Left Atrial Appendage Occlusion Using Cardiac CT Angiography and Intracardiac Echocardiography: A Prospective, Single-Center Study

Objectives. We reviewed the periprocedural events, accuracy of device selection, and outcomes of a series of patients receiving percutaneous left atrial appendage occlusion (LAAO) with cardiac computed tomography angiography (CTA)-guided preprocedural planning and intracardiac echocardiography (ICE)-guided device deployment. Background. Percutaneous LAAO with the Watchman device (Boston Scientific) is approved by the United States Food and Drug Administration for stroke prevention in patients with non-valvular atrial fibrillation with a demonstrated contraindication to oral anticoagulation. Cardiac CTA preprocedural planning with utilization of an ICE-guided deployment may be associated with favorable outcomes. Methods. A prospective analysis of 71 non-consecutive patients who underwent LAAO over an 18-month period with cardiac CTA-guided preprocedural planning and ICE was conducted. Procedures were performed in a single large, academic hospital in the United States. Procedural success, correlation of CTA preprocedural device sizing with final device size utilization, adverse events, length of procedure, and length of stay were evaluated. Results. Preprocedural cardiac CTA-guided device sizing was consistent with the final deployed device in 69 patients (97.2%) evaluated in this case series. Procedure success rate was 100%. All implants were performed using Watchman devices, 45 (63.4%) with the original 2.5 platform and 26 (36.6%) with the Watchman FLX platform. All patients were treated with conscious sedation and the mean length of stay was 1.45 ± 0.72 days. Only 2 adverse events occurred and both resolved. Conclusions. Cardiac CTA-guided preprocedural planning resulted in accurate device sizing in this patient sample and may be used in conjunction with ICE and conscious sedation for a same-day discharge strategy in select patients.

J INVASIVE CARDIOL 2021;33(11):E851-E856. Epub 2021 October 7.

Key words: atrial fibrillation, computed tomography angiography, CTA, ICE, intracardiac echocardiogram, stroke prevention, structural intervention

Imaging guidance to help guide the selection of device size and shape of the delivery system for the left atrial appendage occlusion (LAAO) procedure is essential to success. Most operators have used transesophageal echocardiography (TEE) in conjunction with general anesthesia (GA) for LAAO. However, the use of GA requires anesthesiology support and is associated with longer patient recovery and more postprocedural complications than conscious sedation.¹ In addition, as an aerosol-generating procedure, TEE may pose a risk of COVID-19 exposure to patients and healthcare personnel.² A strategy of cardiac computed tomography angiography (CTA) preplanning with intracardiac echocardiography (ICE) and conscious sedation for LAAO may increase LAAO procedural efficiency, facilitate accurate device sizing, avoid general anesthesia, be safer in the COVID-era, and translate into reduced hospital stay and cost. There have been limited reports on the combined use of CTA preplanning and intraprocedural ICE for the Watchman LAAO device (Boston Scientific). We therefore report on our experience of using a CTA-ICE strategy for successful Watchman device implantation in a single-center, prospective analysis.

Patient population and study design. This is a prospective analysis of patients who underwent the Watchman procedure with a CTA-ICE strategy at University Hospitals, Cleveland Medical Center in Cleveland, Ohio from July 1, 2019 to December 1, 2020. This study was approved by the University Hospitals Cleveland Medical Center institutional review board. All patients referred for closure had a history of non-valvular atrial fibrillation with a documented CHADS-VASc score ≥2 and a rationale that required an alternative to long-term anticoagulation. We reviewed the final device size deployed, the number of devices required, correlation of final device size to preprocedural device size estimation, length of stay, patient demographics, and periprocedural adverse events.

Preprocedural imaging. All patients underwent preprocedural cardiac CTA imaging within 1 month of their device implant and several underwent cardiac CTA on the same day as device implant. Images were acquired using a Somatom Definition Flash dual-source 128-slice CT scanner (Siemens) performed with prospective systolic triggered cardiac gating (Figure 1). Multiplanar reconstruction (MPR) analysis of the LAA was performed on an Aquarius Workstation (TeraRecon). An MPR oblique plane was defined from the level of the circumflex artery to a point 1-2 cm from the tip of the limbus.³ The area of the LAA landing zone was then measured and used to select the initial Watchman device size. If the area indicated that there were 2 possible device sizes that would provide sufficient compression, then the larger size was selected unless the depth available was a concern. LAA device depth was estimated by measuring from the LAA ostium to the end of the most usable appendage lobe. Three-dimensional (3D) rendering was performed to assess LAA morphology, and the interatrial septum was assessed for morphology and thickness as well as best angle of entry to access the LAA.     

Intraprocedural imaging and device deployment. All procedures were performed using conscious sedation through administration of intravenous midazolam and fentanyl. Both Watchman LAAO 2.5 and FLX platforms were used, with transition to FLX following approval by the United States Food and Drug Administration (FDA) and availability at our center. Intravenous heparin was administered to achieve an activated clotting time >250 seconds. An 8 Fr Acuson AcuNav ultrasound catheter (Biosense Webster) was used for ICE imaging. Assessment for pericardial effusion was performed with the AcuNav probe advanced into the right ventricle. Thrombus in the LAA was excluded by positioning the catheter in the right ventricular outflow tract. Using views from the ICE catheter in the right atrium, trans-septal puncture was performed using NRG radiofrequency (Baylis Medical) transseptal needles through a VersaCross transseptal sheath (Baylis Medical). The Watchman sheath was advanced across the septum and then retracted back into the right atrium, thus allowing the ICE probe to enter through the 14 Fr dilated passage. The ICE ultrasound catheter was used to obtain LAA images from the left atrium. A 6 Fr pigtail catheter was advanced through the Watchman delivery sheath and advanced into the LAA to perform LAA angiography, most commonly in anterior-posterior/caudal or right anterior oblique/caudal projections. In all cases, devices were chosen based on preprocedural CTA measurements and confirmed with LAA angiography. The appropriately selected device was then loaded into the delivery sheath and implanted in the LAA. A modified “PASS” (position, anchor, size, seal) criteria was applied for device release and used to define device success. Two primary ICE views were utilized from the LA: (1) mid left atrium with view of the mitral valve, LAA, and left upper pulmonary vein (LUPV); and (2) mitral inflow with an inferior view of the LAA and LUPV (Figure 2). ICE was used to determine the presence of leak around the device. LAA angiography was also utilized to ensure adequate position, compression, and seal during the procedure. Procedure time was defined as time from local anesthetic given to device deployment.

Following device deployment, patients were typically admitted overnight for observation and a repeat limited transthoracic echocardiogram (TTE) was obtained the next day. During the latter aspect of the time frame queried, patients eligible for early discharge underwent postprocedure TTE on the same day as the procedure.

From July 1, 2019 to December 1, 2020, a total of 71 Watchman LAAO devices were implanted at our center utilizing cardiac CTA for preprocedural device sizing as well as an ICE-guided implant strategy. The implants were performed using the Watchman 2.5 platform for 45 patients (63.4%) and the Watchman FLX platform for 26 patients (36.6%); all were treated under conscious sedation. Baseline patient characteristics are shown in Table 1. The mean age was 76.0 ± 8.8 years, CHA₂DS₂-VASc score was 4.2 ± 1.4, and HAS-BLED score was 3.6 ± 1.0. The majority of the patients in this series had paroxysmal AF (59.2%).

The mean procedure time was 58.8 ± 11.1 minutes (Table 2). The device size predicted by cardiac CTA corresponded with the final device used in 69 out of 71 cases (97.2%) and the procedure success rate for this series was 100%. Two cases required deployment of a second device due to device recapture; however, in neither case was a change in device size required. This resulted in 1.06 ± 0.23 devices used per case within this sample.

Only 2 procedural complications (2.8%) occurred in this series. For the first patient, a new pericardial effusion was noted on echocardiography during the procedure, which progressed to cardiac tamponade. The patient underwent successful pericardiocentesis with pigtail drain placement. The effusion resolved and the patient was discharged home 4 days after the procedure with no further complications noted thereafter. The second patient developed a brief period of asystole, requiring a temporary pacemaker placement prior to proceeding with device implant, but without sequela.

From the total, 23 of 40 (57.5%) potential candidates were successfully discharged on the day of the procedure. Most of the patients considered who were not successfully discharged had not been able to receive a post-device implant TTE and were kept overnight to obtain this study on the following day. In addition, a few patients had to stay overnight for surveillance of minor access-site bleeding following the procedure. The majority of patients (96%) were discharged following overnight observation (Figure 3).

At the time of this writing, 63 of the 71 cases were at least 45 days out from the procedure and eligible for transesophageal imaging of the LAAO and have completed follow-up TEE. All of the evaluated patients demonstrated no leak or a jet leak of <4 mm, meeting the threshold to cease oral anticoagulation. No deaths, cerebrovascular accidents, or transient ischemic attacks have been reported since device implant.

We present data from a prospective evaluation of patients at a single large, academic hospital who underwent Watchman LAAO device implantation over a 12-month period. This sample is unique in that all patients in this series underwent preprocedural device sizing with cardiac CTA for LAA anatomic assessment as well as ICE-guided device deployment. The results from our analysis demonstrate this combined imaging strategy to be both safe and effective. Using the CTA + ICE strategy, we had 100% device success rate and good clinical outcomes.

In our series, cardiac CTA demonstrated 97.2% accuracy in predicting the appropriately sized device. Others have compared the accuracy of cardiac CTA with TEE in LAAO device sizing with similar, albeit less impressive, results.^4-6 These findings are not surprising given the superior visualization of the LAA imaging that cardiac CTA provides through the ability to reconstruct the LAA in various planes. The degree of predictive accuracy is likely also reliant on the approach using a scanner with higher temporal resolution in the dual-source mode, acquiring the scans during ventricular systole and ensuring adequate engorgement of the LAA with preprocedure saline infusion. Using a 64-slice scanner, Lopez-Minguez et al reported that CTA predicted accurate LAAO device size in 75.7% of cases.⁷ Even so, in that series, CTA outperformed both TEE (51.4%) and angiography (48.6%) for appropriate device sizing. Cardiac CTA offers several other advantages over TEE in LAAO planning because it provides accurate ostial and depth measurements of the LAA due to the ability to reformat images.⁸ Cardiac CTA also has the ability to provide 3D rendering to aid in the understanding of 3D appendage morphology and presence of thrombus in the LAA, which is critical in planning device placement and in reducing the risk of complications. Finally, cardiac CTA also allows for an understanding of the interatrial septum and its orientation to the LAA for planning the trans-septal approach.

Each case in our series also utilized ICE rather than TEE for device deployment guidance. The use of ICE has been shown not only feasible for guiding LAAO implantation, but in fact may be more advantageous than TEE guidance, with a recent meta-analysis demonstrating a reduction in contrast media utilization for patients treated with an ICE-guided strategy.⁹ Another important advantage of an ICE-guided strategy is the avoidance of GA, which may allow earlier discharge.¹⁰

Although Williams et al have reported a series of 117 patients evaluated for same-day discharge after LAAO using GA, only 66% of those evaluated were discharged on the same day in that series.¹¹ After changing to a same-day discharge strategy, we were able to safely discharge 23 out of 40 patients evaluated and most of the patients that stayed overnight did so due to logistical issues. Although the number of patients in our series who were discharged on the same day was small, we expect that over time, with the accuracy and efficiency afforded by a CTA + ICE strategy, many of our LAAO patients will be discharged on the same day. While same-day discharge is commonly done for percutaneous coronary intervention, patent foramen ovale closure, and catheter ablation procedures, most centers monitor patients overnight following LAAO.^12,13 Changing to a same-day strategy for LAAO may improve patient satisfaction and reduce cost.

Our results are in accordance with other recent studies.¹⁴ Cruz-Gonzales et al¹⁵ published a 4-month experience of 12 highly experienced European centers including 165 LAAO procedures with the Watchman FLX platform and demonstrated similar results, with procedural and technical success in all cases. The average CHA₂DS₂VASc score was 4.4 ± 1.4 and average HAS-BLED score was 3.5 ± 1.2, which are very similar to our numbers. A single device was used 96% of the time, and the average procedure time was 45 minutes. On the other hand, preprocedural CT was only used in 38.2%, local sedation in 39.7%, and ICE guidance in 17.3%. We used the Watchman FLX platform in 26 patients (36.6%), after the FDA approved its use on July 21, 2020.

Our LAAO CTA + ICE strategy began initially as a means to improve procedural efficiency and scheduling logistics.^16,17 We found that the high-quality imaging of cardiac CTA allowed for effective procedural planning, and a conscious sedation approach with ICE improved procedural efficiency and patient recovery. We have been able to further improve our workflow efficiency by having patients undergo cardiac CTA on the same day as their procedure. Although implemented only recently at our center, 24 patients (33.8%) in our series had same-day preprocedural planning and device implant. While there remains the possibility of procedural cancellation if LAA thrombus is found, it is unlikely. In our experience, same-day imaging has been well received and especially desirable to patients who travel long distances for their care. As the COVID-19 pandemic emerged, we found additional benefit in avoiding a strategy that involved esophageal intubation and developed a same-day discharge protocol as our hospital reallocated resources.¹⁷ Reducing overnight stays for patients undergoing elective procedures in the COVID era may not only reduce resource utilization, but also has the additional benefit of reducing possible exposure. Combining CTA with ICE for LAAO is a viable, successful strategy and may be the preferred option for the future of LAAO.   

Study limitations. This was a small, non-randomized study. Non-consecutive patients were chosen as different imaging strategies were employed over the period of time queried. The CTA + ICE strategy represents our evolved and current approach for LAAO. ICE was performed by experienced operators, which may affect the generalizability of our results.

CTA + ICE is a safe, feasible strategy for LAAO that allows for accurate device sizing and may be used for a same-day discharge protocol in select patients. This approach allows for a streamlined procedure with improved scheduling logistics and may have the added benefit of limiting occupational exposure in the COVID era.

Acknowledgments. The authors are grateful to all nurses, technicians, and valve clinic coordinators who care for our patients, and in particular to Nicole Ritter and Stephen Stoycos for their assistance with the images used in this manuscript.