Intraoperative characterization of interventricular mechanical dyssynchrony using electroanatomic mapping system—a feasibility study

Abstract

Background

Interventricular mechanical dyssynchrony (VVMD) is a strong predictor of cardiac resynchronization therapy (CRT) response. However, no simple and reliable clinical method of measuring VVMD during CRT implant is currently available. We tested the hypothesis that the EnSite™ NavX™ system (St. Jude Medical, St. Paul, MN, USA) can be used intraoperatively to determine VVMD, thereby facilitating CRT optimization.

Methods

During CRT implant, the leads in the right atrium (RA), right ventricle (RV), and left ventricle (LV) were connected to the EnSite™ NavX™ system to record the real-time 3D motion of the lead electrodes. The distances from RA to RV lead electrodes (RA–RV) and RA to LV lead electrodes (RA–LV) were computed over ten cardiac cycles during each of RV pacing and biventricular (BiV) pacing, respectively. The degree of synchrony was computed from the distance waveforms between RA–RV and RA–LV by a cross-covariance method to characterize VVMD. Septal-to-posterior wall motion delay (SPWMD) from M-mode echocardiography (echo) was measured for reference at each pacing intervention. VVMD was present in all five patients undergoing CRT implant.

Results

Four of the five patients demonstrated clear improvement in EnSite™ NavX™-derived VVMD during BiV versus RV pacing, which corresponded to the SPWMD results by echo.

Conclusions

It is feasible to characterize VVMD and resynchronization in CRT patients with the EnSite™ NavX™ system during implant, demonstrating its potential as a tool for intraoperative CRT optimization.

Appendix

EMSI was computed as follows. Given the recorded 3D positions of the tip electrodes of each respective lead, RA (x ₁, y ₁, z ₁); RV (x ₂, y ₂, z ₂); and LV (x ₃, y ₃, z ₃), the vector between RA and RV, and RA and LV (abbreviated as RA–RV and RA–LV, respectively) was calculated using the following equations:

$$ \overrightarrow {{\text RA} - {\text RV} } = \overrightarrow {RV} - \overrightarrow {RA} $$

(1)

$$ \overrightarrow {{\text RA} - {\text LV} } = \overrightarrow {LV} - \overrightarrow {{\text RA} } $$

(2)

As described in the text, the average position of the RA electrode over a cardiac cycle is used as a fixed reference point, while the positions of the RV and LV electrodes are functions of time. The distance between the respective electrodes is computed as the magnitude of the vectors.

$$ \left| {\overrightarrow {{\text RA} - {\text RV} } } \right| = \sqrt {{{{\left( {{x_2} - {x_1}} \right)}^2} + {{\left( {{y_2} - {y_1}} \right)}^2} + {{\left( {{z_2} - {z_1}} \right)}^2}}} $$

(3)

$$ \left| {\overrightarrow {{\text RA} - {\text LV} } } \right| = \sqrt {{{{\left( {{x_3} - {x_1}} \right)}^2} + {{\left( {{y_3} - {y_1}} \right)}^2} + {{\left( {{z_3} - {z_1}} \right)}^2}}} $$

(4)

EMSI, which is the cross-covariance coefficient of the RA–RV and RA–LV waveforms, is calculated by

$$ {c_{{xy}}}(m) = \sum\limits_{{n = 0}}^{{N - m - 1}} {\left[ {x\left( {n + m} \right) - \frac{1}{N}\sum\limits_{{i = 0}}^{{N - 1}} {{x_i}} } \right]} \left( {y_n^{ * } - \frac{1}{N}\sum\limits_{{i = 0}}^{{n - 1}} {y_i^{ * }} } \right) $$

where c is the cross-covariance correlation, $x = \left| {\overrightarrow {{\text RA} - {\text RV} } } \right|{, }y = \left| {\overrightarrow {{\text RA} - {\text LV} } } \right|$, N is the length of the RA–RV and RA–LV waveforms, and m = 1, 2,…2 N − 1.