Author + information
- Received July 6, 2000
- Revision received May 21, 2001
- Accepted June 4, 2001
- Published online September 1, 2001.
- David Schwartzman, MD, FACC∗,* (, )
- Eduardo N Warman, PhD†,
- William A Devine, BS‡ and
- Rahul Mehra, PhD†
- ↵*Reprint requests and correspondence: Dr. David Schwartzman, Cardiovascular Institute, Presbyterian University Hospital, B535, 200 Lothrop Street, Pittsburgh, Pennsylvania 15213-2582
We sought to characterize a method of attenuating interatrial conduction using radiofrequency ablated lesions applied to the right atrial septum.
Interatrial conduction occurs in specific zones. Recent data suggest that interatrial conduction can be important in triggering and sustaining atrial fibrillation. Therefore, a method for attenuating interatrial conduction may have therapeutic value.
In 13 healthy pigs, interatrial conduction was evaluated before and after sequential ablation of the right atrial septum, targeting interatrial conduction zones. In six animals, zone 1 (crista terminalis and limbus) was ablated first, followed by ablation of zone 2 (fossa ovalis and coronary sinus ostium). In the other seven animals, the order of ablation was reversed. Electrophysiologic and pathologic findings were correlated.
After ablation of zone 1, interatrial conduction was slowed, but there was no block. After ablation of zone 2, conduction was unchanged. After ablation of both zones, complete block was observed in four animals, and there was left atrial quiescence. In the remaining nine animals, incomplete block was observed, with marked conduction slowing or block during sinus rhythm and pacing. Ablation did not adversely affect atrioventricular node conduction, nor did it facilitate sustenance of an atrial arrhythmia. Pathologic analysis revealed that complete interatrial conduction block was associated with confluent ablation of both targeted zones.
Catheter ablation of the right atrial septum attenuated interatrial conduction without disturbing atrioventricular conduction.
The anatomy of the interatrial septum has been the subject of several reports, some of which have included descriptions of interatrial myocardial connections (1–9). Distinct zones of interatrial conduction have been correlated with these connections (10–12). Previous reports have suggested that intact interatrial conduction is important for triggering and sustaining atrial fibrillation (AF) (13–16). Thus, it has been hypothesized that attenuation of interatrial conduction might have therapeutic utility. A surgical method for eliminating interatrial conduction has been reported, and this method has resulted in a reduction in the incidence of AF (17). However, this method employed a continuous, transmural left atrial incision that leaves the septum intact. In addition, sternotomy and cardiopulmonary bypass were necessary.
The present study is based on the hypothesis that significant attenuation, including complete block, of interatrial conduction could be achieved using percutaneous radiofrequency catheter ablation to target right atrial septal endocardial sites that are associated with interatrial myocardial connections.
Thirteen large (35 to 60 kg) healthy Hanford pigs were used. It has been previously demonstrated that atrial anatomy in this animal is similar to that of humans (18). After endotracheal intubation, a surgical plane of anesthesia was maintained using 1% to 4% isoflurane. Arterial blood was was monitored throughout the procedure. Access to the right atrium was achieved percutaneously through the femoral vein. Access to the left atrium was achieved through the femoral vein and atrial septal puncture.
Two nonfluoroscopic techniques were used: 1) this CARTO (Biosense/Webster, Diamond Bar, California) system employs synchronous extracorporeal magnetic fields to determine the location and orientation of a sensor mounted within an intracardiac catheter (NaviStar, Biosense/Webster) near its distal electrode (19). 2) Intracardiac echocardiography (ICE). The system was composed of a 9F catheter (Boston Scientific, Natick, Massachusetts) incorporating a rotating transducer operating at 9 MHz (20). The catheter was placed into the right atrium. Intracardiac echocardiography provided a detailed image of each of the right atrial septal anatomic landmarks (21).
In addition to the NaviStar and ICE catheters, standard quadripolar catheters were placed in the right atrial appendage (RAA) and left atrial appendage (LAA). These catheters were positioned at the beginning of the experiment and remained in place for its duration. The distal electrode pairs of each catheter were used for pacing, and the proximal pair for measurement of activation times. At baseline, an electrophysiologic study was performed, consisting of several indexes: 1) Sinus rhythm: sinus cycle length (SCL), corrected sinus node recovery time, maximal duration of surface P-wave, derived from the 12-lead electrocardiogram (PDur), PQ interval (P-wave onset to QRS onset), interval between local activation of the right atrial appendage and local activation of the atrium adjacent to this His bundle (RAA-HBA) and interatrial conduction time, measured by interval between RAA and LAA (ICT). 2) RAA pacing: atrioventricular node (AVN) Wenckebach cycle length, ICT at a pacing cycle length of 500 ms (ICT500; pacing was also performed at cycle lengths of 400, 350, 300, 250 and 200 ms; observations at these cycle lengths are reported, but ICTs are not), AVN local atrial effective refractory period (ERP), ICT during extrastimulation (500 ms drive cycle), with coupling intervals between 400 ms and ERP (10 ms steps), and ERP. Inducibility of atrial tachyarrhythmia was evaluated with RAA burst pacing (20 mA, 2 ms pulse width) at a cycle length of 100 ms for 10 s. 3) LAA pacing: same as RAA pacing.
Because of differences in the exact positions of the RAA and LAA electrodes between animals, baseline ICT varied. To combine individual data to assess the relationship between ICT and the atrial extrastimulus coupling interval, conduction times in each animal were expressed as the difference (Δ) from the ICT measured at the longest extrastimulus coupling interval (400 ms). The maximal difference between interatrial conduction times during atrial extrastimulation (Δmax) for each extrastimulation procedure was measured.
The NaviStar catheter was used for radiofrequency energy applications. This catheter is 7F with a 4.5-mm length distal ablation electrode, with a thermocouple embedded near its tip. Radiofrequency energy was applied in unipolar fashion to the right atrial septum. Energy titration was guided by electrode thermometry, with a target ablation temperature of 60°C for 60 s. Intracardiac echocardiography (ICE) and the CARTO system were utilized in tandem: ICE was utilized to define right atrial septal anatomy, to position the ablation electrode relative to anatomic structures and to ensure firm, stable electrode-endocardial contact during each radiofrequency application; CARTO was used to record lesion locations. For purposes of guiding ablation, four anatomic areas were grouped into two “zones”: zone 1 (crista terminalis and limbus) and zone 2 (fossa ovalis and coronary sinus ostium) (Fig. 1). Ablation of each zone was performed using a series of radiofrequency applications. Each radiofrequency application site was demarcated by CARTO as a spherical icon with a diameter of 4 mm (19). Ablation of a septal zone was defined as complete when icons filled the zone and no significant near-field potentials could be recorded within the zone.
Animals were sacrificed immediately. The atria were removed en bloc and immersed in a bath of 1% triphenyl tetrazolium chloride for 60 min. The specimens were photographed within 24 h of death. Endocardial lesion dimensions were measured using photographic planimetry.
After the baseline electrophysiologic study, ablation was performed. In six animals, zone 1 was ablated first; in seven animals, zone 2 was ablated first. The first ablation zone was chosen at random. The electrophysiologic study was repeated after ablation of each zone.
Data are reported as the mean value ± SD, unless otherwise specified. Baseline and postablation data were compared by analysis of variance, with use of the Bonferroni test for multiple comparisons. A p value <0.05 was considered statistically significant.
Complete interatrial conduction block—no interatrial conduction during sinus rhythm or atrial pacing at any cycle length. Atrial fibrillation—tachycardic atrial electrogram varying in cycle length and morphology; defined as “transient” if it persisted for ≥5 s but terminated spontaneously within 60 s. Atrial flutter—tachycardiac atrial electrogram with constant cycle length and morphology; defined as “transient” if it persisted for ≥5 s but terminated spontaneously within 60 s. Atrial tachyarrhythmia(AF, flutter) inducibility—tachyarrhythmia observed after ≥50% of induction attempts.
Electrophysiologic study (Table 1)
Interatrial conduction block was not observed during sinus rhythm or pacing. Interatrial conduction times in either direction were similar. The ERP in the LAA was significantly shorter than that in the RAA. Left atrial extrastimulation was associated with a significantly larger Δmaxrelative to right atrial extrastimulation. However, at extrastimulus coupling intervals that could be tested in both atria (e.g., >RAA ERP), there was no significant difference. Transient “bi-atrial” AF (e.g., AF in both RAA and LAA electrograms) was inducible in each animal from the LAA and RAA. No atrial flutter was induced.
After ablation of zone 1 (crista terminalis and limbus) first
The P-wave duration and right intra-atrial conduction time were both significantly increased relative to baseline (Fig. 2). There was also a significant increase in the interatrial conduction time during both sinus rhythm and pacing, but no conduction block. There were no significant differences in Δmaxvalues relative to baseline. Transient bi-atrial AF was inducible in each animal from the LAA and RAA. No atrial flutter was induced.
After ablation of zone 2 (fossa ovalis and coronary sinus ostium) first
No significant changes were noted in the P-wave duration, right intra-atrial conduction time or interatrial conduction time. There was no interatrial conduction block. There were no significant differences in Δmaxvalues relative to baseline. Transient bi-atrial AF was inducible in each animal from the LAA and RAA. No atrial flutter was induced.
After ablation of both zones
During sinus rhythm, four animals had complete interatrial conduction block, and nine animals had incomplete block.
1) Complete interatrial conduction block: despite a marked increase in the right intra-atrial conduction time, the P-wave duration was significantly decreased due to loss of the left atrial component (Fig. 2). There was no intrinsic left atrial rhythm (Fig. 3). There was no significant change in AVN conduction. During RAA stimulation, neither AF nor flutter was inducible in any animal. During LAA stimulation, only transient left atrial AF was inducible in each animal; the right atrium remained in sinus rhythm (Fig. 4).
Left atrioventricular block, defined by ventricular activation at the SCL despite left atrial pacing at a sub-SCL, was observed in two animals. In the other two animals, intact left atrioventricular conduction was observed. In each case, the electrocardiographic QRS morphology was identical to that in sinus rhythm. In the first animal, the LAA-QRS interval during pacing at 500 ms was 260 ms, whereas the RAA-QRS interval during pacing at 500 ms was 150 ms. Left atrioventricular block occurred in a Wenckebach pattern at a cycle length of 350 ms (right atrioventricular block cycle length of 240 ms). In the second animal, the LAA-QRS interval during pacing at 500 ms was 390 ms, whereas the RAA-QRS interval during pacing at 500 ms was 180 ms. Left atrioventricular block occurred in a Wenckebach pattern at 400 ms (right atrioventricular block cycle length of 300 ms).
2) Incomplete interatrial conduction block: during sinus rhythm, block was observed in two animals. In both of these animals, interatrial conduction was intact during left atrial pacing at cycle lengths similar to those in sinus rhythm. In the other seven animals, interatrial conduction was intact during sinus rhythm, albeit with a markedly prolonged duration. This was associated with a marked increase in the P-wave duration, with the emergence of two discrete complexes (Fig. 2).
During right atrial pacing, block was observed at all pacing cycle lengths in the two animals with block in sinus rhythm. Block was observed in three additional animals at cycle lengths ranging from 400 to 200 ms; in each of these animals, block had a Wenckebach periodicity. During extrastimulation in the seven animals with intact conduction at the drive cycle length, Δmaxwas significantly increased. Block was observed in two of the seven animals. Transient AF, with a mean LAA cycle length significantly longer than the mean RAA cycle length, was inducible in four of the nine animals. No atrial flutter was induced in any animal.
During left atrial pacing in the two animals exhibiting block in sinus rhythm, intact interatrial conduction was observed at a cycle length of 500 ms. In another animal, block was demonstrated at all cycle lengths, despite intact interatrial conduction in sinus rhythm. Block was observed in two additional animals at cycle lengths ranging from 300 to 200 ms; in each animal, block had a Wenckebach periodicity. During extrastimulation in the eight animals with intact conduction at the drive cycle length, Δmaxwas significantly increased. Block was observed in three of the eight animals. Transient AF, with a mean RAA cycle length longer than the mean LAA cycle length, was inducible in all animals (Fig. 5). No atrial flutter was induced in any animal.
In the animals with incomplete block and intact left-to-right interatrial conduction, ICE-guided CARTO activation mapping of the right atrium was performed during LAA pacing (550 ms cycle length) (19). In each animal, the chamber was activated by a single wave front entering the right atrium through the coronary sinus ostium.
The area of the zone 1 ablation lesion was 1.7 ± 0.6 cm2and that of zone 2 was 1.4 ± 0.4 cm2(Fig. 1). There was no damage to the triangle of Koch. The left atrial septum had minimal lesions limited to the area opposite the fossa ovalis. In animals with complete block, the ablation lesion was confluent in each zone. In each of the animals with incomplete block, one or more lesion gaps in the coronary sinus ostium, ranging in total area from 2 to 18 mm2, were observed.
Our data indicate that ablation of the right atrial septum, targeting anatomic areas associated with interatrial myocardial connections, results in attenuation of interatrial conduction, including complete interatrial conduction block. This was associated with preserved sinus node and AVN function.
After ablation of zone 1 but not zone 2, both the sinus P-wave duration and interatrial conduction times were significantly increased relative to baseline. After ablation of either zone, the response of interatrial conduction during pacing or extrastimulation was similar to that at baseline. These data suggest that, based on the proximity to the sinus node, zone 1 connections were primarily responsible for the P-wave duration. The superior locations of both the RAA and LAA recording catheters explain the correlation between the sinus P-wave duration and interatrial conduction time. These data demonstrate the limited utility of the P-wave duration for assessing overall interatrial conduction properties.
After ablation of both zones 1 and 2, interatrial conduction times increased dramatically, and decremental properties were observed. In some animals, complete bi-directional conduction block was observed. In other animals, block was incomplete, occurring uni-directionally or at sub-SCL, or both. In animals with incomplete conduction block, interatrial conduction times were longer than those after zone-1 ablation alone, indicating that ablation of zone 2 added to conduction slowing.
Atrial tachyarrhythmia inducibility
Inducibility of atrial tachyarrhythmias before and after ablation was limited to transient AF. These data suggest that the lesions deployed in the present study did not promote sustenance of fibrillation or flutter. In fact, our observations of left atrial quiescence (Fig. 3), right atrial sinus rhythm during left AF (Fig. 4)asymmetric AF cycle length (Fig. 5)are similar to those reported after surgical left or right atrial isolation or the Corridor procedure, or both (22,23).
Left atrioventricular conduction
In animals with complete conduction block, no intrinsic electrical activity of the left atrium was observed. Quiescence has also been reported after the surgical isolation procedure in humans (17). Complete interatrial conduction block afforded a unique opportunity to probe for left atrioventricular conduction. In two of four animals, such conduction was observed. Relative to sinus rhythm and right atrial pacing, the atrioventricular conduction time during left atrial pacing was markedly prolonged, and atrioventricular block occurred with Wenckebach periodicity. Although inconclusive, these data suggest that there are inputs to the AVN from the left atrium that may be independent of, and less robust than, right atrial inputs in this animal model.
In the present study, confluent ablation of both zones was necessary to achieve complete interatrial conduction block. Incomplete block was consistently associated with one or more lesion gaps, always in the coronary sinus ostium. Lesion gaps limited to the same area in all animals suggest a systematic fault in the ablation procedure, possibly associated with ablation electrode access or stability.
The number of lesions used to ablate each atrial septal zone significantly exceeded the postmortem lesion surface area. Several factors may account for this. First, we were very conservative in lesion spacing. Given that our goal was to pursue attenuation of interatrial conduction, minimizing the number of lesions was a secondary concern. This issue would obviously become more important in human trials employing this technique, during which mapping of right atrial activation during left atrial pacing may further enhance specificity and minimize the number of lesions required. Second, the areas reported herein were calculated from two-dimensional photographic planimetry. Given the complex topography of the right atrial septum, it is likely that this technique significantly underestimated the actual surface area. Finally, specimen shrinkage may have been a factor.
We emphasize several important limitations in our data. First, given the superior location of RAA and LAA electrodes, we cannot exclude the presence of significant changes in interatrial conduction changes after zone 2 ablation alone. Second, although we attributed increases in interatrial conduction time after right atrial septal ablation to trans-septal conduction delay, in this study, we did not map activation between RAA and LAA sites. Therefore, the site(s) of delay cannot be discerned with complete assurance. Third, AF inducibility in an unconditioned porcine model is poorly characterized and may not be predictive of the impact of the ablation procedure on previously reported AF in animal models or in humans. Finally, given that we performed no experiments to investigate the impact of the procedure duration or alternative site ablation on AF inducibility, it is possible that its relationship with septal ablation was coincidental. In this regard, the significant changes in SCL during the study indicate that variation in factors with potential influence on AF inducibility, such as autonomic tone, were likely to occur.
☆ This study was supported in part by a grant from Medtronic, Inc., Minneapolis, Minnesota.
- atrial fibrillation
- atrioventricular node
- maximal difference between interatrial conduction times during atrial extrastimulation
- local atrial effective refractory period
- intracardiac echocardiography
- interatrial conduction time, measured by interval between RAA and LAA
- left atrial appendage
- maximal duration of surface P wave, derived from 12-lead electrocardiogram
- right atrial appendage
- sinus cycle length
- Received July 6, 2000.
- Revision received May 21, 2001.
- Accepted June 4, 2001.
- American College of Cardiology
- Robb J.S
- Chauvin M,
- Shah D.C,
- Haissaguerre M,
- et al.
- Inoue S,
- Becker A.E
- Antz M,
- Otomo K,
- Arruda M,
- et al.
- Sun H,
- Velipasaoglu E.O,
- Wu D.E,
- et al.
- Roithinger F.X,
- Cheng J,
- SippensGroenewegen A,
- et al.
- Kumagai K,
- Khrestian C,
- Waldo A.L
- Allessie M.A,
- Lammers W.J.E.P,
- Bonke F.I.M,
- et al.
- Allessie M,
- Lammers W,
- Smeets J,
- et al.
- Skanes A.C,
- Mandapati R,
- Berenfeld O,
- et al.
- van Hemel N.M,
- Defauw J.J,
- Kingma J.H,
- et al.