Improving Ablation Results in Persistent AF: Is Ethanol the Answer?

Miguel Valderrábano MD

Disclosures

J Cardiovasc Electrophysiol. 2019;30(8):1229-1230. 

Pulmonary vein isolation (PVI) has become established as an effective therapy for paroxysmal atrial fibrillation (AF). The mechanistic basis of PVI, as a strategy to isolate ectopic foci that trigger AF paroxysms, is sound. The therapeutic tools to achieve PVI are robust, and the outcomes, albeit imperfect, are satisfactory. Although PVI failures do exist, the majority are due to technical failures, ie, the inability to achieve durable PVI, and only a relative minority can be attributed to nonpulmonary vein triggers. In essence, the mechanistic paradigm of trigger ablation remains unchallenged for paroxysmal AF. Although purely a temporal-pattern characteristic, paroxysmal AF in its prototypical form occurs in structurally normal hearts, and trigger elimination without extensive substrate modification is generally sufficient.

On the other hand, the treatment of nonparoxysmal forms of AF remains challenging. Persistent AF most commonly arises in the context of aged atria, and often coexists with hypertensive heart disease, diabetes, sleep apnea, and other comorbidities. Even in the presence of structurally normal hearts with normal left ventricular ejection fraction, high left atrial pressures, left atrial scar, and varying degrees of diastolic dysfunction are common in persistent AF. By definition, persistent AF patients are in continuous AF for at least a week, but some for months or even years. The coexistence of structural atrial disease and the prolonged temporal persistence are inconsistent with an exclusive relevance of triggers initiating AF, and suggest that a substrate maintaining AF is perhaps a more valid therapeutic target. Identifying accurately such substrate—regions mechanistically involved in the maintenance of AF—has proven extremely difficult. It is likely that no single region is solely responsible for the maintenance of AF, and that interindividual and intraindividual variations in AF maintenance mechanisms exist. Strategies to improve ablation outcomes by extending lesion sets beyond PVI have failed in randomized clinical trials. Reasons may be technical; if ablation lesion sets are incomplete, no benefit is to be expected, but also could be because mechanistically the additional lesions fail to impact AF.

Although it is likely that no single or fixed anatomical location can be expected to be a universally valid ablation target, the ligament and vein of Marshall (LOM and VOM) appear plausible contributors to AF generation and maintenance. Located in the epicardial aspect of the left atrial ridge, the VOM colocalizes with abundant sympathetic and parasympathetic innervation that can modulate atrial myocyte physiology to create a profibrillatory state. It has associated complex myocardial architecture with discrete LOM-atrial connections. It can be a source of ectopic AF triggers (for review, see Rodríguez-Mañero et al[1]). In addition, the VOM is a true atrial vein with direct communication with the underlying myocardium,[2,3] which encompasses the left atrial ridge, and most importantly, the posterior mitral isthmus, between the coronary sinus and the left inferior pulmonary vein.

Aside from surgical excision of the upstream LOM in its extracardiac portion, the VOM-associated arrhythmogenic substrates could not be specifically targeted for ablation. Even after extensive endocardial radiofrequency ablation, robust signals usually persist in the epicardial VOM.[4] We developed the technique of VOM ethanol infusion[5] and validated its ablative effects in canines, which include not only the neighboring myocardium, but also the regional innervation.[5] In humans, VOM ethanol ablation could be synergistic with radiofrequency in achieving isolation of the left pulmonary veins,[2] and could eliminate the occasional direct LOM-PV connections.[4] Furthermore, effective ablation of the mitral isthmus was demonstrated after VOM ethanol.[6] Finally, although the therapeutic implications are purely speculative, VOM ethanol was shown to eliminate regional atrial innervation in humans.[7] The therapeutic role of VOM ethanol in AF ablation remains to be determined, and is currently the focus of the VENUS and MARS clinical trials. (clinicaltrials.gov NCT01898221).

In this issue of the Journal, Liu et al[8] present a series of nonparoxysmal AF patients treated with three different catheter ablation strategies: patients were stratified into group 1 (PVI, substrate modification, VOM ethanol infusion), group 2 (PVI, substrate modification), and group 3 (PVI alone). VOM-treated patients were compared with subsets of the patients in the other groups that were propensity-score matched for other covariates for a final sample of 128 (32, 64, 32, for the respective groups). After an average follow-up time of 3.9 years, higher success at AF or atrial tachyarrhythmia rhythm control was found in group 1 (28.1% recurrence, vs 54.7% and 43.8%), with VOM ethanol being an independent predictor of freedom from AF.

This is a retrospective series of patients subjected to different ablation strategies. The VOM group included patients specifically selected for this treatment based on the presence of VOM triggers of failed mitral isthmus ablation. Therefore, a selection bias is clearly present. On top of VOM ethanol, PVI and complex potential ablation were also performed, which could have added a confounding factor. Consequently, the implications of the outcome differences should be interpreted with caution.

Nevertheless, this study presents safety and feasibility validation of the VOM ethanol procedure, and support a long-term efficacy. Furthermore, it suggests it may lead to improved AF ablation results, at least for certain patients.

It is critical for any technique to have validation beyond its original proponents. Multiple groups have reproduced the VOM ethanol technique safely.[9–12] Questions remain as to the outcomes when applied to unselected AF patient populations. Although a recent small series seemed encouraging,[12] only the randomized VENUS and MARS trials will answer that question.[13]

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....