Speckle-tracking echocardiography elucidates the effect of pacing site on left ventricular synchronization in the normal and infarcted rat myocardium

Abstract

Background: Right ventricular (RV) pacing generates regional disparities in electrical activation and mechanical function (ventricular dyssynchrony). In contrast, left ventricular (LV) or biventricular (BIV) pacing can improve cardiac efficiency in the setting of ventricular dyssynchrony, constituting the rationale for cardiac resynchronization therapy (CRT). Animal models of ventricular dyssynchrony and CRT currently relay on large mammals which are expensive and not readily available to most researchers. We developed a methodology for double-site epicardial pacing in conscious rats. Here, following post-operative recovery, we compared the effects of various pacing modes on LV dyssynchrony in normal rats and in rats with ischemic cardiomyopathy.

Methods: Two bipolar electrodes were implanted in rats as follows: Group A (n = 6) right atrial (RA) and RV sites; Group B (n = 7) RV and LV sites; Group C (n = 8) as in group B in combination with left coronary artery ligation. Electrodes were exteriorized through the back. Following post-operative recovery, two-dimensional transthoracic echocardiography was performed during pacing through the different electrodes. Segmental systolic circumferential strain (Ecc) was used to evaluate LV dyssynchrony.

Results: In normal rats, RV pacing induced marked LV dyssynchrony compared to RA pacing or sinus rhythm, as measured by the standard deviation (SD) of segmental time to peak Ecc, SD of peak Ecc, and the average delay between opposing ventricular segments. LV pacing and, to a greater extend BIV pacing diminished the LV dyssynchrony compared to RV pacing. In rats with extensive MI, the effects of LV and BIV pacing were markedly attenuated, and the response of individual animals was variable.

Conclusions: Rodent cardiac pacing mimics important features seen in humans. This model may be developed as a simple new tool to study the pathophysiology of ventricular dyssynchrony and CRT.

Figure 1. Experimental setup for LV strain analysis in paced rats.

A: An implanted pacing device consisting of two miniature-bipolar hook electrodes (dotted circles). Long arrow; 8-pin back connector. Short arrows; peripheral electrodes for optional pseudo ECG measurements in conscious animals . B: A rat with an exteriorized back connector following the implantation procedure. C: X-ray lateral view of an implanted rat. The two miniature-bipolar hook electrodes (vertical arrows at the bottom) were inserted on the RV and the LV as described in the text. The X-ray view is shown for illustration and was not utilized to access the position of the electrodes on the heart. D: Short axis mid-level echocardiographic view of an implanted rat and standard separation map of the LV into six segments for circumferential strain (Ecc) analysis.

Figure 2. Segmental Ecc analysis during RA and RV pacing.

A: Representative strain waveforms in an instrumented rat subjected to no pacing (upper), RA pacing (middle) and RV (lower). The colors of the different waveforms match those in the segmental map in Figure 1D. White dotted lines represent the global (average) strain of all six segments. B: Average segmental time to peak Ecc (left) and peak Ecc (right) in seven animals subjected no pacing, RA pacing, and RV pacing.

Figure 3. RV pacing induces marked LV dyssynchrony compared to RA pacing.

Summary of echocardiographic parameters in group A subjected to no pacing, RA pacing, and RV pacing. A: Average time to peak Ecc of all six segments (global). B: Average peak Ecc of all six segments (global). C: LV fractional area shortening. D: Standard deviation of the time to peak Ecc of different segments. E: Standard deviation of the peak Ecc of different segments. F: Analysis of ‘delta TP opposing segments’, i.e., the average time difference between the peak Ecc of opposing segments. * P<0.05, ** P<0.01. Horizontal lines above\below the bar graphs indicate significant differences between pacing modes in the post-hoc analysis. Note that in D the Kruskal-Wallis test was significant. However, the post-hoc analysis could not reveal the source of difference between the groups.

Figure 4. RV pacing induces electrical dyssynchrony compared to LV and BIV pacing.

A: Representative ECG recordings from an instrumented rat during sinus rhythm (no pacing) and during RV, LV and BIV pacing. Arrows indicate the termination of the T wave in each trace. B: Summary of QT interval analysis in six instrumented rats. ** P<0.01. Horizontal lines above the bar graph indicate significant differences between pacing modes in the post-hoc analysis. C: Example of epicardial electrograms during RV pacing (left) and LV pacing (right). In each recording upper trace is a peripheral ECG signal (Lead I), middle trace is the epicardial electrogram and lower trace shows the time of cardiac pacing. Note that the epicardial activation in the RV electrode correlates with a late negative vector in the ECG signal. In contrast, the epicardial activation during RV pacing is correlated with a late positive vector in the ECG signal. Measurements of the time from end of stimulus artifact to the peaks of these late vectors in the ECG (dash vertical lines) indicate shorter activation time during LV pacing. Analysis of epicardial electrograms in an additional animal from group B demonstrated similar findings (not shown).

Figure 5. Segmental Ecc analysis during RV, LV and BIV pacing.

Average segmental time to peak Ecc (A) and peak Ecc (B) in seven animals from group B subjected to no pacing, RV pacing, LV pacing and BIV pacing.

Figure 6. RV pacing induces mechanical dyssynchrony compared to LV and BIV pacing.

Summary of echocardiographic parameters in six animals subjected to no pacing, RV, LV and BIV pacing. A: Average time to peak Ecc of all six segments (global). B: Average peak Ecc of all six segments (global) C: LV fractional area shortening D: Standard deviation of the time to peak Ecc of different segments. E: Standard deviation of the peak Ecc of different segments. F: Average time difference between the peak Ecc of opposing segments (‘delta TP opposing segments’, see Figure 3 for details). * P<0.05, ** P<0.01. Horizontal lines above\below the bar graphs indicate significant differences between pacing modes in the post-hoc analysis.

Figure 7. Combined model of pacing and ischemic heart failure.

A: Anterior image of a heart subjected to electrodes implantation in combination with left coronary ligation. RV electrode and LV electrode are marked by short and long black arrows, respectively. White arrow; coronary artery ligation site. B: Transverse sections of the heart shown in A. Apical to basal sections are ordered from left to right. C: Summary of QT interval analysis in five instrumented rats with MI. * P<0.01. Horizontal line above the bar graph indicate significant difference in the post-hoc analysis. D: Representative strain waveforms of two instrumented rats with MI under control conditions. E: Average segmental time to peak Ecc (left) and peak Ecc (right) in five rats with ischemic heart failure subjected to no pacing, RV, LV and BIV pacing.

Figure 8. Attenuated effects of pacing in rats subjected to extensive MI.

Summary of echocardiographic parameters in five rats with ischemic heart failure subjected to no pacing, RV pacing, LV pacing, and BIV pacing. A: Average time to peak Ecc of all six segments (global). B: Average peak Ecc of all six segments (global) C: LV fractional area shortening D: Standard deviation of the time to peak Ecc of different segments. E: Standard deviation of the peak Ecc of different segments. F: Average time difference between the peak Ecc of opposing segments (‘delta TP opposing segments’, see Figure 3 for details). Kruskal-Wallis test did not reveal a significant difference between the different pacing modes for any of the presented parameters.

Figure 9. Individual variability in the effect of pacing in rats subjected to extensive MI.

A: Average time difference between the peak Ecc of opposing segments in seven individual rats with MI (MI1 to MI7). Note the tendency of increased delta TP in all rats during RV pacing. B: Analysis of delta TP in each animal focused solely on the two opposing segments presenting maximal dyssynchrony during RV pacing. C: Delta TP focused on the two opposing segment presenting maximal dyssynchrony in seven individual rats with MI (as in Figure 9B).