RESUMO
BACKGROUND: Amplatzer Amulet is a frequently used device for left atrial appendage occlusion (LAAO). The current sizing protocol is based on the maximum diameter of the left atrial appendage (LAA) landing zone. However, mean, perimeter-, or area-derived diameter might be more accurate measures for device sizing. METHODS: Retrospective analysis of 150 consecutive patients undergoing LAAO is guided by pre-procedural cardiac CT. A total of 117 patients were included; 7 were excluded due to renal failure and 26 due to closure with the sandwich technique. The maximum, mean, area-, and perimeter-derived diameters of the landing zone were derived from pre-procedural cardiac CT scans, and their accuracy to predict the implanted device size was investigated. The predicted device size was determined based on the currently recommended sizing algorithm. Peri-device leak (PDL) was assessed (grade 1-3) along with the underlying mechanism. RESULTS: Device-sizing accuracy was superior for mean, area-, and perimeter derived diameters compared with the maximal diameter, especially for eccentric landing zones. Mean difference between predicted and actually implanted device size was 0.08 mm (± 2.77), 0.30 mm (± 2.40), - 0.39 mm (± 2.43), and - 2.55 mm (± 2.57) across mean, area-derived, perimeter-derived, and maximal diameter, respectively. Grade 3 peri-device leak was seen in 8.5% of implants without a significant association to the eccentricity of the landing zone. The leading mechanism for PDL was device malalignment. CONCLUSION: Our results indicate mean, area-, and perimeter-derived diameters of the device landing zone to perform similar and superior in device-sizing accuracy compared with the maximum diameter.