A decade of atrial fibrillation ablation

All consecutive patients with symptomatic, drug-refractory or drug-intolerant AF who underwent first PVAI in the University Medical Center Utrecht from 2005 to 2015 were included in this analysis. Baseline patient and procedure characteristics, and 1‑year follow-up data were prospectively collected. This study was approved by the institutional review board.

Prior to the first ablation, baseline characteristics were collected and included: sex, age, type of AF, history of AF (years since AF was diagnosed), risk factors for cardiovascular disease, CHA₂DS₂-VASc score, congestive heart failure, structural heart disease and left atrial (LA) size (end-systolic LA diameter in the parasternal long axis view on echocardiography). AF type was classified as paroxysmal, persistent or longstanding persistent according to the HRS/EHRA/ECAS 2012 Consensus Statement on Catheter and Surgical Ablation of AF [4].

Pre-ablation workup, electrophysiological study and post-ablation care have been described in detail elsewhere [6, 7]. In short, a three-dimensional cardiac mapping system (EnSite NavX and Velocity; St. Jude Medical Inc. or Carto; Biosense Webster) was used to obtain a three-dimensional reconstruction of the left atrium, including the left atrial appendage and the PVs. The left and right PVs were widely encircled at their antrum using an irrigated tip catheter (ThermoCool irrigated tip catheter, Biosense Webster, ThermoCool SmartTouch, Biosense Webster, and TactiCath; St. Jude Medical). During the complete duration of the study, primary ablation strategy was PVAI without additional ablation of complex fractionated atrial electrograms and without linear ablation lesions in the left atrium.

During this study, some major changes in procedural care have been implemented. An overview is shown in Fig. 1.

Patients were seen at the outpatient clinic at 3, 6 and 12 months after the procedure. The patient’s rhythm status was evaluated using patient history and a 12-lead electrocardiogram at every visit and with additional 48-hour Holter recordings at 3 and 6 months. After 3 to 6 months of follow-up, magnetic resonance imaging or computed tomography of the heart was routinely performed to rule out PV stenosis.

One-year success was defined according to the 2012 HRS/EHRA/ECAS consensus statement as complete freedom from AF, atrial flutter (AFl) or atrial tachycardia recurrences following the 3‑month blanking period in the absence of Class I and III antiarrhythmic drug therapy [4].

Patient characteristics were reported as percentages, counts, median (25th to 75th percentile) or mean ± SD, as appropriate. For categorical variables, significant variation over time was determined using the chi-square test. For continuous variables, significant variation over time was analysed using ANOVA in case of homogeneous variance or Kruskal-Wallis H-test if not homogeneous. The CHA₂DS₂VASc score was dichotomised at the clinically relevant cut-off value of ≥2.

Dispersion of patient characteristics over the past decade is shown in Tab. 1. Mean age at ablation increased from 54 ± 9 years in 2005 to 61 ± 10 years in 2014 (p = 0.001). In 2005, 73.4% of patients suffered from paroxysmal AF, which decreased to 45.3% in 2014 (p < 0.001). Incidence of long-standing persistent AF resembled a parabolic form (Fig. 2). History of AF (years since diagnosis) significantly decreased from 7 to 4 years. In 2005 18% of patients had a CHA₂DS₂-VASc score ≥2, in 2014 40.6% (p = 0.023).

Tab. 2 shows procedural characteristics and outcomes of the past 10 years. In 2005, mean procedure duration from femoral vein access to catheter withdrawal was 237 ± 53 min, which decreased to 163 ± 41 min in 2014. The main decrease in procedure duration was noticed in 2013 after introduction of contact force (CF) sensing catheters (Fig. 3). Fluoroscopy time significantly decreased from 41 ± 17 to 19 ± 8 min and total radiation exposure from 465 (263–687) to 210 (118–376) mGy (Fig. 3). A consistent decrease in fluoroscopy time and radiation exposure was observed from 2010 onwards, after implementation of novel three-dimensional navigation systems and later CF catheters (Fig. 3).

Of the 975 included patients, 19 (1.9%) were lost to follow-up. Of the remaining 956 patients, 55.3% were completely free of atrial tachyarrhythmia recurrence off antiarrhythmic drug therapy 1 year after primary PVAI. During the entire study duration, most recurrences were based on AF. Only in 4.7% of patients recurrences we based on left-sided atrial tachycardia or AFl. This was consistent over the years.

Over the past decade, the overall success rate did not change. In 2005, one-year success was 55.6%, in 2014 this was 54.8%. Also in patients with paroxysmal AF, success remained similar. In patients with non-paroxysmal AF, 1‑year success was lower and demonstrated a more fluctuating course (Fig. 2). Independent predictors of recurrences of atrial tachyarrhythmia after primary PVAI were female sex (HR: 1.58, 95% CI: 1.08–2.31, p = 0.018), persistent AF (HR: 1.50, 95% CI: 1.09–2.07, p = 0.013), longstanding persistent AF (HR: 2.23, 95% CI: 1.37–3.62, p = 0.001) and history of AF (HR: 1.04, 95% CI: 1.01–1.06, p = 0.008). Year of ablation was no significant predictor for recurrence (HR: 0.96, 95% CI: 0.91–1.02, p = 0.190). So, independent of patient characteristics, procedure success did not change over a 10-year period.

The number of redo procedures within 2 years after the primary ablation procedure and the number of patients with reconnection of at least one PV observed during these redo procedures (analysed per date of the index procedure) remained unchanged over the past decade (Tab. 2).

The overall complication rate has significantly decreased in the past decade (Tab. 3). Vascular complications, as well as moderate PV stenosis (50–70% diameter reduction), significantly decreased. Severe PV stenosis (>70% diameter reduction) occurred in 6 patients, all in 2011 and 2012. Of all the patients with a moderate or severe PV stenosis, one was symptomatic. This patient suffered from a total occlusion of the left inferior PV and a severe stenosis of the left superior PV. Both PVs were successfully stented. Cerebral vascular accident occurred once in 2006 and a suspected transient ischaemic attack twice in 2007. After 2007, no thromboembolic complications occurred (p = 0.049). Incidence of other complications (e. g. tamponade) did not change over time. In the 10 years, one patient died suddenly 5 weeks after PVAI due to cardiac tamponade caused by a peri-epicarditis. No atrial-oesophageal fistula occurred.

In this study, we have highlighted real-world shifts in patient characteristics and procedural outcomes over the past decade. As the use of RFCA increased and it gained the referring physician’s trust, it was employed in a more heterogeneous patient population and earlier after patients were diagnosed with AF. It is increasingly being performed in older patients with more comorbidity and persistent AF. Procedure duration, fluoroscopy time and radiation dose substantially have decreased, likely due to implementation of novel three-dimensional cardiac mapping systems and CF sensing catheters. Due to a decrease in vascular complications and moderate PV stenosis, the total complication rate has significantly decreased. Remarkably, 1‑year single procedure success and the rate of PV reconnection observed during redo procedures have remained similar.

Ablations have increasingly been performed in patients suffering from persistent AF, while there was a decrease in patients with long-standing persistent AF. In the past decade, disappointing results of RFCA in long-standing persistent AF have been presented in several studies, despite various ablation strategies [6, 8, 9]. These results may have led to a more reluctant approach in this subset of patients. On the other hand, possibly less patients reached 1 year of continuous AF. In more recent years, efforts were made to treat patients with a few months of continuously present AF with priority.

Although the incidence of tamponade and thromboembolic complications were comparable with previous studies, the rate of PV stenosis was higher in our study [10]. In contrast to our approach, other institutions only screened for PV stenosis in case of symptoms. True incidence of PV stenosis is therefore likely underestimated in literature.

In the early years of this study, oral anticoagulation therapy was interrupted preceding the ablation procedure and patients were bridged with subcutaneous low-molecular-weight heparin. Several studies, however, revealed that uninterrupted oral anticoagulation therapy throughout the procedure resulted in less bleeding and thromboembolic complications [11]. We have adopted this strategy, which may have resulted in the decrease of vascular complications.

Regarding the unchanged success rate, two observations stand out: the rate of PV reconnection encountered during redo procedures has not decreased and success rates are lower in patients with non-paroxysmal-AF. To increase the efficacy of the contemporary procedure, two major challenges have to be overcome: 1) achieving durable PV isolation by creating permanent, transmural ablation lesions and 2) sufficient understanding, identification and ablation of the atrial substrate in a subset of patients with non-paroxysmal AF.

Prevention of PV reconnection by creating durable ablation lesions has been extensively investigated and several promising technical and scientific breakthroughs have emerged. First, CF sensing catheters have been developed. In RFCA, insufficient CF may result in ineffective ablation lesions, whereas excessive CF may result in complications such as cardiac tamponade. Despite the fact that safety and applicability of CF catheters has been demonstrated [12], two randomised studies failed to confirm superiority of CF catheters over non-CF catheters [13, 14]. However, both studies indicated in subgroup analysis that sufficient CF resulted in favourable outcomes [13, 14]. Furthermore, two studies showed that procedure duration, fluoroscopy time and radiation dose were significantly reduced by the use of CF catheters.[13, 15]. Besides CF sensing catheters, advanced three-dimensional navigation systems allowing more anatomical details have shown to significantly reduce fluoroscopy time [16].

The use of adenosine formed another encouraging innovation. In the ADVICE trial, adenosine-guided PV isolation led to a significant absolute risk reduction of 27.1% in arrhythmia recurrence in patients suffering from paroxysmal AF [17].

Last, other ablation techniques might increase persistence of ablation lesions. Although robotic-assisted RFCA, second generation cryoballoon and laser balloon ablations form appropriate alternatives for manual RFCA, to date, none of these methods have proven to be superior [18‐22].

We have revealed that success of the RFCA did not improve significantly despite growing experience and technical developments. For over a decade, our procedural endpoint has been persistent isolation of the PVs after a 30-minute observation period. However, this is not an adequate predictor of permanent isolation. Future research should focus on techniques that will allow transmural and permanent ablation lesions. Furthermore, understanding of the changing patient characteristics may facilitate designs of future studies.

This study has some limitations. It is a single-centre cohort study. Our results may not be translated to other institutions. Nevertheless, procedural outcomes may be representative for institutions with a similar approach.

As we did not routinely perform 48-hour Holter recordings after 12 months of follow-up, some asymptomatic AF episodes may have been missed.

Changes in procedure characteristics and outcomes are multifactorial. Since our study was a descriptive study, it was not designed to study the effect of all separate contributing factors. Besides technical developments, other factors such as personal learning curves can attribute. However, it is challenging to estimate the exact influence of personal learning curves. Over the past, decade ablations were performed by starting and experienced electrophysiologists as well as fellows in our tertiary centre.