International Journal of Cardiology 168 (2013) 1023–1027
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Left atrial appendage closure with the Amplatzer™ Cardiac Plug: Impact of shape and device sizing on follow-up leaks☆ Xavier Freixa, Apostolos Tzikas, Ayax Sobrino, Jason Chan, Arsène J. Basmadjian, Réda Ibrahim ⁎ Department of Medicine, Montreal Heart Institute, and Université de Montréal, Quebec, Canada
a r t i c l e
i n f o
Article history: Received 29 May 2012 Received in revised form 31 August 2012 Accepted 28 October 2012 Available online 27 December 2012 Keywords: Left atrial appendage Amplatzer Cardiac Plug Stroke Atrial fibrillation Leaks
a b s t r a c t Background: The Amplatzer Cardiac Plug (ACP) is a relatively new system for left atrial appendage (LAA) closure. In general, the particular design of the device and the variability of the LAA anatomy make the sizing process challenging. The objective of our study was to analyze the impact of the shape and sizing of the device on the presence of follow-up leaks. Methods: The shape was evaluated by angiography and classified as “strawberry-like”, “square-like” and “tire-like”. The presence of peri-device leaks was assessed by transesophageal echocardiography (TEE) and graded as: 1) severe, 2) major, 3) moderate, 4) minor and 5) absent. Results: Twenty five patients were included. Overall, the device was oversized by 3.38 ± 2.26 mm and 1.58 ± 2.75 mm in relation to the largest angiographic and TEE diameter respectively. None of the patients presented any intra-procedural complication. After discharge, one patient was admitted for cardiac tamponade and another for stroke. In both, the device was optimally deployed. Follow-up TEE was done after a mean time of 7.9 ± 5.9 months. The presence of a leak was associated with a lower degree of device over-sizing by TEE (0.06 ± 3.31 mm if leak vs. 2.73 ± 1.85 mm if no leak; p = 0.04) and a trend towards by angiography (2.39 ± 3.06 mm if leak vs. 3.98 ± 1.57 mm if no leak; p = 0.16). The shape of the device was not associated with the presence of TEE leaks. Neither the shape nor the presence of leaks was associated with worse clinical outcomes. Conclusions: Although the shape and sizing of the device were not linked to any particular clinical outcome, we found an association between a lower degree of over-sizing and the presence of follow-up leaks. © 2012 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Percutaneous left atrial appendage (LAA) closure is slowly establishing in reference centers around the world. Currently, the procedure is mostly performed in patients with atrial fibrillation (AF) with a high risk of stroke (CHADS2 score ≥ 1) and a formal contraindication or at high risk for anticoagulation therapy [1–3]. The PLAATO system (ev3 Inc., Plymouth, Minnesota) was the first dedicated device that showed promising results preventing ischemic strokes at short [4] and long-term follow-up [5] although it is not available on the market anymore. PROTECT AF was the first randomized trial comparing warfarin versus percutaneous LAA closure in patients with AF. The study demonstrated the non-inferiority of the WATCHMAN device (Atritech, Plymouth, MN) for the prevention of ischemic stroke as compared with oral anticoagulation [6].
☆ Conflict of interest: Dr. Reda Ibrahim is consultant for St. Jude Medical and Gore Medical. ⁎ Corresponding author at: Intreventional Cardiology Program, Montreal Heart Institute, 5000, East Belanger Street, Montreal, Quebec, Canada H1T 1C8. Tel.: + 1 514 376 3330x3800. E-mail address:
[email protected] (R. Ibrahim). 0167-5273/$ – see front matter © 2012 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ijcard.2012.10.031
The Amplatzer Cardiac Plug (ACP) (St-Jude Medical, Minneapolis, MN) represents a new generation of LAA closure devices. The novel asymmetric design with a distal lobe and a proximal disk makes this device special and unique. In fact, the device conforms to the LAA anatomy adopting different shapes according to the degree of expansion and the anatomy of the appendage [1]. Although initial experience with the ACP has shown satisfactory results [1–3], data on the optimal shape or recommended size are still scarce. Additionally and in contrast with other devices [4,7], few data on the presence and impact of residual leaks have been published to date. Thus, we sought to analyze the role of the ACP shape and sizing after LAA closure on clinical outcomes and residual leaks. 2. Methods 2.1. Patient population Patients who underwent percutaneous LAA closure with the ACP between November 2009 and September 2011 at the Montreal Heart Institute, University of Montreal, Montreal, Quebec, Canada were prospectively included in the study. The clinical, procedural, angiographic and echocardiographic outcome variables of these patients were prospectively collected and analyzed. All patients were more than 18 years old, had AF (paroxysmal, persistent or permanent) with a high risk for stroke (CHADS2 score ≥ 2), and at least one contraindication for oral anticoagulation
1024
X. Freixa et al. / International Journal of Cardiology 168 (2013) 1023–1027
therapy. Exclusion criteria for LAA closure included: LAA thrombus, mobile aortic atheroma or symptomatic carotid artery disease. All patients provided written informed consent before the procedure. In addition, the authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology. 2.2. Amplatzer Cardiac Plug description The ACP is a self expanding dedicated device for LAA closure with a distal lobe and a proximal disk connected by an articulated waste. The distal lobe conforms to the inner LAA wall in a depth of approximately 10 mm, the articulated waste allows a proper orientation of the device into the LAA and the proximal disk seals the LAA ostium [1]. The ACP is retrievable and re-positionable and is implanted through the atrial septum from a femoral vein (9–13 French sheath). The procedure was performed under general anesthesia by a single operator, with simultaneous fluoroscopic and 2D-TEE guidance (iE33 ultrasound system and X7-2t matrix array transducer, Philips Healthcare, Andover, MA, USA). After crossing the atrial septum with the delivery system, via transeptal puncture or through a patent foramen oval (PFO) or atrial septal defect (ASD), all patients received intravenous heparin to keep an activating clotting time between 200 and 250 s throughout the entire procedure. The degree of sealing was assessed by a contrast injection through the delivery system placed in front of the occluder after implantation and classified as: 1) severe leak (complete filling of the LAA), 2) moderate leak (two-thirds filling), 3) mild leak (one-third filling), 4) trace leak (barely detectable filling) and 5) absent leak (no detectable filling) [4]. Successful ACP implantation was considered in case of a final mild, trace or absent angiographic leak. 2.3. Device size selection and shape analysis The size of the device was selected based on angiographic and TEE measurements. Several contrast injections of 8 to 10 mL in the RAO 30° Cranial 20° with a 5 French marked Pigtail were used for the angiographic assessment. For TEE, the LAA was assessed in the mid-esophageal and high-esophageal views from 0° to 180° (in particular at 0°, 45°, 75°, 90°, and 120°). The largest LAA diameter in a depth of 10 mm from the ostium measured in mid diastole of either angiography or TEE was used to choose the size of the device. The degree of device over-sizing chosen by the operator varied based on the anatomy of the LAA and the gradual learning curve. All devices were implanted by the same operator. For every patient, an upsizing ratio was calculated dividing the size of the implanted device by the largest LAA measurement in a depth of 10 mm with both TEE and angiography. After device implantation the shape of the distal lobe was evaluated by two independent reviewers and classified as “tire-like or optimal-deformation”, “strawberry-like or
over-deformation” and “square-like or under-deformation” [1]. Intra and inter-observer variability was 0%. 2.4. In hospital and mid-term follow-up Patient assessment and hospitalization records were reviewed for events prior to discharge or transfer to another medical facility. Clinical follow-up was performed by a programmed clinical visit supplemented by medical record review of any subsequent hospital admissions or additional visit. A questionnaire was used to collect information regarding the clinical status or the occurrence of cardiovascular events. 2.5. Transesophageal echocardiography assessment All procedural and follow-up TEE were analyzed by two experienced observers blinded to the clinical and procedural information and put in consensus in case of disagreement (AS and XF). Device leak was defined by the presence of a Doppler signal inside the LAA after ACP deployment. The presence of peri-device flow was classified as: 1) severe (multiple jets or free flow), 2) major (>3 mm jet), 3) moderate (1 to 3 mm jet), 4) minor (b1 mm jet) and 5) absent (no jet) [4,7] (Fig. 1). Successful LAA sealing was considered in case of moderate, minor, or absent leak [4]. Intra-observer and inter-observer variabilities were 0% and 3.7% respectively. The only disagreement between observers was in the classification of one leak that was considered minor for one observer and moderate for the other. After consensus between both reviewers, the patient was classified as having a “moderate leak”. 2.6. Statistical analysis The results are expressed as mean ± standard deviation (SD) for normally distributed data or as median and interquartile (IQR) range for non-normally distributed data. Comparisons between groups were performed using the unpaired t test or Mann–Whitney U test for continuous variables, and Chi-square or Fisher's exact test for categorical variables. A one-way ANOVA test with a Bonferoni correction was used to define statistical difference between groups when appropriate. Results were considered statistically significant at a p-value b0.05. Statistical analyses were carried out using SPSS package v16.0 (Chicago, IL, USA).
3. Results Within the study period, 25 consecutive patients underwent percutaneous LAA closure with the ACP. In all patients, the procedural indication was based on the inability to receive anticoagulation treatment in spite of a clear clinical indication. The main reason for
Fig. 1. Different examples of peri-device leaks assessed by Transesophageal Echocardiography. Footnote: A: Major leak with a Doppler jet of 4.1 mm. B: Moderate leak with a Doppler jet around 2.2 mm. C and D: Minor Leaks with a 0.8 mm and a b0.5 mm jets.
X. Freixa et al. / International Journal of Cardiology 168 (2013) 1023–1027
contraindicating the oral anticoagulation was intracranial bleeding in 9 (36%), extracranial but non gastrointestinal bleeding in 8 (32%), gastrointestinal bleeding in 6 (24%) and other reasons in 2 (8%). Baseline and echocardiographic characteristics of the study population are shown in Table 1. 3.1. Device implantation The ACP was successfully implanted in all patients. Nonetheless, in 5 (20%), the size of the first device was changed as the operator was not satisfied with the initial result. The selection of the ACP size and the implantation process were guided by angiography and TEE imaging as described in the Methods section. Overall, the device was oversized by 3.38± 2.26 mm and 1.58 ± 2.75 mm in relation to the largest angiographic and TEE diameter respectively. The 22 mm ACP was the most used size (36%). All patients assessed by angiographic contrast injection after device implantation achieved a successful sealing as they showed absent leak in 16 (64%), trace leak in 4 (16%) and mild leak in 4 (16%). Transesophageal echocardiography was done in all patients during and post ACP implantation. However, final valid color Doppler assessment was available in 24 (96%) revealing successful sealing in 22 (91.6%). As shown in Table 2, 11 (45.8%) patients presented no leak, 8 (33.3%) minor leak, 3 (12.5%) moderate leak and 2 (8.3%) major leak. The final shape of the device was optimal in 16 patients (64%), strawberry-like in 6 (24%) and square-like in 3 (12%). There was no significant association between the presence of immediate angiographic or TEE device leak and the degree of oversize or final device shape. Two patients underwent a percutaneous PFO and ASD closure respectively after LAA closure with no additional complications. 3.2. In-hospital outcomes None of the patients presented any intra-procedural air/device embolization, stroke or cardiac tamponade. Trans-thoracic echocardiography 24 h post-procedure ruled out the presence of any device embolization. Three patients presented mild or moderate pericardial effusion pre-procedure but TTE did not reveal any significant change 24 h after the procedure. Vascular access-related bleeding was observed in 3 patients (12%) and successfully controlled with manual compression and/or blood transfusion. All patients were discharged under 80 mg of aspirin indefinitely and 23 (92.5%) received dual antiplatelet therapy with aspirin indefinitely and 75 mg of clopidogrel for 3 months. 3.3. Clinical follow-up Clinical follow-up was available in all 25 (100%) patients. Mean follow-up duration was 15.9 ± 8.1 months. One patient (4%) was admitted with pulmonary embolism 2 weeks after the index procedure. The same patient underwent cardiac tamponade one week after the new admission and pericardial drainage of a non-hematic fluid was Table 1 Baseline characteristics. N = 25 Age (years) Male n (%) Hypertension n (%) Hypercholesterolemia n (%) Diabetes mellitus n (%) Smokers (%) CHADS2 Stroke within 6 months n (%) Coronary artery disease n (%) Previous CABG n (%) Left ventricle ejection fraction (%) Creatinine (mmol/L) CABG, coronary artery bypass graft.
76.64 ± 7.87 15 (60) 15 (60) 12 (48) 17 (68) 4 (16) 3.24 ± 1.09 6 (24) 13 (52) 5 (20) 57.21 ± 7.08 100.32 ± 24.91
1025
Table 2 Prevalence and severity of residual peri-device leaks post ACP implantation and during follow-up as assessed by transesophageal echocardiography.
Severe leaks n (%) Major leaks n (%) Moderate leaks n (%) Minor leaks n (%) Absent leaks n (%)
Post ACP implantation N= 24
Follow-up (7.9 ± 5.9 months) N= 19
0 (0%) 2 (8.3%) 3 (12.5%) 8 (33.3%) 11 (45.8%)
0 (0%) 2 (10.4%) 1 (5.2%) 6 (31.5%) 10 (52.6%)
undergone without complications. Interestingly, the patient presented a chronic moderate pericardial effusion before the LAA closure which did not vary at 24 h but progressed in the context of the pulmonary embolism. Although the pericardial effusion was non-hematic and therefore probably not related with the procedure, it is important to acknowledge that the shape of the ACP was optimal. At 15 month follow-up, one patient (4%) presented a lacunar stroke in a context of uncontrolled hypertension. Of note, the patient presented no residual leak post-implantation and the shape of the device was optimal. A follow-up TEE was however requested in order to evaluate the LAA sealing and to exclude a thrombus on the device but was refused by the patient. The same patient died 6 months later as a result of a colonic cancer. Finally, one more patient (4%) consulted for a mild skin infection at the level of the puncture site 4 months later which was successfully treated with oral antibiotics. No bleeding problems were reported despite dual antiplatelet therapy in all patients. 3.4. Transesophageal echocardiographic follow-up Transesophageal echocardiography was available in 19 (75%) patients during follow-up. The mean time duration between the index procedure and the control TEE was 7.9± 5.9 months. None of the patients presented device embolization or thrombosis. Successful LAA closure was observed in 17 (89.4%). As shown in Table 2, Doppler interrogation showed absent leak in 10 patients (52.6%), minor leak in 6 (31.5%), moderate leak in 1 (5.2%), and major leak in 2 (10.4%). None of the patients with residual leak presented adverse outcomes. Of note, the two patients with major leak after device deployment did not show significant leak at follow-up and the two patients with major leak at follow-up did not show significant leak after device deployment. As shown in Fig. 2, the presence of leak (minor, moderate or major) inside the LAA was associated with a lower degree of device oversize by TEE (0.06± 3.31 mm if leak vs. 2.73± 1.85 mm if no leak; p = 0.04) and a trend towards by angiography (2.39± 3.06 mm if leak vs. 3.98 ±1.57 mm if no leak; p = 0.16). The upsizing ratio for patients without leaks at follow-up was 1.13 by TEE and 1.20 by angiography. The shape of the device was not associated with the presence of leak. 4. Discusion The main findings of our study were 1) the observed low percentage of significant peri-device leaks at follow-up, 2) the association between a lower degree of ACP over-sizing and the presence of follow-up leaks and 3) the absence of association between the shape of the device and the occurrence of residual leaks. In our series, the percentage of successful LAA closure with the ACP was 91.6% post ACP implantation and 89.4% at follow-up. Nonetheless, 12.5% and 5.2% of patients presented a moderate leak and 33.3% and 31.5% a minor leak respectively. The percentage of successful closure and residual leaks was initially reported by Ostermayer et al. [4] with the PLATOO system. In that study, TEE Doppler color flow was assessed and graded on a five-point scale: multiple jets of free flow (grade 1, “severe leak”); >3 mm diameter jet (grade 2, “moderate leak”); 1 to 3 mm diameter jet (grade 3, “mild leak”); b 1 mm diameter jet (grade 4, “trace leak”); no jet (grade 5, “absent leak”). Successful closure was
1026
X. Freixa et al. / International Journal of Cardiology 168 (2013) 1023–1027
Fig. 2. Relationship between ACP over-sizing and the presence of follow-up leaks. Footnote: The presence of ACP leaks at follow-up was associated with a lower degree of device over-sizing compared to TEE measurements (left) and a trend towards compared to angiographic measurements (right). + Difference between the size of the device and the largest diameter measured by TEE in a depth of 10 mm from the ostium of the left atrial appendage. * Difference between the size of the device and the largest diameter measured by angiography in a depth of 10 mm from the ostium of the left atrial appendage. TEE, Transesophageal echocardiography.
defined if the leak was ≤3 mm and was found in 97.7% after the procedure and 100% at 1 month (grade 3 in 13.3%, grade 4 in 51.7% and grade 5 in 35.0%). Levels of closure were similar at 6 months. Interestingly, Park et al. [1] did not observe any case of significant peri-device leak with the ACP at 1 year although patients with non significant leaks were not reported. In PROTECT AF, successful closure of the LAA with the WATCHMAN device leading to warfarin cessation was considered if there was an absence of a significant Doppler jet of at least 5 mm [6]. However, in a recent paper [7], a more detailed analysis of the peri-device flow revealed an incidence of residual leaks of 40.9% by TEE at 45 days decreasing to 32.1% at 1 year. The majority of the leaks were moderate (1 to 3 mm) or major (>3 mm) representing 59.9% and 32.4% respectively of the leaks at 45 days. Of note, although the level of closure was lower compared to our series (67.9% in PROTECT AF vs. 89.4% in our series), it did not impact on future thromboembolic events [7]. One of the most relevant findings of the present study was the association between a lower degree of device over-sizing and the presence of follow-up leaks. In other words, patients with larger device over-sizing presented less risk of residual leak. The clinical impact of this finding is not clear since incomplete closure of the LAA has not been linked to a higher rate of stroke with the PLAATO [4] neither with the WATCHMAN system [7]. In our study, antiplatelet and anticoagulation treatment was not modified in patients with residual leaks as none of them presented free flow inside the LAA (severe leak). Incomplete sealing of the LAA has only been linked to a higher risk of cardio-embolic events in patients undergoing LAA surgical closure [8,9]. In our series, none of the patients with any LAA leak experienced a stroke. Nevertheless, the scarcity of data regarding this question and the intuitive thinking that complete sealing should be better than any degree of leak renders this information still relevant from the author's point of view. In this sense, giving any specific recommendation for ACP sizing is very delicate since the study sample was relatively small. However, based on our experience, a mean device over-sizing around 3 mm using TEE and 4 mm using angiography might be a reasonable option when possible. Another interesting finding of our study was the detection of a major leak at follow-up in two patients without significant leak after device deployment. Of note, there was no change in rhythm between baseline and follow-up TEE. In addition, two more patients presented a reduction of a major leak post-implantation to a minor leak during follow-up. Both phenomena were already observed with
the PLAATO system [4] highlighting the fact that surveillance with TEE might be an appropriate strategy after LAA closure. Percutaneous closure of the LAA using the ACP appears to be safe and is associated with a high procedural success rate. In our series, successful implantation of the device was achieved in all patients with no major in-hospital complications. In agreement with our series, the success implantation rate reported in other previous studies using the ACP was over 95% [1,2] whereas the rate of serious procedural complications ranged between 7 and 10% [1,2]. In other larger trials using other devices, in-hospital success and complications rate varied from 91 to 97% and 2.7 to 6.5% respectively [4,10]. The ACP appears to be effective in preventing strokes in non randomized trials. The annual rate of non-procedural related stroke ranges between 0% and 3.8% depending on the series [2,4–6,11,12]. Based on the CHADS2 score, the population of our study would present an annual risk of stroke close to 6.5% under aspirin treatment and between 2.6 and 3.4% under warfarin [13,14]. In fact, only one patient experienced a lacunar stroke during follow-up. The source of the stroke was however not clear since it happened after a hypertensive crisis. Interestingly, neither the degree of over-sizing nor the shape of the device was associated with a higher incidence of pericardial effusion highlighting the fact that slight over-sizing is probably a safe strategy with the ACP. In fact, the origin of procedure-related pericardial effusion is multi-factorial. Several mechanisms have been proposed including transeptal puncture, catheter or delivery system manipulation, device deployment or even LAA anatomy [10]. To our knowledge, the degree of device overexpansion has not yet been identified as a risk factor for pericardial effusion. The percentage of significant post-procedural pericardial effusion published in previous studies ranges [4,9,11,12] from 0 to 4.1%. In our study, only one patient with a pre-existent pericardial effusion required percutaneous drainage 3 weeks post-implant and 2 weeks after anticoagulation restoration (warfarin bridge with IV heparin) for pulmonary embolism. In this case, it was not clear if the effusion was related with the intervention. 5. Limitations The results must be interpreted with caution as this was an observational study with a small number of patients. Second, the small size of the sample led to a very low number of events that did not allow for a valid assessment of the clinical impact of residual leaks. Nonetheless, the main objective of our paper was helping operators to guide the sizing process in order to reduce the occurrence of follow-up leaks rather than showing the clinical impact of them. Third, follow-up TEE was not available in all patients as this was a cohort of frail and elderly patients in whom invasive follow-up was sometimes difficult. Fourth, although 3D TEE imaging might have been useful for LAA measurements, data was not available in most of the patients. Finally, the results presented in this study referred to the ACP and might not apply for other LAA closure devices. 6. Conclusions The present study shows the initial experience of our group with the first 25 ACP implantations. According to our results, there is an association between a lower degree of device over-sizing and the presence of residual leaks at follow-up. Thus, precise sizing of the LAA based on TEE and angiography seems to play an important role in helping operators to choose the optimal size of the device and therefore reduce the future risk of residual leaks. We did not find any association between the presence of residual leaks and the occurrence of clinical events at follow-up although these results must be interpreted with caution due to observational nature of the study and the small size of the sample. Future studies with larger series will be necessary to confirm our results and clarify the impact of residual leaks on clinical outcomes.
X. Freixa et al. / International Journal of Cardiology 168 (2013) 1023–1027
References [1] Park JW, Bethencourt A, Sievert H, et al. Left atrial appendage closure with Amplatzer Cardiac Plug in atrial fibrillation: initial European experience. Catheter Cardiovasc Interv 2011;77:700-6. [2] Lam YY, Yip GW, Yu CM, et al. Left atrial appendage closure with Amplatzer Cardiac Plug for stroke prevention in atrial fibrillation: initial Asia-Pacific experience. Catheter Cardiovasc Interv 2012;79(5):794-800. [3] Rodes-Cabau J, Champagne J, Bernier M. Transcatheter closure of the left atrial appendage: initial experience with the Amplatzer Cardiac Plug device. Catheter Cardiovasc Interv 2010;76:186-92. [4] Ostermayer SH, Reisman M, Kramer PH, et al. Percutaneous left atrial appendage transcatheter occlusion (PLAATO system) to prevent stroke in high-risk patients with non-rheumatic atrial fibrillation: results from the international multi-center feasibility trials. J Am Coll Cardiol 2005;46:9–14. [5] Block PC, Burstein S, Casale PN, et al. Percutaneous left atrial appendage occlusion for patients in atrial fibrillation suboptimal for warfarin therapy: 5-year results of the PLAATO (Percutaneous Left Atrial Appendage Transcatheter Occlusion) Study. JACC Cardiovasc Interv 2009;2:594-600. [6] 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;374:534-42. [7] Viles-Gonzalez JF, Kar S, Douglas P, et al. The clinical impact of incomplete left atrial appendage closure with the Watchman Device in patients with atrial fibrillation: a PROTECT AF (Percutaneous Closure of the Left Atrial Appendage Versus Warfarin
[8]
[9]
[10]
[11]
[12]
[13]
[14]
1027
Therapy for Prevention of Stroke in Patients With Atrial Fibrillation) substudy. J Am Coll Cardiol 2012;59:923-9. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunick PA, Kronzon I. Surgical left atrial appendage ligation is frequently incomplete: a transesophageal echocardiograhic study. J Am Coll Cardiol 2000;36:468-71. Garcia-Fernandez MA, Perez-David E, Quiles J, et al. Role of left atrial appendage obliteration in stroke reduction in patients with mitral valve prosthesis: a transesophageal echocardiographic study. J Am Coll Cardiol 2003;42:1253-8. Reddy VY, Holmes D, Doshi SK, Neuzil P, Kar S. Safety of percutaneous left atrial appendage closure: results from the Watchman Left Atrial Appendage System for Embolic Protection in Patients with AF (PROTECT AF) clinical trial and the Continued Access Registry. Circulation 2011;123:417-24. Sick PB, Schuler G, Hauptmann KE, et al. Initial worldwide experience with the WATCHMAN left atrial appendage system for stroke prevention in atrial fibrillation. J Am Coll Cardiol 2007;49:1490-5. Bayard YL, Omran H, Neuzil P, et al. PLAATO (Percutaneous Left Atrial Appendage Transcatheter Occlusion) for prevention of cardioembolic stroke in non-anticoagulation eligible atrial fibrillation patients: results from the European PLAATO study. EuroIntervention 2010;6:220-6. Rivaroxaban-once daily, oral, direct factor Xa inhibition compared with vitamin K antagonism for prevention of stroke and Embolism Trial in Atrial Fibrillation: rationale and design of the ROCKET AF study. Am Heart J 2010;159:340-7 [e341]. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med 2009;361:1139-51.
