Author + information
- Received December 23, 2015
- Revision received May 17, 2016
- Accepted May 26, 2016
- Published online November 1, 2016.
- Samuel H. Baldinger, MD∗ (, )
- Jason S. Chinitz, MD,
- Sunil Kapur, MD,
- Saurabh Kumar, BSc(Med)/MBBS, PhD,
- Chirag R. Barbhaiya, MD,
- Akira Fujii, MD,
- Jorge Romero, MD,
- Laurence M. Epstein, MD,
- Roy John, MD, PhD,
- Usha B. Tedrow, MD, MSc,
- William G. Stevenson, MD and
- Gregory F. Michaud, MD
- Cardiac Arrhythmia Center, Cardiovascular Division, Brigham and Women’s Hospital, Boston, Massachusetts
- ↵∗Reprint requests and correspondence:
Dr. Samuel H. Baldinger, Cardiac Arrhythmia Center, Cardiovascular Division, Brigham and Women’s Hospital, 75 Francis Street, Boston, Massachusetts 02115.
Objectives The aim of this study was to categorize arrhythmia mechanisms and to summarize ablation strategies in patients with persistent pulmonary vein isolation (PVI) at the time of redo procedures.
Background Persistent PVI is more frequently seen in patients undergoing redo procedures for recurrent atrial arrhythmias after catheter ablation for atrial fibrillation (AF).
Methods Consecutive patients who underwent their first AF ablation procedures at Brigham and Women’s Hospital were screened and included if they had persistent isolation of all pulmonary veins at the time of redo procedures.
Results Of 300 consecutive patients undergoing first AF ablation procedures, redo procedures were performed in 63 (21%), and 26 patients (9%) had persistent PVI. Of those, 11 had recurred with AF and 15 with organized atrial tachycardia (AT). During the index procedure, linear ablation was performed in 46% of patients with recurrent AF and 93% with recurrent organized AT (p = 0.020). At the time of last follow-up, 2 of 10 patients (20%) in the AF group and 10 of 15 patients (67%) in AT group were in sinus rhythm, without class I or III antiarrhythmic drugs (p = 0.022).
Conclusions Patients with recurrence of atrial arrhythmia despite persistent PVI frequently present with organized AT. Linear ablation during the index procedure is associated with recurrence of organized AT. Recurrence rates after redo procedures were higher if patients had recurrent AF after the index procedure, and these patients often presented with AF again. Patients with recurrent AF despite persistent PVI may represent a population with lower success rates of catheter ablation.
Recurrence of atrial arrhythmias following catheter ablation procedures for paroxysmal or persistent atrial fibrillation (AF) is often associated with recovery of conduction into previously isolated pulmonary veins (PVs) (1,2). However, at our center, persistent PV isolation (PVI) has been encountered more frequently in patients undergoing redo catheter ablation procedures for recurrent atrial arrhythmias, perhaps reflecting our focus on producing more durable lesions. Current guidelines advocate greater use of substrate modification techniques in such situations (3), but the ablation approach is not well defined. In this retrospective analysis of patients with persistent PVI after initial ablation for AF, we sought to identify patient or ablation characteristics that determine the type of atrial arrhythmia recurrence at the time of a redo procedure and to determine an appropriate ablation strategy in such patients.
In total, 300 consecutive patients (mean age 60 ± 11 years, 57% men, 41% with persistent AF) who underwent their first AF ablation (index procedure) for paroxysmal or persistent AF at Brigham and Women’s Hospital performed by G.F.M., J.S.C., S.K., and C.R.B. between January 2011 and June 2014 were screened. Of these 300 patients, 160 (53%) underwent PVI only, 132 (44%) had a roof line, and 71 (24%) had a mitral line; in 6 patients (2%), the superior vena cava was isolated; and 102 patients (32%) had complex fractionated abnormal electrogram (CFAE) ablation.
Patients who subsequently underwent redo procedures for recurrence of any atrial arrhythmia were analyzed and included if they had persistent isolation of all PVs at the time of the first redo procedure.
Written informed consent for the procedure was provided by all patients, and procedures were performed according to protocols approved by the Brigham and Women’s Hospital Human Subject Protection Committee.
Patient categories and comparison
Patients were grouped depending on the type of documented recurrent arrhythmia. The AF group included patients who presented with recurrent AF. The atrial tachycardia (AT) group included patients who presented with organized AT (macro-re-entrant or focal/micro-re-entrant) (4).
Initial clinical presentation and ablation strategy during the index procedure as well as electrophysiological findings and ablation strategy during the redo procedure were compared between the 2 groups.
Antiarrhythmic drugs were discontinued ≥5 half-lives prior to ablation, except for amiodarone. All patients received oral anticoagulant agents for at least 1 month prior to the procedure. Transesophageal echocardiography, when deemed appropriate, was performed within 5 days of the procedure. Vitamin K antagonists were continued uninterrupted; the last dose of direct anticoagulant agent was given the day before the procedure and resumed the night of the procedure.
Radiofrequency ablation procedure
As previously described (5), PVI was performed under general anesthesia, and heparin was administered to maintain an activated clotting time of >350 s. Intracardiac ultrasound was used for visualizing transseptal punctures and improving catheter contact. A duodecapolar catheter (LiveWire, St. Jude Medical, St. Paul, Minnesota) was looped into the coronary sinus and onto the lateral wall of the right atrium. Deflectable (Agilis, St. Jude Medical) and fixed-curve (SR0, St. Jude Medical) transseptal vascular sheaths were inserted into the left atrium (LA) using separate transseptal punctures. Using the CARTO 3 (Biosense Webster, Diamond Bar, California) or the NavX (St. Jude Medical) electroanatomic mapping system, a detailed anatomic map was generated for left atrial anatomy by using a circular or a multipolar, multispline diagnostic catheter (Inquiry AFocus II [St. Jude Medical], Lasso NAV or Pentaray NAV [Biosense Webster]). A 3-dimensional model of the LA reconstructed from prior computer tomographic or magnetic resonance imaging was imported into the mapping system if available. Ablation was performed using a 3.5-mm-tip catheter (Celsius Thermo-Cool [St. Jude Medical], ThermoCool SF or ThermoCool SmartTouch [Biosense Webster]). Radiofrequency (RF) energy (Stockert GmbH, Freiburg, Germany) was delivered in a power-controlled mode at 20 to 35 W at an irrigation rate of 17 to 30 ml/min, aiming to maintain catheter tip temperature <45°C. Goals for individual lesions included an impedance decrease of 10 Ω or greater, except when any esophageal heating occurred, as judged by an esophageal temperature probe (Vital-Temp, Vital Signs, Englewood, Colorado) that was maintained at the level of the ablation catheter tip and an average contact force >10 g if available. Additional endpoints of the procedure were electric isolation of all PVs by antral ablation without carinal lesions, anatomic completion of the RF ablation line with <5-mm gaps between the center of lesion markers, electric unexcitability of the RF ablation line (6), and lack of dormant conduction following 12 mg of intravenous adenosine. Adjunctive ablation at sites of CFAEs, defined as areas of rapid and multiphasic or continuous electrograms (7), and application of left atrial lines were performed in an attempt to terminate AF. Completeness of linear ablation was confirmed by using standard pacing maneuvers (8,9). If AF did not terminate with ablation, intravenous ibutilide (0.5 to 1.0 mg over 5 to 10 min) was given, followed by direct-current cardioversion, if unsuccessful.
A steerable sheath (FlexCath, Medtronic, Minneapolis, Minnesota) was inserted into the LA over an Amplatzer guidewire (Boston Scientific, Minneapolis, Minnesota), and a 28-mm Arctic Front or Arctic Front Advance balloon (Medtronic) was used to perform cryoablation around the antrum of all 4 PVs using intracardiac ultrasound and fluoroscopic guidance. Cooling was performed to temperatures ranging between −41°C and −51°C. At least 2 applications were delivered in the antrum of each vein. Right phrenic nerve pacing was performed using a circular catheter positioned in the superior vena cava. Following cryoapplication, the balloon was removed and a 3.5-mm catheter inserted to the LA to finalize incomplete lesions or add additional lines at the operator’s discretion.
All redo procedures were performed using point-by-point RF ablation as described earlier. Mapping maneuvers as described previously were used to diagnose the mechanism of an induced arrhythmia (10,11). PVs and all previously performed linear lesion sets were checked for conduction recovery. Conduction gaps in lines were ablated to achieve bidirectional conduction block. If patients presented in sinus rhythm (SR), arrhythmia induction was performed with burst pacing and intravenous isoproterenol. Focal or linear ablation and ablation of CFAEs in the left and right atrium were performed depending on the induced arrhythmias. Inducibility for atrial arrhythmias was reassessed at the end of the procedure with burst pacing and intravenous isoproterenol.
Patients were followed up at our center or with referring cardiologists with regular Holter electrocardiography and 12-lead electrocardiography in case of symptoms. Referring cardiologists and primary care physicians were contacted for clinical follow-up if necessary. Recurrence was defined as any documented atrial arrhythmia lasting ≥30 s.
Continuous variables are presented as mean ± SD. Categorical variables are expressed as count (percentage). Differences of continuous and categorical variables were tested for statistical significance using Student t tests and Pearson chi-square tests, respectively. Recurrence curves were plotted using the Kaplan-Meier method, and the log-rank test was used to compare the 2 groups. No blanking period was applied for Kaplan-Meier analysis. A 2-tailed p value <0.05 was considered to indicate statistical significance. Data analysis was performed using IBM SPSS Statistics for Mac version 22.0 (IBM, Armonk, New York).
Redo procedures were performed in 63 of 300 patients (21%), 297 ± 219 days following the index procedures. Persistent isolation of all PVs was demonstrated in 26 patients (9% of the 300 total patients and 41% of redo patients), forming the study population. Baseline and procedural characteristics are summarized in Table 1, and the ablation strategy at the index procedures is summarized in Figure 1. The indications for the index procedures were paroxysmal AF in 9 patients (35%) and persistent AF in 17 patients (65%). Recurrent AF was documented in 11 patients (42%; AF group) and organized AT in 15 patients (58%; AT group) leading to the redo procedures.
AF group (n = 11)
Table 2 provides case-by-case information on patients presenting with recurrent AF, and findings are summarized in Figure 2. At the time of redo procedures, 4 patients were in AF. The other 7 patients presented in SR, but AF or AT was inducible in 5 of them. In 2 patients without inducible arrhythmia, cryoablation had been performed at the index procedures, and empirical RF ablation to include more of the PV antrum was performed in the redo procedures. Of the 9 patients with inducible or spontaneous sustained arrhythmia, 8 patients converted to SR with RF application and were noninducible at the end of the procedures. In 1 patient, AF terminated by cardioversion. In 8 of 9 patients with inducible or spontaneous sustained arrhythmia, organized AT was present or induced at some time during the redo procedures, and 6 of 8 patients had more than 1 AT.
During the index procedures, left atrial linear ablation had been performed in 6 patients (55%) and CFAE ablation in 5 patients (46%). During the redo procedures, 3 patients (27%) were found to have at least 1 line with conduction recovery.
AT group (n = 15)
Table 3 provides case-by-case information on patients presenting with recurrent AT, and Figure 3 summarizes the arrhythmias in this group. Four patients presented in macro-re-entrant AT, 2 patients in locally re-entrant or focal AT, and 9 patients in SR. Of the patients in SR, AF was induced in 2 patients, organized AT in 5 patients, and no inducible sustained atrial arrhythmia could be induced in 2 patients. For the latter 2 patients, empirical ablation was performed to block lines with conduction recovery, and the encircling lesion set was extended to include more of the PV antrum in 1 patient. All present or induced arrhythmias eventually terminated with RF ablation. Any organized AT was present or induced at some time during the procedures in 13 of 15 patients (87%), and 8 patients had more than 1 AT. No atrial arrhythmia could be induced at the end of the procedure in any patient in the AT group.
During the index procedures, linear ablation had been performed in 14 patients (93%; p = 0.020 compared with the AF group), and CFAE ablation had been performed in 9 patients (60%; p = 0.462 compared with the AF group), as shown in Figure 4. One or more lines with conduction recovery were found in 9 patients (64%).
Follow-up after the redo procedure
No follow-up information was available after the redo procedure for 1 patient in the AF group. The remaining patients were followed up for 17 ± 11 months after the redo procedures. As shown in Figure 5, patients in the AT group had a lower recurrence rate after the redo procedures compared with patients in the AF group (p = 0.009). All recurrences occurred within 6 months after the redo procedures in both groups. In the AT group, 1 patient had a single episode of AT with no other atrial arrhythmias without antiarrhythmic drugs since then, and 1 patient presented with AF. In the AF group, 5 patients had additional recurrences of AF. One patient had organized atrial flutter, and 1 had documented asymptomatic runs of AT with rapid ventricular rates and was prescribed dofetilide. No third procedures had been performed in any patient from either study group at the time of last follow-up. Two of 10 patients (20%) in the AF group and 10 of 15 patients (67%) in the AT group had no arrhythmia recurrence, without class I or III antiarrhythmic drugs (p = 0.022).
In the past few years, improvements in technology, such as contact force monitoring, and advances in our appreciation and monitoring of surrogates for lesion formation such as electrogram attenuation, impedance decrease, and unexcitability of the ablation line in addition to acute conduction block have probably led to greater durability of ablation lines and improvement in single-procedure success (12–15). As a consequence, persistent PVI will be more commonly encountered during redo procedures, as was observed in 26 of 63 patients (41%) in this series. Although recurrent atrial arrhythmia after catheter ablation for AF is strongly associated with recovery of PV conduction (1,2), the ablation strategy is less clear when persistent PVI is encountered.
The main findings of this study are as follows: (1) When patients with persistent PVI present with recurrent AT, blocking incomplete lines leads to reasonable medium-term freedom from recurrent atrial arrhythmias (67%). In patients with persistent PVI and AF as the recurrent arrhythmia, this strategy, along with additional substrate ablation, is associated with poor medium-term success (20%). (2) Patients with recurrent AT and persistent PVI are more likely to have recovered conduction through a line of ablation from the index procedure compared with patients with recurrent AF.
The type of arrhythmia recurrence may be an important factor in deciding the ablation strategy in patients with persistent PVI. In this series, a majority of patients (58%) presented with organized AT, which was frequently associated with 1 or more incomplete ablation lines. After the redo procedures focused on re-establishing conduction block, the arrhythmia recurrence rate was relatively lower than for patients who presented with AF. This finding has been reported in prior studies (16). Therefore, in patients with persistent PVI who present with AT, durability of conduction block seems to be the most important factor in determining medium-term freedom from atrial arrhythmia.
Although patients with recurrent AT were more likely to have linear ablation at the index procedure than patients with recurrent AF, many patients with AF also had an initial linear ablation strategy and presented with recurrent AF even after block was re-established in lines and additional substrate was targeted. Likely, there is an inherent difference in these patient groups that cannot be measured easily by demographic or readily obtained electrophysiological parameters. Unfortunately, our data do not provide guidance as to whether some patients should undergo linear or CFAE ablation in addition to PVI at the index procedure. It is not clear that linear ablation prevents recurrence of AF. Instead, it may be that linear ablation is proarrhythmic in certain patients and PVI alone may have sufficed at the index procedure. Even though CFAE ablation was useful in organizing AF to AT during the index procedure, CFAE ablation may create slow conduction, facilitating micro- or macro-re-entrant AT, as is frequently seen in this population. From that perspective, as well as data from STAR AF 2 (Substrate and Trigger Ablation for Reduction of Atrial Fibrillation Part 2) and CHASE-AF (Catheter Ablation of Persistent Atrial Fibrillation Study) (17,18), PVI with attention to durability may be the best initial strategy for AF ablation until additional strategies have been proven in randomized controlled trials. If CFAE ablation is performed, precise documentation of the locations may help the operator find potentially critical sites for re-entry during a redo procedure.
Ablation strategy in patients with durable PVI
Most organized ATs after PVI have been shown to be re-entrant, and nearly all are related to gaps in prior ablation lines (19). We therefore first checked for conduction block across pre-existing lines if patients presented in SR and performed ablation to restore block if needed. If patients presented in atrial arrhythmia, conduction block was tested once SR had been restored. This strategy worked well in patients with recurrent AT, who were subsequently noninducible in nearly all cases.
Whenever fractionated or high-frequency electrograms could be recorded in the PV antrum but outside the encircling ablation line, the line was extended to include more of the LA. This was performed in 3 patients, of whom 1 had another recurrence after the redo procedure. Notably, 2 of these 3 patients had cryoablation in the index procedures. This may indicate that encircling ablation lines applied by cryo-balloon procedures may be too small to include relevant antral sources or substrate in some patients.
When AF was inducible, areas of CFAEs in the LA and right atrium were targeted, aiming for AF termination or organization. Organized ATs were mapped and targeted with focal or linear ablation, dependent on the underlying mechanism, aiming for termination of all induced arrhythmias. Our data suggest that linear but not CFAE ablation during the index procedure predisposes to recurrent AT in patients with persistent PVI. AF termination rates during the index procedure were not significantly different between the 2 groups. Lines with conduction recovery were frequently observed in both groups. This may indicate that these lesion sets often traverse thicker tissue, require deeper lesions, are technically challenging, and thus are more prone to conduction recovery than encircling antral PV ablation lines.
Although the patients were consecutive, the data were analyzed retrospectively, and no pre-specified ablation protocol was used during the index and redo procedures. The initial ablation approach included PVI alone or in combination with a roof line in patients with paroxysmal AF but may involve more ablation in those who failed to terminate with catheter ablation. For persistent AF, adjunctive linear ablation was performed plus ablation of CFAEs if present. This study incorporated patients referred to a tertiary care center for ablation. Results may be different for other centers or operators. No recording devices were implanted, and recurrence was assessed on the basis of symptoms, Holter electrocardiography, and 12-lead electrocardiography, which may have led to an underestimation of the recurrence rate. The study population was relatively small, which may limit the reliability of the results.
Patients with recurrence of atrial arrhythmia despite persistent PVI may present with organized ATs or AF. Patients who presented with organized ATs frequently had incomplete linear ablation from the index procedure but had a relatively low recurrence rate after the redo procedures. Patients who present with recurrent AF also had incomplete linear ablation, but recurrence rates of AF after a redo procedure remain high even when the redo ablation strategy resulted in AF termination. Patients with AF despite PVI may represent a population with more complex underlying mechanisms of arrhythmia, and standard techniques including linear and CFAE ablation do not produce good results. Some patients may not profit from further RF ablation during redo procedures when persistent PVI is documented, and future studies are required to evaluate such an approach.
COMPETENCY IN MEDICAL KNOWLEDGE: Persistent PVI may be encountered more commonly during redo procedures for recurrent atrial arrhythmias after ablation procedures for AF as better ablation techniques and technology are used during catheter ablation procedures, but incomplete linear block is often present outside the PVs. In patients with recurrent AT, completion of the ablation line(s) offers reasonable medium-term freedom from atrial arrhythmias, but the appropriate ablation strategy in patients with recurrent AF is not clear.
TRANSLATIONAL OUTLOOK: Further studies defining the optimal ablation strategy in patients with recurrent atrial arrhythmias despite persistent PVI are needed, especially for patients with recurrent AF.
Dr. Baldinger has received scholarships from the University Hospital of Bern, Switzerland, and the Swiss Foundation for Pacemakers and Electrophysiology. Dr. John has received consulting fees from St. Jude Medical. Drs. Tedrow and Michaud have received consulting fees from St. Jude Medical; and research funding from Boston Scientific and Biosense Webster. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Baldinger and Chinitz contributed equally to this work.
- Abbreviations and Acronyms
- atrial fibrillation
- atrial tachycardia
- complex fractionated abnormal electrogram
- left atrium
- pulmonary vein
- pulmonary vein isolation
- sinus rhythm
- Received December 23, 2015.
- Revision received May 17, 2016.
- Accepted May 26, 2016.
- American College of Cardiology Foundation
- Ouyang F.,
- Antz M.,
- Ernst S.,
- et al.
- Nilsson B.,
- Chen X.,
- Pehrson S.,
- Kober L.,
- Hilden J.,
- Svendsen J.H.
- Calkins H.,
- Kuck K.H.,
- Cappato R.,
- et al.
- Michaud G.F.,
- John R.
- Nademanee K.,
- McKenzie J.,
- Kosar E.,
- et al.
- Hocini M.,
- Jais P.,
- Sanders P.,
- et al.
- Jais P.,
- Hocini M.,
- Hsu L.F.,
- et al.
- Barbhaiya C.R.,
- Kumar S.,
- Ng J.,
- et al.
- Bortone A.,
- Appetiti A.,
- Bouzeman A.,
- et al.
- Steven D.,
- Sultan A.,
- Reddy V.,
- et al.
- Ammar S.,
- Hessling G.,
- Reents T.,
- et al.
- Vogler J.,
- Willems S.,
- Sultan A.,
- et al.
- Chae S.,
- Oral H.,
- Good E.,
- et al.