ADVERTISEMENT
Delayed Pericardial Effusion Following Left Atrial Appendage Closure: A 5-Year Single-Center Experience
Watch the accompanying author interview here.
Abstract
Background. Though uncommon, pericardial effusion and cardiac tamponade are serious complications of left atrial appendage closure (LAAC). There are few data related to delayed pericardial effusions from this procedure. Methods. This is a single-center prospective analysis of 369 patients who underwent LAAC from December 2016 to March 2022 at a large teaching hospital. We compared patients who developed effusion (n = 5) to patients who did not (n = 364) to determine if there were any factors that predispose patients to developing acute (AEs) or delayed pericardial effusions (DEs). We compared patient characteristics, procedural data, and complications. Unadjusted, stepwise multivariate logistic regression was performed. Results. A total of 369 patients underwent LAAC. Of these, 5 patients (1.4%) developed pericardial effusion. Patients in both groups (pericardial effusion vs non-effusion) had similar patient and procedural characteristics. Patients in both groups were older (mean age, 78.4 ± 7.8 years in the effusion group vs 76.3 ± 8.5 years in the non-effusion group; P=.50) and white (60% in the effusion group vs 90.1% in the non-effusion group). CHA2DS2-VASc (4.2 ± 1.1 vs 4.5 ± 1.4; P=.67) and HAS-BLED (3.4 ± 0.5 vs 3.7 ± 0.9; P=.53) scores were similar in the effusion group vs the non-effusion group, respectively. Gastrointestinal bleeding was the most common procedural indication in both groups (80% in the effusion group vs 53.6% in the non-effusion group; P=.23). The majority of the patients in both groups had successful implantation in the first attempt, with the 27-mm device the most commonly used size. There was no significant difference in procedural duration (67 minutes in the effusion group vs 75 minutes in the non-effusion group; P=.16). Among patients who received the Watchman Legacy device, 2 patients developed AEs and no patients had DEs. Of those receiving the Watchman FLX device, 1 patient developed AE and 2 patients developed DEs. All of the patients with effusions had successful recovery. Conclusion. In this 5-year, single-center experience, DEs were uncommon and potentially related to LAA device anchor microperforation. No statistically significant risk factors predisposing patients to pericardial effusions were identified in our analysis.
J INVASIVE CARDIOL 2023;35(1):E1-E6. Epub 2022 November 30.
Complications of left atrial appendage closure (LAAC) have been significantly reduced with the development of new devices and have become infrequent in the contemporary era.1-5 However, pericardial effusion remains a serious potential complication of this procedure.1-11
The National Cardiovascular Data Registry (NCDR) LAAC registry noted an effusion rate requiring intervention of 1.4% for the first-generation device.3 Meanwhile, the PINNACLE (Primary Outcome Evaluation of a Next-Generation Left Atrial Appendage Closure Device) FLX trial, which investigated the second-generation of the Watchman device (Boston Scientific), described an effusion rate requiring intervention of 1%. In the PINNACLE FLX trial, there were no patients with pericardial effusions before day 7; all patients required pericardiocentesis after day 7.5 Pericardial effusions are a known complication from the Amulet occluder device (Abbott Cardiovascular) as well. In a large study comparing these 2 devices, the Amulet occluder device had a higher rate of pericardial effusions compared with the Watchman device (22/903 [2.44%] vs 11/896 [1.23%]).11 In this study, we review our experience with the Watchman Legacy and Watchman FLX devices and their complications at a large academic hospital in a cohort of 369 patients.
Methods
Study design. This study is a single-center, prospective analysis of 369 consecutive patients who underwent LAAC from December 2016 to March 2022. The study was approved by the University Hospitals Cleveland Medical Center institutional review board. These patients all had a history of non-valvular atrial fibrillation (AF) and CHA2DS2VASc score ≥2, in whom oral anticoagulation was indicated but had reason to consider alternative to long-term oral anticoagulation. We analyzed baseline demographic data, procedural characteristics, complications, and postprocedural anticoagulation for both groups. Delayed effusions were defined as effusions occurring >6 hours post procedure.
Procedural details. With developments in new techniques and devices, our procedural workflow changed over the 5-year time period, which has been previously described.6,12-14 In brief, transesophageal echocardiography (TEE) was initially used for procedural planning and intraoperatively under general anesthesia. In 2020, we changed to cardiac computed tomography angiography (CTA) for planning, and after the appropriate device was selected, the LAAC procedure was performed under intracardiac echocardiography (ICE) guidance, using conscious sedation with midazolam and fentanyl. Intravenous heparin was given to achieve an activated clotting time (ACT) >250 seconds. Protamine was administered at the discretion of the operator. As we transitioned to a CTA + ICE strategy, we also initiated a same-day discharge protocol.13 We obtain a limited echocardiogram prior to and 6 hours post procedure to assess for pericardial effusion. Imaging with TEE or CTA was also obtained 45 days post procedure.
Statistical analysis. Descriptive statistics were utilized for data analysis. Categorical variables were represented as numbers and percentages. Continuous variables were represented as means ± standard deviations. Unadjusted, stepwise multivariate logistic regression was performed using Stata, version 15.0.
Results
From December 2016 to March 2022, a total of 369 consecutive patients had implantation of Watchman Legacy and Watchman FLX devices. From December 2016 to September 2020, a total of 164 patients underwent LAAC using the Watchman Legacy device. From September 2020 to March 2022, a total of 205 patients received the Watchman FLX device. Five patients (1.35%) developed pericardial effusion.
Baseline demographics for patients with effusion (n = 5) and without effusion (n = 364) are shown in Table 1, Part 1 and Table 1, Part 2. Both groups were noted to be older (mean age, 78.4 ± 7.8 years vs 76.3 ± 8.5 year; P=.50) and white (60% vs 90.1%), with similar CHA2DS2-VASc (4.2 ± 1.1 vs 4.5 ± 1.4; P=.67) and HAS-BLED (3.4 ± 0.48 vs 3.7 ± 0.9; P=.53) scores in the effusion group vs the non-effusion group, respectively. A majority of patients in both groups had a history of paroxysmal atrial fibrillation (80% vs 51.4%; P=.02) and essential hypertension was the most common comorbid condition (100% vs 88.5%; P=.42). Indication for placing the device was predominantly gastrointestinal bleeding for both groups (80% vs 53.6%; P=.23) in the effusion group vs the non-effusion group, respectively.
Procedural characteristics for both groups are shown in Table 2. Both groups had high rates of successful implantation (100% vs 98.9% in the effusion group vs the non-effusion group, respectively). The majority had success with the first device size attempted (100% vs 92%). The 27-mm sized device was the most common device size implanted. Total procedure duration was similar (67 minutes vs 75 minutes; P=.16). The mean ACT during the procedure in the effusion group was 250 seconds. Protamine was not administered during the procedure for any of the patients with effusion. None of the patients with effusion required repeat trans-septal puncture. There were no device recaptures or size changes in the effusion group.
Antithrombotic regimen at discharge was similar in both groups, with dual-anti platelet therapy (DAPT) being most common (40% vs 44.1% in the effusion group vs the non-effusion group, respectively). Other discharge regimens for patients with pericardial effusion included novel oral anticoagulation (NOAC) only, single-antiplatelet agent (SAPT) with NOAC, and SAPT with warfarin.
Among patients who received the Watchman Legacy device, 2 patients developed AE and none developed DE. One of the effusions happened intraoperatively due to cardiac perforation and required surgical repair. Of those receiving the Watchman FLX device, 1 patient developed AE and 2 patients developed DE. One patient presented 24 hours after the procedure with chest tightness and was found to have an increase in effusion size compared with preprocedure echocardiogram. The effusion gradually increased over the ensuing 48 hours and she underwent a pericardial window due to the posterior location (Figure 1). The second patient complained of lightheadedness 8 hours after the procedure and was found to be hypotensive, requiring urgent pericardial drainage. The remaining patients were also successfully treated by pericardiocentesis. All 5 patients with effusion underwent imaging with CT scan or TEE 45 days after their procedure and did not have a significant peridevice leak, thrombus, or device embolization. None were readmitted after pericardial intervention. One patient with DE had persisting small anterior pericardial effusion on follow-up CT scan. The remaining 4 patients had trivial or no effusion on follow-up imaging.
Discussion
Pericardial effusion is one of the most serious complications of LAAC. We examined patients with and without effusion over a 5-year span. We were not able to identify any major differences between the 2 groups. Patients in the overall cohort were older, female, and white, with average CHA2DS2-VASC score of 4.4 and HAS-BLED score of 3.6. Gastrointestinal bleeding was the most common procedural indication. There were no differences in regard to procedural characteristics. Both groups had successful device implantation, with the majority receiving only 1 device and 27 mm the most common device size used. Both groups had similar procedural duration of approximately 1 hour.
AEs may result from injury at various points in the procedure, typically related to trans-septal puncture or manipulation of wires, catheters, and device in the left atrium or LAA.7 DEs are less common following LAA closure.9,10 However, in the PINNACLE FLX trial, all of the patients with significant pericardial effusions presented after day 7.5 In this experience, there were more DEs with the FLX device compared with the Watchman Legacy device. Two out of 3 patients had DEs after implantation of the Watchman FLX device while zero patients had DEs with the Legacy device. Both patients with DEs had imaging done shortly after the procedure, with no new pericardial effusion visualized. Still, concern for DE should not be a barrier to same-day discharge following LAAC. In fact, outcomes in patients with same-day discharge after LAA have been shown similar to those admitted overnight.15,16 In our analysis, the rate of DE was very low, suggesting that same-day discharge of patients who undergo LAAC is still safe and feasible.
We hypothesize that DE occurs due to microperforation during or just after device implantation. One possible explanation as to why DE was more common with the Watchman FLX device is the increase in the number of J-shaped fixation anchors on this device compared with the Watchman Legacy device. The Watchman FLX device contains 18 peripheral fixation anchors (2 rows of 9) compared with 10 fixation anchors in the earlier-generation device. The addition of more J-anchors was done to reduce device embolization but could have the unintended consequence of DE. The device embolization rate was 0.7% for the first-generation device but decreased to 0% for the second-generation device.3,5 There were no device embolization events noted in our cohort. The interaction of these increased anchors could lead to microperforation of the thin-walled appendage during the procedure, leading to a slow accumulation of blood in the pericardium as opposed to a more brisk accumulation that might occur with catheter or wire injury. Interestingly, both patients with DEs were in sinus rhythm at the time of the procedure. Sinus rhythm makes the appendage and the left atrium more mobile, thus increasing the interaction of the device and the heart tissue and possibly increasing the risk of injury. A tight hemostatic valve on the core wire may add additional tension between the core wire and the appendage, causing the anchors to pull upon appendage tissue and may be observed especially in the setting of hypermobile or hyperdynamic appendage. This phenomenon—commonly referred to as “auto-tug”—is noted when the core wire automatically shifts back and forth within a loosened hemostatic valve (Video 1). Maintaining a loose hemostatic valve until device release can reduce system tension and may reduce the risk of injury in this circumstance.
In addition to anchor number, anchor length may also contribute to the development of pericardial effusions. In the Amulet IDE trial, the Amulet occluder device had a higher rate of pericardial effusion compared with the Watchman Legacy (22/903 [2.44%] vs 11/896 [1.23%]).11 This higher rate of pericardial effusion may be related to longer fixation anchors. The Amulet anchors are 2.2 times longer compared with the effective length of the Watchman anchors (Figure 2 and Figure 3). Since total wall (LAA + pulmonary artery) thickness is approximately 1.55 mm, longer anchors could perforate with less manipulation.
Four of our patients with pericardial effusions developed cardiac tamponade. Identifying these patients early and treating them with pericardiocentesis or pericardial window to drain the effusion is crucial. All patients recovered well after drainage of the effusion. As with all of our patients, we obtained imaging after 45 days and these patients had no significant peridevice leak, thrombus, or device embolization, with stable position of the device. We recommend obtaining imaging on all the patients prior to discharge.
Study limitations. A few limitations should be noted. Because so few patients developed effusions, the study was not sufficiently powered to identify statistically significant risk factors between the 2 groups. There was a change in our procedural workflow over the 5-year span period. Finally, our project required extraction of data from the electronic medical record and there could have been inaccuracy in the data recorded in this record.
Conclusion
In this 5-year, single-center experience, DEs were uncommon and potentially related to LAA device anchor microperforation. No statistically significant risk factors predisposing patients to pericardial effusions were identified in our analysis.
Affiliations and Disclosures
From the 1Department of Medicine, University Hospitals Cleveland Medical Center, Cleveland, Ohio; 2Department of Cardiology, SSM Health St Louis University Hospital, St Louis, Missouri; and 3Division of Cardiology, Harrington Heart & Vascular Institute, University Hospitals Cleveland Medical Center, Cleveland, Ohio.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Filby is a consultant for Boston Scientific. The remaining authors report no conflicts of interest regarding the content herein.
The authors report that patient consent was provided for publication of the images used herein.
Manuscript accepted July 18, 2022.
Address for correspondence: Steven J. Filby, MD, University Hospitals, Cleveland Medical Center, 11100 Euclid Avenue, Lakeside 3rd floor, Cleveland, OH 44106. Email: Steven.Filby@UHhospitals.org
Acknowledgments. The authors are grateful to all nurses, technicians, residents, and fellows who care for our patients.
References
1. Holmes DR Jr, Reddy VY, Gordon NT, et al. Long-term safety and efficacy in continued access left atrial appendage closure registries. J Am Coll Cardiol. 2019;74(23):2878-2889. doi:10.1016/j.jacc.2019.09.064
2. Nadeem F, Igwe C, Stoycos S, et al. A new Watchman sizing algorithm utilizing cardiac CTA. Cardiovasc Revasc Med. 2021;33:13-19. Epub 2021 Jan 12. doi:10.1016/j.carrev.2021.01.009
3. Freeman JV, Varosy P, Price MJ, et al. The NCDR left atrial appendage occlusion registry. J Am Coll Cardiol. 2020;75(13):1503-1518. doi:10.1016/j.jacc.2019.12.040
4. Boersma LV, Schmidt B, Betts TR, et al; EWOLUTION Investigators. Implant success and safety of left atrial appendage closure with the Watchman device: peri-procedural outcomes from the EWOLUTION registry. Eur Heart J. 2016;37(31):2465-2474. Epub 2016 Jan 27. doi:10.1093/eurheartj/ehv730
5. Kar S, Doshi SK, Sadhu A, et al; PINNACLE FLX Investigators. Primary outcome evaluation of a next-generation left atrial appendage closure device: results from the PINNACLE FLX trial. Circulation. 2021;143(18):1754-1762. Epub 2021 Apr 6. doi:10.1161/CIRCULATIONAHA.120.050117
6. Filby SJ, Dallan LAP, Cochet A, et al. Left atrial appendage occlusion using cardiac CT angiography and intracardiac echocardiography: a prospective, single-center study. J Invasive Cardiol. 2021;33(11):E851-E856. Epub 2021 Oct 7.
7. Möbius-Winkler S, Majunke N, Sandri M, et al. Percutaneous left atrial appendage closure: technical aspects and prevention of periprocedural complications with the Watchman device. World J Cardiol. 2015;7(2):65-75. doi:10.4330/wjc.v7.i2.65
8. Reddy VY, Holmes D, Doshi SK, Neuzil P, Kar S. Safety of percutaneous left atrial appendage closure: results from the Watchman left atrial appendage system for embolic protection in patients with AF (PROTECT AF) clinical trial and the continued access registry. Circulation. 2011;123(4):417-424. Epub 2011 Jan 17. doi:10.1161/CIRCULATIONAHA.110.976449
9. Isogai T, Saad AM, Shekhar S, et al. Incidence and outcomes of pericardial effusion/tamponade following percutaneous left atrial appendage closure. Am J Cardiol. 2021;160:126-129. doi:10.1016/j.amjcard.2021.08.028
10. Krishnaswamy A, Kapadia SR. Incidence and outcomes of pericardial effusion/tamponade following percutaneous left atrial appendage closure. Am J Cardiol. 2021;160:126-129. doi:10.1016/j.amjcard.2021.08.028
11. Lakkireddy D, Thaler D, Ellis CR, et al. Amplatzer Amulet left atrial appendage occluder versus Watchman device for stroke prophylaxis (Amulet IDE): a randomized, controlled trial. Circulation. 2021;144(19):1543-1552. Epub 2021 Aug 30. doi:10.1161/CIRCULATIONAHA.121.057063
12. Palma Dallan L, Rana M, Yoon S, et al. CRT-700.47 Novel computed tomography angiography-based sizing methodology for Watchman FLX device in left atrial appendage closure. J Cardiovasc Electrophysiol. 2022;33(8):1781-1787. Epub 2022 May 26. doi:10.1111/jce.15548
13. Dallan LAP, Bezerra HG, Cochet A, et al. Safety, efficacy, and cost-effectiveness of same-day discharge for left atrial appendage occlusion. J Invasive Cardiol. 2022;34(2):E124-E131.
14. Filby SJ, Dallan LAP, Cochet A, Kobayashi A, Arruda M. Novel technique for transseptal passage of intracardiac echocardiogram probe during left atrial appendage closure. Cardiovasc Revasc Med. 2021;28S:150-152. Epub 2021 Apr 24. doi:10.1016/j.carrev.2021.04.021
15. Tan BE, Boppana LKT, Abdullah AS, et al. Safety and feasibility of same-day discharge after left atrial appendage closure with the Watchman device. Circ Cardiovasc Interv. 2021;14(1):e009669. Epub 2021 Jan 11. doi:10.1161/CIRCINTERVENTIONS.120.009669
16. Dallan LAP, Bezerra HG, Cochet A, et al. Safety, efficacy, and cost-effectiveness of same-day discharge for left atrial appendage occlusion. J Invasive Cardiol. 2022;34(2):E124-E131.
