Ablate, or not to ablate left atrial low-voltage areas in atrial fibrillation patients?

This correspondence refers to ‘Win ratio analysis of low-voltage area ablation in persistent atrial fibrillation: sub-analysis of SUPPRESS-AF’, by A. Sunaga et al. https://doi.org/10.1093/ehjopen/oeag024 The authors’ response is available at: https://doi.org/10.1093/ehjopen/oeag113 We read with interest the paper by Sunaga et al.1 that presents the results of a sub-analysis of the SUPPRESS-AF randomized trial on left atrial low-voltage ablation (LVA) in patients undergoing the first ablation for persistent atrial fibrillation (AF). Their results echo one of the main studies that assessed no additional benefit in terms of arrhythmia recurrences by adjunctive ablation of LVAs beyond pulmonary vein isolation (PVI) in patients with persistent AF. Additionally, the win ratio analysis adopted in this sub-study shows no improvement in clinical outcomes considered in a hierarchical sequence prioritizing safety. With the acknowledged limitation that the study was powered to evaluate only the primary endpoint (arrhythmia recurrences), leading to a low number of clinical events comprised secondary endpoints, it should also be considered that the results of a win ratio analysis are inherently influenced by pre-specified analytical choices, including the hierarchy and order of outcomes, the number of considered outcomes, and how ties are handled. Indeed, this study brings up the dilemma of whether additional ablation of LVA beyond PVI could be beneficial in AF patients without increasing complications and unnecessary procedure prolongation. Currently, ablation of LVAs—generally defined as areas with a peak-to-peak bipolar voltage ≤0.5 mV at high-density electroanatomic mapping in sinus rhythm—is an area of uncertainty, rarely performed in first procedures and in paroxysmal cases. In general, the need for adjunctive ablation reflects that in some cases the outcome of PVI, regardless of the energy used, does not meet expectations. If, on one hand, the use of ablation earlier in patients’ history and the removal of risk factors will lead in the future to a reduced burden of atrial cardiomyopathy, on the other, in some patients currently referred for AF ablation, extensive atrial remodeling limits the results of a conservative procedure. Notably, atrial cardiomyopathy includes a complex anatomic and histologic remodeling altering the atrial wall in terms of both arrhythmogenicity and thrombogenicity. The presence of LVAs is per se associated with increased risk of atrial arrhythmia recurrences and cardiovascular events, including heart failure and stroke, in the follow-up after PVI-only procedure.2 Furthermore, the presence of fibroblasts and other non-myocyte cells, observed in atrial cardiomyopathy and detected as LVAs, can be arrhythmogenic based not only on conduction delay or block but also on the cross-talk between myocytes and non-myocyte cells that can perturb the electrical activity of the myocardium. Finally, even minimal extension of the LVAs is associated with altered mechanical and electrical left atrial function. In fact, AF patients with minimal LVAs (<5% of the left atrial surface) may have left atrial mechanical dysfunction associated with post-ablation recurrences, similar to what is observed in those with more extensive LVAs.3 This suggests that the development of atrial cardiomyopathy is a dynamic phenomenon, which progresses from stressors to left atrial mechanical alterations and eventually leads to histologic changes manifest as LVAs. However, in the context of uncertainty, most of the studies, including controlled randomized trials, meta-analysis of randomized trials, and non-randomized studies (Table 1), report a lower recurrence rate of atrial arrhythmia during follow-up when additional LVA ablation is performed, with no significant difference in adverse events compared to patients undergoing PVI only. This seems particularly evident in persistent AF. Among the three studies showing no benefit in additional LVA, important considerations should be made. The DECAAAF II study has two major potential biases. First, there could be a spatial mismatch between fibrosis identified by cardiac magnetic resonance and LVAs. Secondly, the subsequent sub-study showed a reduction of recurrences in patients with limited fibrosis (<20% of the left atrial surface) when radiofrequency energy ablation covers the entire area of fibrosis. Similarly, for the other two randomized trials showing no benefit, namely the STABLE-SR II and SUPPRESS AF trials, subsequent sub-analyses showed possible benefit if a stricter patient selection is used based on age or left atrial diameter. Summary of the main studies comparing pulmonary vein–only ablation vs. pulmonary vein ablation and additional low-voltage area ablation 3D map, three-dimensional mapping; AEs, adverse events; AF, atrial fibrillation; cMR, cardiac magnetic resonance; Cryo, cryoballoon ablation; mos., months; LAD, left atrial diameter; LVAs, low-voltage areas; Pers., persistent; PVI+, pulmonary vein isolation +low-voltage area ablation; RCT, randomized controlled trial; RF, radiofrequency energy; yrs, years. ↓, reduces; Δ, difference; ↑, improves. Finally, definitive conclusions on this topic have yet to be established. In the sub-analysis by Sunaga et al.1 patients with limited LVAs (<5 cm2) were excluded, although they might have benefited more considering both the potential arrhythmogenicity of even small LVAs and the presence of limited atrial cardiomyopathy. Moreover, scar homogenization was obtained in only 78% of the cases, and anticoagulation beyond 3-month follow-up was left to the physician’s discretion. Future studies overcoming these confounders with strict protocols for LVA definition, patient selection, and ablation endpoint should provide evidence of the possible benefit of additional LVA ablation. Roberto De Ponti (Conceptualization, Data curation, Validation, Writing—review & editing [equal], Resources, Writing—original draft [lead]), Carola Griffith Brookles (Conceptualization, Data curation, Resources, Validation, Writing—original draft, Writing—review & editing [equal]), and Matteo Anselmino (Conceptualization, Data curation, Resources, Validation, Writing—original draft, Writing—review & editing [equal]). There was no funding for this paper.