Since 1995 and the work of Haissaguerre et al. demonstrating the venous origin of ectopic impulses triggering atrial fibrillation (AF), pulmonary vein isolation (PVI) has become the cornerstone of AF ablation. At that time, the promise of an effective curative treatment for fibrillatory disease – with codification and a procedural endpoint – was a true therapeutic revolution. However, this mechanism by itself could not explain the complexity of AF electrical mechanisms and clearly identify all AF drivers during ablation. Owing to high recurrence rates after PVI in persistent AF, several teams have studied other hypotheses focusing more precisely on atrial substrate analysis during ablation.

Thus, Nademanee et al. demonstrated, in 2004, that targeting only complex fractionated electrograms (CFEs) among 121 patients with paroxysmal or persistent AF allowed the restoration of sinus rhythm in 95% of cases and provided very good 12-month results (91% of patients were in sinus rhythm after a mean of 1.2 procedures). However, these results were difficult to reproduce on a large scale due to the complexity of endocardial mapping with a bipolar catheter.

Later, Narayan et al. in 2012 and Haissaguerre et al. in 2014 published two very significant studies demonstrating that bi-atrial mapping with sophisticated signal processing made it possible to identify and ablate AF sources. These results were promising and represented an alternative approach to conventional PVI ablation. However, the ablation strategy remained complex and the question about systematic PVI was unresolved.

In an incessant debate about the most appropriate ablation technique, Verma et al. attempted to answer the question in the Substrate and Trigger Ablation for Reduction of Atrial Fibrillation Trial Part II (STAR AF II), published in 2015. This multicentre randomized study compared three ablation strategies in patients with persistent AF: PVI alone, PVI and additional linear ablation (“lines”), and PVI and ablation of CFE (point-to-point mapping). No significant difference was found between the three groups, allowing the authors to conclude that there was no need for a strategy other than PVI alone in AF ablation, whatever AF type is present. When analyzing the results, three important observations arise. First, the overall success rate at 18 months (patients free from any arrhythmia with or without arrhythmic drugs) was relatively low (49% for PVI alone, 37% for PVI and lines and 41% for PVI and ablation of CFE). Second, PVI was performed systematically in the three groups. Third, and most importantly, the per-procedural AF termination rate was low (45%) in the PVI and ablation of CFE group. Per-procedural AF termination is now clearly recognized as an endpoint in persistent AF ablation and has been correlated to mid-term (5-year) prognosis . As this endpoint was particularly low for the PVI and ablation of CFE subgroup in STAR AF II, we cannot consider the long-term success rate in this subgroup as significant.

In line with Nademanee et al. , we reported, in 2016 , results from a series of 47 patients with persistent AF who underwent a CFE-guided ablation strategy that exclusively targeted the pathological substrate, previously identified by a high-density multipolar catheter. In this study, CFE-guided ablation led to a 92% AF termination rate with short radiofrequency time and, also to a large decrease in frequency of pulmonary vein electrograms, without PVI. The authors concluded that a PV modulation approach, instead of complete PVI, might be preferable for persistent AF. This vision of substrate ablation could explain why CFE ablation is sometimes difficult to reproduce with good results.

Recently, van der Does and de Groot screened 348 articles for new definitions of fractionation. They concluded that “It appears that no uniform definition or recording method is used for measuring fractionation of cardiac atrial electrograms. The different electrophysiological causes for fractionation and the influence of recording device properties on fractionation complicate identification of true pathological inhomogeneous conduction. The first step in discrimination between origins for fractionation may be accomplished by relating electrogram morphology to the spatial patterns of activation. Before revisiting ablation of areas with fractionated electrograms, we need to know the correct method for identifying pathological fractionation.” A regional activity analysis is therefore preferable to that of an isolated CFE.

In 2017, in line with the ablation approach exclusively targeting fibrillatory substrate and an entirely patient-tailored approach, we presented the new concept of spatio-temporal dispersion and its role in AF perpetuation in the Substrate Ablation Guided by High-Density Mapping in Atrial Fibrillation (SUBSTRATE HD) study . We decided to base our analysis on regional activation rather than only on local fractionated morphology of electrograms. Thus, we defined the concept of “spatio-temporal” dispersion: clusters of electrograms, either fractionated or non-fractionated, that displayed interelectrode time and space dispersion at a minimum of three adjacent bipoles such that activation spread over all the AF cycle length. This study was conducted in 105 patients and compared with a validation set of 47 patients ablated in a conventional way (PVI for paroxysmal AF; PVI + lines + CFE for persistent AF). Ablation allowed us to acutely terminate AF in 95% versus 63% in the validation set, with shorter procedure and radiofrequency times. Importantly, this ablation protocol is non-extensive (only 17% of the left atrial surface was ablated with a total radiofrequency time of 49 ± 21 minutes). The 18-month success rate (arrhythmia-free status) was also significantly higher in the study group than in the validation set (in which all patients underwent PVI): 85% vs. 59% ( P < 0.001), after a mean of 1.4 ± 0.5 procedures per patient. This study also demonstrated that atrial post-ablation tachycardia was the manifestation of an underlying atrial substrate rather than the consequence of ablative lesions. In this study, we also added a mechanistic input in AF understanding. Our clinical results, as well as optical mapping and numerical simulations, confirm that spatio-temporal dispersion of electrograms is the electrical signature of AF drivers. This study is the only one presenting a true patient-tailored approach, without any systematic atrial venous isolation or line. This approach directly targets areas of interest, allowing non-extensive ablation, with shape, surface and location of the areas varying from one patient to another. The specificity of this new ablation approach makes it possible to combine efficiency and safety, while preserving–as much as possible–the healthy part of the myocardium and the post-ablation atrial systole. It also avoids unnecessary and potentially dangerous injuries (pulmonary vein stenosis or risk of atrio-oesophageal fistulae in case of a systematic posterior line for PVI).

Interestingly, Jadidi et al. – independently from our group – published a very interesting study with very similar data to ours in 2016. Using spatio-temporal dispersion and voltage analysis to guide ablation (in addition to PVI), they also reported that spatio-temporal dispersion is the electrical signature of AF drivers.

Step by step, the understanding of AF mechanisms improves with evolutions in ablative technique. An automated detection of dispersion areas will undoubtedly improve things further, as well as the integration of additional structural parameters of that atrial myocardium, such as fibrosis. The data from these recent studies give hope of a new method of AF ablation and have already revealed interesting, novel insights about arrhythmogenic mechanisms in AF. Although randomized studies are needed to validate our approach, these results are very encouraging, especially in patients with persistent AF, in whom PVI provides poor results. This approach can also be very useful in paroxysmal AF in case of PVI failure, or as an alternative to PVI if a patient-tailored approach is desired.