Introduction
Though atrioventricular nodal re-entry tachycardia remains one of the most common arrhythmias in the paediatric and young adult population, there is no consensus on the most appropriate technique for ablation site selection. Historically, practitioners select the ablation site based on anatomic landmarks and atrial signals within the Triangle of Koch. Reference Young and Niu1 To improve success and decrease complications, providers have been integrating mapping techniques to assist with the isolation of the slow pathway. Some of these techniques include assessing lower voltage areas, evaluating electrogram propagation and late activation during sinus rhythm, and assessing the propagation wave collision in the Triangle of Koch. Reference Bailin, Korthas, Weers and Hoffman2–Reference Van Aartsen, Law, Maldonado and Von Bergen7 Though traditional anatomic mapping and ablation of atrioventricular nodal re-entry tachycardia remain highly successful, a lack of research exists that critically compares various techniques for mapping and/or ablation for atrioventricular nodal re-entry tachycardia. When present, it is typically non-randomised and mostly retrospective. This, the first of its kind, prospective randomised trial sought to compare procedural characteristics and ablation outcomes between an anatomic approach for atrioventricular nodal re-entry tachycardia ablation and an approach guided by low voltage signals, local activation time, and the propagation wave collision during sinus rhythm.
Methods
This is a randomised prospective multi-centre controlled clinical trial at 5 paediatric cardiac electrophysiology centres. Patients 21 years of age or younger, greater than 15 kg, without moderate or complex CHD and being treated with (first-time) ablation therapy for supraventricular tachycardia in participating centres were eligible for enrolment. IRB approval was obtained at each participating institution. A data safety monitoring board was in place. Consent/assent was obtained prior to procedure initiation. 3D mapping was used on all patients (Biosense Webster, Carto or Abbot, EnSite).
After consent was obtained, and if participants were considered to have the diagnosis of atrioventricular nodal re-entry tachycardia based on inducible atrioventricular nodal re-entry tachycardia (typical or atypical) and/or jump and echo beats with a strong clinical suspicion, they were randomly assigned to one of two groups:
Group 1 Voltage and propagation wave evaluation: To determine the site of initial ablation attempts, providers were instructed to identify a site: (1) within the Triangle of Koch, (2) at or superior to the wave collision at a site of later activation, and (3) in an area of low voltage. The collaborators were provided an online video tutorial of this technique to confirm consistency of the approach across institutions (https://avnrt.pediatrics.wisc.edu/stepwise-approach). The mapping catheter was not standardised, as providers utilised catheters including a standard mapping catheter, the ablation catheter, and, in some cases, high-density mapping catheters. Voltage points collected during sinus rhythm were evaluated for accuracy, and errant points were adjudicated or eliminated. For voltage point density, points were collected until coverage was considered adequate by the ablating physician with a minimum recommended 30 points within the Triangle of Koch as demonstrated in a prior publication. Reference Van Aartsen, Law, Maldonado and Von Bergen7 The propagation wave was evaluated for a wave collision, and the site of the wave collision was marked (Figure 1 or online tutorial). This could then be referenced when evaluating voltage mapping (see Figure 2 below).
An example of the propagation wave collision (black line) marked to assist with the ablation site selection. The white line represents the sinus wavefront over three different time points. The more superior wavefront in the TOK slows while the inferior portion continues to advance, ultimately causing the collision between the superior and inferior wavefronts. The black line was placed along the centre of the collision between the superior and inferior atrial signal propagation waves as a marker for the wave collision. Local activation time can also be useful to assess areas of later activation, though this is not shown here. The initial ablation site was recommended at or just superior to the wave collision when using the Voltage-Propagation technique.
Figure 1 Long description
The image presents a cross-sectional view of the heart's Triangle of Koch, highlighting the propagation wave collision to assist with ablation site selection. The black line marks the center of the collision between the superior and inferior wavefronts, which is crucial for determining the ablation site. The white line represents the sinus wavefront at three different time points, illustrating how the superior wavefront slows while the inferior portion continues to advance, ultimately causing the collision. This collision is used as a marker for the wave collision, aiding in the selection of the initial ablation site, particularly when using the Voltage-Propagation technique. The image emphasizes the importance of understanding electrogram propagation and late activation during sinus rhythm for improving the success and decreasing complications of ablation procedures.
Example of a left lateral (slightly anterior) view of the right atrium and coronary sinus looking at the atrial septum and the TOK. The colour gradient map shows low voltage in the red (0.05 mV) to high voltage in the purple colour (>1.5 mV). The first site of ablations (blue circles) was selected at or just superior to the wave collision (white line) overlying a low-voltage (red/orange colouring) area. The gold circles/spheres represent the His signals. Labels and sphere outlines added for clarity. The coronary sinus, Tricuspid Valve (TV) and Inferior Vena Cava (IVC) are also marked.
Figure 2 Long description
The diagram presents a left lateral view of the right atrium and coronary sinus, focusing on the atrial septum and the Triangle of Koch. A color gradient map illustrates voltage levels, with red indicating low voltage (0.05 millivolts) and purple indicating high voltage (greater than 1.5 millivolts). Blue circles mark the initial ablation sites, positioned at or just above the wave collision line, which overlies a low-voltage area. Gold circles and spheres represent the His signals. The coronary sinus, tricuspid valve (TV), and inferior vena cava (IVC) are also labeled. The diagram highlights the relationship between voltage mapping and the selection of ablation sites to manage arrhythmias.
Using the previously obtained voltage points, the voltage map was evaluated by setting the lower voltage gradient bar threshold between 0.05 and 0.1 mV and the high voltage gradient bar between 1.5 and 2 mV. This gradient scale could be adjusted by the electrophysiologist to better highlight low-voltage areas (typically near 0.1 mV), an example of which is seen in Figure 2.
The recommended initial site of ablation application was within the triangle of Koch, at or superior to the wavefront collision, over an area of low voltage. Reference Malloy, Law and Von Bergen4,Reference Somasundaram and Von Bergen5,Reference Van Aartsen, Law, Maldonado and Von Bergen7,Reference Von Bergen and Law8 If multiple areas fit the propagation wave collision and voltage criterion, a long duration (40 m sec) and fractionated atrial electrogram, with a typical A:V ratio, could also be considered with site selection along with position relative to the CS os and the tricuspid valve. Discretion was left to the ablating physician to select the most promising site while optimising safety, often resulting in initial ablations more inferior in the Triangle of Koch and towards the ventricle if there were multiple areas of interest.
Group 2 Anatomical/electrogram approach (anatomic group): Providers utilised 3D mapping systems, but activation times and voltage maps were not utilised with this approach. Instead, the right atrial Triangle of Koch and proximal CS geometry were obtained on the three-dimensional mapping system. The site of initial ablation was selected using traditional anatomic landmarks of the Triangle of Koch and CS. Consideration of the appearance of the atrial and ventricular electrograms (i.e., low-voltage, longer-duration electrograms and a typical atrial-to-ventricular signal ratio) or other anatomic considerations was per the ablating physician’s discretion.
Ablation energy type, radiofrequency or cryotherapy, was chosen at the discretion of the provider. To obtain a more accurate localisation of the effect of the lesion, if cryotherapy was used and the patient tolerated atrioventricular nodal re-entry tachycardia, providers were encouraged to consider ablation during atrioventricular nodal re-entry tachycardia while monitoring for effect. If no effect (of slowing and termination) was noted within 20–30 s of cryotherapy initiation, then the ablation was stopped, and a new site was selected. Testing during and/or after the ablation could be done to evaluate the impact on the slow pathway. At the site of success, a cryotherapy application of 4–5 minutes was recommended with at least one freeze-thaw-freeze at the successful site before considering additional applications in the surrounding tissue. While using radiofrequency, during sinus rhythm, the presence of junctional ectopy was tracked during the ablation. Minimum post-ablation endpoints included no inducible supraventricular tachycardia and one or fewer echo beats, though some providers sought to eliminate all echo beats. Providers had a waiting time of at least 30 minutes after the successful ablation with testing on and off isoproterenol before concluding the procedure.
Patient data were collected, including the patient’s age and demographics. Procedural characteristics included the type of sedation and medications used, procedural timing, and use of fluoroscopy. The procedure start was considered the time of first sheath placement, and all measurements of time to an event (success or final lesion) were measured from that time point. The procedure’s end was the time of study conclusion and/or sheath removal. Mapping and arrhythmia characteristics and response to ablation were evaluated, including inducibility with and without isoproterenol; number of ablations in total and to initial success; as well as post-procedural evidence of persistent slow pathway conduction. All study data was de-identified and entered in a REDCap database. Reference Harris, Taylor and Minor9 Follow up for up to two years was completed. Recurrences were recorded with additional testing as required to diagnose probable or confirmed recurrences.
Descriptive statistics on demographics and clinical characteristics after the procedures were calculated. Comparisons between the voltage-propagation and anatomic groups on treatment effects were analysed using the negative binomial regression models, which were chosen as the modelling framework to accommodate greater levels of dispersion. The analysis was based on the intention-to-treat paradigm, including all randomised participants based on the assigned technique. All analyses were performed with the statistical software STATE/SE 16.1 (StataCorp, College Station, TX). A P-value of 0.05 was considered significant. The research reported in this paper adhered to the Helsinki Declaration guidelines as reviewed in 2013.
Results
In all, 70 patients were recruited and met study criteria. They were randomised to 37 within the voltage-propagation wave approach and 33 in the anatomical group. There was no significant difference between patient demographics or follow-up duration between groups (Table 1).
Patient demographics
Table 1 Long description
The table compares patient demographics between two groups: anatomic/ electrical guided mapping with 33 patients and voltage/ propagation mapping with 37 patients. It includes data on gender, race, age, height, weight, and follow-up time. The table has 6 rows and 4 columns. The columns are labeled Anatomic/ electrical guided mapping, Voltage/ propagation mapping, and P value. The rows are labeled Male percentage, White percentage, Age years median range, Height centimeters mean standard deviation, Weight kilograms mean standard deviation, and Follow-up time months median interquartile range. Notable trends include a higher percentage of white patients in both groups and similar median ages and follow-up times between the groups.
The majority of patients (52/70, 74%) had inducible typical atrioventricular nodal re-entry tachycardia; 5/70 (7%) had inducible atypical atrioventricular nodal re-entry tachycardia (without typical atrioventricular nodal re-entry tachycardia), and the remainder had only a jump and echo beats (13/70, 18%) without inducible atrioventricular nodal re-entry tachycardia. The number of patients randomised to the anatomic vs voltage propagation groups was similar, with typical atrioventricular nodal re-entry tachycardia in 25 vs 27 patients, atypical atrioventricular nodal re-entry tachycardia (without typical atrioventricular nodal re-entry tachycardia) in 2 vs 3, and jump/echo in 6 vs 7, respectively. Of those classified as within the typical atrioventricular nodal re-entry tachycardia group, 1 patient in the anatomical group and two in the voltage propagation group also had atypical atrioventricular nodal re-entry tachycardia.
Mapping for the voltage propagation group was typically performed with either the ablation catheter (if using radiofrequency) or a 2–5–2 mm spacing catheter if the ablations were done with cryotherapy. Three patients had high density mapping (HD grid or Pentarray).
The total procedural timing, including time to first ablation, time to success, and overall duration of the procedure, was not significantly different between techniques. Though the time to first ablation and the successful ablation were slightly (though not significantly) less using the anatomic technique, the time to finish the last application and the time to procedure end were slightly, though not significantly, longer using the same anatomic technique. (Table 2)
Duration from the procedure start to ablation and to procedure end
Table 2 Long description
The table presents a comparison of procedural timing between anatomic guided and voltage-propagation guided techniques. It includes data on the duration from the procedure start to the first ablation, successful ablation, finishing the last ablation, and the procedure end. The table has four rows and four columns, with column headers labeled as Anatomic guided (N=33), Voltage-propagation guided (N=37), and P value. The first row shows the median time to first ablation in minutes with interquartile ranges for both techniques. The second row displays the median time to successful ablation in minutes with interquartile ranges. The third row indicates the median time to finish the last ablation in minutes with interquartile ranges. The fourth row presents the median time to procedure end in minutes with interquartile ranges. The P values indicate the statistical significance of the differences between the two techniques.
While the number of applications to success was not statistically different (P = 0.298), 36% of patients in the voltage-propagation group achieved success on the first application (median: 2 applications) versus 28% in the anatomic group (median: 5 applications).
Most patients (51, 71%) underwent cryoablation with the 6 mm tip catheter. A radiofrequency catheter (solid or irrigated tip) was used in 15 patients, with an irrigated-tip radiofrequency in 10, with some overlap between groups due to multiple types of ablation catheters used. When comparing the time to success, there were significantly more total applications with cryotherapy placed for the voltage-propagation group (median 15.5 lesions) than the anatomical-guided group (median 12 lesions, P = 0.044). This translated to a longer, though not significantly different, total duration of cryoablation lesion time (P = 0.061) within the voltage-propagation group. (Table 3)
Ablation applications and timing data
Table 3 Long description
The table compares ablation applications and timing data between anatomic guided and voltage-propagation guided procedures. It consists of four rows and three columns. The columns are labeled Anatomic guided, Voltage-propagation guided, and P value. The rows detail the number of lesions to first successful application, total number of radiofrequency ablative applications, total number of cryo ablative applications, and total target cryo ablation time. For the number of lesions to first successful application, the median for anatomic guided is 5 with an interquartile range of 1 to 33, and for voltage-propagation guided, it is 2 with an interquartile range of 1 to 7. The total number of radiofrequency ablative applications has a median of 11 for anatomic guided with an interquartile range of 6 to 16, and a median of 12.6 for voltage-propagation guided with an interquartile range of 5.5 to 11. The total number of cryo ablative applications has a median of 12 for anatomic guided with an interquartile range of 9 to 21, and a median of 15.5 for voltage-propagation guided with an interquartile range of 9 to 21. The total target cryo ablation time in seconds has a median of 1808 for anatomic guided with an interquartile range of 1511 to 2379, and a median of 2268 for voltage-propagation guided with an interquartile range of 1795 to 3136. The P values for these comparisons are 0.298, 0.319, 0.044, and 0.061 respectively.
Irrespective of mapping technique, there was no significant difference in the number of applications to success when using only radiofrequency (N = 19, median 2) or cryotherapy (N = 46, median 3) or both (N = 4, median 12.5), though there was a non-significant increase in the number of lesions required for patients that underwent both radiofrequency and cryotherapy (P = 0.072). There was acute success in all but one patient (1.4%) who was initially in the voltage-propagation group but who crossed over to the anatomic approach without success using either approach. This patient had improved symptoms after the procedure but continued to have intermittent supraventricular tachycardia confirmed by monitoring.
Though 69/70 procedures were acutely successful, nine patients crossed over from one mapping technique to the other during the procedure. Of these, six patients started with the voltage-propagation technique and crossed over to the anatomic technique. The risk for crossover was not significantly different between groups (P = 0.48), though it did trend towards more crossover using the voltage-propagation technique. Of the patients that crossed over to both techniques, one patient had a recurrence post-procedure, and another, mentioned above, did not have acute success. The most common reason for cross-over to the alternative technique was lack of success (n = 6). One patient had a crossover due to concerns that the voltage-propagation mapping technique suggested sites closer to the AV node than the provider desired to ablate. After excluding patients that crossed over mapping techniques, there was no significant difference between the two techniques for any measured variable.
After the ablation procedure, patients were followed for up to 2 years. There was no difference between the durations of follow-up, with the average follow-up of 1.2 years for both groups (P = 0.83). Three patients were lost to follow-up after the procedure (one from the anatomic group and two from the voltage-propagation group). Of the 69 acutely successful procedures, there were a total of 4 probable or confirmed recurrences (5.9%). Though not significantly different between groups, 2 of these recurrences were within the voltage-propagation group (7% recurrence risk). One of the recurrences was in the anatomic technique group (3.5% recurrence risk), and one of the recurrences had crossed over to both techniques. Evaluating the ablation therapy, two probable or confirmed recurrences underwent cryotherapy (4%) and two underwent radiofrequency (9%). Both radiofrequency recurrences did have junctional acceleration at the time of ablation. The sensation of palpitations was common after the procedure, with an additional 5 patients that described infrequent palpitations that did not correlate with arrhythmia and had no other evidence of recurrence on monitoring.
When comparing procedural outcomes, patients with typical atrioventricular nodal re-entry tachycardia required significantly fewer applications to achieve success than those with atypical atrioventricular nodal re-entry tachycardia (median: 3 applications for typical atrioventricular nodal re-entry tachycardia vs 13 for atypical atrioventricular nodal re-entry tachycardia vs 3.5 for jump/echo physiology; P = 0.015). Though there were recurrences with each arrhythmia subtype, the risk for recurrence of typical atrioventricular nodal re-entry tachycardia trended toward being less likely to recur in comparison to those with atypical atrioventricular nodal re-entry tachycardia or only a jump/echo (P = 0.158). Of the four confirmed or probable recurrences, two occurred in a patient with typical atrioventricular nodal re-entry tachycardia, a 4% recurrence risk for typical atrioventricular nodal re-entry tachycardia using any technique. One of these recurrences occurred with each mapping technique. The other two recurrences included one patient with atypical atrioventricular nodal re-entry tachycardia (20%) who crossed over to use both techniques and one with a jump/echo (9%) mapped with the voltage-propagation technique.
There were no acute complications in either group. Though not seen during or immediately after the procedure, at follow-up in the anatomic group after cryotherapy, one patient was noted to have a longer intermittent PR interval, and another was noted to have an occasional asymptomatic 2nd-degree AV block at lower rates. In follow-up after ablation for typical atrioventricular nodal re-entry tachycardia from the voltage-propagation group, one patient was noted to have a right bundle branch block after cryotherapy and another intermittent ventricular couplets after radiofrequency ablation for jump/echo beats.
Discussion
This is the first and only prospective, multi-centre randomised trial comparing anatomic versus voltage-propagation mapping approaches for atrioventricular nodal re-entry tachycardia ablation.
The clinically important results demonstrated by this trial include (1) both techniques are excellent at the treatment of typical atrioventricular nodal re-entry tachycardia, (2) voltage-propagation mapping demonstrated a trend towards fewer applications to success, (3) the duration of the procedure and long-term outcomes were similar for both techniques, and (4) results were similar with radiofrequency and cryotherapy ablation therapy. Additionally, the added time required for mapping using the voltage-propagation technique did not prolong procedure duration, reinforcing the technique’s clinical utility. These data underscore the clinical effectiveness of ablation procedures no matter the mapping approach or type of ablation therapy for typical atrioventricular nodal re-entry tachycardia.
The findings of accuracy of the voltage-propagation technique are reinforced by our recommended cryotherapy protocol—discontinuing applications without effect within 20–30 s. Since 71% of ablations used cryotherapy, a rapid response to cryotherapy applications offers a meaningful gauge of targeting accuracy, again adding confidence to the observed trend toward fewer applications with voltage-propagation mapping and suggesting accurate localisation of the slow pathway. In all, 36% of patients in the voltage-propagation group achieved success on the first application (median: 2 lesions), versus 28% in the anatomic group (median: 5 lesions). Even so, the potential benefit was likely obscured by the limited power of this study and the lack of refinement of the techniques. In both approaches, operators often initiated ablations more inferiorly in the Triangle of Koch to optimise safety, which may have led to additional applications before success. Interestingly, voltage-propagation mapping infrequently guided ablation sites higher in the Triangle of Koch than typically chosen using anatomical landmarks, prompting one patient crossover. This highlights the balance required between precision and safety—especially pertinent when using radiofrequency energy. Reference Howard, Valdes and Zobeck10,Reference Pandozi, Ficili and Galeazzi11 Future refinements may help reduce subjectivity in site selection, improve consistency of mapping and further lower the number of lesions to success, though providers may need to incorporate additional site selection criteria such as distance and location from His bundle.
Somewhat paradoxically, there were more total cryoablation applications using voltage-propagation mapping (P = 0.04) despite a trend toward fewer lesions to initial success. Discussion among the authors reinforced that additional applications often targeted areas suggestive of slow pathway characteristics (low voltage, near wave collision…), even after achieving an appropriate endpoint. However, when excluding patients with crossover between mapping techniques, this difference in total lesions was no longer significant (a median of 13 total lesions in both groups), implying that more challenging substrates may necessitate multiple strategies and influence lesion count independently of mapping technique. Again, a larger powered study may be able to better determine the influence on mapping technique and total lesion number.
Interestingly, sub-group analysis showed a trend towards differences in outcomes when comparing typical atrioventricular nodal re-entry tachycardia, atypical atrioventricular nodal re-entry tachycardia and those with a jump/echo only. Both mapping techniques were very successful with typical atrioventricular nodal re-entry tachycardia, with a high acute success rate, a median of only 3 lesions to success, and only a 7% recurrence risk. There were fewer lesions to success for typical atrioventricular nodal re-entry tachycardia than atypical atrioventricular nodal re-entry tachycardia and a trend towards a higher recurrence with atypical atrioventricular nodal re-entry tachycardia irrespective of mapping technique. This data supports both the excellent outcomes with typical atrioventricular nodal re-entry tachycardia, and that we should continue to explore alternative techniques for atypical atrioventricular nodal re-entry tachycardia to achieve a higher success rate.
The number of lesions placed in patients when using radiofrequency or cryotherapy alone was not statistically different unless also comparing radiofrequency only, cryotherapy only, or both cryotherapy and radiofrequency. The number of lesions to success (P = 0.07), total lesions (P = 0.03), and procedure duration (P = 0.01) were greater when the procedure required the use of both radiofrequency and cryotherapy ablation techniques (N = 4) in comparison to a single ablation energy modality.
Despite a few statistically significant differences, this study confirms both techniques are safe and highly effective for typical atrioventricular nodal re-entry tachycardia ablation. Voltage-propagation mapping offers site selection guidance without compromising safety or efficacy and may reduce lesions required for success without increasing procedure duration. Still, the subjectivity of site selection, the more variable outcomes with non-typical atrioventricular nodal re-entry tachycardia substrates, and the lack of consistent first lesion success highlight the need for further refinement or consideration of alternative techniques.
The authors appreciate that techniques continue to evolve. Recent interest in the pivot point and directionality of the atrial signal within the Triangle of Koch, along with other findings such as conduction velocity, peri-CS electrical activity, and/or fractionation, may further increase the chance for success. Reference Bailin, Rhodes, Arter, Kocherla and Kaushik3,Reference Somasundaram and Von Bergen5,Reference Takahashi, Yamaoka, Kujiraoka, Arai, Hojo and Fukamizu6,Reference Howard, Valdes and Zobeck10,Reference Fogarty Iv, Kamp, Eisner, Kertesz and Kumthekar12–Reference O’Leary, Sneider and Przybylski14 Though not evaluated in this study, many of these features overlap spatially with wave collision sites and later activation identified via voltage propagation in the voltage-propagation wave technique.
Importantly, and a potential area of investigation for future investigators, both mapping strategies in this study infer slow pathway location during sinus rhythm using surrogates (e.g., low voltage, late activation, wave collision, electrogram features, and anatomic features) rather than directly mapping the reentrant circuit. Advances in anatomic and mechanistic understanding may soon allow more direct and accurate mapping during tachycardia, reducing reliance on indirect markers. The potential integration of automation, machine learning, and AI may further enhance precision and safety in atrioventricular nodal re-entry tachycardia ablation.
Interpretation of this study is limited by its modest sample size, limited by slow recruitment. We also appreciate inherent subjectivity in lesion selection site, ablation energy selection and slight variations in technique across multiple institutions.
Conclusion
This is the first prospective, randomised evaluation of atrioventricular nodal re-entry tachycardia ablation utilising an anatomic approach in comparison to a voltage-propagation mapping approach. Both techniques demonstrated an excellent acute success rate and a low recurrence rate for typical atrioventricular nodal re-entry tachycardia. This paediatric study suggests that both the traditional anatomical technique and a voltage-propagation technique can both provide excellent clinical outcomes without a difference in complications or procedural duration.
Financial support
This work was supported by an unrestricted research grant from Biosense-Webster.
Competing interests
Biosense-Webster provided an unrestricted research grant for a portion of this study.
Ethical standards
The Institutional Review Board and Safety Monitor were follow per institutional guidelines.
Open access
