pressure

Mechanism of action of nitrates: Role of changes in venous capacitance and in the left ventricular diastolic pressure-volume relation

Nitroglycerin, when administered to patients with heart failure, causes a marked reduction in left ventricular filling pressure but often an increase in stroke volume and stroke work; based on the Frank-Starling principle, such a reduction in “preload” would be expected to result in a decrease in left ventricular end-diastolic volume and, therefore, a decline in stroke volume. Assessment of pressure-volume coordinates, however, has revealed that nitroglycerin produces a downward shift in the pressure-volume relationship. This apparent improvement in left ventricular compliance cannot be attributed to alterations in the elastic properties of the myocardium but rather appear to reflect a reduction in left ventricular external constraint. Recent animal and clinical investigations in our laboratory suggest that nitroglycerin causes venous dilatation (particularly in the mesenteric bed), thereby decreasing venous pressure at any given vascular volume. This decrease in cardiac filling pressure results in a decrease in heart size and, therefore, a reduction in pericardial pressure. Left ventricular transmural (intracavitary minus pericardial) pressure is little changed, however, so that end-diastolic volume and stroke volume are maintained.

Ventricular diastolic pressure-volume shifts during acute ischemic left ventricular failure in dogs

Ischemic left ventricular failure was produced in eight acutely instrumented, anesthetized dogs to study the contribution of changing myocardial compliance and pericardial pressure to shifts in right and left ventricular diastolic pressure-volume relations. Right and left ventricular and pericardial volumes were measured by ungated computed tomography. Cardiac volumes were manipulated by infusion of saline solution, hemorrhage, phenylephrine infusion and, during failure only, nitroglycerin administration. During both control and failure periods, these interventions shifted the left and right ventricular pressure-volume relations by changing pericardial pressure only; that is, these interventions caused no change in the ventricular transmural pressure-volume relation. The induction of failure as such increased pericardial pressure only minimally and did not change the left ventricular or right ventricular transmural pressure-volume relations significantly. Volume loading during the control period caused an apparent pericardial creep which attenuated the pericardial effect on ventricular pressure-volume relations. During failure, volume loading caused an increase of right ventricular volume, but tended to decrease left ventricular volume; this was associated with a leftward displacement of the interventricular septum. In conclusion, in the presence of ischemic left ventricular failure as well as normally, changes in preload, afterload and circulating blood volume shift ventricular diastolic pressure-volume relations by stretching or relaxing the pericardium, thus changing pericardial pressure. In these circumstances, there were no consistent changes in myocardial compliance.

Pericardial influences on ventricular filling in the conscious dog. An analysis based on pericardial pressure.

Twenty-five dogs were chronically instrumented to investigate the effects of the normal pericardium on cardiac function. Pulse-transit ultrasonic transducers were implanted to measure multiple ventricular dimensions. The pericardium was incised transversely at the base of the heart and precisely reapproximated, so as to disturb its characteristics minimally. One week later, the dogs were studied in the conscious state, and left ventricular, right ventricular, pericardial, and pleural pressures were measured with matched micromanometers. Data were recorded before and after blood volume expansion. Absolute end-diastolic pericardial pressure varied directly with pleural pressure during the respiratory cycle. Transpericardial pressure (pericardial-pleural pressure) varied little with respiration and was related directly to ventricular diameter during the cardiac cycle with peak transpericardial pressure uniformly occurring at end-diastole. With volume infusion, normalized end-diastolic minor axis diameter and left ventricular transmural pressure (left ventricular-pleural pressure) increased significantly from 0.14 +/- 0.01 and 9.5 mm Hg +/- 1.0 mm Hg, respectively, in the control state to 0.20 +/- 0.01 and 19.3 mm Hg +/- 1.2 mm Hg after volume loading. End-diastolic transpericardial pressure also increased significantly from 2.3 +/- 0.5 mm Hg to 4.1 +/- 0.3 mm Hg, and represented approximately 21% of transmural left ventricular pressure. When measurements were obtained sequentially after implantation, transpericardial pressure was initially high but decreased with time, presumably due to pericardial creep. After volume loading, right ventricular end-diastolic transmural pressure averaged 9.6 mm Hg, and pericardial pressure constituted 42% of right ventricular pressure. Thus, pericardial restraining effects may predominantly influence right ventricular filling and affect the left ventricle through series interaction. In the normal conscious dog, transpericardial pressure remains low over the entire physiological range, and the direct influence of the normal pericardium on diastolic filling of the left ventricle appears to be minimal.

The role of the pericardium in interactions between the cardiac chambers

Distension of one side of the heart involves both the strium and the ventricle, and such atrioventricular enlargement encoroaches more on pericardial volume than would ventricular distension alone. The influence of distension of the entire right side of the heart on the pressure-volume relationship of the entire left heart was studied in six postmortem canine hearts with intact pericardium. The pressure-volume relation of the left heart was determined when the right heart was empty and when it was filled with saline. The pressure-volume curve of the left heart became steeper when the volume of the right heart was increased. However, after subtracting pericardial pressure from the left heart pressure, the pressure-volume curves were unaffected by increased volume of the right heart. Selective distension of the entire right heart has a considerable effect on the filling characteristics of the left heart when the pericardium is intact, although this is less than that observed in experiments in which the more compliant atria have been excluded. This effect becomes negligible after subtracting pericardial pressure.

Apparent depression of right ventricular function after selective reduction of left ventricular inotropy: an interpretation of experimental data utilising a computer-based circulatory model.

Consequences of ventricular interactions upon commonly used descriptions of ventricular diastolic and systolic properties were inferred from a circulatory model. The model included active heart compartments, vascular interconnections and the pericardium and was adapted on a digital computer. Numerical values assigned to model parameters were such as to resemble the circulatory system of the dog. Model predictions were utilised to analyse experimental observations in a dog preparation with closed chest and pericardium. Dog ventricular and pericardial pressures (P) and volumes (V) were measured in a control situation and after ischaemic depression of left ventricular (l.v.) function. Model predictions and experiment observations were qualitatively identical. During l.v. inotropic depression (in model and experiment the stroke volume was reduced by about 50% and the l.v. end-diastolic pressure (e.d.p.) increased by about 10 mmHg) the right ventricular (r.v.) diastolic P-V relationship shifted leftwards/upwards while that of the l.v. shifted rightwards/downwards. When corrected for the effect of pericardial pressure the r.v. and l.v. diastolic P-V relations fitted single curves. R.v. function curves (stroke work against dastolic pressure or volume) indicated systolic depression of the r.v. The model indicated that the observations were results of encoroachment of a dilating l.v. upon the r.v. The model indicated that the of increased r.v. afterload. The findings imply that the diastolic P-V relationship of one heart compartment cannot be assessed by measurements in this compartment alone, and that function curves of one ventricle are affected by ventricular interactions which cannot be separated from true alterations in the inotropic state.

The hemodynamic derangement associated with right ventricular diastolic collapse in cardiac tamponade: an experimental echocardiographic study.

Clinical reports indicate that right ventricular diastolic collapse (RVDC) is associated with cardiac tamponade. To assess the hemodynamic abnormalities associated with RVDC, we studied six chronically instrumented conscious dogs (group A) with two-dimensional echocardiography during cardiac tamponade induced by continuous saline infusion into the pericardial space. We recorded cardiac output (electromagnetic flowmeter), heart rate, respiration, and aortic, intrapericardial, and right atrial pressures. In four additional animals (group B), we recorded right ventricular pressure and placed a hydraulic occluder around the pulmonary artery so that short-term reversible obstruction to right ventricular outflow could be produced. None of the dogs had RVDC when the pericardial space was empty, but all dogs showed RVDC during cardiac tamponade. The appearance of RVDC in group A was associated with a 21% reduction in cardiac output (p less than .01) and no change in mean aortic pressure. Short-term partial pulmonary artery obstruction led to increased right ventricular pressures and a striking reduction in RVDC in early tamponade, suggesting that RVDC is caused by pericardial pressure exceeding right ventricular pressure in early diastole. An additional animal had right ventricular hypertrophy caused by a severe "heart worm" infestation and did not show RVDC during cardiac tamponade. These observations suggest that in the absence of increased resistance to right ventricular outflow or right ventricular hypertrophy, RVDC occurs early in the course of cardiac tamponade and is associated with a hemodynamically important decrease in cardiac output.

Effects of cardiac tamponade on colonic hemodynamics and oxygen uptake.

The local hemodynamic response of the innervated but vascularly isolated colon to decreased systemic perfusion induced by cardiac tamponade was studied in anesthesized dogs as a model of nonocclusive mesenteric ischemia. Increasing levels of pericardial pressure caused progressive decreases in colonic blood flow associated with substantial increases in colonic vascular resistance. These increases in local colonic resistance were proportionately larger than concurrent increases in systemic resistance. The disproportionate response of the colonic resistance vessels was not diminished by colonic (sympathetic) denervation. Reductions of blood flow to 30 ml . min-1 . 100 g-1 resulted in compensatory increases in colonic oxygen extraction such that colonic oxygen consumption remained constant (flow independent) at about 1.5 ml . min-1 . 100 g-1. At blood flows below 30 ml . min-1 . 100 g-1 colonic oxygen consumption was markedly dependent on blood flow. This fundamental relation of colonic oxygen consumption to blood flow was the same whether ischemia was induced by cardiac tamponade, partial mechanical arterial occlusion, or vasoconstrictor (norepinephrine or digoxin) infusion. Furthermore, this relationship was not altered by vasodilation with isoproterenol after the induction of ischemia by any of the above means.

Documentation of a venous resistance in the baboon

In the steady state, blood flow out of the heart must be the same as blood flow into the heart. Thus, flow in the cardiovascular system should depend not only on cardiac determinants of flow—pericardial pressure, afterload, heart rate, and contractility—but also on systemic venous determinants of flow as well. We studied the venous determinants of flow in nine baboons. The mean systemic pressure, the upstream driving pressure for venous return, was 10 mm Hg higher than the right atrial pressure. Flow generated by this upstream pressure was modulated by a venous resistance. This resistance acted as if it were interposed between the upstream capacitance vessels and the right atrium. These determinants of venous return in the baboon would be inadequately described by classical concepts of cardiovascular physiology. We suspect that classical concepts would be inadequate in man as well.

Effects of Hypo-, Normo-, and Hypercarbia in Dogs with Acute Cardiac Tamponade

The hemodynamic effects of changes in PaCO2 during intermittent positive pressure ventilation (IPPV) were studied in nine dogs with acute cardiac tamponade. During steady state light thiopental anesthesia, measurements were performed during hypocarbia (24.0 +/- 2.6), normocarbia (40.4 +/- 2.4), and hypercarbia (56.8 +/- 3.1 mm Hg; mean +/- SD). The study was carried out at a standardized level of cardiac tamponade that gave a 60% reduction in cardiac output (CO) at normocarbia. Changes in airway pressure were avoided by adding CO2 to the inspiratory gas to obtain the desired PaCO2. Hypercarbia increased pericardial pressure 2-4 mm Hg and significantly decreased CO. During hypocarbia CO increased as pericardial pressure decreased 3-6 mm Hg. These findings are the reverse of changes seen when tamponade is not present. The changes in pericardial pressure most likely influence myocardial tone and cardiac volume and, thus, CO. The results suggest that patients with cardiac tamponade requiring general anesthesia should not breathe spontaneously if there is any danger of respiratory depression and hypercarbia.