Discussion
In this study, we report the feasibility, efficacy, and complication risks related to a moderate volume of VOM ethanol infusions for AF ablation. We demonstrate that VOM ethanol infusion has a learning curve with improved success rates and lowered fluoroscopy times with accumulated experience, even across years of reasonable procedural volume. Rates of VOM perforation, dissection, and overall ethanol infusion success rate were remarkably similar to published reports in a highly experienced center.8
A majority of patients had a CIED or wearable monitors to assess arrhythmia recurrence. In the context of a pandemic follow-up system that was often limited to telemedicine without formal rhythm assessments, and the fact that several procedures were performed in the few months leading up to this study, the average follow-up duration was 9.5 months. We acknowledge that with longer follow-up durations and/or continuous rhythm monitoring, more recurrences may have been detected. Nonetheless, we observed extremely low clinical failure and redo procedure rates, and a one year probability of 80% for arrhythmia freedom. This is prominently viewed in light of a significant proportion of long-standing persistent AF patients and the majority of redo procedures enrolled featuring documented permanent PVI. The left atrial size in this study is similar to that of the CONVERGE trial of hybrid ablation,11 although the proportion of long-standing persistent AF patients is much lower. Studies of similar success rates in persistent AF patients with less comprehensive ablation approaches have only featured subjects with much smaller atrial sizes.12Unfortunately, the degree that any improved efficacy is due to the VOM ethanol infusion per say is not possible to assess by this analysis. The standard lesion set in this study encompassed isolation of the posterior wall, CS, mitral isthmus, and cavotricuspid isthmus in almost all cases. Additional ablation beyond this was not insignificant either, particularly in redo procedures. Nonetheless, other studies have questioned the value of these approaches, particularly for ablation lines which were not found to be helpful in a large randomized trial.3 Even if the VOM ethanol infusion dominates as the marginal difference accounting for the success in this cohort, the mechanism of benefit remains unclear. In VENUS, the benefit was constrained to the patients with successful mitral isthmus block.13It is possible that the VOM ethanol infusion merely ensures more permanent mitral block14-16 or isolation around the left pulmonary vein antra. However additional effects of autonomic denervation17, debulking of atrial mass, or direct suppression of ligament of Marshall AF triggers18 are plausible as well.
Despite excellent arrhythmia freedom in our cohort, review of these patient outcomes unveiled important safety concerns that also may differ in frequency from those after PVI alone. To the best of our knowledge, we note the first demonstration of sinus nodal injury directly from VOM ethanol infusion. The underlying mechanism is not clear. Although not apparent on venography, we suspect that collateral vessels were present and delivered ethanol to the sinoatrial region. We did not observe any AV block, left atrial appendage isolation, or anaphylaxis that have been described with VOM ethanol.8 With respect to the important complication of tamponade we demonstrated a high rate of delayed pericardial effusions. Two of the four patients in this cohort with delayed effusions had unrevealing echocardiograms days after their ablations, only to later present with pericardial tamponade. A third had a hematocrit less than 1% on the pericardial fluid despite a bloody appearance. These findings suggest that the dominant mechanism of delayed tamponade in these cases was inflammatory pericarditis rather than hemorrhagic. Early on in our experience we adjusted our procedural technique to minimize CS related complication. A retention wire is used through the CS guide sheath to avoid CS wall trauma and the IMA catheter is inserted over a wire. Furthermore, we probe for the VOM with a low tip weight angioplasty wire (Suoh wire, Asahi) rather than localize the VOM with contrast puffs. This should lower the risk of VOM perforation/dissection or CS dissection and while we have neared a 100% success rate with VOM ethanol infusion, we have noted a higher rate of delayed tamponade than reported previously in the VENUS study or Bordeaux experience.8,12The rate is more in line with pericardial effusions in the CONVERGE study of hybrid surgical ablation.3 While the low total number of cases leaves open the possibility that our increased rate of delayed tamponade is due to chance alone, interestingly, all four cases of delayed tamponade in our series occurred in the second half of our cohort of patients. We have postulated that as we have become more facile with the procedure, perhaps a brisker workflow with faster ethanol infusions may have contributed. Thus while we typically still instill four ethanol injections, we have minimized the total volume of ethanol infused from an average of 10ml to 4-5ml, and each injection is instilled slowly over 60 seconds. In addition, in order to avoid contrast induced hydraulic dissections of the VOM, we do not any longer systematically perform VOM venography unless there is uncertainty of the vein identity (non-VOM), the quality of venous occlusion as evidenced by balloon movement, or lack of demonstrable ethanol effects on mapping or ICE. Further study is required to tell if this approach will lower the rate of delayed effusions related to VOM ethanol. Importantly, noting this complication has recalibrated our judgement and patient selection in deciding when to employ the technique.
This study is limited by its retrospective nature, selected population, and the fact that all cases were performed at a single center with a single operator. As advances in ablation techniques and energy sources bring us closer to ensuring permanent PVI,9,10the role for VOM ethanol and other adjunctive techniques for AF ablation may take on greater importance. Currently, there is a paucity of published experience with VOM ethanol infusion and it is rarely performed outside of select centers of excellence. We believe that cases series such as this add important insights to the literature of this technique, informing operators already performing or considering VOM ethanol infusion in the future.