Acute right ventricular pressure overload compromises left ventricular function by altering septal strain and rotation

Abstract

While right ventricular (RV) dysfunction has long been known to affect the performance of left ventricle (LV), the mechanisms remain poorly defined. Recently, speckle-tracking echocardiography has demonstrated that preservation of strain and rotational dynamics is crucial to both LV systolic and diastolic function. We hypothesized that alteration in septal strain and rotational dynamics of the LV occurs during acute RV pressure overload (RVPO) and leads to decreased cardiac performance. Seven anesthetized pigs underwent median sternotomy and placement of intraventricular pressure-volume conductance catheters. Two-dimensional echocardiographic images and LV pressure-volume loops were acquired for offline analysis at baseline and after banding of the pulmonary artery to achieve RVPO (>50 mmHg) induced RV dysfunction. RVPO resulted in a significant decrease (P < 0.05) in LV end-systolic elastance (50%), systolic change in pressure over change in time (19%), end-diastolic volume (22%), and cardiac output (37%) that correlated with decrease in LV global circumferential strain (58%), LV apical rotation (28%), peak untwisting (reverse rotation) rate (27%), and prolonged time to peak rotation (17%), while basal rotation was not significantly altered. RVPO reduced septal radial and circumferential strain, while no other segment of the LV midpapillary wall was affected. RVPO decreased septal radial strain on LV side by 27% and induced a negative radial strain from 28 ± 5 to −16 ± 2% on the RV side of the septum. The septal circumferential strain on both LV and RV side decreased by 46 and 50%, respectively, following RVPO (P < 0.05). Our results suggest that acute RVPO impairs LV performance by primarily altering septal strain and apical rotation.

acute right ventricular (RV) pressure overload (RVPO) due to pulmonary hypertension oftentimes progresses to acute RV failure (10, 18). As the RV fails, left ventricular (LV) performance is also compromised, although it is agreed that the mechanism of LV impairment is more complex than a simple drop in LV preload (3, 4, 26). Indeed, when pressure in the RV is significantly elevated, as in clinical conditions such as acute pulmonary embolism, pulmonary hypertension (acute or chronic), and obstructive sleep apnea, the interventricular septum is noted to shift toward the left, leading to decreased LV volumes, as well as decreased systemic and myocardial perfusion (15, 37). Although this ventricular interdependence is commonly observed and well chronicled, the mechanisms for LV mechanical dysfunction due to isolated RV failure are not well defined.

It is recognized that radial strain, circumferential strain, and rotational motion serve as dynamic indexes of LV mechanical function and precisely describe both global and regional ventricular function (6, 7, 14, 20). In fact, alterations in LV rotational motions are sensitive markers for ventricular dysfunction (7). Radial strain quantifies thickening of the myocardium toward the center of the ventricle, while circumferential strain measures shortening along the circumference (16, 27). Currently, no studies exist describing the effect of acute RV failure on regional function of the interventricular septum utilizing radial or circumferential strain, or on global function of the LV as reflected by global circumferential strain or rotation.

Furthermore, recent advances in diffusion tensor imaging have further elucidated the nature of the muscular layers of the LV, with the epicardial and endocardial fibers arranged in helical patterns wrapped in opposite directions (8, 22, 33, 41). From a functional perspective, previous studies have shown that these layers behave differently: the myocardial layer of the septum facing the LV (endocardial fibers) has greater rotation and shorter time to peak rotation compared with the layer facing the RV (epicardial fibers) (23). This may account for some of the changes seen in certain pathological states. However, no experimental data are available to describe the changes in endocardial and epicardial strain in a RV failure model.

Recent advancements in two-dimensional (2D) speckle tracking echocardiography (STE) imaging allow measurement of dynamic LV function. In a porcine model, we used STE to investigate the relationship between indexes of dynamic LV function and intraventricular pressure-volume (PV) relationships (contractility, preload, afterload, relaxation, and filling). Furthermore, we investigated the alteration in LV epicardial and endocardial segmental radial and circumferential strain, and LV apical rotation during RV dysfunction from acute RVPO (39). We hypothesized that 1) RV dysfunction induced by acute RVPO is associated with isolated derangement in septal strain, compared with the other LV myocardial segments where strain is unaltered; 2) septal mechanics of the RV side of the septum (epicardial fibers) and the LV side (endocardial fibers) are differentially altered during acute RVPO; and 3) decreased septal performance is associated with decrease in global LV performance with decreased overall LV circumferential strain, apical rotation, and rotation rate.

METHODS

All animals received humane treatment in compliance with the 1996 National Research Council Guide for the Care and Use of Laboratory Animals, and institutional approval was obtained for each animal used in this study. Seven pigs (40–46 kg) were anesthetized using intramuscular ketamine (25 mg/kg) and xylazine (2 mg/kg), and anesthesia was maintained using inhaled isoflurane (1–2%). Fentanyl was used for analgesia, and pancuronium for muscular relaxation. Respiratory support was initiated via volume-controlled ventilator after endotracheal intubation. Arterial oxygen saturation and electrocardiogram were continuously monitored. Catheterization of the carotid artery provided invasive measurement of arterial blood pressure, and the jugular vein was cannulated for intravenous infusion of fluids and drugs. The heart was exposed via midline incision and sternotomy. A 5F multielectrode PV conductance catheter (Millar Instruments, Houston, TX) was inserted into the LV in a retrograde fashion from access via the right carotid artery; a 3F multielectrode PV conductance catheter (Millar Instruments) was inserted into the RV in an anterograde fashion from access via the internal jugular vein. Both catheters were used to obtain hemodynamic data. Proper electrode position was confirmed both by epicardial echocardiography and examination of segmental volume signals. LV and RV PV data were acquired at baseline, at an intermediate time point (15–20 min), and after 90 min of pulmonary arterial banding to achieve RV pressure of two-thirds to three-fourths of systemic values. The animals were euthanized at the end of each study via intravenous injection of pentobarbital (100 mg/kg).

Experimental protocol.

After recording baseline data, the pulmonary artery (PA) was isolated via blunt dissection and banded. The band was tightened while RV pressures were monitored to achieve RV systolic pressures two-thirds to three-fourths of systemic values (17). Following a 15- to 20-min period and a 90-min period of stabilization after banding of the PA, the same hemodynamic measurements were obtained. After 90 min, RV dysfunction was confirmed by echocardiographic presence of tricuspid regurgitation, decreased RV free wall motion, and decreased tricuspid annular systolic excursion (17, 28, 29, 35, 38). A final equilibration period of 90 min was chosen to account for the initial decrease in LV preload (due to decreased RV ejection), and to allow attenuation in the immediate increase in sympathetic tone after PA banding (9). The goal was thus to grant more accurate measurement of changes due strictly to the new physiological state.

LV PV measurements.

LV function was evaluated by inserting a dual-field micromanometer-tipped pressure-conductance catheter (Millar Instruments) into the ventricular lumen and generating PV loops. PV data were analyzed with PVAN Ultra (Millar Systems, Houston, TX). Correct position of the catheter was confirmed by epicardial echocardiogram.

PV data were obtained during full neuromuscular blockade; all measurements were recorded during end expiration. Indexes of systolic function, myocardial contractility, and diastolic function were assessed. PV data were analyzed for LV end-systolic and end-diastolic pressures and volumes, stroke volume, ejection fraction (EF), systolic change in pressure over change in time, LV elastance, diastolic time constant of relaxation, and diastolic change in pressure over time. PV relationships were derived during serial preload reduction by inferior vena cava occlusion, as previously reported (7). The load-independent contractility can be characterized by a slope [end-systolic elastance (Ees)] and a volume axis intercept (V₀), so that Pes = Ees (Ves − V₀), where Pes and Ves are end-systolic pressure and volume, respectively (7).

Echocardiography.

Epicardial echocardiographic images were acquired at baseline, after 15–20 min, and after 90 min of PA banding. High-resolution 2D echocardiographic images of short-axis views (basal, midpapillary, and apical) were acquired under optimized conditions during end expiration using a 10-MHz sector probe at 80–120 frames/s on a GE vivid 7 platform (GE Medical Systems, Milwaukee, WI) (23). The images were analyzed offline for apical and basal rotation, global circumferential strain, segmental radial strain, and differential radial strain of the right and left septum using an EchoPac (PC version 8.0, GE Vingmed Ultrasound) workstation. Strain curves were interpolated using proprietary software algorithms provided in the GE EchoPAC software. LV rotation measurements were obtained at the basal (identified by the mitral valve) and apical (no papillary muscles noted) levels (21, 32). Short-axis circumferential and radial strain values were evaluated using the same speckle tracing software at the midventricular level, which was identified as the level according to the maximum papillary muscle circumference (1, 2, 25). For analysis of regional differences in strain between the LV side and RV side of the septum, inner and outer regions of interest were placed on the septum for analysis of the two distinct layers separated by an identifiable midseptal line (23). The investigators analyzing echocardiography images were blinded to the hemodynamic data. Tracking adequacy was confirmed by built-in software indicators of tracking quality.

Statistical analysis.

Analyses were performed using Sigma Stat (version 3.1). Data are presented as means ± SD. The hemodynamic and echocardiographic measurements were compared in the experimental stages by use of repeated-measures ANOVA, followed by the post hoc Bonferroni correction. Statistical differences were considered significant at P < 0.05. Intraobserver and interobserver reproducibility were assessed for apical and basal rotation and radial and circumferential strain in five randomly selected studies. Interobserver variability was calculated as the SD of the differences between the measurements of two independent observers. Intraobserver variability was calculated as the differences between the first and second determinations (1-mo interval) for a single observer and expressed as a percentage of the average value.

RESULTS

Figure 1 shows representative PV loops, LV septal r-strain, and apical rotation traces at baseline and following 90 min of RVPO. Acute RVPO increased LV contractility over the intermediate phase, but then decreased LV contractility from baseline values (Fig. 1A) and shifted V₀ to the left. The decrease in Ees after RVPO was mirrored by changes in LV septal strain (Fig. 1B) and apical rotation (Fig. 1C).

Fig. 1.

A: pressure-volume loops. B: left ventricular septal R-strain. C: left ventricular apical rotation traces. Representative tracings illustrating that, after 90 min of pulmonary artery banding, pressure-volume tracings reveal a decrease in end-systolic elastance (Ees; A), and speckle tracking echocardiography reveals decrease in apical rotation with a prolongation in time to peak rotation (B). RVPO, right ventricular pressure overload.

Hemodynamic responses to RVPO are summarized in Table 1. Despite an initial increase in LV systolic function, after the full 90-min equilibration period, acute RV failure resulted in significantly impaired LV systolic and diastolic function. Although heart rate was increased from 73 to 84 beats/min following RVPO, cardiac output was significantly decreased after an initial increase. RVPO markedly increased RV end-systolic pressure and RV end-diastolic pressure.

Table 1.

Hemodynamic responses to RVPO × 90 min

	Baseline	RVPO
Left ventricle
Heart rate, beats/min	73 ± 2	84 ± 2*
End-systolic pressure, mmHg	78 ± 1	63 ± 2*
End-diastolic pressure, mmHg	6.1 ± 0.3	4.1 ± 0.3*
Mean arterial pressure, mmHg	58 ± 2	42 ± 1*
End-systolic volume, ml	35 ± 1	34 ± 2
End-diastolic volume, ml	68 ± 3	53 ± 4*
Stroke volume, ml	34 ± 2	21 ± 2*
Cardiac output, ml/min	2751 ± 151	1733 ± 108*
Systemic vascular resistance, mmHg•l−1•min	1485 ± 67	1788 ± 76*
Ejection fraction, %	51 ± 4	43 ± 2
Ees, mmHg/ml	5.7 ± 0.7	2.6 ± 0.4*
V0, ml	14.2 ± 0.4	5.2 ± 0.6*
dP/dtmax, mmHg/s	1698 ± 95	1300 ± 83*
dP/dtmin, mmHg/s	1437 ± 51	1076 ± 40*
τ, ms	47 ± 2	57 ± 2*
Right ventricle
End-systolic pressure, mmHg	18 ± 1	51 ± 2*
End-diastolic pressure, mmHg	3.8 ± 0.3	19 ± 1*
dP/dtmax, mmHg/s	516 ± 28	614 ± 25*
dP/dtmin, mmHg/s	250 ± 16	356 ± 22*

Values are means ± SE; n = 7. RVPO, right ventricle pressure overload; Ees, LV end-systolic elastance; V₀, volume axis intercept. dP/dt_max, systolic change in pressure over change in time; dP/dt_min, diastolic change in pressure over change in time; τ, diastolic time constant of relaxation.

^*P < 0.05 vs. baseline.

Initially RVPO significantly increased LV apical rotation, but after the 90-min equilibration period, LV apical rotation was significantly impaired (Table 2). Apical peak untwisting/reverse rotation rates were also impaired, although basal rotation remained unaffected. Figure 2 shows that LV Ees closely correlates with apical rotation (R = 0.76, P < 0.05).

Table 2.

LV rotational dynamic responses to RVPO × 90 min

	Baseline	Intermediate	RVPO
Apical rotation, °	13.6 ± 1.2	16.1 ± 1.1*	8.8 ± 1.0*
Time to peak rotation, ms	528 ± 15	496 ± 20	438 ± 19*
Apical peak untwisting rate, °/s	−105 ± 13	−110 ± 14	−65 ± 13*
Basal rotation, °	−3.8 ± 1.1	−4.7 ± 1.3	−4.0 ± 1.2
Basal peak untwisting rate, °/s	31 ± 3	32 ± 4	29 ± 3
Left ventricular twist, °	17.9 ± 0.3	20.8 ± 0.3	12.8 ± 0.4*

Values are means ± SE; n = 7.

^*P < 0.05 vs. baseline.

Fig. 2.

Correlation between apical rotation and left ventricle Ees.

Table 3 and Fig. 3 summarize the LV regional radial and circumferential strain responses to RV dysfunction induced by acute RVPO. Among the six midpapillary segments of the LV, we only observed a significant decrease in septal radial and circumferential strain after 90 min, while the other segments were unchanged from baseline. However, at the intermediate time point, septal strain was unchanged from baseline, while strain of the other segments was slightly increased from baseline, supporting the initial increase in LV function.

Table 3.

Regional strain responses to RVPO × 90 min

	Radial Strain, %			Circumferential Strain, %
	Baseline	Intermediate	RVPO	Baseline	Intermediate	RVPO
Anteroseptal	63 ± 2	75 ± 3	67 ± 2	−16 ± 2	−22 ± 1	−14 ± 1
Anterior	68 ± 2	79 ± 2	69 ± 2	−12 ± 2	−16 ± 1	−14 ± 1
Lateral	45 ± 1	53 ± 2	46 ± 1	−18 ± 2	−19 ± 1	−16 ± 2
Posterior	30 ± 1	44 ± 1	24 ± 1	−16 ± 2	−13 ± 1	−17 ± 2
Inferior	39 ± 1	45 ± 2	33 ± 1	−18 ± 2	−22 ± 2	−17 ± 2
Septal	55 ± 2	55 ± 2	41 ± 2*	−18 ± 2	−12 ± 2	−7 ± 2*

Values are means ± SE; n = 7.

^*P < 0.05 vs. baseline.

Fig. 3.

Screen shot of GE Echopac (GE Healthcare, Waukesha, WI) while analyzing epicardial and endocardial radial strain following 90 min of RVPO. The red tracing represents septal radial strain, which becomes negative in the epicardium.

Comparison of septal radial and circumferential strain between LV and RV sides of the septum are shown in Table 4. When the STE region of interest was specifically applied to the endocardium and epicardium, RV dysfunction after 90 min of RVPO similarly reduced LV-sided septal (endocardial fibers) circumferential strain, but reversed the RV-sided septal (epicardial fibers) radial strain, explaining the paradoxical motion observed in septal wall (Table 4, Fig. 4). No other statistically significant differences were noted in the endocardium or epicardium of any other LV wall segments after 90 min. At intermediate time point, a decrease in RV-sided septal radial strain to nearly zero and significant impairment in RV-sided septal circumferential strain are seen. The septal layer facing the LV also displayed, increasing impairment with time, but no reversal in radial strain or circumferential strain.

Table 4.

LV side and RV side septal strain responses to RVPO × 90 min

	Baseline	Intermediate	RVPO	%Change
LV side radial strain, %	53 ± 2	49 ± 2	35 ± 3*	27
Time to peak, ms	464 ± 20	475 ± 16	503 ± 2*	13
RV side radial strain, %	29 ± 2	9 ± 1	−16 ± 1*	reversed
Time to peak, ms	492 ± 17	459 ± 15	497 ± 17	1.4
LV side circumferential strain, %	−26 ± 1	−19 ± 2	−12 ± 1*	46
Time to peak, ms	447 ± 16	470 ± 18	526 ± 17*	23
RV side circumferential strain, %	−21 ± 2	−11 ± 2	−14 ± 0.6*	50
Time to peak, ms	511 ± 16	514 ± 13	490 ± 17	3.9

Values are means ± SE; n = 7. LV, left ventricle; RV, right ventricle.

^*P < 0.05 vs. baseline.

Fig. 4.

Regional changes in radial and circumferential left ventricular strain in the endocardial and epicardial layers following 90 min of RVPO. S, septal; AS, anteroseptal; A, anterior; L, lateral; P, posterior; I, inferior. Shaded areas marked with asterisks reflect P < 0.05 vs. baseline.

There was very good reproducibility and concordance of the echocardiographic measurements. The mean differences of all echocardiographic measurements were 4.6 ± 2.8% (interobserver) and 4.2 ± 2.9% (intraobserver).

DISCUSSION

Our experimental study shows that acute RV dysfunction from RVPO led to a decrease in LV systolic and diastolic function, as confirmed by both hemodynamic assessment and STE. This decrease in LV function in the setting of acute RV failure is described as “ventricular interdependence” and is seen in many clinical settings, including acute pulmonary embolism, following cardiac surgery with cardiopulmonary bypass severe hypoxemia, and during thoracic surgical repairs that require PA clamping, such as lung transplants (4, 5, 13). While the changes in RV mechanics after RVPO have been described, including decreases in RV free wall velocity, circumferential, and longitudinal strain (17, 21, 30, 31, 37), our study was specifically aimed at further elucidating the factors contributing to the ventricular interdependence and alteration of LV mechanics and performance. We have shown that, as RVPO progresses to failure, the LV initially compensates for the new physiological state by augmenting overall contractility before developing impaired radial and circumferential strain of the septal wall, while strain in the remaining LV segments remains preserved. This, in turn, is associated with derangements in LV function reflected by impairment of global circumferential strain, apical rotation, and overall cardiac performance, as measured by hemodynamic parameters. When further assessing function of the septum, we have found that septal dysfunction is apparent even in the early phase of RVPO, when rests of the LV segments demonstrate enhanced function. With established RV dysfunction, the myocardium facing the LV (endocardial fibers) thickens less than at baseline, and the myocardium facing the RV (epicardial fibers) fails to thicken entirely (radial strain values become negative), creating a paradoxical septal motion.

Structure of the interventricular septum and its role in ventricular interdependence.

Advances in diffusion tensor imaging have enabled greater understanding of the arrangement of myocardial fibers in the interventricular septum and, as such, may indeed improve our understanding of its function (8, 22, 33, 41). In short, the fibers of the LV are arranged as two helices wrapped in opposite directions, leading not only to thickening of the ventricle, but also to the torsional, or “twisting”, motion of the heart during systole. This fiber arrangement holds true throughout all LV walls, including the interventricular septum (22).

Furthermore, in the setting of septal displacement due to chronic right heart failure, cardiac MRI studies have shown these helical fibers to align in a more parallel fashion along the LV circumference (5, 34). Despite this structural understanding, no studies have assessed the functional effects of myocardial fiber derangement in the setting of acute RVPO and RV dysfunction, although Hayabuchi et al. (19) have studied strain changes in children with chronic RV overload. Our baseline data agree with those in this previous study showing that left-sided strain values tend to be higher than those on the right. However, as their study involved children with chronic RV overload (without RV failure), their findings were more in line with Cho et al. (9), who found that, in RVPO without failure, the heart increases work via twist and strain. In the Cho et al. study, measurements were taken ∼10 min after milder increases in RV afterload, which did not result in RV failure. Our findings after 15–20 min of RVPO agree with those of this study, in that, over a shorter time period, we noted an augmentation in LV function. We speculate that the transient augmentation in LV function may be due to catecholamine release as a part of compensatory mechanism. In our study, the final 90-min equilibration time allowed for dissipation of the immediate sympathetic response to RVPO and created a more complete model of RVPO induced RV failure, which was not present in the Cho et al. model.

Our study performed in an acute RV failure model shows that, when measuring radial strain while placing the region of interest specifically on the RV side of the septum, the value approaches zero early after onset of RVPO, then becomes negative, the RV side of the septum fails to thicken, and instead becomes thinner during systole in the setting of RV failure due to acute RVPO. This is in distinction to the LV side, which still thickens, although less than at baseline. Such impairment is not noted in either muscle layer of any other wall segment and has not been evaluated in any previous study. In fact, our findings suggest that this return to baseline in the remaining segments occurs only after an initial increase in performance in the nonseptal segments before progression to RV failure.

The interplay between the RV and LV is complex and remains poorly understood. Theories exist regarding impairment of LV filling during RVPO due to leftward shift of the interventricular septum. This decrease in LV preload is thought to result in a decrease in LV EF due to the Frank-Starling mechanism in a phenomenon referred to as “series interaction.” However, we show that LV Ees, a preload-independent indicator of systolic function, is impaired during acute RVPO, thus obviating the LV filling hypothesis as the sole mechanism for ventricular interdependence. Other hypotheses include direct ventricular interaction mediated by the pericardium and interventricular asynchrony. In an intact pericardium, both ventricles are contained within a nondistensible space, and thus an increase in RV size will lead to a decrease in LV size via a mechanism referred to as “direct” or “pericardial interaction.” However, as all of our experimental data were collected in pigs after pericardiotomy, our observed impairment in LV function cannot be solely attributed to this mechanism either.

LV functional response to acute RVPO and proposed mechanism for LV dysfunction.

We found that, after an initial increase, the LV experiences a decrease in systolic change in pressure over change in time and Ees, both indicators of systolic performance. LV diastolic function suffers as well, as evidenced by changes in diastolic time constant of relaxation and diastolic change in pressure over time. STE analysis of the LV reveals that, while at first the septal wall experiences no change in radial or circumferential strain, overall LV performance is augmented via an increase in strain of the remaining walls. After an adequate equilibration period for the new physiological state, RVPO causes the septal wall to suffer a significant decrease in radial and circumferential strain, while strain in the other walls returns to baseline (no significant change). In turn, the global indexes of circumferential strain, apical rotation, and rotation rate are initially improved, but then become significantly impaired, reflecting the decrease in LV systolic performance, even while EF, a traditional marker of LV systolic function, fails to detect a significant change. Markers of diastolic function via STE confirm the hemodynamic data, with impairment in average as well as peak untwisting/reverse rotation rates at the final time point. No previous data exist measuring the relationship between radial and circumferential strain or LV apical rotation in RV failure induced by acute RVPO in a manner independent of both series and direct interaction.

As it is established that RVPO leads to displacement of the interventricular septum toward the left (11, 13), and that this displacement is associated with a distortion of the normal, helical arrangement of myocardial fibers in the septal wall (33), we propose that this abnormal, circumferential orientation of septal fibers may be responsible for the regional decrease in septal thickening. This may be similar to the aberrant fiber deformation that has been recognized in myocardial infarction and cardiomyopathy (8, 10). This would seem to be consistent with our findings that, while LV function augmented over a shorter time period, the septum is unable to mount a similar increase in strain to that of the remaining walls. And while velocity vector imaging studies have shown decreased septal tissue velocities in the setting of chronic pulmonary hypertension (31), our study reveals that, at least in the acute setting, this dysfunction is more pronounced on the right side of the septum, as evidenced by initial loss of radial strain leading to complete reversal and significantly decreased circumferential strain. This progressive deterioration of septal contraction may, in turn, account for the LV dysfunction noted after equilibration via impaired ability of the LV wall to shorten along its circumference and to perform its usual torsional motion. As there is a significant correlation between Ees and apical rotation, this change in LV torsional mechanics may serve to explain the impairment of LV performance in the setting of RV failure and serve as the predominant factor in ventricular interdependence.

Physiological implications.

The present study serves to describe the physiological degradation in regional and global strain that occurs once acute RVPO progresses to RV failure. Our findings may help to improve understanding of the underlying physiological changes associated with the progression of RV disease in settings such as marked, sudden changes in RV afterload (acute pulmonary embolism, cardiac surgery with cardiopulmonary bypass, surgical clamping of PAs in thoracic surgery), or even acute worsening of chronic RVPO, and hence may aid in early diagnosis, risk stratification, and treatment strategies. Indeed, as our results suggest that the initiating event in LV dysfunction due to ventricular interdependence is most likely distortion of septal architecture, interventions aimed at restoring more physiological septal geometry and function in this setting may, in turn, lead to improved LV global mechanics (e.g., strain and torsion) and thus restore overall cardiac performance (11, 26, 38). Furthermore, as apical rotation is a key factor in LV performance, our study may provide physiological evidence as to the mechanism by which restoration of sequential activation via electrical stimulation of the ventricle in a torsional fashion, similar to what has been found in septal pacing and cardiac resynchronization therapy, is beneficial (29, 36, 38). Indeed, our findings may suggest that further studies aimed at reversing these derangements in this setting would be useful.

Limitations.

As with any assessment of global function based on a single 2D image, both LV apical rotation and circumferential strain must be interpreted with caution in the setting of regional functional abnormalities. The present study does not compare STE with other imaging modalities, such as tagged MRI, three-dimensional echo, or radio nucleotide angiography (40). However, no gold standard is established for such imaging, and each of these methods has its own set of limitations (16, 32, 41). This study was performed on open-chest and open-pericardium pigs, and therefore these findings may have limited extrapolation to intact humans. A further limitation of this model is the difficulty in obtaining consistently adequate longitudinal views, thus precluding us from including longitudinal strain in our analysis. Although some studies (3, 12, 24) suggest that the pericardium enhances ventricular interaction, interdependence does exist in the setting of an open pericardium. Further studies may be aimed at reproducing this model over a chronic timeframe and could seek to define underlying mechanisms by assessment of microscopic structural alterations in different regions of the myocardium, as well as elucidating signaling pathways triggered by acute or chronic RVPO.

Conclusion.

Ventricular interdependence causes a decrease in LV function when the RV experiences pressure overload. In this study, we demonstrate that RV failure induced by acute RVPO diminishes septal strain, with the right side of the septum being affected more than the left side. Isolated septal dysfunction is associated with a decrease in LV global strain and rotational dynamics, with apical rotation being the most affected rotational parameter. Furthermore, 2D speckle tracking parameters of LV function correlate well with LV PV relationships during RVPO and can be used to assess cardiac performance.

GRANTS

This study is supported by National Heart, Lung, and Blood Institute Grant RO1-HL084261.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: J.H.C., W.Z., J.K.H., G.B.M., and A.M. conception and design of research; J.H.C., W.Z., J.K.H., N.A.P., and A.M. performed experiments; J.H.C., W.Z., J.K.H., N.A.P., G.B.M., and A.M. analyzed data; J.H.C., W.Z., J.K.H., N.A.P., G.B.M., and A.M. interpreted results of experiments; J.H.C., W.Z., and A.M. prepared figures; J.H.C., W.Z., N.A.P., G.B.M., and A.M. drafted manuscript; J.H.C., W.Z., J.K.H., N.A.P., G.B.M., and A.M. edited and revised manuscript; J.H.C., W.Z., J.K.H., N.A.P., G.B.M., and A.M. approved final version of manuscript.