https://orcid.org/0000-0003-2932-6525
Abstract
In patients with non-valvular atrial fibrillation, the vast majority of thrombi originate in the left atrial appendage (LAA). Thus, occluding the LAA significantly reduces one’s risk for developing an ischaemic stroke. To date, many different surgical methodologies in LAA occlusion (LAAO)/exclusion have been studied and utilized. Unfortunately, patients are often left with incomplete closure of their LAA, leaving behind residual lobes that continue to allow thrombus formations. With the recent rise in percutaneous approaches and devices such as the WATCHMAN FLX, there have been proven success rates in achieving total closure of the LAA. Reports and investigations regarding the utilization of WATCHMAN FLX devices in patients with surgically incomplete LAAO remain limited.
We present three cases of patients who had previously undergone surgical exclusion of the LAA yet unfortunately were left with residual LAA that continued to place them at high risk for an ischaemic stroke. Percutaneous LAAO with the WATCHMAN FLX was utilized to successfully achieve complete sealing of the residual lobes in failed LAA surgical closures.
Our multicentre case series elucidates that an increased risk of stroke due to surgical LAAO failure is a real-world possibility that is likely to be encountered in clinical practice. We demonstrate in this series how the WATCHMAN FLX may provide a feasible and safe method to supplement a surgically incomplete LAAO to allow for improved ischaemic stroke and systemic embolization risk reduction.
To elucidate that a failed or incomplete surgical left atrial appendage (LAA) exclusion is a real-world possibility that is likely to be encountered in clinical practice.
To demonstrate the feasibility of achieving complete LAA closure of previously incomplete residual lobes utilizing the WATCHMAN FLX.
To demonstrate how the high success rate of achieving complete LAA closure can be attributed to the endocardial nature of this approach with concomitant intraprocedural imaging and direct visualization of the appendage.
Introduction
Non-valvular atrial fibrillation is one of the leading causes of cardioembolic ischaemic strokes worldwide.1 The vast majority of the thrombi that lead to these strokes form in the left atrial appendage (LAA).2 For decades, clinicians have attempted various surgical and non-surgical approaches for LAA occlusion (LAAO).3 These approaches include but are not limited to suture exclusion or excision, suture/snares loops, and clipping.4 To this day, certain surgical approaches have shown relatively unfavourable results with higher failure and complication rates compared with the newer percutaneous endocardial approaches as seen with the WATCHMAN FLX implant.5–7 Often times, patients undergoing surgical LAA closure are left with incomplete occlusion, leaving behind residual lobes that continue to allow thrombi formation.8,9
Many studies have shown relatively suboptimal success rates with surgical closure in achieving complete LAAO.6,10,11 A variety of approaches, including ligation, stapling, excision with sutures, have been utilized. Surgical clipping (AtriClip) has thus far shown the greatest outcomes compared with traditional suturing and stapling. However, even clipping may result in incomplete closures in ∼6% of all cases.12 We believe that the endocardial approach in sealing these residual lobes with the WATCHMAN FLX is a feasible and efficacious method in providing these patients a proper and complete LAA sealing and cerebrovascular accident (CVA) risk reduction. As of now, this concept has not been extensively reported or studied in the literature yet. We present three cases of patients who had previously undergone surgical LAA closure yet unfortunately were left with residual LAA that continued to place them at high risk for an embolic CVA. We were able to fully occlude these residual lobes endocardially with the utilization of the WATCHMAN FLX device.
Timeline
| Patient 1 | Patient 2 | Patient 3 |
|---|---|---|
| Year 49 triple-vessel coronary artery bypass with left atrial appendage (LAA) surgical ligation | Year 63 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 56 new-onset atrial fibrillation. Anticoagulated with apixiban |
| Year 55 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 64 patient undergoes the MitraClip and Cox Maze procedure with incomplete LAA ligation. Remains on Warfarin | Year 58 surgical aortic valve repair with LAA occlusion (LAAO) with AtriClip |
| Year 65 warfarin switched to Apixiban | Years 66–70 persistent heavy bruising with symptomatic iron-deficiency anaemia | Year 59 first hospitalization due to lower GI bleed |
| Year 73 mechanical fall (first) with heavy bruising | Year 71 transoesophageal echocardiogram (TEE) confirms a large LAA remnant and a previously incomplete ligation. Complete LAAO residual successfully achieved with the WATCHMAN FLX | Year 60 s hospitalization due to lower GI bleed |
| Year 74 mechanical fall (second) with subdural haematoma. Anticoagulation held. The patient suffers ischaemic cerebrovascular accident | Year 61 complete LAAO residual successfully achieved with the WATCHMAN FLX | |
| Year 75 TEE reveals a large LAA remnant. Complete LAAO residual successfully achieved with the WATCHMAN FLX |
| Patient 1 | Patient 2 | Patient 3 |
|---|---|---|
| Year 49 triple-vessel coronary artery bypass with left atrial appendage (LAA) surgical ligation | Year 63 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 56 new-onset atrial fibrillation. Anticoagulated with apixiban |
| Year 55 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 64 patient undergoes the MitraClip and Cox Maze procedure with incomplete LAA ligation. Remains on Warfarin | Year 58 surgical aortic valve repair with LAA occlusion (LAAO) with AtriClip |
| Year 65 warfarin switched to Apixiban | Years 66–70 persistent heavy bruising with symptomatic iron-deficiency anaemia | Year 59 first hospitalization due to lower GI bleed |
| Year 73 mechanical fall (first) with heavy bruising | Year 71 transoesophageal echocardiogram (TEE) confirms a large LAA remnant and a previously incomplete ligation. Complete LAAO residual successfully achieved with the WATCHMAN FLX | Year 60 s hospitalization due to lower GI bleed |
| Year 74 mechanical fall (second) with subdural haematoma. Anticoagulation held. The patient suffers ischaemic cerebrovascular accident | Year 61 complete LAAO residual successfully achieved with the WATCHMAN FLX | |
| Year 75 TEE reveals a large LAA remnant. Complete LAAO residual successfully achieved with the WATCHMAN FLX |
| Patient 1 | Patient 2 | Patient 3 |
|---|---|---|
| Year 49 triple-vessel coronary artery bypass with left atrial appendage (LAA) surgical ligation | Year 63 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 56 new-onset atrial fibrillation. Anticoagulated with apixiban |
| Year 55 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 64 patient undergoes the MitraClip and Cox Maze procedure with incomplete LAA ligation. Remains on Warfarin | Year 58 surgical aortic valve repair with LAA occlusion (LAAO) with AtriClip |
| Year 65 warfarin switched to Apixiban | Years 66–70 persistent heavy bruising with symptomatic iron-deficiency anaemia | Year 59 first hospitalization due to lower GI bleed |
| Year 73 mechanical fall (first) with heavy bruising | Year 71 transoesophageal echocardiogram (TEE) confirms a large LAA remnant and a previously incomplete ligation. Complete LAAO residual successfully achieved with the WATCHMAN FLX | Year 60 s hospitalization due to lower GI bleed |
| Year 74 mechanical fall (second) with subdural haematoma. Anticoagulation held. The patient suffers ischaemic cerebrovascular accident | Year 61 complete LAAO residual successfully achieved with the WATCHMAN FLX | |
| Year 75 TEE reveals a large LAA remnant. Complete LAAO residual successfully achieved with the WATCHMAN FLX |
| Patient 1 | Patient 2 | Patient 3 |
|---|---|---|
| Year 49 triple-vessel coronary artery bypass with left atrial appendage (LAA) surgical ligation | Year 63 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 56 new-onset atrial fibrillation. Anticoagulated with apixiban |
| Year 55 new-onset atrial fibrillation. Anticoagulated with warfarin | Year 64 patient undergoes the MitraClip and Cox Maze procedure with incomplete LAA ligation. Remains on Warfarin | Year 58 surgical aortic valve repair with LAA occlusion (LAAO) with AtriClip |
| Year 65 warfarin switched to Apixiban | Years 66–70 persistent heavy bruising with symptomatic iron-deficiency anaemia | Year 59 first hospitalization due to lower GI bleed |
| Year 73 mechanical fall (first) with heavy bruising | Year 71 transoesophageal echocardiogram (TEE) confirms a large LAA remnant and a previously incomplete ligation. Complete LAAO residual successfully achieved with the WATCHMAN FLX | Year 60 s hospitalization due to lower GI bleed |
| Year 74 mechanical fall (second) with subdural haematoma. Anticoagulation held. The patient suffers ischaemic cerebrovascular accident | Year 61 complete LAAO residual successfully achieved with the WATCHMAN FLX | |
| Year 75 TEE reveals a large LAA remnant. Complete LAAO residual successfully achieved with the WATCHMAN FLX |
History of presentation
Patient 1
A seventy-five-year-old male with a past medical history of non-valvular paroxysmal atrial fibrillation (CHADS2-VASc = 6; HAS-BLED 4) anticoagulated with rivaroxaban, and coronary artery bypass grafting presented with a recent subdural haematoma after a fall (BARC Type 3c). His physical examination was unremarkable. He was deemed a poor candidate for long-term oral anticoagulation. A shared decision between neurology, cardiology, and the patient himself was made to perform LAAO with the WATCHMAN FLX device. During pre-procedural imaging with transoesophageal echocardiogram (TEE), it was incidentally discovered that the LAA had previously been surgically ligated. Unfortunately, a relatively large LAA remnant still remained (Figure 1AandB). Intraprocedural intracardiac echocardiography (ICE) revealed a residual lobe with an ostial measurement of 17.3 mm with a maximum depth of 16.2 mm. A 20 mm WATCHMAN FLX was selected. Initial deployments of the device were not well seated and were quite proximal. We recaptured and reattempted deployment by realigning the device with the most anterior lobe of the appendage in the right anterior oblique caudal projection. Once in this position, we kept a counter torque on the entire delivery sheath and steadily unsheathed the device. We were able to maintain an anterior trajectory and keep it in the upper dome of the residual appendage (Figures 1CandD and 2A). This time, position, anchor, size, and seal (PASS) criteria were successfully met along with 15–20% compression and the device was released. The patient tolerated the procedure well. He was subsequently safely discharged home on daily aspirin and rivaroxaban. A follow-up TEE 45 days’ status post WATCHMAN FLX revealed a well-sealed LAA with no peri-device leakage or device-related thrombi (Figure 2B). Rivaroxaban was then discontinued indefinitely, and he was initiated daily dual anti-platelet therapy (DAPT) with aspirin and clopidogrel for 6 months followed by daily aspirin monotherapy thereafter.
Deployment of a 20 mm WATCHMAN FLX into a left atrial appendage remnant with fluoroscopic views. (A and B) Demonstration of left atrial appendage residual lobes with exposed trabeculae pre deployment. (C and D) Post deployment revealing a successful closure of left atrial appendage remnants. WFD, WATCHMAN FLX™ device; ICE, intracardiac echocardiography.
Echocardiography of a left atrial appendage remnant pre- and post deployment of a 20 mm WATCHMAN FLX. (A) Visualization of a WATCHMAN FLX™ implant on intracardiac echocardiography. (B) Transoesophageal echocardiogram at 45 days showing the WATCHMAN FLX™ device firmly in place with no perivalvular leak. LA, left atrium; LV, left ventricle; WFD, WATCHMAN FLX™ device.
Patient 2
A seventy-one-year-old female with non-valvular paroxysmal atrial fibrillation (CHADS2-VASc 3; HAS-BLED 4) status post the Cox Maze procedure with a surgical LAA clip and prior MitraClip was suffering from recurrent bleeding (BARC Type 2) and easy bruising. A physical examination revealed scattered senile purpura throughout all her extremities. She was deemed a poor candidate for long-term oral anticoagulation. Preprocedural TEE demonstrated a small LAA with residual lobes anteriorly and posteriorly that were not occluded. The neck of the appendage was patent as well. Ostial measurements at 0, 45, 90, and 130° were 12.1, 12.0, 14.0, and 16.6 mm, respectively, with a maximum depth of 14.5 mm. A shared decision between our team and the patient herself was made to perform LAAO via the WATCHMAN FLX procedure. Angiography again revealed a large remnant of the LAA at the level of the previously placed surgical clip (Figures 3A and 4AandB). We initially prepped a WATCHMAN FLX 20 mm device according to the manufacturer’s recommendations and appropriately de-aired the device. Once deployed, it was clear that it was too small in width to seal the residual appendage. Therefore, we elected to recapture the entire device and switch to a larger 24 mm WATCHMAN FLX device. We lined the device up with the appropriate fluoroscopic markers and deployed the device to ball and then advanced the device out with the delivery sheath at the ostium. A tug test demonstrated incomplete engagement of the anchors. We elected to partially recapture and reattempt deployment with a more coaxial and anterior approach. Position, anchor, size, and seal criteria were successfully met, and the device was released. There was ∼10–15% compression. Transoesophageal echocardiogram confirmed good positioning with no peri-device colour flow (Figures 3B and 4CandD). The patient tolerated the procedure well and was discharged on daily aspirin and warfarin. A follow-up TEE 45 days’ status post WATCHMAN FLX revealed a well-sealed LAA with no peri-device flow or device-related thrombi (Figure 4C). Warfarin was then discontinued indefinitely, and she initiated daily DAPT with aspirin and clopidogrel for 6 months followed by daily aspirin monotherapy thereafter.
Deployment of a 24 mm WATCHMAN FLX™ into a left atrial appendage remnant with fluoroscopic views. (A) Demonstration of the left atrial appendage remnant pre deployment. (B) Post deployment with a good seal. WFD, WATCHMAN FLX™ device.
Transoesophageal echocardiogram views of a residual left atrial appendage remnant pre- and post deployment of a 24 mm WATCHMAN FLX™. (A and B) Demonstration of left atrial appendage remnants pre deployment. (C) Post deployment revealing a good seal with no leak. (D) Successful closure of the left atrial appendage remnant post deployment. LA, left atrium; LCx, left circumflex artery; PA, pulmonary artery; WFD, WATCHMAN FLX™ device; NCC, non-coronary cusp; RCC, right coronary cusp; LCC, left coronary cusp.
Patient 3
A sixty-one--year-old female with heart failure with preserved ejection fraction and chronic persistent atrial fibrillation (CHADS2-VASc = 3; HAS-BLED 4) anticoagulated with Apixiban was evaluated for multiple episodes of gastrointestinal bleed from duodenal ulcers and AV malformations (BARC Type 3). A physical examination again revealed scattered senile purpura throughout her extremities as well as a persistent irregularly irregular heart rhythm. She previously also underwent surgical aortic valve replacement for severe aortic valve stenosis, pulmonary vein isolation, and LAAO with AtriClip. She was initially evaluated for a percutaneous LAAO with a WATCHMAN 2.5 device. A TEE assessment demonstrated incomplete closure of the LAA with AtriClip (Figure 5). Ostial measurements at 40, 90, 130, and 180° were 14.6, 13.8, 16.6, and 14.6 mm, respectively. Computed tomography angiography (CTA) with three-dimensional reconstruction confirmed a residual appendage with an ostium of 17 mm and a maximum depth of 15.4 mm (Figure 6). Initial attempts at device deployment were unsuccessful due to a non-compatibility of the WATCHMAN 2.5 with the appendage’s shallow depth and smaller ostium, resulting in device instability and a higher risk of device embolization. Once the second-generation WATCHMAN FLX device was commercially available, we were able to successfully achieve complete LAAO with a 20 mm WATCHMAN FLX device (Figure 7). Sufficient compressibility of 15–20% without any peri-device leak or device-related thrombi were noted (Figure 8). The patient was safely discharged home on daily aspirin and low-dose Apixiban.
Transoesophageal echocardiogram view with evidence of a residual left atrial appendage and post-surgical left atrial appendage occlusion. LAA, left atrial appendage; LAAO, left atrial appendage occlusion.
Three-dimensional computed tomography angiography reconstruction of a model LA with evidence of a residual left atrial appendage and the AtriClip placed during surgical left atrial appendage occlusion. LA, left atrium; LAA, left atrial appendage; LAAO, left atrial appendage occlusion.
Angiography pre- and post deployment of a 20 mm WATCHMAN FLX™ revealing the presence of the AtriClip and a residual left atrial appendage with fluoroscopic views. LAA, left atrial appendage.
Echocardiographic images of a left atrial appendage post deployment of a 20 mm WATCHMAN FLX obtained at various angles. WFD, WATCHMAN FLX™ device.
Results and follow-up
All three patients were safely discharged home on oral anticoagulation and daily aspirin. Patients 1 and 2 underwent a follow-up TEE at 45 days, which revealed well-seated and well-anchored devices, with no peri-device leaks (<5 mm) or any device-related thrombi. After 45 days, they transitioned to DAPT with aspirin and clopidogrel, followed by de-escalation to daily aspirin after Month 6. All patients now undergo a repeat TEE at 1-year post-LAAO according to protocol. Patient 1 is now 12 months’ status post complete LAAO and has continued to enjoy a course free of any bleeding complications, while maintaining adequate stroke prophylaxis. He is due for 1-year post-LAAO follow-up TEE. Patient 2 is now 14 months’ status post LAAO and has also enjoyed a course free of any bleeding complications, while maintaining adequate stroke prophylaxis, with serial transoesophageal echocardiography monitoring at 45 days and 1 year (see Supplementary material online, Figure S3) revealing a well-seated device with adequate seal and no peri-device leaks (<5 mm) or any device-related thrombi. Patient 3 was unable to undergo a 45-day TEE due to a traumatic mechanical fall with subsequent hip fracture. Shortly afterwards, she succumbed to SARS-CoV-2-induced respiratory failure and cardiac arrest 3 months after her procedure.
Discussion
Surgical and percutaneous LAAO have become widespread approaches in providing non-pharmacological CVA prophylaxis in patients with non-valvular atrial fibrillation. Various surgical techniques have been previously described, such as suture ligation (i.e. Cox-MAZE), stapling, clipping (i.e. AtriClip), and surgical excision. These techniques often result in residual LAA remnants that continue to contribute to thrombus formation. When compared with surgical LAAO, utilization of the WATCHMAN FLX device has the distinct advantage in that it seals the appendage endocardially to avoid any ambiguity in ensuring complete exclusion of all lobes of the LAA from the blood pool. This self-expanding occlusion device is engineered with a nitinol frame with a permeable polyester fabric covering the atrial-facing surface and 18 intricate peripheral fixation anchors arranged in 2 layers. The design has significantly improved stability and reduced embolization risk in comparison with the former WATCHMAN 2.5. Current data suggest that the use of WATCHMAN FLX is at least equivalent to using oral anticoagulation in providing CVA prophylaxis for patients with atrial fibrillation.13
In 2016, Boersma et al. conducted a large study with 1021 subjects who appropriately underwent the WATCHMAN procedure. The study concluded that the WATCHMAN had a 99.3% success rate in achieving a total closure of the LAA.14 It also provides a much less-invasive approach that allows patients to enjoy swifter recovery, less risk for infections, and same-day discharge. In contrast, Kanderian et al. revealed that surgical LAA closures were successful at a rate of roughly only 40%. Nearly half of the patients with unsuccessful surgical LAA closure developed LAA thrombus. Of these, one in seven can be expected to eventually suffer from an ischaemic stroke or transient ischaemic attack.6 Surgical excision approaches have had the highest success rates, with complete LAA closure rates of up to 73%, whereas only 23% of suture exclusion cases revealed successful closure on follow-up TEE. Recanalization of the LAA in cases utilizing sutures or staplers is possible due to the fact that they may erode through the LAA tissue over time.15 Some external ligation devices such as the AtriClip have proved to be more successful than traditional suturing and stapling. However, the AtriClip requires that the surgeon mobilize the entire LAA and place the clip at the level of the ostium from an external approach without confirmation from the endovascular side due to the fact that the heart is decompressed and is on bypass at the time of clip placement, thus still leading to possible incomplete closure at the level of the LAA ostium.
Despite issues with incomplete sealing, the LAAOs three trial demonstrated that with surgical ligation, patients are less likely to experience atrial fibrillation–associated systemic embolism or ischaemic stroke. However, almost 5% of the occlusion group patients suffered an ischaemic stroke/systemic embolization event.16 An in-depth review of the trial demonstrates that approximately three-fourths of the trial participants were continued on oral anticoagulation for 3 years after LAA exclusion. On the other hand, the PINNACLE FLX trial studying the WATCHMAN FLX device demonstrated that 99% of patients’ LAAs were successfully excluded with a 0.5% risk of ischaemic stroke. Furthermore, 96% of patients were liberated from oral anticoagulation at 45 days.5 Similar results were witnessed in the real-world SURPASS registry that included over 16 000 patients.17
WATCHMAN FLX device size selection in patients with residual LAAs remains a complex decision that predominantly depends on remnant depth and may require multiple fine procedural adjustments (partial/complete recaptures, sheath torque, altering trans-septal access site) to navigate these surgically altered appendages. As of now, there is very limited literature on specific criteria for patients in whom risks would outweigh benefits. However, as noted above, a unified structural heart team approach is advised to be able to delineate procedural risks based on operator experience weighed against the potential benefit of being off oral anticoagulation. We recommend comprehensive pre-implant imaging with either TEE or CTA (or both). Experientially, we believe that a minimal residual depth of at least 12 mm is necessary for appropriate landing of the smallest WATCHMAN FLX device. Furthermore, we believe selecting a device size that allows for 15–20% compression would optimize implantation success rates and outcomes.
Our case series elucidates that CVA in the setting of failed or incomplete surgical LAA exclusion is a real-world possibility that is likely to be encountered in clinical practice. We have demonstrated with three cases that the utilization of the WATCHMAN FLX has potential to adequately address this issue by achieving complete LAAO in such patient populations, thus providing optimal CVA risk reduction that was sought after in the first place. Further studies with larger sample sizes will help validate the use of the WATCHMAN FLX in surgically altered LAA anatomies.
Physician perspective
In September 2022, the United States Food and Drug Administration made changes to the WATCHMAN FLX labelling and approved DAPT as an alternative to oral anticoagulation and aspirin combination for the first 45 days post-WATCHMAN LAAO, regardless of bleeding risk. The revised labelling for the WATCHMAN FLX provides physicians with more flexibility to exercise their clinical judgment based on individual patient characteristics to determine the most appropriate post-procedural regimen. In the USA, due to an ongoing controversy and a lack of unequivocal data, post-WATCHMAN antithrombotic regimens continue to be tailored uniquely towards the individual patient. In the event of bleeding complication(s), a de-escalation of the regimen itself as well as the methodology of doing so remains at the discretion of the physician. The aforementioned patients in this study were all moderate to high risk for bleeding. Despite this, the authors would like to clarify that American guidelines have no specific recommendations in adjusting their post-procedural regimens. Therefore, our patients were provided with traditional post-procedural regimens consistent with those of the PREVAIL and PINNACLE FLX trials.
However, it is crucial that the readership understands the current variations of the European vs. American guidelines. In Europe, Calabrò et al. summarize with an undetermined level of evidence that patients post-WATCHMAN LAAO who are deemed to be at high risk for bleeding complications are recommended to initiate Clopidogrel 75 mg for the first 1–6 months or until adequate LAA sealing (leak <5 mm) is achieved followed by daily aspirin 75–325 mg indefinitely. Oral anticoagulation is NOT recommended.18 Additional large-scale studies are needed and ongoing in order to better understand ways to optimize such post-implantation regimens.
Lead author biography
Dr Hafez Golzarian is a resident physician at Mercy Health—St. Rita’s Medical Center in Ohio, USA. In 2016, he partook in a year-long research internship at the National Institute of Health in Maryland, USA, while pursuing an accelerated pre-health Master’s in biotechnology at Johns Hopkins University. He graduated from medical school from Philadelphia College of Osteopathic Medicine in 2021. He has interests in clinical research, particularly in interventional cardiology, electrophysiology, cardiac imaging, heart failure, and critical care.
Supplementary material
Supplementary material is available at European Heart Journal – Case Reports.
Slide sets: A fully edited slide set detailing this case and suitable for local presentation is available online as Supplementary data.
Consent: Both written and verbal consent was obtained by the subjects involved in this case series in accordance with COPE guidelines and are available upon request.
Funding: None declared.
Data availability
All data and images supporting the findings of this study are available upon request.
References
Author notes
Conflict of interest: The authors have no conflicts of interest to disclose.
Comments