Defibrillation threshold modeling to refine implant recommendations for an extravascular ICD system

Background The extravascular ICD (EV ICD) is a novel defibrillator that couples a substernal lead to a device under the left axilla. The electrode locations for best defibrillation performance are unknown. In this study we use data from a 21 patient pilot study to address this question. Objective To create computational torso models to study the effect of device and lead position on defibrillation threshold (DFT). Methods Finite element models were developed from CT and intraoperative imaging of the 21 patients enrolled in the EV ICD Pilot Study. Between 150 and 200 combinations of electrode locations were created for each patient, including the actual implant locations as determined from post-implant CT. The shock amplitudes from DFT testing and multiple vector impedance measurements from a large subset of patients and their as-implanted models were used to calibrate global model parameters. In total, 3330 implant models were automatically meshed, simulated and analyzed to compute predicted DFT values. For each patient an acceptable implant zone was calculated in which it is predicted that they would pass implant DFT testing. The zones for all patients were aligned and combined using common anatomical landmarks so that a common zone for acceptable DFT performance could be calculated for the cohort. Results Two patients were omitted due to incomplete CT data and 16 models were used during the calibration phase. Final analysis was performed using 19 patient models. The model results recommend positioning the lead slightly left of the sternal midline with the most proximal electrode 10 mm above the xiphisternal junction. Similarly, the device should be positioned so that its upper margin is at the height of the leftmost projection of the heart as seen in the AP view, and its posterior margin aligned with the posterior margin of the ventricles. The figure illustrates these positions. The model predicts that using these locations will result in acceptable DFT performance for all 19 patients - including one who failed testing at implant. Conclusion Modeling showed that DFT is dependent on both lead and device location. Furthermore, the predicted DFT and impedance values are consistent with clinical measurements. Specific lead and device locations were determined that are both clinically identifiable and provide acceptable DFTs in all patients. Implant positioning Funding Acknowledgement Type of funding source: Private company. Main funding source(s): Medtronic PLC