Potential impact of dynamic automated CT aortic annular measurements on outcomes for transcatheter aortic valve replacement sizing.

Abstract

To determine the potential impact of automated computed tomography (CT) software used for aortic annular sizing for transcatheter aortic valve replacement (TAVR) on paravalvular leak (PVL) and major adverse cardiovascular events (MACE) as compared to standard CT manual measurement. In 60 TAVR patients (84 ± 7years, 60% male), we evaluated the preprocedural CT scans. For the standard manual measurement, we measured the perimeter and area from a single cardiac phase deemed to be of maximum systolic opening. Valve type and size were determined by a multidisciplinary TAVR team per clinical routine. From the dynamic automated software, we determined the aortic annular perimeter and area as the maximum value from an entire cardiac cycle. Valve size was readjudicated by a blinded interventional cardiologist who was provided with valve type and automated values. Clinical endpoints were adjudicated for presence of at least mild PVL and MACE at 30days. There were 16 (28%) patients with PVL and 4 (7%) with 30-day MACE. When reclassifying valve size using dynamic automated values, 12 (20%) patients were undersized and 3 (5%) patients were oversized. Undersized patients were more likely to have mild-to-moderate PVL at 30days (27% vs 4%, p = 0.04) than those not undersized. Of the 5 (45%) undersized patients with at least mild PVL, all were balloon-expandable valves. Automated dynamic CT annular measurements have the potential to reclassify patients with PVL with larger TAVR valve size, particularly balloon-expandable valves.