Long-term oral anticoagulant after catheter ablation for atrial fibrillation

Abstract

Catheter ablation is superior to antiarrhythmic therapy for the reduction of symptomatic atrial fibrillation (AF), recurrence, and burden. The possibility of a true ‘rhythm’ control strategy with catheter ablation has re-opened the debate on rate vs. rhythm control and the subsequent impact on stroke risk. Some observation studies suggest that successful AF catheter ablation and maintenance of sinus rhythm are associated with a decrease in stroke risk, while the CABANA trial had demonstrated no apparent reduction. Other observational studies have demonstrated increased stroke risk when oral anticoagulation (OAC) is discontinued after catheter ablation. When and in whom OAC can be discontinued after ablation will need to be determined in properly conducted randomized control trials. In this review article, we discuss our current understanding of the interactions between AF, stroke, and anticoagulation following catheter ablation. Specifically, we discuss the evidence for the long-term anticoagulation following successful catheter ablation, the potential for OAC discontinuation with restoration of sinus rhythm, and novel approaches to anticoagulation management post-ablation.

Introduction

Catheter ablation is an increasingly employed rhythm control intervention in patients with symptomatic atrial fibrillation (AF). Multiple randomized controlled trials have shown that catheter ablation is superior to antiarrhythmic drug therapy for the reduction of AF recurrence and AF burden, and improving quality of life.^1–3 Furthermore, in select populations such as patients with concomitant heart failure, catheter ablation of AF may be potentially useful therapy for the reduction of hospitalization and mortality.^4–6

Despite the well-known benefits of catheter ablation, less clear is about the association between AF ablation and subsequent stroke risk. Some observation studies have suggested reduced risk of stroke and systematic embolism following maintenance of sinus rhythm with catheter ablation. However, the CABANA trial demonstrated no reduction in stroke risk following ablation. Nonetheless, there are still limited prospective data to address the necessity of oral anticoagulation (OAC) following successful catheter ablation.⁷

In this review article, we will discuss our understanding of the nexus between AF, stroke, and anticoagulation following catheter ablation. Specifically, we will examine the current evidence on the long-term anticoagulation post-ablation, and the potential for OAC discontinuation with restoration of sinus rhythm.

Historical perspective on rhythm control and subsequent stroke risk

Over two decades ago, the prevailing paradigm favoured a rhythm-control approach with the belief that maintenance of sinus rhythm could circumvent the adverse consequences of AF such as an increased risk for stroke from left atrial thrombi.⁸ Intuitively, restoration of sinus rhythm, or at least a decrease in AF burden, would reduce the deleterious effects of AF that contribute to an increased risk of stroke, such as atrial stunning, progressive atrial enlargement and remodelling, circulatory stasis in the left atrial appendage and increased platelet reactivity.^9–12

This paradigm was challenged by several landmark randomized clinical trials that failed to demonstrate that the strategy of rhythm control with AADs was superior to a rate control strategy for reducing stroke risk and other ‘hard’ cardiovascular endpoints such as mortality (Table 1).^13–18 In the largest of these trials, the AF Follow-up Investigation of Rhythm Management (AFFIRM), cardioembolic events actually were numerically more common in those treated with rhythm control. Most cardioembolic events occurred when warfarin was discontinued or in the setting of a subtherapeutic INR, which underscores our current clinical approach of continued anticoagulation regardless of a rate or rhythm control strategy.¹⁴

However, the failure of a rhythm control strategy in reducing stroke may indicate more about the inadequacies of pharmacologic antiarrhythmic therapy rather than the rhythm control strategy itself. Importantly, only two-thirds of patients maintain sinus rhythm at 1 year on amiodarone, our most effective antiarrhythmic currently available.²⁰ Other antiarrhythmic drugs are even less effective with only 30–50% of patients maintaining sinus rhythm at 1 year.^21–23 Interestingly, post hoc analyses suggest that successful maintenance of sinus rhythm may modify subsequent stroke risk. For example, in a post hoc analysis of AFFIRM, the presence of AF (modelled as a time-dependent covariate) was associated with a 60% increased risk in stroke [hazard ratio (HR) 1.60, 95% confidence interval (CI) 1.11–2.30; P = 0.01].²⁴ When examining stroke after catheter ablation, it is important to consider that there are many factors that influence its likelihood of occurrence, including the patient’s underlying risk, atrial substrate, the degree of AF burden after ablation, the impact of ablation on left atrial function, and the impact of any therapies that might impact stroke risk (Figure 1). It is also important to understand that like stroke prevention therapy in the general AF population, a one-size fits all approach is not likely as risk factors and amount of AF will likely impact stroke risk (Figure 2).

Rate vs. rhythm in the era of catheter ablation

The development of radiofrequency catheter ablation figuratively re-ignited the debate between rate vs. rhythm control, and there has been renewed interested in the potential benefits of a more-effective rhythm control strategy. Consistent with this renewed interest, several large observational studies have demonstrated that catheter ablation may be associated with lower risk of stroke when compared with antiarrhythmic drug therapy. Using US claims data from IBM Marketscan databases to derive a propensity-score matched community sample of 1602 AF patients, Reynolds et al.²⁵ found that AF ablation was associated with a reduced risk of any stroke or transient ischaemic attack over 3 years of follow-up (HR 0.62, 95% CI 0.44–0.86; P = 0.005).

Friberg et al.²⁶ conducted a similar study and compared outcomes between those treated with catheter ablation vs. medical therapy alone in a 1:1 propensity-score matched cohort of 5672 AF patients from the national Swedish Patient Register. Over a mean follow-up of 4.4 ± 2.0 years, the overall annual rates of stroke were low: 0.70% for catheter ablation and 1.01% for medical therapy alone. After multivariable adjustment for CHA₂DS₂VASc score, time in therapeutic INR range and baseline medications, catheter ablation was associated with a 31% lower stroke risk compared with those patients who did not undergo ablation (HR 0.69, 95% CI 0.5–0.93, P = 0.016). However, the benefit of catheter ablation was observed primarily among patients at a higher stroke risk as reflected by a CHA₂DS₂VASc score ≥2 (HR 0.39, 95% CI 0.19–0.78, P = 0.008). There was no statistically significant difference in stroke rate among patients with lower stroke (CHA₂DS₂VASc score 0 or 1) regardless if they received catheter ablation or medical therapy alone.

In contrast to the results of large administrative studies, data available from clinical trials do not support the concept that catheter ablation reduces the risk of stroke. CABANA (Catheter Ablation vs. Antiarrhythmic Drug Therapy for Atrial Fibrillation), the largest clinical trial of catheter ablation to-date, randomized 2204 patients with symptomatic AF to catheter ablation or medical therapy (rate or rhythm control). CABANA found that ablation was not superior to drug therapy with respect to all-cause mortality, disabling stroke, or serious bleeding according to the intention to treat analysis.^3,27 However, CABANA was not powered to assess for differences in stroke and the overall rate of disabling stroke over the trial follow-up was small at 0.5%. This very low rate of stroke was largely in part due to the fact that guideline-directed OAC was mandated. Similarly, a recent meta-analysis identified four clinical trials of catheter ablation that reported stroke as an outcome. The overall pooled stroke rate was less than 1% among the 2911 patients included in the meta-analysis. Comparing catheter ablation to medical therapy, there was no difference in stroke risk [relative risk (RR) 0.56, 95% 0.26–1.22; P = 0.14].¹

The recently published EAST-AFNET 4 (Early Treatment of Atrial Fibrillation for Stroke Prevention Trial) study enrolled 2789 patients with recent onset AF (<1 year), and tested the hypothesis that early, structured rhythm control therapy with antiarrhythmic drugs and/or catheter ablation could prevent AF-related complications (including stroke) in patients with AF when compared to usual care.^28,29 Patients who were randomized to early rhythm control had a lower risk of cardiovascular death, stroke, or hospitalization with worsening of heart failure or acute coronary syndrome (HR 0.79, 96% CI 0.66–0.94; P = 0.005) compared to usual care where patients were treated primarily with rate control agents (i.e. 84% rate control at 2 years). While the incidence of stroke was low in this study population (<1%), the early rhythm control strategy was surprisingly associated with a lower risk of ischaemic stroke (HR 0.65, 95% CI 0.44–0.97). Nevertheless, extrapolating the findings of EAST-AFNET4 to the AF population undergoing catheter ablation requires caution. That is, the trial allowed for both antiarrhythmic drug therapy and catheter ablation in the early rhythm control strategy where the overall rates of catheter ablation were low (19% at 2 years). Importantly, despite the significant value of the observations in CABANA nor EAST, neither trial was designed to compare OAC management strategies after catheter ablation. Thus, these studies will not answer all of the relevant clinical questions about stroke risk and its prevention after ablation.

Discontinuation of long-term oral anticoagulation following ‘successful’ ablation

Currently, the evidence evaluating the safety of discontinuing anticoagulation in AF patients following successful catheter ablation is limited to observational studies.^30–35 Among the largest of these observation studies, Karasoy et al.³⁴ reported the outcomes of 4050 AF patients in Denmark undergoing first time radiofrequency catheter ablation. Among the 1507 patients with increased stroke risk (CHA₂DS₂VASc score ≥ 2), OAC was discontinued in 30% of patients at 1 year. The overall rate of thromboembolism was low and comparable between patients with discontinued OAC (0.93 per 100 patient years) and continued OAC use (0.97 per 100 patient years). Yang et al.³⁵ reported the outcomes of 4512 consecutive AF patients post-ablation using data from the Chinese Atrial Fibrillation Registry, where 3149 discontinued OAC 3 months after ablation. The incidence rates of thromboembolism beyond 3 months post-ablation were overall low, and similar between the OAC and discontinued OAC groups regardless of stroke risk. For example, among high stroke risk patients (defined as CHA₂DS₂VASc ≥ 2 in men or ≥3 in women), the incidence rate of thromboembolism was 1.11 per 100 patient years with continued OAC use and 0.69 per 100 patient years with discontinued OAC (P = 0.11). Themistoclakis et al.³⁰ reported the outcomes of 3335 AF patient post-ablation enrolled from five high-volume AF centres. At the discretion of local institutional policy, OAC was discontinued 3–6 months post-ablation in 2692 patients, including 346 patients that had a CHADS₂ score ≥2. After a mean follow-up of 2 years, stroke risk was low among patients that discontinued OAC and the 663 patients that continued OAC (0.07% vs. 0.45%, respectively, P = 0.06).

In an attempt to combine the available evidence, a meta-analysis of 3436 patients examined the more controversial practice of discontinuing long-term anticoagulation post-ablation in the high-risk cohort of patients with CHADS₂ or CHA₂DS₂VASc ≥2 scores.⁷ The pooled analysis found no significant difference between the 1815 patients continued OAC and 1621 discontinued OAC after 3 months with regard to the risk of cerebrovascular events (RR 0.9, 95% CI 0.4–1.7, P = 0.64) or systemic thromboembolism (RR 1.2, 95% CI 0.7–2.2, P = 0.54). However, continued OAC use was associated with an increased risk of major bleeding (RR 6.5, 95% CI 2.5–16.7, P = 0.0001).

In contrast to these findings, in a study of 6886 patients from the Optum database, Noseworthy et al.³⁶ found that the risk of thromboembolic events beyond 3 months was increased when OAC was discontinued after AF ablation among high-risk patients with a CHA₂DS₂VASc score of 2 or higher (HR 2.48, 95% CI 1.11–5.52, P < 0.05) but not in lower-risk patients.

A key issue when examining the need or impact of OAC after ablation is whether those treated continue to have AF. Most studies focusing on this issue have limited ability to determine the degree of AF burden following ablation. Ghanbari et al. 2014⁶⁰ reported the outcomes of 3058 patients with paroxysmal or persistent AF undergoing catheter ablation. When sinus rhythm post-ablation was modelled as a time-dependent covariate, there was no significant reduction in stroke events among patients who remained in sinus rhythm (HR 0.79, 95% CI 0.48–1.29; P = 0.34).

These data need to be interpreted within the limitations of observational data. The observational design of prior studies is limited by the inherent inability to adjust for unknown confounders, short duration of follow-up, and unclear adjudication of stroke events. Finally, the variation in definition of ‘successful’ catheter ablation in terms of electrocardiogram monitoring post-ablation further complicates study interpretation and application to clinical practice. While prior meta-analyses suggest a favourable risk-benefit profile for discontinuation of OAC following successful AF ablation, these same studies report substantial publication bias. Indeed, the findings of Noseworthy et al.³⁶ give pause when considering discontinuation of OAC in the post-ablation period.

Several upcoming randomized clinical trials will help us determine whether successful AF ablation can obviate the need for long-term OAC (Table 2), the largest of which is the Optimal Anti-Coagulation for Enhanced-Risk Patients Post-Catheter Ablation for Atrial Fibrillation (OCEAN) trial (ClinicalTrial.gov, NCT02168829). OCEAN aims to enroll 1572 AF patients with CHA₂DS₂VASc score ≥1 and no atrial arrhythmia recurrences for at least 12 months following catheter ablation as determined by serial Holter monitoring to either OAC therapy with rivaroxaban 15 mg daily or aspirin 75–160 mg daily. The primary outcome is a composite of clinically overt stroke, systemic embolism, and covert stroke on brain magnetic resonance imaging with 3 years of follow-up.³⁷

Divergence between consensus recommendations and clinical practice

Given the paucity of high-quality evidence, the major cardiovascular society guidelines have generally favoured the approach that has demonstrated effective patients with AF without ablation, such as the use of validated risk scores (i.e. CHADS₂ or CHA₂DS₂-VASc scores). Specifically, current guidelines counsel that AF ablation and subsequent maintenance of sinus rhythm should not inform treatment decisions and that continued OAC following successful catheter ablation should be based on the patient stroke risk profile (Table 3). Anticoagulation post-ablation is arbitrary divided into two periods (Figure 3). In the immediate post-ablation period, consensus supports OAC for at least 2–3 months due to an increased thrombotic risk from post-ablation inflammation and delayed recovery of atrial function.⁴⁰ This recommendation is consistent with the available observational data. In a US cohort of AF patients who underwent catheter ablation followed for a median of 1.2 years (interquartile range 0.5–2.4 years), approximately a quarter of thromboembolic events occurred within the first 3 months post-ablation.³⁶ Additionally, there was an eight-fold risk of thromboembolism following premature discontinuation of OAC within the first 3 months of ablation compared to patients who continued OAC during the same time period.

Regarding continuing OAC in the long-term (i.e. beyond 2 months following catheter ablation), the European guidelines recommend that OAC therapy be continued indefinitely regardless of follow-up rhythm status in patients at high-risk of stroke.³⁸ Similarly, the 2017 Heart Rhythm Society/European Heart Rhythm Society/European Cardiac Arrhythmia Society/Asian Pacific Heart Rhythm Society/Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología expert consensus on AF catheter ablation recommend that decisions to discontinue OAC should be based on a patient’s stroke risk profile rather than the apparent success or failure of catheter ablation.⁴⁰

Despite these recommendations, surveys of Canadian and European electrophysiologists indicate variation in clinical practice, where 14–16% of physicians indicate that they would discontinue OAC even among high stroke risk patients (CHADS₂ ≥ 2 or CHA₂DS₂-VASc ≥ 2) within the year of successful catheter ablation and maintenance of sinus rhythm.^43,44 Data from registries and electronic health records suggest an even higher rate of OAC discontinuation in the real world (Figure 4). In a US cohort of 6886 patients who underwent catheter ablation, only 38% of high-risk patients (CHA₂DS₂-VASc ≥ 2) remained on OAC.³⁶ Data from the Outcomes Registry for Better Informed Treatment of Atrial Fibrillation (ORBIT-AF) registries also corroborate the substantial practice variation with regard to OAC use following AF ablation.⁴⁵ Among 982 AF patients from the ORBIT-AF I and II registries who underwent catheter ablation and were on OAC prior to ablation, 29% discontinued OAC post-procedure including 23% who were at moderate to high risk of stroke. Despite current guideline recommendations, approximately one in four patients at moderate to high risk of stroke discontinue OAC after ablation. The divergence between guideline recommendation and practice may reflect the current clinical equipoise, physician perceptions regarding rhythm control and stroke risk, and/or patient expectations regarding OAC use following ablation.

Alternate approaches to stroke prevention following AF ablation

Recognizing that certain AF patients at high risk of stroke may be highly motivated to discontinue OAC following ablation, several alternatives to continued OAC use post-catheter ablation are currently being studied. For example, left atrial appendage occlusion following ablation is being studied as an alternative to OAC following ablation. The Comparison of Anticoagulation With Left Atrial Appendage Closure After AF Ablation (OPTION) trial (ClinicalTrial.gov, NCT03795298) is a prospective, multi-centred, open-labelled, randomized control trial that will enroll 1600 participants to determine if left atrial appendage closure with the WATCHMAN FLX device is a reasonable alternative to OAC in patients after AF ablation. The OPTION trial has two co-primary outcomes: (i) an efficacy endpoint of composite of stroke, all-cause mortality, and systemic embolism at 36 months that will be tested for non-inferiority, and (ii) a safety endpoint of non-procedural bleeding that will be assessed for superiority.

Two pilot studies have explored a strategy of intermittent anticoagulation triggered by AF burden as detected on cardiac implantable electronic devices (i.e. ‘pill-in-the-pocket’ anticoagulation). The TACTIC AF pilot study enrolled 48 patients with non-permanent AF, CHADS₂ ≤ 3 and cardiac implantable electronic devices that were capable of continuous rhythm monitoring. Direct oral anticoagulant (DOAC) re-initiation was triggered by recurrence of AF defined as episodes lasting ≥6 min with a total AF burden >6 h/day. This strategy decreased OAC utilization by 75%. No thromboembolic events were observed during the follow-up of 12 months.⁴⁷ Similarly, the REACT.COM pilot study was a multi-centred, single armed study that enrolled 59 patients with non-permanent AF at moderate risk of stroke (i.e. CHADS₂ score of 1 or 2). DOACs were re-initiated for any AF episode ≥1 h. This feasibility study was able to reduce the exposure to DOAC by 94%, although the study was not powered to assess the safety of subsequent stroke risk. A randomized trial of pill-in-the-pocket DOAC use guided by implantable cardiac monitoring is currently being planned for a moderate risk non-permanent AF (REACT AF trial).⁴⁸

Areas of uncertainty

Although it may be tempting to causally link stroke risk and atrial rhythm, recent studies have questioned our understanding of the mechanistic link between AF and stroke risk. For example, in patients with continuous atrial rhythm monitoring enabled through cardiac implantable electronic devices, there does not appear to be a consistent temporal association between episodes of AF and stroke.^49–51 The evolving concept of ‘atrial myopathy’ ascribes AF as a secondary manifestation of the underlying processes that contribute stroke.^52–54 In this model of AF, the benefit of catheter ablation on stroke is questionable given the focus on maintenance of sinus rhythm without addressing the contributing conditions involved in the pathogenesis of progressive atrial myopathy. Nevertheless, to address the need for long-term OAC post-ablation, further research is required to better delineate the interaction between stroke, rhythm, and conditions related to atrial myopathy (such as obesity, or sleep apnoea).

Another area of uncertainty, further complicating post-ablation OAC treatment decisions, is our increasing attention towards the clinical significance of asymptomatic cerebral emboli associated with AF. While most studies have focused on clinical thromboembolic events as outcomes, there is increasing recognition of covert stroke detected by magnetic resonance imaging with an estimated prevalence of 10–50% in AF patients.^55–58 Covert stroke may be clinically important given the possible association between AF and increased risk of dementia.^58,59 It is unclear whether catheter ablation modifies the risk of covert stroke, as well as clinical stroke, and the potential link with cognitive impairment warrants further study. As previously described, several ongoing clinical studies, such as OCEAN and ODIn-AF, should provide further insights into the association between covert stroke (as detected by magnetic resonance imaging) and the discontinuation of OAC with maintenance of sinus rhythm post-ablation.

Conclusions

While catheter ablation offers a more durable rhythm control option compared to anti-arrhythmic therapy, the mechanistic links between sinus rhythm maintenance and stroke remain uncertain. Furthermore, the intensity and duration of rhythm monitoring following ablation to define ‘procedural success’ is unclear. Currently, continued long-term OAC guided by the stroke risk factor profile is the only proven strategy to prevent stroke. The evidence for the safety of OAC discontinuation following catheter ablation and durable maintenance of sinus rhythm is limited to observational data, although several upcoming large randomize clinical trials aspire to address this area of clinical equipoise.

Supplementary material

Supplementary material is available at Europace online.

Conflict of interest: D.C. is supported by a Canadian Institutes of Health Research Banting Fellowship and an Arthur JE Child Cardiology Fellowship. J.P.P. receives grants for clinical research from Abbott, American Heart Association, Association for the Advancement of Medical Instrumentation, Bayer, Boston Scientific, NHLBI, and Philips and serves as a consultant to Abbott, Allergan, ARCA Biopharma, Biotronik, Boston Scientific, LivaNova, Medtronic, Milestone, Sanofi, Philips, and Up-to-Date.

References

Notes

Biography: Dr. Derek Chew received his medical degree from the University of Toronto, Canada in 2011. He went on to complete his residency and fellowship training in Internal Medicine, Cardiology and Cardiac Electrophysiology at the University of Calgary, Canada. He has pursued additional research training and attained his Master of Science degree in Health Economics and Outcomes in Cardiovascular Sciences at the London School of Economics. Currently, he is completing a clinical research fellowship at Duke Clinical Research Institute funded by a Canadian Institutes of Health Research Banting Postdoctoral Fellowship. His research focuses on applied health economics, health technology assessment, processes of care delivery, and clinical arrhythmia research.

Biography: Jonathan P. Piccini, MD, MHS, FACC, FAHA, FHRS is a clinical cardiac electrophysiologist and Associate Professor of Medicine at Duke University Medical Center and the Duke Clinical Research Institute. He is the Director of the Cardiac Electrophysiology section at the Duke Heart Center. His focus is on the care of patients with atrial fibrillation and complex arrhythmias, with particular emphasis on catheter ablation, left atrial appendage occlusion, and lead extraction. His research interests include the conduct of clinical trials and the assessment of innovative cardiovascular therapeutics for the care of patients with heart rhythm disorders. Dr. Piccini has more than 425 publications in the field of heart rhythm medicine.