Results
From April 2019 to April 2020, 56 patients (57.1% male; mean age, 69.4 ± 7.5 years) underwent combined radiofrequency ablation and LAAC with ICE guidance. The patients’ baseline demographic characteristics are summarized in Table 1. In terms of stroke risk, the median CHA2DS2-VASc score was 4.0 (3.0–5.0); in terms of hemorrhage risk, the median HAS-BLED score was 3.0 (2.0–3.0). Ischemic stroke or TIA occurred in 46.4% of patients.
The acute ablation endpoints and device implantation were achieved in all patients. First‐pass PVI was achieved in 49/56 (87.5%) patients. The mean fluoroscopy time was 6.3 ± 5.3 minutes with zero fluoroscopic exposure during radiofrequency ablation. The mean radiofrequency ablation time was 14.6 ± 2.7 minutes. The mean time for assessment of the device position and peri-device leakage with the FLAVOR approach was 2.1 ± 0.4 minutes. After device release, ICE showed complete sealing of the LAA in 46/56 (82.1%) patients, and the remaining 10/56 (17.9%) patients had minimal residual peri-device flow of ≤3 mm. The mean post-detachment peri-device leakage was 1.6 ± 0.5 mm, with the majority of leaks occurring in the 45° or 90° ICE view (Table 2, Figure 4). The total periprocedural complication rate was 3.6%, including one case of pericardiac tamponade that occurred during transseptal puncture and was resolved with pericardiocentesis and one case of a groin hematoma. No stroke/TIA/SE, device embolization, or death occurred.
During a mean follow-up of 12 months, 77.8% of patients were free from AF recurrence. Five patients underwent repeated catheter ablation and maintained sinus rhythm until the last visit. A first follow-up TEE was performed at 2 months post-procedure in 49/56 (87.5%) patients. LAA occlusion was complete in 26 patients, while small peri-device leakage of ≤3 mm was present in the remaining 23 patients. A total of 48/56 (85.7%) patients underwent TEE and 2 other patients who refused repeat TEE exam underwent cardiac CT angiography at least 12 months post-procedure. A complete or satisfactory (residual peri-device flow of ≤3 mm) seal of the LAA was found in 32 and 18 patients, respectively. In accordance with the findings from ICE assessment after release, subsequent TEE evaluation showed similar sites of residual peri-device flow (commonly detected in 90° or 135° TEE views, which resembled the 90° and 45° ICE views) (Table 3). Two DRTs which were absent at 2-month TEE were detected at the 12-month TEE follow-up. The patients were subsequently continued on novel oral anticoagulant therapy. One patient died of a hepatoma 10 months after the procedure.
All 56 patients were discharged with an oral anticoagulant (rivaroxaban in 41, dabigatran in 10, and warfarin in 5). After the 3-month follow-up, oral anticoagulant therapy was discontinued in all of the implanted patients. At 12 months, almost all patients were still withdrawn from oral anticoagulants except two patients with DRTs.
Notably, there was no evidence of thromboembolic events during 12 months of follow-up despite an expected stroke rate of 4.8% without anticoagulation12 (Figure 5). Only one patient developed severe gastrointestinal bleeding, representing a 68% relative risk reduction for major bleeding events based on the expected bleeding rate according to the HAS-BLED score (1.8% vs. 5.7%).13
An iatrogenic ASD was detected in 57.9% of patients at 2 months; this proportion decreased to 4.2% at 12 months.