Dynamic Right and Left Ventricular Interactions in the Pig

Abstract

We characterize dynamic factors determining ventricular interdependence with and without the pericardium. We simultaneously measured right (RV) and left ventricular (LV) pressures and volumes using conductance catheters in 7 pentobarbital-anesthetized open-chested 5–7-week old piglets. We studied these effects during apnea, inferior vena caval occlusion and rapid partial aortic and pulmonary arterial occlusions. Conductance catheter-defined long axis regional volumes were assessed to define regional contractile synchrony. Closed pericardium measures were made from an initial (baseline) volume then after two 20 ml/kg fluid loads followed by an open pericardium step. Baseline RV and LV volumes were similar. Aortic occlusion increased LV pressures and volumes and RV end-systolic pressure such that RV end-systolic elastance increased without changes in RV contraction synchrony and not affected by the pericardium. Pulmonary artery occlusion increased RV end-systolic pressure but not end-systolic volume. On the subsequent beat RV end-diastolic pressure increased and LV end-diastolic volume and diastolic compliance decreased. These effects were attenuated by opening the pericardium. Contraction synchrony across longitudinal segments was unaltered by either aortic or pulmonary artery occlusion. We conclude that the determinants of systolic and diastolic ventricular interdependence are different. Increasing RV pressures causes diastolic RV to LV interdependence decreasing LV diastolic compliance and dependent on an intact pericardium. Increases in LV end-systolic pressure increases RV end-systolic elastance independent of pericardium, and minimally alters RV diastolic function or contraction synchrony.

Introduction

Ventricular interdependence may alter either diastolic or systolic function of the opposite ventricle. It is influenced by pericardial restraint and accounts for many of the differences seen in dynamic versus steady-state ventricular function as loading conditions to one ventricle varies (Bove & Santamore 1981). Furthermore, selective changes in ejection pressure of one ventricle may occur in disease states, although, the function of both ventricles often appears to be altered. Previous studies have examined the effects of either selective RV loading on LV function or LV loading on RV systolic function (Badke 1982, Brinker et al. 1980, Chow & Farrar 1989, Fenely et al. 1985, Moulopoulos et al. 1965, Santamore et al. 1976, Santamore & Dell’Italia 1998, Visner et al. 1983). These studies have demonstrated that selective increases in LV end-diastolic volume or ejection pressure will increase the force of RV ejection (Santamore et al. 1976), presumably because of direct mechanical coupling between the two ventricles via the common myocardial fibers and intraventricular septum. Unfortunately, few studies have attempted to study the effects of selective changes in loading conditions of one ventricle on the pressure-volume relation of both ventricles simultaneously and on the impact of such loading conditions of contraction synchrony within each ventricle. In part, this lack of investigation is due to inherent difficulties in measuring RV volumes in the intact animal. Although estimates of both RV and LV volumes can be made using ultrasonic crystal methodologies, they may not reflect complex structural changes seen during the cardiac cycle, are highly invasive and time consuming in their application. Similarly, the associated effects selective increases in RV or LV ejection pressure have on systolic function of both ventricles using ventricular pressure-volume relations is poorly characterized. Such an analysis is important because the two ventricles are mechanically coupled, such that changes in loading conditions of one ventricle may alter the performance of both ventricles. Accordingly, any analysis of the determinants of a change in ventricular function of one ventricle is more accurate if assessed for both ventricles simultaneously.

Dynamic evaluation of LV pressure-volume loops using conductance catheter technology (Baan et al. 1984) can also be applied to assessing RV performance (McKay et al. 1984, Santamore et al. 1976). However, the asymmetrical RV cross-sectional area and U-shaped internal longitudinal structure make accurate measurement of RV volumes using this technology difficult (Geddes et al. 1999). To circumvent this problem, we previously limited our application of conductance catheter technology to hearts whose internal RV dimensions and volume are small enough to minimize this concern (Pinsky et al. 1996, Solda et al. 1992). Those data demonstrate that changes in loading conditions of one ventricle alters bi-ventricular function, with diastolic interactions being primarily from right to left, whereas systolic interactions being from left to right. Previous workers have demonstrated that selective increases in LV systolic wall tension increases RV developed pressure (Santamore et al. 1996), presumably through mechanical linkage of epicardial myofibrils and not through intraventricular septal thickening. We postulated that the instantaneous ventricular interdependence processes from left to right and right to left are fundamentally different. Thus, we hypothesized the left to right systolic ventricular interdependence would be unaffected by the pericardium and primarily decrease RV end-systolic volume by shifting its end-systolic pressure-volume relationship (ESPVR), with volume on the x-axis upward and to the left. Whereas, increases in RV ejection pressure though inducing similar increases in RV ESP would not improve RV ejection. However, the associated increase in RV end-diastolic volume (EDV) on the subsequent diastole would decrease LV diastolic compliance reducing LV systolic performance by the Starling mechanisms. Furthermore, this effect would be both intravascular volume and pericardial restraint dependent because RV diastolic compliance is very high (Dhainaut et al. 1997, Pinsky et al. 1996, Schwiep et al. 1988, Solda et al. 1992) and changes in RV EDV can occur with minimal changes in RV EDP. Accordingly, LV ejection is dependent on LV EDV, but RV ejection is not. Pericardiectomy, by removing the external pericardial constraint should selectively reduce RV to LV diastolic ventricular interdependence but minimally affect LV to RV systolic interdependence. Finally, the impact of instantaneous systolic loading on either ventricle’s contraction synchrony is unknown. Since ventricular contraction asynchrony can alter ESV (Strum & Pinsky 2000a, Strum & Pinsky 2000b) and the position of the ESPVR (Johnson et al. 2009), some of the increase in RV ejection pressure due to increasing LV ejection pressure may result from improvement in RV contraction synchrony.

Thus, we studied the instantaneous biventricular effects of step increases in univentricular systolic pressure on both ventricles simultaneously in an open-chested porcine model with both an intact and open pericardium. We tested two related hypotheses based on the findings of our original study (Pinsky et al. 1996). First, that the changes in RV systolic function are fundamentally different whether RV ejection pressure or LV ejection pressure are selectively increased. Second, that pericardial restraint alters only the magnitude of the diastolic RV to LV interaction, not its directional effects and that systolic LV to RV interactions that improve RV ejection performance are not influenced by either the pericardium or RV contraction synchrony.

Materials and Methods

Ethical Approval:

We received approval of the protocol by the Institutional Animal Care and Use Committee of the University of Pittsburgh (PRO-HL0731980) and all experiments were carried out according to the guidelines laid down by the institution’s animal welfare committee, and conform to the principles and regulations of this journal.

Preparation:

Ten 5 to 7-week old piglets (17 ± 1.5 kg body wt, x ± SD) were acquired from an FDA-approved pig farm (Whippo Farms, Inc, Saltsburg, PA), and transported to our animal holding facility. They were allowed free access to food until 6 hours before surgery and free access to water up until the time of surgery. At the start of the experiments, the animals were anesthetized with ketamine (10 mg i.m.) followed by isoflurane inhalational anesthesia (1–4% for effect) for endotracheal intubation and intravenous catheter insertion. Maintenance anesthesia was a continuous infusion of pentobarbital at 2–4 mg/kg/hr. Once pentobarbital was started, isoflurane was discontinued. Subsequent boluses of pentobarbital sodium (50–100 mg iv.) were given as necessary to abolish spontaneous movements, acute elevations in systolic arterial pressure and the eye blink reflex. The animals were placed supine on a surgical cradle and kept warm on a heating pad at 38° C, as measured by a rectal probe. A continuous recording of standard lead II from the electrocardiogram was made to determine heart rate (HR) and to identify abnormally conducted ventricular beats. Each animal had a tracheotomy performed and the trachea was intubated with a 7-Fr cuffed endotracheal tube, and was ventilated (Siemens-Servo 900B, Selna, Sweden) at a tidal volume of 12 ml/kg and a frequency sufficient to maintain an arterial PCO₂ between 38 and 42 mmHg. End-tidal CO₂ was monitored continually (Hewlett-Packard, Palo Alto, CA) and arterial blood gases were periodically sampled (Radiometer ABL-30, Copenhagen, Denmark), and arterial pH was maintained between 7.37 and 7.42 by bolus intravenous infusion of NaHCO₃ as necessary. A 5-Fr catheter with multiple side holes was inserted into the right femoral artery for measurement of arterial pressure and for arterial blood sampling. A 7-Fr polyvinylchloride triple lumen catheter was also inserted into the intra-abdominal inferior vena cava via the right femoral vein for the administration of a continuous infusion of pentobarbital, fluid resuscitation and for venous blood sampling. The arterial vascular catheter was connected to a MP-50 pressure transducer (Gould Inc., Cleveland OH) and the airway catheter to a 23dB pressure transducer (Gould Inc., Cleveland OH). A 3-Fr high-fidelity pressure transducer (Millar Micro-tip catheter model SPC-350, Houston, Texas) was inserted into the left ventricle via the left carotid artery under fluoroscopic guidance to record LV pressure. A 3-Fr, 7-pole single field conductance catheter (custom built, Webster Laboratories, Irvine, CA) was placed in the left ventricle through the aortic valve under fluoroscopic guidance to measure LV volume on a continuous basis.

A midline sternotomy was performed. A small 0.5 cm incision was made in the pericardium over the great vessels and umbilical tape snare ligatures were placed loosely around the pulmonary arterial outflow tract and the ascending aorta. Care was taken to ensure that this incision did not overlay the myocardium nor grossly alter LV ejection, as assessed by continual observation of the LV conductance catheter signal. An additional umbilical tape snare ligature was placed around the supra-diaphragmatic inferior vena cava (IVC). A second 3-Fr high-fidelity pressure transducer was placed into the RV through the tricuspid valve from an internal jugular vein approach to measure RV pressure. A second impedance catheter was placed into the RV via the pulmonary outflow track insertion distal to the pulmonic valve, to measure RV volume. This catheter was held in place by a purse string suture on the anterior surface of the pulmonary artery. Positioning of the LV and RV conductance catheters was verified prior to data collection by direct palpation of the heart once the chest was opened, repeat fluoroscopy after completion of all surgical manipulations, and by repeated inspection of the regional LV and RV segmental volume signals with respect to time. Finally, the position of all vascular catheters was verified by direct inspection at necropsy following the experiment. Due to difficulties in the surgical preparation only 7 of the 10 animals were able to start the protocol. Two animals were excluded because of severe damage to their hearts or great vessels during instrumentation. A third animal developed intractable ventricular arrhythmias. The remining seven animals were hemodynamically intact, without evidence of continual surgical bleeding, arrhythmias, or hypotension throughout the study intervals. A schematic diagram of the experimental preparation is shown in Figure 1.

Figure 1.

Schematic presentation of the experimental preparation, measures made at each step and the four sequential volume loading or pericardiotomy steps.

After initial data collection with the pericardium closed (infra vide), the pericardium was widely incised over the anterior pericardial surface and electromagnetic flow probes (Carolina Medical, King, NC) were placed around either the pulmonary outflow tract (n=4) or aortic outflow tract (n=3). The site of flow probe insertion was determined by anatomical limitations. These flow probes were used to compare RV and LV stroke volume measures from the respective conductance catheter volume signals, prior to completing the open pericardium part of the protocol. These flow probes were not removed during the subsequent open-pericardium studies but were turned off to minimize electrical interference with the conductance catheter signals. In one animal (Pig #4), the RV conductance catheter was inadvertently removed during flow probe insertion and needed to be re-inserted. Post-reinsertion RV volume signal was smaller than the initial segmental volume, consistent with incomplete insertion. Thus, the data from this animal was not used to compare open to closed pericardium states but was used to compare the effects of interventions during closed pericardium states. Thus, within the closed pericardium group all 7 animal’s data were used, but between open and closed pericardium conditions only the 6 paired animals’ data were used for statistical analysis and reported in the two summary tables.

All vascular pressures were measured relative to atmosphere. RV and LV EDV were defined as the maximal ventricular volumes prior to ejection with LV and RV EDP defined as LV and RV pressures at their respective EDV. End-systolic volumes represented the minimal volumes during ejection with LV and RV ESP taken at their respective ESV. End-systolic and peak systolic pressures were further determined by inspection of the pressure-volume curves of the respective ventricles.

Before starting the experimental protocols, the animals were hemodynamically stable with a LV EDP of 5–12 mmHg, with no change in HR or peak systolic LV pressure, and no arrhythmias for 15 minutes. If fluid resuscitation was needed to maintain a LV EDV >5 mmHg, then a mixture of normal saline and 5% dextran was infused after it was blended to derive an identical resistivity (Rho) as the animal’s own blood as described below.

Conductance Catheter Volume Signals:

The RV and LV time-volume signals for each segment of both conductance catheters were inspected individually, as recommended by the manufacturer. Briefly, segments distal to the apex of the heart may be outside the ventricular lumen and thus not contribute to ventricular volume changes. Segmental volume signals were excluded from the sum signal if they either displayed paradoxical volume changes during systole (increasing volume as pressure increased) or if their volume change during the cardiac cycle was less than 10% of their baseline volumes. However, once a segment demonstrated normal volume behavior with respect to time, all segments proximal to it and the catheter tip were included for subsequent analysis. This quality control process is important because over the course of an experiment a conductance catheter can migrate into or out of the ventricle depending of changes in ventricular size, necessitating adding or removing segmental volumes in the final pressure-volume loop analysis. In practice, no animal altered its LV segmental volume contributions, but three animals increased segmental contributions of their right ventricles from 3 to four segments upon opening the pericardium. In these animals, common segmental data were used for maneuvers within each condition, but different segmental data were used between closed and open pericardium conditions.

The conductance catheter method for measuring ventricular volume has been described and validated previously (Baan et al. 1984). Conductance catheter signals were collected and processed using two matched conductance catheter data processor and signal conditioners (Leycom Sigma 5DF, Leyden, Netherlands) whose frequency of activation was altered by the manufacturer to allow them to operate in close proximity without electrical cross talk at 20-kHz, constant-amplitude current of 30 μA RMS in a single-field format. A four-electrode calibration chamber was used to determine blood conductivity (α value). The blood Rho (reciprocal of conductivity) was measured three times during the study: once prior to the first series of measurements, again after the pericardium was opened, and then at the end of the experiment. Parallel conductance artifact for both ventricles was determined separately using a single injection of 5 ml of 10 N saline injected into the inferior vena cava by the saline dilution method (Foex & Leone 1994).

To further validate the RV and LV volume signals, we measured RV and LV stroke volumes from both conductance catheter and electromagnetic flow probe during both steady state apnea and as bi-ventricular stroke volumes rapidly decreased during transient total IVC occlusion. The paired conductance catheter and flow probe data during IVC occlusion were compared by linear regression whereas five sequential beats during apnea for both ventricles were compared by a Bland-Altman analysis for bias. We defined accurate conductance catheter volume data if they were <10% different from flow probe data during apneas before and after the IVC occlusion test. If not, the conductance catheter was repositioned and recalibrated. In practice, this repositioning was required in 3 of the 7 pigs prior to data collection. The comparison was done over two IVC occlusion maneuvers, one with flow probe on and other with conductance catheters on to minimize electrical interference between the two measuring systems. All reported data displayed tight correlation between conductance catheter and flow probe stroke volume estimates with <10% SD.

Since the single field conductance catheter format for measuring ventricular volumes does not minimize parallel conduction artifact from one ventricle to the other (Steendijk 1993), we further processed each ventricular volume signal as follows. We (Pinsky et al. 1996, Solda et al. 1992) and others (Kun & Peura 1994, Stamato et al. 1995) have previously shown that 5% of the volume of the opposite ventricle is sensed as “self-blood” volume of the measured ventricle due to parallel conduction artifact of blood when using the single field conductance catheter technique. Thus, we simultaneously subtracted 5% of total ventricular volume from one ventricle from the other on a point-by-point basis throughout the cardiac cycle. These data are referred to as volume corrected data. Volume corrected stroke volumes more accurately reflected simultaneously measured stroke volumes by flow probe than did the raw data. Accordingly, all volume data reported here are volume corrected data.

Finally, we analyzed the regional volume signals that are used to construct global LV and RV volume histories. The conductance catheter produces an electrical field within the ventricular cavity in which equipotential planes may be defined between each of the intervening electrodes at right angles to the long axis of the catheter. The volume of blood measured between any two sensing electrodes is a disc with boundaries defined by the endothelial surfaces between to the electrodes. The sum of all such discs within a chamber reflects total chamber volume. We used the individual segmental time-volume signals to assess asynchrony of contraction among regions of each ventricle, as previously validated by us (Strum & Pinsky 2000a).

Data Processing:

We recorded arterial blood pressure, airway pressure, LV pressure, LV dP/dt, total and regional LV conductance (volume), RV pressure, total and regional RV conductance (volume) and lead II of the ECG. All data were recorded on an eight-channel physiologic recorder (Gould, Cleveland, Ohio) to monitor for hemodynamic stability and to follow the effects of our interventions. All data were also digitized at 256 Hz for subsequent analysis using a Hewlett-Packard/Apollo DN4000 Unix workstation and a SignifiCAT^® RTS-132 A/D subsystem.

Protocol:

The protocol was performed twice: at three levels of increasing intravascular volume with the pericardium intact, referred to as closed pericardium and again at the final blood volume immediately following incision of the anterior surface of the pericardium and excising it widely. This final state is referred to as open pericardium.

All studies were performed during steady state apneic conditions with 5 cm H₂O positive end-expiratory pressure. RV and LV pressure-volume loops were continuously measured over a 10 second interval during apnea to define baseline conditions. Stable baseline conditions were defined as sinus rhythm, during which apnea RV and LV stroke volumes were similar and the EDV and ESV of each ventricle was constant from one beat to the next. Next, the effect of complete IVC occlusion for 5 seconds on RV and LV pressure-volume relations was observed over the preceding 10-second interval. To document hemodynamic stability, repeat baseline apnea and IVC occlusion maneuvers were performed between and following the two maneuvers described below.

Partial but rapid transient aortic occlusion (AO) was performed to increase LV ESP by 30 to 40 mmHg within mechanical diastole, such that on the proceeding ejection occurred from a constant EDV but against an elevated aortic pressure. The effect of this AO maneuver on the subsequent cardiac cycle was compared to the average of the preceding five apneic beats. Partial but rapid pulmonary arterial occlusion (PAO) was also performed to increase RV ESP 10 to 15 mmHg within mechanical diastole, such that on the proceeding ejection occurred from a constant EDV but against an elevated pulmonary arterial pressure. The effect of this PAO maneuver on the subsequent cardiac cycle was compared to the average of the preceding five apneic beats. Our goal in these AO and PAO maneuvers was to rapidly and selectively increase the outflow pressure of one ventricle by approximately 50% over one beat but not to totally occlude the outflow tract or alter the filling of either ventricle. Care was taken to occlude each vessel consistently between runs. In general, two to three AO and PAO per open or closed pericardium run were necessary to identify occlusion maneuvers of a similar degree and appropriate timing during diastole. The hemodynamic response was followed after release from each partial vessel occlusion and the animals were allowed to recover for 1 minute prior to any subsequent challenges. At recovery all animals were hemodynamically stable as defined by a common heart rate, RV and LV EDV and arterial pressure as compared to baseline values.

Following completion of the above protocols in both closed and open pericardium conditions the animals were killed by an overdose of pentobarbital sodium i.v. and a partial necropsy was performed. The position of all catheters in the heart was defined, and the presence or absence of cardiac abnormalities associated with the surgical preparation noted. Blood from the animal was also collected for a final blood Rho measurement. In practice, once the protocol was started Rho remained constant between 120–145 ohms/ml over the measurement interval of approximately 60 minutes presumably because we infused only isoconductive replacement fluids made from a mixture of normal saline and 5% dextran.

Data Analysis:

The simultaneous RV and LV pressure-volume loops for sequential beats were analyzed during baseline, IVC occlusion, AO and PAO. To minimize error due to beat-to-beat variations in steady state function, the pressure and volume data of five cardiac cycles prior to IVC occlusion, AO or PAO were averaged to generate mean steady state values that were used to construct RV and LV pressure-volume histories. This was accomplished by using the R wave as time marker to align the sequential cardiac cycles. All subsequent analyses were done on these averaged pressure-volume loops unless otherwise stated. The initial slope of the diastolic filling phase prior to rapid increases in luminal pressure of these steady-state pressure-volume loops was calculated by linear regression and was used to define diastolic compliance. Maximal and minimal RV and LV volumes were used to define EDV and ESV, respectively, with their associated EDP and ESP taken as the pressures at those volumes. The IVC occlusion maneuver was used to calculate the baseline RV and LV ESPVR by the least squares linear regression method as described (Kono et al. 1984). These ESPVR data were used to assess changes in the RV and LV end-systolic pressure-volume point induced by the AO and PAO maneuvers. In general, if AO and PAO altered either the RV or LV end-systolic pressure-volume data pairs, they resulted in their migration upward and/or to the right of the baseline ESPVR with volume on the x-axis. To quantify this shift, we measured the apparent decrease in ESV from that predicted by the baseline ESPVR for the resultant ESP. This shift is referred to as δESV. For example, if the slope of the LV ESPVR, referred to as end-systolic elastance (Ees) was 1.5 mm Hg/mL and AO induced a selective increase in ESP of 15 mm Hg without a change in ESV, then the “apparent” decrease in ESV would be 10 mL. If, using this example, AO had also increased ESV by 7 mL, then the “apparent” decrease in ESV would have only been 3 mL.

Asynchrony of contraction within each ventricle was estimated by inspection of the time to minimal volume of individual segmental volume signals. If these segmental volumes reached minimal volume at different times, then then the degree of dispersion of these segmental end-systoles would define the degree of dyssynchrony as previously described by us (Strum & Pinsky 2000a, Strum & Pinsky 2000b). Baseline asynchrony and its change with AO and PAO maneuvers was also quantitated as the sum of all segmental stroke volumes not contributing to global stroke volume.

Comparisons between the initial steady state beats of each AO and PAO maneuver prior to occlusion with the first beat of sequent AO and PAO maneuvers were made using an analysis of variance for between-group differences for animals and type of occlusion and repeated measures analysis of variance for specific sequential end-diastolic or end-systolic relations. Changes in the RV or LV end-diastolic or end-systolic pressure-volume data pairs from the baseline diastolic compliance and ESPVR were taken to represent changes in diastolic compliance or ejection performance, respectively.

Results

Baseline bi-ventricular pressure-volume relations:

Although LV and RV EDV, EDP, and ESV were similar, RV EDV and ESV tended to be higher (P=0.08). Both LV E_es and E_max were significantly greater than RV values (Tables 1 and 2). LV ejection (decrease in LV volume) started and ended earlier than RV ejection. However, time to peak pressure from the QRS signal was similar between the LV and RV. Peak LV pressure occurred prior to end-systole. Peak RV pressure was similar in time to peak elastance. Thus, RV and LV ejection reached peak pressure at a similar time despite starting and ending ejection at different times. IVC occlusion decreased all RV and LV pressures and volumes by design. These data are included only to document temporal stability of the preparation since they were used to calculate baseline E_es and E_max prior to the AO and PAO maneuvers. Figures 2 through 15 represent the graphic LV and RV pressure volume relations at final volume, with IVC occlusion from baseline and then the immediate beats before and during AO and PAO for LV and RV paired studies with both intact (above) and excised pericardium (below) states. In all these figures the baseline ESPVR line is superimposed for reference. As previously reported, LV pressure-volume loops have a rectangular shape whereas RV pressure-volume loops are more triangular. An example of the segmental volume relations for pig #2 during baseline and AO is shown in Figure 16. Pericardiotomy did not alter baseline or IVC occlusion-based measures relative to those created with an intact pericardium (Table 3).

Table 1.

Effects of partial aortic, pulmonary artery, and inferior vena caval occlusion on ventricular function^♦

		Left Ventricle						Right Ventricle
Step	Condition	EDV	EDP	ESV	ESP	Ees	Emax	EDV	EDP	ESV	ESP	Ees	Emax
initial	baseline	3.29 ±0.55	3.0 ±1.2	2.43 ±0.33	42.0 ±7.3	19.9 ±4.0	22.9 ±3.9	3.38 ±1.56	3.3 ±2.0	2.84 ±1.36	15.6 ±3.2	8.7 ±6.7	9.9 ±7.6
	AO	4.03A ±1.06	11.0A ±5.7	3.73A ±0.96	90.9A ±21.5	26.6 ±9.2	36.4A ±14.1	2.85 ±1.20	3.6 ±2.1	2.72* ±1.28	15.5 ±3.7	9.8 ±6.4	10.9 ±6.6
	PAO	2.82A ±0.59	2.2 ±1.6	2.29 ±0.35	24.7A ±7.6	15.4A ±8.9	17.3 ±10.6	3.50 ±1.28	4.8 ±1.9	2.96 ±0.74	23.7A ±8.3	13.6 ±13.3	13.9 ±13.1
	IVC	2.92A ±0.64	1.2A ±1.0	2.37 ±0.34	25.5A ±5.9	13.4 ±3.0	18.6 ±6.7	2.72 ±1.48	2.3 ±2.2	2.30 ±0.79	11.1 ±3.3	7.8 ±4.7	7.9 ±4.4
mid	baseline	3.59 ±0.48	3.7 ±1.6	2.75 ±−0.31	43.3 ±9.2	17.8 ±3.8	20.1 ±3.6	3.47 ±1.54	5.1 ±1.9	3.32 ±2.26	17.1 ±7.3	11.3 ±6.4	12.4 ±6.7
	AO	3.95 ±0.54	9.9 ±8.3	3.26 ±0.47	70.6 ±12.1	27.8* ±8.5	31.5 ±10.5	3.24 ±1.84	5.2 ±2.3	3.31 ±2.33	22.4 ±9.6	12.9 ±8.3	14.5 ±9.2
	PAO	3.09 ±0.18	2.3 ±0.8	2.58* ±0.21	24.0 ±8.8	13.0 ±5.4	14.9 ±6.0	3.49 ±1.33	7.1 ±3.3	3.49* ±1.78	26.4 ±14.7	12.5 ±11.0	13.6 ±10.6
	IVC	3.13 ±1.90	1.5 ±1.1	2.29 ±0.39	38.9 ±8.8	25.2 ±12.6	28.7 ±14.7	3.13 ±1.90	2.1 ±2.7	3.03 ±1.89	13.3 ±3.6	6.4 ±4.8	8.3 ±5.0
final	baseline	3.55 ±0.79	5.8 ±3.3	2.69 ±0.65	41.5 ±9.0	20.0 ±6.3	22.1 ±5.9	3.64 ±1.38	8.6 ±2.7	3.14 ±1.28	22.7 ±12.7	11.8 ±11.1	12.5 ±11.0
	AO	4.23 ±0.98	9.9 ±8.8	3.43 ±0.45	90.3 ±20.0	29.8* ±5.9	32.1 ±10.3	3.53 ±1.42	6.8 ±3.0	3.19 ±1.34	23.5 ±7.5	11.6 7.6	13.1 ±8.6
	PAO	3.17±0.46	4.3 ±3.4	2.52 ±0.29	24.9 ±9.7	15.1* ±3.9	18.2 ±4.1	3.78 ±1.33	9.9 ±4.8	3.33 ±1.46	27.0 ±15.9	14.1 ±15.8	15.2 ±16.7
	IVC	3.29 ±0.25	1.2 ±0.7	2.48 ±0.42	34.2 ±6.6	14.5 ±4.1	17.7 ±5.2	3.17 ±1.62	5.8 ±6.0	2.78 ±1.23	12.2 ±2.6	7.5 ±5.1	8.4 ±5.5

^♦All data mean ± SD, n = 6. Step notes volume loading level: initial: following surgery; mid: after 20 mL/kg volume infusion; final: following an additional 20 mL/kg volume infusion. Baseline reflects mean individual animal data averaged from three consecutive apneic beats, partial aortic occlusion (AO), partial pulmonary artery occlusion (PAO) and inferior vena caval occlusion (IVC) reflect mean group data from the initial beats of AO and PAO, and the final beat of IVC. All pressures in mm Hg, volumes in mL, and elastances in mm Hg/mL.

^*notes a difference (P<0.05) between and baseline and either AO, PAO, or IVC values for a specific step. “A” connotes a difference (P<0.05) between baseline and either AO, PAO, or IVC across all steps. “B” connotes a difference (P<0.05) between right and left ventricles for paired data. All pressures are in mmHg, volumes in mL, and elastances in mm Hg/mL.

Table 2.

Effects of partial aortic, pulmonary arterial and inferior vena caval occlusion on ventricular ejection timing^*

		Left Ventricle					Right Ventricle
Step	Condition	start ejection	peak pressure	peak elastance	end ejection	stroke volume	start ejection	peak pressure	peak elastance	end ejection	stroke volume
initial	baseline	35B±15	93±11	118±26	195B±42	.86±.33	68±34	100±26	104±29	215±39	.83±.35
	AO	68*±37	124A±17	123±19	216A±39	.33A,B±.15	50±23	110±17	116±25	218±24	.64±.70
	PAO	61±32	113±20	119±23	192B±60	.53B±.31	56±33	106±10	110±15	226±31	.22A ±.28
	IVC	45B±28	136*±48	145*±53	175B±26	.53±.10	64±30	137*±47	139±48	217±57	.59B±.77
mid	baseline	40±25	99±14	120±25	183±30	.91±.06	50±29	123±9	128±41	195±32	.55±.33
	AO	34±26	118±13	129±13	211±34	.69±.43	59±24	118±37	122±42	212±40	.53±.38
	PAO	39±31	104±14	116±19	156±50	.63±.14	62±26	102±22	106±24	194±50	.39±.23
	IVC	28±17	100±20	135±18	178±27	.85±.43	64±36	103±28	112±28	217±49	.22±.06
final	baseline	29±23	93±10	124±28	166±30	1.0±.44	46±22	101±25	103±27	186±33	.89±.66
	AO	52*±35	114±6	118±9	208±43	.86±.67	57±11	105±27	114±36	216±19	.82±.41
	PAO	27±14	102±16	119±15	200±56	.82±.41	49±21	98±13	121±18	195±22	.35±.21
	IVC	51±30	100±20	124±16	206±60	.81±.28	73±33	113±25	126±20	202±36	.60±.66

^♦All data mean ± SD, n = 6. Step connotes volume loading level: initial: following surgery; mid: after 20 mL/kg volume infusion; final: following an additional 20 mL/kg volume infusion. Baseline reflects mean individual animal data averaged from three consecutive apneic beats, whereas partial aortic occlusion (AO), partial pulmonary artery occlusion (PAO) and inferior vena caval occlusion (IVC) reflect mean group data from the initial beats of AO and PAO, and the final beat of IVC. All data in msec from QRS wave, except stroke volume in mL.

^*notes a difference (P<0.05) between and baseline and either AO, PAO, or IVC values for a specific step. “A” connotes a difference (P<0.05) between baseline and either AO, PAO, or IVC values across all steps for a ventricle. “B” connotes a difference (P<0.05) between LV and RV values for paired data. All pressures are in mmHg, volumes in mL, and elastances in mmHg/mL.

Figure 2.

Left ventricular (LV) pressure volume relations of pig #1 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion. The IVC occlusion-defined LV end-systolic pressure-volume relation and its associated beat-to-beat end-systolic pressure-volume points per beat are superimposed on the other graphs for reference. This format is used for figures 3–13.

Figure 15.

Right ventricular (LV) pressure volume relations of pig #7 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 16.

Effect of partial pulmonary artery occlusion on the synchrony of segmental right ventricular volumes for pig #2 during closed pericardium (left) and open pericardium (right) conditions.

Table 3.

Effects of partial aortic and pulmonary artery compared to baseline on ventricular function during closed and open pericardium conditions^♦

		Left Ventricle					Right Ventricle
Pericardium	Step	EDV	EDP	ESV	ESP	Ees	EDV	EDP	ESV	ESP	Ees
Closed	Baseline	3.55 ±0.79	5.8 ±3.3	2.69 ±0.65	41.5 ±9.0	20.0 ±6.3	3.64 ±1.38	8.6 ±2.7	3.14 ±1.28	22.7 ±12.7	11.8 ±11.1
	AO	4.23 ±0.98	9.9 ±8.8	3.43 ±0.45	90.3 ±20.0	29.8* ±5.9	3.53 ±1.42	6.8 ±3.0	3.19 ±1.34	23.5 ±7.5	11.6 7.6
	PAO	3.17±0.46	4.3 ±3.4	2.52 ±0.29	24.9 ±9.7	15.1* ±3.9	3.78 ±1.33	9.9 ±4.8	3.33 ±1.46	27.0 ±15.9	14.1 ±15.8
Open	Baseline	3.25±1.17	4.5± 3.3	1.68*± 0.40	79.3± 10.4	14.8± 3.0	3.14± 0.57	6.5± 3.0	1.83± 0.27	42.8± 11.0	13.4± 3.4
	AO	3.57± 1.32	5.0± 2.9	2.31*± 0.75	103.5*±15.8	15.7± 3.1	2.84± 0.86	6.5± 3.0	1.70± 0.35	47.0± 16.6	14.4± 3.3
	PAO	2.98± 0.83	4.0± 3.2	1.85± 0.51	77.8± 14.9	12.4± 3.9	3.03± 0.66	7.3± 0.34	2.58± 0.64	51.3*± 11.4	18.3± 5.2

^♦All data mean ± SD, n = 6. Step notes pericardium status during final volume load condition: Closed: following final fluid load; Open: immediately after pericardial excision. Baseline reflects mean individual animal data averaged from three consecutive apneic beats, partial aortic occlusion (AO) and partial pulmonary artery occlusion (PAO) reflect mean group data from the initial beats of AO and PAO, and the final beat of IVC. All pressures in mm Hg, volumes in mL, and elastances in mm Hg/mL.

^*notes a difference (P<0.05) between and baseline and either AO or PAO.

Partial aortic occlusion:

AO increased LV EDV, EDP, ESV, ESP, E_es, and E_max. RV ESV tended to decrease; however, this decrease was significant only for the initial and final volume steps. This decrease in RV ESV resulted in a shift in the RV ESPVR to a point to the left of the baseline E_es described by the IVC occlusion maneuver. AO also delayed significantly both the time to start LV ejection and the time to peak pressure but did not prolong total LV ejection time. The time to RV peak pressure also increased significantly, such that the time to peak pressure for the RV was similar to that for the LV despite being increased relative to baseline. RV EDV and EDP were similar from baseline during AO. Opening the pericardium did not alter the shift in the RV ESPVR during AO.

Partial pulmonary artery occlusion:

PAO did not consistently alter RV EDV, EDP, or ESV, but significantly increased RV ESP. RV E_es and E_max tended to increase, but the response was variable among animals and not significant. In 5 of the 6 animals both E_es and E_max increased during PAO in all steps whereas in one animal they decreased greatly during PAO in the initial and mid steps. Time to start of RV ejection, peak pressure and end-ejection were not significantly different from baseline, although RV stroke volume decreased significantly. Both LV EDV and ESP decreased significantly whereas both EDP and ESV did not change from baseline. PAO did not significantly alter the timing of LV ejection (Fig. 4). Opening the pericardium abolished the observed decrease in LV EDV to PAO.

Figure 4.

Left ventricular (LV) pressure volume relations of pig #2 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Contraction synchrony:

None of the maneuvers altered the segmental LV or RV volume relations to their respective end-systole though ejection times were altered by selective ventricular loading (Table 2).

Discussion

We studied the effects of selective changes in ejection pressure of one ventricle on bi-ventricular function when the pericardium was either intact or widely excised. Selective increases in aortic pressure impedes LV ejection as manifest by a delayed LV end-ejection and LV stroke volume but simultaneously increases RV ESP without altering contraction synchrony of either ventricle. Selective increases in pulmonary arterial pressure causes LV diastolic compliance to decrease on the subsequent beat without altering contraction synchrony of either ventricle. This effect is decreases significantly following pericardiectomy. These data support our previous findings in of systolic ventricular interdependence in the open-pericardium open-chest rabbit (Solda et al. 1992) and expand them to both open and closed pericardial conditions and to the potential impact of ejection pressures of contraction synchrony.

Diastolic ventricular interdependence:

Our data demonstrate that acute increases in RV EDV induced by PAO induce marked decreases in LV EDV and subsequently, LV stroke volume. These changes we report may be even greater if the chest wall was closed as the cardiac fossa would also function as a limiting structure to biventricular volume increase. As noted above previous investigators have demonstrated that increases in either LV or RV EDV decreases diastolic compliance of the opposite ventricle (Taylor et al. 1967). However, our data agrees with others (Bove & Santamore 1981, Janicki & Weber 1980, Taylor et al. 1967) that the magnitude of this effect is greatest from RV to LV and when the pericardium is intact. Large changes in RV EDV can occur during ventilation by ventilation-induced alterations in either systemic venous return (RV preload) or pulmonary vascular resistance (RV afterload), or by constriction of both ventricles within the cardiac fossa by hyperinflation (Butler 1983, Lloyd Jr. 1982) or massive pulmonary embolism. Ventilation-associated changes in RV EDV are the primary processes creating LV pulsus paradoxus (Bove & Santamore 1981, Brinker et al. 1980). Our data also underscores the importance of an intact pericardium, because this RV to LV diastolic interaction is attenuated by pericardiotomy as previously reported (Janicki & Weber 1980).

Systolic ventricular interdependence:

We found that increases in RV developed pressure (PAO) have complex effects on LV systolic function. Although LV ESP and ESV decrease, they do so on the basal E_es relationship, such that LV E_es is similar to that produced by IVC occlusion. E_es is a load-independent measure of mechanical ejection efficiency (Suga et al. 1973). Since RV volumes are not reduced by PAO and are reduced by IVC these data suggest that the reductions in LV systolic function during PAO reflect a primary decrease in LV diastolic compliance which in subsequent beats reduces LV EDV. These data agree with previous investigators of a minimal RV to LV systolic interaction (LeWinter & Pavelec 1982, Maughan et al. 1981, Olsen et al. 1983). Previous studies from our group using both canine (Schertz & Pinsky 1993) and rabbit (Solda et al. 1992) models demonstrated that LV isovolumetric contractions result in simultaneous increases in RV stroke volume and RV developed pressure for a constant RV EDV. The present study expands these findings to document that this increased RV ESP due to partial pulmonary arterial obstruction is not due to improved RV contraction synchrony. We previously showed that LV contraction dyssynchrony can cause LV chamber dilation without changes E_es (Johnson et al. 2015). Normal RV contraction is known to be dyssynchrony with contraction progressing from base to outflow track, increased pulmonary artery pressure by delaying early contractile segments might cause RV synchrony to improve, allowing for a more efficient RV ejection in the setting of an increased afterload. However, we saw no changes in contraction synchrony with PAO.

Increases in LV developed pressure (AO) directly augment RV ejection as noted by a decrease in RV ESV and a shift of RV peak ejection in parallel to the IVC occlusion-generated RV E_es relationship without a change in the synchrony of occurrence of RV and LV peak ejection pressure. Santamore et al. (Santamore et al. 1976) demonstrated that most of the RV developed pressure is generated by parallel increases in LV developed pressure. Our studies support these observations and extend them into the regional pressure-volume domain. Our data demonstrate that the increase in LV pressure and volume not only simultaneously increase RV pressure but induce a greater RV ejection (smaller RV ESV) as defined by a shift of the RV end-systolic pressure-volume point to the left of the baseline RV E_es without altering RV contraction synchrony. The clinical implications of this hypothesis are that in the setting of marginal RV systolic function, the sudden selective unloading of the LV by ventricular assist devices may induce acute RV failure in an otherwise compensated state.

Clinical Relevance:

Although RV to LV diastolic interactions are well described and explain volume overload cor pulmonale induced cardiac decompensation, the unidirectional nature of LV to RV systolic interdependence also has significant clinical implications. Two clinical examples illustrate this point. First, biventricular interdependence explains both the mechanism of cardiovascular collapse associated with massive pulmonary embolism and effective resuscitation therapies. The immediate effect of pulmonary embolism is pulmonary vascular obstruction leading to sudden increases in pulmonary vascular resistance causing acute pulmonary artery pressure overload. RV ESV increases and on the next beat as RV EDV also increases, LV diastolic compliance decreases, such that LV EDV decreases decreasing LV stroke volume and LV ESP. The combined impaired RV ejection, increased RV EDP and decreased LV diastolic compliance causes hypovolemic LV-induced decreased LV SV causing systemic hypotension. Hypotension decreases coronary blood flow causing myocardial ischemia. Rapid fluid resuscitation alone under these conditions will only serve to further dilate the right ventricle, worsening acute cor pulmonale. The primary non-thrombolytic non-thrombectomy method to restore cardiovascular function is the use of a systemic vasopressor agent, like norepinephrine, to increase mean arterial pressure, with the goal of increasing mean arterial pressure greater than pulmonary artery pressure. By increasing systemic arterial pressure both right ventricular coronary flow increases minimizing RV ischemia, and, more importantly, LV contraction and ejection pressures are increased, directly increasing RV ejection (Vieillard-Baron et al. 2018). Our pig data suggest that this effect is primarily due to direct systolic LV to RV mechanical support rather than either decreasing RV contraction dyssynchrony (Fig. 4) or increasing RV coronary blood flow. Second, patients with severe LV failure and cardiogenic shock often receive LV assist device (LVAD) implantation as a bridge to recovery or cardiac transplantation. However, if such LVAD patient also have preexisting pulmonary hypertension, as commonly occurs in the setting of chronic obstructive pulmonary disease, then acute cor pulmonale may develop upon decompressing the diseased LV by the LVAD. Such patients need either biventricular assist device support, pulmonary vasodilator therapy or both (Viellard-Baron et al. 2018). Finally, many cardiac surgery patients who appear stable post-cardiopulmonary bypass, decompensate upon chest closure. Since chest closure restricts the cardiac fossa within the mediastinum, it functions as a pericardial limiting membrane, even though an anterior pericardial window is usually created to allow for pericardial fluid drainage. Such post thoracotomy cardiac insufficiency, if not associated with primary surgical failure (e.g. coronary artery graft obstruction) is most likely due to RV failure. Based on our pig studies, we predict that this condition would be associated with high RV filling pressure or at the least diastolic equalization of filling pressures. Many etiologies of such RV failure exist and require echocardiographic or pulmonary artery catheter-derived measures to separate, including inadequate RV cardioplegia during bypass leading to selective stunned RV myocardium and reactive pulmonary vasoconstriction. The former is treated by reopening the chest and prolonging bypass perfusion to allow RV contractility to be restored and potentially providing pharmacologic support. Whereas the latter requires focusing on pulmonary vasodilator therapy, limiting lung hyperinflation and maintaining a high mean arterial pressure.

Limitations of study:

The preparation used for our study was an anesthetized, open chest, 5–7-week old porcine model in which rapid decreases in ventricular return and selective increases in LV and RV ejection pressures were induced by manual snare occlusion maneuvers. Excepting massive pulmonary embolism, there are very few naturally occurring situations which are modeled by this preparation. Furthermore, the model used to analyze bi-ventricular function assumed that each ventricle could be assessed by monitoring the change in regional volume modeled as three cylinders stacked on themselves. Although prior studies by our group have validated this technique for the rabbit RV (Schwiep et al. 1988) and others for the LV (Baan et al. 1984), it is not clear if animals with larger sized hearts would behave similarly. Potentially, these young animals may have residual neonatal RV hypertrophy, which may cause them to behave differently. However, Stamato et al. (Stemato et al. 1995) also measured RV volumes using conductance catheter technology in the closed chest pig preparation. They described similar RV pressure-volume behavior to those we report, suggesting that if any residual RV hypertrophy was present, it did not affect diastolic compliance. However, in general conductance catheter derived absolute ESV, EDV and stroke volume may vary from real values, if the sensing electrode pairs to not fully measure volume up to the level of the aortic and pulmonic valves (Schwiep et al. 1988). Still their instantaneous changes during maneuvers remain directionally accurate. The stress of anesthesia and surgery may have altered cardiac function, but the baseline IVC occlusion data remained constant throughout the study duration suggesting that no additional deterioration occurred during the study interval. Changes in ejection pressure were induced by snare occlusion of the aorta and pulmonary artery. These methods of increasing ejection pressure do not model vasoconstriction-induced changes in ejection pressure because they do not include elements of increased input impedance (Fitzpatrick & Grant 1990). Still, selective changes in impedance at a constant ejection pressure did not appreciably alter the LV ESPVR in an isolated, perfused heart preparation (Maughan et al. 1981). Increases in aortic pressure induced by snare occlusion must increase coronary artery input pressure, especially to the unobstructed RV (Chow & Farrar 1989). RV function during the AO could have been altered by changes in coronary blood flow. We cannot exclude this possibility. However, we saw an improved ejection on the same beat as AO systole, a time before increased coronary flow could have altered contraction.

In summary, our data demonstrates that the known ventricular interdependence effects are not associated with changes in contraction synchrony and are primarily unidirectional being RV to LV for diastolic effects and augmented by an intact pericardium and LV to RV for systolic effects and not influences by the pericardium.

Figure 3.

Right ventricular (LV) pressure volume relations of pig #1 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 5.

Right ventricular (LV) pressure volume relations of pig #2 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 6.

Left ventricular (LV) pressure volume relations of pig #3 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 7.

Right ventricular (LV) pressure volume relations of pig #3 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 8.

Left ventricular (LV) pressure volume relations of pig #4 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 9.

Right ventricular (LV) pressure volume relations of pig #4 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 10.

Left ventricular (LV) pressure volume relations of pig #5 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 11.

Right ventricular (LV) pressure volume relations of pig #5 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 12.

Left ventricular (LV) pressure volume relations of pig #6 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 13.

Right ventricular (LV) pressure volume relations of pig #6 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

Figure 14.

Left ventricular (LV) pressure volume relations of pig #7 during closed pericardium (top) and open pericardium (bottom) conditions during four sequential states: apneic baseline, it’s subsequent pressure-volume loops during inferior vena caval (IVC) occlusion, and then for the first few beats before and during partial aortic occlusion and partial pulmonary artery occlusion.

New Findings.

Instantaneous selective increases in aortic pressure causes an immediate increase in right ventricular end-systolic pressure independent of the pericardium. Whereas selective increases in pulmonary artery pressure decreases left ventricular diastolic compliance due to a subsequent increasing right ventricular end-diastolic volume as a function of an intact pericardium limiting bi-ventricular volume. Finally, changes in contraction synchrony of either ventricle do not appear to be causing these effects.