Between 2009 and 2020, 88 patients with AF underwent LAAO. Of those, 45 were selected to undergo LAAO with ICE guidance, at the operator's discretion and according to the availability of an anesthesiologist. All patients had a CHA2DS2-VASc score ≥2 and contraindication to OACs (due to high bleeding risk, major or recurrent bleeding history under anticoagulation, bleeding dyscrasia secondary to hematologic disorder or labile international normalized ratio) or inefficacy of OACs as demonstrated by the presence of an LAA thrombus or history of cardioembolic events despite therapeutic oral anticoagulation. Stroke was defined as a focal neurologic defect, with rapidly developing clinical signs and symptoms, lasting for more than 24 hours or leading to death, of probable vascular origin (World Health Organization definition, 2006). Bleeding complications were defined according to the Bleeding Academic Research Consortium criteria (types 2–5).

To characterize the anatomy and morphology of the LAA and to exclude the presence of thrombus, all patients underwent a cardiac computed angiography (CTA) (example in Figure 1) imaging assessment 48 hours before the procedure or TEE with conscious sedation in the cath lab immediately before LAAO, when CTA imaging was not available.

Figure 1.

Pre-procedural cardiac computed angiography scan to assess left atrial appendage morphology and dimensions.

(0.38MB).

Three types of device (Figure 2) were used for LAAO: Amplatzer Cardiac Plug (ACP) or Amulet (St. Jude Medical), WATCHMAN (Boston Scientific), and LAmbre (Lifetech Scientific). Device sizing was mainly based on the pre-procedural CTA scan or TEE assessment, bearing in mind the angiographic and ICE-acquired dimensions.

Figure 2.

Three types of implanted left atrial appendage occlusion devices.

(0.13MB).

All patients underwent LAAO under local anesthesia with or without mild sedation. The procedure was performed using two venous femoral accesses (left femoral for ICE introduction and right femoral for the LAAO device) or two punctures on the same side.

We used the following ICE imaging protocol for LAAO (example images in Figure 3):

• 1.

  The probe is placed in the mid right atrium with clockwise rotation and posterior tilt to better visualize the fossa ovalis and to guide transseptal puncture in an inferoposterior position. In this view, the LAA is visualized in far field to ensure alignment in the anteroposterior plane and direction of puncture toward the LAA.

• 2.

  The transseptal puncture is enlarged with the LAAO device delivery sheath to facilitate the passage of the ICE catheter to the left through the same orifice.

• 3.

  The ICE catheter is placed parallel to the LAO with posterior tilt to guide device delivery. This position is used to guide the LAAO device landing zone with the LAA, assess correct positioning and stability before release, rule out significant peridevice leaks and monitor for possible embolization.

• 4.

  Finally, after LAAO, the probe is retracted to the right side of the heart and is placed in the right ventricle to exclude pericardial effusion.

Figure 3.

Fluoroscopic images of the intracardiac echocardiography probe (black arrows) and its respective echographic images during a left atrial appendage (LAA) occlusion (LAAO) procedure, showing its utility in guiding transseptal puncture (A and B); device deployment (C and D); assessment of device position, significant peridevice leaks using color Doppler and embolization risk (E and F) and exclusion of pericardial effusion (G and H). IAS: interatrial septum.

(0.62MB).

Post-procedural antithrombotic therapy was decided on an individual basis according to bleeding risk. The most frequently used antithrombotic strategy consisted of six weeks of 2.5 mg apixaban twice daily or dual antiplatelet therapy (DAPT) with 75 mg clopidogrel once daily and 100 mg aspirin once daily. Depending on the clinical course, in the absence of significant complications patients were discharged 24–48 hours after the procedure.

All patients underwent transthoracic echocardiography (TTE) before discharge to rule out device displacement, significant peridevice leak or pericardial effusion, and TEE at six weeks after the procedure to assess device endothelization, the presence of device thrombus, peridevice leaks, or significant atrial septal defects. Patients were followed for a mean of 19±10 months, minimum 12 months. No patients were lost to follow-up.

Baseline characteristics, procedural data (regarding technical success, complications and procedure-related parameters such as radiation and contrast use), six-week TEE data and follow-up outcomes (stroke, bleeding and all-cause mortality) were noted retrospectively. Continuous variables are presented as means and standard deviation and categorical variables are presented as absolute and relative values (percentage rounded to one decimal place). The recorded stroke and bleeding rates were compared with the rates predicted using the CHA2DS2-VASc and HAS-BLED scores, respectively. Comparisons between TEE and ICE patients were performed using the Student's t test and the chi-square test, and the results presented as odds ratios (OR), 95% confidence intervals and p-values. The statistical analysis was carried out using IBM SPSS version 25.0.

A total of 88 patients underwent LAAO (43 patients guided with TEE and 45 guided with ICE), mean age 74.9 years, mainly male (68.2%, n=60), and most with permanent AF (71.6%, n=63). The prevalence of CAD was 20.5%, 31.8% had diabetes, and 28.4% had CKD (stages 3–5 of the Kidney Disease Improving Global Outcomes classification). The prevalence of chronic liver disease was 11.4% and 76.1% had hypertension. Mean CHA2DS2-VASc and HAS-BLED scores were 4.0 and 3.6, respectively, predicting a stroke risk of 4.0% per year and a major bleeding risk of 8.7% per year. The most frequent indication for LAAO was a history of major bleeding (65.9%, n=58). Of note, in some patients the contraindication to oral anticoagulation was established due to the presence of esophageal varices (some with previous upper gastrointestinal bleeding). Baseline characteristics, comorbidities, risk scores and the type of device used for occlusion were similar between the TEE-guided and ICE-guided populations (Table 1).

The LAAO devices were implanted successfully in 94% of patients (n=83), and in 93% of the ICE-guided population (n=42). In the ICE group there was procedural failure in three cases: complete occlusion of the LAA could not be achieved with the available occlusion devices in one patient and device embolization to the left ventricle occurred in the other two, of whom one required urgent heart surgery and the other resulted in cardiac arrest and in-hospital death. In the TEE group, complete LAAO could not be achieved with the available occlusion devices in one patient and in the other, a significant gas embolism occurred, leading to myocardial infarction and death.

The major complication rate was 8.8% (n=4), consisting in perforation with pericardial effusion requiring pericardiocentesis in 4.4% (n=2) and device embolization (see example in Figure 4) in 4.4% (n=2, one of them leading to cardiac arrest and death), and no major vascular complications occurred. No other procedure-related deaths occurred. The 30-day mortality rate was 2.2% (n=1). Mean length of hospital stay after the procedure was 5.5 days.

Figure 4.

Images of left atrial appendage occlusion (LAAO) device embolization to the left ventricle, attached to the mitral valve apparatus, causing significant mitral regurgitation. This patient underwent urgent cardiac surgery with device removal and surgical LAAO.

(0.08MB).

All patients were followed for a mean of 19.0 months. During this period, another death (2.2%) occurred from non-cardiac cause, and the rehospitalization rate was 26.7% (n=12), 11.1% (n=5) for cardiac causes. During the same period, one stroke (2.2%) and six significant bleeding events (13.3%) occurred (yearly rates of 1.4% and 8.4%, respectively).

Table 2 summarizes the comparative analysis between ICE- and TEE-guided LAAO. Technical success was achieved in 95.3% (n=41) and 93.3% (n=42) of patients in the TEE and ICE groups, respectively (p=0.69). There was no significant difference in procedure-related complications (major vascular complications, perforation or device embolization) between the TEE and ICE groups (14.0% vs. 8.9%, OR 1.66, p=0.46). Fluoroscopy time was shorter in the TEE group (29.1±13.6 min vs. 44.1±17.4 min, p=0.001), while radiation dose (2761±1555 mGy vs. 3397±2118 mGy, p=0.113) and contrast volume (220.3±104.1 ml, vs. 204.0±100.9 ml, p=0.469) showed no significant differences. TTE at 45 days showed no significant differences between the TEE and ICE groups regarding peridevice leaks (14.0% vs. 24.4%, p=0.212; all of them minor, i.e. less than 5 mm in diameter), device thrombus (2.3% vs. 0%, p=0.99) or persistent iatrogenic atrial septal defects, all mild (4.7% vs. 13.3%, p=0.174). One-year outcomes showed no significant differences regarding stroke (9.3% vs. 4.4%, p=0.186), major bleeding (9.3% vs. 2.2%, p=0.78) or all-cause mortality (9.3% vs. 11.1%, p=0.38) between the TEE and ICE groups, respectively.

This is a single-center study and the small sample size limits statistical analysis. The data collection and analysis were performed retrospectively. The ICE and TEE groups were not randomized and were thus prone to selection bias. As a result, the likelihood of learning-curve complications may have been lower in the ICE group, thereby limiting direct comparisons between the two groups. The recommended post-procedural antithrombotic regimen in not well established and in our study depended largely on physician criteria. This may have had an impact on stroke and bleeding rates.