The n e w e ng l a n d j o u r na l of m e dic i n e

Original Article

Left Atrial Appendage Closure after Ablation

for Atrial Fibrillation
O.M. Wazni,1 W.I. Saliba,1 D.G. Nair,2 E. Marijon,3 B. Schmidt,4 T. Hounshell,5
H. Ebelt,6 C. Skurk,7 S. Oza,8 C. Patel,9 A. Kanagasundram,10 A. Sadhu,11
S. Sundaram,12 J. Osorio,13 G. Mark,14 M. Gupta,15 D.B. DeLurgio,16 J. Olson,17
J.E. Nielsen‑Kudsk,18 L.V.A. Boersma,19,20 J.S. Healey,21 K.P. Phillips,22 F.M. Asch,23
K. Wolski,1 K. Roy,24 T. Christen,24 B.S. Sutton,24 K.M. Stein,24 and V.Y. Reddy,25
for the OPTION Trial Investigators*​​

A BS T R AC T

BACKGROUND
Oral anticoagulation is recommended after ablation for atrial fibrillation among The authors’ full names, academic de‑
patients at high risk for stroke. Left atrial appendage closure is a mechanical alterna- grees, and affiliations are listed at the
end of the article. Dr. Wazni can be con‑
tive to anticoagulation, but data regarding its use after atrial fibrillation ablation are tacted at ­waznio@​­ccf​.­org or at the Heart
lacking. and Vascular Institute, Cleveland Clinic,
9500 Euclid Ave., Cleveland, OH 44195.
METHODS *A list of the OPTION Trial Investigators
We conducted an international randomized trial involving 1600 patients with is provided in the Supplementary Ap‑
atrial fibrillation who had an elevated score (≥2 in men and ≥3 in women) on the pendix, available at [Link].
CHA2DS2-VASc scale (range, 0 to 9, with higher scores indicating a greater risk of This article was published on November 16,
stroke) and who underwent catheter ablation. Patients were randomly assigned in 2024, at [Link].
a 1:1 ratio to undergo left atrial appendage closure or receive oral anticoagulation. N Engl J Med 2025;392:1277-87.
The primary safety end point, tested for superiority, was non–procedure-related DOI: 10.1056/NEJMoa2408308
Copyright © 2024 Massachusetts Medical Society.
major bleeding or clinically relevant nonmajor bleeding. The primary efficacy end
point, tested for noninferiority, was a composite of death from any cause, stroke, CME
or systemic embolism at 36 months. The secondary end point, tested for noninferior-
ity, was major bleeding, including procedure-related bleeding, through 36 months.
RESULTS
A total of 803 patients were assigned to undergo left atrial appendage closure, and
797 to receive anticoagulant therapy. The mean (±SD) age of the patients was
69.6±7.7 years, 34.1% of the patients were women, and the mean CHA2DS2-VASc
score was 3.5±1.3. At 36 months, a primary safety end-point event had occurred
in 65 patients (8.5%) in the left atrial appendage closure group (device group) and
in 137 patients (18.1%) in the anticoagulation group (P<0.001 for superiority); a pri-
mary efficacy end-point event had occurred in 41 patients (5.3%) and 44 patients
(5.8%), respectively (P<0.001 for noninferiority); and a secondary end-point event
had occurred in 3.9% and 5.0% (P<0.001 for noninferiority). Complications related
to the appendage closure device or procedure occurred in 23 patients.
CONCLUSIONS
Among patients who underwent catheter-based atrial fibrillation ablation, left atrial
appendage closure was associated with a lower risk of non–procedure-related major
or clinically relevant nonmajor bleeding than oral anticoagulation and was nonin-
ferior to oral anticoagulation with respect to a composite of death from any cause,
stroke, or systemic embolism at 36 months. (Funded by Boston Scientific; OPTION
[Link] number, NCT03795298.)

n engl j med 392;13 [Link] April 3, 2025 1277

 The n e w e ng l a n d j o u r na l of m e dic i n e

C
atheter ablation for atrial fibril- patients provided written consent to participate.
lation is an effective strategy for treating An independent data and safety monitoring com-
symptomatic atrial arrhythmia. However, mittee oversaw patient safety and trial conduct
because of the risk of recurrence of atrial fibril- (Table S2). A clinical events committee adjudicated
lation, which may be minimally symptomatic, cur- all outcome events (Table S3), and staff at an in-
A Quick Take rent guidelines recommend indefinite continua- dependent core laboratory assessed all imaging;
is available at tion of oral anticoagulation in patients who are at the members of the clinical events committee and
[Link] moderate or high risk for stroke, regardless of the the laboratory staff were unaware of the group
perceived outcome of the ablation procedure.1 Oral assignments. The sponsor collected and moni-
anticoagulant therapy has important limitations, tored the trial data and performed outcome analy-
including a risk of bleeding, patient anxiety, and ses that were independently validated by data ana-
cost considerations, which result in a quarter of lysts and statisticians at the site of the principal
patients stopping oral anticoagulants within a investigator according to the statistical analysis
year after starting treatment.2-5 plan (available with the protocol). The principal
Catheter-based left atrial appendage closure is investigator had unrestricted access to the data and
an alternative strategy for stroke prophylaxis. The wrote the first draft of the manuscript. All the
safety and efficacy of left atrial appendage closure authors provided critical review of the manuscript
as compared with warfarin has been established, and vouch for the accuracy and completeness of
but data comparing left atrial appendage closure the data and for the fidelity of the trial to the pro-
devices with contemporary oral anticoagulants are tocol. Confidentiality restrictions between the spon-
limited.6 Furthermore, because patients may have a sor and authors were in place between the time
lower risk of stroke after atrial fibrillation abla- data became available and publication.
tion,7 whether the benefits of left atrial appendage
closure are outweighed by the associated risks, Patients
including short-term complications related to the Patients who underwent catheter ablation and
procedure and device-related thrombosis, is un- had a CHA2DS2-VASc score of at least 2 for men
clear.8 Accordingly, the Comparison of Anticoag- or at least 3 for women were eligible for random-
ulation with Left Atrial Appendage Closure after ization if the ablation had been performed 90 to
Atrial Fibrillation Ablation (OPTION) trial was 180 days before randomization or was scheduled
designed to determine whether left atrial append- to be performed within 10 days after randomiza-
age closure can safely decrease the risk of bleed- tion. Additional inclusion and exclusion criteria
ing associated with oral anticoagulants while are shown in Table S4. The CHA2DS2-VASc scale
maintaining a low risk of stroke among patients is used to assess the risk of stroke among patients
with atrial fibrillation who have undergone cath- with atrial fibrillation; scores range from 0 to
eter ablation and are at moderate or high risk for 9, with higher scores indicating a greater risk
stroke. of stroke. Patients were assigned in a 1:1 ratio
to undergo left atrial appendage closure (device
group) or receive oral anticoagulation (anticoagu-
Me thods
lation group), with randomization stratified ac-
Trial Design cording to site and timing of the catheter ablation
The OPTION trial was a multicenter, randomized procedure.
clinical trial. The trial protocol has been published
previously 9 and is available with the full text of Treatment Protocol and Follow-up
this article at [Link]. The trial was funded by Catheter ablation was performed before the left
the manufacturer of the left atrial appendage clo- atrial appendage closure device was implanted.
sure device (WATCHMAN FLX, Boston Scientific). After implantation, patients received oral anti-
The trial was designed by a steering committee coagulants and aspirin for 90 days, followed by
(Table S1 in the Supplementary Appendix, avail- aspirin alone until 12 months after randomization.
able at [Link]) in collaboration with the spon- Follow-up imaging of the device (by transesoph-
sor and the Food and Drug Administration. The ageal echocardiography [recommended] or com-
trial was approved by the ethics or institutional puted tomography) was performed at 3 months
review board at each participating site, and all and 12 months. Patients assigned to the antico-

1278 n engl j med 392;13 [Link] April 3, 2025

 Left Atrial Appendage Closure after Ablation for AF

agulation group started or continued receiving of 14% in the device group and 20% in the anti-
market-approved agents. The choice of oral anti- coagulation group. We increased the sample size
coagulant was at the discretion of the physician. to 1600 patients to allow for withdrawals and loss
Follow-up visits occurred at 3, 12, 24, and to follow-up. Testing for superiority at a two-sided
36 months after randomization. Clinical informa- alpha level of 0.05 was performed in the inten-
tion was recorded at each visit, including the HAS- tion-to-treat population (all patients who under-
BLED score (an assessment of major bleeding risk went randomization, grouped according to their
among patients with atrial fibrillation receiving assigned treatment) with the use of a log-rank
anticoagulants; range, 0 to 9, with higher scores P value based on the Kaplan–Meier estimation.
indicating a greater risk of bleeding) and recur- The trial had 85% power to detect noninferi-
rences of atrial arrhythmia. A recurrence of atrial ority with respect to the primary efficacy end
arrhythmia was defined as a documented episode point in the intention-to-treat population, with a
of atrial fibrillation or new-onset atrial flutter or noninferiority margin of 5 percentage points.
an atrial tachycardia event (≥30 seconds in dura- We assumed that a primary efficacy event would
tion or from a 10-second 12-lead electrocardio- occur in 10% of patients in each group. A non-
gram) or electrical or pharmacologic cardiover- inferiority margin of 5 percentage points, repre-
sion for atrial flutter or atrial tachycardia since a senting a relative risk of 1.5, was chosen because
previous visit. left atrial appendage closure is already considered
to be a safe and effective alternative to oral an-
End Points ticoagulation for the composite of death from
The primary safety end point was non–procedure- any cause, stroke, or systemic embolism in pa-
related bleeding — a combination of major bleed- tients with reasons to consider an alternative to
ing (as defined by the International Society on long-term anticoagulant use. The noninferiority
Thrombosis and Haemostasis [ISTH]) and clinically margin in this trial is similar to that used in the
relevant nonmajor bleeding (bleeding that required trials that established noninferiority of left atrial
medical intervention, led to hospitalization or in- appendage closure to oral anticoagulation.12-14
creased level of care, or prompted a face-to-face Noninferiority was assessed with the use of the
evaluation) — through 36 months (Table S5).10,11 Farrington–Manning method at a one-sided type
The primary efficacy end point was a composite of I error rate of 0.025 with the use of a one-sided
death from any cause, stroke, or systemic embo- z test. If the criteria were met for both superior-
lism at 36 months after randomization. The second- ity regarding the primary safety end point and
ary end point was ISTH major bleeding, including noninferiority regarding the primary efficacy end
procedure-related bleeding, through 36 months.11 point, the trial would be considered successful.
Data on peridevice leaks and device-related throm- If the criteria for success regarding the pri-
bosis were obtained at 3 months and 12 months. mary end points were met, the statistical analy-
The EuroQol Group 5-Dimension 5-Level ques- sis plan allowed for hierarchical testing for non-
tionnaire (EQ-5D-5L) and the 12-Item Short-Form inferiority of the secondary end point, with a
Health Survey (SF-12) were completed at the time noninferiority margin of 5.25 percentage points.
of enrollment and at 12 months and 36 months. Noninferiority was assessed with the use of a one-
Scores on the EQ-5D-5L index range from −0.59 to sided z test. The proportional-hazards assump-
1, with 1 indicating the best possible health state; tion for analyses reporting hazard ratios was
scores on the SF-12 range from 0 to 100, with assessed with the use of the Wald test from
higher scores indicating better health status. Ad- chi-square distribution for linearity of the log
ditional details regarding definitions of end point hazard ratio. Sensitivity analyses were conducted
are provided in the Supplementary Appendix. to assess the primary and secondary end points
in the per-protocol population (which included
Statistical Analysis patients in each treatment group regardless of
For the primary safety end point, we estimated adherence to anticoagulant therapy) and the on-
that a sample size of 1280 patients would provide treatment population (which included patients who
the trial with 86% power to show the superiority had ≥80% adherence to the protocol medication
of left atrial appendage closure to oral anticoagu- until a primary or secondary end-point event oc-
lation alone, assuming an incidence of bleeding curred or the trial ended). No statistical techniques

n engl j med 392;13 [Link] April 3, 2025 1279

 The n e w e ng l a n d j o u r na l of m e dic i n e

a device and continued treatment with oral anti-

coagulation, and 1 patient who was assigned to
1710 Patients were assessed for eligibility
the anticoagulation group received a device in
error. A total of 82 patients in the anticoagulation
110 Patients were excluded group crossed over to the device group; 67 of these
71 Did not meet eligibility criteria
26 Withdrew consent patients (82%) did so after a primary end-point
7 Were withdrawn by investigator event occurred (Table S6).
5 Had other reason
1 Was lost to follow-up The mean (±SD) age of the patients was
69.6±7.7 years and 34.1% were women. The mean
CHA2DS2-VASc score was 3.5±1.3, and the mean
1600 Underwent randomization (1:1)
HAS-BLED score was 1.2±0.8. Baseline character-
istics appeared to be well balanced between the
two trial groups (Table 1 and Table S7), and the
803 Were assigned to undergo ablation 797 Were assigned to undergo ablation
groups were representative of patients with atrial
and left atrial appendage closure and receive oral anticoagulation fibrillation who undergo catheter ablation pro-
(intention-to-treat population) (intention-to-treat population)
cedures (Table S8). Follow-up data at 36 months
were available for 758 patients (94.4%) in the
753 Received assigned intervention 796 Received assigned intervention
device group and 737 (92.5%) in the anticoagula-
(treated as assigned) (treated as assigned) tion group. A total of 28 patients (3.5%) in the
50 Did not receive assigned interven- 1 Received a left atrial appendage
tion closure device in error after
device group and 43 (5.4%) in the anticoagulation
41 Received no left atrial appendage randomization group withdrew consent, and 17 (2.1%) and 28
closure device
9 Had unsuccessful implantation
(3.5%) patients, respectively, were lost to follow-
up (Fig. 1 and Table S9).
Catheter ablation was performed after ran-
60 Discontinued prematurely 84 Discontinued prematurely domization in 40.9% of the patients (Table 1).
28 Withdrew consent
17 Were lost to follow-up
43 Withdrew consent
28 Were lost to follow-up
Radiofrequency ablation was used in 59.4% of
9 Had other reason 7 Had other reason the patients, and cryoablation in 33.2%. A return
4 Were withdrawn by investigator
2 Died
5 Were withdrawn by investigator
1 Died
of sinus rhythm after the ablation procedure oc-
curred in 88.1% of the patients (Table S10).
Overall, 95.0% of the patients received a non-
758 Had a 36-mo follow-up 737 Had a 36-mo follow-up warfarin anticoagulant (59.3% received apixaban,
visit or had an event visit or had an event 27.2% rivaroxaban, 4.3% edoxaban, 3.9% dabiga-
(intention-to-treat population) (intention-to-treat population)
tran, and 0.3% other). Throughout the trial, 84.8%
of the patients in the anticoagulation group con-
Figure 1. Eligibility, Randomization, and Follow-up.
tinued to receive oral anticoagulation. In the de-
vice group, 10.1% of patients were receiving oral
anticoagulation at 36 months (Fig. S1). The rea-
were used to impute missing data. Additional de- sons for continuing treatment with anticoagulants
tails regarding the statistical analysis are provided are shown in Table S11.
in the Supplementary Appendix. In the device group, implantation was consid-
ered by investigators to be successful in 753 of the
762 patients (98.8%) in whom implantation was
R e sult s
attempted. Complications related to the device or
Patients and Follow-up procedure occurred in 22 patients in the device
Between November 6, 2019, and June 28, 2021, group and in 1 patient in the anticoagulation group
a total of 1600 patients underwent randomization who crossed over to receive a device (Table S12). A
at 106 sites in 10 countries (a full list of sites is complete seal of the left atrial appendage with the
provided in the Supplementary Appendix). Of these device was observed in 81.0% of the patients at 3
patients, 803 were assigned to the device group months and in 79.7% of the patients at 12 months.
and 797 to the anticoagulation group (Fig. 1). The 12-month incidence of device-related thrombus
Fifty patients in the device group did not receive was 1.9% (Table S13).

1280 n engl j med 392;13 [Link] April 3, 2025

 Left Atrial Appendage Closure after Ablation for AF

Table 1. Characteristics of the Patients at Baseline.*

Device Anticoagulation
Group Group
Characteristic (N = 803) (N = 797)
Age — yr 69.7±7.4 69.4±7.9
Sex — no. (%)
Male 520 (64.8) 533 (66.9)
Female 283 (35.2) 263 (33.0)
Intersex 0 (0) 1 (0.1)
Race or ethnic group — no. (%)†
White 673 (83.8) 686 (86.1)
Black 14 (1.7) 11 (1.4)
Hispanic or Latino 13 (1.6) 12 (1.5)
Asian 4 (0.5) 1 (0.1)
American Indian or Alaskan Native 2 (0.2) 1 (0.1)
Other 3 (0.4) 3 (0.4)
Not disclosed 94 (11.7) 83 (10.4)
CHA2DS2-VASc score‡
Score 3.5±1.3 3.5±1.3
Distribution — no. (%)
1 5 (0.6) 7 (0.9)
2 to 3 428 (53.3) 422 (52.9)
4 to 5 310 (38.6) 307 (38.5)
≥6 60 (7.5) 61 (7.7)
HAS-BLED score§ 1.2±0.8 1.2±0.8
Score 1.2±0.8 1.2±0.8
Distribution — no. (%)
0 148 (18.4) 117 (14.7)
1 to 2 605 (75.3) 624 (78.3)
≥3 25 (3.1) 56 (7.0)
Atrial fibrillation pattern — no. (%)
Persistent 326 (40.6) 296 (37.1)
Paroxysmal 477 (59.4) 501 (62.9)
Present for <1 yr 257 (32.0) 251 (31.5)
Present for ≥1 yr 546 (68.0) 546 (68.5)
Sequential LAAC: ablation 90 to 180 days before 475 (59.2) —
randomization — no. (%)
Concomitant LAAC: ablation within 10 days after 328 (40.8) —
randomization — no. (%)¶
Ablation performed 90 to 180 days before random‑ — 471 (59.1)
ization — no. (%)
Ablation performed within 10 days after random- — 326 (40.9)
ization — no. (%)

* Plus–minus values are means ±SD. LAAC denotes left atrial appendage closure.
† Race or ethnic group was reported by the patient.
‡ CHA2DS2-VASc scores (an assessment of the risk of stroke among patients with atrial fibrillation) range from 0 to 9,
with higher scores indicating a higher risk of stroke.
§ HAS-BLED scores reflect the risk of major bleeding among patients with atrial fibrillation receiving anticoagulants and
range from 0 to 9, with higher scores indicating a greater risk of bleeding.
¶ Ablation occurred within 10 days after randomization and on the same day as the left atrial appendage closure except
in 3 patients.

n engl j med 392;13 [Link] April 3, 2025 1281

 The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Primary and Secondary End Points at 36 Months (Kaplan–Meier Estimates).*

Difference
Device Anticoagulation (one-sided 97.5%
Group Group upper confidence
End Point Analysis (N = 803) (N = 797) limit) P Value

no. of patients (%) percentage points

Primary end points
Safety: non–procedure-related bleeding† Superiority 65 (8.5) 137 (18.1) — <0.001
Efficacy: death from any cause, stroke, Noninferiority, with 5.0-per‑ 41 (5.3) 44 (5.8) −0.5 (1.8) <0.001
or systemic embolism‡ centage-point margin
Secondary end point
Major bleeding event§ Noninferiority, with 30 (3.9) 38 (5.0) –1.1 (1.0) <0.001
5.25-percentage-point
margin

* The analyses were performed in the intention-to-treat population, which included all patients who underwent randomization, grouped
according to their assigned treatment group; the start time of follow-up for the intention-to-treat analysis was the day of randomization.
Testing was performed in a hierarchical manner, with each step needing to reject the null hypothesis in order to proceed to the next step.
Step 1 was superiority testing of the primary safety end point and noninferiority testing of the primary efficacy end point, step 2 was nonin‑
feriority testing of the secondary end point, step 3 was superiority testing of the secondary end point, and step 4 was superiority testing of
the primary efficacy end point. The trial was considered to be successful if the criteria were met for both superiority regarding the primary
safety end point and noninferiority regarding the primary efficacy end point.
† Non–procedure-related bleeding was a composite of ISTH major bleeding or clinically relevant nonmajor bleeding. Non–procedure-related
events were those that occurred after 3 days following the procedure in the device group. The P value was calculated with the use of the log-
rank test and is based on the Kaplan–Meier estimation for the superiority testing.
‡ The P value for the primary efficacy end point (a composite of stroke, all-cause death, or systemic embolism) was calculated with the use of
the z test and is based on the standard normal distribution for the noninferiority testing and log-rank test for the superiority testing.
§ Major bleeding was defined as ISTH major bleeding, including procedure-related bleeding. The P value was calculated with the use of the z test
and is based on the standard normal distribution for the noninferiority testing and log-rank test for the superiority testing.

End Points point were similar in the per-protocol and on-

All analyses were performed in the intention-to- treatment populations (Tables S15 and S16).
treat population. Non–procedure-related major ISTH major bleeding, including procedure-
bleeding or clinically relevant nonmajor bleeding related bleeding, through 36 months (second-
(primary safety end point) occurred in 65 patients ary end point) occurred in 30 patients (Kaplan–
(Kaplan–Meier estimate, 8.5%) in the device group Meier estimate, 3.9%) in the device group and in
and in 137 patients (Kaplan–Meier estimate, 18.1%) 38 patients (Kaplan–Meier estimate, 5.0%) in the
in the anticoagulation group (hazard ratio, 0.44; anticoagulation group, for a difference of −1.1
95% confidence interval [CI], 0.33 to 0.59; P<0.001 percentage points (hazard ratio, 0.77; 95% CI,
for superiority) (Table 2, Fig. 2, and Table S14). 0.48 to 1.24; one-sided 97.5% upper confidence
Results of sensitivity analyses (assessed in the per- limit, 1.0), which met the criterion for noninfe-
protocol and on-treatment populations) were con- riority (P<0.001) but not superiority (P = 0.28)
sistent with the results of the primary end-point (Table 2, Fig. 2, and Fig. S2). Results of pre-
analysis (Table S15 and S16). specified subgroup analyses of the primary and
Death from any cause, stroke, or systemic em- secondary end points appeared to be consistent
bolism at 36 months (primary efficacy end point) across subgroups stratified according to age, sex,
occurred in 41 patients (Kaplan–Meier estimate, CHA2DS2-VASc score, HAS-BLED score, and type
5.3%) in the device group and in 44 patients of atrial fibrillation (paroxysmal or persistent)
(Kaplan–Meier estimate, 5.8%) in the anticoagu- (Fig. S3).
lation group, for a difference of −0.5 percentage By 36 months, death from any cause had oc-
points (hazard ratio, 0.91; 95% CI, 0.59 to 1.39; curred in 29 patients (Kaplan–Meier estimate,
one-sided 97.5% upper confidence limit, 1.8; 3.8%) in the device group and 34 patients (Ka-
P<0.001 for noninferiority) (Table 2, and Fig. 2). plan–Meier estimate, 4.5%) in the anticoagula-
Results of analyses of the primary efficacy end tion group, and ischemic stroke had occurred in

1282 n engl j med 392;13 [Link] April 3, 2025

 Left Atrial Appendage Closure after Ablation for AF

9 patients (Kaplan–Meier estimate, 1.2%) and ly result in a lower incidence of bleeding than oral
10 patients (Kaplan–Meier estimate, 1.3%) in the anticoagulation, while maintaining a low risk of
two groups, respectively. Hemorrhagic stroke had stroke, in patients with atrial fibrillation who
occurred in 3 patients (Kaplan–Meier estimate, underwent catheter ablation, with the majority
0.4%) in each group. Systemic embolism had oc- (95%) of patients receiving nonvitamin K antago-
curred in 2 patients (Kaplan–Meier estimate, 0.3%) nists (direct oral anticoagulants). The observed
in the device group and in 1 patient (Kaplan–Meier percentage of patients with bleeding at 36 months
estimate, 0.1%) in the anticoagulation group, and (18.1%) in the anticoagulation group was high
pericardial effusion requiring intervention had oc- despite the relatively low HAS-BLED score, which
curred in 2 patients (Kaplan–Meier estimate, 0.3%) emphasizes the challenges with long-term oral
and 5 patients (Kaplan–Meier estimate, 0.7%), anticoagulation. The 56% lower risk of non–
respectively. Additional safety end points through procedure-related bleeding with left atrial ap-
36 months are shown in Table S17, and a full list pendage closure than with oral anticoagulation
of adverse events is provided in Table S18. was driven largely by clinically relevant nonma-
The percentage of patients with clinical recur- jor bleeding (bleeding that required medical in-
rence of atrial fibrillation and of patients who tervention, led to hospitalization or increased level
received antiarrhythmic agents at 36 months is of care, or prompted a face-to-face evaluation).
shown in Table S19 and Figure S4. A descriptive Clinically relevant nonmajor bleeding is one of the
analysis of EQ-5D-5L and SF-12 quality-of life many causes of suboptimal adherence to long-
measures is shown in Table S20. term oral anticoagulation.17 Even with the inclu-
sion of procedure-related bleeding events, there
were numerically fewer major bleeding events
Discussion
with left atrial appendage closure than with oral
In this international trial involving patients at anticoagulation.
moderate-to-high risk for stroke who underwent The continued use of oral anticoagulants ex-
catheter ablation for atrial fibrillation, left atrial poses patients to a higher risk of bleeding than
appendage closure was associated with a lower the long-term aspirin therapy received by most
risk of non–procedure-related bleeding than oral patients who undergo left atrial appendage clo-
anticoagulation and was noninferior with respect sure. This observation is highlighted by the re-
to the composite end point of death from any sults of the ARTESIA trial (Apixaban for the Re-
cause, stroke, or systemic embolism at 36 months. duction of Thrombo-Embolism in Patients with
The fact that left atrial appendage closure can be Device-Detected Subclinical Atrial Fibrillation).
completed safely at the time of atrial fibrillation In the ARTESIA trial, in which patients had a
ablation makes it a possible alternative to long- CHA2DS2-VASc score of 3.9±1.1, which is similar
term oral anticoagulation. to that in the OPTION trial, patients who received
Randomized trials have shown that among treatment with apixaban had higher rates of ma-
patients with atrial fibrillation at elevated risk for jor bleeding than patients who received aspirin
stroke, left atrial appendage closure is associated (1.71% per patient-year vs. 0.94% per patient-year;
with a lower incidence of postprocedure bleeding, hazard ratio, 1.80; 95% CI, 1.26 to 2.57; P = 0.001).18
hemorrhagic stroke, and death from cardiovascu- In the OPTION trial, left atrial appendage clo-
lar causes than oral anticoagulation, with a simi- sure was noninferior to oral anticoagulation with
lar incidence of ischemic stroke.15,16 However, the respect to the primary efficacy composite end
oral anticoagulant used in these studies was most point of death from any cause, stroke, or sys-
often warfarin and the patients who were enrolled temic embolism. Deaths from any cause account-
in these trials were at high risk for stroke and ed for most of the efficacy end-point events and
bleeding. After ablation for atrial fibrillation, pa- the incidence appeared to be similar in the two
tients are not perceived to be at high risk for groups. None of the deaths were considered to be
bleeding, and the results of several nonrandom- related to the device. The incidence of ischemic
ized studies have suggested that catheter ablation stroke was low in both trial groups. The two groups
may reduce the incidence of ischemic stroke. appeared to be balanced with respect to sequential
The OPTION trial was designed to assess ablation (ablation performed 90 to 180 days before
whether left atrial appendage closure would safe- randomization), and the incidence of recurrence

n engl j med 392;13 [Link] April 3, 2025 1283

 The n e w e ng l a n d j o u r na l of m e dic i n e

A Non–Procedure-Related Major Bleeding or Clinically Relevant Nonmajor Bleeding (primary safety end point)
100 20 Anticoagulation
90
80

Percentage of Patients
10 Device
70
60
50
0
40 0 3 12 24 36
30
Hazard ratio, 0.44 (95% CI, 0.33–0.59)
20 P<0.001 for superiority
10
0
0 3 12 24 36
Months since Randomization
No. at Risk
Anticoagulation 797 753 701 657 598
Device 803 776 749 728 681

B Composite of Death from Any Cause, Stroke, or Systemic Embolism (primary efficacy end point)
100 10
90
80 Anticoagulation
Percentage of Patients

70 5

60 Device
50
0
40 0 3 12 24 36
30
Hazard ratio, 0.91 (95% CI, 0.59–1.39; one-sided 97.5% upper confidence limit, 1.8)
20 P<0.001 for noninferiority
10
0
0 3 12 24 36
Months since Randomization
No. at Risk
Anticoagulation 797 775 754 740 701
Device 803 782 772 757 722

C Major Bleeding (secondary end point)

100 10
90
80 Anticoagulation
Percentage of Patients

70 5

60 Device
50
0
40 0 3 12 24 36
30
Hazard ratio, 0.77 (95% CI, 0.48–1.24; one-sided 97.5% upper confidence limit, 1.0)
20 P<0.001 for noninferiority
10
0
0 3 12 24 36
Months since Randomization
No. at Risk
Anticoagulation 797 772 749 726 678
Device 803 778 763 746 708

1284 n engl j med 392;13 [Link] April 3, 2025

 Left Atrial Appendage Closure after Ablation for AF

Figure 2 (facing page). Primary and Secondary End

may be attributed to the use of computed tomog-
Points at 36 Months (Kaplan–Meier Analysis). raphy for monitoring in the OPTION trial (com-
Shown is the incidence, at 36 months after randomiza‑ puted tomography was used in 16% of patients at
tion, of non–procedure-related major bleeding or clini‑ 12 months), which may be more sensitive in detect-
cally relevant nonmajor bleeding (as defined by the In‑ ing leaks than echocardiography, which was used in
ternational Society on Thrombosis and Haemostasis the PINNACLE FLX study. The incidence of device-
[ISTH]), which was the primary safety end point (Panel A);
a composite of death from any cause, stroke, or systemic
related thrombosis appeared to be low at 12 months.
embolism, which was the primary efficacy end point Although the percentage of patients with a com-
(Panel B); and major bleeding (as defined by the ISTH), plete seal was lower in the OPTION trial than the
including procedure-related bleeding, which was the PINNACLE FLX study at 12 months, the observed
secondary end point (Panel C). The two primary co‑ incidence of ischemic stroke in our trial appeared
horts were patients who underwent catheter ablation
and implantation of a left atrial appendage closure de‑
to be similar in the two groups (occurring in 1.2%
vice and those who underwent catheter ablation and re‑ of patients assigned to undergo left atrial append-
ceived oral anticoagulation. The insets show the same age closure and in 1.4% of those assigned to oral
data on an expanded y axis. anticoagulation).
Left atrial appendage closure proved to be safe
in this trial. Although the attribution of specific
of atrial arrhythmias after the ablation procedure complications to left atrial appendage closure as
also appeared to be similar in the two groups, so compared with catheter ablation is confounded
it is unlikely that the outcomes could be attributed when these procedures are performed concomi-
to changes in the burden of atrial fibrillation. tantly, the most serious procedure-related compli-
The relatively low risk of stroke in both treat- cation, pericardial tamponade, occurred in 0.3%
ment groups may be related to a lower burden of of patients assigned to undergo left atrial append-
atrial fibrillation after ablation. Several studies age closure and in 0.7% assigned to receive oral
have shown that a higher atrial fibrillation burden anticoagulation. The low incidence of complica-
is associated with an increased risk of stroke.19-21 tions is consistent with advances in device tech-
Data have been mixed or underpowered to deter- nology.25,26
mine whether ablation reduces stroke rates.4,22,23 Our trial has limitations. At the time of the
Although the results of this trial appeared to be trial, pulsed field ablation was not available; how-
consistent across subgroups, the percentage of ever, the type of ablation technology used is un-
patients with a primary efficacy end-point event likely to alter the strategies used for stroke preven-
appeared to increase in both groups with increas- tion. Patients with a left ventricular ejection fraction
ing CHA2DS2-VASc score. In the absence of any of 30% or less were excluded from the trial, and
compelling data, current guidelines are uniform our results may not be applicable to these patients.
in advocating the continuation of oral anticoag- Although all bleeding events were adjudicated by
ulation after ablation on the basis of underlying an independent clinical events committee, the
stroke risk and not on the basis of the perceived open-label design could have led patients in the
success of the ablation procedure.24 In this con- anticoagulation group to seek medical attention
text, the low risk of death, stroke, or systemic more frequently.
embolism associated with both left atrial ap- The results of this trial showed that among
pendage closure and oral anticoagulation pro- patients at moderate-to-high risk for stroke who
vides alternatives for patients after atrial fibrilla- underwent catheter-based ablation for atrial fi-
tion ablation. brillation, left atrial appendage closure was associ-
Complete seal of the left atrial appendage at ated with a lower risk of non–procedure-related
early time points with the device used in the major or clinically relevant nonmajor bleeding
OPTION trial was similar to that in the PINNACLE than oral anticoagulation and was noninferior to
FLX study and the SURPASS Registry.25,26 However oral anticoagulation with respect to a composite
at 12 months, a complete seal was present in 80% end point of death from any cause, stroke, or
of patients in the OPTION trial, as compared with systemic embolism at 36 months.
90% of patients in the PINNACLE FLX study. Most Supported by Boston Scientific. No extramural funding was
observed leaks were 3 mm or less. The difference used to support this work.

n engl j med 392;13 [Link] April 3, 2025 1285

 The n e w e ng l a n d j o u r na l of m e dic i n e

Disclosure forms provided by the authors are available with partment, Georges Pompidou European Hospital, Paris; 4 Car‑
the full text of this article at [Link]. dioangiologisches Centrum Bethanien, Agaplesion Markus
A data sharing statement provided by the authors is available Krankenhaus, Frankfurt am Main, Germany; 5 Iowa Heart Center,
with the full text of this article at [Link]. West Des Moines; 6 Catholic Hospital, Sankt Johann Nepomuk,
Erfurt, Germany; 7 Deutsches Herzzentrum der Charité (DHZC),
Author Information Campus Benjamin Franklin, Berlin; 8 Ascension St. Vincent’s
The authors’ full names and academic degrees are as follows: Medical Center, Jacksonville, FL; 9 UPMC Pinnacle, Harrisburg,
Oussama M. Wazni, M.D.,1 Walid I. Saliba, M.D.,1 Devi G. Nair, PA; 10 Vanderbilt University, Nashville; 11 Phoenix Cardiovascular
M.D.,2 Eloi Marijon, M.D., Ph.D.,3 Boris Schmidt, M.D.,4 Troy Research Group, Phoenix, AZ; 12 South Denver Cardiology, Little‑
Hounshell, D.O.,5 Henning Ebelt, M.D.,6 Carsten Skurk, M.D.,7 ton, CO; 13 Grandview Medical Center, Birmingham, AL; 14 Heart
Saumil Oza, M.D.,8 Chinmay Patel, M.D.,9 Arvindh Kanagasun‑ House-Cooper University, Camden, NJ; 15 Lindner Center for Re‑
dram, M.D.,10 Ashish Sadhu, M.D.,11 Sri Sundaram, M.D.,12 Jose search and Education at Christ Hospital, Cincinnati; 16 Emory
Osorio, M.D.,13 George Mark, M.D.,14 Madhukar Gupta, M.D.,15 University, Medicine, Atlanta; 17 St. Vincent Heart Center of Indi‑
David B. DeLurgio, M.D.,16 Jeffrey Olson, D.O.,17 Jens Erik ana, Indianapolis; 18 Department of Cardiology, Aarhus Universi‑
Nielsen‑Kudsk, M.D., [Link].,18 Lucas V.A. Boersma, M.D., ty Hospital, Aarhus, Denmark; 19 Amsterdam University Medical
Ph.D.,19,20 Jeff S. Healey, M.D.,21 Karen P. Phillips, M.B., B.S.,22 Center, Amsterdam; 20 St. Antonius Hospital, Nieuwegein, the
Federico M. Asch, M.D.,23 Katherine Wolski, M.P.H,1 Kristine Netherlands; 21 Population Health Research Institute, Hamilton,
Roy, Ph.D.,24 Thomas Christen, M.D., Ph.D.,24 Brad S. Sutton, ON, Canada; 22 Brisbane AF Clinic, Greenslopes Private Hospital,
M.D.,24 Kenneth M. Stein, M.D.,24 and Vivek Y. Reddy, M.D.25 Brisbane, QLD, Australia; 23 Medstar Health Research Institute,
Medstar Washington Hospital Center, Washington, DC; 24 Bos‑
1
Cleveland Clinic, Cleveland; 2 St. Bernards Medical Center and ton Scientific, Marlborough, MA; 25 Cardiac Electrophysiology,
Arrhythmia Research Group, Jonesboro, AR; 3 Cardiology De‑ Mount Sinai Fuster Heart Hospital School of Medicine, New York.

References
1. Joglar JA, Chung MK, Armbruster AL, in atrial fibrillation: current status and of randomized trials. J Am Coll Cardiol
et al. 2023 ACC/AHA/ACCP/HRS guideline perspectives. Eur Heart J 2024;​45:​2914-32. 2020;​76:​2795-7.
for the diagnosis and management of atri- 9. Wazni OM, Boersma L, Healey JS, et al. 16. Kapadia SR, Krishnaswamy A, Whisenant
al fibrillation: a report of the American Comparison of anticoagulation with left B, et al. Concomitant left atrial appendage
College of Cardiology/American Heart As- atrial appendage closure after AF ablation: occlusion and transcatheter aortic valve re-
sociation Joint Committee on Clinical Prac- rationale and design of the OPTION ran- placement among patients with atrial fi-
tice Guidelines. J Am Coll Cardiol 2024;​83:​ domized trial. Am Heart J 2022;​251:​35-42. brillation. Circulation 2024;​149:​734-43.
109-279. 10. Kaatz S, Ahmad D, Spyropoulos AC, 17. Capiau A, Mehuys E, Van Tongelen I,
2. Chew D, Piccini JP. Long-term oral anti- Schulman S. Definition of clinically rel- et al. Community pharmacy-based study
coagulant after catheter ablation for atrial evant non-major bleeding in studies of of adherence to non-vitamin K antagonist
fibrillation. Europace 2021;​23:​1157-65. anticoagulants in atrial fibrillation and oral anticoagulants. Heart 2020;​106:​1740-6.
3. Freeman JV, Shrader P, Pieper KS, et al. venous thromboembolic disease in non- 18. Healey JS, Lopes RD, Granger CB, et al.
Outcomes and anticoagulation use after surgical patients. J Thromb Haemost 2015;​ Apixaban for stroke prevention in sub-
catheter ablation for atrial fibrillation. 13:​2119-26. clinical atrial fibrillation. N Engl J Med
Circ Arrhythm Electrophysiol 2019;​12(12):​ 11. Schulman S, Kearon C. Definition of 2024;​390:​107-17.
e007612. major bleeding in clinical investigations 19. Steinberg BA, Hellkamp AS, Lokhny-
4. Noseworthy PA, Yao X, Deshmukh AJ, of antihemostatic medicinal products in gina Y, et al. Higher risk of death and
et al. Patterns of anticoagulation use and non-surgical patients. J Thromb Haemost stroke in patients with persistent vs. par-
cardioembolic risk after catheter ablation 2005;​3:​692-4. oxysmal atrial fibrillation: results from the
for atrial fibrillation. J Am Heart Assoc 12. Holmes DR, Reddy VY, Turi ZG, et al. ROCKET-AF Trial. Eur Heart J 2015;​36:​
2015;​4(11):​e002597. Percutaneous closure of the left atrial ap- 288-96.
5. Okumura Y, Nagashima K, Arai M, et pendage versus warfarin therapy for pre- 20. Al-Khatib SM, Thomas L, Wallentin L,
al. Current status and clinical outcomes vention of stroke in patients with atrial et al. Outcomes of apixaban vs. warfarin
of oral anticoagulant discontinuation af- fibrillation: a randomised non-inferiority by type and duration of atrial fibrillation:
ter ablation for atrial fibrillation in Japan trial. Lancet 2009;​374:​534-42. results from the ARISTOTLE trial. Eur
— findings from the AF Frontier Ablation 13. Holmes DR Jr, Kar S, Price MJ, et al. Heart J 2013;​34:​2464-71.
Registry. Circ J 2019;​83:​2418-27. Prospective randomized evaluation of the 21. Uittenbogaart SB, Lucassen WAM,
6. Osmancik P, Herman D, Neuzil P, et al. Watchman left atrial appendage closure de- van Etten-Jamaludin FS, de Groot JR, van
Left atrial appendage closure versus direct vice in patients with atrial fibrillation versus Weert HCPM. Burden of atrial high-rate
oral anticoagulants in high-risk patients long-term warfarin therapy: the PREVAIL episodes and risk of stroke: a systematic
with atrial fibrillation. J Am Coll Cardiol trial. J Am Coll Cardiol 2014;​64:​1-12. review. Europace 2018;​20:​1420-7.
2020;​75:​3122-35. 14. Lakkireddy D, Thaler D, Ellis CR, et al. 22. Karasoy D, Gislason GH, Hansen J, et
7. Friberg L, Tabrizi F, Englund A. Cath- Amplatzer amulet left atrial appendage oc- al. Oral anticoagulation therapy after ra-
eter ablation for atrial fibrillation is as- cluder versus watchman device for stroke diofrequency ablation of atrial fibrillation
sociated with lower incidence of stroke prophylaxis (amulet IDE): a randomized con- and the risk of thromboembolism and
and death: data from Swedish health reg- trolled trial. Circulation 2021;​144:​1543-52. serious bleeding: long-term follow-up in
istries. Eur Heart J 2016;​37:​2478-87. 15. Turagam MK, Osmancik P, Neuzil P, nationwide cohort of Denmark. Eur Heart J
8. Landmesser U, Skurk C, Tzikas A, Dukkipati SR, Reddy VY. Left atrial ap- 2015;​36:​307-14a.
Falk V, Reddy VY, Windecker S. Left atrial pendage closure versus oral anticoagu- 23. Packer DL, Mark DB, Robb RA, et al.
appendage closure for stroke prevention lants in atrial fibrillation: a meta-analysis Effect of catheter ablation vs antiarrhythmic

1286 n engl j med 392;13 [Link] April 3, 2025

 Left Atrial Appendage Closure after Ablation for AF

drug therapy on mortality, stroke, bleed- fibrillation: a report of the American Col- from the PINNACLE FLX trial. Circulation
ing, and cardiac arrest among patients lege of Cardiology/American Heart Asso- 2021;​143:​1754-62.
with atrial fibrillation: the CABANA ran- ciation Joint Committee on Clinical Prac- 26. Kapadia SR, Yeh RW, Price MJ, et al.
domized clinical trial. JAMA 2019;​321:​ tice Guidelines. Circulation 2024;​149(1):​ Outcomes with the WATCHMAN FLX in
1261-74. e1-e156. everyday clinical practice from the NCDR
24. Joglar JA, Chung MK, Armbruster AL, 25. Kar S, Doshi SK, Sadhu A, et al. Prima- Left Atrial Appendage Occlusion Registry.
et al. 2023 ACC/AHA/ACCP/HRS guideline ry outcome evaluation of a next-generation Circ Cardiovasc Interv 2024;​17(9):​e013750.
for the diagnosis and management of atrial left atrial appendage closure device: results Copyright © 2024 Massachusetts Medical Society.

journal archive at [Link]

Every article published by the Journal is now available at [Link], beginning
with the first article published in January 1812. The entire archive is fully searchable,
and browsing of titles and tables of contents is easy and available to all.
Individual subscribers are entitled to free 24-hour access to 50 archive articles per year.
Access to content in the archive is also being provided through many institutional
subscriptions.

n engl j med 392;13 [Link] April 3, 2025 1287

Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.