Detection of endo-epicardial atrial low-voltage areas using unipolar and omnipolar voltage mapping

Abstract

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): NWO-Vidi Medical Delta Background Endocardial bipolar voltage mapping has emerged as an invasive tool for defining potential target sites for ablation therapy of atrial tachyarrhythmia. Bipolar low-voltage areas (LVA) are commonly considered surrogate markers for areas of conduction disorders. However, bipolar LVAs can be located exclusively at either the endocardium or epicardium. Furthermore, identification of bipolar LVAs is hampered by considerable directional differences in bipolar voltages. Omnipolar voltage mapping, on the other hand, provides maximal bipolar voltage from a collection of electrograms and is directional independent. Unipolar is preferred to identify intramural or epicardial substrates as unipolar electrograms comprise a larger region of myocardial electrical activity. It, however, remains unknown whether simultaneously recorded endo- and epicardial unipolar and omnipolar voltages are complementary or contradictory on identifying LVAs at the opposite layer. Purpose To 1) examine endo-epicardial characteristics in unipolar and omnipolar voltages, 2) explore the relation between various types of voltages in identification of LVAs from an endocardial and epicardial point of view and 3) examine whether characteristics of LVAs can be predictive for LVAs at the opposite layer. Methods Intra-operative simultaneous endo-epicardial mapping (interelectrode distances 2mm) of the right atrium was performed in 93 patients (67±9 years, 73 male). Cliques of 4 electrodes (2x2 mm) were used to define maximal omnipolar (Vomni,max) and unipolar (Vuni,max) voltages. LVAs were defined as Vomni,max ≤0.5 mV or Vuni,max ≤1.0 mV. Results Epicardial Vuni,max were larger than corresponding endocardial Vuni,max (8.2 [4.9–11.6] mV vs 6.1 [3.2–10.9] mV, P<0.001) and they were strongly correlated (R2=0.85, P<0.001). Epicardial Vomni,max were also larger than endocardial Vomni,max (6.7 [3.2–10.9] mV vs 3.3 [1.1–8.6] mV, P<0.001), although there was no correlation between both layers. The majority of unipolar and omnipolar LVAs occurred at either the endocardium (74.2% and 82.0%) or epicardium (52.7% and 47.6%) alone and the opposite areas frequently contained normal voltages. Epicardial Vuni,max and Vomni,max were most accurate in identifying endo-epicardial-LVAs (respectively 47.3% and 52.4% vs 25.8% and 18.0%). Conclusions Unipolar and omnipolar LVAs are frequently located exclusively at either the endocardium or epicardium and could be undetectable when measuring from the opposite layer only. Endo-epicardial LVAs are most accurately identified using combined epicardial unipolar and omnipolar voltages. Therefore, a combination of simultaneously recorded endo-epicardial low unipolar and low omnipolar voltage may be more indicative of dual-layer LVAs and probably arrhythmogenic substrates.