MI reconnection after EIVM and considerations for associated RF lesions
The success rate of RF alone to achieve acute MI block ranged from 32% to 91% in several studies.19–22, 25 Also, MI reconnection rate late after endo-epicardial RF has been reported as high as 73%.23 These variable and overall low acute and late outcomes are due to the heterogeneous and complex structure of the MI, consisting of a thick myocardium, vascular and nervous structures, epicardial muscular fibers sometimes requiring extensive ablation within the CS and difficulty to achieve catheter-tissue contact and stability.22 Added to RF, EIVM significantly increases MI block success, PMLAF termination rate and long-term freedom from PMLAF relapse.7,8,11,12, 16 In a recent study of 84 patients undergoing EIVM added to RF for MI ablation, MI block was successful in 93% of patients.18 In our study, EIVM added to RF achieved successful MI block in 83% of cases. One case of intra-EIVM scar reentrant AT has been reported,24 but EIVM has consistently been found to create homogenous lesions — in terms of endocardial voltage — and additional RF mostly needed at the mitral annulus edge of the MI.12 Our results are in line with Nakashima’s et al study25, showing same rate of MI reconnection and sites of reconnection. In our study of 24 consecutive patients referred for CA of recurrent AF or AT after EIVM, the MI was reconnected in 25%. The site of MI reconnection was not within the EIVM-related scar but rather at its mitral annulus border in 40% of patients and within the CS in 60%. Additionally, in patients with acute MI ablation failure, residual MI conduction was found at the ridge aspect of the MI at procedure redo.
MI reconnection at its mitral annulus extremity after EIVM could be explained by 3 phenomena: 1) MI mitral annulus extremity being outside of VOM drainage territory, 2) incomplete VOM alcoolization related to balloon-occlusion of the most proximal branches and 3) progressive retraction of the EIVM-created lesion at its border. Scar measurement by using EAM systems may not be accurate enough to reliably assess the last hypothesis, and histopathological works in animals late after EIVM may shed light on this matter. Further works are also needed to optimize the EIVM technique. Ethanol infusion of the entire VOM network may be incomplete in some patients because of the great interindividual anatomical variability and the exclusion of most proximal VOM branches when using balloon occlusion. Development of a technique not involving balloon occlusion could be useful for that purpose. Finally, our results should encourage physicians to consolidate RF lesions at both extremities of the EIVM-created lesion.