Author + information
- Received September 16, 2015
- Revision received December 18, 2015
- Accepted January 28, 2016
- Published online May 23, 2016.
- Carola Gianni, MDa,b,∗ (, )
- Luigi Di Biase, MD, PhDa,c,d,e,
- Chintan Trivedi, MD, MPHa,
- Sanghamitra Mohanty, MD, MSa,
- Yalçın Gökoğlan, MDa,
- Mahmut F. Güneş, MDa,
- Rong Bai, MDa,
- Amin Al-Ahmad, MD, CCDSa,
- J. David Burkhardt, MDa,
- Rodney P. Horton, MDa,d,
- Andrew K. Krumerman, MDc,
- Eugen C. Palma, MD, CCDSc,
- Miguel Valderrábano, MDf,
- Douglas Gibson, MDg,
- Matthew J. Price, MDg and
- Andrea Natale, MDa,c,d,g,h,i,j,k
- aTexas Cardiac Arrhythmia Institute, St. David’s Medical Center, Austin, Texas
- bDepartment of Clinical Sciences and Community Health, University of Milan, Milan, Italy
- cMontefiore Medical Center, Albert Einstein College of Medicine, Bronx, New York
- dDepartment of Biomedical Engineering, University of Texas, Austin, Texas
- eDepartment of Clinical and Experimental Medicine, University of Foggia, Foggia, Italy
- fDivision of Cardiac Electrophysiology, The Methodist Hospital, Houston, Texas
- gDivision of Cardiovascular Diseases, Scripps Clinic, La Jolla, California
- hMetroHealth Medical Center, Case Western Reserve University School of Medicine, Cleveland, Ohio
- iDivision of Cardiology, Stanford University, Stanford, California
- jElectrophysiology and Arrhythmia Services, California Pacific Medical Center, San Francisco, California
- kDell Medical School, University of Texas, Austin, Texas
- ↵∗Reprint requests and correspondence:
Dr. Carola Gianni, Texas Cardiac Arrhythmia Institute, St. David’s Medical Center, 3000 N. IH-35, Suite 720, Austin, Texas 78705.
Objectives The aim of this study was to evaluate the incidence and clinical implications of leaks (acute incomplete occlusion, early and late reopenings) following LAA ligation with the LARIAT device.
Background Percutaneous LAA ligation with the LARIAT device may represent an alternative for stroke prevention in high-risk patients with atrial fibrillation with contraindications to oral anticoagulation.
Methods This was a retrospective, multicenter study of 98 consecutive patients undergoing successful LAA ligation with the LARIAT device. Leaks were defined as the presence of flow as evaluated by transesophageal echocardiography (TEE). TEE was performed during the procedure, at 6 and 12 months, and after thromboembolic events.
Results Leaks were detected in 5 (5%), 14 (15%), and 19 (20%) patients at the 3 time points. During follow-up, 5 patients developed neurological events (4 strokes and 1 transient ischemic attack). Two occurred early (1 fatal stroke and 1 stroke with multiple recurrences in the following months), and TEE was not repeated after the events. The remaining 3 occurred late (after 6 months) and were associated with small leaks (<5 mm). In 2 of 3 cases, such a small leak was missed by the standard evaluation on 2-dimensional TEE, being evident only with the aid of 3-dimensional imaging.
Conclusions Incomplete occlusion of the LAA after LARIAT ligation is relatively common and may be associated with thromboembolic events. Proper long-term surveillance with careful TEE should be considered to detect leaks, which can be managed with either resumption of oral anticoagulation or percutaneous transcatheter closure.
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and a major cause of morbidity and mortality, mainly because it is an independent risk factor for stroke (1–4). Left atrial appendage (LAA) thromboembolism is the dominant cause of stroke in patients with AF (5,6). Oral anticoagulation (OAC) reduces thromboembolic events in selected patients, but it is associated with a significant risk for bleeding in high-risk patients (7,8). Therefore, several techniques to exclude the LAA have been developed over the years, from surgical ligation to excision through out transcatheter closure devices (5,9–14).
The LARIAT device (SentreHeart, Redwood, California) allows the percutaneous ligation of the LAA through the delivery of a suture via a combined endocardial and epicardial approach (15). As with surgical ligation, there may be residual or de novo leaks that potentially reduce the long-term efficacy of this procedure (16–19). We report our experience of the incidence and clinical implications of leaks following LAA ligation with the LARIAT device.
This was a retrospective, multicenter, observational cohort study of consecutive patients with AF who underwent successful transcatheter LAA ligation with the LARIAT device for stroke prevention at 4 centers across the United States. LAA ligation was indicated in patients at risk for stroke (high CHADS2 [cardiac failure, hypertension, age ≥75 years, diabetes mellitus, prior stroke or transient ischemic attack] or CHA2DS2-VASc [cardiac failure, hypertension, age ≥75 years, diabetes mellitus, stroke or transient ischemic attack or thromboembolism, vascular disease, age 65 to 74 years, female sex] score, previous LAA isolation) in whom OAC was contraindicated (history of significant bleeding), poorly tolerated, or failed (thromboembolic event despite anticoagulation). Patients were excluded if they had symptomatic carotid artery disease, aortic arch or descending aorta complex atheroma, prior open heart surgery, or unsuitable LAA anatomy (LAA ostial diameter >40 mm or LAA posterior to the pulmonary artery) (20). Moreover, patients were not included in the study if LARIAT deployment was attempted but the procedure was unsuccessful because of LAA thrombus (n = 2), pericardial adhesions (n = 5), or a residual leak ≥5 mm (n = 1).
Data were collected from patient records using structured case reports forms. Clinical events were site reported and not independently adjudicated. The study was approved by the hospitals’ institutional review boards.
The LARIAT procedure has previously been described in detail (20). In brief, patients undergo LARIAT ligation under general anesthesia. Pericardial and transseptal access are obtained with fluoroscopy and transesophageal echocardiographic or intracardiac echocardiographic guidance. Unfractionated heparin is administered to achieve a goal activated clotting time of 250 to 300 s. A magnet-tipped endocardial guidewire is advanced into the LAA, and a magnet-tipped epicardial guidewire is advanced into the pericardium and connected to the endocardial guidewire. The epicardial snare and suture are advanced over the wire and closed at the mouth of the LAA. Intraprocedural transesophageal echocardiography (TEE) and contrast-enhanced fluoroscopy are used to assess for incomplete closure.
Post-procedural medical therapy (including antithrombotic therapy) was prescribed according to the operator’s preference. All patients underwent clinical follow-up visits at 6 and 12 months. Transesophageal echocardiographic complete follow-up to assess for residual or new leaks was available in 96 of 98 patients (98%); adjunctive 3-dimensional (3D) TEE was available in 47 of 96 patients (49%) across all 3 time points and in 96 of 98 (98%) at the end of follow-up. An incomplete occlusion, or leak, was defined as evidence of flow into the LAA as assessed by TEE. It was considered small if <5 mm and large if ≥5 mm. A leak was defined as acute when it was detected at the time of the procedure, early when it was evident before or at 6-month follow-up (early reopening), and late when it was first evident after 6-month follow-up (late reopening).
Continuous variables are presented as mean ± SD and categorical variables as counts and percentages. Analyses were performed using Excel 2010 (Microsoft, Redmond, Washington) and GraphPad 6 (GraphPad Software, La Jolla, California).
A total of 98 consecutive patients undergoing successful LAA ligation with the LARIAT device were included. The mean age was 73 ± 8 years, and 35% of patients were women. The mean CHADS2 score was 2.7 ± 1.2, the mean CHA2DS2-VASc score was 4.1 ± 1.4, and the mean HAS-BLED (uncontrolled hypertension, abnormal liver or renal function, stroke, bleeding history or predisposition, labile international normalized ratio, elderly (age >65 years), drugs or alcohol) score was 3.2 ± 1.1 (Table 1). The rate of periprocedural complications (driven by major bleeding) was 9%; most of the complications occurred at 1 center.
Post-procedure, 85% of patients were discharged on antithrombotic therapy: 50% on antiplatelet therapy (aspirin in 18%, clopidogrel in 2%, and dual-antiplatelet therapy in 30%) and 35% on OAC (warfarin in 26%, rivaroxaban in 5%, and dabigatran in 4%). At 6-month follow-up, 21% of patients were on OAC and 44% on antiplatelet therapy, whereas at 12-month follow-up, 19% were on OAC and 48% on antiplatelet therapy. Patients were kept on OAC either because leaks were observed or because they underwent AF ablation procedures at follow-up.
At the time of TEE at the end of the procedure, 93 patients (95%) had no residual Doppler flow, whereas in 5 patients, small leaks (<5 mm) were detected (Figure 1). At 6-month follow-up, 3 of the 5 patients with acute small leaks showed no residual leaks, and in the remaining 2 patients, the leaks were stable. New leaks were present in 12 of 91 patients with no leaks at implantation: 8 had leaks <5 mm, and 4 developed leaks ≥5 mm (up to 1 cm in 1 case). At 12-month follow-up, in 12 of 12 patients with early reopening, the leaks persisted and were stable. New leaks were present in 5 of the 79 patients with neither acute leaks nor early reopening: 4 patients developed small leaks and 1 patient a leak ≥5 mm. In all patients with leaks ≥2.5 mm, this was successfully treated with percutaneous closure with an endovascular occlusion device. Left atrial thrombus at the site of LAA occlusion was detected in 3 patients (2 on aspirin and 1 without any antithrombotic therapy). OAC was resumed in all patients, and in 1 patient, the thrombus resolved, revealing a small leak (Figure 2).
During a mean follow-up period of 16.1 ± 2.5 months, 5 patients developed neurological thromboembolic events (4 strokes and 1 transient ischemic attack). Brain computed tomography or magnetic resonance imaging was done in all patients and showed no hemorrhages. All patients were not on OAC at the time of the clinical event. One patient experienced recurrent strokes starting 1 month after the procedure, and although intraprocedural TEE showed complete occlusion, he refused to undergo repeat transesophageal echocardiographic follow-up. One patient had a stroke after 2 months; it was associated with mesenteric ischemia, and although no acute leak was detected during the procedure, the patient died before repeat TEE could be done. In the remaining 3 patients, the neurological events were late, occurring after 6-month follow-up, and in all, small leaks were present. In 2 of 3 patients, the small leaks were missed by standard 2-dimensional (2D) TEE and were evident only on 3D TEE (Figure 3). Of those with transesophageal echocardiographic follow-up, 3 of 19 patients (16%) with leaks had thromboembolic events, compared with 0 of 77 (0%) with persistent LAA closure. In the 4 patients with neurological events who survived, OAC was resumed either long term (n = 3) or as a bridge to endovascular leak closure (n = 1) (Figure 4).
The main findings of this study are as follows: LAA leaks after the LARIAT procedure are common, they can develop at a later stage, they may be undetected with standard 2D TEE, and they can be associated with thromboembolic events.
Our study showed that 23% of patients had various degrees of LAA leaks after the LARIAT procedure, as demonstrated by serial TEE (Figure 1). Leaks can be evident at the time of the procedure (acute leaks; 5%), and these intraprocedural leaks can persist (2%) or close during follow-up (3%). Acute leaks are likely caused by suboptimal tightening at the time of the procedure, and they might close because of endothelialization or fibrosis filling the gap (21). Among patients with no acute leaks, new leaks were evident at follow-up (18% of the overall population) and could occur de novo even after 6 months from implantation (early reopening, 13%; late reopening, 5%). The mechanism of reopening after LARIAT LAA ligation could be related to knot slippage (the so-called gunny sack effect) due to suboptimal tightening or tissue necrosis, similarly to what happens in surgical ligation (22). Our study also shows that leaks that develop after implantation are less likely to spontaneously close during follow-up, similarly to what happens in late reopening that occurs in the Watchman population (21). Previous studies on LAA ligation with the LARIAT device have reported heterogeneous rates of leaks; whereas some centers have low rates of leak detection, others report rates as high as 24% (Table 2) (13,23–27). Compared with other studies that followed patients long term, we report a higher rate of leaks (20%) at follow-up. This can be explained partly by the use of 3D TEE, which allowed us to detect leaks in 3 additional patients with thromboembolic events, in whom the results of standard 2D TEE were negative (Figure 2). It has been shown that 3D TEE is more accurate than 2D TEE for the assessment of the LAA orifice, and it is comparable with 64-slice computed tomography (28,29). Two-dimensional images are not adequate to evaluate the geometry of noncircular structures, and the LAA orifice, especially in patients with AF, is highly eccentric (30); multiple views are needed for accurate measurements, and it is not possible to have an “en face” view. This leads to frequent misestimation of diameters, and it is possible to miss small reopenings, as was the case in our population. Accordingly, in the LARIAT study that reported the highest rate of leaks at follow-up (24%), LAA imaging was performed with either TEE or computed tomographic angiography (25).
Leaks remain an inherent complication of LAA occlusion procedures, and their clinical implications are still debated. Although leaks have been reported in as many as 59%, 47%, and 32% following PLAATO, Amplatzer Cardiac Plug, and Watchman device implantation, respectively, no study to date has proved that their presence is associated with thromboembolic events (31–33). By contrast, after surgical ligation, available studies point toward an association between leaks and an increase in thromboembolic events (16–18). Earlier studies that followed patients after LAA ligation with the LARIAT device show either no thromboembolic events or no evidence of association with leaks (Table 2). In a recent study comparing leaks during 1-year follow-up after the LARIAT and Watchman procedures, no association was found in both device groups. More specifically, 2 neurological events occurred in the LARIAT population (n = 259), 1 in a patient with a leak and 1 in a patient without. It is possible that no association could be found in previous studies because of underdetection of leaks with standard 2D TEE or because of the low event rate. In our population, 5 patients developed thromboembolic events during 1-year follow-up, all of them in the absence of OAC. In the 3 patients in whom TEE follow-up available, small leaks were detected, whereas no patient with persistent LAA closure had a thromboembolic event. Moreover, the 2 patients without TEE follow-up had clinical presentations (fatal stroke associated with mesenteric ischemia and recurrent strokes) compatible with cardioembolic events. Although causation is not certain, given the nature of the study and no direct evidence of thrombus in the LAA vicinity, there appears to be an association between leaks and thromboembolic events in our population. Although we cannot exclude that those events originated from other sites, it is more likely that the incompletely closed LAA played a central role. First, high-risk atherosclerosis (aortic arch or descending aorta complex atheroma or symptomatic carotid disease) was a contraindication to LARIAT implantation. Second, it has been shown that enhanced stasis due to reduced blood flow and/or electromechanical dysfunction promotes thrombus formation within the LAA (16,34), which we did not specifically look for; moreover, the thrombus could have dislodged at the time of the transesophageal echocardiographic evaluation. The presence of a leak provides a communication through which thrombi can pass and embolize. Interestingly, in the only patient with evidence of left atrial thrombus at follow-up, this was noted just above the ligation site, and after its resolution, a small leak was detected in its place.
There are no standard recommendations for post-procedural OAC after the LARIAT procedure. It is not easy to determine the need and duration of OAC, especially because patients who undergo LAA ligation are selected because of contraindications to or intolerance of these drugs. Our study suggests that it might not be safe to discontinue anticoagulation after the LARIAT procedure without proper surveillance with follow-up TEE, preferably 3D, and for at least 12 months after the procedure. By contrast, if a leak is detected, it can be managed by continuing or resuming OAC, if possible, or with a percutaneous procedure, either a repeat epicardial ligation with LARIAT or placement of an endovascular occlusion device (22,27,35).
The main limitations of this study were its retrospective observational nature and the absence of a control group for comparison. Moreover, 3D TEE was not performed in all patients across all 3 time points, and its results were not systematically compared with those of 2D TEE; a prospective study blindly comparing 2D and 3D TEE for the assessment of leaks can clarify the role of 3D TEE in this population.
Incomplete occlusion of the LAA after ligation with the LARIAT device is relatively common. Leaks can occur de novo even after 6 months after the procedure and appear to be associated with an increased risk for thromboembolic events. Although prospective studies are needed to confirm the present findings, this observed association is potentially clinically significant. Thus, in the LARIAT population, long-term surveillance with careful TEE (preferably 3D) should be considered, and when detected, leaks should be addressed with either resumption of OAC, if possible, or percutaneous closure with endovascular occlusion devices.
WHAT IS KNOWN? LAA closure with the LARIAT device may be an alternative to long-term OAC for patients with AF.
WHAT IS NEW? Reopenings (leaks) are common and appear not to be benign. Moreover, they might be undetected with standard 2D TEE.
WHAT IS NEXT? Prospective clinical trials are needed to assess the efficacy of the LARIAT device compared with long-term OAC or other LAA closure techniques.
Dr. Di Biase has received honoraria from Biosense Webster, Boston Scientific, Biotronik, Medtronic, Stereotaxis, and St. Jude Medical. Dr. Al-Ahmad has a received honoraria from Medtronic. Dr. Burkhardt has received honoraria from Biosense Webster, Stereotaxis, and St. Jude Medical. Dr. Horton has received honoraria from Boston Scientific and St. Jude Medical. Dr. Valderrábano has received honoraria and research funding from Boston Scientific, Hansen Medical, Medtronic, St. Jude Medical; honoraria from Biosense Webster, SentreHeart. Dr. Gibson has received honoraria from Boston Scientific and SentreHeart. Dr. Price has received honoraria from Boston Scientific, St. Jude Medical, and W.L. Gore & Associates. Dr. Natale has received honoraria from Biosense Webster, Medtronic, and St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atrial fibrillation
- left atrial appendage
- oral anticoagulation
- transesophageal echocardiography
- Received September 16, 2015.
- Revision received December 18, 2015.
- Accepted January 28, 2016.
- American College of Cardiology Foundation
- Wolf P.,
- Abbott R.,
- Kannel W.
- Miyasaka Y.,
- Barnes M.,
- Gersh B.,
- et al.
- Johnson W.D.,
- Ganjoo A.K.,
- Stone C.D.,
- Srivyas R.C.,
- Howard M.
- Hughes M.,
- Lip G.Y.H.
- Ostermayer S.H.,
- Reisman M.,
- Kramer P.H.,
- et al.
- Sick P.,
- Schuler G.,
- Hauptmann K.,
- et al.
- Bartus K.,
- Han F.T.,
- Bednarek J.,
- et al.
- Lee R.J.,
- Bartus K.,
- Yakubov S.J.
- Katz E.S.,
- Tsiamtsiouris T.,
- Applebaum R.M.,
- Schwartzbard A.,
- Tunick P.A.,
- Kronzon I.
- Kanderian A.S.,
- Gillinov A.,
- Pettersson G.B.,
- Blackstone E.,
- Klein A.L.
- García-Fernández M.A.,
- Pérez-David E.,
- Quiles J.,
- et al.
- Price M.J.,
- Gibson D.N.,
- Yakubov S.J.,
- et al.
- Nucifora G.,
- Faletra F.F.,
- Regoli F.,
- et al.
- Viles-Gonzalez J.F.,
- Kar S.,
- Douglas P.,
- et al.