Organization of atrial fibrillation using a pure sodium channel blocker: Implications of rotor ablation therapy

Abstract Background Rotors are the source of atrial fibrillation (AF). However, the ablation of rotors for persistent AF is challenging. The purpose of this study was to identify the dominant rotor by accelerating the organization of AF using a sodium channel blocker and detecting the rotor's preferential area that governs AF. Methods Overall, 30 consecutive patients with persistent AF who underwent pulmonary vein isolation and still sustained AF were enrolled. Pilsicainide 50 mg was administered. An online real‐time phase mapping system (ExTRa Mapping™) was used to identify the meandering rotors and multiple wavelets in 11 left atrial segments. The time ratio of non‐passive activation (%NP) was evaluated as the frequency of rotor activity in each segment. Results Conduction velocity became slower—from 0.46 ± 0.14 to 0.35 ± 0.14 mm/ms (p = .004)—and the rotational period of the rotor was significantly prolonged—156 ± 21 to 193 ± 28 ms/cycle (p < .001). AF cycle length was prolonged from 169 ± 19 to 223 ± 29 ms (p < .001). A decrease in %NP was observed in seven segments. Additionally, 14 patients had at least one complete passive activation area. Of them, the use of high %NP area ablation resulted in atrial tachycardia and sinus rhythm in two patients each. Conclusions A sodium channel blocker organized persistent AF. In selective patients with a wide organized area, high %NP area ablation could convert AF into atrial tachycardia or terminate AF.


| INTRODUC TI ON
Even in the computational mapping era, 1 mechanism-based ablation for persistent atrial fibrillation (AF) is challenging. [2][3][4][5][6] Though the reason for sustained AF is not completely clear, the concept that rotors are the source of sustained AF is widely supported based on basic and clinical research in recent years. 3,7,8 The self-existing spiral wave reentry-a rotor-generates succeeding wavefronts that propagate rapidly in various directions and result in fragmentation by anatomical or functional obstacles, which generate new wavelets; of these, some result in new rotors. 9 Therefore, detecting rotor activity, multiple wavelets activity, and distinguishing these activities from passive activity is a clue to mechanism-based AF ablation. 7,10 In contrast, it is difficult to distinguish "dominant rotors" from "subordinate rotors" during ongoing fibrillatory conduction using a real-time mapping system. 11,12 If a wide area of the atrium is wellorganized without fragmentation and the rotor activity is limited to a small area, the AF may be terminated by ablating the small area and, consequently, organizing AF to atrial tachycardia. [13][14][15][16] Furthermore, accelerating the organization of AF using a sodium channel blocker will help map them easily without considering wavefront fragmentation that becomes subordinate bystander rotors.
We hypothesized that the dominant rotor activity can be identified by accelerating the organization of AF using a sodium channel blocker and easily detecting the rotor area that governs the AF.
ExTRa Mapping, a novel online real-time phase mapping system, is now commonly used in Japan to detect rotor activity and multiple wavelets. 17,18 2 | MATERIAL S AND ME THODS This observational prospective study was conducted between August 2020 and December 2021. We enrolled patients with persistent AF who underwent pulmonary vein (PV) isolation within 2 hours of entering the catheterization laboratory without subsequent termination of AF. All patients were refractory to drug therapies before the ablation procedure. Patients who were on amiodarone, had a history of sinus bradycardia or moderate chronic kidney disease (Creatinine >1.5 mg/dL), or were older than 85 years of age were excluded from this study. Anti-arrhythmic drugs were discontinued before the procedure in all patients, while direct oral anticoagulants or warfarin were continued. General anesthesia and control of anti-coagulation used in the procedures have been described previously. 19 Before starting this study, we performed rotor ablation as additional ablation for paroxysmal and persistent AF patients using ExTRa Mapping between Aug 2019 and July 2020. During this period, 20 patients exhibited persistent AF that was resistant to repetitive ablation. The outcome of these 20 patients is shown Supporting Information Figure S1. In this preliminary observation, most patients (14/20) had already undergone pulmonary vein isolation (PVI) and roofline ablation by cryoballoon or PVI and posterior wall isolation by radiofrequency (RF) ablation. We used these preliminary observation groups as the control group to confirm the clinical impact of this study.

| Ablation procedure
A 20-pole catheter (BeeAT, Japan Lifeline Shinagawa, Tokyo Japan) was inserted via the internal jugular vein and positioned into the distal coronary sinus along the right atrial lateral wall. After insertion of the 8-Fr sheath into the left atrium (LA), it was exchanged for an 8.5-Fr deflectable sheath (Agilis™ NxT, Abbott) over the guidewire. NavX mapping system (EnSite Precision™, Abbott) or CARTO mapping system (CARTO® 3, Biosense Webster, Diamond Bar, CA, USA) were used to obtain the voltage map and RF ablation.
After an electro-anatomical map during the AF was obtained, extensive encircling PV isolation was performed according to indexed ablation. On the NavX system, RF energy was delivered using a 4-mm irrigated-tip catheter (TactiCath™ SE, Abbott) with the following parameters: energy, 35 W; contact force, 5-20 g; targeting lesion size index, 4.5-5. On the CARTO system, using a similar 4-mm irrigated-tip catheter (THERMOCOOL SMARTTOUCH™ SF, Biosense Webster), the parameters were the following: energy, 40 W; contact force, 5-20 g; ablation index, 400-450. Wide area ablation was confirmed based on the elimination of bipolar signals on the voltage map (<0.1 mV) or the detection of a dissociated signal from the isolated area.

| ExTRa mapping
ExTRa Mapping™ is an online real-time phase mapping system, 17 which is classified as contact mapping and is well-validated in optical mapping. 18 In brief, 32 real bipolar signals and nine virtual bipolar signals were recorded using a 20-pole spiral-shaped deflectable catheter (Reflexion HD; Abbott). The concept of this phase map is based on the restitution property of human atrial action potentials during non-paroxysmal AF. It is closely related to the preceding diastolic interval, 20,21 and we can chart a phase map from the local bipolar signals. Based on this theory, action potentials were simulated from bipolar signal timings, 17,18 and 5-s spatio-temporal phase map (approximate region of interest, 2.5 cm in diameter; 4 ms per frame) was automatically created using wavefronts and phase singularities. The shadow of Reflexion HD was projected on the left atrial electro-anatomical map, and the acquisition of signal was ob-  Figure 1B). The catheter was moved to each segment so that most of the real bipolar signals exceeded 0.05 mV. After the catheter was stabilized, the shadow of the catheter was confirmed in 3-dimensional geometry. More than 5 areas were sampled in each patient. According to the number of phase singularities, meandering rotors, which are defined as within 2 phase singularities, and multiple wavelets, defined as more than 2 phase singularities, were automatically classified. The periods of meandering rotors and multiple wavelets were summed respectively and indicated on the phase map. These meandering rotors and multiple wavelets were defined as non-passive activation. The value of "non-passively activated ratio" (%NP) was automatically calculated as the ratio of the non-passive activation period to the recording time.

| Reproducibility of ExTRa mapping
Rotors usually meander, and phase singularities drift out of the region of interest (i.e., out of the spiral catheter) within a short period. Before the present study, we repeated the measurement of %NP 10 times in 6 patients in various areas (Supporting Information   Table S1). In this series, a Shapiro-Wilk test indicated that most data were consistent with a normal distribution and that the moving average for the 15 consecutive seconds was enough to represent the mean %NP value. Therefore, we repeated the measurement of %NP thrice (15 s) in each area and used the mean value. In this series, no area showed %NP 0% repetitively. In contrast, atrial tachycardia always shows 0% %NP because all waves are planar (not shown). Thus, we defined newly developed repetitive 0% % NP as AF organization.

| Administration of pilsicainide and measurement of electrophysiological data
Pilsicainide was administered at a rate of 25 mg/min for a total dose of 50 mg over 2 min. Electrophysiological data were collected before and 5 min after pilsicainide was administered. The mean duration of the AF cycle was calculated from the coronary sinus electrogram using manual annotation for 10 consecutive signals. The conduction velocity (CV) during AF was measured on the maps. CV is greatly influenced by the wavefront curvature, that is, as the curvature becomes steep, the CV decreases. We chose a low %NP area both before and after pilsicainide administration in an identical area and selected a planar wavefront that propagated from one edge to the opposite edge with the least rotation and defined this duration required as the wave crossing time. CV was calculated as the diameter of the catheter divided by the wave crossing time (Figure 2A).
A rotor is a spiral wave of excitation that rotates around a phase singularity for one or more cycles. 9,14 The rotational period was also determined on the maps ( Figure 2B). In each patient, two identical areas with relatively high %NP were chosen and phase singularity meandering near the center of the region of interest was carefully selected for evaluation. The period of wavefront rotation around the phase singularity (i.e., rotor) in one cycle was calculated by the frame count.
AF organization is characterized by a reduced number of wavefronts in comparison with the original AF and is detected as reduced wavelets or decreased phase singularities. If %NP approaches 0% area after pilsicainide administration, we classified the case in the AF organization group ( Figure 1A).
The dominant frequency (DF) was obtained at the same time in the same area of ExTRa mapping using Reflexion HD catheter. All signals were collected for 5 s at a sampling rate of 1000 Hz and analyzed using software embedded in the polygraph (RMC 5000 Nihon Kohden Co. Tokyo, Japan). Thirty-two points of DF values and regularity indices were shown simultaneously at one area, and the highest DF value with a regularity index ≥0.2 represented the DF value of the area. The DF gradient of each patient was defined as the maximal value-minimal value of the highest DF among 11 areas.

| Rotor ablation in organized AF
If the rotational activity is observed only in one or two areas and other large areas are low %NP areas, the rotational activity area (high %NP area) is regarded as the source of AF controlling the other areas. 13 Therefore, we performed a high %NP area ablation in only  the organized group ( Figure 1A). We defined %NP > 50% as high %NP area. 17

| Study follow-up
Follow-up appointments were scheduled at 1, 3, and 6 months after the procedure. At each visit, we collected data regarding the medical history and performed a detailed physical examination including a 12-lead electrocardiogram (ECG). During the follow-up, the patients were free to visit our hospital at any time in case of arrhythmia symptoms, at which time a 12-lead ECG was recorded. A Holter ECG was obtained at 6 months post-procedure or at other time points during symptoms suggestive of a recurrence of arrhythmia.

| Statistical analysis
Data are expressed as mean ± standard deviation or median (25th-75th percentile). Comparisons between intra-patient data were performed using the Wilcoxon signed-rank test or χ 2 analysis. Comparisons between patient groups were performed using the Kruskal-Wallis test.
Arrhythmia-free survival curves for each group are presented as Kaplan-Meier plots, and time-to-event analysis was performed using the log-rank test. Dates were censored in April 2022. Patients who were lost to follow-up were censored at the date of the last visit.

F I G U R E 3 (A)
A 65-year-old male patient with wide area organization before and after pilsicainide administration. Left pulmonary vein is the common trunk. %NP value of septal, inferior, and lower posterior areas reduced drastically after pilsicainide administration, while those of anterior and lateral sites remained high. (B) Ablation point of the same patient and intracardiac electrogram. Two high %NP areas were targeted. During the lateral site encircling ablation, atrial fibrillation was converted into atrial tachycardia (asterisk). Red dot indicates the converted point. CS, coronary sinus; DF, dominant frequency. CS1-2 is located in the distal CS and CS7-8 is located in the proximal CS.

| Patient characteristics and procedural results
We recruited 30 consecutive patients in the present study. The baseline patient characteristics are summarized in Table 1. Three patients were excluded from the analysis due to termination of AF after PV isolation (n = 1) and attenuation of local signal (<0.05 mV) after pilsicainide administration and unreliable ExTRa Mapping data (n = 2). Fourteen patients had approximately 0% %NP at least in one area and were classified in the organized group. Thirteen patients showed no organization (sustained group). One of the patients in the sustained group developed transient atrial tachycardia but degenerated into AF in a short period. In the sustained group, sinus rhythm was achieved in all patients using cardioversion at the end of the procedure, and no ExTRa Mapping-guided ablation was performed in this group.

| Effect of pilsicainide on persistent AF
The effects of pilsicainide on all persistent AF are summarized in Table 2.
The AF cycle length was prolonged from 169 ± 18 to 223 ± 29 ms (p < .001). The CV slowed from 0.46 ± 0.13 to 0.35 ± 0.15 mm/ms (p = .004), The rotational period of the rotor was prolonged significantly from 156 ± 21 to 192 ± 28 ms/cycle (p < .001). There was no area with an increase in the %NP. The %NP derived from the meandering rotors decreased significantly in 3/11 areas, whereas the %NP derived from the multiple wavelets decreased in 9/11 areas. Total %NP value decreased in 7/11 areas, and most of them (6/7) were consistent with the decreased area as per the multiple wavelets.
The mapping data of the organized and sustained groups are summarized in Table 3. At baseline, the AF cycle length, CV, and rotational period were not different between the groups. The AF cycle length was prolonged significantly in both groups (sustained group, 172 ± 22 to 215 ± 26 ms; organized group, 166 ± 14 to 229 ± 31 ms).

F I G U R E 3 (Continued)
The CV in the organized group decreased significantly with pilsic-

| Change in high %NP
High %NP areas were observed in 19 patients (70%, not shown in the table). Of these, the high %NP areas changed in 5 patients and The number of high %NP areas decreased from 1.6 ± 1.6 to 0.9 ± 1.0 (p = .029) ( Table 2).

| Change in highest DF
The highest DF in each area at baseline was not significantly different between the organized group and sustained group except for the septal area. The highest DF markedly decreased in every area after pilsicainide administration in both groups (Supporting Information   Table S2).
At baseline, the DF gradient was slightly higher in the organized group than in the sustained group (1.1 ± 0.9 Hz, vs. 0.9 ± 0.4 Hz p = .07) but the difference became non-significant between two groups (organized 1.0 ± 0.9 Hz, sustained 0.9 ± 0.2 Hz p = .15). In this study, 11 patients received rotor ablation. The ablation area (which is a high %NP area after pilsicainide administration) was coincident with the highest DF area in three patients (3/11, 27%) before pilsicainnide administration and in three patients (3/11, 27%) after pilsicainide administration.

| Outcome of ExTRa mapping-guided ablation
In 3 patients in the organized group, AF organized into atrial tachycardia before rotor ablation. Atrial tachycardia was confirmed by regular intracardiac signals and a %NP value of approximately 0% in any area. In these patients, sinus rhythm was achieved using linear ablation. The remaining 11 patients received rotor ablation. In 2 patients, AF organized to atrial tachycardia after rotor ablation ( Figure 3), which was mapped by local activation timing and returned to sinus rhythm using linear ablation. In two patients, AF returned to sinus rhythm after rotor ablation ( Figure 4). Therefore, sinus rhythm was returned in 4/14 (29%) patients in the organized group using ExTRa Mapping-guided catheter ablation. Overall termination of AF or atrial tachycardia using catheter ablation was achieved in seven patients in the organized group.
The number of organized areas in these seven patients was significantly higher than that in the non-termination patients (4.9 ± 2.6 areas vs. 1.0 ± 1.2 areas, p < .01).
Recurrence of atrial arrhythmia beyond the 1-month blanking pe-  Figure 5A). To confirm the effects of the termination of AF or atrial tachycardia using ablation, we compared the groups with and without AF termination. The 9-month arrhythmia-free survival was 100% in the group with termination and 49.9% in the group without termination (log-rank chi-squared, 4.38, p = .04; Figure 5B).

| DISCUSS ION
This is the first study to evaluate the effects of a pure sodium channel blocker using a contact phase map in human persistent AF. The major finding of this study is that pilsicainide decreased the CV, slowed the rotational period of rotors and prolonged the AF cycle length. Another finding is that organization of AF into atrial tachycardia (or atrial flutter) is closely related to a reduction in rotor activity in several areas. Lastly, ablation of high rotor activity observed after pilsicainide can organize AF into atrial tachycardia or sinus rhythm in some cases.
In an induced canine AF model, pilsicainide decreased CV and increased the effective refractory period. [22][23][24] Although we could not measure the refractory period, CV during AF has been reported to be decreased in vitro human atrium with AF. 25 In most animal AF models, the longevity of persistence is extremely shorter than that in human persistent AF, and it is uncertain if their findings apply to long-lasting human AF. These results demonstrate that pharmacological effects can be achieved in human persistent AF with a clinical dose. Generally, an enlarged atrium includes structural remodeling because of fibrosis and electrical remodeling with various types of ionic current abnormalities. 26,27 Prolongation of AF cycle length and slowing of the rotational period observed in our series also suggest that the use of sodium channel blockers is a potential pharmacological intervention in persistent AF. However, AF cycle length or rotational period alone cannot become a surrogate for the organization of AF.
In this study, pilsicainide significantly reduced multiple wavelets in several areas. On the ExTRa Mapping system, wave dynamics manifesting more than 2 phase singularities in a region of interest is classified as multiple wavelets. The reduction of multiple wavelets implies that less phase singularities are observed in a region of interest. Theoretically, the rotor will continue to rotate even if the CV or excitability is reduced. Two reasons have been postulated for the decreased number of phase singularities in a region of interest.
One is that the rotor core size becomes large and the rotor (with phase singularity) meanders out of the region of interest. Pilsicainide is reported to make the original rotor circuit size large and slow the rotational period. 14,22 If the tissue excitability decreases, a steep curvature of the wave will not excite the adjacent tissue and the ability to propagate is consequently decreased; therefore, the wave tip should rotate around a large core. 15 The second postulated mechanism is that the generation of a new rotor is inhibited by pilsicainide.
Pilsicainide shortens the wavelength and widens the excitable gap. 23 Consequently, the wavefront becomes difficult to meet the wave tail and the wavefront-wavetail interaction will occur less.
In our series, the termination of AF using rotor ablation was still challenging. We could terminate AF in 50% (7 /14) 13,32 ; however, such an ideal situation was rarely observed in our series. In intention to treat analysis, the 9-month arrhythmia-free rate was higher in the organized group, but the difference was not statistically significant between the groups. In contrast, once termination was achieved, most patients developed no recurrence of atrial tachyarrhythmia. Regular atrial tachycardia is sometimes observed after PV isolation, some of which are macro-reentry tachycardia. 33 Treating pilsicainide-induced atrial tachycardia may mimic atrial tachycardia following PV isolation.
The clinical benefit of administering additional pilsicainide is that it aids in determining, which patients are likely to change to AT from persistent AF or to organize AF into AT with a limited ablation area, which leads to termination of arrhythmia. In contrast, the use of anti-arrhythmics may mask the non-PV foci. A pilsicainide challenge might be useful, especially in patients who could not be induced or in whom it is difficult to determine non-PV foci.
The DF gradient was slightly higher in the organized group. Some patients in the organized group showed a high DF area before pilsicainide administration (Figure 4), which implies that a dominant rotor might exist in some cases in the organized group. However, the ablation area did not always correlate with the highest %NP area. Since the rotor meanders, the location of the highest NP area may vary.
Accordingly, both indices should be used complementarily to reveal the AF maintenance mechanism. 34

| Study limitations
This study was a single-center retrospective observational analysis and included a small number of patients. Furthermore, the %NP value after rotor ablation was not clear. We tried to confirm the activity after encircling the NP area; however, reliable ExTRa Mapping was often difficult due to low signal amplitude. It was difficult to map all the left atrial areas in patients with a small atrium. Some electrodes of the spiral catheter did not contact the tissue sufficient enough to record the local signal or the catheter deformed in the atrium, especially in the low septal area and right inferior area. Additionally, buprenorphine can inhibit the sodium current, and propofol can inhibit sodium, calcium, and potassium currents and may interact with sodium channel blockers, although the concentrations needed for inhibition is much higher than concentrations that are clinically achievable. 35 Finally, low-voltage areas were not determined during sinus rhythm at baseline; thus, the role of low-voltage areas in rotor initiation or maintenance is unclear.

| CON CLUS IONS
We investigated the effects of a sodium channel blocker in persistent AF. The reduction of the rotors was closely related to the organization of AF. Intensive organization is achieved in 50% of cases at a clinical dose. Instantaneous termination of AF using rotor ablation was achieved in only 28% of patients with well-organized AF.
However, termination of AF resulted in improved outcomes.

CO N FLI C T O F I NTER E S T S TATEM ENT
The authors declare no conflict of interests for this article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author (Tadafumi Nanbu), upon reasonable request.

E TH I C S A PPROVA L
The study protocol was approved by the institutional review board of our hospital.

PATI ENT CO N S ENT
Patient consent was obtained with written form and consent for inclusion was obtained by an opt-out method.