Down But Not Out—Addressing the Scourge of Late Pulmonary Vein Reconnection

Konstantinos C. Siontis MD; Peter A. Noseworthy MD


J Cardiovasc Electrophysiol. 2019;30(6):824-826. 

Pulmonary vein isolation (PVI), the mainstay ablative approach for atrial fibrillation (AF), is an effective management strategy, but its Achilles' heel continues to be the poor durability of the lesion sets which results in a relatively high recurrence rate of the arrhythmia during long-term follow-up.[1] This is in spite of acute isolation rates being virtually 100% in most labs worldwide. Beyond the traditional procedure endpoint of entrance block into the PVs with or without exit block, there is no consensus on the optimal approach to assess the quality and predict the durability of the PVI. Several methods have been used with variable levels of supporting evidence and adoption in clinical practice, including adenosine testing to reveal dormant conduction (DC) and pace-capture testing along the ablation line to identify ablation gaps, among others.

Adenosine can unmask DC in up to half of patients[2] by transiently hyperpolarizing the membrane potential and restoring excitability of cells that are reversibly injured by ablation. PVs with DC demonstrated with adenosine administration are more likely to recover conduction spontaneously 30 to 60 minutes after initial PVI.[3] Despite promising initial results in observational studies, randomized studies of adenosine-guided PVI have not shown consistent superiority over standard PVI, possibly due to limited specificity and sensitivity of adenosine in revealing reconnections that are likely to be clinically relevant in the long run. Indeed, Ghanbari et al[4] demonstrated that in a group of 61 patients who underwent adenosine-guided PVI, 12 patients had redo procedures and 29 PVs with reconnections were identified, but only four of these PVs had previously shown transient adenosine-induced reconnection during the initial procedure. In the largest randomized controlled trial (RCT) to-date with greater than 2000 patients, adenosine administration revealed DC in almost one in three patients but additional ablation was not associated with improved survival free of AF recurrence compared with a non-adenosine-guided approach. Similar results were seen in two other smaller RCTs,[4,5] whereas a study comparing the procedural endpoints of elimination of PV-left atrium (LA) DC vs ablation of extra-PV triggers found the latter approach to lead to superior long-term outcomes[6] suggesting that the durable PVI is not enough on its own. However, in the ADVICE trial, additional ablation to eliminate adenosine-induced DC resulted in improved arrhythmia-free survival compared with a strategy of no further ablation. Even with the inclusion of the ADVICE study in a meta-analysis of all RCTs of adenosine testing, the overall effect is neutral with respect to the long-term outcome of arrhythmia recurrence (Figure 1).

Figure 1.

Random-effects meta-analyses of randomized controlled trials of adenosine testing and pace-capture testing in patients undergoing pulmonary vein isolation (endpoint of atrial tachyarrhythmia recurrence)

The pace-capture approach has been studied less extensively. Initial promising clinical results with this approach[7] were not replicated in a more recent, larger RCT in which contact-force sensing catheters were used.[8] A nonrandomized comparison between adenosine and pace-capture testing revealed similar DC rates and long-term arrhythmia recurrences between the two approaches.[9] Both testing approaches were given a lukewarm class IIb recommendation in the 2017 HRS/EHRA/ECAS/APHRS/SOLAECE expert consensus statement on catheter and surgical ablation of AF.[10] In fact, only 61% and 24% of the writing group members routinely perform adenosine and pace-capture testing, respectively.

Against this background, the combination of adenosine and unexcitability testing along the PVI line was studied by Zeng et al[11] in a single-center RCT of 107 patients with paroxysmal AF reported in this issue of the journal. In one group (n = 57), adenosine was administered first. If DC was uncovered, ablation was performed until adenosine testing was negative. Then, pace-capture testing on the PVI line was performed followed by additional ablation until there was no capture on the PVI line. In the other group (n = 54), the tests were performed in the opposite sequence. Of note, all patients underwent ablation with contact-force sensing irrigated-tip catheters and the procedural endpoint was negative adenosine testing plus negative pace-capture testing in both study groups. The investigators also included a historical, nonrandomized control group of 107 patients with paroxysmal AF from their institution in whom only an adenosine-guided approach was used.

Regardless of the order in which the two tests were applied, each test identified patients in whom the PVI was incomplete despite the other (initial) test being negative—suggesting that the two modalities assess different aspects of the adequacy of the PVI lesion set. In the first group, pace-capture testing identified gaps in almost half of the patients with a negative adenosine test. In the second group, adenosine testing uncovered DC in nearly one in five patients with a negative pace-capture test. At 12-month follow-up, there was no difference between the two groups in freedom from atrial tachyarrhythmias. However, the combined study group (all 107 patients) had significantly better survival free of atrial tachyarrhythmias compared with the historical control group (91.6% vs 81.3%, P = 0.031).

The current study adds to the body of literature evaluating the quality of PVI lesions toward optimizing long-term outcomes of ablation. One of the most important merits of the study is that it assesses the combination of adenosine and pace-capture testing in all patients, which provides an opportunity to better appreciate the limitations of each approach when used in isolation. In particular, the nearly 50% rate of excitable myocardium on seemingly complete PVI lines after adenosine-guided ablation is remarkable. Conversely, even though a previous study suggested that pace-capture testing may make adenosine testing expendable,[11] Zeng et al reported that approximately one in five patients with negative pace-capture test were found to have DC by subsequent adenosine testing. Notably, only two patients had positive pace-capture and positive DC at the same anatomic sites. This is in keeping with prior investigations where there was little overlap of PVs with adenosine-induced DC and with positive pace-capture, and a significant number of patients had residual DC despite the loss of excitability along the PVI line.[9,12] Thus, the two approaches may complement rather than substitute each other because they identify distinct mechanisms of PV to LA reconnections. Unlike much of the prior literature regarding adenosine and pace-capture testing after PVI, the current study is also notable in that all procedures used contact-force sensing catheters. Ablation with contact-force sensing catheters has been shown to lead to more transmural lesions and reduction in the incidence of DC.[13] Clearly, this study demonstrates that even with optimized tissue contact reconnections along the PVI line can still exist.

What do these results mean for those of us who perform AF ablations? Are both tests necessary in every patient? The study by Zeng et al cannot address the latter question definitively as it did not include a randomized control arm with a single test or no testing. However, the impressive survival free of AF recurrence of about 90% at 1 year in both study groups is enviable. This long-term success rate is higher than that of the historic control group and that reported in the seminal contact-force sensing catheter trials. Whether this is attributable to the combination of pace-capture plus adenosine testing approaches or other factors is unclear. It is possible that the additional ablation dictated by the combined approaches resulted in more durable PVIs or even collateral elimination of other targets such as ganglionated plexi and rotors, hence leading to improved efficacy. The current study lacks information from redo procedures, thus it is unclear what the mechanisms of recurrences may be. Nevertheless, employing both approaches to assess the quality of the PVI may be justifiable in some patients, including, those with difficult first-pass PVI, multiple areas with small impedance drops during PVI, patients undergoing redo procedures with demonstrated PVI reconnections, among others. The results of the current study indicate that the order in which the two tests are applied is less important, though it should be cautioned that the small sample size and number of clinical events likely render the analysis underpowered to detect differences between the two groups.

Despite the encouraging clinical results with the combined approach, safety and cost considerations should also be considered. Zeng et al reported no significant difference between the study groups and the historical control group in terms of ablation and fluoroscopy time, but combining adenosine and pace-capture testing in the real world has the potential to prolong procedures significantly and further increase ablation-related risks.

The authors should be congratulated for conceiving and executing an elaborate protocol aimed at tackling a complex problem. The thought-provoking results should rekindle interest in the topic and future studies incorporating contemporary and evolving catheter technologies will be welcome. Randomized trials comparing the combined approaches vs each one individually and vs no testing controls can have a significant impact and potential to inform practice.