Article Text

other Versions

Heart rhythm disorders
Dissociated pulmonary vein potentials following antral pulmonary vein isolation for atrial fibrillation: impact on long-term outcome
  1. Geoffrey Lee1,
  2. Jonathan M Kalman1,3,
  3. Jitendra K Vohra1,
  4. Andrew Teh1,2,
  5. Caroline Medi1,2,
  6. Liang-Han Ling1,2,
  7. Peter M Kistler1,2,3,4
  1. 1Department of Cardiology, The Royal Melbourne Hospital, Melbourne, Australia
  2. 2Department of Cardiology, The Alfred Hospital, Melbourne, Australia
  3. 3Department of Medicine, University of Melbourne, Melbourne, Australia
  4. 4Baker IDI Diabetes and Heart Research Institute, Melbourne, Australia
  1. Correspondence to Dr Peter M Kistler, The Heart Centre, The Alfred Hospital, Commercial Road, Melbourne 3004, Australia; peter.kistler{at}


Background Pulmonary vein (PV) reconnection is the Achilles heel of pulmonary vein isolation (PVI) for atrial fibrillation (AF). Dissociated pulmonary vein potentials (dPVP) may reflect abnormal PV automaticity, indicate more extensive PV muscular sleeve or may simply be an epiphenomenon.

Objective This study sought to determine the incidence, characteristics and prognostic significance of dPVP following PVI for AF.

Methods 89 consecutive patients (mean age 58.2±8.4 years, 75% male, 74% paroxysmal, 26% persistent AF) underwent antral PVI using three-dimensional mapping systems with image integration with the endpoint of bidirectional PV block. Following PV electrical isolation the presence and characteristics of dPVP were recorded. Holter monitoring was performed at 3, 6 and 12 months. Acute PV reconnection was assessed over a 30-min waiting period.

Results Electrical isolation was achieved in all 372 PV targeted for ablation. 69 of 372 isolated PV (19%) demonstrated dPVP after acute electrical isolation. Sites of dPVP origin were the left superior in 36%, left inferior in 20%, right superior in 31% and right inferior in 12%. All 69 dPVP demonstrated slow activity (cycle length >1500 ms) with only four persisting more than 30 min after acute isolation. There was no difference in the clinical characteristics between dPVP-positive vs dPVP-negative patients. At a mean follow-up of 21±8 months the single procedure success was 25/33 (76%) in dPVP-positive versus 39/60 (64%) in dPVP-negative patients (p=−0.3). In the eight dPVP-positive patients who underwent a second procedure, 11 of the 14 (79%) veins with initial dPVP demonstrated PV–left atrial reconnection.

Conclusion dPVP are present in 19% of PV following acute antral electrical isolation. The presence of dPVP did not predict recurrent AF following PVI.

  • Atrial fibrillation
  • radiofrequency ablation

Statistics from

Pulmonary vein (PV) electrical isolation with demonstration of entrance and exit block represents the cornerstone of atrial fibrillation (AF) ablation procedures.1 The appearance of dissociated pulmonary vein potentials (dPVP) originating from PV musculature is generally a welcome sign of PV isolation, and their observed incidence ranges from 9% following segmental isolation to 40% following antral isolation.2–4 Conversely, the Achilles heel of ablation strategies for paroxysmal AF is PV reconnection with an incidence of 80−100%,5 6 in patients with AF recurrence. A number of markers have been used to identify which PV are likely to reconnect. Among these, studies have suggested that PV with active triggers are more likely to have extensive circumferential muscular PV–left atrium (LA) connections7 and more likely to be associated with early recurrence.8 However, whether these observations are applicable to PV that demonstrate dPVP following isolation is unknown. Such potentials may reflect abnormal PV automaticity, may indicate a more extensive PV muscular sleeve with inadequate debulking of the antral region by catheter ablation or may simply be an epiphenomenon without clinical implication.

Therefore, the aim of the study was to investigate whether the presence of dPVP following antral electrical isolation is associated with a higher AF recurrence rate.


Eighty-nine consecutive patients (mean age 58.2±8.4 years, 75% males, 74% paroxysmal AF) with drug refractory AF were included who underwent first-time catheter ablation between 2008 and 2009. All patients were followed up for a minimum of 12 months.

Paroxysmal AF was defined as recurrent AF (two or more episodes) that spontaneously reverted within 7 days. Persistent AF was defined as recurrent AF lasting more than 7 days or requiring pharmacological or electrical cardioversion.

Electrophysiological study

Anti-arrhythmic medications were ceased 5 days before the ablation procedure. All procedures were performed under general anaesthesia. Femoral venous access was obtained with two 6 French and two 8 French sheaths. A decapolar catheter was positioned within the coronary sinus and a quadripolar catheter was positioned at the His bundle location. Two separate transeptal punctures were performed using transoesophageal echocardiography and fluoroscopy guidance. Patients received a bolus of 100 IU/kg of intravenous heparin after the first transeptal puncture and further heparin to achieve an activated clotting time of more than 350 s after the second puncture. In all patients 3.5/4 mm irrigated ablation catheters and a circular mapping catheter were deployed in the LA.

Ablation procedure and endpoints

Left atrial geometry was created using a three-dimensional (3D) electro-anatomical mapping system (CARTO-XP, Biosense-Webster or NavX; St Jude Medical Inc., St Paul, MN, USA). The PV antrum was defined with angiography and with the segmented left atrial MRI or CT scan imported into the mapping system (CARTO-Merge or NavX Fusion).

Ablation was performed to encircle the left and right-sided PV in pairs 1−2 cm from their ostia as defined by PV angiography and the 3D map. At the anterior aspect of the left PV, ablation was performed along the ridge between the left atrial appendage and the PV ostia. Energy was delivered through a 3.5 mm irrigated tip catheter with power limited to 30 Watts and temperature to 48°C for left atrial ablation with flow between 17 and 20 ml/min. The PV were continuously assessed for electrical disconnection using the circular mapping catheter. If venoatrial electrical connections persisted further ablation was performed at the ablation line guided by the activation sequence on the circular PV catheter until electrical isolation was achieved (figure 1A). If this was not successful then further applications (power 20−25 Watts and temperature 50°C) were made at the veno-atrial junction typically in the region of the intervenous ridge. Following electrical isolation of the superior PV the circular mapping catheter was left in situ to look for dPVP. After 5 min the circular mapping catheter was then repositioned in the inferior PV to assess for electrical isolation and the presence of dPVP. This process was then repeated for the contralateral PV. Antral isolation was confirmed with a circular mapping catheter by demonstration of entrance and exit block by using pacing manoeuvres in all cases. Entrance block was defined as the absence of discrete PV potentials or the presence of dPVP (figure 1B). Exit block was assessed by pacing from the ablation catheter positioned within the PV distal to the ablation line with the circular PV catheter in the same vein to assist in assessing local capture (figure 1C). In persistent AF patients a roof line (30 w) joining the superior aspects of each wide encirclement ablation ring was performed. If AF continued electrical cardioversion was performed to restore sinus rhythm. Bidirectional block from each PV was then assessed. Conduction block at the roof line was confirmed by demonstrating a corridor of widely spaced double potentials and caudocranial posterior wall activation during pacing from the left atrial appendage.

Figure 1

(A) Acute isolation (left) demonstrates the position of the multi-polar Lasso catheter within the right superior pulmonary vein (RSPV) and ablation catheter at the pulmonary vein (PV)–left atrium (LA) junction at the time of acute PV isolation. (Right) Corresponding intracardiac signals at time of RSPV isolation performed during coronary sinus (CS) pacing. PV catheter signals on the left of the tracing consist of two rows of signals; first, LA signals and second, PV signals. There is increasing PV–LA conduction delay during radiofrequency delivery before PV isolation (fourth beat). The secondary PV component then disappears (right of tracing) as the RSPV is acutely isolated. (B) Dissociated pulmonary vein potential (dPVP) (left) demonstrates the position of the multipolar Lasso catheter within the RSPV and ablation catheter at the PV–LA junction just after acute PV isolation. (Right) Corresponding intracardiac signals during SR. PV catheter signals show two signals; first LA signals in sinus rhythm (in red, arrows) and second, the appearance of dPVP (blue), which typically manifested as a single isolated beat or slow cycle length (CL) activity (CL >1500 ms). (C) Exit block (left) position of the catheters during testing for exit block. (Right) Corresponding intracardiac recording demonstrating exit block. Ablation catheter is positioned within the recently isolated PV and pacing at CL 700 ms. From the PV electrogram recordings, there is local capture (green arrow) of the PV musculature, while the LA signals are clearly dissociated (circled in red). ABL, ablation electrograms; CS, coronary sinus electrograms; HIS, His bundle electrograms; PV, pulmonalry vein electrograms.

A minimum waiting period of 30 min following PV isolation was employed in all patients with continuous sequential PV assessment to detect PV reconnection. Isoprenaline and adenosine were administered in a subgroup of patients at the discretion of the operator. In the setting of acute reconnection the PV were re-isolated and the waiting period of 30 min recommenced.

Definition of PV potentials

dPVP were defined in the setting of: bidirectional PV–LA block and dissociation from far field LA and RA potentials. The presence, location and characteristics of dPVP were recorded using a circular catheter placed at the PV–LA ostium. A dPVP was defined as two or more PV potential that demonstrated exit block into the LA.

For those patients who underwent a repeat procedure, the nature and characteristics of sites of PV–LA reconnection were recorded as well as the presence/absence of dPVP after re-isolation.

Post-ablation anticoagulation

Low molecular weight heparin and warfarin were administered during the first 3−5 days after the procedure until an international normalisation ratio of 2.0−3.0 was achieved. Warfarin was administered for at least 3 months or indefinitely in patients with CHADS2 scores of 2 or greater. Anti-arrhythmic therapy was restarted immediately after the procedure and continued for 3 months. After that period, patients without symptoms and without documented arrhythmias discontinued all anti-arrhythmic drugs.


Clinical, electrocardiographic and Holter follow-up was performed routinely at 3, 6, 9 and 12 months and at least every 6 months thereafter. Additional Holter monitoring or 7-day monitoring was performed when patients reported symptoms suggestive of recurrence. In addition to routine clinical review, all patients had telephone interviews performed by a trained electrophysiology nurse for assessment of symptom recurrence every 6 months. Procedural success was defined as freedom from symptomatic AF or arrhythmia recurrence lasting more than 30 s beyond a 2-month post-ablation blanking period (on or off anti-arrhythmic medication).

Statistical analysis

All continuous variables are presented as mean±SD and all categorical variables are presented as number (%) of patients in each group. The overall comparison of the frequencies between groups was performed using χ2 statistics or Fisher's exact testing. Comparison of continuous variables between the groups was performed using an independent samples t-test. A p value less than 0.05 (two tailed) was considered statistically significant.


All patients had failed at least two anti-arrhythmic mediations. CHADS2 scores of 0.1 and 2 were present in 70%, 25% and 5%, respectively. Electrical isolation was achieved acutely in all 372 PV targeted for ablation. There were two vascular access-related complications (one femoral pseudo-aneurysm and one retroperitoneal haematoma) managed conservatively. There were no other major complications. The characteristics of the dPVP seen are shown in table 1. dPVP was seen during antral electrical isolation in 37 (41%) patients. dPVP were seen in 69 of 372 (19%) PV targeted at the index procedure with a mean of 1.5 dPVP per patient. The distribution of the dPVP was more common in the upper veins compared with lower veins (left superior pulmonary vein (LSPV) 36%, right superior pulmonary vein (RSPV) 31%, left inferior pulmonary vein (LIPV) 20%, right inferior pulmonary vein (RIPV) 13%). All 69 dPVP were seen immediately after achieving acute PV isolation and all demonstrated slow activity (cycle length >1500 ms). No new PV tachycardias or non-PV triggers were observed in response to isoprenaline.

Table 1

Characteristics of dPVP

The incidence of dPVP did not differ between patients who underwent ablation on the intervenous ridge 48.5% dPVP-positive vs 41.7% dPVP-negative, p value 0.5. At the completion of a 30-min waiting period only four of the 69 (6%) dPVP persisted.

Acute reconnection

Acute PV reconnection was seen in 24 (6.5%) acutely isolated PV after the 30-min waiting period with all undergoing successful re-isolation. Twelve of the 69 (17%) veins that had dPVP at initial isolation had acute PV–LA reconnection. There was no significant difference in the acute reconnection rate between dPVP-positive and negative veins.

Chronic reconnection

At a mean follow-up of 21±8 months the overall single procedure success for catheter ablation was 60 of the 89 patients (single procedural success rates were 70% and 58% for paroxysmal and persistent AF, respectively); 87.5% of the PV potential-positive group were off anti arrhythmic drugs at follow-up compared with 75.9% in the PV potential-negative group, p=0.36. The minimum follow-up for all patients was 12 months. Time to median arrhythmia recurrence was 5.5 months. All patients with symptomatic recurrence presented with clinical AF and there were no cases of atrial flutter. The recurrence rate did not differ between the two groups (24% dPVP-positive vs 36% dPVP-negative, p=0.30). There was no difference in the clinical or procedural characteristics between the patients who did and did not have dPVP. In particular radiofrequency ablation delivery time was the same between the two groups (53.8±19.7 min vs 49.1±17.9 min, p=−0.9; table 2).

Table 2

Patient demographics and procedural data

Seventeen of 29 patients underwent a second procedure with 12 refusing due to infrequent symptoms controlled on previously ineffective anti-arrhythmic drugs. PV reconnection was the mechanism of recurrence found in all patients. A mean of 2.5±0.9 PV per patient demonstrated PV–LA reconnection (LSPV 22%, LIPV 22%, RSPV 24%, RIPV 30%). Eight patients with dPVP underwent a second procedure. In these patients, 11 of 14 (79%) dPVP-positive PV demonstrated PV–LA reconnection. No dPVP were observed with electrical isolation of reconnected PV.


The present study demonstrates that PV potentials following PV electrical isolation are not associated with AF recurrence. Such potentials may indicate a more extensive PV muscular sleeve, which although acutely isolated is more susceptible to chronic reconnection; or reflect a more active trigger; however, the findings from the present study suggest an epiphenomenon without clinical implication. To our knowledge no earlier studies have determined the impact of dPVP on freedom from AF following PV antral isolation. In addition: (1) dPVP were seen in up to 19% of PV after PV antral isolation and more frequently originate from the superior PV; (2) dPVP are a transient phenomenon seen acutely with the achievement of electrical isolation with just 6% present 30 min after electrical isolation; (3) dPVP were not a useful marker for acute or chronic PV reconnection. However, at repeat procedure reconnection was seen in up to 79% of PV that originally demonstrated dPVP.

Incidence of dPVP

Patients with highly active veins although acutely isolated may be expected to present with clinical recurrence should reconnection occur. dPVP were first described in an era of catheter ablation confined to ostial isolation of the ‘arrhythmogenic’ vein. Weerasooriya et al2 demonstrated the presence of dPVP following ostial ablation in 12% in an era when the majority of patients underwent isolation confined to the ‘arrhythmogenic’ vein. Marrouche et al3 reported a dPVP incidence of 2.8%; however, the definition required sustained activity beyond 10 min and were confined to observations from the triggering vein. Acceleration was observed in response to isoprenaline and suppression in the presence of adenosine. Willems et al4 reported dPVP in 9% of veins isolated with an ostial approach with a similar response to isoprenaline. Two cases of isolated PV fibrillation were observed in left common PV following local automatic activity. Takahashi et al9 was able to induce sustained re-entrant PV tachycardia in 2.6% of acutely isolated PV with intra-PV programmed stimulation. Using a high-resolution multipolar PV catheter, Nakagawa et al7 demonstrated PV firing to be more frequent in veins with circumferential as compared with broad or discrete PV–LA connections. Slow dissociated PV activity was observed in 19% of PV acutely after isolation and was more frequent in veins that were considered ‘arrhythmogenic’. Similarly, in the present study we observed dPVP in 19% of veins, which largely dissipated over a 30-min observation period. The higher incidence may be partly explained by more proximal four-vein electrical isolation with image integration into a 3D mapping system to assist in addressing the PV antrum rather than ostium and a longer waiting period following electrical isolation.

Recently Kabra et al10 reported dissociated PV activity in 91% of acutely isolated veins using a strategy of ipsilateral antral isolation. PV activity varied from single ectopics in 43% to sustained PV rhythms in 20% and PV fibrillation in 4.5%. The higher incidence of dissociated PV rhythms may partly be explained by less ablation on the intervenous ridge (18%) compared with the 48% in the current study.

Although there have been a number of reports defining the incidence of dPVP, to our knowledge there is a paucity of data determining the impact of PV potentials on long-term outcome following catheter ablation. Weerasooriya et al2 did not observe an effect on clinical outcome; however, the earlier study was in the era of ostial isolation with PV assessed using either an ablation catheter, circular mapping or basket catheter for 2−5 min and follow-up confined to 9 months.

Electroanatomic origin of dPVP

dPVP probably originate from remnant PV musculature located either distally 2–4 cm from the PV ostium as originally described by Haissaguerre et al11 or more proximally at the PV ostium. The PV–LA junction is invested in epicardial muscle sleeves, which contribute to focal triggers for AF. In the present study we observed a higher proportion of dPVP originating from the superior PV consistent with longer muscle sleeves in the superior compared with the inferior veins.12 The muscle sleeves are thickest at the PV–LA junction and up to twice as thick at the intervenous ridge.13 Rapid or incessant dissociated PV activity was not present in the current study probably due to more extensive ablation at the PV–LA junction compared with previous studies. This included ablation at the intervenous ridge in 40−50% when electrical isolation could not be achieved by wide encirclement alone.14–16 The incidence of dPVP did not differ in the present study between patients who underwent ablation on the intervenous ridge.

The underlying mechanism of PV arrhythmogenesis is yet to be defined, with evidence suggesting multiple possible mechanisms from PV automaticity,17–19 to triggered activity20or localised re-entry.21 Dixit et al22 reported sustained PV firing in response to overdrive pacing, adenosine, isoproterenol and carotid sinus massage. Overdrive pacing led to temporary suppression of the PV activity with rapid (<5 s) recovery of PV firing on cessation of pacing. Isoproterenol led to an increase in PV activity, while adenosine resulted in a mixed response. The PV response to these stimuli supports the role of triggered activity or abnormal automaticity as important in PV arrhythmogenesis.

Overdrive pacing led to temporary suppression of the PV activity with rapid (<5 s) recovery of PV firing on cessation of pacing. Isoproterenol led to an increase in PV activity, while adenosine resulted in a mixed response (40% augmented, 20% suppressed and no effect in 40%). Carotid sinus massage had no effect in 86% of patients. The response of PV activity to these stimuli argues against sustained re-entry for the mechanism of PV arrthymogenesis and suggests that triggered activity or abnormal automaticity may be more likely.

Similarly, the mechanisms responsible for dPVP are similarly unclear. The cycle length shortens in response to isoprenaline consistent with automaticity or triggered activity. The transient nature of the dPVP is intriguing, with just 6% present at 30 min in the present study. dPVP may represent ‘arrhythmogenic’ PV; however, earlier studies in which the ablation was confined to ostial isolation of the arrhythmogenic vein observed an incidence in up to 12%. One may also speculate that dPVP may be due to the acute stimulatory effect of radiofrequency energy on local PV muscle automaticity at the site of electrical isolation, which may explain the transient nature of the observation.

Lack of impact on clinical outcome of dPVP

Given earlier reports associating dPVP with arrhythmogenic veins the present study aimed to determine the impact of dPVP on long-term clinical outcome. The present study did not demonstrate an association between clinical outcome following antral PV isolation and dPVP. Rather PV reconnection was present in all patients who underwent repeat procedures for recurrent AF. Known predictors of acute reconnection include non-paroxysmal AF, hypertension, enlarged LA and sleep apnoea.23 Strategies and technologies continue to involve in an attempt to achieve enduring PV isolation. Kim et al24 randomly assigned patients following antral vein isolation to ablation of residual PV potentials as identified by mapping distal to the ablation line and demonstrating local PV capture without LA connection. Residual PV potentials were observed in 34−47% with no improvement in clinical outcome with adjunctive ablation. The incidence or effect on spontaneous dPVP was not presented.

Study limitations

Given the transient nature of dPVP it is possible that some dissociated PV activity was missed during repositioning of the circular mapping catheter at the time of acute isolation. This may have led to a lower reported frequency of observed dPVP in this study. Twenty-four hour holter cardiac monitoring was performed at 3, 6, 9 and 12 months and at least every 6 months thereafter. Additional Holter monitoring or 7-day monitoring was performed when patients reported symptoms suggestive of recurrence. Although trans-telephonic or 7-day Holter monitoring would be expected to increase the yield of detecting asymptomatic AF, this is unlikely to alter the findings of the present study significantly as the intensity of follow-up duration did not differ between the two groups. In addition, all patients in this study were highly symptomatic before the ablation and it is unlikely that patients had significant ongoing episodes without being aware. The sample size in the present study may have underpowered the ability to detect smaller differences on the impact of DPVP on recurrent AF.


dPVP are a transient phenomenon present in 19% of PV following antral electrical isolation in patients undergoing catheter ablation for AF. The presence of dPVP did not predict recurrent AF following antral electrical isolation.


View Abstract


  • Funding PMK is supported by a research investigatorship from the Cardiac Society of Australia and New Zealand. GL is the recipient of a postgraduate research scholarship from the Cardiac Society of Australia and New Zealand. AT and CM are the recipients of a postgraduate research scholarship from the National Heart Foundation of Australia. AT, CM and GL are recipients of a cardiovascular and lipid research grant from Pfizer.

  • Competing interests None to declare.

  • Ethics approval This study was conducted with the approval of The Alfred Hospital Ethics Department.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.