Aman Chugh, MD
A 76-year-old man was referred to for catheter ablation of atrial tachycardia (AT) that occurred after an ablation procedure of atrial fibrillation (AF) performed 9 months ago. Four years ago, he underwent mitral and aortic valve replacement (both porcine) for rheumatic heart disease (ejection fraction, 45%; left atrial [LA] diameter, 6.1 cm). He developed postoperative AF for which he was prescribed amiodarone. He remained free of AF for approximately 18 months at which time he developed symptomatic-persistent AF. He was pretreated with amiodarone and underwent transthoracic cardioversion. However, sinus rhythm only lasted for 2 weeks. He then underwent catheter ablation of persistent AF, which consisted of antral pulmonary vein (PV) isolation, ablation of complex fractionated electrograms, and linear ablation at the left atrial (LA) roof.
The electrocardiogram (ECG) during AT showed that the initial portion of the flutter wave in the inferior leads was slightly negative, followed by notching (Figure 41-1). In addition, apart from lead V1, the remaining precordial leads were isoelectric. This pattern is consistent with an LA origin. Insertion of a multipolar catheter into the coronary sinus (CS) revealed distal-to-proximal activation, verifying LA origin. The cycle length of the tachycardia was 265 ms. A high-density activation map of the LA using more than 500 points confirmed macroreentry around the mitral valve in a clockwise fashion (Figure 41-2). The mechanism was confirmed by entrainment mapping around the mitral valve (Figure 41-3). Linear ablation was commenced from the lateral annulus to the left-sided PVs. Radiofrequency (RF) energy slowed and terminated the tachycardia to sinus rhythm. Pacing from anterior to the ablation line showed a proximal-to-distal activation of the CS, consistent with bidirectional block. Programmed atrial stimulation and rapid atrial pacing during isoproterenol infusion failed to reveal any arrhythmias. Eighteen months after the ablation procedure, there is not evidence of recurrent arrhythmia on long-term ambulatory monitoring. His symptoms of dyspnea and fatigue, along with the ejection fraction, have also improved.
FIGURE 41-1 Electrocardiogram during atrial tachycardia (AT) that occurred following catheter ablation of atrial fibrillation. The lack of a “saw-tooth” morphology of the P waves in the inferior leads and the presence of isoelectric P waves in the lateral precordial leads point to a left atrial (LA) origin. Also, note the notching (*) after the initial negative deflection (arrow).
FIGURE 41-2 Activation map of the LA constructed using more than 500 points suggests that the mechanism of the AT (cycle length = 265 ms) is macroreentry around the mitral valve in a clockwise direction. The color spectrum (261 ms) shows that mapping accounted for almost the entire cycle length of the tachycardia, confirming a macroreentrant mechanism. The appendage (LAA) has been removed for clarity. Abbreviations: LI, left inferior; LS, left superior; PV, pulmonary vein; RS, right superior.
FIGURE 41-3 Entrainment mapping from the mitral isthmus (region between the lateral mitral annulus and the left inferior PV). The tachycardia is accelerated to the pacing rate, and the postpacing interval upon cessation of pacing approximates the cycle length of the tachycardia, confirming the findings on activation mapping shown in Figure 41-2. Abbreviations: Abl, ablation; CS, coronary sinus.
Patients undergoing catheter ablation of atrial fibrillation (AF) may develop atrial tachycardia (AT) during follow-up.1,2 The incidence of postablation AT depends on a number of factors, including the underlying atrial substrate, the ablation approach, and possibly treatment with antiarrhythmic medications. In patients undergoing only pulmonary vein isolation, the incidence of postablation AT is low.3Patients who undergo empiric linear ablation, the incidence is about 33%,2 with about one-fourth of the patients requiring a repeat ablation procedure for AT. AT is quite common when a step-wise approach is utilized in an attempt to acutely terminate AF during radiofrequency ablation. In fact, about 40% to 50% of such patients require a repeat ablation for AT.4
FIGURE 41-4 (A) Termination of mitral isthmus-dependent flutter during radiofrequency energy delivery along the line (black circles in Figure 41-2). (B) Pacing anterior to the line (from the region of the LAA) results in a proximal-to-distal activation of the coronary sinus, confirming the presence of linear block at the mitral isthmus. Abbreviation: S, stimulus.
ETIOLOGY AND PATHOPHYSIOLOGY
Broadly speaking, there are two main hypotheses that help explain the emergence of AT after catheter ablation of AF.5 It is possible that postablation AT represents an underlying “driver” that is only unmasked after eliminating AF or more precisely, fibrillatory conduction. A prior study using spectral analysis showed that the frequency (ie, activation rate) of the resultant AT after elimination of AF matched the frequency of a spectral component that was present during the initial periodogram during AF.6 This finding suggests that the substrate for AT—macroreentrant in the majority—is already present during AF and that the AT is destined to be encountered at some point. Although this hypothesis is certainly plausible, it may not be applicable to all patients. The other hypothesis posits that postablation AT is a result of the ablation approach, that is, a proarrhythmic complication of the initial AF procedure. There is a wealth of data that supports this contention. First, it is well known that both antral PVI and circumferential PV ablation (ie, circular lesions around the PVs, along with linear lesions at the mitral isthmus and the posterior wall/roof) are very effective in eliminating AF in patients with paroxysmal AF.7However, the former does so without intervening macroreentrant AT, which is quite common with the latter approach. So if patients were destined to develop macroreentrant AT, why should they only do so with one technique and not both? In fact, one could argue that they should be less likely to develop macroreentrant AT with an approach that involves linear lesions than with one that does not. Second, a recent study showed that the area of slowest conduction velocity, requisite for the development of macroreentry, during either roof or perimitral flutter was found at the precise location where ablation was performed during the initial AF procedure.8 In other words, the area of slowest velocity during a roof- or mitral-isthmus-dependent AT was identified at the most cranial aspect of the roof, and the lateral mitral annulus, respectively. If macroreentrant AT were due to the underlying substrate, why would the area of slowest conduction be confined to the site that was ablated, and not other sites along the reentrant path? When sources of AF are mapped, as opposed to an empiric lesion set for all patients, ablation of AF often yields sinus rhythm without an intervening atrial tachycardia.9 Lastly, a prior study showed that >95% reentrant AT following circumferential PV ablation arose from the prior ablation targets/lines for AF.10 Therefore, AT in most patients is likely related to a proarrhythmic effect of ablation.
Establishing a diagnosis of postablation AT is straightforward. The diagnosis can be made on the ECG or ambulatory monitoring, or by analysis of stored electrograms in the patient with an implantable device. The ECG shows evidence of organized atrial activity, and that the P or flutter waves are identical with a consistent cycle length. This may require careful analysis of the ECG to distinguish between organized AF and AT. If the P waves are even slightly dissimilar in rate or appearance, the rhythm is likely AF and not AT. Not infrequently, the P waves may not be apparent due to low amplitude or are obscured by the QRS or T waves during rapid ventricular rates. An ECG during postablation AT is sometimes misdiagnosed as sinus tachycardia. A closer inspection of the ECG usually shows a P wave that is obscured by the QRS/T wave. Resting sinus tachycardia, in the absence of a coexisting illness in these patients—recall that patients with persistent AF have evidence of structural and sinus node remodeling—is uncommon.
The ECG may also be helpful in elucidating the mechanism of the atrial tachycardia. The P-wave duration during macroreentrant tachycardias (ie, “atrial flutter”) is longer owing to continuous atrial activation. A recent study suggested that a cut-off value >185 ms is consistent with macroreentry.8 During focal tachycardias or in situations in which reentry is confined to a small region of the atrium (ie, localized reentry or small reentrant circuits), the activation time is shorter, and the diastolic interval is longer as compared to large circuits. These features result in a shorter P-wave duration and a presence of an isoelectric interval between successive P waves. It should be stressed that in atria with multiple areas of conduction block or extensive scarring, these observations may not be applicable.
Not all patients with post-AF atrial tachycardia need to undergo a repeat ablation procedure. It is not unreasonable to perform transthoracic cardioversion (for persistent AT) or recommend antiarrhythmic medications (for paroxysmal AT) in the acute phase (within the first 3 months) following the AF procedure. In some of these patients, AT may not recur during long-term follow-up.2 For AT that occurs after 3 months of the AF procedure, our practice is to recommend catheter ablation. Note that most of these patients were referred to for the initial AF procedure after failure or intolerance of antiarrhythmic medications. Thus, to prescribe antiarrhythmic medications on an indefinite basis is not an attractive option.
We prefer to perform the repeat procedure for AT during the arrhythmia, as opposed to sinus rhythm, as some clinical ATs may not be readily inducible in the electrophysiology laboratory. If the clinical arrhythmia is not inducible, the patient may experience recurrence, despite elimination of possible triggers (eg, reisolation of connected PVs) or empiric linear ablation at the major isthmi. Patients with paroxysmal tachycardia while taking antiarrhythmic medications are instructed to discontinue rhythm-controlling drugs a few weeks prior to the scheduled procedure. If the arrhythmia fails to recur, the procedure is postponed. Amiodarone should be discontinued at least 2 months prior to the ablation procedure given its long half-life. Another reason to discontinue antiarrhythmic medications prior to the ablation procedure is the possibility that the drug may be masking AF. The ablation procedure may eliminate AT, but underlying drivers may not be manifest and hence cannot be mapped and ablated.
One should be aware of the rapid atrioventricular (AV) nodal conduction that may occur in patients with AT as compared to patients with AF. As patients are asked to discontinue rate-controlling (and of course, rhythm-controlling) medications in preparation for the procedure, unchecked AV nodal conduction may result in tachycardia. Occasionally, 1:1 AV nodal conduction may occur leading to hemodynamic collapse and syncope, and may require urgent transthoracic cardioversion. In such patients, our practice is to admit them to the hospital a day prior to the procedure allowing for drug washout and initiation of intravenous rate-controlling medications as needed.
Our preference is to perform the procedure on therapeutic oral anticoagulation to prevent not only thromboembolism but also access site complications, which are more likely related to intravenous or subcutaneous heparin. We ask patients who are taking one of the novel oral anticoagulants (dabigatran, rivaroxaban, or apixaban) to discontinue the drug 36 hours prior to the procedure. All patients who present to the electrophysiology laboratory in AT undergo transesophageal echocardiography to rule out LA thrombus prior to the procedure.
Mapping and Ablation of AT
A question often arises whether to map the right atrium first prior to performing transseptal catheterization or whether LA mapping may be obviated in some patients altogether. From a practical standpoint, all patients who have previously undergone LA ablation of AF deserve LA mapping at the repeat session to ensure that the PVs remain disconnected and that any linear lesions that were deployed previously are complete. Nonetheless, there are some clues on the ECG that suggest a right atrial source. Although cavotricuspid isthmus dependent flutter is not uncommonly seen after AF ablation, its ECG signature is frequently obscured by the attendant LA ablation. The end result is that the ECG may not show the classic saw-tooth pattern of the flutter waves in the inferior leads.11 Initial negativity of the flutter waves, especially in the absence of notching, is suggestive a right atrial source. Further, a transition of positive to negative in the precordial leads is also suggestive (but not diagnostic) of flutter from the right atrial isthmus. Isoelectric flutter waves across the precordium are also suggestive of an LA origin.
A minimum of two venous sheaths is required for mapping and ablation of post-AF atrial tachycardia. A multipolar catheter is placed into the coronary sinus (CS). A distal-to-proximal activation confirms the presence of a LA source, whereas a proximal-to-distal activation is not as informative. The latter could be consistent with a right atrial, septal, or even a lateral left atrial origin. For example, during counterclockwise mitral isthmus dependent flutter (which, unlike typical flutter, is equally prevalent as the clockwise version) or even during AT from the lateral LA in a patient with prior mitral isthmus conduction block, the CS is activated in proximal-to-distal fashion.
After LA access, the operator has a choice to map the AT using activation mapping using a 3-dimensional mapping system or entrainment mapping. There are advantages and disadvantages with either approach. A criticism of activation mapping is that it is time consuming and needs to be repeated in case of multiple tachycardias, which is not uncommon. However, a high-density map using a multipolar catheter (either a ring catheter [Lasso]or Pentaray, Biosense Webster, Diamond Bar, CA) can be constructed within 10 to 15 minutes. Currently, however, the operator still needs to review each point to accurate annotation, which may be time consuming. Further refinements and other promising technologies should help reduce the need of user intervention in assigning timing to activation points.
The advantage of entrainment mapping is that it quickly informs the operator if the site in question is involved in the reentry circuit. If the postpacing interval (PPI) is within 20 to 30 ms of the tachycardia cycle length (without a change in atrial activation), the site in question is part of the reentry circuit. A PPI that approximates the TCL but with a change in atrial activation during pacing or a long PPI identifies a site that is not critical to the reentry circuit. A single “good” return cycle is not enough to establish the diagnosis or mechanism of the tachycardia. For large circuits, one has to show good PPIs from the opposite walls of the LA (anterior and posterior or superior and inferior in the case of roof flutter, and analogously around the mitral valve for perimitral flutter). The major drawback of such an approach is that pacing may change or terminate the tachycardia, which may not be inducible. If the clinical tachycardia cannot be reinduced, it may recur during follow-up. Occasionally, the return cycle may be significantly longer than the TCL despite that the fact target site lies in the reentrant path. As has been shown for typical flutter, delayed conduction may result in a spuriously long PPI.12 Another problem with entrainment mapping is that it presupposes that the mechanism of the AT is macroreentry. The other possible mechanisms, eg, small reentrant circuits (during which the reentrant activity is restricted to a small segment of the atrium) and focal tachycardias, are probably best targeted by identifying the diastolic or presystolic activity, respectively, on the ECG.
The prevalence of these various mechanisms probably depends on the ablation approach that was utilized during the AF procedure. If multiple linear lesions were deployed, macroreentry is a good bet. In general, a macroreentrant mechanism involving the mitral isthmus, LA roof, the cavotricuspid isthmus, and other, less common areas is found in the majority of patients. In about 20% to 30% of patients, small reentrant or focal sources are found. For large circuits, the operator strives to design an ablation line that connects nonconductive areas, such as the two upper veins in the case of roof reentry, and lateral mitral valve to the left-sided PVs for perimitral reentry. The critical site within small reentrant circuits is characterized by very fractioned electrograms13 that cover about 50% of the TCL, and these sites are exquisitely responsive to radiofrequency (RF) energy or even pacing.8 For focal arrhythmias, the site with the earliest activation with respect to the P wave is targeted.
The endocardial mitral isthmus is typically ablated using about 35 watts of irrigated RF energy. The majority of patients with perimitral flutter require RF energy delivery within the distal CS (at 20 watts) for either tachycardia termination14 or creation of isthmus block or both. If a lateral approach is ineffective, an anterior approach (from the anterior annulus to the right-side PVs) or a superior approach (from the anterolateral annulus, along the left lateral ridge, to the left upper PV) usually results in tachycardia termination. After tachycardia termination, pacing from anterior to the line (usually from the LA appendage) is performed to determine whether linear block is present. With incomplete block, one observes the anticipated activation of the CS, that is, from distal-to-proximal. With block, the CS is activated from proximal-to-distal, with prolongation of the conduction time (Figure 41-5). Linear block at the mitral isthmus can be achieved in about 90% of patients. Recently, an alternative to CS ablation has been introduced in patients undergoing mitral isthmus ablation.15 The vein of Marshall, an atrial branch that is identified on CS venography, is injected with ethanol, which may reduce the amount of RF energy required (especially within the CS) for linear block.
FIGURE 41-5 Linear block at the mitral isthmus usually requires radiofrequency ablation within the CS, as shown in this tracing from another patient. Note that during LAA pacing, the activation of the CS abruptly changes from the expected (distal-to-proximal) to the unexpected (proximal-to-distal), heralding conduction block at the mitral isthmus. Concomitantly, there is a sudden increase in the mitral isthmus conduction time from 115 to 215 ms.
The left atrial roof is ablated using 25 to 30 watts of irrigated RF depending upon the degree of sheath support. If an initial line at the most cranial aspect of the roof is ineffective, a slightly anterior approach may be effective. If tachycardia or conduction persists, RF energy is delivered at the high posterior wall and even the low posterior wall at the level of the PV antra. With a more posterior approach, one has to be aware of the location of the esophagus, and the line may be altered as appropriate. Linear block is confirmed by observing ascending activation of the posterior LA during pacing from the appendage or sinus rhythm (Figure 41-6). Linear block at the roof may be demonstrated in about 95% of patients, but it may be difficult to distinguish from slow conduction.
FIGURE 41-6 Confirmation of linear block at the LA roof from the same patient as in Figure 41-5. During LAA pacing, the wavefront (solid red arrow) blocks at the roof, and then descends the anterior wall (solid, white, downward pointing arrow). The latter wavefront then ascends the posterior wall, confirming conduction block at the roof. With persistent conduction across the roofline, one would expect to seedescending activation of the posterior wall. Also shown are representative electrograms from the mid posterior LA (A) and high posterior LA, near the roofline (B). The increase in conduction time from A → B (110 to 150 ms) is consistent with ascending activation of the posterior wall and linear block at the roof. Also, note the presence of widely split double potentials (solid arrows on top electrogram) recorded on the line. In this patient, there was also block at the mitral isthmus (dashed arrows) LA.
Procedural End Points
After tachycardia termination and demonstration of linear block, the next step is to ensure that all PVs remain isolated. Then, we perform an induction protocol using programmed atrial stimulation and rapid atrial pacing during isoproterenol infusion (10-20 mcg/min). It is important to note that despite tachycardia termination, PV reisolation, and noninducibility, some patients still experience recurrence.8 In roughly one-half of the patients, recurrence is due to resumption of conduction along previously ablated lesions, and in the other half, emergence of a new tachycardia substrate that was not apparent at the previous session. To minimize the prevalence of the latter, we perform rapid atrial pacing to atrial refractoriness (to cycle lengths as short as 160 ms) during isoproterenol infusion, which may be the only way to induce ATs that may manifest themselves clinically during follow-up. With such an aggressive induction protocol, AF may occasionally be encountered. At this point, isoproterenol is discontinued to see if AF gives way to AT, which is then mapped and ablated. The clinical import of inducible AF that persists in this setting is not clear, but it is probably not unreasonable to perform cardioversion if the patient has not demonstrated AF clinically. Empiric ablation of the cavotricuspid isthmus is also reasonable to prevent recurrent arrhythmias.
Clinical AT can be eliminated in >90% of patients. Reasons for an unsuccessful ablation procedure include multiloop ATs that involve multiple circuits/mechanisms, extensive atrial scar, and involvement of the septum (especially involving the atrioventricular junction). Linear block at the roof and the mitral isthmus may be demonstrable in roughly 90% of patients. Nonetheless, 20% develop recurrence necessitating a repeat procedure.8 After multiple procedures, long-term freedom from AT/AF can be achieved in about 80% of patients without antiarrhythmic medications.8 The risk of a serious complication including vascular site issues that require transfusion or surgery, tamponade, or thromboembolism is about 1%. Anticoagulation is discontinued in most patients (except those with a prior stroke) in whom long-term ambulatory monitoring fails to show atrial arrhythmias. In case of procedural failure, other options include medical therapy (with either rhythm or rate-controlling medications) or AV junction ablation with a permanent pacemaker if the ventricular rates cannot be adequately managed. In the latter case, a pacemaker with atrial antitachycardia pacing is recommended as pacing is much more efficacious in terminating organized tachycardia as opposed to AF.
Some patients may not be able to distinguish symptoms during AT versus AF. However, others are particularly bothered by the higher heart rates that are present during AT as opposed to AF despite AV nodal medications. Some patients with facile AV nodal conduction may experience extremely fast heart rates during AT leading to syncope and may require urgent cardioversion. For these reasons and the fact that recurrences are common (even after an acutely successful procedure), patients with postablation AT require an experienced and dedicated team of outpatient clinical care coordinators, inpatient nurses and midlevel providers, and physicians to ensure optimal outcomes.
1. Mesas CE, Pappone C, Lang CC, et al. Left atrial tachycardia after circumferential pulmonary vein ablation for atrial fibrillation: electroanatomic characterization and treatment. J Am Coll Cardiol. 2004;44(5):1071-1079.
2. Chugh A, Oral H, Lemola K, et al. Prevalence, mechanisms, and clinical significance of macroreentrant atrial tachycardia during and following left atrial ablation for atrial fibrillation. Heart Rhythm. 2005;2:464-471.
3. Gerstenfeld EP, Callans DJ, Dixit S, et al. Mechanisms of organized left atrial tachycardias occurring after pulmonary vein isolation. Circulation. 2004;110:1351-1357.
4. O’Neill MD, Wright M, Knecht S, et al. Long-term follow-up of persistent atrial fibrillation ablation using termination as a procedural endpoint. Eur Heart J. 2009;30:1105-1112.
5. Chugh A. Atrial tachycardia after ablation of persistent atrial fibrillation: is it us or them? Circ Arrhythm Electrophysiol. 2013;6:1047-1049.
6. Yoshida K, Chugh A, Ulfarsson M, et al. Relationship between the spectral characteristics of atrial fibrillation and atrial tachycardias that occur after catheter ablation of atrial fibrillation. Heart Rhythm. 2009;6:11-17.
7. Oral H, Scharf C, Chugh A, et al. Catheter ablation for paroxysmal atrial fibrillation: segmental pulmonary vein ostial ablation versus left atrial ablation. Circulation. 2003;108: 2355-2360.
8. Yokokawa M, Latchamsetty R, Ghanbari H, et al. Characteristics of atrial tachycardia due to small vs large reentrant circuits after ablation of persistent atrial fibrillation. Heart Rhythm. 2013;10:469-476.
9. Narayan SM, Krummen DE, Shivkumar K, Clopton P, Rappel WJ, Miller JM. Treatment of atrial fibrillation by the ablation of localized sources: CONFIRM (Conventional Ablation for Atrial Fibrillation With or Without Focal Impulse and Rotor Modulation) trial. J Am Coll Cardiol. 2012;60:628-636.
10. Chae S, Oral H, Good E, et al. Atrial tachycardia after circumferential pulmonary vein ablation of atrial fibrillation: mechanistic insights, results of catheter ablation, and risk factors for recurrence. J Am Coll Cardiol. 2007;50:1781-1787.
11. Chugh A, Latchamsetty R, Oral H, et al. Characteristics of cavotricuspid isthmus-dependent atrial flutter after left atrial ablation of atrial fibrillation. Circulation. 2006;113:609-615.
12. Vollmann D, Stevenson WG, Luthje L, et al. Misleading long post-pacing interval after entrainment of typical atrial flutter from the cavotricuspid isthmus. J Am Coll Cardiol. 2012;59:819-824.
13. Jais P, Matsuo S, Knecht S, et al. A deductive mapping strategy for atrial tachycardia following atrial fibrillation ablation: importance of localized reentry. J Cardiovasc Electrophysiol. 2009;20:480-491.
14. Chugh A, Oral H, Good E, et al. Catheter ablation of atypical atrial flutter and atrial tachycardia within the coronary sinus after left atrial ablation for atrial fibrillation. J Am Coll Cardiol. 2005;46:83-91.
15. Baez-Escudero JL, Morales PF, Dave AS, et al. Ethanol infusion in the vein of Marshall facilitates mitral isthmus ablation. Heart Rhythm. 2012;9:1207-1215.