Routine application of three-dimensional transesophageal echocardiography-based device selection for transcatheter atrial septal defect closure

Background A standardized global framework for device size selection in transcatheter atrial septal defect (ASD) closure is lacking. Purpose We aimed to evaluate the outcomes and insights from the routine application of a 3D transesophageal echocardiography (TEE)-based device selection strategy in consecutive patients undergoing ASD closure. Methods The study population was derived from the second phase of the prospective ASD_3D_Device_Size registry, where device sizing was exclusively guided by a 3D TEE-based formula (0.964×maximum diameter–2.622×circular index+7.084). Computational simulations for oval ASDs (circular index ≥ 1.5) were conducted using a U-Net-based segmentation algorithm. All patients were clinically followed for a median of 1.6 years after the procedure. Results Between September 2016 and May 2024, a total of 748 patients (median age, 47.0 years) were included. The median estimated device size was 19.6 (IQR 15.5–24.6). Interobserver agreement in device size estimation between the junior and senior physicians was excellent, with an intraclass correlation coefficient of 0.95 (95% CI: 0.93–0.97). The median procedural and fluoroscopy time was 18.0 (13.0–24.0) and 3.7 (2.9–5.3) minutes, respectively. One patient required intraprocedural device size reselection, and three experienced device embolization, two of which were attributed to delivery system malfunction. Consequently, procedural success rate was 99.6%, and resizing or complication caused by device size selection occurred in 0.27% of patients. During follow-up, one patient required surgery for cardiac erosion and another for improper position of the device caused by unexperienced decision. Among 186 (24.9%) oval ASDs, long axis orientations were horizontal in 21.4%, anterosuperior in 36.0%, posterosuperior in 16.9%, and vertical in 25.8% (Figure). Computational simulations demonstrated that, assuming balloon sizing in the fluoroscopic AP projection, horizontally-oriented oval ASDs had the largest degree (31.7–36.9%) of maximal diameter underestimation. Conclusions Routine use of a 3D TEE -based sizing strategy for ASD device closure is safe and effective. Considering diversity of maximal diameter directions, balloon-sizing of ASD using fluoroscopy might underestimate the maximal diameter of ASD which can be accurately measured by 3D TEE images.