Author + information
- Received May 27, 2008
- Revision received June 23, 2008
- Accepted July 1, 2008
- Published online October 21, 2008.
- Ermengol Valles, MD⁎,†,
- Roger Fan, MD⁎,
- Jean François Roux, MD⁎,
- Christopher F. Liu, MD⁎,
- John D. Harding, MD⁎,
- Sandhya Dhruvakumar, MD⁎,
- Mathew D. Hutchinson, MD⁎,
- Michael Riley, MD⁎,
- Rupa Bala, MD⁎,
- Fermin C. Garcia, MD⁎,
- David Lin, MD⁎,
- Sanjay Dixit, MD⁎,
- David J. Callans, MD⁎,
- Edward P. Gerstenfeld, MD⁎ and
- Francis E. Marchlinski, MD⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Francis E. Marchlinski, 9 Founders Pavilion, Hospital of University of Pennsylvania, 3400 Spruce Street, Philadelphia, Pennsylvania 19104
Objectives This study sought to identify the origin within the pulmonary vein (PV) of reproducible atrial fibrillation (AF) triggers.
Background Triggers for AF frequently originate from PVs. However, a systematic evaluation of the location of origin within the PV orifice and associated techniques for eliciting triggers has not been performed.
Methods Spontaneous triggers and those provoked with isoproterenol (up to 20 μg/min) and/or cardioversion in 45 patients with AF were identified using multipolar catheter recordings. In identifying origin, PVs were divided into 17 equal segments from ipsilateral PVs with “carina zone” (CZ) (7 segments between the PVs) and 10 “noncarina zone” (NCZ) segments.
Results Sixty-three reproducible triggers were noted in 37 of the 45 (82%) patients with 57 from PV and 6 (10%) from non-PV sites. Although triggers were identified from 26 of 34 distinct PV segments, most PV triggers (36, 63%) originated from CZ segments (p < 0.05) from both right (17 triggers) and left (19 triggers) PVs. The CZ triggers were more often spontaneous (11 of 36 in CZ vs. 2 of 21 in NCZ; p < 0.05) or elicited with CV (17 of 36 in CZ vs. 6 of 21 in NCZ; p < 0.05). In contrast, NCZ triggers were more likely to require isoproterenol to be provoked (13 of 21 [62%] vs. 8 of 36 [22%], p < 0.05).
Conclusions Reproducible spontaneous and provoked PV triggers initiating AF can be observed in most patients undergoing AF ablation. These triggers most commonly originate from the carina region of both right and left PVs. Noncarina PV triggers more commonly require provocation with isoproterenol infusion.
Atrial fibrillation (AF) is typically initiated by triggers originating in the pulmonary veins (PVs) (1). Our understanding of the anatomic and physiologic basis for these triggers is incomplete. It is assumed that these triggers can occur from any region of the ostia of the PVs and almost all current strategies for AF ablation include circumferential PV ablation with the goal of disconnecting the entire muscle sleeve entering the PV from the left atrium (LA). If PV triggers are localized to particular PV regions, then it might be possible to preferentially target these anatomic regions with additional radiofrequency lesions to prevent AF recurrence during PV isolation. The purpose of this study was to identify the location of PV ectopy triggering AF. We hypothesized that PV triggers might originate preferentially from specific anatomic locations, and that the location of origin of the trigger might also dictate a relatively unique response to maneuvers used to provoke the triggers.
Forty-five consecutive patients with paroxysmal or persistent AF refractory to at least 1 antiarrhythmic drug and referred for ablative therapy were included in this prospective study. There were no exclusion criteria. All patients provided written informed consent. All procedures were performed following the institutional guidelines of the University of Pennsylvania Health System.
All patients underwent transthoracic echocardiogram and a computerized tomography. Patients with persistent AF or inadequate anticoagulation therapy underwent transesophageal echocardiogram to exclude the presence of LA thrombus. Antiarrhythmic drug therapy was discontinued for at least 5 half-lives before the procedure.
Our technique has been described previously (2–4). Briefly, a 7-F decapolar catheter with 2/8-mm interelectrode spacing and proximal shocking coil for atrial defibrillation was placed in the coronary sinus (CS) via the right internal jugular vein (St. Jude Medical, St. Paul, Minnesota). Another 6-F decapolar catheter with 2/5-mm interelectrode spacing was placed vertically along the posterior right atrium with the distal electrode in the superior vena cava (SVC) (Bard, Murray Hill, New Jersey). An intracardiac ultrasound catheter (AcuNav, Acuson Inc., Mountain View, California) was placed in the right atrium at the level of the fossa ovalis for guidance and monitoring during the procedure. Two transseptal punctures were performed. An 8-mm tip or a 3.5-mm irrigated tip mapping/ablation catheter (Biosense Webster Inc., Diamond Bar, California) and a decapolar circular mapping catheter (Biosense Webster Inc.) were advanced into the LA. The anatomy of the LA and the PVs was defined using the magnetic electroanatomic mapping merged with computerized tomography (Carto Merge, Biosense Webster Inc.) and intracardiac ultrasound in all patients.
Segmentation of the PV ostium
To categorize the trigger location, each PV was subdivided into 8 approximately equal segments as shown in Figure 1. In addition, we defined the zone between both ipsilateral PVs as carina, which was designated as a separate segment (Fig. 1). In the case of common ostium, we placed the circular mapping catheter further into the vasculature to perform the segmentation in the same manner as described previously.
We defined the origin of a trigger for AF from the PV when we were able to record the earliest local activation from the circular mapping catheter placed at the ostium of the PV with activity recorded before the surface P-wave or any other intracardiac atrial electrogram from noncircular catheter recordings by at least 60 ms (5). To consider the response as a reproducible AF trigger, the trigger needed to initiate AF at least 3 times with the same intracardiac activation sequence with at least 1 of the responses sustained (>30 s). Reversal in activation sequence on any of the bipolar pairs of the circular mapping catheter with the PV potential preceding the LA potential was often observed when we defined a trigger and was used as a corroborative finding (6,7). We divided the PV sites into 2 categories: carina zone (CZ) and noncarina zone (NCZ). We defined CZ as all segments of the superior and inferior PVs in contact with each carina, as well as the carina itself. We defined NCZ as all segments not in contact with the PV carina. Overall CZ consisted of 7 segments per ipsilateral PV pair (41% of segments), and the NCZ consisted of 10 segments per ipsilateral PV pair (59% of segments) (Fig. 1). Reproducible non-PV triggers were identified using multipolar mapping/ablation catheters as previously described (8).
Identification of triggers
Patients in Sinus Rhythm
Initially, the circular mapping catheter was placed at the left superior PV ostium and the ablation catheter at the right superior PV ostium. Catheters were then manipulated to record the earliest onset of atrial activation triggering AF, guided by the activation sequence on the intracardiac catheters (9). A distal-to-proximal activation sequence in the CS was suggestive of origin in left PVs. We localized the trigger in the left superior PV when the first electrogram in the circular mapping catheter preceded by at least 60 ms the rest of electrograms in both atria (Fig. 2). If the surrounding electrograms followed the first one by <60 ms, we moved the circular mapping catheter to the left inferior PV. To further delineate the exact origin of the PV trigger within an ipsilateral vein set, we also recorded the trigger with the circular mapping and the quadripolar ablation catheters in each ipsilateral PV. By requiring reproducible triggers with a stable catheter position, we eliminated the recording of any triggers provoked by catheter manipulation. An early activation recorded from the quadripolar catheter placed in the right PVs coupled with a proximal-to-distal activation sequence in the CS was suggestive of origin in the right PVs and/or less frequently in the right atrium. The origin of the trigger in the right superior versus inferior PVs was identified as described on the left PVs with direct and stable electrode recordings using the circular mapping and quadripolar ablation catheters. Assessment of the anatomic stability of the catheter was confirmed using biplane fluoroscopy and intracardiac ultrasound (Fig. 2).
Patients in AF
If the patient presented in AF during the procedure, electrical cardioversion (CV) was performed and any triggers leading to reinitiation of AF during the subsequent 5 min were identified.
Induction of triggers
Patients in Sinus Rhythm
In the absence of spontaneous ectopy, a standardized protocol was used to provoke triggers. Incremental isoproterenol infusion (starting at 3, and increasing to 6, 12, and 20 μg/min) was performed, limited only by hypotension, patient intolerance, or onset of AF (10). If no ectopy was induced with isoproterenol, AF was induced with burst atrial pacing (15 beats at 250 ms cycle length, decreasing until shortest cycle length with 1:1 capture) followed by CV to provoke and identify the origin of AF triggers. If needed, we repeated the burst atrial pacing and CV while the patient was on isoproterenol infusion at a 2 to 3 μg/min infusion rate. We have previously documented (11) that we can elicit triggers in 95% of patients using this approach.
Patients in AF
If no spontaneous triggers occurred after the electrical CV of patients presenting in AF, then the isoproterenol infusion protocol noted earlier was followed.
After trigger localization, complete isolation of all PVs was performed. Isolation was defined by the presence of entrance and exit block using previously defined criteria (4). Non-PV triggers were targeted with focal ablation (12) unless the trigger was from the SVC, in which case the SVC was isolated (13).
Results are reported as mean ± SD for all continuous data. We compared the relationship between the location of origin of triggers and several different characteristics of the patient population and the procedure using univariate analysis. The chi-square test was used for categorical variables. When n ≤ 5 for 1 or more variables, we used the Fisher exact test. The Student t test was used for comparison of normally distributed quantitative variables. The results were considered statistically significant when p < 0.05.
Forty-five patients (35 men, 78%) with paroxysmal (38 patients, 84%) or persistent (7 patients, 16%) AF were studied. The mean age of the population was 56 ± 9 years. Thirteen patients (28%) had already undergone at least 1 prior PV isolation procedure (Table 1).
Overall, we identified 63 triggers in 37 (82%) patients (1.7 triggers per patient) during the AF mapping procedure. In 2 patients, triggers occurred but did not meet the strict definition of a “consistent” trigger. In 6 patients, there were no triggers identified.
Location of Triggers
We identified 57 triggers arising from the PVs (90%), and 6 triggers (10%) originating from non-PV sites (2 from crista terminalis, 2 from SVC, 1 from CS, and 1 from the posterior LA). The largest number of triggers came from the left superior PV (21 triggers), followed by the right superior PV (18 triggers) (Figs. 3 and 4).⇓⇓
The CZ demonstrated more triggers per segment compared with the NCZ (2.6 ± 2 triggers vs. 1 ± 0.8 triggers per segment, respectively; p < 0.05). Overall, 63% of the PV triggers (36 of the 57 triggers) were located in the CZ despite the lesser number of CZ segments compared with NCZ (7 CZ segments per PV pair vs. 10 NCZ segments per PV pair). There was an additional 160% risk for PV triggers originating from the CZ. Of note, triggers were more common from the CZ in both right (17 triggers) and left (19 triggers) PVs (Fig. 3).
Relationship of Presence of Triggers and Trigger Location to Clinical Variables
Triggers were observed more frequently in older patients (1.7 ± 1 triggers vs. 0.9 ± 0.9 triggers per patient in patients age ≥55 years vs. age <55 years, respectively, p < 0.05). There were no significant differences between patients undergoing initial and repeat PV isolation with respect to the total number of triggers (Table 2), the presence of non-PV triggers (12.5% patients undergoing initial procedure vs. 15.5% patients undergoing repeat PV isolation; p = NS), or the number of triggers per patient originating from the CZ (0.8 ± 0.8 CZ triggers vs. 0.8 ± 1 CZ triggers per patient in patients undergoing initial procedure vs. repeat PV isolation, respectively; p = NS). There were no differences in any other clinical or echocardiographic characteristic between patients with AF triggers from the CZ or NCZ (Table 2).
Induction of Triggers
Of the 63 triggers, 15 were observed spontaneously and 18 during isoproterenol infusion. The remainder of the 30 triggers were first identified in patients who were in AF at the time of the procedure and required electrical CV (range 3 to 9 CV per patient) to identify the origin of the trigger. Four of these triggers required CV during low-dose isoproterenol infusion to be elicited.
Differences in Induction
The average isoproterenol infusion rate for attempted trigger provocation was 12.6 μg/min. The need for isoproterenol depended on the localization of triggers. Triggers localized to the CZ were more likely to be spontaneous (11 of the 36 CZ triggers vs. 2 of the 21 NCZ triggers; p < 0.05) and were more likely to require CV to be elicited (17 of the 36 CZ triggers vs. 6 of the 21 NCZ triggers; p < 0.05). The NCZ triggers were more likely to need isoproterenol to be provoked (13 of the 21 NCZ triggers [62%] vs. 8 of the 36 CZ triggers [22%]; p < 0.05).
Patients undergoing a repeat procedure were more likely to have spontaneously triggers than patients undergoing a first procedure (8 of 18 triggers [44%] vs. 7 of 45 triggers [16%], respectively; p < 0.05), with 71% of the spontaneous PV triggers originating from the CZ. The proportion of triggers induced after CV and after isoproterenol was not statistically different in patients undergoing a repeat versus a first procedure (33% vs. 53% with CV, p = NS, and 28% vs. 37% with isoproterenol, p = NS, respectively).
All of the procedures included proximal PV isolation, which was successfully performed for 178 of 180 veins. In 9 patients, we successfully isolated the left-sided PVs as a unit at its junction with the LA because of left common vein anatomy. We also performed focal ablation of a posterior wall trigger in 1 patient, crista terminalis in 2 patients, and CS in 1 patient, and isolated the SVC in 2 patients. In 6 patients found to have typical flutter, we performed cavo-tricuspid isthmus ablation, obtaining bidirectional block in all cases. After the ablation procedure, isoproterenol infusion up to 20 μg/min was administered, with no AF trigger or AF provoked in any patient. There were no procedural complications.
Ablation of Triggers
Although this study was not designed to test a new ablation approach, 4 patients with triggers in the CZ precipitating recurrent bursts of AF during the procedure had initial ablation lesions applied to the CZ directed at the AF trigger in the process of PV isolation. In all 4 patients, AF terminated with the initial radiofrequency application. Because this was a study not meant to guide ablative therapy in terms of ablation strategy, we completed the ablation isolating the entire PV in all cases.
Although triggers from PVs are common, no systematic evaluation to define origin within the PV orifice has been performed. We report a systematic prospective evaluation of the origin of the exit site of PV and non-PV triggers in patients undergoing AF ablation. We observed that spontaneous or induced triggers were present in 82% of patients undergoing AF ablation. Most patients had more than 1 trigger, with an average of 1.7 anatomically distinct triggers per patient. Thus, similarly to initial studies by Haïssaguerre et al. (6), we observed that the majority of the patients have multiple triggers. Overall, 57 of 63 triggers originated from PVs or their boundaries, supporting previous findings (6,14) that identified the PVs as the most important sites triggering AF. As previously described, the upper PVs contained the majority of the AF triggers (1,6).
One of the most important findings in our current report is the differential distribution of triggers within the PVs. The percentage of sites containing AF triggers was significantly greater in the CZ than in the NCZ region of each PV. The CZ accounted for 63% of all the triggers arising from PV (36 of 57 triggers) despite the lesser number of regions compared with NCZ regions (7 vs. 10 per ipsilateral vein pair, respectively).
Overall, 33 of the reproducible triggers (52%) were observed without the need for CV. Many (18 of 33 triggers; 55%) required isoproterenol infusion to be provoked. The remaining 30 triggers required CV of AF to occur, with 4 requiring low-dose isoproterenol infusion coupled with the CV. Haïssaguerre et al. (6) observed in patients with paroxysmal AF that 65% required CV to provoke reproducible triggers. Moreover, according to our findings, most of the patients in sinus rhythm required isoproterenol infusion to localize triggers. Of note, we found that triggers from the CZ were significantly more likely to be induced without isoproterenol, suggesting greater activity in basal conditions for inducing sustained AF. Furthermore, we observed that triggers were more likely to be observed spontaneously in patients undergoing a repeat compared with a first procedure. These region-specific observations point to the carina region as a location with an apparently unique electrical behavior.
It is reasonable to speculate on the possible anatomic and physiologic basis for the described observations. Pulmonary veins are complex structures under the influence of the autonomic system. Both sympathetic and parasympathetic stimulation modulates firing within PVs (15–18). Kholová and Kautzner (19) performed an elegant anatomic study with human autopsies, observing myocardial extensions (sleeves) in the majority of patients with AF. They found significant interindividual and intraindividual variability in the presence of sleeves as well as in their arrangement and thickness (19). Unfortunately this study did not describe the regional heterogeneity in terms of sleeve characteristics along the different segments into the PV. Better understanding of the autonomic input and its relationship with the anatomy of the PVs will require additional investigation to further understand the unique behavior of CZ as a trigger.
Complete isolation of PVs is the standard technique incorporated in most ablation procedures for AF (4,20). However, reconnection of the PV at the time of repeat procedures appears to be the rule, rather than the exception (21,22). We have observed reconnection in 61% of previously isolated arrhythmogenic PVs triggering atrial tachyarrhythmias and in 82% of patients at the time of repeated PV isolation procedures (4). Lim et al. (23) observed that electrical reconnection occurred in all patients undergoing repeat ablation, with an average of 3 PVs reconnected per patient. Of note, more recently, Udyavar et al. (24) have reported very successful acute PV isolation results (84 of 84 ipsilateral PV pairs) by targeting the carina in those patients with persistence of conduction into the PV after circumferential ablation. In our opinion, it is frequently necessary to effectively isolate PVs to target within the circumferentially ablated ipsilateral veins in the carina region. Although this is performed with caution and concern for the potential for PV orifice narrowing, this may appear to be justified not only because it is required for effective vein isolation but also because it may be effective for targeting the source of the PV triggers. Given the current difficulty in achieving permanent PV isolation and the promising results of the last study by Udyavar et al. (24), it may be prudent to concentrate efforts toward the most arrhythmogenic zones of the PVs, delivering additional lesions to those areas most likely to serve as the source of AF triggers. This hypothesis is at least consistent with some clinical observations that suggest that procedures not assessing conduction block can be associated with a good clinical response (25). Our finding that the CZ is an important source of such PV triggers would suggest that additional ablation directed at these regions might decrease AF recurrence after ablation. Certainly additional study is warranted based on these unique observations.
When recording PV triggers, we identified the site of earliest activation (exit) at the PV ostium as the trigger location using a circular catheter. We did this because this technique at the very least identifies the earliest exit from a target originating deeper within the vein and, therefore, is an appropriate potential ablation target. This study did not attempt to identify the exact location of origin of a PV trigger inside the branching PV architecture. Because identification of the trigger origin was based on a single lasso recording coupled with the ablation catheter, it is possible in selected patients that this recording technique may have not identified the exact origin of the PV trigger. More detailed recording techniques such as the use of the multipolar basket catheter and/or a double lasso catheter technique may further enhance the ability to precisely identify the trigger origin.
We also required a very reproducible pattern of trigger initiation that was observed remarkably in the majority of patients. In doing so, we excluded the analysis of the origin of the minority of less reproducible triggers. Despite our meticulous research, we made the assumption that triggers induced by isoproterenol are clinically relevant, although this has not been validated by recording of spontaneous events.
Finally, we did not perform a randomized study to assess the efficacy of delivering additional lesions in the CZ if a trigger was evident. Therefore, we cannot conclude that such a strategy will decrease AF recurrence after ablation, and this hypothesis will require additional prospective evaluation.
Triggers from the PVs triggering AF can be commonly identified in patients undergoing AF ablation procedures. These triggers most frequently originate in the carina region of the PVs. In addition, the noncarina PV triggers more typically require the administration of isoproterenol, frequently at high doses, to be initiated. These unique anatomic and physiologic observations related to AF triggers provide potential insight into the mechanism of efficacy of partial PV disconnection and suggest an appropriate target for additional ablation that may enhance the efficacy of PV isolation procedures.
- Abbreviations and Acronyms
- atrial fibrillation
- coronary sinus
- carina zone
- left atrium
- noncarina zone
- pulmonary vein
- superior vena cava
- Received May 27, 2008.
- Revision received June 23, 2008.
- Accepted July 1, 2008.
- American College of Cardiology Foundation
- Dixit S.,
- Gerstenfeld E.P.,
- Callans D.J.,
- et al.
- Haïssaguerre M.,
- Jaïs P.,
- Shah D.C.,
- et al.
- Hu X.S.,
- Thomas S.P.
- Gerstenfeld E.P.,
- Marchlinski F.E.
- Lin D.P.J.,
- Beshai J.F.,
- Lewkowiez L.,
- et al.
- Hou Y.,
- Scherlag B.J.,
- Lin J.,
- et al.
- Scherlag B.J.,
- Yamanashi W.,
- Patel U.,
- Lazzara R.,
- Jackman W.M.
- Pappone C.,
- Santinelli V.,
- Manguso F.,
- et al.
- Ouyang F.,
- Antz M.,
- Ernst S.,
- et al.
- Cappato R.,
- Negroni S.,
- Pecora D.,
- et al.
- Pappone C.,
- Oreto G.,
- Rosanio S.,
- et al.