Prediction of Atrioventricular Block During Radiofrequency Ablation of the Slow Pathway of the Atrioventricular Node
Background Selective radiofrequency (RF) ablation of the slow pathway is an effective treatment for atrioventricular (AV) nodal reentry tachycardia. A previous report showed that rapid junctional tachycardia (JT) caused by RF associated with loss of ventriculoatrial (VA) conduction is related to increased risk for AV block. However, this can be difficult to detect during energy delivery, and more importantly, it cannot be measured before the onset of RF energy delivery. The aim of our study was to determine whether measurements made from electrograms could be used to predict the risk of AV block before RF energy is delivered.
Methods and Results Fifty-eight patients underwent 63 selective slow pathway RF ablation procedures. In 46 (26.9%) of 172 JTs caused by RF, VA block was observed, and in 11 this was followed by AV block of various degrees. Electrograms before each application of RF were analyzed for the interval between the atrial signals in the His bundle catheter and in the distal mapping catheter [A(H)-A(Md)], the interval between the atrial signals in the His bundle catheter and in the proximal coronary sinus catheter [A(H)-A(CS)], the AV ratio, and the presence of a slow pathway potential or a fractionated atrial signal in the distal mapping catheter. Mean cycle length (CL) of JT was calculated if it consisted of at least 10 beats. These parameters were compared between patients with JT who developed VA block and subsequent AV block (group 1), patients with JT and VA block but without subsequent AV block (group 2), and patients with JT without VA block (group 3). The A(H)-A(Md) interval was significantly shorter in group 1 (17±8 ms) than in groups 2 (33±8 ms, P<.001) and 3 (32±10 ms, P<.001), whereas the A(H)-A(Md) intervals of groups 2 and 3 did not differ from each other. CL of JT, A(H)-A(CS) interval, AV ratio, presence of a slow pathway potential, or a fractionated atrial electrogram were not related to the occurrence of AV block.
Conclusions The A(H)-A(Md) interval provides an electrophysiological marker that can be used in addition to the radiological catheter position to assess the risk for AV block before onset of RF delivery. CL of JT and occurrence of VA block are not related to the risk of AV block.
Selective RF ablation of the slow pathway of the AV node is a well-established and highly effective treatment for paroxysmal AVNRT. However, it carries a small risk of creating heart block, with the consequence of pacemaker dependency, in patients who are often young.1 2 3
JT is frequently observed during RF ablation of the slow pathway, although it is also seen during fast pathway modification and total AV node ablation. It seems to be a response to thermal injury of the compact AV node and/or the perinodal tissue forming the input of fast and slow pathway into the AV node4 5 6 and is a valuable marker for the effectiveness of RF energy delivery.
Thakur et al7 reported that a relatively fast rate of JT and loss of VA conduction are associated with an increased risk of inadvertent AV block. However, these markers for impending heart block do not allow the risk of AV block to be determined before RF energy delivery. AV block may occur concurrently with the development of retrograde fast pathway block during JT. Therefore, even if RF energy is discontinued immediately, disturbance of anterograde AV conduction may result. The aim of our study was to identify a more reliable parameter that can be used as a predictor of the risk of inadvertent heart block before RF energy delivery is started.
Fifty-eight consecutive patients, 44 women and 14 men (mean age, 43±17 years; range, 8 to 88 years), with the common type of AVNRT who underwent 63 selective RF ablation procedures of the slow pathway were studied. All patients had a physical examination, 12-lead resting surface ECG, and two-dimensional echocardiogram without any evidence for structural heart disease.
After informed written consent had been obtained, electrophysiological testing was performed, with antiarrhythmic drugs being discontinued for at least five half-lives. Diazepam was used for sedation, lidocaine for local anesthesia, and morphine for systemic pain relief. The diagnosis of common AVNRT was confirmed by our standard technique (incremental atrial and ventricular pacing and programmed stimulation using at least two basic CLs in the atrium and the ventricle). Isoproterenol was used if necessary to induce sustained AVNRT.
RF ablation targeting exclusively the slow pathway of the AV node was performed with a 7F quadripolar steerable 4-mm-tip electrode ablation catheter. Catheter location was confirmed by monoplanar or biplanar fluoroscopy in the RAO and LAO views. Ablation was guided primarily by a search for slow pathway potentials or a fractionated atrial electrogram at the basis of the triangle of Koch, but anatomic considerations were also used for identification of a suitable site for RF ablation. RF energy pulses of 20 to 30 W for a maximal duration of 30 seconds were delivered between the tip electrode of the ablation catheter and a large cutaneous patch on the posterior left chest with an HAT 200S (Oscor Medical Corp) or an Atakr (Medtronic Corp) RF energy generator. Only RF energy deliveries during sinus rhythm inducing JT or at least single junctional beats were studied; there were no episodes of AV block without prior occurrence of JT. Ablation during AVNRT or initiation of AVNRT within 5 seconds after onset of RF energy application was excluded. Episodes with junctional rhythm or pacing at any site immediately before RF delivery were also excluded from analysis. JT or single junctional beats were identified by low to high right atrial activation and, in the case of tachycardia, by irregularity. All episodes of junctional rhythm caused by RF energy during sinus rhythm, even if only a single beat was noted, were analyzed for VA conduction. If loss of VA conduction was observed, RF delivery was stopped immediately. RF energy application was repeated until JT ceased or slowed appreciably.
The mean CL of the JT preceding loss of VA conduction was calculated only if it consisted of at least 10 beats. Electrograms before ablation were analyzed for the presence of slow pathway potentials or a fractionated atrial electrogram in the signal recorded from the distal electrodes of the mapping catheter. The interval between the atrial component of the His bundle electrogram (when there was a distinct and stable His bundle potential) and the atrial signal of the distal mapping catheter [A(H)-A(Md) interval] was measured. In addition, the interval between the atrial component of the His bundle electrogram and the atrial signal in the proximal CS catheter positioned near the CS ostium [A(H)-A(CS) interval] was measured. Analysis of the A(H)-A(CS) interval was performed only if a proximal position of the catheter in the CS could be confirmed by fluoroscopy. Patients with variation of the A(H)-A(CS) interval during the ablation procedure of more than 5 ms were also excluded from analysis of this interval, even if the stored images showed a correct, proximal position of the CS catheter, because variation of the interval was considered to be evidence for lack of catheter stability. The AV ratio recorded in the distal mapping catheter was also analyzed. If the ratio showed beat-to-beat inconsistency, the mean value of the last five recorded beats was calculated. The type of AV block following the energy delivery was identified. The patients were divided into three groups: group 1, patients with VA block during JT and consequent AV block; group 2, patients with VA block during JT but without consequent AV block; and group 3, patients without VA block during JT and without consequent AV block. AV block was considered transient if AV nodal conduction returned to normal values within 30 minutes.
To standardize the analysis of signals, only ablations using a computerized electrophysiology system (Bard EP LabSystem, Bard Electrophysiology) were studied. The intracardiac intervals were measured at a speed of 100 mm/s and a gain setting of 16. If the signal in any of the tracings was too small to measure at this amplification, the gain was increased to 32 in all tracings. The onset of an intracardiac activation was defined as deviation of the tracing by >45° from the isoelectric. Control by a second investigator revealed a difference of ±5 ms in the measured intervals.
The fluoroscopic position of the ablation catheter was defined by the scheme introduced by Jazayeri et al,8 dividing the triangle of Koch marked by His bundle and CS catheter in the RAO 30° view into six areas, labeled A1, A2, M1, M2, P1, and P2 from anterior to posterior. Stored fluoroscopic images were not available for each site of RF energy delivery. The image for the last successful delivery was available for analysis in 36 cases (57%), and these ablation catheter positions were correlated to the A(H)-A(Md) interval. In addition, the absolute distance of the tip of the ablation catheter from the pair of electrodes on the His bundle catheter recording a distinct and stable His bundle signal was correlated to the A(H)-A(Md) interval. This distance was measured only if one of the catheters introduced was oriented exactly in the RAO. The known interelectrode spacing of this catheter was used as a scale for the measurement. For this reason, images of only 17 cases (27%) were suitable for analysis.
Patients were followed for 2 to 3 months in our outpatient clinic and thereafter by their referring physicians. Follow-up included history, physical examination, and 12-lead surface ECG. If there were any rhythm-related symptoms, ambulatory Holter monitoring was performed.
Univariate and multivariate logistic regression analyses were performed to define the determinants of AV block. In addition, comparisons between different groups were performed with two-tailed Student’s t test, one-way ANOVA, and ANCOVA (using the CL of sinus rhythm before RF delivery as the covariate) for the continuous variables and χ2 test for the categorical variables. A least-squares linear regression analysis was performed to examine the relationship between catheter position and A(H)-A(Md) interval. A value of P<.05 was considered statistically significant. The data are expressed as mean±SD unless otherwise indicated.
The characteristics and outcome of the patients included in our study are shown in Table 1⇓. Of a total of 179 episodes of junctional ectopy or tachycardia caused by RF delivery at 126 distinct catheter positions, 172 (mean, 3.4±3.3 per patient) were suitable for analysis. Twelve episodes were excluded from analysis because of junctional rhythm or pacing at the start of RF energy delivery. In 46 episodes (26.7%), loss of VA conduction was noted, and RF energy delivery was stopped immediately if VA block was recognized during the energy application. Eleven of these episodes (23.9% of JTs with VA block, or 6.4% of all JTs) were followed by heart block of various degrees. First-degree AV block was observed in 3 cases, second-degree block in 2 cases, and third-degree block on 6 occasions. In only 1 case did first-degree heart block persist; in all other cases, normal conduction was restored, with a PR interval <200 ms within 30 minutes of RF energy delivery. Permanent heart block during the follow-up period did not occur (Table 1⇓).
Logistic regression analysis revealed that the A(H)-A(Md) interval was the only significant predictor of AV block (P<.0001) even if adjusted for all the other variables analyzed (P<.01). The A(H)-A(Md) interval in group 1 (17±8 ms; range, 5 to 28 ms) was significantly shorter than in group 2 (33±8 ms; range, 14 to 52 ms; P<.001) or group 3 (32±10 ms; range, 12 to 78 ms; P<.001) (Table 2⇓, Fig 1A⇓). The A(H)-A(Md) intervals of group 2 and group 3 did not differ significantly from each other. The result remained the same even when adjusted for heart rate before RF delivery. Representative examples of short and long A(H)-A(Md) intervals are shown in Fig 2⇓. As shown in Fig 3⇓, the risk of AV block was significantly increased if the A(H)-A(Md) interval was <20 ms.
Groups 1, 2, and 3 did not differ from each other with respect to the number of RF applications, A(H)-A(CS) interval, AV ratio, registration of a slow pathway potential or a fractionated atrial electrogram, or mean CL (Table 2⇑, Fig 1B⇑ and 1C⇑). Analysis based on calculation of a mean value of all episodes for each patient to avoid influences of patient-specific properties of the AV node revealed no differences compared with the results based on analyses of all single episodes.
In four patients, a comparison could be made between RF energy delivery that resulted, on the one hand, in episodes of JT with loss of VA conduction and subsequent AV block, and on the other hand, episodes that were not followed by AV block. The number of patients was too small to permit statistical analysis. However, the range of A(H)-A(Md) intervals was 5 to 20 ms when there was AV block and 22 to 36 ms when not followed by AV block.
Although RF energy delivery was discontinued immediately when loss of VA conduction during JT was recognized, our retrospective analysis revealed 13 of a total of 46 episodes of JT with VA block (28.3%) in which VA block was not recognized during the ablation procedure. Therefore, in these cases, RF energy delivery had been completed despite VA block. The mean value of the A(H)-A(Md) interval before these episodes (33±5 ms) and the mean value of the A(H)-A(Md) interval preceding episodes with loss of VA conduction followed by immediate cessation of energy delivery (32±9 ms) were not significantly different. None of the unrecognized episodes of VA block during JT was followed by impairment of AV conduction.
There were two episodes of transient heart block, one second-degree AV block and one complete AV block, that were not preceded by loss of VA conduction during JT and therefore were not included in our study. The A(H)-A(Md) intervals were 20 and 16 ms, respectively. One patient with permanent second-degree AV block was excluded because of pacing in the high right atrium during mapping before RF application. The A(H)-A(Md) interval during pacing in the high right atrium before onset of RF delivery followed by heart block was 18 ms. The A(H)-A(Md) interval during sinus rhythm was not available for analysis in this episode.
Because of our practice of storing the fluoroscopic image of the ablation catheter only at the site that resulted in successful ablation of tachycardia, there were only five patients in whom the catheter position at a site that resulted in AV block could be compared with our electrophysiological measurements. The tip of the ablation catheter was found in areas M1, M2, and P1, respectively, according to the scheme introduced by Jazayeri et al.8 The correlation between A(H)-A(Md) interval and radiological ablation catheter position was not significant (r=.32, P=.054) (Fig 4A⇓). The correlation between A(H)-A(Md) interval and the absolute distance of the distal mapping catheter from the His bundle catheter was significant (r=.61, P=.005) (Fig 4B⇓).
Cycle Length of Junctional Tachycardia
JT caused by RF energy application is characteristically irregular, and even during successful RF applications, it may occur only in short runs, making it difficult to estimate the mean CL during the ongoing RF delivery. In some episodes, JT may accelerate a few beats before loss of VA and AV conduction, therefore being a very late marker of impending AV block. Thakur et al7 reported a series of 59 ablation procedures in 53 consecutive patients in which they studied the characteristics of JT caused by either fast pathway or slow pathway ablation with and without subsequent heart block. JT followed by AV block was characterized by a shorter CL and loss of VA conduction, but they did not describe how the rate of this typically irregular tachycardia was assessed. In our series, we found that only a small proportion of episodes of JT consisted of at least 10 beats and were therefore sufficient for calculation of a mean CL. In addition, we were not able to demonstrate a relation between mean CL of JT and occurrence of heart block (Table 2⇑, Fig 1C⇑). A study by Natale et al9 also did not reveal a statistically significant difference in CL of JT with and without subsequent heart block. This discrepancy might be due in part to the different groups of patients used for analysis. Thakur et al included fast pathway and slow pathway ablation procedures, whereas in the study of Natale et al and also our study, only slow pathway ablation procedures were analyzed.
For these reasons, we consider analysis of CL of the JT not to be a reliable and safe parameter for the risk of impending AV block.
Loss of VA Conduction
Loss of VA conduction during JT is a more useful marker, but even loss of retrograde conduction for only one single beat of JT, followed by immediate cessation of RF energy delivery, was followed by AV block in some of our patients. In our study, the proportion of JTs with loss of VA conduction appears to be relatively high. This may be because it is impossible to discriminate between dynamic changes in VA conduction due to progressive damage to the AV conduction system and VA dissociation due to the CL of JT being shorter than the retrograde effective refractory period of the fast pathway. We did not routinely study the retrograde effective refractory period of the fast pathway before RF delivery. However, even if the retrograde effective refractory period had been assessed before RF application, it would be difficult to discriminate between true VA block and VA dissociation during an ongoing RF delivery and to base the decision to stop or to continue the RF delivery thereon. In addition, in our study only 11 (23.9%) of 46 episodes of JT with loss of VA conduction were followed by impairment of AV conduction. This is consistent with the report of Natale et al,9 who in their series of 20 patients with transient heart block found only 9 patients (45%) with JT and loss of VA conduction preceding heart block but also observed a large number of JTs with loss of VA conduction without consequences for AV nodal conduction. In a fashion analogous to the mean CL of JT, loss of VA conduction is also a marker that can occur too late to recognize the increased risk of producing inadvertent heart block. Additionally, heart block occurred in two patients in our series without preceding loss of VA conduction during JT. Intriguingly, the heart block in both of these patients was transient.
Blanck et al10 reported a series of 19 patients, of whom 4 developed complete heart block undergoing selective fast pathway RF ablation. They studied the A(H)-A(CS) interval during retrograde activation of the fast pathway before applying RF energy. They found that an A(H)-A(CS) interval of ≤10 ms was associated with the development of complete heart block. They concluded that a short A(H)-A(CS) interval reflects a more posteriorly located fast pathway in closer proximity to the slow pathway input into the compact AV node. In this case, lesions created to ablate the fast pathway may damage both the fast and the slow pathways, resulting in AV block. We measured this interval during sinus rhythm. However, our analysis revealed no relation between the A(H)-A(CS) interval and the risk of causing heart block by selective RF ablation of the slow pathway (Table 2⇑, Fig 1B⇑).
We postulated, however, that the A(H)-A(Md) interval should reflect the distance of the distal electrodes of the mapping catheter to the region of the compact AV node marked by the His bundle catheter. The longer the interval, the greater the distance between the mapping catheter and the compact AV node should be. Therefore, a long A(H)-A(Md) interval should predict a low risk of creating inadvertent AV block. According to this hypothesis, the A(H)-A(Md) interval would be significantly shorter before RF energy deliveries that resulted in AV block. The ranges of A(H)-A(Md) intervals measured in patients with AV block after energy delivery and in patients without impairment of AV conduction after energy delivery were sharply demarcated, with only very few overlaps (Figs 1A⇑ and 4⇑). This makes the A(H)-A(Md) interval a very reliable and clinically valuable parameter for assessing the a priori risk of inducing damage to anterograde AV conduction.
The A(H)-A(Md) intervals in the two RF energy deliveries leading to transient heart block without preceding VA block during JT were 20 and 16 ms, respectively, and were similar to the A(H)-A(Md) interval of group 1. Thus, this interval also seems to allow prediction of occurrence of heart block even in the absence of VA block during JT. The A(H)-A(Md) interval in a patient with second-degree AV block (18 ms) who was excluded from our study because of pacing in the high right atrium before RF delivery was also in the range of group 1. Since there may be similar inputs into the AV node during sinus rhythm as well as during high right atrial pacing, the A(H)-A(Md) interval may also be suitable for prediction of the risk for heart block during pacing.
As mentioned above, when loss of VA conduction during JT was recognized, RF energy delivery was stopped immediately. For these episodes, we are not able to determine the effect that would have resulted on anterograde conduction had RF energy delivery been continued. However, in 13 (28.3%) of a total of 46 episodes of JT with loss of VA conduction without subsequent AV block, the occurrence of VA block was not recognized at the time, and energy delivery was completed. The fact that the A(H)-A(Md) interval before these episodes (33 ms) and the A(H)-A(Md) interval before episodes with VA block followed by immediate cessation of RF energy (32 ms) were similar suggests that if the A(H)-A(Md) interval exceeds 30 ms, energy delivery can be completed safely despite loss of VA conduction during JT. On the other hand, energy delivery at a site at which the A(H)-A(Md) interval is <20 ms signifies a high risk for AV block (Fig 3⇑).
Correlation of A(H)-A(Md) Interval and Radiological Ablation Catheter Position
Our practice has been that, in general, only the radiographs corresponding to the ultimately successful ablation sites were available for analysis. Therefore, the finally successful and therefore radiographically stored ablation catheter position was associated with the induction of AV block in only five patients. Comparison of A(H)-A(Md) intervals and successful radiological ablation catheter positions by the scheme introduced by Jazayeri et al8 revealed only a poor correlation between these two (r=.32, P=.054) (Fig 4A⇑). The ranges of A(H)-A(Md) intervals measured in patients with AV block after energy delivery and in patients without impairment of AV conduction after energy delivery were sharply demarcated, with only very few overlaps, whereas radiological ablation catheter positions resulting in AV block and those not resulting in AV block were not well demarcated. In contrast, the correlation between A(H)-A(Md) interval and the absolute distance of the distal mapping catheter from the His bundle catheter was significant (r=.61, P=.005) (Fig 4B⇑). However, the number of episodes that could be correlated was small because of limitations of the measurement of an absolute distance from two bidimensional planes (RAO and LAO), which may not reflect the full three-dimensional complexity of this area. Consistent with our analysis using the scheme of Jazayeri et al,8 the radiological catheter positions resulting in impairment of AV conduction and those that did not were not well demarcated.
Therefore, the A(H)-A(Md) interval appears to be superior to the radiological ablation catheter position in predicting AV block after energy delivery.
This study was a retrospective analysis of our experience in slow pathway ablation for AV nodal reentrant tachycardia. The use of electrophysiological measures to indicate anatomic proximity can be confounded by a variety of other factors, including fiber orientation, conduction time, and catheter orientation. None of these procedures were performed with temperature-controlled RF energy delivery, which would be expected to produce more uniform lesions between patients. It could be argued that making the measurement we have suggested is simply a complicated method of determining whether the mapping catheter is more or less anteriorly or posteriorly located. That this is too simplistic a view of the complex electrophysiological relationships in this region is suggested by our finding that there was not a good correlation between the A(H)-A(Md) interval and the anatomic position of the mapping catheter as indicated by fluoroscopy.
Although this was not a prospective study, we subsequently used the A(H)-A(Md) interval as a predictor for AV block in 13 RF slow pathway ablation procedures. RF energy was not delivered if the A(H)-A(Md) interval was <20 ms. None of these procedures produced either transient or permanent AV block. It would be possible to evaluate the use of this measurement as a predictor for AV block in a prospective study. However, provided that the success rate for slow pathway ablation can be maintained, as in our series, there seems to be little to gain.
Our study suggests that the A(H)-A(Md) interval provides an electrophysiological marker for the distance of the ablation catheter from the compact AV node that can be used, in addition to the radiological ablation catheter position, to assess the risk for AV block caused by selective RF ablation of the slow pathway before onset of RF delivery. An A(H)-A(Md) interval of <20 ms is associated with increased risk of inadvertent AV block. In contrast, junctional beats or JT does not reliably herald impending heart block, even when VA block occurs, if the A(H)-A(Md) interval exceeds 30 ms (Fig 3⇑). The CL of JT caused by RF delivery is not related to the risk of AV block. Occurrence of VA block during JT is very common even in the absence of subsequent AV block and also is not related to the risk of AV block.
Our data do perhaps provide some further support for the suggestion that the slow pathway can be interrupted at various points along its length, a more anterior ablation approach simply increasing the risk of inadvertent AV block.
Selected Abbreviations and Acronyms
|AVNRT||=||AV nodal reentry tachycardia|
|LAO||=||left anterior oblique|
|RAO||=||right anterior oblique|
Dr Hintringer was supported by a grant from the Hans und Blanca Moser Foundation, Vienna, Austria. We are grateful to Dr Joseph Kautzner, Dr Ravi Kishore, and Dr Yukihiko Momiyama for very valuable discussions and to Ian Little for assistance in collecting data.
Presented in part at the 16th Annual Scientific Sessions of the North American Society of Pacing and Electrophysiology, Boston, Mass, May 3-6, 1995, and the 17th Congress of the European Society of Cardiology, Amsterdam, Netherlands, August 20-24, 1995.
- Received February 22, 1995.
- Revision received August 1, 1995.
- Accepted August 6, 1995.
- Copyright © 1995 by American Heart Association
Jackman WM, Beckman KJ, McClelland JH, Wang X, Friday KJ, Roman CA, Moulton KP, Twidale N, Hazlitt A, Prior MI, Oren J, Overholt ED, Lazzara R. Treatment of supraventricular tachycardia due to atrioventricular nodal reentry by radiofrequency catheter ablation of slow-pathway conduction. N Engl J Med. 1992;327:313-318.
Haissaguerre M, Gaita F, Fischer B, Commenges D, Montserrat P, d’Ivernois C, Lemetayer P, Warin J-F. Elimination of atrioventricular nodal reentrant tachycardia using discrete slow potentials to guide application of radiofrequency energy. Circulation. 1992;85:2162-2175.
Jackman WM, Wang X, Friday KJ, Fitzgerald DM, Roman C, Moulton K, Margolis PD, Bowman AJ, Kuck K-H, Naccarelli GV, Pitha JV, Dyer J, Lazzara R. Catheter ablation of atrioventricular junction using radiofrequency current in 17 patients: comparison of standard and large-tip catheter electrodes. Circulation. 1991;83:1562-1576.
Jazayeri MR, Hempe SL, Sra JS, Dhala AA, Blanck Z, Deshpande SS, Avitall B, Krum DP, Gilbert CJ, Akhtar M. Selective transcatheter ablation of the fast and slow pathways using radiofrequency energy in patients with atrioventricular nodal reentrant tachycardia. Circulation. 1992;85:1318-1328.
Natale A, Sra J, Jazayeri M, Blanck Z, Wase A, Bremner S, Koch K, Akhtar M. Clinical significance of AV nodal block during radiofrequency modification with the posterior approach. Circulation. 1994;90(suppl I):I-126. Abstract.
Blanck Z, Makal J, Hanson L, Strommen P, Jazayeri M, Akhtar M. Prediction of atrioventricular block during radiofrequency ablation of the fast pathway in atrioventricular nodal reentry. Circulation. 1991;84(suppl II):II-235. Abstract.