Catheter ablation to prevent recurrent atrial fibrillation: Anticoagulation

INTRODUCTION — Ischemic stroke and systemic embolization are major causes of death and disability in patients with atrial fibrillation (AF). This topic will focus on the role of anticoagulant therapy to prevent embolization in patients scheduled to undergo catheter ablation (CA). The role of anticoagulant therapy in the broad population of patients with AF is discussed separately. (See "Atrial fibrillation in adults: Use of oral anticoagulants".)

Other aspects of CA are discussed elsewhere. (See "Atrial fibrillation: Catheter ablation" and "Maintenance of sinus rhythm in atrial fibrillation: Catheter ablation versus antiarrhythmic drug therapy" and "Overview of catheter ablation of cardiac arrhythmias" and "Patient education: Catheter ablation for the heart (The Basics)" and "Catheter ablation for the treatment of atrial fibrillation: Technical considerations for non-electrophysiologists".)

OUR APPROACH TO ANTICOAGULATION — There are three periods when a decision or decisions have to be made about anticoagulation in a patient scheduled for catheter ablation (CA).

●Preprocedural – We effectively anticoagulate most patients, irrespective of CHA₂DS₂-VASC score (table 1) or presence or absence of sinus rhythm, with either a vitamin K antagonist (VKA) or a direct oral anticoagulant (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]) for at least three weeks prior to CA. It is reasonable to not use preprocedural anticoagulation in AF patients in sinus rhythm (and who are likely to remain in sinus rhythm for three weeks prior to the procedure) with a CHA₂DS₂-VASC score of 0. (See 'Preprocedural issues' below.)

●Periprocedural – We continue VKA in the periprocedural period. For most patients taking once-a-day DOACs, we either continue them uninterrupted or hold the dose the day before and the morning of the procedure. For twice-a-day DOACs, some of our experts hold both doses the day before the procedure while others hold only the evening dose before the procedure; no drug is given the morning of the procedure in these cases. Uninterrupted DOAC is commonly used and has been shown to have similar rates of thromboembolic and bleeding events as minimally interrupted DOAC use [1]. Studies support the fact that uninterrupted DOACs may be superior to uninterrupted warfarin for patients who require continued anticoagulation due to high risk of thromboembolism [2]. All patients receive a continuous infusion of unfractionated heparin (UFH); the activated clotting time is maintained at greater than 300 seconds during the procedure. (See 'Periprocedural issues' below.)

●Postprocedural – UFH is stopped at the end of the procedure and the sheaths are pulled when the activated clotting time is <180 to 200 seconds.

For patients previously taking a VKA, the next dose is given approximately 24 hours after the prior dose. For those patients in whom the international normalized ratio was <2.0 prior to the procedure, we restart UFH without a bolus six hours after sheath pull, increase the oral warfarin the night of the procedure, and we continue UFH until the INR is 2.0. Another reasonable approach is to stop UFH the morning after the procedure and start low molecular weight heparin, usually at half the normal dose (0.5 mg/kg twice daily) to avoid bleeding. Substituting DOAC for VKA can also be considered at this time. (See 'Postprocedural anticoagulation' below.)

For patients previously taking a DOAC, we suggest restarting it six hours after sheath removal (and in the absence of any related bleeding). Some experts give intravenous heparin (no bolus; drip at 1000 to 1200 units per hour) or low molecular weight heparin (enoxaparin 0.5 mg/kg) starting six hours after sheath pull, that is uncomplicated by bleeding, and continue until the morning after the ablation. Most experts no longer give a heparin after sheath pull.

●Long-term – We continue oral therapy with the previously prescribed oral anticoagulant for two to three months regardless of CHA₂DS₂-VASc score. After this period, the decision to continue on long-term anticoagulation is based on the patients underlying stroke risk regardless of whether rhythm control has been achieved. For those patients whose risk for embolization is very low, such as those with a CHA₂DS₂-VASc score of 0 (table 1), we stop anticoagulation at the two-to-three-month visit.

INCIDENCE, TIMING, AND MECHANISM OF EMBOLISM — The risk of stroke, transient ischemic attack, or other manifestation of embolization is increased at the time of catheter ablation (CA) and is in the range of 0.4 to 2.0 percent [3-5]. These rates come from studies of patients who are receiving anticoagulant therapy and would be higher off such treatment. Most strokes occur within 24 to 48 hours after the procedure [5]. However, embolic events thought attributable to the procedure have been reported to occur for up to one week [6].

The following are potential causes of periprocedural embolization:

●Withdrawal of anticoagulation before the procedure

●Catheter manipulation within the left atrium, which may dislodge preexisting thrombus

●Catheter trauma to the left atrial endothelium, which increases the risk of thrombus formation

●Thrombus formation on the ablation catheters or left atrial guide sheaths

●Conversion to sinus rhythm during the procedure in some patients (see "Prevention of embolization prior to and after restoration of sinus rhythm in atrial fibrillation")

Asymptomatic embolism — Not all emboli to the brain are symptomatic. Multiple magnetic resonance imaging (MRI) studies performed within 24 hours after CA have demonstrated new cerebral lesions in 7 to 44 percent of asymptomatic patients [7-12]. These lesions are presumed secondary to microemboli [13].

Studies of the subsequent impact of these lesions on neurocognitive function have come to somewhat differing conclusions as to the significance of these lesions:

●The prevalence of cognitive impairment after radiofrequency CA (RFA) was evaluated in a study of 150 patients: 60 undergoing ablation for paroxysmal atrial fibrillation (AF), 30 for persistent AF, 30 for supraventricular tachycardia, and 30 matched AF patients awaiting RFA (the control group) [14]. All CA patients received periprocedural enoxaparin and most patients with AF had a CHADS₂ score of 0 or 1 (table 1). All patients underwent eight neuropsychological tests at baseline and at 2 and 90 days after RFA. The prevalence of neurocognitive dysfunction at day 90 was 13, 20, 3, and 0 percent, respectively, in these four groups of patients.

●In a study of 37 patients with paroxysmal AF who underwent 41 CA procedures, MRI performed within 48 hours showed new brain lesions in 41 percent of patients and 44 percent of procedures [10]. Follow-up MRI at six months found glial scar in about 12 percent of those with lesions. However, there was no decline of neurocognitive function on testing performed after six months.

PREPROCEDURAL ISSUES — All patients not at low risk of stroke should be treated with long-term anticoagulant therapy using one of the novel oral anticoagulants or warfarin. Thus, many patients will be receiving anticoagulation when scheduled for catheter ablation (CA) and should continue their anticoagulant. (See "Atrial fibrillation in adults: Use of oral anticoagulants", section on 'Summary and recommendations'.)

For patients not taking long-term anticoagulant therapy due to a low risk of stroke, there are no studies that have compared differing anticoagulant strategies prior to CA; thus, the optimal preprocedural anticoagulation strategy is not known. For these patients, including those in sinus rhythm, most of our experts carry out a minimum of three weeks of effective oral anticoagulation prior to the procedure. The rationale for doing so is that many episodes of atrial fibrillation (AF) are asymptomatic and these episodes will have placed the patient at risk of embolization at the time of catheter manipulation.

We also believe it is reasonable to not use preprocedural anticoagulation in AF patients in sinus rhythm (and who are likely to remain in sinus rhythm for three weeks prior to the procedure) with a CHA₂DS₂-VASc score of 0.

When three weeks of effective anticoagulant therapy has not been carried out, preprocedural transesophageal echocardiography (TEE) should be performed (see 'Role of transesophageal echocardiography' below); patients with evidence of left atrial thrombus are not candidates for CA unless it resolves with anticoagulation.

Choice of anticoagulant — For patients started on oral anticoagulant therapy prior to catheter ablation, direct oral anticoagulants (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]) are preferred over warfarin. This preference is based on our preference for these agents in the general AF population [15]. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation".)

Most but not all observational studies comparing one NOAC to warfarin have found similar efficacy [16] and safety [17-24]. At least three meta-analyses of observational studies comparing warfarin to dabigatran have come to similar conclusions [25-27]. A 2016 meta-analysis of 25 studies (11,686 patients) comparing DOACs with uninterrupted VKAs found no significant difference in the rate of stroke or transient ischemic attacks (odds ratio 1.35, 95% CI 0.62-2.94) and major bleeding (odds ratio 0.80, 95% CI 0.65-1.00) [28].

Switching oral anticoagulant — As stated directly above, we prefer one of the DOAC group to warfarin for patients undergoing catheter ablation. For patients receiving long-term warfarin therapy, there is no evidence that switching to DOAC prior to catheter ablation improves outcomes. Thus, we do not routinely switch from warfarin to an DOAC.

We do not have a preference for one DOAC over another and thus we do not switch DOAC.

Role of transesophageal echocardiography — Most experts will image the left atrial appendage prior to ablation to ensure the absence of thrombus even in those with effective preprocedural oral anticoagulation, as the presence of left atrial thrombus is a contraindication to the procedure [29,30]. TEE is often performed prior to (generally the day before) CA. Two reasons to perform TEE prior to (generally the day before) CA are that it adds significant length to the CA, and some complications of TEE, such as a retropharyngeal hematoma, can be aggravated by the unfractionated heparin used during the procedure. (See "Echocardiography in detection of cardiac and aortic sources of systemic embolism", section on 'LA/LAA thrombi'.)

We acknowledge that some experts will omit a TEE in the lowest-risk patients who have been taking effective anticoagulant therapy for at least three weeks, such as those in sinus rhythm who have no significant structural heart disease or those with a CHA₂DS₂-VASc score of 0 (table 1) [30]. These experts often prefer that a pre-ablation magnetic resonance image confirms the absence of left atrial appendage thrombus in these patients who do not undergo preprocedural TEE. In a study of 97 patients undergoing pulmonary vein isolation, contrast-enhanced MRI demonstrated 100 percent concordance with TEE for the presence and absence of left atrial appendage thrombus [31,32].

An attempt to determine the need for preprocedural TEE in patients at low risk for embolization was made in an analysis of 1058 patients who had TEE performed within 24 hours of pulmonary vein isolation [33]. The frequency of left atrial thrombus or sludge was evaluated according to the CHADS₂ score. (See "Echocardiography in detection of cardiac and aortic sources of systemic embolism".) A CHADS₂ score of 0 was present in 47 percent of patients. Left atrial or left atrial appendage thrombus or sludge was found in 0.6 and 1.5 percent of all patients and the frequency increased with ascending CHADS₂ scores (percents in parentheses): 0 (0), 1 (2), 2 (5), 3 (9), 4 to 6 (11).

We do not use intracardiac echocardiography or computed tomography as a substitute for TEE. Each of these has been shown to be inferior in this setting [34,35].

PERIPROCEDURAL ISSUES — The two principal periprocedural anticoagulant issues are how to manage the oral anticoagulant and whether/how to use parenteral anticoagulant.

Management of oral anticoagulants — For patients taking long-term oral anticoagulant who present for catheter ablation, the approach depends on which anticoagulant the patient has been taking.

Patients taking long-term vitamin K antagonist — For patients taking a VKA prior to catheter ablation, we prefer the strategy of uninterrupted VKA to a strategy of a heparin bridge. We do not hold doses of VKA unless the international normalized ratio (INR) is >3.0. (See "Perioperative management of patients receiving anticoagulants", section on 'Bridging anticoagulation'.)

One randomized trial [36] and most observational studies [5,37,38] have shown that continuous anticoagulation with warfarin, compared with warfarin discontinuation with a heparin bridge, is associated with a lower rate of embolization and an equivalent or lower bleeding rate [39]. In the COMPARE trial, 1584 patients with paroxysmal or persistent atrial fibrillation (AF) (see "Atrial fibrillation: Overview and management of new-onset atrial fibrillation", section on 'Classification and terminology') and CHADS₂ score ≥1 were randomly assigned to warfarin discontinuation two to three days before ablation and bridging with low molecular weight heparin (1 mg/kg enoxaparin twice daily until the evening before the procedure) or continuation of therapeutic warfarin (three to four weeks with an INR 2.0 to 3.0) [36]. The primary end point of the incidence of thromboembolic events (stroke, transient ischemic attack, or systemic thromboembolism) in the 48 hours after ablation occurred more frequently with warfarin discontinuation (4.9 versus 0.25 percent; odds ratio 13, 95% CI 3.1-55.6). The incidence of major bleeding complications was similar in the two groups (0.76 versus 0.38 percent, respectively). The majority of events occurred in patients with persistent AF. One limitation of the trial is that operators were not blinded to the anticoagulation strategy.

For patients in whom a strategy of continuous anticoagulation with warfarin has been chosen, the optimal immediate-preprocedural range for the INR is not known. In a retrospective study of 1113 patients undergoing radiofrequency catheter ablation for AF, bleeding and vascular complications were less prevalent when the INR was ≥2.0 and ≤3.0 (5 percent), compared with ≤2.0 (10 percent) or ≥3.0 (12 percent) [40]. The optimal INR range was calculated to be 2.1 to 2.5.

Patients taking DOACs — In most studies that have evaluated periprocedural outcomes in patients taking direct oral anticoagulants (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]), the DOAC was held prior to the procedure. In one trial, 326 patients undergoing AF ablation were randomized to uninterrupted DOAC, procedure day single-dose skipped DOAC, or 24-hour skipped DOAC [41]. The intraprocedural heparin dose was higher in the 24-hour skipped group, but the incidence of major bleeding and postprocedural hemoglobin levels were not significantly different among the treatment groups and different DOACs. There were no fatal events or thromboembolic complications.

For patients taking once-a-day DOACs in the morning, some operators will hold the dose the day before and also the day of the procedure. For those that take once-a-day DOACs with the evening meal or later, they may hold only a single dose the day before the procedure. For twice-a-day DOACs, some of our experts hold both doses the day before the procedure while others hold only the evening dose before the procedure and the morning of the procedure.

For many patients, especially those at particularly high risk of a periprocedural stroke, including those with a high CHA₂DS₂-VASc score in whom intraprocedural cardioversion is planned, no interruption of (continuous) oral anticoagulation with a DOAC is the preferred alternative to interruption. (See 'Choice of anticoagulant' above.)

Small randomized trials of dabigatran, rivaroxaban, edoxaban, and apixaban suggest that outcomes with uninterrupted DOACs are similar to those with uninterrupted warfarin [42-45].

Use of intravenous heparin — As catheter ablation (CA) is associated with an increase (from baseline) in the risk of a periprocedural thromboembolic event, we recommend that all patients receive intraprocedural heparin. (See 'Incidence, timing, and mechanism of embolism' above.)

We start with a loading dose of 100 units/kg at the beginning of the procedure. Others start the loading dose before transeptal puncture, while others give half the dose before and half the dose after transeptal puncture [46]. A continuous infusion is used to maintain the activated clotting time greater than 300 seconds; the first activated clotting time is performed 10 to 15 minutes after the loading dose. Heparin is stopped at the end of the procedure and the sheaths are pulled when the activated clotting time is <180 to 200 seconds. Protamine can be given after the procedure before removing vascular sheaths.

Intracardiac echocardiography is done at the end of the CA procedure to ensure that there is no pericardial effusion. If a pericardial effusion is found we take the following approach:

●For small effusions, we observe and continue with the anticoagulation protocol.

●If moderate, we reverse anticoagulation and observe.

●If large, we reverse anticoagulation and perform pericardiocentesis. (See "Pericardial effusion: Approach to management".)

●For patients with hemodynamic compromise, regardless of the size of the effusion, we perform pericardiocentesis and reverse anticoagulation.

POSTPROCEDURAL ANTICOAGULATION — The approach to anticoagulation within the first 24 hours after a successful procedure is determined in large part by preprocedural anticoagulant approach (see 'Preprocedural issues' above). There have been no studies comparing one approach to another. In the absence of any related bleeding, we suggest the following approach:

●For those previously taking warfarin, and for whom the management of the international normalized ratio was not problematic, we suggest continuing warfarin. The first postprocedural dose should be the day after the procedure, assuming a dose was given the morning of the procedure. (See 'Patients taking long-term vitamin K antagonist' above.)

●For patients taking warfarin whose procedure was done with a subtherapeutic international normalized ratio, we restart intravenous heparin without a bolus six hours after sheath pull and start low molecular weight heparin the morning after the procedure. Low molecular weight heparin is continued until the international normalized ratio is therapeutic. When low molecular weight heparin is used, some of our experts give the first dose at 50 percent.

●For patients previously taking a direct oral anticoagulant (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]), DOAC may be restarted four to six hours after sheath pull.

Alternatively, some experts delay restarting these newer agents until the morning after the procedure. If the DOAC is restarted the morning after the procedure, we give either intravenous unfractionated heparin (no bolus; drip at 1000 to 1200 units per hour starting six hours after sheath pull and continued until the morning after the procedure) or low molecular weight heparin (enoxaparin 0.5 mg/kg; typically a single dose administered six hours after sheath pull).

●For patients previously not taking any oral agent, we suggest starting either a DOAC or warfarin. A first dose of either agent can be given six hours after an uncomplicated procedure. Patients started on warfarin will need to receive bridging treatment for a few days with low molecular weight heparin, as described directly above.

LONG-TERM ANTICOAGULATION — We continue oral therapy with the previously prescribed oral anticoagulant [47] for at least two months to ensure that the increased risk of embolization associated with the procedure, which lasts for about four weeks, has returned to a baseline risk and that there has been adequate time to document an absence of recurrence of atrial fibrillation (AF) for those patients in whom practitioners and patients are contemplating discontinuing anticoagulation [48]. (See 'Incidence, timing, and mechanism of embolism' above.)

After this two-month period of mandatory oral anticoagulation, we generally restore the anticoagulant regimen in place prior to the procedure. For those patients without evidence of recurrent AF and whose risk for embolization is very low, such as those with a CHA₂DS₂-VASc score of 0 (table 1), we stop anticoagulation. (See "Atrial fibrillation in adults: Selection of candidates for anticoagulation", section on 'CHA2DS2-VASc score'.)

Some experts are comfortable stopping anticoagulation in patients with a CHA₂DS₂-VASc score of 1 (table 1) after sufficient documentation of the absence of recurrent episodes of AF. We believe this approach has not been adequately tested. Therefore, we tell patients and referring physicians that the desire to stop long-term anticoagulation is not an indication for catheter ablation (CA) by itself.

For all patients with a CHA₂DS₂-VASc score of >1 (table 1) after CA, irrespective of whether or not recurrence has been documented, we maintain the patient on long-term oral anticoagulation [49,50].

The optimal approach to chronic anticoagulation after successful CA, defined as no evidence of recurrence, is uncertain [51]. It is known that late recurrent AF occurs in 20 to 30 percent of patients, but the methods used in some studies likely underestimate the incidence [52-55]. (See "Atrial fibrillation: Catheter ablation", section on 'Efficacy'.)

In a 2019 meta-analysis of five studies with nearly 4000 patients that evaluated safety and efficacy of long-term oral anticoagulation (OAC) compared with no OAC, the following was found during a mean follow-up of nearly 40 months [56]:

●In patients with a CHA₂DS₂-VASc score ≥2, OAC continuation was associated with a decrease in the risk of thromboembolic events (risk ratio [RR] 0.41, 95% CI 0.21-0.82) but an increased risk of intracranial hemorrhage (ICH; RR 5.78, 95% CI 1.33-25.08). The absolute risk decrease in thromboembolic risk was 1.14 percent, while the increase in ICH was 0.7 percent.

●In patients with a CHA₂DS₂-VASc score of 0 or 1, the risk of ICH from OAC exceeded any potential decrease in thromboembolic risk.

The issue of the role of long-term anticoagulation was indirectly addressed by at least three studies that found a lower incidence of stroke comparing successful CA to antiarrhythmic drug therapy [57-59]. In a retrospective study of 174 matched pairs of AF individuals with a CHA2DS2-VASc score ≥1 who were treated with either antiarrhythmic drug therapy or CA and treated for at least three months with warfarin, the rate of stroke/transient ischemic attack was lower with the CA group (0.59 versus 2.21 percent per year) [57]. In those individuals treated with CA and in whom there was no AF recurrence, the stroke rate was very low compared to those with recurrence (0.8 versus 5.4 percent) after a mean follow-up period of 47 months.

In addition, it is not known if asymptomatic recurrences result in a persistent thromboembolic risk in patients who have undergone CA. A low risk of stroke was reported in a series of 755 patients with longstanding persistent AF who underwent CA or a tailored approach [60]. In this cohort, anticoagulation was discontinued three months after the procedure in the majority of the 522 patients who did not have evidence of recurrent AF. During a median follow-up of 25 months, none of the patients who stopped anticoagulation had a stroke. Although these results are encouraging, the study cohort had a low baseline thromboembolic risk, with most having a CHADS2-VASc score of 0 to 2 (table 1). Some of these patients, such as those with lone AF, would not require chronic oral anticoagulation whether or not they had a successful CA procedure.

The following recommendations were made in the 2019 update of the 2014 American Heart Association/American College of Cardiology/Heart Rhythm Society AF guideline [61,62]:

●Systemic anticoagulation was recommended for at least two months in all patients following CA.

●After two months, the decision to continue anticoagulation should be based on the patient’s risk factors for stroke and risk of bleeding, and not on the type of AF. The guideline acknowledges that recurrent episodes of AF, which may be asymptomatic, occur.

The 2019 focused update from the American Heart Association/American College of Cardiology/Heart Rhythm Society AF guideline recommends a long-term oral anticoagulation CHA₂DS₂-VASc score of ≥2 for men and ≥3 for women [15]. The position document on antithrombotic management in patients undergoing electrophysiological procedures states "…the decision for oral anticoagulation depends on the patient’s stroke risk profile and not the perceived success or failure of ablation…" [46].

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Atrial fibrillation" and "Society guideline links: Arrhythmias in adults" and "Society guideline links: Catheter ablation of atrial fibrillation".)

SUMMARY AND RECOMMENDATIONS

●Incidence, timing, and mechanism – The risk of stroke, transient ischemic attack, or other significant manifestation of embolization is increased, compared to baseline risk, at the time of catheter ablation (CA) for atrial fibrillation. (See 'Incidence, timing, and mechanism of embolism' above.)

●Role of transesophageal echocardiography – We perform a preprocedural transesophageal echocardiogram or cardiac magnetic resonance imaging (MRI) in most patients undergoing CA. The finding of intracardiac thrombus is a contraindication to the procedure. (See 'Role of transesophageal echocardiography' above.)

●Preprocedural issues – Most patients scheduled to undergo CA who have been receiving long-term oral anticoagulation should continue to do so until the procedure. For those low-risk patients who have not been receiving long-term anticoagulation, including those in sinus rhythm at the time of the procedure, we suggest at least three weeks of effective anticoagulation prior to CA rather than no preprocedural anticoagulation (Grade 2C). An alternate approach is presented above. (See 'Preprocedural issues' above.)

●Periprocedural issues – All patients should receive intraprocedural anticoagulation with intravenous heparin, irrespective of the patient’s baseline thromboembolic risk and whether or not the procedure is performed on uninterrupted warfarin. (See 'Periprocedural issues' above.)

For patients taking long-term oral anticoagulant with warfarin who present for CA, we recommend continuing such therapy rather than stopping and using a heparin bridge (Grade 1A).

For patients taking long-term oral anticoagulant with a newer oral anticoagulant who present for CA, we suggest discontinuing such therapy before the procedure rather than continuing it (Grade 2B). (See 'Periprocedural issues' above.)

●Patients taking DOACs – For the uncommon patient who is at very high risk of a periprocedural stroke, it is reasonable to not discontinue direct oral anticoagulants (DOAC; also referred to as non-vitamin K oral anticoagulants [NOAC]). (See 'Patients taking DOACs' above.)

DOAC may be restarted four to six hours after sheath pull. Alternatively, DOAC may be restarted the morning after the procedure in patients who are treated overnight with either intravenous unfractionated heparin (no bolus; drip at 1000 to 1200 units per hour starting six hours after sheath pull and continued until the morning after the procedure) or low molecular weight heparin (enoxaparin 0.5 mg/kg; typically a single dose at six hours after sheath pull). (See 'Patients taking DOACs' above.)

●Long-term anticoagulation – Oral anticoagulation is continued for at least two months after the procedure in all patients. After two months, the decision to continue anticoagulation should be based on the patient’s risk factors for stroke and risk of bleeding and not on the type of AF or outcome of the procedure. (See 'Long-term anticoagulation' above.)