Electrophysiological findings during atrial fibrillation reablation: Extending from pulmonary vein reconnection to sequential bipolar voltage map information

Left atrial substrate modification targeting low voltage zones (LVZ) is an ablation strategy that—in addition to pulmonary vein (PV) isolation—tries to eliminate arrhythmogenic mechanisms harbored in such tissue. Electrophysiological findings at reablation include (a) PV reconnection, (b) reconnection over previous substrate ablation, and (c) de‐novo LVZ.

areas are frequent findings besides classical PV reconnection-without a clear leading cause for recurrences. These findings impact reablation strategies as well as the strategic focus during index procedures. transformation is thought to be a progressive process that increases over time and is associated with age and comorbidity such as hypertension, diabetes, and heart failure. 4 Individualized LA substrate modification (LASM) targeting LVZ in addition to pulmonary vein isolation (PVI) is an ablation strategy that tries to eliminate arrhythmogenic mechanisms harbored in such tissue areas. 5,6 Although this approach is superior against sole PVI, recurrences do occur. 7,8 Electrophysiological findings at reablation include (a) pulmonary vein (PV) reconnection, (b) reconduction over previously ablated LVZ, and (c) spontaneous de-novo LVZ. The overall contribution of each of these three arrhythmia-generating entities to the actual recurrence of AF is unclear. That limits strategies for reablation procedures but also impacts the strategic focus during the index ablation.
The aims of the present study were (a) to systemically analyze all electrophysiological findings at reablation, (b) to compare voltage map characteristics between the index and re-do procedures, and (c) to explore long-term outcomes according to those electrophysiological characteristics.

| Subjects
We included consecutive patients with highly symptomatic AF who underwent consecutive index (1st) and re-do (2nd) HD-BVM guided AF ablations at Heart Center Dresden between January 2014 and January 2017. Patients were followed intensively with continuous monitoring as reported previously. 8 All participants had (a) paroxysmal or persistent symptomatic AF, (b) previously ineffective antiarrhythmic drug therapy (at least one antiarrhythmic drug), and (c) LA diameter of less than 60 mm (transthoracic echocardiography, parasternal long axis). We excluded patients with (a) the previous ablation affecting the LA, (b) previous thoracic surgery, or (c) previous thoracic radiotherapy or chemotherapy.
The study was approved by the institutional ethical review board (EK 284092012) and conforms to the principles outlined in the Declaration of Helsinki. All data were collected, managed and analyzed at Heart Centre Dresden and the Steinbeis Research Institute, Electrophysiology and Cardiac Devices.
All mapping points were taken in sinus rhythm. For each mapping point, stable contact between the local atrial tissue and each pair of electrodes of the circular mapping catheter was required. Extra care was taken while collecting voltage points on the border of LVZ. 4 Sufficient quality of the acquired voltage points was verified by the following criteria 4 : (a) P-wave morphology, (b) coronary sinus (CS) activation sequence, (c) cycle length, (d) local bipolar electrogram morphology, (e) local activation time, and (f) reproducibility.

| Bipolar voltage map analysis
Quantitative analysis of voltage maps was performed as described previously. 4 The LA was divided into five regions, that is, septum, anterior, posterior, inferior and lateral walls, omitting the LA appendage. Each region was further divided into nine equally sized blocks. The median voltage value within each block was recorded and used for further offline analysis. Contiguous areas of bipolar voltage less than 0.5 mV were considered as an area of LVZ. Within each region size and localization of LVZ were analyzed.
Eventually, the LVZs found at reablation were further categorized as (a) spontaneous de-novo LVZ, and (b) ablation associated LVZ.
The bipolar voltage maps at the index procedures served as a reference and compared to voltage maps at re-do procedures. The spontaneous de-novo LVZ at re-do procedures was defined as a contiguous area with bipolar voltage less than 0.5 mV, which located in a predefined segment without previous substrate modifications.
The enlarged LVZs in the segment, which contained previous substrate modification, were excluded from further analysis. Ablation endpoints were (a) lack of local pace capture and (b) bidirectional conduction block over linear lesions. 4,5,8 A standard induction protocol was carried out at the end of the procedure using a burst stimulation (300, 250, and 200 ms, each for 10 seconds) and a ramp stimulation (300-200 ms for 10 seconds).
Further induced regular atrial tachycardias (ATs) were also mapped and ablated. A duration of induced AF longer than 30 seconds was defined as "AF inducible." However, no further ablations were conducted if PVI or PVI + LASM due to LVZ were achieved.
Ablation of the right atrial isthmus was only performed in case of documented/induced typical atrial flutter. Premature atrial contractions (PACs) were mapped and ablated after the achievement of re-PVI and/or LASM if necessary.

| Follow-up postprocedural management
Antiarrhythmic medications were discontinued, and patients remained on ß-blocker. In the case of arrhythmia recurrences, antiarrhythmic drugs were reinitiated upon an individual decision.
Reablation for symptomatic drug-refractory recurrences of AF and AT was scheduled after at least 3 months from the index procedure.
Oral anticoagulation was continued for at least 3 months and thereafter according to CHA 2 DS 2 -VASc-Score with deviations upon patient and physician's discretion.
Postinterventional rhythm assessment was based on implantable device monitoring (Reveal LINQ, Medtronic, Inc) and 4-day Holter every 3 months. 8

| Data management and statistical analysis
Continuous variables were tested for normal distribution using the Shapiro-Wilk test. Data with normal distribution are presented as mean ± SD and data without normal distribution are presented as the median and interquartile range. Categorical variables are expressed with a number and percentage of patients.
Differences between continuous normally distributed data were tested with the Student t test and differences between categorical data were tested with a χ 2 test. A P < .05 was considered significant.
The data were collected and managed by the investigators (MK and YH) and all statistics were performed using Stata version 12 (Stata Corporation, College Station, TX).

| Patient characteristics
During the study period, 1758 patients underwent HD-BVM guided AF ablation at our institution. Of these, 129 patients underwent both index and re-do procedures. In 16 patients, the HD-BVM was insufficient at one of the procedures and a total of 113 patients was included in the study (mean age 67 ± 9, 59% male, 40% with paroxysmal AF). Meantime between the index and re-do procedures was 16 ± 10 months. Baseline characteristics of patients are shown in Table 1. Patients with spontaneous de-novo LVZ after the index procedure were significantly older (65 ± 10 vs 70 ± 6 years, P = .009) and had a lower left ventricular ejection fraction (LVEF) (56% ± 10% vs 50% ± 14%, P = .03) as compared with patients without spontaneous de-novo LVZ, otherwise baseline characteristics were comparable. Among the study patients, 82% (n = 93) were followed with continuous monitoring from the implantable device and 18% (n = 20) with periodic 4-day Holter (Figure 1).
3.2 | Characteristics of documented atrial arrhythmias between the index and re-do procedures and ablated regular ATs at the re-do procedure Out of 113 patients, 33 (29%) paroxysmal AF, 66 (58%) persistent AF, and 14 (12%) regular AT were documented between the index and re-do procedures. There were no significant differences in the regular ATs recurrences between patients with and without spontaneous de-novo LVZ (n = 6/36 vs n = 8/77, P = .35). However, in the patients with spontaneous de-novo LVZ at re-do procedures, the incidence of persistent AF was significantly greater than in patients without de-novo LVZ (n = 28/36 vs n = 38/77, P = .001; Abbreviations: AF, atrial fibrillation; AFL, atrial flutter; LVZ, low voltage zones. A total of 7 patients with a regular AT at the beginning of the re-do procedures and further 5 were fully mapped and ablated successfully, including 2 perimitral ATs, 2 roof-dependent ATs, and 1 localized reentrant AT originating at LA anterior wall. The rest 2 ATs were terminated during mapping. At the end of the re-do procedures, 3 perimitral ATs originating at LA were induced due to acute re-conductions over anterior ablation lines (large LVZ-LA anterior wall) and further bidirectional blocks were re-achieved.  In patients (n = 51) with no LVZ at the index and re-do procedures, overall bipolar LA voltage declined significantly from the index to re-do procedures (2.6 ± 0.9 vs 2.2 ± 0.7 mV, P = .013). On a regional level, a decline in voltage was only seen in the anterior and septal regions (Table 3).

| What makes AF recurrences?
PV electrophysiology has recognized a cornerstone of the AF pathophysiology. 9 Therefore, PV isolation is an accepted ablation strategy and endpoint. 10 Subsequently PV reconnection has been identified as a frequent finding during reablation and is considered the main reason for AF recurrences. 11,12 Our data are in line with these findings of PV-reconnection. In the attempt to improve the long-term efficacy of AF ablation a focus has been put on improving durable PV isolation.
Outcome data in large-scale clinical trials, however, are conflicting. 13,14 Therefore, the question needs to be raised, whether alternative arrhythmia-generating mechanisms are involved in the development of AF recurrences.
Under that light, the evolution in the understanding of AF pathophysiology has to be considered. 15,16 Today structural remodeling in histoarchitecture is recognized to alter cellular coupling with subsequent conduction impairment. 1,17,18 Affected myocardial regions are substrates for reentry-a keystone in AF development. 18,19 Voltage mapping during sinus rhythm provides electrophysiological surrogates for such tissue pathologies. 15 Our study used sequential bipolar voltage map information to describe the evolution of these tissue and conduction abnormalities over longer time periods, which for the first time allows insights into mechanisms for AF recurrences beyond PV reconnection.

| Disease progression
We detected a significant amount of spontaneous de-novo low voltage substrates during the re-do procedure. That finding is especially impressive for the group of patients with no low voltage at the index procedure. One-third of those patients showed de-novo LVZ during re-do procedures.

| Clinical outcome
Given the intense follow-up with continuous device monitoring the reported clinical outcome after the re-do procedure was acceptable.
More than half of the patients did not have a single arrhythmia episode within the first 12 months.
In univariate analysis reconnection of PVs, reconduction over previous lines, presence of LVZ, type of AF, and various clinical characteristics were not independently associated with the success of reablation. These findings suggest that none of these factors alone can be considered as the main reason for recurrences. They may also hint to other so far unrecognized arrhythmia mechanisms-not included in the analysis-to play a role for recurrences. In that context, it is interesting to mention recent data on exclusive epicardial arrhythmia substrates responsible for AF in re-do patients. 24

| Limitations
The present study could suffer from the inherent limitations of a nonrandomized study including selection bias, confounding and lack of control group. The sample size was moderate, and a minor proportion of the patients was excluded due to insufficient voltage maps from one of the two procedures. Furthermore, we did not perform any routine imaging or histology to support the presence of fibrofatty tissue underlying LVZs. On the other hand, this is the first study reporting findings from sequential voltage maps over long time periods in patients undergoing ablation for AF.
The fact, that we only mapped patients with AF recurrences limits our ability to judge the individual contribution of the various electrophysiological abnormalities to the actual arrhythmia recurrence. The fact that PACs were mapped and ablated after Re-PVI

| CONCLUSION
In patients undergoing reablation we could show that reconduction over previous substrate ablation as well as the development of new low voltage areas are frequent findings besides classical PV reconnectionwithout a clear leading cause for recurrences. These findings impact reablation strategies as well as the strategic focus during index procedures.