Characterization of Reentrant Circuit in Macroreentrant Right Atrial Tachycardia After Surgical Repair of Congenital Heart Disease
Isolated Channels Between Scars Allow “Focal” Ablation
Background—The purpose of this study was to characterize the circuit of macroreentrant right atrial tachycardia (MacroAT) in patients after surgical repair of congenital heart disease (SR-CHD).
Methods and Results—Sixteen patients with atrial tachycardia (AT) after SR-CHD were studied (atrial septal defect in 6, tetralogy of Fallot in 4, and Fontan procedure in 6). Electroanatomic right atrial maps were obtained during 15 MacroATs in 13 patients, focal AT in 1 patient, and atrial pacing in 2 patients without stable AT. A large area of low bipolar voltage (≤0.5 mV) involved most of the free wall in all patients and contained 2 to 7 dense scars or lines of double potentials, forming 29 narrow channels (width ≤2.7 cm) between scars in all but 1 patient, who had a single scar and only focal AT. All 15 MacroATs were propagated through narrow channels. Ablation within the channel eliminated all 15 MacroATs with 1 to 3 (median 1) radiofrequency applications. Ablation was performed in 9 other channels identified during MacroAT (5 patients) and in 5 channels identified during atrial pacing (2 patients). Conduction block was obtained across 28 of 29 channels. After ablation, reproducible sustained right AT was not induced in any patient. During follow-up (median 13.5 months), new MacroATs, atrial fibrillation, or palpitations occurred in 3 of 16 patients.
Conclusions—MacroAT after SR-CHD requires a large area of low voltage containing ≥2 scars forming narrow channels. Ablation within the channels eliminates MacroAT.
Macroreentrant right atrial tachycardia (MacroAT) is a late complication that occurs after the surgical repair of congenital heart disease.1 2 3 4 Catheter ablation has been useful in only selected patients1 2 3 4 because of the presence of multiple or unstable (unmappable) atrial tachycardias (ATs).
Two approaches have been proposed for ablation. In the first approach, an isolated diastolic atrial potential is identified during MacroAT, and entrainment pacing is used to confirm participation within the reentrant circuit.1 4 The second approach is based on the concept that MacroAT results from reentry around the atriotomy scar. A linear lesion is created between the atriotomy scar and an anatomic barrier (tricuspid annulus or inferior vena cava [IVC] or superior vena cava [SVC]).2 3 These approaches ablated at least 1 tachycardia in 73% to 93% of patients.1 2 3 However, tachycardia recurred in 33% to 53% of patients after acute success.2 3 4 We hypothesized that the high recurrence after ablation might be related to incomplete identification of all possible reentrant circuits.
The purpose of the present study was to localize all components of MacroAT reentrant circuits in patients after the surgical repair of congenital heart disease by use of an electroanatomic mapping system5 and to identify and eliminate other potential reentrant circuits that might be related to otherwise unmappable tachycardias.
All patients referred for ablation of AT after the surgical repair of congenital heart disease were included. Patients were not excluded for multiple ATs by history or at electrophysiological study or for noninducibility of stable sustained AT. Patients with only typical atrial flutter (AFL, reentry around the tricuspid annulus) were excluded.
Electrophysiological study was performed with patients under general anesthesia (propofol). Multielectrode catheters were inserted transvenously and positioned in the coronary sinus, in the high right atrium, around the tricuspid annulus, and (except Fontan patients) in the His bundle region and right ventricle. Right atrial mapping was performed during each sustained AT by use of a quadripolar electrode catheter with 0.5- to 5- to 2-mm spacing and a 4-mm tip electrode (Navistar, Biosense Webster). The average distance between mapped sites was ≈5 mm (2 to 4 mm in areas of interest).
Electroanatomic Mapping System
The electroanatomic mapping system (CARTO, Biosense Webster) has been described previously.5 With use of a magnetic sensor, the system displays the location of the mapping/ablation catheter relative to the location of a reference sensor taped to the posterior chest (accuracy ≤1 mm).5
A coronary sinus or right atrial electrogram was used as a timing reference. A window of time, equal to the tachycardia cycle length, was established around the reference activation time. The activation time at each site was displayed in color relative to its timing within the window (earliest, red; latest, purple). The voltage of the bipolar atrial potential recorded at each site was also displayed in color (lowest, red; greatest, purple). Areas with no atrial potential distinguishable from noise (generally ≤0.03 mV) were considered dense scars and were displayed in gray.
Definition of MacroAT
MacroAT was defined as follows: (1) continuous sequence of activation, with earliest activation (red) adjacent to latest activation (purple), and (2) range of activation times equal to the tachycardia cycle length.
A focal pattern of activation was defined as follows: (1) radiation in all directions from a single site of earliest activation and (2) range of activation times less than the tachycardia cycle length.
Ablation was directed at the site of earliest activation for focal AT and directed at isolated channels for MacroAT. Radiofrequency (RF) current was initiated at ≈10 W, was increased until the impedance decreased 3 to 5Ω, and was maintained until the unipolar atrial potential recorded from the ablation electrode decreased by 80% or split into 2 small potentials, indicating transmural necrosis. The catheter was held stationary (≈70%) or was dragged across the channel (≈30%).
Conduction block across the channel was verified by pacing close to the ablation line and by demonstration of marked delay and reversal in the direction of activation on the opposite side of the ablation line. Programmed stimulation was repeated to induce the same or another AT.
When AT was nonsustained, variable in activation sequence, or nonreinducible (unmappable AT), mapping was performed during atrial pacing to identify the channels between scars. Ablation was performed across as many narrow (<3.0 cm in width) channels as possible.
Transesophageal echocardiography was performed within 24 hours after ablation. Patients with repair of an atrial septal defect (ASD) or tetralogy of Fallot (ToF) received aspirin (325 mg/d) for 6 weeks. Patients with a Fontan procedure or history of atrial fibrillation received warfarin. The patients were discharged on no antiarrhythmic drugs.
The study population consisted of 16 patients (Table 1⇓). Six had repair of secundum ASD; 4 had correction of ToF; and 6 had the Fontan procedure.
AT, present for 1 to 13 years, was paroxysmal in 10 patients and incessant in 6. Multiple ATs had been documented in 12 (75%) patients. AT was unresponsive to 2 to 6 (median 3) antiarrhythmic drugs in all patients, including amiodarone in 9 patients. Nine (56%) patients had prior unsuccessful ablation procedures for AT. Six patients had previously undergone ablation of AFL.
Sixty-nine ATs (1 to 7 per patient, median 5) were present incessantly (6 ATs) or were induced by programmed stimulation (63 ATs, Figure 1⇓). Four tachycardias originated from the left atrium, and 65 originated from the right atrium. Seventeen right ATs in 14 of the 16 patients were sustained and reproducible, allowing a complete right atrial map during tachycardia. Fifteen of 17 ATs (13 of 14 patients) exhibited a macroreentrant activation pattern, and 2 ATs (2 patients) exhibited focal activation originating at the crista terminalis.
In the remaining 2 patients (Nos. 6 and 16, Table 1⇑) with unmappable AT, a right atrial map was obtained during atrial pacing to localize potential arrhythmogenic channels.
A complete right atrial map (182 to 318 points) was obtained during 5 ATs in 5 of 6 ASD patients (Nos. 1 to 5) and was consistent with MacroAT in all (Figure 2⇓). The circuit was located within a large area of low voltage (bipolar voltage ≤0.5 mV) involving the free wall and extending to the septum (median height 7.8 cm, width 10.3 cm; Table 2⇓) in all patients. The low-voltage area contained many sites exhibiting fragmented or double atrial potentials. It also contained 2 adjacent areas with no bipolar atrial potential distinguishable from noise (dense scars) in 2 patients (Figures 2⇓ and 3⇓) or a narrow scar (line of double potentials) adjacent to a dense scar in 3 patients (Figures 4⇓ and 5⇓). The dense scars or areas of double potentials measured 1.0 to 4.9 (median 2.4) cm in length and 0.5 to 4.4 (median 1.6) cm in width. The lower scar appeared continuous with the IVC in all 5 patients. The upper scar was larger than the lower scar and located at the mid free wall in 4 of 5 patients (Figure 3⇓). The MacroAT circuit propagated around the upper scar and through the narrow channel between the 2 adjacent scars in all 5 patients. The channels were ≈0.7 to 1.6 cm in width and 0.8 to 3.4 cm in length (Table 2⇓).
All 5 MacroATs were terminated by 1 to 3 (median 2) stationary (Figures 2⇑ and 3⇑) or dragging (Figure 4⇑) RF applications applied within an isolated channel. After ablation of the single channel, the original AT was not induced in any of the patients. No other AT was induced in 3 of 5 patients (Nos. 1, 2, and 4). In patient 3, 1 episode of another AT was induced. This tachycardia could not be reinduced, preventing mapping and ablation. Two left ATs were also induced in this patient. Neither left atrial mapping nor ablation was attempted. In patient 5, 6 other ATs were induced that were nonsustained or not reinduced. After ablation across 2 other channels selected from the initial map (Figures 4⇑ and 5A⇑), no AT was induced.
In the remaining patient (No. 6), entrainment pacing performed before completion of the map terminated the MacroAT. The original tachycardia could not be reinduced. Stimulation induced multiple episodes of 3 other nonsustained ATs. A map was then obtained during atrial pacing, revealing a large area of low voltage containing 2 dense scars forming a channel 1.5 cm in width (Figure 5A⇑). Ablation produced block across the channel. No AT was induced after ablation.
Eight channels were targeted in the 6 ASD patients. Block was confirmed in 7 patients and not tested in 1 (Table 2⇑).
The septum did not participate in the circuit in any of the 5 mapped MacroATs. Although reduced, the septal voltage was greater than the free wall voltage in all 6 patients (Figure 4⇑). The septum contained a small dense scar or double potentials in only 2 patients (Nos. 3 and 5). The closest area of block to these small scars was 4.4 and 3.7 cm. These wide channels were not associated with MacroAT, and ablation was not attempted.
A map (175 to 223 points) was obtained during sustained AT in all 4 ToF patients. Three (Nos. 7 to 9) had macroreentry confined to the free wall within a large area of low voltage (Table 2⇑), containing 2 dense scars or a dense scar and a line of double potentials (Figure 5B⇑). The lower scar was continuous with the IVC in 2 patients, and there was a line of double potentials (block) between the lower scar and IVC in 1 patient (Figure 6⇓). In all 3 patients, reentry was propagated through the channel (0.7 to 1.1 cm wide) and around the upper scar. Only 1 to 3 RF applications terminated tachycardia, produced block across the channels, and prevented reinduction of any sustained AT.
The fourth patient (No. 10) had only focal AT (crista terminalis). One RF application eliminated the AT. The voltage map showed a narrower area of low voltage (width 3.2 cm), containing a single scar extending to the IVC (no channel).
In the 6 Fontan patients, maps containing 257 to 419 (median 321) points were obtained during 2 MacroATs (2 patients), during 1 MacroAT (3 patients), and during atrial pacing (1 patient). The maps demonstrated very large right atria with large areas of low voltage involving the free wall and septum, containing 3 to 7 (median 5.5) dense scars or lines of double potentials (Table 2⇑). The size and location of the scars varied among patients, resulting in different reentrant circuits (Figure 5C⇑).
In 2 patients (Nos. 11 and 14), the scars were continuous with SVC or IVC, producing reentry around the right atrium in the transverse plane (Figure 7⇓). The other 4 patients had 3 to 6 channels and 3 to 7 MacroATs. All channels were identified in the first map obtained either during a tachycardia (patients 12, 13, and 15) or atrial pacing (patient 16). For example, in patient 12, the first map obtained during a MacroAT showed a small circuit (2-cm diameter) incorporating 2 channels plus a third uninvolved channel (Figure 8⇓). The voltage throughout the circuit was extremely low (0.04 to 0.06 mV), which may explain the long cycle length (435 ms). A single stationary RF application within 1 channel terminated the tachycardia. One additional application was delivered within the same channel (RF2, Figure 8B⇓). Three applications were delivered across the second channel, guided by the initial map (RF3 to RF5). The third channel was subsequently shown to produce another MacroAT, which was terminated by a single linear (1.0-cm length) RF application within the channel (RF6), followed by an additional application (RF7).
At least 6 channels were identified from a map (367 points) obtained during atrial pacing in patient 16, who had multiple ATs that were unmappable because of the frequently changing activation sequence. Four of 6 channels were selected for ablation (Figure 5C⇑). Block was obtained across 3 of the 4 channels. Block was not obtained across the fourth channel with the highest voltage (1.48 mV), consistent with thicker atrial myocardium.
Block was obtained across 19 of 20 channels in the 6 Fontan patients (Figure 9⇓, Table 2⇑). After ablation, only 6 of 33 ATs were induced in 3 of 6 patients. These tachycardias were nonsustained (3 ATs), left atrial (2 ATs), or nonreproducible (1 AT).
Entrainment pacing was performed in the first 3 ASD patients, 2 ToF patients, and 1 Fontan patient. Entrainment confirmed the sites suspected to be inside or outside the reentrant circuit on the basis of the activation map (Figures 2⇑, 3⇑, 6⇑, and 7⇑). Many sites within the circuit, but outside the isolated channel (“outer loop” site), exhibited a diastolic potential, entrainment with concealed fusion (same P wave and activation sequence), and a postpacing interval equal to the tachycardia cycle length (PPI=TCL). Ablation at one such site failed to terminate tachycardia, whereas 1 application of RF current delivered within the channel (identified from the activation map) terminated the tachycardia, even though pacing at that site failed to capture. (Figure 3⇑).
Because entrainment provided only confirmatory information and could terminate the tachycardia before ablation, entrainment was not used in the last 9 patients. The number of RF applications required to terminate MacroAT was not different between the groups using and not using entrainment, 1 to 3 (median 1.5) versus 1 to 3 (median 1) applications.
Characteristics of Channel Potentials
The timing of the atrial potential at the 15 ablation sites terminating MacroAT was distributed throughout the tachycardia cycle length (Figure 10⇓). The electrogram exhibited single (6 sites), double (5 sites), or fragmented (4 sites) atrial potentials. The bipolar atrial potential amplitude ranged from 0.05 to 0.41 mV. The amplitude was extremely low (≤0.1 mV), mimicking dense scar, at 7 of the 15 sites (Figure 8C⇑).
Ablation of AFL
AFL was also ablated in 12 of 16 patients, 7 in this procedure and 5 in a prior procedure.
The 16 patients were followed up for 1 to 29 (median 13.5) months after ablation. Fifteen received no antiarrhythmic drug. Patient 1 eventually received sotalol to suppress ventricular tachycardia. Sotalol has previously failed to prevent MacroAT.
All 16 patients have had a significant improvement in symptoms. Atrial tachyarrhythmia recurred in 3 (19%) of 16 patients. One ASD patient (No. 3) had 3 episodes of AT (lasting 6 hours) over 7 months. At repeat study, a new MacroAT with a longer cycle length (330 versus 208 ms) was induced. The map demonstrated reentry around a large scar (formed by fusion of the original upper and lower scars) and through a narrow channel between the fused scar and IVC (Figure 5A⇑, patient 3, dotted line). The conduction time through the channel was 155 ms, simulating block during the first ablation procedure. Tachycardia was eliminated by a single RF application within the new channel. Mapping during atrial pacing after ablation confirmed block across the new and previous channels.
One Fontan patient (No. 12) had a single episode of atrial fibrillation 12.5 months after ablation. Electrophysiological study failed to induce any of the original ATs. Mapping confirmed block across the 3 original channels. Programmed stimulation induced a focal AT originating in the posterior right atrium, eliminated by 2 RF applications.
Another Fontan patient (No. 13) had a single episode of palpitations lasting 15 minutes and several other episodes of palpitations lasting <15 seconds. None of the episodes were recorded.
There were no complications.
Substrate for MacroAT
The entire reentrant circuit was identified in the 15 sustained reproducible MacroATs. All circuits were located within a large area of low voltage (depressed conduction) containing a narrow channel (≤2.7 cm in width) between 2 close dense or thin scars (lines of block). The lines of block (double potentials) during tachycardia were also recorded during atrial pacing (long cycle length) after ablation, suggesting fixed anatomic block rather than functional block. Similarly, areas with no atrial potential during tachycardia exhibited no atrial potential during pacing.
The arrhythmogenic substrate was found in the free wall and not the septum, even in patients with ASD closure. The cause of the large area of low voltage (“atrial myopathy”) is unclear. Possible explanations include interruption of arterial supply and insufficient protection during cardioplegia. The multiple dense scars may result from the atriotomy, venous cannulation sites, conduit sites, and other surgical trauma.
The single patient with only focal AT had a narrower area of low voltage and a single dense scar extending to IVC, with no channel through the scar (no channel–no macroreentry).
Focal Ablation Within an Isolated Channel
Each MacroAT used at least 1 narrow channel, allowing a single point or very short linear (focal) ablation. The 15 mapped MacroATs were terminated by 1 to 3 (median 1) stationary or short dragging RF applications delivered within an isolated channel.
Only 1 isolated channel was identified in 7 of 9 ASD and ToF patients. In contrast, 5 of 6 Fontan patients had multiple channels and macroreentrant circuits. Because all channels could be identified from a single high-density map (2- to 5-mm spacing), the strategy used in later patients was to obtain 1 complete map during a stable MacroAT. One channel was ablated, terminating the tachycardia. Guided by the initial map, as many other channels as possible were ablated before attempting to reinduce tachycardia. This approach reduced the number of maps required. The ability to identify the channels from a map during atrial pacing allowed the ablation of multiple channels in 2 patients with unmappable ATs. The lower success rate for ablation in Fontan patients in previous studies may relate to the inability to identify all of the channels by use of conventional mapping techniques.1 2 3 4
The atrial potential at 7 of the 15 ablation sites terminating MacroAT was extremely small (≤0.1 mV), mimicking dense scar (Figures 8⇑ and 10⇑). In the present study, the maps included every distinguishable bipolar potential. Many of the channels between 2 scars would have been obscured (displayed as a single scar) if the scar had been defined by an arbitrary voltage criterion (ie, ≤0.1 mV).
The combination of a diastolic potential and entrainment pacing showing concealed fusion with PPI=TCL has been considered to identify a “protected isthmus” within the circuit (ideal ablation site).1 2 In the present study, potentials within channels occurred throughout the atrial cycle, not just in diastole. In addition, because most of the cycle length involved conduction in broad low-voltage areas, many sites outside of channels exhibited diastolic potentials and concealed entrainment (because the paced impulse exited the low-voltage area in the same region) with PPI=TCL (Figure 3⇑). Neither electrogram morphology, timing, nor entrainment differentiated between sites within or outside a narrow channel. Entrainment pacing may also terminate or change the tachycardia and therefore was not helpful in the presence of a high-density map.
After ablation of the channels, reproducible sustained right AT was not induced by programmed stimulation in any of the 15 patients. This compares favorably with the 73% to 93% acute success rate (defined as elimination of at least 1 AT) with the use of conventional catheter mapping techniques.1 2 3 More important, confirmed sustained MacroAT recurred during follow-up in only 1 patient, in whom a slowly conducting channel was initially mistaken as a block. In previous studies using conventional mapping techniques, AT recurred in 33% to 53% of patients with acute success.2 3 4
AT has not recurred in either of the 2 patients in whom ablation was performed in channels identified from a map during atrial pacing, suggesting that a high-resolution map during atrial pacing may be as useful as a map during MacroAT.
Limitations of the Study
The principal limitation of the present study was the time required to obtain high-density maps by use of this point-by-point technique. This limitation is mitigated somewhat by the ability to identify all channels from a single map during 1 tachycardia or pacing.
Drs Nakagawa and Jackman serve as consultants to Biosense Webster.
- Received July 19, 2000.
- Revision received September 22, 2000.
- Accepted September 26, 2000.
- Copyright © 2001 by American Heart Association
Triedman JK, Saul P, Weindling SN, et al. Radiofrequency ablation of intra-atrial reentrant tachycardia after surgical palliation of congenital heart disease. Circulation. 1995;91:707–714.
Kalman JM, VanHare GF, Olgin JE, et L. Ablation of “incisional” reentrant atrial tachycardia complication surgery for congenital heart disease. Circulation. 1996;93:502–512.