Safety and efficacy of left atrial appendage occlusion with the ACP or Watchman device guided by intracardiac echocardiography from the left atrium

Abstract Background There is a paucity of randomized data regarding the safety and efficacy of the use of intracardiac echocardiography (ICE) from the left atrium (LA) to guide left atrial appendage occlusion (LAAO) procedures under local anesthesia using either of the available devices. Hypothesis The aim of this study was to compare the efficacy and safety of ICE from the LA with transesophageal echocardiography (TEE) for guidance during transcatheter LAAO procedures. Methods Single‐center, cohort study of patients undergoing LAAO with the Amplatzer Cardiac Plug or Watchman device. Procedures were guided by ICE from the LA with local anesthesia (n = 175) or TEE under general anesthesia (n = 49). Efficacy outcomes were procedural success and peri‐device leaks 6 weeks after LAAO. The safety outcome was a composite of procedure‐related complications. Results Procedural success was similar between groups: 100% in the TEE‐guided group, and 98% in the ICE‐guided group. Procedure‐related complications such as death, embolism, migration, or major vascular complications occurred similarly between groups (p = 0.590). The rate and degree of peri‐device leaks or presence of a thrombus on the device did not differ between groups on follow‐up CT. Turnover time in the catheter laboratory and use of contrast agent were reduced with ICE. Conclusions ICE in the left atrium to guide LAAO procedures appears to be as effective and safe as TEE. There was no increase in procedure‐related complications, whatever the device used. ICE resulted in similar procedural success while decreasing procedure time and requiring only local anesthesia.


| INTRODUCTION
The left atrial appendage (LAA) occlusion technique is safe for the prevention of ischemic events in patients with non-valvular atrial fibrillation (AF), and is an increasingly common procedure. 1,2 Efficacy and safety has been demonstrated in randomized clinical trials 3 and several observational "real-world" studies. 4,5 While transesophageal echocardiography (TEE) under general anesthesia remains the most frequently used approach for ultrasound guided implantation, 6 recent papers have reported that intracardiac echocardiography (ICE) can be used to guide LAAO under local anesthesia. The use of ICE from the right atrial (RA) position has been previously described, 7-10 but the visualization of the LAA has been inconsistent and suboptimal because of the right-sided position of the probe, especially when the atria are dilated. The left atrial position of the ICE to guide LAAO was recently found to be a suitable approach, and has thus been compared with TEE. The results suggest that this new approach is effective and safe, and does not increase procedure-related complications. 11,12 However, the authors suggest that there is a learning curve, which indicates the need for a real long-term development policy for this technique. Moreover, these studies evaluated the feasibility and efficacy of ICE-guided LAAO closure using a single device (either the Amplatzer Cardiac Plug/Amulet device or the Watchman device) and with a short follow-up period. Thus, we sought to compare the clinical efficacy and safety LAAO procedures guided by either ICE or TEE, using both available devices, with in-hospital and mid-term follow-up.

| PATIENTS AND METHODS
The participants in this non-randomized study were prospectively recruited between January 2014 and April 2019. Initially, all patients underwent LAAO with TEE guidance (TEE cohort). In March 2015, after 10 procedures had been performed with a dual imaging modality (ICE and TEE), the ICE technique was used alone for all subsequent LAAO (ICE cohort) to increase patient comfort and facilitate procedural logistics. The dual technique imaging procedure patients were integrated into the TEE cohort. Patients eligible for LAAO were known to have non-valvular AF (chronic, persistent, or paroxysmal), an increased risk of stroke (elevated CHA 2 DS 2 -VASc score) 13 with a history of or predisposition for bleeding (elevated HAS-BLED score), 14 along with an absolute or relative contraindication for OAC.
We excluded patients with the following criteria: presence of an intracardiac thrombus on the preprocedural TEE or CT (n = 4), history of previous implanted device for atrial septal defect or patent foramen ovale (n = 2). This study complied with the Declaration of Helsinki and was approved by the ethics committee of the University Hospital of Dijon. All patients gave written consent before participation.

| Procedure and transseptal puncture
The LAAO technique using the Watchman or ACP devices did not differ between ICE and TEE guidance, except for the fact that all TEE-guided interventions were performed on intubated patients under general anesthesia compared with only local anesthesia for ICEguided LAAO.

| ICE and ICE-guided LAAO
All patients, regardless of TEE or ICE guidance, underwent cardiac CT before LAAO for anatomic analysis, device sizing, and exclusion of LAA thrombus. ICE imaging was performed using the 9-F ViewFlex Xtra ICE

| Data collection
We collected data relative to comorbidities, including type of AF, prior congestive heart failure, history of CAD, CABG, or TAVR, cardiovascular risk factors such as history of hypertension or treated hypertension, and diabetes. The CHA 2 DS 2 -VASC and the HAS-BLED risk scores were calculated with the standard admission variables 13,14 to estimate individual annual risk of bleeding or stroke. Blood samples drawn at admission were used to assess plasma N-terminal pro B-type natriuretic peptide (NT-proBNP) and creatinine levels. Creatinine clearance was calculated with the Cockcroft formula. Outcomes gathered during and after the procedure followed the Munich consensus paper. 15 Procedural success was defined according the Munich consensus, which specifies that the device should be implanted in the correct position without device-related complications and no peri-device leaks >5 mm on color Doppler. Following implantation, a dual antiplatelet regimen (clopidogrel 75 mg and aspirin 75 mg daily) was prescribed for a 4-week period, and aspirin was maintained at the same dose. The following safety outcomes were gathered during the in-hospital period: major periprocedural complications including stroke (hemorrhagic and ischemic) or transient ischemic attack, pericardial effusion requiring drainage, device embolization, major bleeding or access-related complications. The following clinical endpoints were recorded at a median of 326 (145-869) days after LAA exclusion in our cohort: death, stroke (hemorrhagic and ischemic) or transient ischemic attack, major bleeding, late device embolization and heart failure symptoms during the follow-up visit.
Finally, the rate and degree of residual peri-device leaks on TEE or CT-scan performed between 6 and 8 weeks after LAAO procedure were categorized as follows: no leak, small (< 5 mm) or large (>5 mm). No patient was lost to follow-up.

| Statistical analysis
All analyses were performed using Sigmastat Software (Jandel Inc.).
The results are expressed as means ± SD or median values (25-75th percentile) for continuous variables, or as percentages for qualitative variables. The normality of distribution for each variable was analyzed using the Kolmogorov-Smirnov test. Comparisons between the two   Mean age was 76 years, and both groups were predominantly male.
The primary indication for LAAO was prior major bleeding events in 91% and 84% of the ICE and TEE groups, respectively. Baseline characteristics were comparable between groups (Table 1).
Procedural success rates with the device delivered to the appropriate position during the initial procedure was high in both groups, 97% for ICE and 100% for TEE (p = 0.895) ( Table 2). There was no difference according the device used.
Cardiac CT interpretation, which was used to select the best device for to the anatomy and diameter of the LA, improved over time. This explains why the first device selected was implanted in 88% of TEE-guided procedures and 96% of ICE-guided procedures (p = 0.566). Though the fluoroscopy time was similar between groups, the overall time spent in the catheter laboratory was significantly reduced with ICE guidance (p < 0.001), facilitating the execution of more LAAO procedures. Finally, the quantity of iodinated contrast media used was significantly decreased with ICE (p = 0.028). The length of hospital stay was similar between groups, except for the fact that patients from the TEE group spent time in the post-anesthesia care unit before returning to the ward, whereas the ICE group returned directly to the ward.
The complete exclusion of the left appendage, evaluated either by CT or FU TEE was also comparable between groups during post-procedure evaluation.

| Procedural adverse outcomes
Procedural serious adverse events (SAE) are described in Table 2.
In the TEE group, cardiac tamponade occurred in two patients: the first was percutaneously drained, and the second occurred before the device was implanted and required emergency surgery during which the left appendage was surgically closed. In the ICE group, cardiac tamponade occurred in one patient, and aortic puncture occurred in another patient during the transeptal step, leading to surgical closure of the left appendage. Another patient experienced device migration with embolization at the end of the procedure, requiring successful percutaneous retrieval and immediately followed by re-implantation. In-hospital SAE occurred in the same proportion between the two groups (Table 2). Interestingly, we observed the same proportion of significant hematoma at access the site but without surgical repair.
Pericardial effusion with tamponade occurred equally in the two groups (p = 0.532). Note: Data are presented as n (%),or Mean ± SD. SAE: Death, stroke, embolism, major bleed, device migration, major vascular complication.

| Follow-up adverse outcomes
During the clinical follow-up, rates of stroke/TIA, major bleeding or device migration were similar between groups. In addition, we found no significant difference in mortality at 30 days of follow-up in both groups (1.5% vs 1.8%). Finally, we found a trend toward a higher rate of heart failure symptoms in the TEE group compared to the ICE group (16.7% vs. 6%, p = 0.053).

| DISCUSSION
In most cases and centers, the evaluation of the LAAO procedures have been done using TEE imaging. The key to successful and safe procedures is an understanding of the transseptal puncture, LAAO size selection for the device-specific landing zone, and post- Moreover, we improved out approach over time by favoring the use of the same sheath to move the two systems, all while using a vascular closure system in order to improve patient recovery.

| Study limitation
Our study has several limitations. First, we described a single center experience, so the efficacy of our approach needs to be proven in larger and multicenter cohorts. Furthermore, we excluded patients with a history of atrial septal defect or PFO closure because the manipulation of catheters through the implanted device could be dangerous and induce migration. Imaging guidance was not randomly chosen, but the cohort consisted of two consecutive groups.
When learning a new procedure, performance tends to improve with experience, which explains why the results are slightly different for the first procedures with any approach. As the first group underwent TEE guidance, a learning curve may be in favor of the ICE group. On the other hand, a learning curve associated with introduction and handling of the ICE probe is also expected. Nevertheless, from the introduction of ICE approach, it was used almost exclusively in our center. Consequently, our two groups have unequal numbers with potential biases linked in particular to the small number of the TEE group. Because the device sizing was based on preprocedural CT analysis, TEE and/or intraprocedural angiography and not on ICE imaging, no comparisons between ICE and TEE could be made for LAA measurements. Finally, the results are currently limited to procedural, in-hospital outcomes and one-year follow-up, but long-term follow-up is ongoing.

| CONCLUSION
Our study provides additional data suggesting that ICE guidance for LAAO, whatever the device used, is feasible and comparable to TEE guidance. Furthermore, ICE guidance precludes the need for general anesthesia or sedation and has the potential to become a preferred imaging modality for LAAO. Left atrial navigation of the ICE probe may also potentially be used to assist other technique such as transcatheter mitral valve replacement.