Methods
This prospective, single-center study involved patients aged ≥18 years with documented drug-resistant non‑valvular AF at high risk of stroke treated by ICE-guided AF catheter ablation and LAAC. Informed consent was obtained from all patients, and the study procedure was approved by the ethics committee of the first affiliated hospital of Wenzhou medical university.
Catheter ablation was performed prior to LAAC under conscious sedation with fentanyl and midazolam. The procedure was performed via both the left and right femoral veins. A decapolar coronary sinus electrode (Abbott, Saint Paul, MN, USA) was advanced through the left femoral vein access using a long 25-cm 7-Fr insertion sheath. Double right femoral vein access was obtained for introducing the 10-Fr ICE catheter (SoundStar; Biosense Webster, Irvine, CA, USA) and transseptal puncture. The geometry map of the left atrium (LA) including the LAA was created by ICE probe manipulation at the right median atrial septum and at the entrance of the coronary sinus using a three-dimensional mapping system (CARTO 3; Biosense Webster). Single transseptal puncture was guided by ICE. A contact force-sensing catheter (Thermocool SmartTouch; Biosense Webster) supported by a steerable sheath (Agilis NxT; Abbott) was used to perform circumferential pulmonary vein isolation (PVI). The ablation power was set to 45 W with a target ablation index (AI) of 450 for the anterior wall and 40 W with a target AI of 350 for the posterior wall. The Visitag (CARTO3; Biosense Webster) settings included: inter-lesion distance ≤5 mm, minimum force 5 g, force‐over‐time 50%, minimum time 3 seconds, maximal range 2 mm, and lesion tag size 2 mm (Figure 1). In patients with persistent AF, cardioversions were delivered to restore sinus rhythm after PVI. Substrate mapping and additional ablation were then conducted.
A 6-Fr pigtail catheter was advanced to the LAA to perform LAA angiography, with LA pressure of ≥12 mmHg, at right anterior oblique and caudal projections. The Agilis NxT sheath was further exchanged with a 10-Fr delivery sheath to pre-dilate the transseptal puncture site. The delivery sheath was then retracted into the inferior vena cava while keeping the guidewire in the left superior pulmonary vein (LSPV) to allow the ICE catheter advance into the LA through the same site, which was tagged on the image guided by the CARTOSOUND system without fluoroscopic exposure. The echocardiographic LAA ostium and landing zone diameters were measured using the “Four Long-Axis Views around Orifice” (FLAVOR) approach at 90°, 135°, 0°, and 45° (Figures 2 and 3). The delivery sheath was cautiously advanced again to the LA over the guidewire. An appropriately sized LAmbre device was selected based on both angiographic and echocardiographic measurements. Fluoroscopy- and ICE-guided LAAC with the LAmbre device was performed using our previously described modified method.11 The position of the device and the degree of residual peri-device flow were assessed with ICE (Figure 3). The device was released when the COST criteria were met (umbrella deployed beyond Circumflex artery, umbrella fully Open, optimal peri-device Sealing (leak of ≤3 mm), and device stability confirmed by Tug test) and the cover of the device was not impinging the mitral valve and/or pulmonary vein ostium. Otherwise, recapture and/or resizing of the device was attempted. Repeated ICE assessment was performed after release of the device.
The constructed LAA geometry map on the anteroposterior projection was adjusted rightward and upward until the LAA orifice was adequately exposed. Guided by Carto-Sound feature, the ICE probe was manipulated to achieve the proposed four imaging planes (the angle between adjacent imaging planes was approximately 45°) for comprehensive evaluation of the LAA or device. Each plane was adjusted to show the long-axis view of the LAA or device. The ICE probe was initially advanced to the top of the left atrium with slight posterior flexion to obtain the first view similar to the 90° TEE view which we defined as the 90° ICE view showing the LAA, short-axis view of the left circumflex artery [LCX] and mitral valve. The second view similar to the 45° TEE view which was defined as the 135° ICE view could be achieved through rightward and anteflexion of the ICE probe. The ICE probe was usually placed at the entrance of the LSPV. The pulmonary vein ridge, LAA, short-axis view of the LCX and short-axis view of aortic sinus were most commonly visualized in this view. Next, the ICE probe was frequently positioned at the entrance of the left inferior pulmonary vein (LIPV) on the LA posterior wall by further rightward flexion to form the third view similar to the 0° TEE view which was defined as the 0° ICE view with excellent display of the LAA, short-axis view of the LCX, short-axis ascending aorta and concomitant pulmonary artery. Finally, the ICE catheter was rotated anticlockwise and advanced to the mitral isthmus with posterior flexion to form the fourth view resembling the 135° TEE view which was defined as the 45° ICE view. This view showed the LAA and the pulmonary artery. This systematic approach allowed comprehensive evaluation of the LAA or device by mimicking TEE echo windows (Figures 2 and 3).
Outpatient follow-up was scheduled at 3, 6, 12 months post-procedure with a detailed documentation of clinical events. A standard 12-lead electrocardiogram and 24h Holter recording was performed at each visit and 7-day Holter monitoring was obtained at the 12-month follow-up for recurrence of atrial tachyarrhythmia. TEE was conducted at 2 and 12 months after implantation to assess LAA occlusion and rule out thrombi. Successful sealing of the LAA was defined as complete occlusion or residual peri-device flow of ≤3 mm. If the sealing criteria were met, anticoagulation was interrupted at 3 months, followed by dual antiplatelet therapy for 3 months and finally lifelong single antiplatelet therapy.
The primary efficacy endpoints were (1) the rate and degree of residual peri-device flow, (2) iatrogenic atrial septal defect (ASD) at the TEE evaluation, and (3) the combined endpoint of stroke/transient ischemic attack (TIA)/systemic embolism (SE) over the 12-month follow-up. The primary safety endpoints were (1) assessment of bleeding events, (2) all-cause mortality, and (3) the fluoroscopy time during the procedure. The secondary endpoints included (1) AF/atrial tachycardia-free survival, (2) periprocedural complications, and (3) device-related thrombosis (DRT). AFrecurrence was defined as any documented AF, atrial flutter, or atrial tachycardia lasting >30 seconds after a 3-month blanking period.
All statistical analyses were performed using SPSS version 20.0 (IBM Corp., Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation. Categorical variables are reported as median with interquartile range. A P-value of <0.05 was considered statistically significant.