Clinical Evaluation of Two Different Evoked Response Sensing Methods for Automatic Capture Detection in the Left Ventricle

Abstract

This acute feasibility study compared two different automatic capture detection methodologies, the reduced coupling capacitor (RCC) and the independent pace/sense (IPS) methods, for the left ventricle (LV). LV threshold tests were performed in DDD mode, with LV-only and bi-ventricular (BiV) pacing using an external cardiac resynchronization therapy (CRT) defibrillator. Evoked response (ER) signals from LV leads were recorded using the LV(Tip) (LV(Tip)-->Can) and LV(Ring) (LV(Ring)-->Can) to empty pulse generator (Can) housing sensing vectors to evaluate the two methodologies. Pacing vector, pulse duration, atrioventricular delay, and interventricular delay were varied to assess their effects on ER. The minimum ER amplitude (ER(min)), signal-to-artifact ratio (SAR), and ER amplitude voltage dependence were evaluated. ER(min)>2 mV and SAR(min)>2 define potential automatic LV capture detection for the two methodologies. Data collected from 43 patients (63.7 +/- 11.0 years) were analyzed, including unipolar and bipolar (14/29) LV leads. Neither ER sensing method was affected by changing the pacing vector. The LV(Tip)-->Can ER(min) was significantly decreased at the 1.0-ms pulse duration when compared to 0.4-ms (p < 0.05). During BiV pacing, LV(Tip)-->Can ER(min) increased at negative interventricular delays and decreased at positive interventricular delays relative to simultaneous pacing. LV(Tip)-->Can resulted in fewer patients with sufficient ER characteristics for capture detection, albeit only significantly at the extended pulse duration (79% vs 97%, p < 0.05) and at simultaneous and positive interventricular delays (81% vs 97%, p < 0.05). Though LV capture detection was feasible using both investigated methods, the RCC method (LV(Tip)-->Can) sensitivity to the evaluated pacing parameters suggests the IPS method (LV(Ring)-->Can) provides a more robust performance.