Atrial bipolar voltage analysis to assess vein of Marshall ethanol infusion acute effectiveness and its implication in predicting mitral isthmus block

Background: The vein of Marshall (VoM) is a promising therapeutic target for persistent atrial fibrillation (PeAF) treatment. Ethanol infusion into the VoM (VoM-EI) has demonstrated high effectiveness in facilitating mitral isthmus (MI) block, a fundamental part of the anatomical ablation setup for the treatment of PeAF. However, reliable indexes for optimal ethanol delivery are still lacking, and the role of voltage analysis after VoM-EI in predicting MI block has been poorly explored. Purpose: To evaluate the role of voltage analysis after VoM-EI in predicting acute bidirectional MI block and the need of coronary sinus (CS) radiofrequency (RF) applications. Methods: Fifty-three PeAF patients who underwent catheter ablation were retrospectively enrolled. Left atrial (LA) high-density bipolar voltage mapping was performed before and after VoM-EI. Low voltage areas (LVA) were assessed for all LA maps using the area measurement tool, and the difference in area width between pre- and post-VoM-EI was defined as ∆LVA. An anatomical lesion set including VoM-EI, pulmonary vein isolation (PVI) and linear lesion for dome, lateral MI, and cavo-tricuspid isthmus (CTI) was performed. In case of residual conduction across MI, additional endocardial and/or epicardial ablations approaching the CS musculature were performed. The time required to achieve bidirectional MI block (AblTime-MI) was collected and the VoM length was measured. Results: Forty-eight out of 53 patients enrolled (90.5%) achieved acute bidirectional MI block. After VoM-EI, mean bipolar ∆LVA was 6.6 ± 4.4 cm2, and mean AblTime-MI was 14.7 ± 10.4 min. RF applications targeting CS musculature were required in 22/53 patients (41.5%). Linear regression showed a strong inverse correlation between ∆LVA and AblTime-MI (r = -0.70, β = -128.2; 95% CI -165.3, -91.2; p < 0.001). Patients with higher ∆LVA were less likely to need CS applications [OR = 0.70 (95% CI 0.56-0.88); p = 0.002]. Patients requiring CS applications had significantly longer AblTime-MI (21.0 ± 9.0 min vs 8.8 ± 8.2 min; p < 0.001), and larger left atrial volume index (LAVI) (37.0 ± 10.0 ml/m2 vs 31.8 ± 6.6 ml/m2; p = 0.03). ∆LVA showed the best performance in predicting the need for CS applications (AUC 0.79) with a threshold of < 5.30 cm2 associated with a higher risk of requiring additional RF applications. Conclusions: VoM-EI-induced LA lesions align with VoM trajectory and anatomy. Larger ∆LVA correlates with shorter ablation time for bidirectional MI block and predicts fewer residual epicardial MI gaps requiring RF application in CS musculature. ∆LVA represents a reliable indicator of VoM-EI effectiveness, predicting MI block failure due to epicardial gaps.