Author + information
- Received March 1, 2012
- Revision received May 25, 2012
- Accepted June 12, 2012
- Published online July 9, 2013.
- Krzysztof Bartus, MD, PhD⁎,
- Frederick T. Han, MD†,
- Jacek Bednarek, MD, PhD‡,
- Jacek Myc, MD, PhD⁎,
- Boguslaw Kapelak, MD, PhD⁎,
- Jerzy Sadowski, MD, PhD⁎,
- Jacek Lelakowski, MD, PhD‡,
- Stanislaw Bartus, MD, PhD⁎,
- Steven J. Yakubov, MD§ and
- Randall J. Lee, MD, PhD†∥,¶,⁎ ()
- ⁎Department of Cardiovascular Surgery and Transplantology, Jagiellonian University, John Paul II Hospital in Krakow, Krakow, Poland
- †Department of Medicine, University of California San Francisco, San Francisco, California
- ‡Department of Electrocardiology, Jagiellonian University, John Paul II Hospital in Krakow, Krakow, Poland
- §Riverside Methodist Hospital, Columbus, Ohio
- ∥Cardiovascular Research Institute, University of California San Francisco, San Francisco, California
- ¶Institute for Regeneration Medicine, University of California San Francisco, San Francisco, California
- ↵⁎Reprint requests and correspondence:
Dr. Randall J. Lee, University of California–San Francisco, Box 1354, San Francisco, California 94143
Objectives The purpose of the study was to determine the efficacy and safety of left atrial appendage (LAA) closure via a percutaneous LAA ligation approach.
Background Embolic stroke is the most devastating consequence of atrial fibrillation. Exclusion of the LAA is believed to decrease the risk of embolic stroke.
Methods Eighty-nine patients with atrial fibrillation were enrolled to undergo percutaneous ligation of the LAA with the LARIAT device. The catheter-based LARIAT device consists of a snare with a pre-tied suture that is guided epicardially over the LAA. LAA closure was confirmed with transesophageal echocardiography (TEE) and contrast fluoroscopy immediately, then with TEE at 1 day, 30 days, 90 days, and 1 year post-LAA ligation.
Results Eighty-five (96%) of 89 patients underwent successful LAA ligation. Eighty-one of 85 patients had complete closure immediately. Three of 85 patients had a ≤2-mm residual LAA leak by TEE color Doppler evaluation. One of 85 patients had a ≤3-mm jet by TEE. There were no complications due to the device. There were 3 access-related complications (during pericardial access, n = 2; and transseptal catheterization, n = 1). Adverse events included severe pericarditis post-operatively (n = 2), late pericardial effusion (n = 1), unexplained sudden death (n = 2), and late strokes thought to be non-embolic (n = 2). At 1 month (81 of 85) and 3 months (77 of 81) post-ligation, 95% of the patients had complete LAA closure by TEE. Of the patients undergoing 1-year TEE (n = 65), there was 98% complete LAA closure, including the patients with previous leaks.
Conclusions LAA closure with the LARIAT device can be performed effectively with acceptably low access complications and periprocedural adverse events in this observational study.
Atrial fibrillation (AF) is the most common cardiac arrhythmia in the world, with an estimated prevalence of over 3 million people in the United States (1). AF carries a significantly increased risk of morbidity and mortality, with embolic stroke being the most severe manifestation. In patients with AF, there is a 5-fold increased incidence of embolic stroke (2). The risk of embolic stroke in the general population increases with age. Therefore, AF is one of the most important causes of embolic stroke in people over the age of 75 years (3).
Currently, oral anticoagulation therapy is the most effective available prophylactic approach in patients with AF at high risk of thromboembolic events (4,5). Unfortunately, only 50% to 60% of patients on warfarin are in the therapeutic range (6,7), whereas the overall withdrawal rate from warfarin therapy is 10% to 38% after 1 year (8,9). Additionally, contraindications to or refusal of warfarin therapy adds to the difficulty in providing effective long-term anticoagulation (10,11). Dabigatran, apixaban, and rivaroxaban are new oral anticoagulant agents that have demonstrated noninferiority and/or superiority to warfarin in embolic stroke prevention (8,12–14). However, these agents are still associated with potential bleeding complications, high cost, and unlike warfarin, there is no antagonist. As the risk of major bleeding events increases with age, potential contraindications to anticoagulation therapy increases (15).
Surgical exclusion of the left atrial appendage (LAA) has been advocated for patients intolerant to warfarin therapy and as a routine part of the maze procedure due to the LAA's propensity for thrombus formation in AF patients (16). Prophylactic exclusion of the LAA is also recommended during mitral valve surgery as a means of eliminating a potential source of thromboembolic events (16). However, successful LAA surgical exclusion can be variable, with incomplete closure failing to eliminate the risk of thromboembolic events (17–19). In addition, surgical LAA exclusion can be complicated by post-operative bleeding events, perioperative stroke, and LAA laceration with the application of traction to the appendage (20–22).
A percutaneous approach for exclusion of the LAA is desirable due to the invasive nature and morbidity associated with surgical exclusion (20,21). The Watchman trial demonstrated that an LAA occlusion device is noninferior to warfarin for preventing embolic stroke in AF patients; however, concerns in regard to device embolization, tamponade, and periprocedural stroke have delayed its approval by the U.S. Food and Drug Administration (23). Recently, a percutaneous approach using an epicardial suture has been developed for exclusion of the LAA (24–26). The procedure has been shown to be feasible in humans, but its long-term efficacy and safety remains unknown. In this study, we report the efficacy and 1-year clinical results of LAA closure via a percutaneous epicardial suture ligation approach.
Patients ages 35 to 81 years were identified and enrolled between December 2009 and December 2010 for closure of the LAA. The protocol was conducted with the approval of the Polish Ministry of Health and the ethics committee at John Paul II Hospital, Krakow, Poland.
Eligible patients met all of the following inclusion criteria: 1) age 18 years or older; 2) nonvalvular AF; 3) at least 1 risk factor of embolic stroke (CHADS2 ≥1); 4) a poor candidate or ineligible for warfarin therapy (e.g., labile international normalized ratio level, noncompliant, contraindicated) and/or a warfarin failure (i.e., transient ischemic attack or stroke while on warfarin therapy); and 5) a life expectancy of at least 1 year (Table 1).
Patients were excluded from the study if they met any of the following exclusion criteria: 1) history of pericarditis; 2) history of cardiac surgery; 3) pectus excavatum; 4) recent myocardial infarction within 3 months; 5) prior embolic event within the last 30 days; 6) New York Heart Association functional class IV heart failure symptoms; 7) left ventricular function <30%; and 8) history of thoracic radiation. Patients meeting the criteria for the study enrollment underwent a screening contrast cardiac computed tomography (CT) scan. Additional exclusion criteria based on LAA anatomy included: 1) a LAA width >40 mm; 2) a superiorly oriented LAA with the LAA apex directed behind the pulmonary trunk; 3) bilobed LAA or multilobed LAA in which lobes were oriented in different planes exceeding 40 mm; and 4) a posteriorly rotated heart.
Percutaneous suture ligation of LAA
The device (SentreHEART, Redwood City, California) for exclusion of the LAA consists of 3 components: 1) a 15-mm compliant occlusion balloon catheter (EndoCATH); 2) 0.025-inch and 0.035-inch magnet-tipped guidewires (FindrWIRZ); and 3) a 12-F suture delivery device (LARIAT) (24–26) (Fig. 1). The procedure involves 4 basic steps: 1) pericardial and transseptal access; 2) placement of the endocardial magnet-tipped guidewire in the apex of the LAA with balloon identification of the LAA os; 3) connection of the epicardial and endocardial magnet-tipped guidewires for stabilization of the LAA; and 4) snare capture of the LAA with closure confirmation and release of the pre-tied suture for LAA ligation.
Patients were prepped and draped with sterile preparation of the subxiphoid and bilateral groin regions. Once the patient was anesthetized, transesophageal echocardiography (TEE) was performed to rule out LAA thrombus. The 3-dimensional cardiac CT reconstruction was used to guide pericardial access (Fig. 2). Pericardial access using a 17-G epidural needle was performed as previously described (27). The epicardial puncture was performed with the goal of achieving access to the anterior surface of the heart. An anterior epicardial puncture facilitated an anterior and inferior approach with the LARIAT snare over the LAA apex to its base. Anterior-posterior and lateral fluoroscopic views were used to guide the needle for epicardial puncture. Once epicardial access was confirmed, a 0.35-inch guidewire was left in the pericardial space while a transseptal catheterization was performed. Five thousand units of heparin were administered intravenously once the transseptal catheterization was completed. An additional 2,000 U of heparin were administered intravenously every 45 min for prolonged procedures. An activated clotting time was not obtained due to the lack of available equipment. The transseptal sheath was flushed with heparinized saline every 5 min. An 8.5-F SL1 catheter (St. Jude Medical, St. Paul, Minnesota) was directed anteriorly in the left atrium toward the LAA. A left atriagram was performed in the right anterior oblique view to delineate the ostium and body of the LAA (Fig. 3A). The 15-mm occlusion balloon catheter (EndoCATH) was back-loaded with a magnet-tipped 0.025-inch endocardial guidewire (FindrWIRZ) and inserted to the end of the SL1 transseptal catheter. The endocardial magnet-tipped guidewire was advanced to the apex of the LAA under fluoroscopic guidance. To navigate the endocardial guidewire, a mild bend was placed at the end of the guidewire to allow steerability of the endocardial guidewire. The balloon catheter was then advanced over the wire into the LAA. An LAA angiogram (appendagram) was performed through the lumen of the balloon catheter to evaluate the endocardial guidewire position, with a distal LAA position being the favored location. The epicardial access site was then sequentially dilated over the guidewire for placement of the 14-F soft-tipped epicardial guide cannula (SentreHEART). The 0.035-inch epicardial magnet-tipped guidewire was placed through the epicardial sheath for the goal of achieving an end-to-end magnetic union with the endocardial guidewire (Fig. 3B). The attached magnet-tipped guidewires act as a controlled pathway for delivery of the LARIAT snare to the base of the LAA without the need for traction or grasping of the delicate LAA tissue (Fig. 3C). Once the LARIAT suture delivery device was positioned over the LAA, the endoCATH balloon was used to position the snare at the ostium of the LAA. TEE was used to verify the anatomic position of the endoCATH balloon at the ostium of the LAA (Fig. 4B). After confirmation of the balloon catheter at the LAA ostium, the snare was closed (Fig. 3D). A left atriagram was performed to confirm complete capture of the LAA and rule out the existence of a remnant trabeculated LAA (Fig. 3E). After verifying LAA capture, the pre-loaded suture was released from the snare and tightened to exclude the LAA. Tightening was completed using a suture-tightening device that eliminates operator variability during tightening (TenSURE, SentreHEART). The LARIAT snare was removed from the pericardial space. The suture was cut near the LAA with a suture cutter (SentreHEART) that was passed over the suture. A pigtail catheter was placed in the pericardial space for at least 6 h and more commonly overnight. The pericardial catheter was attached to low suction to assess for any periprocedural pericardial effusions. A transthoracic echocardiogram was performed to rule out a pericardial effusion before the pericardial catheter was removed.
TEE was performed at 1 day, 30 days, 90 days, and 1 year post-LAA ligation. Following the procedure, the patients were observed in the intensive care unit overnight. The patients were then sent to a telemetry floor for 2 days before discharge. Clinical follow-up by the investigators were via phone contact at 1 month, and outpatient appointments performed at 6 months and 1 year post-ligation. Additionally, patients were followed by their referring physicians within 1 month of the procedure. Patients with a contraindication to warfarin remained off warfarin. It was recommended that patients with a CHADS2 score of 2 or higher who could tolerate warfarin (i.e., noncompliant or labile international normalized ratio level) continue warfarin. Warfarin use in patients with a CHADS2 score of 1 was left to the discretion of the referring physician. For patients not on warfarin, aspirin therapy was recommended.
Normally distributed continuous variables are expressed as mean ± SD. Continuous variables that were not normally distributed are expressed as median (interquartile ranges [IQR]).
The initially screened population consisted of 119 patients satisfying inclusion criteria for LAA ligation. After evaluation of the pre-procedural CT scan, 16 patients were excluded due to LAA size (8 patients with LAA width >40 mm) or morphology (n = 8) contraindications (Fig. 5). The predominant LAA shape that led to exclusion was a superior-posterior orientation of the LAA apex, commonly with the apex behind the pulmonary artery. Of the remaining 103 patients, 11 patients were excluded due to an LAA thrombus on pre-procedural TEE. Of the 92 patients who proceeded to LAA ligation, 3 patients were excluded because of pericardial adhesions precluding pericardial access. These adhesions were noted to obstruct the passage of the epicardial guidewire in the pericardial space. Pericardial adhesions were confirmed with contrast injection into the pericardial space. The procedure was subsequently aborted for these 3 patients. Table 1 describes the baseline characteristics of the 89 patients for whom LAA ligation was attempted. Table 2 describes the thromboembolic and bleeding event risk characteristics of our cohort as reflected in the CHADS2, CHA2DS2-VASc (28), and HAS-BLED (29) scores.
Immediate LAA closure was achieved in 85 of 89 patients attempted (Table 3). Complete closure was achieved in 96% of successful LAA ligations. Complete closure was defined as a <1-mm jet by color flow Doppler. TEE performed 1 day, 30 days, and 90 days after the procedure revealed complete closure in 95% of patients in whom LAA ligation was attempted. Four patients had a 2-month TEE and refused the 3-month TEE follow-up. There was only 1 patient that had an increase in leakage 1 day after the procedure (Table 3). In 1 patient, color Doppler echocardiography increased from <1- to a <2-mm color Doppler jet, but remained stable at 1-month and 3-month follow-up. No patients had a >3-mm jet by color Doppler at 3 months. At the 3-month TEE, there was no color Doppler evidence of an increase in the size of the leak in any of the patients. The patients with a <2- or 3-mm residual channel at day 1, day 30, and day 90 were the same patients. Complete closure of the LAA was seen in 98% of patients undergoing 1-year follow-up TEE. This included all patients that had a 2- to 3-mm color Doppler leak noted at the 3-month TEE. One patient that had a <2-mm color Doppler leak at 3 months was found to still have a <2-mm color Doppler leak. Thrombus formation was not detected at the site of occlusion by TEE throughout the follow-up period. In 1 patient at the 1-year follow-up TEE, there was a thrombus in the left atrium at a site distant from the occlusion site. The patient had stopped her warfarin after the initial LAA ligation procedure and was not on aspirin. The patient was treated with warfarin to resolve the thrombus.
Immediate closure was not achieved in 4 patients. One patient had pericardial adhesions in the LAA sulcus precluding placement of the LARIAT device at the LAA base; the procedure was aborted in this patient. Another patient had his procedure aborted due to the inability to perform a transseptal puncture. In 2 patients, LAA ligation was aborted due to pericardial access–related complications. One patient had a right ventricular puncture (which was dilated over the guidewire, with consequent hemopericardium), requiring drainage and observation. The second epicardial access complication was a laceration of a superficial epigastric vessel leading to bleeding at the skin puncture site. This epigastric vessel laceration required cauterization to stop the bleeding. A third complication occurred during the transseptal catheterization, resulting in perforation and hemopericardium. Since epicardial access was in place, the hemopericardium was successfully drained. After observation to assess for additional hemopericardium, the LAA ligation was completed, and heparin was subsequently reversed with protamine sulfate. Significant pericardial effusion was defined as >100 ml of fluid aspirated during the procedure. There were no device-related complications.
Sixty-eight of 85 patients required only 1 pericardial access attempt, whereas 17 of 85 patients required 2 pericardial access attempts to obtain the proper orientation to deliver the LARIAT snare over the LAA (Fig. 6). The predominant reason for a second pericardial access attempt was the initial pericardial access being too posterior or too medial.
All LAAs had discrete necks with 1 to 4 lobes of multiple shapes and orientations. Successful connection of the endocardial and epicardial magnet-tipped guidewires was accomplished immediately in 68% of the cases on the first attempt; and within 1 min in 91% of the cases. The mean time to connect the magnet-tipped guidewires was 1.4 ± 0.64 min. Time to connect the magnet-tipped guidewires began with the insertion of epicardial guidewire into the guide sheath.
Only 1 case took more than 3 min to connect the magnet-tipped guidewires. TEE guidance was used to properly position the LARIAT snare in all patients. Proper positioning of the LARIAT snare to allow LAA capture was achieved in 68%, 94%, and 97% of the patients on the 1st attempt, 2nd attempt, and 3rd attempt, respectively. A maximum of 4 attempts at positioning the LARIAT snare was required in 3% of patients. There was no change in color flow Doppler within the left upper pulmonary vein during inflation of the balloon or closure of the snare, consistent with initial findings (22).
The median time to capture the LAA was 4.0 min (IQR: 2 to 8 min). Time to capture was defined as the time of introduction of LARIAT suture delivery device into the pericardial space to placement of the LARIAT over the LAA. An LA angiogram was performed after closure of the LARIAT snare and prior to release of the pre-tied suture to ensure that no lobe or remnant trabeculated LAA was present (Fig. 3E). LA angiogram after LAA ligation did not reveal any remnant trabeculated LAA or diverticulum (Fig. 3F). The median procedural time was 45 min (IQR: 36 to 55 min), whereas the mean fluoroscopy time was 13.6 ± 6.5 min.
The rhythm prior to and at the conclusion of the procedure was not significantly different (Table 3). Prior to the procedure, 57 (64%), 29 (33%), and 3 (3%) patients were in AF, atrial flutter, and sinus rhythm, respectively. At the conclusion of the procedure, 56 (63%), 30 (34%), and 2 (2%) patients were in AF, atrial flutter, and sinus rhythm, respectively. The post-procedure rhythm for 1 patient was not available after the patient's ligation was complicated by a right ventricular puncture with an aborted ligation.
Post-operatively, 20 of the 85 patients developed chest pain following the procedure. However, once the pigtail catheter was removed the day following the LAA ligation, the chest pain resolved in all but 2 of the patients. The diagnosis of pericarditis was made in these 2 patients. The patients were treated with a nonsteroidal anti-inflammatory drug with resolution of their pain in 2 to 3 days. One of the patients diagnosed with pericarditis had severe chest pain associated with ST-segment elevation in the precordial leads. The patient underwent coronary angiography and was found to have normal coronary anatomy. There were no significant pericardial effusions noted during the post-operative hospitalization. One patient developed a late pericardial effusion 2 weeks after the LAA ligation. The patient presented with shortness of breath. Echocardiography did not reveal tamponade physiology. Pericardiocentesis was performed with resolution of the patient's symptoms. Clinical follow-up at 6 months revealed 2 significant adverse events. One patient had a sudden cardiac death following a hospitalization unrelated to the procedure. The event occurred 3 months after the LAA ligation procedure. The patient presented with weakness. The patient was not in congestive heart failure. Subsequent work-up did not find any abnormal findings. The patient was discharged to home. No autopsy was performed following the patient's death. The second adverse event was a hemorrhagic stroke that occurred 6 months after the patient's LAA was ligated. The hemorrhagic stroke was documented by magnetic resonance imaging of the head and was presumably secondary to an aneurysm. Warfarin was stopped after the procedure, and the patient was in sinus rhythm. The patient recovered from the neurological event.
At the 1-year clinical follow-up, there were 2 additional adverse events. A second patient death occurred in a patient who had previously been diagnosed with significant bradycardia and who refused pacemaker implantation. Death in this patient was presumed to be an arrhythmic death and occurred 12 months after LAA ligation. The other patient had a lacunar stroke related to a hypertensive crisis 14 months after LAA ligation. The patient had stopped her antihypertensive medications and was in sinus rhythm. Fifty-five percent of patients were on warfarin therapy at the end of 1 year. There were no thromboembolic events during the follow-up period.
The percutaneous LAA ligation approach produced immediate complete closure of the LAA in over 95% of our patients, with the long-term follow-up suggesting the closure is permanent. To our knowledge, this study is the largest, prospective series of LAA suture closure (surgical or percutaneous) verified with TEE with follow-up to 1 year.
Surgical literature has raised questions regarding the reliability of suture ligation to completely exclude the LAA. Endocardial suture ligation is incomplete in 10% to 30% of patients, predisposing the patient to thromboembolic events (30–32). Potential reasons for incomplete closure include: 1) the procedure is performed when the heart is in a flaccid state during cardiopulmonary bypass; 2) the access for suturing can be awkward; and 3) LAA closure cannot be verified until the patient is off cardiopulmonary bypass (22,30–32). The morbidity of cardiopulmonary bypass with the potential for stroke, rebleeding, and respiratory failure (33) can be prohibitive and preclude revision of a suboptimal closure result.
Exclusion of the LAA with epicardial suture ligation or surgical staplers is possible without cardiopulmonary bypass (34). However, epicardial suture closure ranges from 23% to 100% (20,30–32,35–37), whereas reported LAA closure rates using surgical staplers range from 0% to 80% (20,35). Procedural and anatomical variables leading to incomplete epicardial suture ligation include operator variability with suture tightening, the presence of prosthetic mitral valves or annuloplasty rings, and the complex anatomy of the LAA ostium (38). The linear orientation of the surgical stapler may not be ideal for LAA exclusion if the desired outcome is closure at the LAA ostium. Studies have demonstrated a post-closure diverticulum of >1 cm is not unusual, with concern that this diverticulum may predispose to thrombus formation (20,22). Complications most commonly observed with the epicardial approach are tearing of the LAA (7% to 25%) leading to bleeding complications (20,31,34,36,39). The percutaneous approach used in this study avoids any grasping or traction of the LAA and avoids this risk.
The success rate in complete closure of the LAA using the LARIAT suture delivery device is likely multifactorial, with: 1) intraprocedural TEE using color flow Doppler to assess for an LA–LAA leak; 2) contrast fluoroscopy to rule out any remnant LAA; 3) the ability to open the snare and reposition the snare before tightening the pre-tied suture; and 4) the reduction in operator variability with use of the TenSURE knot-tightening device. Finally, the screening CT also provides valuable data regarding the LAA anatomy. The CT identified 16 of 119 (13.4%) patients whose LAA size, shape, and/or orientation excluded them from LAA ligation. In comparison, in the LAAOS (Left Atrial Appendage Occlusion Study), 21% of patients were excluded from surgical exclusion based on the surgeon's assessment of the LAA—due to size or inability to access (20). The approach described in this study is superior to existing surgical methods that define successful closure as no communication of flow and a <1-cm nontrabeculated stump (20,35). Compared with percutaneous implants that consider closure success as <2- to 3-mm residual flow, the approach described in this study is 100% effective versus implant effectiveness of 86% (23).
The safety profile of the percutaneous LAA ligation procedure is favorable compared with other LAA exclusion procedures (23,35–37). The predominant periprocedural adverse event associated with the percutaneous LAA ligation procedure was chest pain. However, the incidence of persistent pericarditis was 2.4%. This was treated with nonsteroidal anti-inflammatory drugs. The late pericardial effusion was most likely due to an inflammatory reaction similar to Dressler's syndrome.
The LARIAT snare device avoids potential complications anticipated with other nonsurgical approaches to exclude the appendage. Although initial experience with the Watchman device (Atritech, Plymouth, Minnesota) was complicated by device embolization, perforation, or erosion resulting in pericardial effusions, the Continued Access Protocol for the Watchman device had fewer complications with increased operator experience (23,24). Warfarin therapy is necessary to prevent thrombus formation as the implant surface endothelializes. Additionally, oral anticoagulation may be crucial for prevention of chronic thrombus formation and thromboembolic events in patients with a significant peri-device leak (41). Although the risk of device embolization is averted with epicardial LAA ligation, obtaining pericardial access for the LARIAT procedure is associated with its own inherent risks, including right ventricular puncture, epicardial vessel injury, sheath-related trauma, and post-procedure pericarditis (42,43). Patient selection (exclusion of patients with previous cardiac surgery and pectus excavatum) and use of pre-procedural cardiac CT scans may account for the lower incidence of major pericardial access complications (2.2%) in this study compared with published multicenter experience for epicardial VT ablation (7%) (43). Other percutaneous epicardial approaches for LAA closure require grasping or traction to snare the LAA (44,45). Potential complications include laceration of the LAA, trauma to the epicardial surface of the myocardium, and incomplete closure due to the inability to reposition the snare.
Prevention of thromboembolic events has become one of the cornerstones of AF treatment regardless of a rhythm or rate control strategy (46). At least 20% of all ischemic strokes and more than a third of ischemic strokes in the elderly are associated with AF (47–50). About 1 in 3 people with AF will experience a stroke in their lifetime (51–53). Despite the development and greater awareness of newer oral anticoagulation agents, there will always be patients with a contraindication to oral anticoagulation therapy. Long-term bleeding risks can be mitigated with successful suture exclusion of the LAA. Additionally, complete closure of the LAA may allow for safer anticoagulation strategies, such as apixaban, low-dose dabigatran and/or aspirin, which are associated with lower bleeding rates than high-dose dabigatran or warfarin (8,12,14).
The medical justification for the LAA closure procedure is based on the lack of alternative therapies for stroke prevention in patients with a contraindication to anticoagulation therapy. In patients with a higher risk for stroke (CHADS2 score >2), recommended guidelines state that patients should be treated with anticoagulation therapy if it can be tolerated, regardless of the successful maintenance of sinus rhythm (16,54). Novel algorithms, such as the CHA2DS2-VASc score, may successfully identify “low-risk” patients at risk of thromboembolic events (28); however, it does not assess a patient's risk of bleeding complications with anticoagulation therapy.
A prospective, randomized clinical trial would provide definitive proof that LAA exclusion would prevent thromboembolic events. However, there is suggestive surgical data that complete LAA ligation is effective in preventing embolic stroke (55). It is well known that post-operative AF is a highly independent risk factor for stroke, with rates between 2% and 7% in patients undergoing coronary bypass (56,57). As part of the maze procedure, the LAA is surgically ligated and excised. There was a large percentage of post-operative AF in these patients, but only a 0.7% rate of stroke (55). A follow-up study on 178 patients undergoing a Cox Maze III procedure with discontinuation of warfarin reported that none of the patients had a stroke 10 years after surgery (21). Additionally, the Watchman device for LAA exclusion compared with warfarin therapy demonstrated noninferiority of the occlusion device (23). Given the current data, percutaneous LAA ligation may be a complementary alternative to oral anticoagulation and the Watchman device for thromboembolic event risk reduction. However, no definitive statement can be made regarding stroke reduction without prospective studies. A noninferiority trial comparing the LARIAT device to oral anticoagulation and/or the Watchman device may provide atrial fibrillation patients with supportive data for LAA ligation as an alternative therapy for thromboembolic event risk reduction.
The study is a nonrandomized, single-center trial. Much of the experience was concentrated with several operators, thus potentially skewing the efficiency of performing the procedures. The assessment of LAA closure by TEE might be overestimated due to AF resulting in decreased inflow and outflow velocities in the LAA. This may result in the lack of detection of small communications between the LA and a sutured LAA by color flow or spectral Doppler flow assessment. However, the initial operators did not have experience with either pericardial access or transseptal catheterization at the onset of the study, thus demonstrating the ease of adoption of the procedure. Long-term results for thromboembolic event reduction are confounded by the high proportion (61%) of patients on warfarin at the time of last follow-up. Future prospective studies will require a systematic protocol for discontinuing anticoagulation by a pre-specified time point to accurately assess the long-term risk reduction for thromboembolic events following LAA ligation.
The percutaneous catheter-based LAA ligation procedure using the LARIAT suture delivery device is feasible and effective in humans and produces complete closure of the LAA verified by serial TEE. This observational study provides evidence of the reliability of LAA exclusion with acceptably low access complications and adverse events; enabling this percutaneous LAA ligation procedure to be used in future randomized clinical trials to determine whether LAA exclusion prevents thromboembolic events in patients with AF.40
Drs. Yakubov and Lee are consultants to SentreHEART, Inc., with equity in the company. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Bartus and Han contributed equally to the preparation of the manuscript.
- Abbreviations and Acronyms
- atrial fibrillation
- computed tomography
- interquartile range
- left atrial appendage
- transesophageal echocardiography
- Received March 1, 2012.
- Revision received May 25, 2012.
- Accepted June 12, 2012.
- 2012 American College of Cardiology Foundation
- Rosamond W.,
- Flegal K.,
- Furie K.,
- et al.
- Hart R.G.,
- Halperin J.L.
- Stafford R.S.,
- Singer D.E.
- Wallentin L.,
- Yusuf S.,
- Ezekowitz M.D.,
- et al.,
- RE-LY investigators
- Fuster V.,
- Rydén L.E.,
- Cannom D.S.,
- et al.
- Lee R.J.,
- Bartus K.,
- Yakubov S.J.
- Lip G.Y.,
- Frison L.,
- Halperin J.L.,
- Lane D.A.
- Katz E.S.,
- Tsiamtsiouris T.,
- Applebaum R.M.,
- Schwartzbard A.,
- Tunick P.A.,
- Kronzon I.
- Hannan E.L.,
- Wu C.,
- Smith C.R.,
- et al.
- Kanderian A.S.,
- Gillinov A.M.,
- Pettersson G.B.,
- Blackstone E.,
- Klein A.L.
- Reddy V.Y.,
- Holmes D.,
- Doshi S.K.,
- Neuzil P.,
- Kar S.
- Sacher F.,
- Roberts-Thomson K.,
- Maury P.,
- et al.
- Britton M.,
- Gustafsson C.
- Olsen T.S.,
- Skriver E.B.,
- Herning M.
- Calkins H.,
- Brugada J.,
- Packer D.L.,
- et al.
- Stamou S.C.,
- Hill P.C.,
- Dangas G.,
- et al.