Wolff–Parkinson–White ablation after a prior failure: a 7-year multicentre experience

Abstract

Aims

Catheter ablation for Wolff–Parkinson–White syndrome (WPW) can be challenging and is associated with failure in ∼1–5% of cases. We analysed the reasons for failure.

Methods and results

All patients (89 patients, 28 ± 16 years old) referred for WPW ablation after a prior failure were studied. Reasons for the prior failure as well as for the acute success were analysed. The repeat procedure was successful in 81 (91%) patients. Multiple (2.7 ± 0.9) or large accessory pathways (APs) were seen in 13 patients. For left lateral APs, inaccurate mapping and lack of transseptal access during the index procedure accounted for failure ( n = 15). An irrigated-tip catheter was required for epicardial APs ( n = 7). In addition, seven posteroseptal APs required bi-atrial and coronary sinus (CS) applications in order to succeed. For parahisian and midseptal APs, radiofrequency was cautiously titrated from 5 to 30 W, eliminating the AP in three patients. Cryoablation was used in seven patients (acute success in six but delayed recurrences in three of these). For patients with CS AP, irrigated ablation in the CS was crucial to deliver adequate power. For anteroseptal and right lateral APs, a successful outcome was achieved with long sheaths ( n = 5) or a left subclavian approach (anteroseptal, n = 4).

Conclusion

Failure in WPW ablation may be due to a variety of reasons but catheter manipulation and inaccurate mapping remain the two major causes. Knowledge of the reasons for failure depending on the location of the WPW may facilitate a successful outcome.

Introduction

Catheter ablation is the treatment of choice for Wolff–Parkinson–White syndrome (WPW) with a long-term success rate of ∼95%. ^1–6 However, ablation of accessory pathways (APs) can be challenging. Some of the factors leading to prolonged procedures, procedural failure, or recurrence of APs have previously been reported in the 1990s. ^7–9

Since then, there have been a number of developments that can potentially make the ablation of APs easier. Relatively recent improvements include: steerable sheaths for improved catheter stability, irrigated-tip catheters that are capable of delivering more energy to the tissue safely, and also cryoablation catheters. In addition, most electrophysiologists are now comfortable in performing transseptal puncture for left atrial access, given the rise in numbers of patients referred for atrial fibrillation. However, given that patients still have recurrences, we sought to investigate the underlying reasons for failure of WPW ablation in the modern era.

In this multicentre study, all consecutive patients referred for WPW ablation after a prior failure were systematically studied and reason(s) for failure investigated.

Methods

Study population

From 2000 to 2006, 1171 patients underwent radiofrequency (RF) catheter ablation of WPW (AP with antegrade conduction) at three international tertiary electrophysiological (EP) centres (Bordeaux University Hospital, Bordeaux, France; St Mary's Hospital, London, UK; and Brigham and Women's Hospital, Boston, MA, USA). All patients with a prior acute failed WPW ablation were included in the present study (patients with acute success but recurrence were not included), whether or not the original ablation had been performed in the reference centre.

Procedure

Techniques, catheters, and settings differed depending on the centre. However, ablation in the coronary sinus (CS) system, when performed with RF energy, used an irrigated-tip catheter with a power limited to 20–25 W.

The procedure was performed under conscious sedation (midazolam and morphine if needed). An EP evaluation was performed to determine anterograde and retrograde AP refractoriness at baseline ± after isoprenaline.

Study design

This retrospective study was performed reviewing all EP procedure reports for patients with a prior failed WPW ablation. Data analysed were patients' characteristics, number of previous ablations, localization of the pathway, type of access (venous, transseptal, arterial, or epicardial), type of catheter and energy used, RF, fluoroscopy, and procedure duration. The reason for previous failure was determined by the operator based on the prior report and the current procedure. All differences between the initial failed procedure and the final successful procedure were carefully recorded.

Statistical analysis

Values are presented as mean ± SD. However, continuous variables such as number of previous ablations, RF, fluoroscopy, and procedure duration are expressed as median and quartile (first and third). Fisher's exact test was used to compare categorical variables, whereas the Kruskal–Wallis test was used to compare continuous data according to the different AP locations. A P -value <0.05 was considered significant.

Results

Patient characteristics

Eighty-nine patients (61 males, 28 ± 16 years old) (8% of WPW ablation in these centres) had WPW ablation after a prior failed ablation. Eighty-two (92%) had been referred from another institution following a failed ablation, with a mean of 1.8 ± 0.8 procedures prior to being referred ( Table  1 ). The repeat procedure was successful in 81 patients (91%). There was no evidence of structural heart disease as assessed by transthoracic echocardiography in 82 patients, 4 patients had Ebstein's anomaly, 2 patients had sickle-cell disease, and 1 patient had a congenitally corrected transposition of the great vessels. The locations of the APs are summarized in Table  1 . Multiple (2.7 ± 0.9) or large APs were seen in 13 patients.

Pathway location

From the 2000–03 to the 2004–06 period, pathway location of patients with a prior failed ablation did not evolve much except for left lateral pathways that became less frequent (respectively, 34 vs. 14% of failed pathway, P = 0.03) with the development of transseptal access in the EP laboratory ( Table  2 ). Pathway locations and their ablation characteristics are summarized in Table  1 . Concerning the two patients classified as ‘other’, one had a left anterior pathway and the other a pathway situated in the cavotricuspid isthmus.

Left lateral accessory pathways

Twenty-two patients attended due to a prior failed ablation of a left lateral AP. In five cases (2000–03 period), transseptal access had not been attempted during the initial procedure. At the repeat procedure, transseptal access was performed in all of these patients and the AP was successfully eliminated with a single application. Inaccurate mapping of the AP was thought to be the reason for failure in a further 10 patients. Changing from ventricular to atrial mapping (or vice versa) or changing the pacing site to distal CS or left ventricle was judged useful in these cases. In five patients, despite correct pathway localization, ablation required irrigated-tip catheters, suggesting that the APs were located deeper within the myocardium, rather than superficially. In two of these patients, epicardial ablation was within the great cardiac vein in a position contiguous to the best location endocardially. In two other patients, the AP could not be eliminated from the endocardial surface with either antegrade or retrograde approaches using irrigated-tip catheters and no favourable site was found within the great cardiac vein. The mean RF duration during the repeat procedure was 6 ± 6 min. Transient ischemic attack occurred after the procedure in one patient with a left lateral pathway ablation. No other complication occurred in our studied population.

Posteroseptal and coronary sinus accessory pathways

Redo ablations of posteroseptal WPW ( n = 12) were more complicated, with 19 ± 24 min of RF needed. Irrigated-tip catheters were used in all patients. Seven required bi-atrial and CS applications to achieve success, suggesting a pathway located deep within the septum. In one patient with a left posteroseptal pathway, stability was mentioned in the initial report as the reason for failure. However, the redo procedure performed via a retrograde aortic approach with a different catheter and a more experienced electrophysiologist was successful within l h. In another patient, epicardial access via a percutaneous subxyphoid approach was performed and epicardial ablation was successful after failed CS ablation. In the remaining three patients, the pathway could not be ablated.

Endocardial RF applications were ineffective in patients with CS WPW ( n = 10). In six patients, non-recognition of the location of the AP within the CS had been the reason for the previous failed procedure. In a further two patients, the referring electrophysiologists did not use an irrigated-tip catheter within the CS and, despite correctly locating the pathway within the CS, were unable to deliver enough energy with a standard RF catheter to eliminate the AP. In the two remaining patients, the AP was within a CS diverticulum which was identified at the repeat procedure, but which had not been recognized at the index procedure. The repeat procedure was successful in all patients using either an irrigated-tip catheter with a power limited to 25 W ( n = 7) or a cryoablation catheter ( n = 3).

Parahisian and midseptal accessory pathways

For parahisian and midseptal WPW ( n = 11), previous failure was due to the close proximity of the AP to the atrioventricular (AV) node and the risk of subsequent heart block. In four patients, RF energy, cautiously titrated from 5 to 30 W, eliminated the AP without AV node damage in three and was stopped in one due to an excessive risk of AV block. Cryoablation was used in seven patients with acute success in six but delayed recurrences in three of these (within 1 month). The procedure was abandoned in one young patient due to the risk of AV block (transient AV block during cryomapping).

Right lateral accessory pathways

For right lateral APs, the main reason for previous failure had been lack of catheter stability at the index procedure (five patients). Inaccurate mapping of the AP was the reason for failure in three patients due to trabeculation and/or oblique pathways. Changing mapping from earliest V to earliest A or vice versa was useful in these patients. In one patient, percutaneous subxyphoid epicardial ablation was successfully attempted after extensive endocardial ablation.

Anteroseptal accessory pathways

Nine patients had a previous failed anteroseptal AP ablation. The reasons identified were a lack of catheter stability and concern regarding the proximity of the AV node. At the redo-procedure, a successful outcome was achieved using either a long sheath ( n = 5) or a left subclavian approach ( n = 4) ( Figure  1 ).

Multiple or large accessory pathways

Thirteen patients had multiple pathways or an AP with multiple ventricular or atrial insertions. The major reason for ablation failure at the index procedure was the non-recognition of the presence of several APs which had resulted in mapping inaccuracy ( Figure  2 ). As expected, these patients required more than a single RF application with a mean ablation time of 17 ± 16 min. All four patients with Ebstein's anomaly had multiple pathways and/or an AP with multiple ventricular or atrial insertions. Three other patients had very large pathways. A pathway extending from 7 to 10 o'clock on the tricuspid annulus was present in one patient with sickle-cell disease and two other patients presented a large pathway extending from the right atrial appendage to the right ventricle (RV). The correct ablation site was identified in these patients from a change in the surface electrocardiographic (ECG) morphology and/or a delay in the local activation time ( Figure  3 ).

Failed ablation at the repeat procedure

In eight (9%) patients, even at the repeat procedure with an experienced operator and all technical assistance possible, the AP could not be ablated. The locations of these APs were posteroseptal ( n = 3), parahisian ( n = 2), left lateral ( n = 2), and multiple pathways ( n = 1). All of the patients with a failure in a posteroseptal pathway had congenital heart disease (two patients with Ebstein's anomaly and in one patient with congenitally corrected transposition of the great vessels). In the latter patient, the mid- and distal CS could not be accessed. In the two patients with failure of a parahisian pathway, the procedure was aborted due to the very high risk of AV junction (junctional rhythm as soon as RF application started at 5 W or transient AV block during cryomapping). In the patients with failed ablation of left lateral pathways, endocardial ablation at the earliest site did not result in loss of the AP, and the great cardiac vein was not close enough to the putative site (no percutaneous subxyphoid pericardial approach was performed in these patients).

Discussion

Main findings

The present study demonstrates that even after a prior failure, a successful outcome may be obtained in up to 91% of the cases in an experienced centre. The primary reasons for the initial failure were difficulties in catheter manipulation and inadequate mapping of the AP. However, in one-third of all patients (34%), AP ablation was very challenging, primarily due to AP localization (parahisian, epicardial), multiple pathways >3, multiple insertions, or structural heart disease (congenital heart disease).

Challenging catheter manipulation

Despite the advances in EP over the last decade, ablation of APs remains a challenging procedure in some cases. Morady et al . ⁷ also demonstrated in a series of 619 patients, 65 of whom had previously undergone an unsuccessful procedure, that difficulties in catheter manipulation were responsible for failure in 48% of cases. Although a left lateral pathway is the most prevalent location for earlier failure in this study, this does not indicate that this was the most challenging location. Left lateral pathways are the commonest AP, whereas Ebstein's anomaly and multiple AP are far less prevalent. However, in terms of time taken and number of individual applications, patients with congenital heart disease are the most difficult pathways to ablate. Accessory pathways with multiple atrial or ventricular insertions and some posteroseptal pathways may be much more challenging than left lateral pathways. Depending on the location of the AP, the ‘tricks’ employed at the repeat procedure to improve catheter manipulation differed. For left-sided APs, changing from retrograde aortic to transseptal approach was very useful to improve stability and facilitate atrial mapping (especially in failed left lateral pathways). ¹⁰

For posteroseptal pathways, the irrigated-tip catheter was particularly important. In this location, pathways were sometimes deep in the myocardium and their elimination required right and left atrial as well as CS application in seven cases. In one patient, percutaneous subxyphoid access was performed and ablation was carried out in the pericardium after coronary angiography, as described by Valderrabano et al . ¹¹

Challenging mapping

For APs with an oblique course, instead of mapping the earliest ventricular activation, mapping the earliest atrial activation was essential in difficult cases, as well as using both ventricular and atrial pacing from different sites. ¹²

For epicardial posteroseptal pathways, the most important point is to recognize them. A posteroseptal pathway within the CS has a specific pattern (negative delta wave in lead II and the combination of a steep delta wave in aVR with a deep S-wave in V6—R-wave ≤ S wave) as described. ¹³ Once recognized, and after correct mapping in the coronary vein, the main issue is energy delivery due to poor blood flow. An irrigated-tip catheter (power limited to 20–25 W) or cryoenergy is then mandatory to achieve long-term success. Another issue with this location was CS diverticulum, which was present in two of our patients. A CS angiogram may be useful in these cases to improve mapping according to the anatomy. ^{14 , 15}

In the context of anteroseptal pathways, left subclavian vein access is of particular interest for stability ( Figure  1 ). The use of a long sheath may also be a solution, as used for right free wall pathways.

Atrioventricular block

The problem with ablation of parahisian pathways is the risk of AV block. Mapping has to be very accurate and stability is crucial to eliminate the AP safely without altering AV conduction. Cryoablation is useful in these cases as thermal mapping can be carried out safely and is reversible. However, in the present study, we also used thermal mapping but with RF as the energy source (starting at 5 W and titrated in 5 W increments to 30 W). However, there are two important caveats. First, stability may not be as good as with cryoablation (where an iceball forms and fixes the catheter to the endocardium). Therefore, more so even than with cyoablation, careful monitoring of junctional rhythm is mandatory so that the application can be immediately stopped (a particular pitfall is that as pre-excitation disappears during junctional rhythm, this may be interpreted as a successful ablation by less experienced operators). The other ‘trick’ that was employed during some repeat ablations was to target the ventricular insertion of the pathway with a small amplitude atrial potential, as the His bundle is more robust than the compact AV node. Once at the supposed elective site, atrial extra-stimulus was performed to block the pathway and look for a His potential. However, some limitations exist with this technique: (i) the extra-stimulus may change the catheter position slightly, (ii) AP refractoriness may be shorter than AV node refractoriness, and (iii) a small His potential may be recorded on the effective site (it was the case in 3 of our 11 patients).

Multiple and/or large pathways may be much more challenging, especially when they are not recognized. During the EP study preceding the AP ablation, the morphology of the surface 12-lead ECG must be carefully examined while assessing the refractory period of the AP (multiple pathways often have different refractory periods, and hence a different surface ECG appearance). In cases of AF, the minimal pre-excited R–R interval is important to determine prognosis but this is not the only valuable piece of information that may be gained as different morphologies of the surface ECG may indicate the presence of multiple pathways ( Figure  2 ). Large pathways may be more challenging to detect. During ablation, delayed local activation is a sign of effective application, as is a (slight) change in QRS morphology ( Figure  3 ).

Study limitations

One limitation of this study is that the reason for an ineffective ablation procedure may have been oversimplified in some patients. The reports from referring institutions were not always specific as to the cause of failure or the exact location of the AP. Although an attempt was made to identify the primary and secondary reasons for failure, it is possible that in some patients, the reasons were multiple and not well documented. Another limitation is that the 8% failure rate for WPW ablation is largely overestimated, as we were unable to obtain complete records of all the patients ablated with WPW in all referring centres over the study period. The patients were referred to our centres after an unsuccessful ablation procedure from a wide range of geographic areas by university- and community-based electrophysiologists with a broad range of training and experience in ablation techniques. However, we can extrapolate from these data that 15 of 1171—1.3% (7 patients with a prior failure in one of the reference centres + 8 failures during redo cases = 15) are close to the real acute failure rate in experienced centres.

Conclusions

Ablation of left-sided APs in WPW requires familiarity with both the transseptal and retrograde aortic approach. In cases of left inferior WPW, a middle cardiac vein AP should be ruled out. For right lateral to anteroseptal pathways, long sheaths and a left subclavian approach should be used. In all cases, careful appreciation of the 12-lead surface ECG is mandatory.

Conflict of interest: none declared.

Funding

F.S. was supported by a grant from the French Federation of Cardiology.

References