atrial fibrillation: stopping anticoagulation 1 yr after ablation

 

A recent South Korean study found that some patients who had catheter ablation for atrial fibrillation were able to discontinue anticoagulation therapy, in the ALONE-AF trial (Anticoagulation One year after Ablation of Atrial Fibrillation in Patients with Atrial Fibrillation (see afib ablation stop anticoag after 1 yr JAMA2025 in dropbox, or doi:10.1001/jama.2025.14679)

 

Details:

-- 850 adults aged 19 to 80 were enrolled in this open-label randomized controlled trial from 2020 the 2023 at 18 hospitals in South Korea, with follow-up until 2025

    -- All enrollees had atrial fibrillation (AF) with catheter ablation and were put on anticoagulation; they were in intermediate or high risk of thromboembolism with a CHA₂DS₂-VASc score of at least 1 for men or 2 for women

-- All had at least one non-sex-related stroke risk factor as determined by their CHA₂DS₂-VASc score and no documented recurrence of atrial arrhythmia for at least one year after catheter ablation for AF

    -- the CHA₂DS₂-VASc score assesses stroke risk in individuals with atrial fibrillation (AF), based on the presence of congestive heart failure, hypertension, being at least 75 years of age/or between 65 and 74 years of age, diabetes, history of stroke or transient ischemic attack, vascular disease, and sex category (though the latter was not included in this study)

-- there was no upper time limit on the period from ablation to randomization

-- atrial arrhythmia recurrence was defined as any documented episode lasting 30 seconds or longer for AF, atrial flutter, or atrial tachycardia that was assessed after the ablation using 2 sessions of 24- to 72- hour Holter monitoring and ECG monitoring, with at least one session of Holter monitoring and ECG monitoring performed within 2 months prior to enrollment

    -- All patients had routine ECG monitoring at 6-monthly follow-up visit and 24- to 72-hour Holter monitoring at least every 6 months when asymptomatic, more so if potentially symptomatic of arrhythmia recurrence. Any evidence of recurrence led to reinitiating anticoagulant therapy in the placebo group

-- patients were randomized to discontinue oral anticoagulation therapy or continue with a direct oral anticoagulant, with either 5 mg of apixaban twice a day or 20 mg of rivaroxaban once a day (with dosage changes as appropriate for age, weight, renal function). Antiplatelet therapy was discouraged but allowed as clinically indicated, such as for those who had percutaneous coronary interventions or acute coronary syndromes (anti-platelet therapy was used in 8.6% in the no-anticoagulant group versus 5.0% in the anticoagulant group)

    -- In those on the oral anticoagulants, 78% were on apixaban 5 mg, 8.7% were on 15 mg of rivaroxaban, 7.3% on rivaroxaban 20 mg (reduced doses were prescribed overall for 14.4%)

-- mean age 64, 25% women, CHA₂DS₂-VASc score 2.1 (29% had score <2, 40% score of 2, 20% score of 3, 11%  4 or more), HAS-BLED score 2 (HAS-BLED assesses bleeding risk in patients with AF on anticoagulants, based on hypertension, kidney or liver disease, stroke history, prior bleeding, unstable INR, >65 years old, and drug or alcohol use, with a score of 0-9, and the higher the score, the greater risk of bleeding)

-- 67.6% had paroxysmal AF, 33.8% had persistent AF

-- hypertension in 69%, dyslipidemia 28%, diabetes 18%, heart failure 15%, stroke or transient ischemic attack 6%, chronic kidney disease 2%, myocardial infarction 1%, peripheral arterial disease 1%

-- current drinking 28%, smoking 10%, blood pressure 128/76, BMI 25

-- left atrial diameter 40 mm, left ventricular ejection fraction 61%, mitral inflow: mitral annulus tissue velocity ratio 9, creatinine clearance 86

 

-- primary outcome: first occurrence of a composite of stroke, systemic embolism, and major bleeding at 2 years

-- secondary outcomes: individual components of this primary outcome

-- mean duration between catheter ablation and randomization was 3.6 years

  

Results:

-- adherence to therapy at 2 years follow-up; 86.8% in the no-anticoagulant group and 90.5% in the anticoagulant group

-- recurrent atrial arrhythmias: 40 patients (9.6%) in the no-anticoagulant group versus 37 patients (8.7%) in the oral anticoagulant group

 

-- followup and monitoring done on the patients:

   -- 6-month completion rate of followup 94%, 30% for Holter monitoring

   -- 12-month completion rate of followup 92%, 30% for Holter monitoring

   -- 18-month completion rate of followup 90%, 30% for Holter monitoring

   -- 24-month completion rate of followup 88%, 30% for Holter monitoring

-- recurrence rate of atrial arrhythmias:

    -- median time to recurrence after randomization: 11 months

    -- recurrence by method used:

        -- regular Holter monitoring: 27.5% in those not on oral anticoagulants, 13.5% in those on oral anticoagulants

        -- regular 12-lead ECGs: 65% in those not on oral anticoagulants, 83.8% in those on oral anticoagulants

        -- intermittent devices (smartwatches or event recorders): 7.5% in those not on oral anticoagulants, 2.7% in those on oral anticoagulants

 

-- Primary outcome by 2 years (cumulative incidents):

    -- no oral anticoagulant group: 1 of 417 patients (0.3%)

    -- oral anticoagulant group: 8 of 423 patients (2.2%)

        -- absolute difference: -1.9 percentage points (-3.5 to -0.3), p=0.02

        -- number-needed-to-treat to prevent one primary outcome at 2 years: 53 patients discontinuing versus continuing anticoagulant therapy

 

--Ischemic stroke or systemic embolism by 2 years:

    -- no oral anticoagulant group: 0.3%

    -- oral anticoagulant group: 0.8%

        -- absolute difference -0.5 percentage points (-1.6 to 0.6)

--types of stroke: ischemic in 1 person on no anticoagulants, 3 on anticoagulants; hemorrhagic in 2 and only if on anticoagulant; severity of stroke was nondisabling in 1 person on no anticoagulant and 4 people on anticoagulant; disabling in 1 on anticoagulant

 

-- transient ischemic attack by 2 years:

    -- no oral anticoagulant group: 2 patients (0.6%)

    -- oral anticoagulant group: 0 patients

        -- absolute difference 0.6 percentage points (-0.2 to 1.3)

 

-- Major bleeding by 2 years:

    -- no oral anticoagulant group: 0 patients

    -- oral anticoagulant group: 5 patients (1.4%) with 2 intracranial, 2 GI, 1 genitourinary; 1 needed transfusion and 1 had life-threatening bleeding, both in anticoagulation group

        -- absolute difference -1.4 percentage points (-2.6 to-0.2)

 

-- Clinically relevant bleeding by 2 years:

    -- no oral anticoagulant group: 5 patients (1.4%)

    -- oral anticoagulant group: 7 patients (1.9%), 1 had hemorrhagic stroke

        -- absolute difference -1.4 percentage points (-2.6 to -0.2)

    -- no cases of MI or all-cause mortality

 

-- all of the above differences in the graphs (which i could not copy into this blog) confirmed the benefits of no oral anticoagulation:

    -- for the cumulative incidence of the composite of stroke, systemic embolism and major bleeding (the primary outcome): benefit begins at 2 months and increases over time

    -- for the cumulative incidence of ischemic stroke or systemic embolism: benefit begins at 6 months and increases over time

    -- for the cumulative incidence of major bleeding: benefit begins at 18 months and increases over time

-- sensitivity analyses:

-- no significant difference in primary outcome or major ischemic and bleeding events in the per-protocol vs with the intention-to-treat analysis

-- the treatment effect of oral anticoagulant discontinuation in the primary outcome was not different across all subgroups

    -- in the supplement there is a Forrest plot of subgroups, finding very small, overlapping but statistically significant differences favoring more benefit in the primary endpoint when off  anticoagulants in those >65yo, male sex, not having heart failure, having hypertension, not having diabetes, not having vascular disease, having higher HAS-BLED scores of at least 2, but no difference by  CHA₂DS₂-VASc scores. 

-- no difference in results by adjusting for age and sex

  

 Commentary:

-- AF is most common heart arrhythmia, has profound effects on the lives of those developing AF, and poses a large healthcare burden internationally

-- several observational studies have found that catheter ablation for AF is better than medical therapy for maintaining sinus rhythm and improving quality of life, though current guidelines do recommend continuing anticoagulation after successful ablation in those felt to have significant risk of thromboembolism.

    -- however, the randomized controlled CABANA study (see afib CABANA study ablation vs meds no diff JAMA2019 in dropbox, or doi:10.1001/jama.2019.0693) did not find any difference between ablation and medical therapy for AF, though 27.5% of those assigned to meds in this study did subsequently get ablation, thereby muddying their results a bit

-- a retrospective Japanese observational study with 231,374 patients with mostly low CHA₂DS₂-VASc scores (92% had scores of 2 or less) found that the risks of bleeding from continuing anticoagulants outweighed the 14% and almost significant (p=0.06) increase in thromoboembolic events, but only in those with CHA₂DS₂-VASc scores <3):  https://pubmed.ncbi.nlm.nih.gov/38117227/

-- another older retrospective data-mining study done in the US of 6886 patients found that stopping anticoagulants (initially vitamin K antagonists in 2005 and changing to mostly non-vit K antagonists by 2014) had increased cardioembolic events in the first 3 months after ablation in those with CHA₂DS₂-VASc scores >1: https://pubmed.ncbi.nlm.nih.gov/26541393/

-- in this setting of conflicting observational study results done with very different methodologies and reaching different conclusions, it is especially important to have a true randomized controlled trial, as per this current study

 

-- this study found the perhaps unexpected finding of a marginally decreased risk of stroke and systemic embolism after discontinuation of oral anticoagulants (though nonstatistically significant) when patients who had AF catheter ablation and at least 2 tests in the following year for atrial arrhythmia (Holter monitor and ECG), along with the expected finding of decreased bleeding (statistically significant).

-- those with even high-risk CHA₂DS₂-VASc scores of at least 4 (though only 11% of those in the study) did not have an increased stroke risk by discontinuing anticoagulants, though with a strict protocol of assessing 1- to 3-day Holter monitors/ECGs every 6 months to confirm the lack of atrial arrhythmias

-- the bleeding risk in this study may have been overstated, since 8.6% of patients in the no anticoagulant group were on anti-platelet drugs, and these by themselves confer increased risk of bleeding

-- it was encouraging that the intermittent testing for postablation AF burden and symptoms was sufficient in this study, given that the real standard for security in this regard would be continuous invasive monitoring (though this is expensive and invasive)

 

-- it is a bit curious that there were more primary outcomes in those continuing the anticoagulation, though that was primarily driven by bleeding events with a non-statistically significant trend for embolic/cerebrovascular events. One possible issue is the use of rivaroxaban in 16% of the patients. There have been a couple of studies finding that rivaroxaban is associated with an increase in both thrombotic and bleeding events (and probably should really be avoided as an anticoagulant….). eg see https://gmodestmedblogs.blogspot.com/2022/01/atrial-fib-apixaban-outperforms.html for detail showing the significant benefit of apixaban over rivaroxaban and the increased bad outcomes of both bleeding and thrombosis with rivaroxaban (and reviews how rivaroxaban got FDA approval through drug company shenanigans)

    -- unfortunately, this current South Korean study and its supplementary material did not include a subgroup assessment of the anticoagulant used and the associated study endpoints

-- there was also no information on different outcome stratified by the platelet inhibitors used (51 on aspirin, 21 on clopidogrel) with 20 on aspirin and 1 on clopidogrel in the oral anticoagulation group. It would be useful to know if there were different outcomes dependent on the antiplatelet agent used

  

 Limitations:

-- this study was in one country (South Korea), though in several sites there, but its results may not be generalizable to other areas of the world, given possible differences in major demographics, culture (different diet, perhaps exercise (though this was not measured), social relationships, stressors...)), comorbidities (few smokers, low levels of dyslipidemia, heart failure, stroke, kidney disease, prior cardiovascular disease, hypertension). it would be important to have other studies in different areas of the world to confirm the results

-- as per above, there was no information about which anticoagulant or which antiplatelet drug was used and potential relationships to the outcomes. One issue this brings up is why did the clinicians choose which ones to prescribe?? Was there something in the individual’s clinical presentation that drove one or another?? and this choice perhaps reflected clinical issues that made one group of patients different from the other? Was it the clinician’s decision based on their own experience? Or that of a drug company? since the participants of this study were not randomized to anticoagulation (or by which agent), this could affect our interpretation of the results

-- the researchers note that the overall number of events found was fewer than what they had predicted in their study design, and the number of people with high CHA₂DS₂-VASc scores was lower. this could affect the statistical rigor of their results. Also there was a relatively small number of women (25%) in the study

-- this study also had patients with relatively low CHA₂DS₂-VASc and HAS-BLED scores, with followup of 2 years. this might not represent the results in patients who had higher scores and over longer periods of times

 

so, an impressive study finding that for many individuals with atrial fibrillation who have catheter ablation, monitoring them closely every 6 months for 1-2 years (1 year seemed to be sufficient in this group) might lead to discontinuing oral anticoagulation with a direct acting oral anticoagulant (especially apixaban), with potential benefit for both thrombotic and bleeding complications.

-- as with many studies done in one location, these results should be confirmed in other, diverse settings prior to latching onto their results

-- but the results of actually being able to stop anticoagulation is really profound, with potentially dramatic effects for the patients, their families, their communities, their work, their self-esteem, as well as decreasing health care utilization and costs

 

geoff

-----------------------------------

If you would like to be on the regular email list for upcoming blogs, please contact me at gmodest@bidmc.harvard.edu

to get access to all of the blogs:  go to http://gmodestmedblogs.blogspot.com/ to see the blogs in reverse chronological order

or you can just click on the magnifying glass on top right, then type in a name in the search box and get all the blogs with that name in them