Pulmonary Vein Stenosis After Atrial Fibrillation Ablation: An Iatrogenic Problem Larger Than the Primary Problem

Catheter-based ablation procedures for patients with symptomatic atrial fibrillation (AF) have steadily increased over the past decade in the United States. Despite significant technological advances and better understanding of the anatomic substrate, AF ablation can still result in rare, yet life-threatening complications. One such complication is pulmonary vein stenosis (PVS) and can incur substantial patient morbidity with recalcitrant symptoms. The incidence of severe PVS has fallen dramatically from 42.4% reported in an early study from 1999¹ to between 0.29% and 3.4% in studies after 2000.^2–7 A high index of suspicion is required as presentation is often delayed (weeks to months after ablation), atypical, and can mimic a case of bronchitis, pneumonitis, or malignancy.^8,9 Symptoms depend on the number of pulmonary veins involved, lesion severity, and presence and extent of collaterals. Diagnosis is often delayed or missed, as routine imaging after ablation is not mandated by the current Heart Rhythm Society consensus statements.¹⁰ If not recognized early, PVS may progress rapidly and result in total pulmonary venous occlusion, chronic pulmonary hypertension, irreversible lung parenchymal damage, and its sequelae.^11,12 Treatment of PVS is also challenging and far from ideal. Pulmonary venous interventions using balloon angioplasty or stenting have been reported from single-center studies, but outcomes have remained suboptimal.^13–17 Restenosis rates requiring reintervention are significant and observed in 56% to 72% of the balloon angioplasty cases compared with 27% to 33% of the stenting cases.^13,14 Recently, surgical repair of severe PVS has been reported in a small series, with a restenosis rate of 38% over long-term follow-up.¹⁸

In this issue of the Circulation: Arrhythmia and Electrophysiology, Raeisi-Giglou et al¹⁹ present the Cleveland Clinic experience of outcomes and management of patients with severe PVS from prior AF ablation. A total of 10 368 patients who underwent AF ablation during the study period (2000–2015) were included. Intracardiac echocardiography and a circular mapping catheter (Lasso) were used to guide pulmonary vein and left atrial ablation. All patients underwent contrast-enhanced spiral computed tomographic (CT) scans routinely at 3 to 6 months post-ablation to assess for PVS. Fifty-two patients (0.5%) developed severe PVS (defined as luminal narrowing >70%) involving a total of 63 pulmonary veins. Interestingly, 93% of the pulmonary veins involved were left sided (51% left superior, 36% left inferior, and 6% left common). The mean age of the cohort was 56 years, and majority of patients were men (85%). Before the ablation procedures that resulted in PVS, 32% had a prior AF catheter ablation and 4% had a prior MAZE procedure. PVS symptoms included shortness of breath (46%), decrease in exercise capacity (42%), cough (25%), chest pain or discomfort (9%), and hemoptysis (5%).

Of 52 patients with severe PVS, 43 (83%) had pulmonary venous interventions performed and consisted of pulmonary vein stenting in 37 patients and balloon dilatation in 6 patients. A total of 53 veins were stented in 37 patients. Over a median of 14 months, 9 veins (17%) had stent restenosis, which required reintervention. There were no procedure-related deaths, but 5 patients had major complications (3 pulmonary venous tears requiring emergent surgical repair, 1 procedure-related stroke, and 1 phrenic nerve injury).

Of 52 patients with severe PVS, 11 (21%) had arrhythmia recurrence over a median follow-up of 25 months. Of these, 7 patients underwent a redo ablation procedure along with 8 patients with severe PVS and symptomatic drug-refractory arrhythmia recurrence after a prior ablation at other centers (total n=15). Pulmonary vein reconnections were noted in most of the veins with severe PVS or stents. Antral ablation was performed to reisolate the veins without progression of PVS in most patients on repeat CT scans. There were no major redo procedure-related complications. Lasso entrapment within pulmonary vein stents occurred in 2 patients but was eventually freed without needing cardiac surgery.

The authors have to be commended on the meticulous follow-up of >10 000 patients spanning 16 years. This is the first study of its kind to examine longitudinally the outcomes and management of patients with severe PVS and adds to the literature on redo ablation procedures in patients with severe PVS or stents. The incidence of severe PVS in this study from a large tertiary care center with experienced operators was 0.5%.¹⁹ However, the true incidence of PVS in the real world is not known as routine imaging is not performed post-ablation in majority of the centers and diagnosis is often missed. Nonetheless, the decline in overall incidence of PVS can be attributed to several factors including modifying ablation techniques with wide area circumferential ablation and antral isolation, using lower temperature/power during ablation, pulmonary venography, and using multimodality imaging including intracardiac ultrasound, CT, or magnetic resonance imaging for accurate delineation of the pulmonary vein ostial anatomy. Particular attention is required for patients undergoing redo AF ablation procedures, as significant baseline ostial narrowing may be present in the absence of clinical symptoms. As the authors suggested in this study, a routine CT scan before a redo procedure¹⁹ or a pulse wave Doppler to document the pulmonary vein velocity before ablation can identify occult PVS and may caution from further ablation.

All patients in the present study underwent routine CT scans 3 to 6 months post-ablation, and 0.5% were noted to have severe PVS, that is, for every 200 CT scans performed, 1 patient was diagnosed with severe PVS. With >100 000 AF procedures performed annually in the United States, is it cost effective to perform routine CT scans in all patients? Perhaps future studies should focus on identifying individuals who are at a high risk of PVS after AF ablation and hence benefit from routine imaging. For example, individuals who require extensive or segmental/ostial ablation to achieve pulmonary vein isolation, those who had small pulmonary veins before ablation, or those who had significant increase in pulsed wave Doppler velocities in the pulmonary veins pre- and post-ablation (>1 m/s). In an elegant study by Teunissen et al⁷ conducted between 2005 and 2016, among 976 patients who underwent pulmonary vein antral isolation, routine pre- and post-ablation screening by magnetic resonance imaging or CT scan demonstrated severe PVS in 0.7% of cases and severe symptomatic PVS in only 0.1%. Not only do routine CT scans expose patients to radiation, they can in some instances overestimate the severity of PVS. In a study by Fender et al,¹³ 6% of the cases were found to have <50% narrowing on pulmonary venography with insignificant pressure gradient despite the preprocedure CT showing severe (>75%) narrowing. In a study by Prieto et al,¹⁴ CT overestimated the rate of complete pulmonary vein occlusion as 13 of 27 occluded veins, were in fact patent at pulmonary artery wedge angiography. These data suggest that it may be appropriate to only screen those patients with signs or suggestive symptoms of PVS.

Patients with moderate-to-severe PVS typically present with cough, dyspnea, chest pain, decreased exercise tolerance, hemoptysis, or rarely with an x-ray finding that is more consistent with a bronchitis or pneumonitis.⁹ A high index of suspicion is hence needed. A delay in diagnosis or misdiagnosis can lead to progression of the PVS that may result in persistent, irreversible venous and arterial morphological changes throughout the lung, including areas both close to and remote from the site of catheter ablation.¹¹ Stenting chronic severe PVS may not be effective in treating the lung damage once small vessel disease has developed.¹¹ Hence, it is crucial to educate the patients, primary care physicians, pulmonologists, and cardiologists about the signs and symptoms of PVS and trigger work-up for PVS if patients experience these symptoms after an ablation procedure so as to avoid delay in diagnosis. When suspected, a dedicated CT scan or a magnetic resonance imaging must be performed promptly to rule out PVS.⁹

Management of patients with severe PVS can be challenging because of high rates of restenosis and recalcitrant symptoms. In the largest series published to date, Fender et al¹³ reported 113 patients with severe PVS involving 178 veins treated with balloon angioplasty and stenting. Over a long-term follow-up of 4.6 years, 56% of the 92 veins treated with balloon angioplasty developed restenosis compared with 27% of the 86 veins treated with stenting. In the stenting group, bare metal stents approved for biliary or peripheral interventions (10 mm diameter) were used in 82 veins, whereas drug-eluting stents (largest commercially available diameter is 4 mm) were used in only 4 veins that were small or had tight stenosis.¹³ In another study, Prieto et al¹⁴ demonstrated that among 34 patients with severe PVS involving 55 veins, restenosis rates were 72% with balloon dilatation compared with 33% with stenting over a median follow-up of 25 months. Risk factors for restenosis included small reference vessel diameter and longer time from AF ablation to intervention for PVS. Notably, long-term patency rates of 80% were observed in veins stented with stents that were at least 10 mm in diameter.¹⁴ These findings suggest that early diagnosis of PVS and prompt referral for stenting when the vessel diameter is large is critical for improved outcomes. De Potter et al,²⁰ in a small series of 5 patients with severe PVS, reported excellent patency rates with drug-eluting stents (mean stent diameter 4.3 mm), with only 1 patient required restenting (restenosis rate of 14.3%) during a mean follow-up of 12 months.

In the present study, the restenosis rates were low at 17% compared with the published literature.¹⁹ The low rates are likely because of routine imaging of all patients after ablation, thus enabling early identification of PVS and prompt intervention. Although there were no procedure-related deaths, major complications were noted in 9.6% of the cases (3 pulmonary venous tears, 1 procedure-related stroke, and 1 phrenic nerve injury), higher than the major complication rate of 3.5% (3 pulmonary venous tears and 2 cardiac tamponade) observed in the series by Fender et al.¹³

Although a small sample size, the present study also provides an important fund of knowledge concerning the arrhythmia burden in patients with severe PVS or stents.¹⁹ Approximately 21% of the patients with severe PVS experienced recurrent symptomatic AF/atrial flutter. During the redo ablation procedure in 15 drug-refractory patients, majority of the stenotic veins were noted to be reconnected, and isolation was achieved by antral ablation, suggesting that prior ablations were ostial. Only 1 patient had progression to severe PVS on repeat CT imaging after redo ablation, requiring reintervention.

In summary, PVS is a rare but dreaded complication after catheter ablation of AF. The incidence in the current era is low (<1%), and there is no robust data to suggest routine imaging after the ablation. A heightened awareness of PVS signs and symptoms among all providers caring for a patient with a recent AF ablation procedure is imperative for early diagnosis and prompt management before the development of chronic occlusion and its sequelae. Pulmonary vein stenting offers better long-term patency rates compared with balloon angioplasty and should be the first-line strategy. Recurrent restenosis rates are high; hence, these patients should be monitored closely for development of symptoms, and routine imaging might be necessary. Further research studies are needed to improve percutaneous intervention outcomes in patients with PVS.