Influence of atrial cycle length and site of origin on the peak frequency of bipolar electrograms to discriminate isthmus conduction in atrial scar

Abstract Funding Acknowledgements Type of funding sources: Private company. Main funding source(s): Abbott Background Peak frequency (PF) of conventional bipolar electrograms is a novel parameter which may distinguish between near-field (NF) and far-field (FF) signals. However, the influence of activation rate and the direction of the activation front on it is unknown. Purpose 1. To study the influence of atrial cycle length and wavefront origin on PF to detect residual conduction in a well-defined model of narrow isthmus of conduction such as subacute pulmonary vein (PV) reconnection following atrial fibrillation ablation. 2. To compare this influence with that found for Bipolar voltage (BiV). Methods Baseline maps were acquired in redo PV isolation (PVI) procedures with a 16-pole grid catheter (HD-Grid) during low rate sinus rhythm (SR), low rate coronary sinus pacing at 500 ms cycle length (CSLR) and high rate coronary sinus pacing at 300 ms cycle length (CSHR). PFs were retrospectively computed. PVI sites requiring ≤3 radiofrequency applications were included, with EGMʼs ≤1cm from the PVI site classified as GAP (vs >1cm, No-GAP). Results 28 GAPs were found in 12 of 14 consecutive patients in the study. 3976 EGMʼs (1547 GAP vs 2429 No-GAP) were analyzed. In both GAP and No-GAP, PF was similar in SR vs CSLR (GAP: 363±132 Hz -SR vs 345±135 Hz -CSLR, P=NS ; No-GAP: 196±110 Hz -SR vs 181±116 Hz -CSLR, P=NS). PF was slightly higher for CSLR vs CSHR. (GAP: 345±135 Hz -CSLR vs 317±154 Hz -CSHR, P<0.001; No-GAP 181±116 mV -CSLR vs 162±129 mV -CSHR, P<0.01). (Panel A). BiV was significantly higher in SR vs CS pacing but similar between CSLR and CSHR (GAP: 1.86±2.11 mV -SR vs 0.87±1.19 mV -CSLR vs 0.74±0.92 mV -CSHR, P= NS, No-GAP: 0.62±1.22 mV -SR vs 0.33±0.68 mV - CSLR vs 0.27±0.48 mV - CSHR, P=NS. (Panel B). ROC GAP discrimination for PF was similar between SR and CSLR (AUCʼs: 0.86-SR, 0.84-CSLR) and slightly lower for CSHR (AUC: 0.81) with optimal cutoffs of 260, 240 and 220 Hz, respectively. ROC GAP discrimination for BiV was lower vs PF in all rhythm modes (AUCʼs: 0.77-SR, 0.74-CSLR, 0.75-CSHR), with optimal cutoffs of 0.3, 0.2 and 0.15 mV, respectively. Conclusion ROC gap discrimination was significantly higher for all rhythm modes with PF vs BiV. PF showed slightly decreasing cutoff values for SR, CSLR and CSHR respectively, whereas BiV showed a greater relative decrease in cutoffs for SR, CSLR and CSHR.