Curr Cardiol Rep (2014) 16:448 DOI 10.1007/s11886-013-0448-1
ECHOCARDIOGRAPHY (RM LANG, SECTION EDITOR)
Left Atrial Appendage Exclusion for Prevention of Stroke in Atrial Fibrillation: Review of Minimally Invasive Approaches Joshua D. Moss
Published online: 10 January 2014 # Springer Science+Business Media New York 2014
Abstract Stroke prevention is of vital importance in the management of atrial fibrillation (AF), though the proven strategy of systemic anticoagulation for thromboembolic prophylaxis is underutilized for a variety of reasons. The left atrial appendage (LAA) has long been suspected as the principal source of arterial emboli, particularly in nonvalvular AF, and a variety of techniques for its exclusion from the circulation have been developed. This review highlights the history of the LAA as a target of intervention, and the parallel advances in three minimally invasive strategies for its exclusion: percutaneous occlusion of the LAA orifice from within the left atrium, closed-chest ligation via a percutaneous pericardial approach, and minimally invasive thoracoscopic surgery. While further study is necessary, available evidence suggests that effective LAA exclusion is becoming a viable alternative to anticoagulation for stroke prevention in nonvalvular AF. Keywords Atrial fibrillation . Stroke . Anticoagulation Left atrial appendage occlusion . Left atrial appendage ligation . Left atrial appendage exclusion . Percutaneous closure of left atrial appendage . WATCHMAN . Amplatzer Cardiac Plug . LARIAT suture delivery device
Introduction Atrial fibrillation (AF) is the most commonly encountered arrhythmia in clinical practice, affecting an estimated 2.3 million patients in North America [1]. In addition to This article is part of the Topical Collection on Echocardiography J. D. Moss (*) Section of Cardiology, Department of Internal Medicine, University of Chicago, 5758 S. Maryland Ave, MC 9024, Chicago, IL 60637, USA e-mail: [email protected]
hemodynamic impairment and uncomfortable or debilitating symptoms, AF is associated with a significantly increased risk of stroke – about six times that of patients in sinus rhythm [2]. It has been estimated that atrial fibrillation is responsible for 20 % of all strokes [3–5]. Large trials have convincingly demonstrated the efficacy of oral anticoagulation (OAC) with warfarin for both primary and secondary prevention of thromboembolism in nonvalvular AF [6–13]. Alternative OAC agents have also shown to be non-inferior [14, 15] or superior to warfarin [16] in this population. Simple, widely used risk models have thus been derived to predict thromboembolic risk and likelihood of a favorable risk-benefit profile with OAC, including the popular CHADS2 and CHA2DS2-VASc scores [17–20]. However, many patients with AF and moderate-to high risk of stroke are not treated with OAC despite its proven benefit, or they are unable or unwilling to use it when recommended [21–24]. In fact, several studies have shown a “risk-treatment paradox:” lower use of OAC in patients with higher stroke risk [24, 25]. While patient preference and lack of physician awareness of OAC benefits contribute to under-treatment, reasons that clinicians cite for not prescribing OAC generally align with negative predictors of warfarin use in retrospective studies: bleeding risk (especially gastrointestinal and intracranial hemorrhage) and typically ill-defined “fall risk” [21, 22, 26, 27]. Physicians and patients also cite difficulty and inconvenience of INR monitoring for warfarin, as well as drug and dietary interactions. The newer OAC agents have advantages with respect to convenience, and they generally carry similar or lower risks of major bleeding than warfarin [14–16]. However, renal dysfunction can preclude their safe use, and concerns exist about the speed and ease with which their effects can be reversed. There is thus a clear need for a safe and effective alternative to OAC for patients in whom it is contraindicated, poorly tolerated, or not prescribed or taken correctly.
448, Page 2 of 10
Surgical Exclusion of the Left Atrial Appendage as the Source of Cerebral Emboli The left atrial appendage (LAA) has long been suspected as a source of thrombus formation in AF, based in part on early studies of rheumatic mitral stenosis [28]. After the feasibility of LAA resection was demonstrated in dogs [29], the first two cases of surgical LAA resection for prophylaxis of recurrent arterial emboli in humans were reported in 1949 [30]. Unfortunately, one patient suffered peri-operative stroke, and the other died suddenly on post-operative day nine. Over time, various surgical techniques for suture ligation of the LAA evolved. An automatic stapling technique and a number of other closure devices have also been developed in an effort to improve upon the safety and efficacy of intraoperative LAA ligation and/or removal [31–34]. However, while exclusion of the LAA became a routine part of operative treatments for AF [35], it remained for many years an unproven strategy for preventing post-operative stroke after rheumatic mitral valve surgery [36]. The LAA was more convincingly implicated in nonvalvular AF by a widely cited review by Blackshear and Odell in 1996. The authors analyzed 23 studies—including more than 4800 patients with AF—in which the presence and location of left atrial thrombus were identified using operative findings, transesophageal echocardiography (TEE), and/or autopsy. In non-rheumatic AF, 91 % of identified left atrial thrombi were isolated to or originated in the LAA, compared with only 57 % of thrombi associated with rheumatic heart disease. These findings formed the basis of a hypothesis that LAA “obliteration” could play a role in stroke prophylaxis for nonvalvular AF. Subsequent studies examining the effect of surgical LAA obliteration on incidence of thromboembolism were primarily retrospective case series, with historical or non-randomized controls [37–46]. A comprehensive review of published results by Chatterjee et al. [47] concluded that the majority of these studies showed a positive effect of LAA closure on stroke prevention. However, closure success rates as assessed by TEE varied widely—from 17 % to 90 % with endocardial suturing, 23 % to 100 % with epicardial suturing, and 0 % to 80 % with stapler techniques—and residual stumps were common. In one randomized trial, the LAA Occlusion Study (LAAOS), the rate of LAA closure success was only 45 % in cases using a suturing technique and 72 % in cases using a stapler. Notably, residual Doppler flow was the more common mode of failure with suture ligation, whereas a residual LAA stump greater than 1 cm in length was more common when a stapler was used. Prophylactic LAA occlusion was not associated with a significant reduction in neurologic events in that cohort of 77 coronary artery bypass patients [44]. The American College of Cardiology/American Heart Association 2006 Guidelines for the Management of Patients
Curr Cardiol Rep (2014) 16:448
with Valvular Heart Disease, which recommend amputation of the LAA at the time of mitral valve surgery to reduce the likelihood of postoperative thromboembolism, cited one study as evidence. Garcia-Fernandez et al. retrospectively evaluated 205 patients with prior mitral valve replacement who were referred for TEE for a variety of indications. Over a mean follow-up of 69.4 months between surgery and TEE, 27 patients experienced an embolic event, most within one week prior to the TEE. While only 7.4 % of patients with an embolic event had undergone LAA ligation at the time of valve replacement, 31.5 % of patients without an embolic event had the prophylactic procedure. Absence of “effective” LAA ligation—no ligation attempted or incomplete ligation noted—was independently associated with a nearly 12-fold risk of an embolic event in multivariate analysis [39].
Minimally Invasive Techniques for Exclusion of the Left Atrial Appendage As both interest in and skill with percutaneous techniques for structural heart procedures have evolved, so have methods and devices for minimally invasive LAA exclusion. Two broad strategies have emerged, each with advantages and disadvantages: delivery of an endocardial “plug” to occlude the LAA from within the heart and prevent its communication with the left atrium, and external ligation of the LAA in a manner that approximates open surgical exclusion (Table 1). Endocardial Left Atrial Appendage “Plug” Technologies The ideal endocardial device for occlusion of the LAA would be safe and simple to implant, would fit accurately in a wide range of LAA anatomies, and could be removed and replaced if not well seated. Once implanted, the device would not migrate, dislodge, or embolize, erode into the pericardial space or nearby structures such as the pulmonary vein or circumflex artery, or interfere with atrial function or diastolic transmitral blood flow. Perhaps most importantly, the device would itself not be thrombogenic. A multitude of technologies have been designed and tested, with the most published data available for three: the PLAATO system, the WATCHMAN Left Atrial Appendage Closure Device, and the Amplatzer Cardiac Plug. PLAATO A percutaneous LAA transcatheter occlusion (PLAATO) system (ev3/Covidien, Plymouth, MN) was the first minimally invasive technique to be trialed clinically. The original device consisted of a self-expanding nickel-titanium (nitinol) cage, covered with an occlusive membrane of expanded polytetrafluoroethylene (ePTFE). The device was available
Curr Cardiol Rep (2014) 16:448
Page 3 of 10, 448
Table 1 Minimally invasive options for exclusion of the left atrial appendage Method
Major publications Regulatory approval
Endocardial occlusion PLAATO None WATCHMAN CE Mark Holmes, 2009 [48] Reddy, 2013 [50••] Reddy, 2011 [49•] Reddy, 2013 [51] Boston Scientific, 2013 [53]
Amplatzer (ACP)
Epicardial ligation: percutaneous LARIAT
No longer clinically available • Percutaneous • Multiple sizes • Only device with available long-term randomized data, suggesting non-inferiority to (and possibly superiority to) OAC
CE Mark
• Percutaneous • Multiple sizes
None CE Mark* CE Mark*
Minimal human data available Minimal human data available Minimal human data available
None
Minimal human data available
FDA 510(k), CE Mark
• Percutaneous • Suitable for wide-range of LAA sizes, without endovascular hardware • High rate of complete LAA exclusion (< 1 mm Doppler flow)
Park et al. [54] Urena et al. [55••]
WaveCrest Cheng, 2012 [56] Occlutech Kanthan, 2013 [57] Transcatheter Patch Toumanides, 2011 [58] LAmbre Device Lam Y-Y [59]
Major advantages
Bartus, 2011 [60] Bartus, 2013 [61••]
Potential disadvantages
• High complication rate in early experience, decreased in Continued Access Protocol [49•] • Therapeutic OACrecommended for ≥45 days post implant, though non-randomized ASAP study suggested low stroke rate w/ dual-antiplatelet therapy [51] • Residual peri-device flow common, though retrospective analysis did not suggest association with stroke risk [52•]. • High complication rate in early experience • Residual device flow common • Randomized comparison against OAC still in progress
• Patients with prior sternotomy contraindicated due to pericardial adhesions • Large LAA or unfavorable LAA position contraindicated • Frequent post-procedure pericarditis • No prospective data compared with oral anticoagulation
EGM-guided
None
No human data available
Friedman, 2009 [62] Bruce, 2011 [63] Epicardial ligation: thoracoscopic Thoracoscopic Blackshear, 2003 snare [41]
Thoracoscopic appendectomy
Ohtsuka, 2013 [64•]
FDA, CE Mark • Direct visualization • Surgeon prepared for conversion to open procedure
• Requires thoracoscopic ports and left lung deflation • Minimal human data available
FDA, CE Mark • Direct visualization • Complete LAA removal may minimize residual stump/recanalization
• Requires thoracoscopic ports and left lung deflation • Limited human data available
*CE Marked for other cardiac indications; not commercially available for LAA occlusion
in multiple sizes and could be deployed in the orifice of the LAA via a 12-French transseptal sheath. It could also be recaptured and repositioned after initial deployment. The first
published animal experience demonstrated that by 3 months post-implant, the atrial-facing surface of the device was covered with organized neointima that was continuous with the
448, Page 4 of 10
atrial walls [65]; initial use in humans was published in 2002 [66]. In early international multi-center feasibility trials, implantation was attempted in 111 patients with nonvalvular AF and contraindication to anticoagulation. Implantation was successful in 108 patients (97.3 %), with two major adverse events within the first 30 days, both in the same patient. Three other patients experienced hemopericardium requiring in-hospital pericardiocentesis. Two patients experienced stroke during average follow-up of 9.8 months [67]. Subsequent single-center outcomes studies showed rates of serious procedural complications ranging from 2.7 % to 5.0 %, with very low embolic stroke rates over follow-ups ranging from two to four years [68, 69]. Five-year results of a US multicenter observational study were published in 2009, and results of the multi-center European PLAATO study published in 2010. In the former, the annualized stroke/TIA rate was 3.8 %, and in the latter 2.3 %, compared to an anticipated rate of 6.6 % per year via the CHADS2 scoring method [70, 71]. Despite promising early, intermediate, and long-term results, the PLAATO program did not proceed to phase II or phase III trials. Follow-up in the European PLAATO study was halted prematurely for financial reasons, and the device is not available for clinical use. In 2007, Atritech, Inc. completed an acquisition of intellectual property from ev3 Endovascular Inc., manufacturer of the PLAATO device, to support development of the WATCHMAN LAA Closure Device. WATCHMAN The WATCHMAN device is similarly based on a selfexpanding nitinol frame structure with fixation barbs,
Curr Cardiol Rep (2014) 16:448
implanted via transseptal approach through a 12-French delivery sheath, though it is covered with a permeable polyester membrane rather than ePTFE (Fig. 1). It is available in multiple sizes ranging from 21 to 33 mm in diameter. In the PROTECT AF trial, 707 patients with nonvalvular AF and a CHADS2 score of at least 1 were randomized in a 2:1 ratio to percutaneous LAA occlusion or to conventional warfarin therapy. Patients undergoing WATCHMAN implantation received warfarin for a minimum of 45 days after the procedure to facilitate device endothelialization; warfarin was then replaced by several months of clopidogrel and lifelong aspirin if a TEE showed satisfactory LAA closure (
ECHOCARDIOGRAPHY (RM LANG, SECTION EDITOR)
Left Atrial Appendage Exclusion for Prevention of Stroke in Atrial Fibrillation: Review of Minimally Invasive Approaches Joshua D. Moss
Published online: 10 January 2014 # Springer Science+Business Media New York 2014
Abstract Stroke prevention is of vital importance in the management of atrial fibrillation (AF), though the proven strategy of systemic anticoagulation for thromboembolic prophylaxis is underutilized for a variety of reasons. The left atrial appendage (LAA) has long been suspected as the principal source of arterial emboli, particularly in nonvalvular AF, and a variety of techniques for its exclusion from the circulation have been developed. This review highlights the history of the LAA as a target of intervention, and the parallel advances in three minimally invasive strategies for its exclusion: percutaneous occlusion of the LAA orifice from within the left atrium, closed-chest ligation via a percutaneous pericardial approach, and minimally invasive thoracoscopic surgery. While further study is necessary, available evidence suggests that effective LAA exclusion is becoming a viable alternative to anticoagulation for stroke prevention in nonvalvular AF. Keywords Atrial fibrillation . Stroke . Anticoagulation Left atrial appendage occlusion . Left atrial appendage ligation . Left atrial appendage exclusion . Percutaneous closure of left atrial appendage . WATCHMAN . Amplatzer Cardiac Plug . LARIAT suture delivery device
Introduction Atrial fibrillation (AF) is the most commonly encountered arrhythmia in clinical practice, affecting an estimated 2.3 million patients in North America [1]. In addition to This article is part of the Topical Collection on Echocardiography J. D. Moss (*) Section of Cardiology, Department of Internal Medicine, University of Chicago, 5758 S. Maryland Ave, MC 9024, Chicago, IL 60637, USA e-mail: [email protected]
hemodynamic impairment and uncomfortable or debilitating symptoms, AF is associated with a significantly increased risk of stroke – about six times that of patients in sinus rhythm [2]. It has been estimated that atrial fibrillation is responsible for 20 % of all strokes [3–5]. Large trials have convincingly demonstrated the efficacy of oral anticoagulation (OAC) with warfarin for both primary and secondary prevention of thromboembolism in nonvalvular AF [6–13]. Alternative OAC agents have also shown to be non-inferior [14, 15] or superior to warfarin [16] in this population. Simple, widely used risk models have thus been derived to predict thromboembolic risk and likelihood of a favorable risk-benefit profile with OAC, including the popular CHADS2 and CHA2DS2-VASc scores [17–20]. However, many patients with AF and moderate-to high risk of stroke are not treated with OAC despite its proven benefit, or they are unable or unwilling to use it when recommended [21–24]. In fact, several studies have shown a “risk-treatment paradox:” lower use of OAC in patients with higher stroke risk [24, 25]. While patient preference and lack of physician awareness of OAC benefits contribute to under-treatment, reasons that clinicians cite for not prescribing OAC generally align with negative predictors of warfarin use in retrospective studies: bleeding risk (especially gastrointestinal and intracranial hemorrhage) and typically ill-defined “fall risk” [21, 22, 26, 27]. Physicians and patients also cite difficulty and inconvenience of INR monitoring for warfarin, as well as drug and dietary interactions. The newer OAC agents have advantages with respect to convenience, and they generally carry similar or lower risks of major bleeding than warfarin [14–16]. However, renal dysfunction can preclude their safe use, and concerns exist about the speed and ease with which their effects can be reversed. There is thus a clear need for a safe and effective alternative to OAC for patients in whom it is contraindicated, poorly tolerated, or not prescribed or taken correctly.
448, Page 2 of 10
Surgical Exclusion of the Left Atrial Appendage as the Source of Cerebral Emboli The left atrial appendage (LAA) has long been suspected as a source of thrombus formation in AF, based in part on early studies of rheumatic mitral stenosis [28]. After the feasibility of LAA resection was demonstrated in dogs [29], the first two cases of surgical LAA resection for prophylaxis of recurrent arterial emboli in humans were reported in 1949 [30]. Unfortunately, one patient suffered peri-operative stroke, and the other died suddenly on post-operative day nine. Over time, various surgical techniques for suture ligation of the LAA evolved. An automatic stapling technique and a number of other closure devices have also been developed in an effort to improve upon the safety and efficacy of intraoperative LAA ligation and/or removal [31–34]. However, while exclusion of the LAA became a routine part of operative treatments for AF [35], it remained for many years an unproven strategy for preventing post-operative stroke after rheumatic mitral valve surgery [36]. The LAA was more convincingly implicated in nonvalvular AF by a widely cited review by Blackshear and Odell in 1996. The authors analyzed 23 studies—including more than 4800 patients with AF—in which the presence and location of left atrial thrombus were identified using operative findings, transesophageal echocardiography (TEE), and/or autopsy. In non-rheumatic AF, 91 % of identified left atrial thrombi were isolated to or originated in the LAA, compared with only 57 % of thrombi associated with rheumatic heart disease. These findings formed the basis of a hypothesis that LAA “obliteration” could play a role in stroke prophylaxis for nonvalvular AF. Subsequent studies examining the effect of surgical LAA obliteration on incidence of thromboembolism were primarily retrospective case series, with historical or non-randomized controls [37–46]. A comprehensive review of published results by Chatterjee et al. [47] concluded that the majority of these studies showed a positive effect of LAA closure on stroke prevention. However, closure success rates as assessed by TEE varied widely—from 17 % to 90 % with endocardial suturing, 23 % to 100 % with epicardial suturing, and 0 % to 80 % with stapler techniques—and residual stumps were common. In one randomized trial, the LAA Occlusion Study (LAAOS), the rate of LAA closure success was only 45 % in cases using a suturing technique and 72 % in cases using a stapler. Notably, residual Doppler flow was the more common mode of failure with suture ligation, whereas a residual LAA stump greater than 1 cm in length was more common when a stapler was used. Prophylactic LAA occlusion was not associated with a significant reduction in neurologic events in that cohort of 77 coronary artery bypass patients [44]. The American College of Cardiology/American Heart Association 2006 Guidelines for the Management of Patients
Curr Cardiol Rep (2014) 16:448
with Valvular Heart Disease, which recommend amputation of the LAA at the time of mitral valve surgery to reduce the likelihood of postoperative thromboembolism, cited one study as evidence. Garcia-Fernandez et al. retrospectively evaluated 205 patients with prior mitral valve replacement who were referred for TEE for a variety of indications. Over a mean follow-up of 69.4 months between surgery and TEE, 27 patients experienced an embolic event, most within one week prior to the TEE. While only 7.4 % of patients with an embolic event had undergone LAA ligation at the time of valve replacement, 31.5 % of patients without an embolic event had the prophylactic procedure. Absence of “effective” LAA ligation—no ligation attempted or incomplete ligation noted—was independently associated with a nearly 12-fold risk of an embolic event in multivariate analysis [39].
Minimally Invasive Techniques for Exclusion of the Left Atrial Appendage As both interest in and skill with percutaneous techniques for structural heart procedures have evolved, so have methods and devices for minimally invasive LAA exclusion. Two broad strategies have emerged, each with advantages and disadvantages: delivery of an endocardial “plug” to occlude the LAA from within the heart and prevent its communication with the left atrium, and external ligation of the LAA in a manner that approximates open surgical exclusion (Table 1). Endocardial Left Atrial Appendage “Plug” Technologies The ideal endocardial device for occlusion of the LAA would be safe and simple to implant, would fit accurately in a wide range of LAA anatomies, and could be removed and replaced if not well seated. Once implanted, the device would not migrate, dislodge, or embolize, erode into the pericardial space or nearby structures such as the pulmonary vein or circumflex artery, or interfere with atrial function or diastolic transmitral blood flow. Perhaps most importantly, the device would itself not be thrombogenic. A multitude of technologies have been designed and tested, with the most published data available for three: the PLAATO system, the WATCHMAN Left Atrial Appendage Closure Device, and the Amplatzer Cardiac Plug. PLAATO A percutaneous LAA transcatheter occlusion (PLAATO) system (ev3/Covidien, Plymouth, MN) was the first minimally invasive technique to be trialed clinically. The original device consisted of a self-expanding nickel-titanium (nitinol) cage, covered with an occlusive membrane of expanded polytetrafluoroethylene (ePTFE). The device was available
Curr Cardiol Rep (2014) 16:448
Page 3 of 10, 448
Table 1 Minimally invasive options for exclusion of the left atrial appendage Method
Major publications Regulatory approval
Endocardial occlusion PLAATO None WATCHMAN CE Mark Holmes, 2009 [48] Reddy, 2013 [50••] Reddy, 2011 [49•] Reddy, 2013 [51] Boston Scientific, 2013 [53]
Amplatzer (ACP)
Epicardial ligation: percutaneous LARIAT
No longer clinically available • Percutaneous • Multiple sizes • Only device with available long-term randomized data, suggesting non-inferiority to (and possibly superiority to) OAC
CE Mark
• Percutaneous • Multiple sizes
None CE Mark* CE Mark*
Minimal human data available Minimal human data available Minimal human data available
None
Minimal human data available
FDA 510(k), CE Mark
• Percutaneous • Suitable for wide-range of LAA sizes, without endovascular hardware • High rate of complete LAA exclusion (< 1 mm Doppler flow)
Park et al. [54] Urena et al. [55••]
WaveCrest Cheng, 2012 [56] Occlutech Kanthan, 2013 [57] Transcatheter Patch Toumanides, 2011 [58] LAmbre Device Lam Y-Y [59]
Major advantages
Bartus, 2011 [60] Bartus, 2013 [61••]
Potential disadvantages
• High complication rate in early experience, decreased in Continued Access Protocol [49•] • Therapeutic OACrecommended for ≥45 days post implant, though non-randomized ASAP study suggested low stroke rate w/ dual-antiplatelet therapy [51] • Residual peri-device flow common, though retrospective analysis did not suggest association with stroke risk [52•]. • High complication rate in early experience • Residual device flow common • Randomized comparison against OAC still in progress
• Patients with prior sternotomy contraindicated due to pericardial adhesions • Large LAA or unfavorable LAA position contraindicated • Frequent post-procedure pericarditis • No prospective data compared with oral anticoagulation
EGM-guided
None
No human data available
Friedman, 2009 [62] Bruce, 2011 [63] Epicardial ligation: thoracoscopic Thoracoscopic Blackshear, 2003 snare [41]
Thoracoscopic appendectomy
Ohtsuka, 2013 [64•]
FDA, CE Mark • Direct visualization • Surgeon prepared for conversion to open procedure
• Requires thoracoscopic ports and left lung deflation • Minimal human data available
FDA, CE Mark • Direct visualization • Complete LAA removal may minimize residual stump/recanalization
• Requires thoracoscopic ports and left lung deflation • Limited human data available
*CE Marked for other cardiac indications; not commercially available for LAA occlusion
in multiple sizes and could be deployed in the orifice of the LAA via a 12-French transseptal sheath. It could also be recaptured and repositioned after initial deployment. The first
published animal experience demonstrated that by 3 months post-implant, the atrial-facing surface of the device was covered with organized neointima that was continuous with the
448, Page 4 of 10
atrial walls [65]; initial use in humans was published in 2002 [66]. In early international multi-center feasibility trials, implantation was attempted in 111 patients with nonvalvular AF and contraindication to anticoagulation. Implantation was successful in 108 patients (97.3 %), with two major adverse events within the first 30 days, both in the same patient. Three other patients experienced hemopericardium requiring in-hospital pericardiocentesis. Two patients experienced stroke during average follow-up of 9.8 months [67]. Subsequent single-center outcomes studies showed rates of serious procedural complications ranging from 2.7 % to 5.0 %, with very low embolic stroke rates over follow-ups ranging from two to four years [68, 69]. Five-year results of a US multicenter observational study were published in 2009, and results of the multi-center European PLAATO study published in 2010. In the former, the annualized stroke/TIA rate was 3.8 %, and in the latter 2.3 %, compared to an anticipated rate of 6.6 % per year via the CHADS2 scoring method [70, 71]. Despite promising early, intermediate, and long-term results, the PLAATO program did not proceed to phase II or phase III trials. Follow-up in the European PLAATO study was halted prematurely for financial reasons, and the device is not available for clinical use. In 2007, Atritech, Inc. completed an acquisition of intellectual property from ev3 Endovascular Inc., manufacturer of the PLAATO device, to support development of the WATCHMAN LAA Closure Device. WATCHMAN The WATCHMAN device is similarly based on a selfexpanding nitinol frame structure with fixation barbs,
Curr Cardiol Rep (2014) 16:448
implanted via transseptal approach through a 12-French delivery sheath, though it is covered with a permeable polyester membrane rather than ePTFE (Fig. 1). It is available in multiple sizes ranging from 21 to 33 mm in diameter. In the PROTECT AF trial, 707 patients with nonvalvular AF and a CHADS2 score of at least 1 were randomized in a 2:1 ratio to percutaneous LAA occlusion or to conventional warfarin therapy. Patients undergoing WATCHMAN implantation received warfarin for a minimum of 45 days after the procedure to facilitate device endothelialization; warfarin was then replaced by several months of clopidogrel and lifelong aspirin if a TEE showed satisfactory LAA closure (
