Chronic ambulatory sensing comparison between intercostal extravascular versus transvenous ICD lead systems

Background A novel extravascular (EV) ICD lead designed for compatibility with commercial DF-4 compatible ICD pulse generators (PG) is under development. Objective We compare within-patient, simultaneous ambulatory sensing and detection between concomitantly implanted transvenous (TV) and EV-ICD systems with matched programming and sensing algorithms. Methods Two iterative EV-ICD studies, SECURE EV (n=16) and STEP ICD (n=20), were combined to assess ambulatory EV-ICD sensing compared to a matched TV-ICD system. The sensing electrodes, distances and orientation were largely the same in both studies. Subjects underwent concomitant implantation of a TV-ICD System and investigational EV-ICD lead connected to a record only ICD PG. Whenever possible, the same ICD manufacturer was used for both systems. Devices were interrogated periodically over a 90-day period to assess recorded episodes. Results Thirty-six subjects were implanted and ambulatory with a concomitant EV-ICD (record only) and TV-ICD System (Active). No appropriately detected episodes requiring therapy were detected by either system during follow-up. Matched events (recorded by both systems) include 4 non-sustained appropriate ventricular tachyarrhythmia (VT) detections in 3 subjects and 4 rapidly conducted supraventricular tachyarrhythmias in 4 subjects. Unmatched events include T-wave oversensing by the TV-ICD only (1 subject) and non-cardiac noise detected by the EV-ICD alone (n=5). The causes of these events include: 1) algorithm limits (TV-ICD), 2) lead fracture, 3) air in the mediastinum, 4) EMI, 5) diaphragm noise, and 6) sensitivity programming in response to low (<1 mV R waves) amplitude signals. Of the 5 subjects with non-cardiac noise on the EV-ICD lead, 4 came from the 1st generation lead. The 2nd generation lead design, procedure improvements and implant criteria related to sensing were implemented to mitigate the potential for similar occurrences. Two appropriately detected non-sustained VT episodes were recorded by the EV-ICD alone in a subject with unmatched ICD PG manufacturers (due to limited ICD PG availability) with different detection criteria, so no inferences can be made about the TV-ICD mal-detection comparatively. Conclusion Feasibility studies of an investigational intercostal EV-ICD lead demonstrated accurate sensing and detection of ambulatory VT. Design enhancements to the 2nd generation EV ICD lead decreased inappropriate detections related to extracardiac noise. Further development is planned.