Author + information
- Received December 27, 2016
- Revision received March 16, 2017
- Accepted March 28, 2017
- Published online December 4, 2017.
- Christopher R. Ellis, MDa,∗ (, )
- Sam G. Aznaurov, MDa,
- Neel J. Patel, MDa,
- Jennifer R. Williams, MDa,
- Kim Lori Sandler, MDa,
- Steven J. Hoff, MDb,
- Stephen K. Ball, MDa,
- S. Patrick Whalen, MDc and
- John Jeffrey Carr, MDa
- aVanderbilt University Medical Center, Nashville, Tennessee
- bOrlando Health, Cardiovascular and Thoracic, Surgery, Orlando, Florida
- cWake Forest Baptist Health, Winston-Salem, North Carolina
- ↵∗Address for correspondence:
Dr. Christopher R. Ellis, Vanderbilt University Medical Center, Cardiac Electrophysiology Laboratory, LAA Closure Program, Medical Center East, South Tower, Suite 5209, 1211 Medical Center Drive, Nashville, Tennessee 37232-8802.
Objectives This study sought to assess long-term left atrial appendage (LAA) closure efficacy of the Atriclip applied via totally thoracoscopic (TT) approach with computed tomographic angiography.
Background LAA closure is associated with a low risk for atrial fibrillation–related embolic stroke. The Atriclip exclusion device allows epicardial LAA closure, avoiding the need for post-operative oral anticoagulation. Previous data with Atriclip during open chest procedures show a high efficacy rate of closure >95%.
Methods Three-dimensional volumetric 2-phase computed tomographic angiography ≥90 days post-implantation was independently assessed by chest radiology for complete LAA closure on all consented subjects identified retrospectively as having had a TT-placed Atriclip at Vanderbilt University Medical Center from June 13, 2011, to October 6, 2015.
Results Complete LAA closure (defined by complete exclusion of the LAA with no exposed trabeculations, and clip within 1 cm from the left circumflex artery) was found in 61 of 65 subjects (93.9%). Four cases had incomplete closure (6.2%). Two clips were placed too distally, leaving a large stump with exposed trabeculae. Two clips failed to address a secondary LAA lobe. No major complications were associated with TT placement of the Atriclip. Follow-up over 183 patient-years revealed 1 stroke in a patient with complete LAA closure and no thrombus (hypertensive cerebrovascular accident).
Conclusions Angiographic LAA closure efficacy with a TT-placed Atriclip is high (93.9%). The clinical significance of a remnant stump is unknown. Confirmation of complete LAA occlusion should be made before cessation of systemic anticoagulation.
Autopsy studies have implicated the left atrial appendage (LAA) as the source of thrombus in >90% of atrial fibrillation (AF)-related embolic stroke. Observation of a lower than expected stroke rate after open chest surgery with concomitant LAA ligation formed the scientific rationale that obliteration of the LAA could potentially reduce AF-related embolic stroke (1). In 2015, the Watchman LAA closure device (Boston Scientific, St. Paul, Minnesota) was approved for stroke risk reduction in patients with nonvalvular AF who can tolerate short-term warfarin, but who are deemed unsuitable for long-term oral anticoagulation (OAC) treatment (2). Complete LAA closure seems essential to provide stroke protection, surgical data point to an increased risk for embolic stroke in the setting of an incomplete LAA closure, and leaks with the Lariat device (SentreHeart, Redwood City, California) have also been implicated in late onset of AF-related embolic stroke (3–5). It is unclear whether small leaks of contrast on computed tomography (CT) or transesophageal echocardiography (TEE) with the Watchman or the Amplatz Cardiac Plug device (St. Jude Medical, St. Paul, Minnesota) also confer an increased embolic risk (6–9).
The Atriclip and Atriclip PRO (Atricure, West Chester, Ohio) LAA occlusion devices have been evaluated for LAA closure at the time of concomitant open heart surgery under direct visualization, and high closure success was demonstrated in the EXCLUDE (Exclusion of the Left Atrial Appendage with a Novel Device: Early Results of a Multicenter Trial) by TEE or computed tomographic angiography (CTA) at 90-day follow-up (10). The Atriclip can be placed by a totally thoracoscopic (TT) approach as a stand-alone off-label therapy for LAA closure in absolute OAC contraindicated patients or in a hybrid endocardial-epicardial catheter ablation strategy mimicking the Cox-Maze IV to treat long-standing persistent AF (11). However, LAA closure efficacy by placement of the Atriclip device via a TT approach has not previously been published nor formally evaluated. We sought to confirm closure efficacy by CT angiography in long-term follow-up on patients who underwent a TT-placed Atriclip or Atriclip PRO at our institution.
The study protocol was designed as a single-arm CTA assessment of LAA closure in patients who had previously had placement of an Atriclip LAA occlusion device through a completely thoracoscopic approach. The protocol was approved by Vanderbilt University’s Institutional Review Board and all subjects provided informed consent. All subjects undergoing LAA closure at Vanderbilt University Medical Center available for consent agreed to participate in the VaLAAR (Vanderbilt LAA Registry), which is inclusive of all LAA closure technologies. TT-placed Atriclip subjects whose device was implanted at Vanderbilt University Medical Center (operative placement of the Atriclip or Atriclip PRO between June 13, 2011, and October 6, 2015) were eligible. Eighty-one patients were identified and available for long-term follow-up at Vanderbilt Medical Center, and 65 subjects consented to participate in the formal CTA study protocol. A dedicated 3-dimensional (3D) CTA for LAA closure assessment was then performed prospectively on those 65 subjects alone between November 11, 2015, and June 29, 2016 (Figure 1).
CTA was performed ≥90 days post-implantation. Patients with iodine contrast allergy were pre-treated to avoid anaphylactoid reactions in accordance with Vanderbilt Department of Radiologic Sciences protocol. Then 3D volumetric 2-phase CTA was performed to assess the presence of any contrast leak or exposed LAA trabeculations, to define the characteristics of a residual stump, and to objectively assess clip position in relation to the left circumflex coronary artery. CTA interpretation and data were independently obtained and reviewed by Vanderbilt University Department of Radiologic Science chest radiologists. LAA ostial location was assigned not by any single fixed anatomic landmark, but rather by use of multiplanar reformats (coronal, sagittal, 3D) simultaneously to best identify the epicardial LAA to LA body junction. All measurements were made in the axial plane. Successful LAA closure was defined as complete exclusion of the entire trabeculated LAA. Patient demographics, baseline medication use, and outcomes regarding stroke and AF were obtained through direct patient interview at the time of CTA and review of the electronic medical record (StarPanel, Vanderbilt University Department of Informatics, Nashville, Tennessee). The decision to continue or withhold OAC treatment after review of the CTA result was at the discretion of the managing cardiac electrophysiologist. Baseline CTA, cardiac magnetic resonance imaging (MRI), or pre-procedure and intraoperative TEE were used to categorize LAA anatomy before clip placement.
Surgical description of placement of the Atriclip/Atriclip PRO
Total thoracoscopic unilateral or bilateral access was performed in a hybrid electrophysiology lab operating room with single lung ventilation via a dual lumen endotracheal tube. Dissection and direct visualization of the pulmonary veins and LAA were accomplished in sequential fashion (in setting of hybrid AF ablation, right thorax followed by left-side approach). Epicardial ablation was performed in cases undergoing hybrid AF ablation before LAA closure. LAA was occluded epicardially with a specially designed atrial appendage clip and applicator (Atriclip, Atriclip PRO). Complete LAA occlusion was confirmed using intraoperative TEE and direct visualization at the completion of surgery in all subjects, epicardial landmarks for proper closure included the visible LAA neck to LA interface and location of the ligament of Marshall, LAA ridge to left-sided pulmonary veins, and left circumflex coronary artery. In all subjects, intraoperative TEE was used to confirm closure of the LAA to the satisfaction of the implanting cardiac surgeon and cardiac anesthesiologist.
This prospective CT angiographic LAA closure study was designed, overseen, and analyzed by the authors as an investigator-initiated study. The protocol was approved by and reviewed by the radiation safety committee of Vanderbilt University Institutional Review Board and Department of Radiology and Radiologic Sciences.
Patients who had previously undergone a TT-placed Atriclip device and were followed at Vanderbilt University Medical Center from June 13, 2011, to October 6, 2015, were identified (n = 81). All 65 subjects who consented for the study completed a dedicated 3D CTA assessment for efficacy of the Atriclip for LAA closure. Demographics of the study cohort are listed in Table 1. There were 50 male subjects (76.9%) representing a significant sex bias as seen previously in patients referred for AF ablation. Average LA diameter was 5.6 ± 7.5 mm, with LA volume 140 ± 46.2 ml representing a cohort with high prevalence of chronic or long-standing AF. The mean CHADS2-VASc score of the cohort was 2.48 ± 1.54, accounted for by type I or II diabetes (n = 15), essential or malignant hypertension (n = 51), systolic or diastolic congestive heart failure (n = 8), prior stroke or transient ischemic attack (n = 7), age ≥65 years (n = 40), age ≥75 years (n = 6), vascular disease (n = 9), and female (n = 15). Mean HAS-BLED (Hypertension, Abnormal renal and liver function, Stroke, Bleeding, Labile international normalized ratio, Elderly, prior Drug or alcohol usage history) score was 2.22 ± 1.24. Mean duration from Atriclip placement to CTA assessment was 1,029.4 ± 375.0 days, representing a chronic stable Atriclip position. Mean duration of follow-up from TT Atriclip placement to study completion was 3.48 ± 1.03 years.
Simultaneous or staged hybrid endocardial/epicardial ablation for persistent AF was the indication for initial Atriclip placement in 56 subjects (86.2%). Patients undergoing same-day hybrid AF ablation remained on OAC for 90 days post-procedure, and staged hybrid patients remained on OAC for an additional 90 days after endocardial catheter ablation. Adequate LAA closure (defined by complete exclusion of the LAA with no exposed trabeculations, and clip placement within <1 cm from the left circumflex artery) was found by CTA in 61 of 65 subjects (93.9%). Closure confirmation and detailed radiologic assessment of the clip position and atrial volume are detailed in Table 2. A smooth-walled residual LAA stump was noted in 35 subjects (54.7%) with mean dimensions of 2.1-cm width by 1.4-cm depth. Only 2 subjects had a stump with any visible trabeculations. There were no identified thrombi in any subject regardless of stump presence. Representative cases of complete LAA closure and no stump, or smooth-walled <2-cm stump are shown in Figure 2. Typical orientation of the Atriclip in relation to the left circumflex artery is shown in Figure 3.
There were 4 cases with incomplete LAA closure (6.15%). In 2 cases, the clip was placed too distally, leaving a large stump with exposed trabeculated segments (Figure 4), 2 clips failed to address a secondary or posteriorly rotated LAA lobe (Figure 5). In patients undergoing hybrid staged ablation, electroanatomic and voltage mapping of the residual stump demonstrated atrial electrograms present in the stump; however, spontaneous ectopy or AF induction was not seen. Standard intraprocedural testing included adenosine (12- to 24-mg bolus) and isuprel challenge (5- to 20-μg/min infusion when tolerated hemodynamically). Endocardial ablation at the LAA stump was not part of the index lesion set in the patients undergoing hybrid ablation.
Of the remaining 16 subjects who did not undergo the dedicated 3D CTA protocol, 12 had consented for VaLAAR, but they did not consent to the additional CTA to assess closure. All 12 had complete LAA closure by previously available and clinically indicated TEE (n = 5), cardiac MRI (n = 2), or CT chest with contrast (n = 5) performed outside of the formal CTA study protocol, but at least 90 days from TT Atriclip implantation. An additional 4 subjects did not provide consent for evaluation. Review of intraprocedural TEE data and operative notes state all subjects had complete closure by echocardiography, at least to the satisfaction of the cardiac anesthesia and cardiac surgical teams at time of implantation. This was reviewed and confirmed, although clip positon and LAA closure was not formally imaged at 0°, 45°, 90°, and 135° across the LAA ostium and body as is now standard during percutaneous LAA closure.
Baseline LAA anatomy was verified by CTA, cardiac MRI, and TEE data evaluated before placement of the Atriclip. There were 17 windsock (single dominant lobe of sufficient length), 12 cauliflower/broccoli (complex internal characteristics), 11 anterior rotated chicken wing, 13 posterior rotated chicken wing (behind the pulmonary artery), and 12 superior-C loop LAA morphologies. Incomplete closure occurred in 2 superior-C loop, 1 posterior rotated chicken wing, and 1 windsock morphologies. The posterior rotated chicken wing case was a stand-alone TT Atriclip placement. Baseline anatomy of the LAA did not predict failure to completely occlude the LAA with Atriclip.
Complications of TT placement of the Atriclip device
Placement of the Atriclip appeared to be safe. Major surgical complications at 30 days after Atriclip placement were ascertained retrospectively. There were no procedure-related myocardial infarctions, no evidence of damage or compression of the left circumflex coronary artery in any subject clinically, nor in long-term follow-up by CTA. Additionally no acute pericardial effusion, pleural effusion, myocardial perforation, or LAA perforation occurred with Atriclip or Atriclip PRO placement. There was 1 case of hypotension requiring pressors post-procedure due to significant volume shifts with hyponatremia. One subject had thoracic bleeding within 24 h post-operatively, requiring transfusion presumed due to bleeding from intercostal vessel after thoracoscopy port placement. One subject had access site bleeding not requiring transfusion or intervention (hybrid AF ablation case). For those subjects undergoing stand-alone TT Atriclip placement (n = 9), 3 were placed in the setting of failed subxiphoid Lariat ligation attempt, 3 were before available percutaneous technologies in patients with contraindication to long-term OAC, and 3 in the setting of a posterior chicken wing (behind pulmonary artery) location. The mean total procedure time for stand-alone TT Atriclip placement was 92 min, mean anesthesia time was 207.6 min, and average length of stay was 4.88 days. There were no complications associated with the stand-alone TT Atriclip procedure. Mean length of stay was prolonged primarily from 2 cases of Atriclip placement following failed Lariat placement due to availability of CT surgeon scheduling at index hospitalization.
At last follow-up, 12 patients remained on OAC treatment (warfarin, n = 6; rivaroxaban, n = 2; apixaban, n = 2; dabigatran, n = 2) for a variety of clinical indications including malpositioned Atriclip (n = 3), history of deep vein thrombosis or pulmonary embolism (n = 2), prior embolic events and thrombus on OAC therapy (n = 2), persistent arrhythmia undergoing cardioversion or ablation (n = 4), and patient preference (n = 1). Before placement of the Atriclip, 57 of 65 patients were on OAC therapy. In follow-up, there was a single stroke (n = 1) over 183 patient-years that occurred in a patient with confirmed complete closure of the LAA by Atriclip and no cardiac thrombus at TEE. Hypertensive urgency with lacunar infarct was confirmed by cerebral diffusion-weighted brain MRI; this subject was not on OAC therapy when the stroke occurred. Antiplatelet therapy consisted of 81 mg daily acetylsalicylic acid (ASA) in 35 subjects; no subject was on 325-mg dose ASA; 4 patients were on long-term clopidogrel 75 mg daily; and 1 subject was on long-term ticlopidine. The decision to resume OAC was made after the CTA results were evaluated in 2 of the subjects with incomplete LAA closure. Two subjects with incomplete LAA closure had absolute contraindication to OAC medications and remained on antiplatelet therapy alone.
At the final follow-up, 54 patients were in sinus rhythm on electrocardiography, 8 patients were in paced rhythm (cardiac resynchronization therapy or right ventricular pacing due to prior atrioventricular nodal junction ablation with permanent AF in 3 patients, atrial paced rhythm due to sinus node dysfunction in 5 patients). An additional 3 patients without devices were in AF on electrocardiography.
Evolving nonpharmacologic stroke prevention strategies in nonvalvular AF patients have become a viable alternative to OAC treatment; however, the U.S. Food and Drug Administration–labeled options for patients in the United States are limited to the Watchman LAA Occluder. The Watchman is specifically not labeled for patients who are contraindicated to OAC treatment, and thus, off-label use of the Lariat suture ligation tool and thoracoscopic Atriclip device have been pursued as alternatives (12–16). The Watchman has been used clinically in patients with contraindications to OAC therapy in the ASAP (ASA Plavix Feasibility Study With Watchman Left Atrial Appendage Closure Technology) and will be tested formally in a randomized fashion against antiplatelet therapy alone in the ASAP-TOO (Assessment of the Watchman Device in Patients Unstable for Oral Anticoagulation) trial (12).
Our study represents the first formal angiographic assessment of chronic position of the Atriclip and evaluation for contrast leaks or ineffective LAA closure. We found a high success rate for closure by 3D CTA with no associated thrombus or embolic strokes in a cohort of 65 patients over 183 patient-years. Previous published studies demonstrate peridevice or central leaks by TEE or CTA in up to 32% of Watchman devices, 20% of Lariat devices, and 36% of Amplatz Cardiac Plug LAA closures (5–9). Though no head-to-head clinical trials for LAA closure techniques exist, the TT Atriclip closure efficacy compares favorably, and our results show complete closure is likely higher than for percutaneous device–based or suture-based approaches.
Previously published data have established a lower than expected rate of stroke or transient ischemic attack in follow-up on patients undergoing LAA ligation at the time of cardiac surgery; however, more recent data show that surgical closure is often ineffective. In a recent 2016 pilot study, incomplete LAA closure by surgical excision, stapled excision, or internal suture ligation was 57% when formally assessed by TEE in all subjects post-operatively (4). A recent evaluation of LAA surgical closure at a mean of 44 months of follow-up showed ischemic stroke or systemic embolism occurred in 2% of patients with complete LAA closure versus 24% with incomplete LAA closure (all subjects had confirmed AF and were not on systemic anticoagulation) and 0% with smooth LAA stump (p = 0.006) (17). The importance of documentation of complete LAA closure before a decision to stop systemic anticoagulation in patients with known AF cannot be understated. Additionally, failure to completely excise the LAA, when LAA closure is performed during open chest procedures, has been associated with increased late neurologic events in long-term follow-up, supporting the need for a device that can completely seal the LAA orifice at the time of surgery (18). The EXCLUDE study used a variety of methods to assess closure during concomitant or open chest placement of the clip (10). However, open chest surgery solely for ligation of the LAA likely is not a viable methodology in most cases. Additionally, the EXCLUDE study did not include Atriclips placed by TT approach, and in the verification for LAA closure assessment, there was no formal imaging protocol.
The ability to place an Atriclip through thoracoscopic approach alone is appealing as it can be performed as a stand-alone therapy with an average case time in our series of 92 min and does not mandate any post-operative systemic anticoagulation therapy. Our experience has largely been with Atriclip placement during thoracoscopic hybrid AF ablation or following failed attempts at other LAA closure technologies (11,13). The majority of patients implanted in our series were challenging persistent or long-standing persistent AF cases or those with LAA anatomy unsuitable for closure by subxiphoid suture ligation in the setting of contraindications to long-term anticoagulation treatment. The experience with TT placement during hybrid AF ablation can likely be extrapolated to clip closure for stand-alone cases, with an expected reduction in overall adverse events through avoidance of entry into the right thorax. Patient selection for LAA closure with a thoracoscopic approach is limited if the patient cannot tolerate single lung ventilation or has undergone previous sternotomy or significant thoracic irradiation.
Our study does demonstrate that despite direct visualization of the LAA during TT placement of an Atriclip, with intraoperative TEE visualization, closure may not be completely effective. We have seen that direct manual compression of a lobe of the LAA using the Atriclip applicator can appear on TEE as complete occlusion and only after the compression is released is the lobe allowed to refill with blood. Follow-up imaging with TEE or CTA will confirm closure was not complete (13). This may take several minutes to hours and should warrant careful TEE assessment intraoperatively, likely in multiple angles through the LAA as with endocardial occlusion device implantation. Additionally, the shape and application of the Atriclip device lends itself to a closure location more typically at the neck of the LAA, rather than true ostial occlusion, as evidenced by the high prevalence of a smooth-walled stump in our series (54.7%, mean dimensions of 2.1-cm width by 1.4-cm depth). However, presence of this stump was not associated with stroke or systemic embolic events, regardless of the use of systemic anticoagulation therapy. Whether closure efficacy would be improved in a lower risk AF cohort with smaller LA volume and windsock only LAA anatomy is unknown. Online Videos 1 and 2 demonstrate a typical Atriclip TT placement from our series with careful identification and closure of a second LAA lobe, which was initially not appreciated on intraoperative TEE.
Due to the retrospective nature of patient selection, all subjects having a previously implanted Atriclip by TT approach at Vanderbilt did not consent to the additional CTA. There were 81 subjects identified with TT-placed Atriclip between June 13, 2011, and October 6, 2015; however, only 65 of 81 available subjects consented to study after review of the risks and benefits of additional dedicated CTA for LAA closure. This may introduce bias in the results given the small sample size if those subjects did not follow similar trends in angiographic closure. Additional modality imaging (MRI, TEE) was available for 12 of the remaining 16 subjects and supported a high rate of LAA closure efficacy. Stroke data were ascertained by chart review and as such are subject to potential bias in reporting. Whether exposed trabeculae or smooth-walled stumps after Atriclip provide a nidus for LAA or LA thrombus formation when subjects stop OAC therapy is unknown and warrants investigation. The primary objective of this study was to assess LAA closure efficacy and is hypothesis-generating only, as there was no formal internal comparator cohort to alternative methods of LAA closure.
Angiographic LAA closure efficacy with a TT-placed Atriclip is high (93.9%). The clinical significance of a smooth-walled remnant stump is unknown. Confirmation of complete occlusion should be made to ensure effective LAA closure before cessation of systemic anticoagulation treatment.
COMPETENCY IN MEDICAL KNOWLEDGE: Closure of the LAA for prevention of AF-related embolic stroke has become a viable option for patients unsuitable for long-term OAC therapy. Methods for ligation, occlusion, or removal of the LAA vary in efficacy when assessed for residual leaks into the remnant LAA and incomplete surgical ligation has been associated with increased stroke risk.
TRANSLATIONAL OUTLOOK: This CTA study of thoracoscopic-placed Atriclip device demonstrates a high rate of effective closure (93%), but caution remains to ensure the entire LAA is captured when the clip is released. A decision regarding cessation of OAC following Atriclip placement should be guided by formal imaging to confirm complete closure of the LAA.
The authors acknowledge Amanda Carroll, RN, who was integral in conduct of the trial.
External funding for the computed tomographic angiography scans was provided by Atricure Inc. Dr. Ellis has received research funding from Atricure, Medtronic, and Boston Scientific; and consulting fees from Atricure, SentreHeart, and Boston Scientific. Dr. Hoff has received consulting fees from Atricure. Dr. Whalen has received consulting fees from Atricure. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
All authors attest they are in compliance with human studies committees and animal welfare regulations of the authors’ institutions and Food and Drug Administration guidelines, including patient consent where appropriate. For more information, visit the JACC: Clinical Electrophysiology author instructions page.
- Abbreviations and Acronyms
- atrial fibrillation
- acetylsalicylic acid
- computed tomography
- computed tomographic angiography
- left atrium/left atrial
- left atrial appendage
- magnetic resonance imaging
- oral anticoagulation
- transesophageal echocardiography
- totally thoracoscopic
- Received December 27, 2016.
- Revision received March 16, 2017.
- Accepted March 28, 2017.
- 2017 American College of Cardiology Foundation
- Lee R.,
- Vassallo P.,
- Kruse J.,
- Malaisrie S.C.,
- Rigolin V.,
- Andrei A.C.,
- McCarthy P.
- Gianni C.,
- Di Biase L.,
- Trivedi C.,
- et al.
- Lee R.J.
- Viles-Gonzalez J.F.,
- Kar S.,
- Douglas P.,
- et al.
- Richardson T.D.,
- Shoemaker M.B.,
- Whalen S.P.,
- Hoff S.J.,
- Ellis C.R.
- Reddy V.Y.,
- Möbius-Winkler S.,
- Miller M.A.,
- et al.
- Aznaurov S.G.,
- Ball S.K.,
- Ellis C.R.
- Sievert H.,
- Rasekh A.,
- Bartus K.,
- et al.