Basic/Translational Science -> Computer Modeling/Simulation D-PO02 - Poster Session II (ID 47) Poster

D-PO02-034 - Optimal LV Pacing Locations In Endocardial Cardiac Resynchronization Therapy Can Be Predicted With Non-invasive Electrical Indices (ID 990)

Abstract

Background: Endocardial CRT is a novel technology that allows for pacing at any location inside the LV. It is important to identify the optimal LV lead location, as sub-optimal placement leads to non-response. Clinically, optimal pacing sites can be identified from invasive acute hemodynamic response (AHR) measurements. However, testing all potential endocardial pacing sites is unrealistic, putting undue burden and risks on patients.
Objective: To assess the ability of electrical simulations to non-invasively predict optimal LV lead locations for endocardial CRT.
Methods: Clinical data (CT images and QRSd) acquired from 8 endocardial CRT patients were used to personalize the anatomy and electrical properties of cardiac models. The AHR was clinically measured with pacing at multiple sites (2-7) per patient, with positive response defined as AHR≥10%. We simulated electrical activation with RV [Fig A] and biventricular (BiV) pacing [Fig B].
Results: We tested the ability of the simulated LV delay with RV pacing and change in QRSd with BiV pacing in predicting optimal response (AHR≥10%). Area under the ROC curves were assessed for relative and absolute values of LV delay (0.74 and 0.66, respectively) and QRSd change (0.56 and 0.49, respectively) [Fig C]. We were able to predict positive response with absolute LV delay≥45ms or relative LV delay≥60% (p<0.05) [Fig D], while QRSd change was not predictive.
Conclusion: Non-invasive electrical models can predict the optimal LV endocardial lead location. Prospective analysis of these models could allow clinicians to test the AHR at targeted sites reducing time, costs and risks of the procedure.

Collapse