Original Contribution

Left Atrial Appendage Closure in Patients With Atrial Fibrillation at Very High Bleeding Risk Without Postimplantation Antithrombotic Therapy

Arthur Darmon, MD; Etienne L. Couture MD, MPH; Gauthier Stein MD; Bertrand Cormier, MD; Bernard Chevalier, MD; Thierry Lefèvre, MD; Antoinette Neylon, MD; Francesca Sanguineti, MD; Jerome Horvilleur, MD; Philippe Garot, MD

Arthur Darmon, MD; Etienne L. Couture MD, MPH; Gauthier Stein MD; Bertrand Cormier, MD; Bernard Chevalier, MD; Thierry Lefèvre, MD; Antoinette Neylon, MD; Francesca Sanguineti, MD; Jerome Horvilleur, MD; Philippe Garot, MD

Abstract: Objectives. We assess the proportion, baseline characteristics, and outcomes of a cohort of very high bleeding risk (HBR) patients discharged with no antithrombotic therapy after left atrial appendage closure (LAAC). Background. The optimal antithrombotic therapy after LAAC remains controversial. However, a substantial proportion of patients have HBR and are contraindicated to any antithrombotic therapy. Data regarding the feasibility and safety of such a strategy are scarce. Methods. All patients who underwent LAAC at our institution between October 2013 and December 2018 were included. Clinical, procedural, and imaging data were collected prospectively, and patients receiving no antithrombotic therapy were compared with those receiving at least 1 agent. Results. A total of 152 patients were included. At discharge, 72 (47.3%) received single-antiplatelet therapy (SAPT), 57 (37.5%) received dual-antiplatelet therapy (DAPT), and 22 (14.5%) received no antithrombotic therapy (NATT). One patient received a combination of aspirin and vitamin K antagonist. There were no differences regarding baseline characteristics between NATT patients and the others. At a median follow-up of 16 months, 25 patients died. There were no differences between NATT patients and others regarding all-cause mortality (9.1% vs 18.8%, respectively; P=.07) or ischemic stroke (0.0% vs 4.2%, respectively; P=.20). There were 4 episodes of major bleeding, but none occurred in the NATT group, and no device-related thrombosis was seen on computed tomography scan performed within 12 weeks after LAAC. Conclusions. In highly selected patients at very HBR, discontinuation of any antithrombotic therapy after LAAC appears safe and feasible.

J INVASIVE CARDIOL 2020;32(10):385-391. 

Key words: antithrombotic therapy, atrial fibrillation, high bleeding risk, left atrial appendage closure

Percutaneous left atrial appendage closure (LAAC) is an alternative to oral anticoagulants (OACs) for prevention of stroke and thromboembolic events in patients with non-valvular atrial fibrillation (AF) and with contraindications to OAC therapy.1 Antiplatelet therapy is often prescribed after successful percutaneous LAAC in order to prevent device-related thrombosis (DRT) and thromboembolic events, while promoting endothelialization of the device. 

However, the optimal antithrombotic therapy after percutaneous LAAC remains uncertain, particularly regarding the type and duration of the antiplatelet agent regimen. In the PROTECT2 and PREVAIL3 trials, patients received warfarin alone or in combination with aspirin for 45 days, followed by prolonged dual-antiplatelet therapy (DAPT). In patients with contraindications to OAC, the ASAP trial4 evaluated DAPT for 6 months followed by prolonged single-antiplatelet therapy (SAPT), whereas patients in the Amplatzer trial5 received either SAPT or DAPT for 1 to 6 months. However, patients eligible for percutaneous LAAC may be at high bleeding risk, and while they have contraindications to OAC, some of them also have contraindications to antiplatelet therapy and, therefore, receive no antithrombotic therapy (NATT) at discharge. In the EWOLUTION6,7 and FLAAC8 registries, NATT patients accounted for 6% and 3.3% of the population, respectively. At our institution, patients referred for LAAC were highly selected according to the presence of severe comorbidities with very recent or active bleeding. Some of them were consequently discharged without any antithrombotic agents and we aimed to study their clinical outcomes with specific attention to the risk of recurrent thromboembolic events, stroke, and DRT.


Study design and patient selection. All patients who underwent LAAC at our institution between October 2013 and December 2018 were included in the analysis. In order to be eligible for LAAC, patients had to fulfill the following criteria: (1) documented non-valvular AF; (2) high thromboembolic risk (CHA2DS2VASc score ≥4) with high bleeding risk (HASBLED score ≥3); and (3) a contraindication to OAC. In France, reimbursement has been obtained for patients with an absolute and definitive contraindication to any anticoagulant. For this reason, patients were referred to our center because of severe comorbidities with very recent or sometimes active bleeding. Similar to the heart team approach in multivessel coronary artery disease patients, we have designed a multidisciplinary consensus involving other physicians (ie, neurologist, hematologist, gastroenterologist, anesthesiologist, etc) in an “LAAC heart team consensus” to discuss the indication and the postprocedural antithrombotic treatment of these patients. Before the procedure, patients underwent computed tomography (CT) scan of the LAA, which was completed by transesophageal echocardiography in patients with suspected LAA thrombus. Baseline characteristics, risk assessment, and procedural data were collected both prospectively and retrospectively. 

The two devices used were the Watchman (Boston Scientific) and the Amplatzer Amulet (Abbott Vascular).9 All procedures were performed according to international current practice consensus10,11 by trained physicians in order to ensure optimal procedural success and safety. From 2015 procedures were guided by fusion between angiography and CT in order to enhance transseptal puncture, as well as device sizing and optimal positioning. We performed transthoracic echocardiography 1 hour after the procedure in order to rule out periprocedural complications (ie, pericardial effusion, device migration). In the absence of procedural complications, patients were discharged home within 24-48 hours.

Follow-up was conducted at our institution, and each patient was prescribed a control CT scan to be performed between 6 and 12 weeks after the index procedure (with systematic double interpretation by two distinct and blinded physicians) for assessment of DRT, peridevice leaks, and quantitative opacification of LAA. Mid-term follow-up was conducted by phone interviews with either the patient, his/her relatives, and/or their general practitioner or cardiologist. The study flow chart is displayed in Figure 1.

Antithrombotic therapy after LAAC. Antithrombotic management of patients after successful LAAC was at the operator’s discretion, in accordance with the LAAC heart team, including the specialist of the organ where bleeding occurred. In general, OACs were avoided, and strategies were mostly based on antiplatelet agents. The various antithrombotic therapies at discharge included DAPT (combination of low-dose aspirin [75-160 mg once daily] and clopidogrel [75 mg once daily]) for 6 weeks to 6 months followed by lifelong SAPT, lifelong SAPT only, or NATT. 

Outcomes definition. The main outcomes for the present analysis were all-cause death, cardiovascular death, ischemic stroke, systemic embolism, and bleeding classified from 3-5 as defined by the Bleeding Academic Research Consortium (BARC).12 Each outcome was assessed during the first follow-up visit scheduled after the first CT scan, and during phone interviews.

Statistical analysis. Baseline characteristics are expressed as mean ± standard deviation for continuous variable and count (percentage) for categorical variables. Baseline characteristics, periprocedural outcomes, and short-term to mid-term outcomes were compared between the NATT group and other patients. Continuous variables were compared by analysis of variance and categorical variables by Chi-squared test. A P-value <.05 was considered to be statistically significant. We used JMP software, version 9.0 (SAS). 


Baseline characteristics of the study population. Between October 2013 and December 2018, a total of 152 patients underwent successful LAAC. Baseline characteristics of the population are reported in Table 1. Mean age was 79.3 ± 7.2 years, and 106 (69.7%) were male. LAAC was indicated for intracranial or gastrointestinal bleeding in 35.5% of the patients, high bleeding risk without a history of bleeding in 10.0% and urinary bleeding in 6.0%. Mean CHA2DS2-VASc score was 4.4 ± 1.1 and mean HASBLED score was 3.8 ± 1.1. 

At discharge, 72 patients (47.3%) were prescribed SAPT with ≤160 mg aspirin, 57 (37.5%) were on DAPT, and 22 (14.5%) were on NATT. Only 1 patient received a combination of ≤160 mg aspirin and OAC (Figure 2).

There were no significant differences regarding clinical baseline characteristics between NATT patients and SAPT/DAPT patients in terms of age (80.6 ± 6.6 years vs 79.1 ± 7.2 years, respectively; P=.36), sex (68.2% men vs 70.0% men, respectively; P=.86), hypertension (81.8% vs 73.1%, respectively; P=.37), history of heart failure (33.3% vs 20.6%, respectively; P=.21), stroke (42.9% vs 40.5%, respectively; P=.83), or cancer (30.0% vs 24.0%, respectively; P=.57). However, there was a trend toward lower rates of previous percutaneous coronary intervention in NATT patients vs SAPT/DAPT patients (9.1% vs 25.2%, respectively; P=.06).

NATT patients were more frequently referred because of recent (<3 months) intracranial hemorrhage compared with SAPT/DAPT patients (52.2% vs 33.1%, respectively; P=.048). There was no significant difference with respect to other bleeding sites between groups (Figure 3). The main reasons precluding prescription of any antithrombotic therapy after LAAC were a history of recent intracranial bleeding, angiodysplasia, mobility disorders, cognitive decline or loss of autonomy, and repeated bleeding upon resumption of OAC treatment (Table 2).

Finally, there were no differences in thromboembolic and bleeding risk assessment by CHA2DS2VASc and HASBLED scores between NATT and SAPT/DAPT patients (P=.21 and P=.94, respectively).

Periprocedural outcomes. Periprocedural data and outcomes are reported in Table 3. The large majority of devices implanted (109/152; 71.7%) were Amplatzer Amulet occluders. There were no significant differences regarding periprocedural complications between NATT patients and those receiving SAPT or DAPT. There were 9 pericardial effusions (6.0%) requiring percutaneous drainage, 5 cases of major bleeding requiring transfusion (3.3%), and 1 periprocedural stroke with complete recovery. Overall, any BARC 3-5 bleeding occurred in 3 patients (13.6%) in the NATT group and 9 patients (5.9%) in the SAPT/DAPT group, without any statistical difference (P=.31). One occurrence of device migration required uncomplicated surgical retrieval. There was no intrahospital death. 

CT follow-up at 6-12 weeks. Between 6 and 12 weeks following successful LAAC, 112/152 patients (73.6%) had a repeat CT scan (Table 4). No patient had device-related thrombosis on its atrial side. On the appendage side, there was a complete thrombosis of the device in 72/112 patients (64.3%). LAA opacification, reflecting its degree of occlusion, was partial in 38/112 (33.9%), total in 18/112 (16.1%), and absent in the majority of patients (60/112; 53.5%). There were no statistical differences in any of the CT outcomes between the NATT and SAPT/DAPT groups.

Clinical outcomes. Vital status was obtained for 100% of the patients at 3 months and 95.3% at a median follow-up of 16 months (mean follow-up, 20.6 ± 15.4 months) (Table 5). Four of the 152 patients (2.6%) died within 3 months, without no difference between the NATT and SAPT/DAPT groups (1/22 vs 3/130, respectively; P=.39). Half of these patients died of cardiovascular causes, but no recurrent stroke or systemic embolism occurred. No bleeding occurred during the first 3 months of follow-up, regardless of the antithrombotic therapy strategy.

Twenty-five of 145 patients (17.3%) died during the follow-up period; 4 (2.9%) died from cardiovascular causes, while the majority died from non-cardiovascular causes. Five patients had ischemic stroke (3.6%), but no systemic embolism occurred. In addition, 4 patients receiving antiplatelet agents had BARC 3-5 bleeding (intracranial bleeding) that was fatal in 2 patients. There were no differences between NATT patients and SAPT/DAPT patients regarding all-cause mortality (9.1% vs 18.8%, respectively; P=.07), ischemic stroke (0.0% vs 4.2%, respectively; P=.20), or death from cardiovascular causes (0.0% vs 3.4%, respectively; P=.24).


The main result of our analysis is that in highly selected patients referred for LAAC because of very recent or active bleeding conditions, the decision not to prescribe any antithrombotic therapy after the procedure seems a feasible and safe strategy that may not be associated with any excess risks of device-related thrombosis, ischemic stroke, or systemic embolism. Twenty-two patients were deemed at such a high bleeding risk by the medical heart team that any antithrombotic therapy was considered contraindicated. Although usually ineligible for clinical trials, such patients are relatively frequently encountered in clinical practice. For these reasons, in the present report, the proportion of patients discharged without antithrombotic therapy was 14.5%, which is markedly higher than the percentages reported in the literature. In 2 recent registries evaluating the Amulet device, the proportions of NATT patients were only 2.0%13 and 2.2%.14 In the EWOLUTION registry6,7 and in a recent French cohort,15 this subgroup of patients accounted for only 6.4% and 7.5% of the entire study population, respectively. The EWOLUTION registry6 reported no differences regarding ischemic stroke or thrombus on the device in the NATT subgroup of patients at 3-month follow-up exam. However, patients included in the EWOLUTION registry may not be similar to our “real-life” high bleeding risk cohort, and epidemiological data and longer-term follow-up are lacking for this subgroup of patients. 

The rate of major bleeding following LAAC remains high, ranging from 5.6% for patients on SAPT regimen16 to 10.7% in patients treated with DAPT.17 In our cohort, 4 episodes of major bleeding occurred during follow-up, of which 2 were fatal (both intracranial bleeding). No major bleeding event occurred in NATT patients, which highlights the crucial role of the ischemic/bleeding events balance, especially in very high bleeding risk patients, who are also at very high risk of ischemic events.18

Optimal antithrombotic therapy after LAAC is yet to be defined, given the large number of clinical trials that have evaluated different combinations and different durations of antithrombotic agents. However, the strategies evaluated over time have shown a systematic trend toward a reduction in duration and intensity of antithrombotic treatment. In the first trials,2,3,19 patients were treated using a combination of warfarin and aspirin for 45 days, followed by prolonged DAPT. Patients with contraindications to vitamin K antagonists in the ASAP study were treated with DAPT for 6 months, and received either SAPT or DAPT for 1-6 months in the Amplatzer trial.5 More recently, a single-center trial17 demonstrated the feasibility of short DAPT for 6 weeks after LAAC. Finally, 2 studies16,20 have prospectively evaluated aspirin monotherapy after LAAC. The rates of device-related thrombosis (1.9% and 3.3%, respectively) and recurrent stroke (2.3% per year and 0.0% per year, respectively) were remarkably low in this high-risk population. 

Overall, thanks to procedure simplification and device improvement — and in parallel with the operators’ learning curve — the rates of periprocedural and long-term adverse events, particularly thromboembolic events, have been continuously decreasing despite the scaling down of post-LAAC antithrombotic strategies. Our study confirms this trend, since no DRT on the atrial side occurred at 3 months after the procedure, without any alarming signal pointing to the incidence of thromboembolic events during short- and mid-term follow-up in the NATT group. Of note, in the recent Amulet registry,14 70% of the DRT cases were diagnosed within 100 days of the procedure. 

In order to optimize antithrombotic management after LAAC, clinicians need to quantitatively address the overall risks of their patients. The risk scores most used in this context are the CHA2DS2VASc and the HASBLED,21,22 which evaluate the ischemic and bleeding risks, respectively. There were no statistically significant differences regarding these 2 risk scores between the NATT group and the SAPT/DAPT group. Although the HASBLED score has proved to be the most accurate predictor of bleeding events in the entire population of patients with AF,23-25 the performance of this score in stratifying patients referred for LAAC is questionable. Indeed, these patients are by definition at high bleeding risk (mean HASBLED score was 3.8 ± 1.1), and other tools guided by clinical judgment should be used. In particular, in our cohort of NATT patients, there was a high prevalence of clinical factors (ie, history of intracranial bleeding, amyloid angiopathy, history of multiple bleeding, bleeding of unknown localization) and general conditions (ie, high risk of falls, cognitive impairment) that were classified by the heart team for LAAC as being associated with increased risk of bleeding. These conditions are not included in the calculation of these scores and could be useful for evaluating the risk of bleeding in these patients.

Study limitations. Certain limitations to our study are worth highlighting. First, this study is limited by the relatively low number of patients in the NATT group, and event count, which may result in a lack of statistical power. In that sense, our results add some pieces of evidence to the current literature, and should be seen as hypothesis generating, in order to be confirmed in future randomized trials. Second, this was a non-randomized, single-center study, which carries inherent bias due to its methodology, and events were not adjudicated by a clinical events committee. Third, the choice of antithrombotic strategy at discharge was at the operator’s discretion and was not pre-established based on validated criteria. However, this limitation could be seen as a strength, reflecting “real-life” management of high-risk patients. There was no occurrence of DRT on CT scan following LAAC, which is inconsistent with previous publications on this subject, even if the rate remains quite low (1.7%) in the most recent publication.14-16,20 This difference can be explained by the lack of a consensus regarding the definition of device-related thrombosis. In our analysis, we only considered thrombosis on the atrial side of the device, given that thrombosis of the appendage can be induced by the treatment. This emphasizes the need for a consensus regarding the definition of the different types of DRT. Finally, the relatively small number of patients in each treatment subgroup (ie, SAPT, DAPT, NATT, and SAPT plus OAC) led us to analyze NATT patients compared with all other patients “pooled” into a single group in order to gain power, but this, in turn, might have introduced a bias. However, events were so rare that it seems unlikely that analyzing such groups separately would yield different results.


In the present study of 152 patients, 22 highly selected patients (14.5%) with active, very recent, or recent bleeding did not receive antithrombotic therapy after LAAC. This strategy appears safe and feasible, without increased risk of device-related thrombosis, ischemic stroke, or systemic embolism. Further studies are warranted in order to confirm these results prior to its implementation in real-life practice. 

Acknowledgment. We thank Miss Catherine Dupic for her skillful technical assistance.

From the Hôpital Privé Jacques Cartier, Institut Cardiovasculaire Paris-Sud, Ramsay-Santé, Massy, France.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Darmon reports grant support from Abbott Vascular; personal fees from Bayer. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript accepted March 23, 2020.

Address for correspondence: Philippe Garot, MD, Institut Cardiovasculaire Paris Sud, 6 Avenue du Noyer Lambert, 91300 Massy, France. Email: pgarot@angio-icps.com

  1. Kirchhof P, Benussi S, Kotecha D, et al. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Europace. 2016;18:1609-1678.
  2. Holmes DR, Reddy VY, Turi ZG, et al. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial. Lancet. 2009;374:534-542.
  3. Holmes DR Jr, Kar S, Price MJ, et al. Prospective randomized evaluation of the Watchman left atrial appendage closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol. 2014;64:1-12.
  4. Reddy VY, Mobius-Winkler S, Miller MA, et al. Left atrial appendage closure with the Watchman device in patients with a contraindication for oral anticoagulation: the ASAP study (ASA Plavix feasibility study with Watchman left atrial appendage closure technology). J Am Coll Cardiol. 2013;61:2551-2556.
  5. Urena M, Rodes-Cabau J, Freixa X, et al. Percutaneous left atrial appendage closure with the Amplatzer cardiac plug device in patients with nonvalvular atrial fibrillation and contraindications to anticoagulation therapy. J Am Coll Cardiol. 2013;62:96-102.
  6. Bergmann MW, Betts TR, Sievert H, et al. Safety and efficacy of early anticoagulation drug regimens after Watchman left atrial appendage closure: three-month data from the EWOLUTION prospective, multicentre, monitored international Watchman LAA closure registry. EuroIntervention. 2017;13:877-884.
  7. Boersma LV, Schmidt B, Betts TR, et al. Implant success and safety of left atrial appendage closure with the Watchman device: periprocedural outcomes from the EWOLUTION registry. Eur Heart J. 2016;37:2465-2474.
  8. Teiger E, Thambo JB, Defaye P, et al. French National Left Atrial Appendage Closure Registry I: percutaneous left atrial appendage closure is a reasonable option for patients with atrial fibrillation at high risk for cerebrovascular events. Circ Cardiovasc Interv. 2018;11:e005841.
  9. Asmarats L, Rodes-Cabau J. Percutaneous left atrial appendage closure: current devices and clinical outcomes. Circ Cardiovasc Interv. 2017;10:e005359. 
  10. Saw J, Lempereur M. Percutaneous left atrial appendage closure: procedural techniques and outcomes. JACC Cardiovasc Interv. 2014;7:1205-1220.
  11. Tzikas A, Gafoor S, Meerkin D, et al. Left atrial appendage occlusion with the Amplatzer Amulet device: an expert consensus step-by-step approach. EuroIntervention. 2016;11:1512-1521.
  12. Mehran R, Rao SV, Bhatt DL, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. Circulation. 2011;123:2736-2747.
  13. Landmesser U, Schmidt B, Nielsen-Kudsk JE, et al. Left atrial appendage occlusion with the Amplatzer Amulet device: periprocedural and early clinical/echocardiographic data from a global prospective observational study. EuroIntervention. 2017;13:867-876.
  14. Aminian A, Schmidt B, Mazzone P, et al. Incidence, characterization, and clinical impact of device-related thrombus following left atrial appendage occlusion in the prospective global Amplatzer Amulet observational study. JACC Cardiovasc Interv. 2019;12:1003-1014.
  15. Fauchier L, Cinaud A, Brigadeau F, et al. Device-related thrombosis after percutaneous left atrial appendage occlusion for atrial fibrillation. J Am Coll Cardiol. 2018;71:1528-1536.
  16. Korsholm K, Nielsen KM, Jensen JM, et al. Transcatheter left atrial appendage occlusion in patients with atrial fibrillation and a high bleeding risk using aspirin alone for post-implant antithrombotic therapy. EuroIntervention. 2017;12:2075-2082.
  17. Weise FK, Bordignon S, Perrotta L, et al. Short-term dual antiplatelet therapy after interventional left atrial appendage closure with different devices. EuroIntervention. 2018;13:e2138-e2146.
  18. Urban P, Mehran R, Colleran R, et al. Defining high bleeding risk in patients undergoing percutaneous coronary intervention: a consensus document from the Academic Research Consortium for high bleeding risk. Eur Heart J. 2019;40:2632-2653.
  19. Reddy VY, Sievert H, Halperin J, et al; the PROTECT AF Steering Committee and Investigators. Percutaneous left atrial appendage closure vs warfarin for atrial fibrillation: a randomized clinical trial. JAMA. 2014;312:1988-1998.
  20. Rodriguez-Gabella T, Nombela-Franco L, Regueiro A, et al. Single antiplatelet therapy following left atrial appendage closure in patients with contraindication to anticoagulation. J Am Coll Cardiol. 2016;68:1920-1921.
  21. Kirchhof P, Benussi S, Kotecha D, et al; the ESC Scientific Document Group. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS. Eur Heart J. 2016;37:2893-2962.
  22. Pisters R, Lane DA, Nieuwlaat R, et al. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: the Euro Heart Survey. Chest. 2010;138:1093-1100.
  23. Apostolakis S, Lane DA, Guo Y, Buller H, Lip GY. Performance of the HEMORR(2)HAGES, ATRIA, and HAS-BLED bleeding risk-prediction scores in patients with atrial fibrillation undergoing anticoagulation: the AMADEUS (evaluating the use of SR34006 compared to warfarin or acenocoumarol in patients with atrial fibrillation) study. J Am Coll Cardiol. 2012;60:861-867.
  24. Esteve-Pastor M, Rivera-Caravaca J, Roldan V, et al. Long-term bleeding risk prediction in ‘real world’ patients with atrial fibrillation: comparison of the HAS-BLED and ABC-Bleeding risk scores. Thromb Haemost. 2017;117:1848-1858.
  25. Roldan V, Marin F, Manzano-Fernandez S, et al. The HAS-BLED score has better prediction accuracy for major bleeding than CHADS2 or CHA2DS2-VASc scores in anticoagulated patients with atrial fibrillation. J Am Coll Cardiol. 2013;62:2199-2204.