“Thinking Outside the Box”

Ever since it was recognized that atrial fibrillation (AF) could be triggered by rapid atrial firing from ectopic foci, radiofrequency ablation targeted at eliminating those foci has emerged as an effective method for the treatment of AF. In particular, Haissaguerre et al. described that for patients with lone, paroxysmal AF, most of these ectopic foci were located within the pulmonary veins (PVs) and that electrical isolation of the PVs could not only eliminate these triggers, but also prevent AF recurrence.1 As a result of this seminal observation, most of the approaches for AF ablation employed today are designed around the electrical isolation of all four PVs.2 Evidence has also shown that the success of such ablation procedures is directly related to eliminating conduction between the PVs and the left atrium (LA). Verma et al. studied patients post-PV antrum isolation and found that those with successful outcomes had significantly more PVs isolated, compared with those who failed.3 Furthermore, patients who were responsive to antiarrhythmic medications had more conduction delay between the LA and PVs versus those who were not responsive.3 Ouyang et al. also found that recurrent LA-PV conduction was the predominant finding in patients with recurrent arrhythmia post-PV antrum isolation.4 In both studies, patients were successfully cured by re-isolating the PVs. The techniques for PV isolation have evolved over time, however, as the importance of the LA posterior wall in the triggering and maintenance of AF has been understood.5 Earlier studies done at a time when lesions were restricted to the PV ostia only suggested that 10–20% of patients could have non-PV triggering foci, the majority of these on the posterior LA wall.6 Experimental data also demonstrated that maximal dominant frequencies during AF, perhaps representing rotors critical to AF maintenance, were also located on the posterior LA wall.7 In a human surgical series, Todd et al. showed that a “box-type” lesion set encircling the four PVs and the posterior LA wall could successfully prevent AF in a paroxysmal AF population.8 They also found that they could induce and sustain AF within the isolated posterior wall and PV region, while the rest of the LA could not. Thus, they concluded that the PV-posterior wall was essential to AF maintenance. Anatomically, however, it is actually difficult to separate the posterior wall from the posterior extensions of the PVs. The PVs are not simply tubes that emerge from the posterior LA, but are funnel-shaped structures that fan out into a proximal “cup” or “antrum” that blends into the posterior atrial wall.9 These antra often converge so that there is very little “free” posterior wall between them. Thus, to fully isolate these antral structures, much of the posterior LA wall must be ablated. Consequently, as AF ablation evolved, lesion sets migrated away from the tubular ostium of the PVs and out into the LA to encompass more of these posterior antral extensions. Clinical trials suggested that these wider lesion sets were more efficacious than previous ostial lesions.10 Although there are many variations to AF ablation described (including double Lasso technique, intracardiac echocardiography-guided, or wide circumferential left atrial lines), most of the posterior LA ends up being ablated in all of them and with good effect.2 Of course, the “Achilles heel” of such extensive posterior LA ablation includes the risk of longer ablation time in the LA, but also the risk of esophageal injury. The esophagus is directly posterior to the LA, often running within 2–4 mm of the posterior LA wall.11 With the extent and intensity of ablation performed on the posterior LA, it is not surprising that increasing numbers of left atrial esophageal fistulae are being reported.12 Although this complication appears exceedingly rare (less than 1 in 1,000), it is nearly always fatal and its magnitude is compounded by the fact that AF ablation has never been shown to reduce patient mortality. Unfortunately, avoidance of ablation over the esophagus is not practical since the esophagus frequently runs very close to the PV-LA junction—a key region for ablation. Furthermore, esophageal position may change over the course of the procedure.13 While some authors have suggested various techniques to avoid esophageal injury, such as visualizing the esophagus, limiting maximum power output, monitoring esophageal luminal temperature, or even cooling the esophagus, no technique to date has been shown to reduce the risk of injury. Given the very low incidence of the complication, it is unlikely that any clinical trial will ever be able to show definitively a reduction in risk. However, investigators have tried to propose alternative methods of ablation, in an effort to reduce posterior LA ablation and hopefully mitigate the risk. In the current issue of the Journal, Kumagai et al. describe one such alternative approach to left atrial ablation for AF.14 In a prospective single-center cohort of 91 patients with mostly lone, paroxysmal AF, the authors perform linear lesions to surround completely and isolate the entire posterior LA wall including the four PVs. These lesions form a square set around the four PVs, very similar to the surgical lesion set described by Todd et al.,8 and hence the use of the term “box isolation.” Continuous lesions were first performed along the anterior borders of the PVs on both the right and left sides, and then a roof line, followed by an inferior LA line were created to complete the box. Lesions along the esophageal aspect of the posterior LA were avoided, theoretically reducing the risk of esophageal injury. Successful box isolation was acutely achieved in 90% of the patients, with ablation resulting in a 65% rate of AF termination and 71% of patients becoming non-inducible for AF. At just over 1 year of follow-up, 95% of the patients were arrhythmia-free without antiarrhythmic drugs, with only six (7%) of patients undergoing a second ablation procedure. Based on these results, the authors concluded that box isolation is feasible and effective and offers the advantage of not having to create lesions along the esophageal aspect of the posterior LA. The authors' intent of reducing posterior LA ablation is clearly a good one since the best way to avoid esophageal injury is to avoid ablation. Furthermore, the technique seems to have a very good long-term success rate in a lone, paroxysmal AF population, with a high rate of patients becoming non-inducible for AF. This is not the first study to describe a “box” lesion set for AF using percutaneous catheter ablation. Ernst et al. proposed an almost identical lesion set many years ago, but because of limitations in catheter technology, these linear ablations could not be completed in most patients, and there was nearly a 100% recurrence rate.15 Within the last couple of months, Thomas et al. proposed a similar lesion set, and the clinical outcome was almost identical to a control group using the more traditional lesion set of surrounding the four PVs by extensive posterior LA ablation.16 Obviously, the findings of Kumagai et al. lend further support to the box isolation technique and in turn support the fact that the posterior LA wall, with its antral extensions, plays an important role in both the initiation and maintenance of AF. However, whether the technique will completely avoid esophageal injury and whether it offers benefit over traditionally applied techniques remains in question. The most distinguishing feature of box isolation is that it purports to avoid—or at least minimize—ablation along the esophageal aspect of the posterior LA. As mentioned earlier, given the extremely small incidence of serious esophageal injury post-ablation, it would be impossible to determine if box isolation successfully avoids this complication in a study of only 91 patients. Furthermore, the present study demonstrated rises in esophageal temperature of almost 1°C during ablation of the inferior line of the box. The authors do not specify in how many patients this rise was seen, but the fact that the rise occurred within 10 seconds of ablation and reached about 38°C in at least one patient suggests that the esophageal course is definitely transected by this line. The roof line may avoid the esophageal course, but it must be directed very cranially to do so. When performed in this fashion, roof line ablation becomes technically very challenging because the LA roof is very thick and catheter stability may be very difficult to achieve.17 In the study by Thomas et al., for example, very long procedural and radiofrequency times were needed to complete the box. In spite of this, complete block across the roof and total posterior LA isolation could not be achieved in 40% of the patients without further ablation along the posterior LA within the box and close to the esophagus.16 If roof ablation is moved more posteriorly, it becomes much easier to perform, but the line may then transect the esophageal path. In the present study, the roof line only took 9 minutes to perform with an 8-mm tip catheter at 30-35 W. This suggests that the line was drawn more posteriorly, since other investigators have shown that higher powers with irrigated-tip catheters are often required to achieve cranial roof line block. The representative figures for the present study also show a more posterior roof line. Again, such ablation may reduce the ability of the box technique to reduce esophageal risk. Apart from the theoretical reduction of ablation over the esophagus, box isolation may not offer much advantage compared with wide antral PV isolation techniques being used today. Radiofrequency, fluoroscopy, and procedural times in the present study were all similar to those reported for wide antral PV isolation.3 The success rates were also similar to success rates reported for other techniques for lone, paroxysmal AF patients with minimal structural heart disease and normal LA size.2 Roof and floor lines are often inadvertently created by the convergence of posterior ablation lines while isolating the PV antra.9 Importantly, previous evidence has suggested that performing linear ablations with incomplete block may be pro-arrhythmic as opposed to curative. Incomplete roof line block, for example, could alter LA conduction in such a way as to encourage perimitral reentry.17 In the present study, complete block lines could not be achieved acutely in 10% of patients and left atrial flutter was inducible in 22% post-ablation. Fortunately, only 10% of patients presented with arrhythmia after one procedure, but the recurrence rates in Thomas et al. were considerably higher. In the present study, pacing was performed within the box to demonstrate exit block, and electrical silence was confirmed by mapping within the box. Whether additional pacing maneuvers are required to confirm block across the roof line—as has been described elsewhere17—is not known, but would definitely add further complexity to the procedure. It is of interest that no AF was seen in the isolated posterior LA segment in this study. In Todd et al.,8 4 of 14 patients (29%) had spontaneous AF or atrial tachycardias within the posterior isolated segment that did not conduct into the rest of the atria. This raises an important question about the effects of box isolation on LA contractility and the ongoing risk of thromboembolism. It is unknown whether ongoing fibrillatory activity or dyssynchronous activity within this segment changes embolic risk compared with techniques in which the entire posterior wall is ablated. Unfortunately, the present study does not report on any assessments of LA function and further study into this question is required. Finally, it should be noted that the present study included only patients with lone, paroxysmal AF. Box isolation does not include several areas of the atria that have been shown to be important for AF maintenance, namely the superior vena cava, the interatrial septum, the base of the left atrial appendage, and the anterior LA wall. Other studies have shown that complex fractionated electrograms,18 AF nests,19 and autonomic ganglia20 may occur in these areas outside the box. For patients with lone, paroxysmal AF, inclusion of these other targets may not be as important given the high procedural success rates in this study. However, whether such success rates will translate to persistent or permanent AF populations is still unknown. Despite its limitations, the study of Kumagai et al. makes an important contribution in expanding the way in which LA ablation may be performed to treat AF. They have shown that a box-type lesion set is not only feasible, but can be associated with very high procedural success rates in a lone, paroxysmal AF population. Furthermore, the isolation may be completed in a similar time frame to wide antral PV isolation techniques that are commonly used today. The study confirms the importance of the posterior wall and its antral extensions in the triggering and maintenance of AF. Whether or not this technique truly reduces esophageal risk and by what magnitude is unclear, and no definitive answer will be forthcoming short of performing a study with thousands of patients to account for the rarity of this complication. As well, it is not known how well this technique would apply to other AF populations and the effects it may have on LA function. As we move forward to answer these questions, this study reminds all of us about the importance of exploring alternative ablation techniques and thinking outside of the box.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.