Author + information
- Received November 22, 2016
- Revision received February 9, 2017
- Accepted February 10, 2017
- Published online June 19, 2017.
- Cas Teunissen, MDa,∗ (, )
- Birgitta K. Velthuis, MD, PhDb,
- Rutger J. Hassink, MD, PhDa,
- Jeroen F. van der Heijden, MD, PhDa,
- Evert-Jan P.A. Vonken, MD, PhDb,
- Nick Clappers, MDa,
- Pieter A. Doevendans, MD, PhDa and
- Peter Loh, MD, PhDa
- aDepartment of Cardiology, University Medical Center Utrecht, Utrecht, the Netherlands
- bDepartment of Radiology, University Medical Center Utrecht, Utrecht, the Netherlands
- ↵∗Address for correspondence:
Dr. Cas Teunissen, Department of Cardiology, University Medical Center Utrecht, Heidelberglaan 100, 3584CX, Utrecht, the Netherlands.
Objectives This study aimed to determine incidence of pulmonary vein stenosis (PVS) and evaluate PVS-related symptoms.
Background The real-life incidence of PVS after radiofrequency catheter ablation (RFCA) of atrial fibrillation (AF) is unknown.
Methods All patients who underwent RFCA of AF from 2005 to 2016 with routine pre- and post-ablation screening by magnetic resonance imaging or computed tomography were included. Primary ablation strategy was PV antrum isolation alone in all patients. PVS, defined as a significant reduction in the superoinferior or anteroposterior PV diameter, was classified as mild (30% to 50%), moderate (50% to 70%), or severe (>70%).
Results Sufficient quality imaging of the PV anatomy before ablation and during follow-up (mean 6 ± 4 months) was performed in 976 patients (76.4% men, 59.1% paroxysmal AF). Of these patients, 306 (31.4%) showed mild stenosis, 42 (4.3%) revealed moderate stenosis, and 7 (0.7%) had a severe stenosis in at least 1 PV. Incidence of PVS fluctuated over the past decade. All severe PVS cases were likely caused by ablations being performed inside the PVs. Only 1 (0.1%) patient reported PVS-related symptoms of severe dyspnea during follow-up. Computed tomography revealed a subtotal occlusion of the left inferior PV and a severe stenosis of the left superior PV, requiring stenting.
Conclusions Although mild PVS was frequently observed after RFCA in this large cohort, incidence of severe PVS was <1% and incidence of symptomatic PVS necessitating intervention was negligible. Based on these findings, it seems appropriate to only screen for PVS in patients with suggestive symptoms.
Pulmonary vein stenosis (PVS) is a well-known complication of radiofrequency catheter ablation (RFCA) of atrial fibrillation (AF). It is defined as a reduction of a PV diameter and is likely caused by a vascular response to RF energy application leading to replacement of necrotic myocardium by collagen (1,2). The clinical presentation of PVS, normally appearing 3 to 6 months after RFCA, varies from asymptomatic in the majority of patients to occasional symptoms of severe dyspnea, cough, chest pain, and hemoptysis (3,4). During follow-up, PVS can be detected by computed tomography (CT), magnetic resonance imaging (MRI), or transesophageal echocardiography (TEE) (5–8). Treatment of severe PVS depends on the symptoms, and varies from no treatment to balloon dilatation or stenting (3). If such interventions fail, severe PVS may even necessitate lobectomy (9).
Although PVS is a feared complication, it is not clearly defined and its true incidence remains unknown (10). Among the studies of AF ablations reviewed for the 2012 expert consensus statement, PVS was reported in <10% of studies (1). In studies that did report PVS, various imaging modalities and definitions for PVS were used and the reported incidence varied widely. Incidence up to 42% was shown in an earlier study (11). In contrast, incidence of only 0.5% was reported in a large systematic review on complications of RFCA (12). Yet, many of the studies included in this review only screened for PVS in case of symptoms suspected to be related to PVS. Because moderate and even severe PVS can remain asymptomatic (3), true incidence is likely to be underestimated (10).
In this study we report a large single-center cohort of patients who were routinely screened for PVS after RFCA of AF. Goals of this study were to determine incidence of PVS and assess PVS-related symptoms. Changes in incidence of PVS over the past decade were analyzed.
This study was approved by the local ethics committee. All consecutive patients with symptomatic, drug-refractory, or drug-intolerant AF who underwent first PV antrum isolation (PVAI) in the University Medical Center Utrecht over an 11-year period from January 2005 to January 2016 were analyzed. As part of standard clinical care, all patients routinely underwent pre-ablation imaging of the left atrium (LA) and PV anatomy and post-ablation imaging mostly after 4 to 6 months (range 3 to 12 months) to screen for PVS. The main reason for this wide range is that if a redo ablation procedure was scheduled, post-ablation imaging was performed prior to this procedure.
Patients who did not undergo imaging of the PV anatomy before or after ablation (e.g., due to claustrophobia or refusal of the control scan) or those in whom PV diameters could not be accurately measured due to insufficient scan quality were excluded. Baseline patient characteristics were prospectively collected and comprised sex, age, AF type, history of AF (years since AF was diagnosed), cardiovascular risk factors, structural heart disease, CHA2DS2-VASc (congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack or thromboembolism, vascular disease, age 65 to 74 years, sex category) score, and LA size measured on echocardiography.
Imaging protocols before ablation and during follow-up
From 2005 to 2011, patients underwent gadolinium-enhanced MRI of the heart ≤6 months before and around 6 months after PVAI. Our MRI protocol has been described previously (13). In short, a 1.5-T MRI system (Philips Healthcare, Best, the Netherlands) was used to obtain a MR angiogram with single breath-hold 3-dimensional fast spoiled gradient echo imaging in coronal view. Scan parameters were as follows: repetition time = 4 ms, echo time = 1 ms, radiofrequency flip angle 35°, field of view 400 mm, matrix size 272 × 173, slice thickness 3 mm, and gap + slice 1.5 mm. As MRI could be planned up to 1 month before PVAI, all these patients underwent TEE just prior to the procedure to rule out LA and LA appendage (LAA) thrombus.
From 2011 onward, an electrocardiography-gated cardiac CT angiography (CTA) was performed ≤7 days prior to and around 6 months after ablation. Our CTA protocol has been described before (14,15). In short, CTA was performed using a 256-slice CT system (Philips Healthcare). The scan parameters were as follows: collimation 128 × 0.625 mm, tube voltage 80 to 120 kV, tube current 195 to 210 mAs, rotation time 0.27 s. Images were reconstructed with a slice thickness of 0.9 mm and a reconstruction increment of 0.45 mm. TEE was only performed if LA or LAA thrombus could not be ruled out by CT. In the transition period of MRI to CT imaging follow-up imaging was performed with the same imaging modality as the initial scan.
PV anatomy and diameters
Source images were transferred to a commercially available workstation (Vitrea 2, Vital images or Intellispace, Philips Healthcare) for 3-dimensional reconstructed images of the LA, LAA, and PVs and measurement of PV diameters. The number and anatomy of PVs and presence of common trunks were assessed in all patients. A common trunk was defined as a superior and inferior PV that have joined before entering the LA. Both superoinferior (SI) and anteroposterior (AP) diameters were measured of the PVs at the level of the ostium and the common trunks (Figure 1). During follow-up, PV diameters were measured in the same views and on identical locations using the same software as in the initial scan.
Classification of PVS
In the 2012 Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert Consensus Statement, PVS is defined as a reduction of a PV diameter and classified as mild (<50%), moderate (50% to 70%), and severe (>70% diameter reduction) (1). We classified mild stenosis as a reduction of 30% to 50% and moderate and severe stenosis according to the Consensus Statement.
In the 2012 Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert definition of PVS, it is not specified whether a PV diameter reduction should be present in 1 or both measured directions to meet the definition (1). In our study, PV diameters were measured in the SI and AP direction and reduction in both diameters are presented. To classify for PVS, a reduction of at least 30% in 1 of the PV diameters was sufficient. Diameter reduction in both directions was also noted.
In all patients, the primary ablation strategy was PVAI without additional ablation of complex fractionated atrial electrograms and without linear ablation lesions in the LA. Pre-ablation workup, electrophysiological study and post-ablation care have been described in detail elsewhere (16,17). In short, a 3-dimensional cardiac mapping system (CARTO 3, Biosense Webster, Diamond Bar, California; or EnSite NavX, St. Jude Medical, St. Paul, Minnesota) was used to obtain a 3-dimensional reconstruction of the LA, including the LAA and the PVs. The right and left PVs were widely encircled at their antrum using an irrigated tip catheter (ThermoCool irrigated tip catheter, Biosense Webster) from 2005 to the end of 2013, and contact force (CF) sensing catheters (ThermoCool SmartTouch, Biosense Webster; or TactiCath, St. Jude Medical) from the end of 2013 onward. Radiofrequency energy was delivered with a maximum power of 40 W and a maximum temperature of 42°C. At the posterior wall, maximum power was limited to 30 W. Irrigation flow rate was 17 ml/min with conventional irrigated tip catheters and 30 ml/min with CF sensing catheters.
Patients were seen in the outpatient clinic 3, 6, and 12 months after ablation. Recurrence of AF and possible PVS-related symptoms such as cough, dyspnea (on exertion), chest pain, hemoptysis, and recurrent lung infections were assessed during every visit.
Categorical patient variables are shown as number and percentage. Continuous variables are expressed as mean ± SD or median (25th to 75th percentile), depending on the normality of the data distribution. PV diameter reduction was calculated for all PVs in all patients. Statistical analyses were performed using SPSS version 21.0 (IBM, Armonk, New York).
From January 2005 to January 2016, 1,167 patients underwent primary PVAI. Of these 1,167 patients, 976 (83.6%) patients with sufficient imaging quality for assessing the PV anatomy both before ablation and during follow-up were included in this study (Figure 2). None of the excluded patients reported PVS-related symptoms.
Patient characteristics are shown in Table 1. Of the 976 included patients, most were men (76.4%) and suffered from paroxysmal AF (59.1%). Mean LA size before ablation was 42 ± 7 mm on echocardiography.
PV anatomy prior to ablation
Prior to PVAI, 429 patients underwent MRI and 547 patients underwent CTA. PV characteristics are shown in Table 2. Most patients (88.9%) had 4 PVs entering the LA. More than 5 PVs were rare (0.9%) and only 1 patient had 7 PVs, with 2 left-sided PVs and 5 PVs on the right side. In 138 patients (14.1%) a left-sided common trunk was present. Overall, the largest PV was the right superior, followed by the right inferior, left superior, and the left inferior PV.
Incidence of PVS
Follow-up CTA or MRI was performed with a mean duration of 6 ± 4 months after PVAI. Table 3 shows an overview of the results. On average, all PVs significantly decreased in size, with a greater reduction seen in the SI diameter than in the AP diameter. The largest mean percentage reduction in diameter was found in the left common trunk (SI diameter 16 ± 13%, AP diameter 14 ± 15%).
Incidence of PVS in all 976 patients is shown in Table 3. Overall, PVS occurred more frequently in the left PVs than in the right PVs. In total, 355 (36.3%) patients showed mild, moderate, or severe PVS in at least 1 PV (SI or AP diameter). In 306 (31.4%) patients, a total of 514 PVs (including the common trunks) revealed mild stenosis. Of those 514 PVs, 130 (25.3%) showed a diameter reduction of 30% to 50% in both measured diameters, with the greatest incidence of mild PVS occurring in the left common trunk (14.5%) followed by the left superior PV (14.2%). Among patients with mild PVS, 2 different patterns in PV narrowing could be distinguished; either a waist-like tapering or a lumen narrowing with a smooth transition to the atrium (Figure 3). Of all 514 PVs with a mild stenosis, 430 (83.7%) revealed a waist-like tapering and 84 (16.3%) showed a smooth transition narrowing.
A moderate stenosis in 1 diameter was seen in 42 (4.3%) patients, in a total of 50 PVs. All 50 moderate PVS showed waist-like tapering, and 17 (34%) PVs showed a diameter reduction of 50% to 70% in both diameters. Incidence of moderate stenosis varied between 0.5 and 2.2% in all PV locations, and was greatest in the left common trunk.
Seven (0.7%) patients had a severe stenosis in 1 diameter in a total of 9 PVs. All 9 PVs revealed a waist-like tapering, and 8 of 9 (88.9%) PVs showed a >70% diameter reduction in both diameters. Severe PVS occurred in the left superior, left inferior, right superior, and right middle PVs.
PVS-related symptoms were reported by 1 of the 7 patients with a severe stenosis, and by none of the patients with mild and moderate PVS. One patient (0.1%) suffered from dyspnea on exertion 6 months after PVAI. Routine follow-up CT revealed a subtotal occlusion of the left inferior PV and a severe stenosis of the left superior PV (SI 74% reduction, AP 60% reduction) in this patient. Balloon dilation was performed twice in both left-sided PVs, but restenosis occurred after both attempts. Subsequently, both PVs were successfully stented, leading to relief of symptoms. Both PVs remained free of restenosis after 3 years of follow-up.
Incidence of PVS over time
The number of mild and moderate PVS showed fluctuations over the past decade, with a gradual decrease from 2005 to 2008, followed by an increase to 2011 (Figure 4). Six of the 7 cases of severe stenosis occurred in 2011 and 2012. After 2013, there was again a minor increase in mild to moderate PVS and another case of severe PVS occurred in 2015 (Figure 5). Retrospectively, in all patients with moderate and severe PVS, some ablations were performed inside the affected PV. We have also reviewed 50 procedures of patients without PVS. In 40% of these patients, at least 1 ablation has been applied inside a PV. Mostly, comparable to the patients that developed PVS, this occurred in the left PVs on the ridge between the PV and the LAA.
To the best of our knowledge, this is the first study to systematically analyze incidence of PVS in a large cohort of patients who were routinely screened after RFCA. It revealed that mild stenosis occurs frequently after PVAI. Yet, incidence of moderate and severe stenosis is low, and occurrence of symptomatic PVS necessitating intervention is negligible. Incidence and severity of PVS fluctuated over the past decade.
Incidence of PV stenosis
After identification of the PVs as a potential origin of AF almost 2 decades ago, RFCA treatment focused on focal trigger ablation inside the PVs (18). Unfortunately, ablation in the PVs resulted in a high incidence of PVS (2,11). To prevent PVS, circumferential ablations at the PV ostium were introduced (19). Later, ablations shifted to the PV antra in the LA body (20). The shift of ablation therapy from inside the PVs toward the PV antra has resulted in a decrease in the incidence of PVS over the past 2 decades. From 1999 to 2004, reported incidence ranged from 0% to 44% (median 5.4%), whereas later studies reported incidence from 0% to 19% (median 3.1%) (10). However, various imaging modalities and different criteria for PVS (mostly luminal narrowing >50%) were used (10).
In our study, incidence of PVS was higher in the left PVs compared to the right PVs. Retrospectively, this was probably caused by ablations performed inside the left PVs on the ridge between the PV and the LAA. However, when we reviewed ablation procedures of patients that did not develop PVS, in 40% of patients ablations were also performed inside the PV. Therefore, development of PVS is apparently multifactorial. Beside ablation inside the PV, other ablation–related factors such as the CF, ablation duration, and power settings as well as individual patient-related factors such as PV anatomy and tissue characteristics may influence the development of PVS.
Incidence of PVS fluctuated over the past decade. Interestingly, most cases of severe stenosis occurred in 2011 and 2012. In these years, several ablations were performed using a long sheath to create better contact in the LA. In a few of these patients, ablations were performed inside the PVs and with the good contact created by the long sheath this probably resulted in the increase of PVS. After this was noted, RFCA was performed without long sheaths and ablations in the PVs were strictly avoided resulting in a drop of all PVS. After introduction of CF sensing catheters at the end of 2013, another small increase in PVS was noted. Hypothetically, CF feedback may result in ablations being applied with greater force. Therefore, extra caution should be taken with CF sensing catheters not to ablate inside the PVs. A randomized controlled study comparing CF sensing catheters and non–CF sensing catheters did not reveal any difference in incidence of PVS (21).
Catheter ablation treatment with different catheters or energy sources has resulted in varying incidence of PVS. After circumferential ablation with the PV ablation catheter system, moderate PVS occurred in approximately 15% of patients (22). First-generation cryoballoon ablation showed a very low risk of PVS (23–25). Data on the second-generation cryoballoon are limited and restricted to case reports (26).
Difficulties in defining PVS
In our study we used current definitions of PVS according to the 2012 Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert Consensus Statement. Some difficulties arose using these definitions. We measured PV diameters in 2 directions: SI and AP. This is not described in the definitions of the Consensus Statement. In our study, reduction in 1 of both diameters was noticed frequently, whereas reduction in both diameters was only seen in one-quarter of the mild PVS subgroup and one-third of the moderate PVS subgroup. Only in severe PVS, almost all PVs showed a >70% diameter reduction in both diameters. The severity of PVS is dependent on reduction of both diameters. Expressing PVS in area reduction may solve this issue.
Another difficulty was the definition of a mild stenosis. In our center, mild stenosis of at least 1 PV was found in 31.4% of patients using a modified definition of mild stenosis to only include stenosis within the 30% to 50% range. In a previous small study (n = 41) that used the 2012 Expert Consensus Statement criteria for PVS, 59% of patients showed mild stenosis (<50% diameter reduction on MRI) of at least 1 PV (27). Whether all <50% reductions in PV diameter represent PVS is debatable. It is known that LA size significantly decreases after PVAI, irrespective of recurrence of AF (16). Most likely, this is caused by a combination of shrinking due to ablation-induced fibrosis and reverse remodeling due to a decreased AF burden (16). A similar response might be expected for the PV lumen. We noticed 2 different patterns in PV narrowing in mild stenosis, either a waist-like tapering, probably related to fibrosis at the ablation site, or a decrease in lumen size with smooth transition to the atrium that may also be explained by global shrinkage of the atria. Because of the high incidence of mild PVS and the fact it has no clinical implications, mild PVS should perhaps not be seen as a complication of AF ablation but rather as an ablation effect.
The occurrence of PVS-related symptoms depends on several aspects. It is related to the number of affected PVs, severity of PVS, response of the pulmonary vasculature to the lesion, presence of collaterals, clinical setting and time course of PVS (3). Patients requiring treatment of severe, symptomatic PVS decreased over the past years. A large worldwide survey studied 16,300 patients who underwent AF ablation. From 1995 to 2003, 0.71% of patients needed intervention for severe PVS compared with 0.29% from 2003 to 2006 (28). A further decline to 0.1% is seen in our center from 2005 to 2016.
Management of PVS
Currently, treatment of PVS is performed in patients with PVS-related symptoms. PVS-related symptoms include cough, dyspnea (on exertion), chest pain, decreased exercise tolerance, hemoptysis, and recurrent lung infections. As these symptoms are nonspecific, they should be thoroughly evaluated. In case of symptomatic PVS, stenting may be preferred over balloon angioplasty. In a recent, large observational study of 124 patients with severe PVS, stenting resulted in significantly less restenosis compared to balloon angioplasty. Nevertheless, restenosis still occurred in 24% of patients treated with a stent (29).
Management of patients with asymptomatic PVS is poorly studied and not mentioned in guidelines. Because the incidence of severe, asymptomatic PVS is very low, patient care should be individualized. It may be desirable to discuss patients in a multidisciplinary team to consider additional diagnostics such as a ventilation–perfusion scan or pulmonary function test. Furthermore, in all patients, symptoms should be closely monitored during follow-up (3).
Significance of study findings
We demonstrated that mild PVS occurs frequently after PVAI and that incidence of moderate and severe stenosis is low. Yet, all patients with mild and moderate PVS and most patients with severe stenosis remained asymptomatic during 1 year of follow-up. Whether early diagnosis and treatment of asymptomatic PV stenosis would result in any prognostic benefit to the patient is unknown. This questions the need of routine screening for asymptomatic PVS. To reduce costs, and radiation dose in case of CT imaging, screening for PVS only in patients with suggestive symptoms seems appropriate.
This was a single-center cohort study. Our results cannot be translated to other institutions and other ablation techniques. However, the reported PVS rate may be indicative for centers using RFCA with a comparable ablation strategy.
Due to the study design, no predictors for PVS could be identified.
Despite PVS was asymptomatic in almost all patients after 1 year of follow-up, long-term pulmonary effects are unknown. Conclusions on long-term outcomes cannot be drawn from our study.
Although overall incidence of mild PVS after PVAI by RFCA was high, moderate to severe PVS rarely occurred. Only 1 (0.1%) patient with severe PVS was symptomatic and required intervention. Ablations performed inside the PVs with good tissue contact were most likely the cause of severe PVS. Moreover, they also probably caused fluctuations in incidence of PVS over the past decade. It may be advisable to only screen for PVS in patients with suggestive symptoms.
COMPETENCY IN MEDICAL KNOWLEDGE: After PVAI with radiofrequency energy, incidence of mild PVS was high. Nevertheless, incidence of severe PVS was <1% and incidence of symptomatic PVS necessitating intervention was negligible.
TRANSLATIONAL OUTLOOK: Based on our findings, it seems appropriate to only screen for PVS in patients with suggestive symptoms. Nevertheless, long-term outcomes of untreated, asymptomatic PVS are unknown. Future research is needed to identify the risk of complications resulting from asymptomatic PVS.
The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- computed tomography angiography
- left atrium
- left atrial appendage
- magnetic resonance imaging
- pulmonary vein
- pulmonary vein antrum isolation
- pulmonary vein stenosis
- radiofrequency catheter ablation
- transoesophageal echocardiography
- Received November 22, 2016.
- Revision received February 9, 2017.
- Accepted February 10, 2017.
- 2017 American College of Cardiology Foundation
- Calkins H.,
- Kuck K.H.,
- Cappato R.,
- et al.
- Taylor G.W.,
- Kay G.N.,
- Zheng X.,
- Bishop S.,
- Ideker R.E.
- Holmes D.R. Jr..,
- Monahan K.H.,
- Packer D.
- Packer D.L.,
- Keelan P.,
- Munger T.M.,
- et al.
- Dill T.,
- Neumann T.,
- Ekinci O.,
- et al.
- Saad E.B.,
- Rossillo A.,
- Saad C.P.,
- et al.
- Wittkampf F.H.,
- Vonken E.J.,
- Derksen R.,
- et al.
- Rostamian A.,
- Narayan S.M.,
- Thomson L.,
- Fishbein M.,
- Siegel R.J.
- Chen S.A.,
- Hsieh M.H.,
- Tai C.T.,
- et al.
- Gupta A.,
- Perera T.,
- Ganesan A.,
- et al.
- Teunissen C.,
- Habets J.,
- Velthuis B.K.,
- Cramer M.J.,
- Loh P.
- Pappone C.,
- Rosanio S.,
- Oreto G.,
- et al.
- Reddy V.Y.,
- Dukkipati S.R.,
- Neuzil P.,
- et al.
- De Greef Y.,
- Tavernier R.,
- Raeymaeckers S.,
- et al.
- Tse H.F.,
- Reek S.,
- Timmermans C.,
- et al.
- Packer D.L.,
- Kowal R.C.,
- Wheelan K.R.,
- et al.
- Luik A.,
- Radzewitz A.,
- Kieser M.,
- et al.
- Miyazaki S.,
- Ichihara N.,
- Iesaka Y.
- Dong J.,
- Vasamreddy C.R.,
- Jayam V.,
- et al.
- Cappato R.,
- Calkins H.,
- Chen S.A.,
- et al.
- Fender E.A.,
- Widmer R.J.,
- Hodge D.O.,
- et al.