Feasibility of C-arm computed tomography for transcatheter aortic valve replacement planning

The C-arm used for fluoroscopy during transcatheter aortic valve replacement (TAVR) may also be used to acquire 3-dimensional data sets similar to multidetector row CT (MDCT).

The aim of this study was to evaluate the feasibility of C-arm CT (CACT) for aortic annulus and root (AoA/R) measurements in TAVR planning compared with MDCT.

Twenty patients who were studied for TAVR underwent MDCT and CACT. Two independent observers measured predicted perpendicular projection to annular plane, diameters of the aortic annulus, sinus of Valsalva, sinotubular junction and ascending aorta, distance of coronary ostia to annular plane, sinus of Valsalva height, and leaflet length. Correlation between MDCT and CACT and interobserver variability were analyzed.

MDCT and CACT showed strong correlation for all the measurements of the AoA/R (r ranging from 0.62 to 0.94; P between <.001 and .042) and also for the predicted perpendicular projection (left/right anterior oblique: r = 0.96, P = .002; cranial/caudal: r = 0.83, P = .043). Interobserver variability analysis showed disagreement for the measurements of the aortic annulus structures with CACT (intraclass correlation coefficient [ICC], <0.25) but not for the rest of the variables (ICC between 0.47 and 0.97). MDCT showed no interobserver variability for all the measurements (ICC between 0.45 and 0.93).

CACT showed strong correlation with MDCT for the measurement of all AoA/R structures. However, CACT showed also important interobserver variability for the assessment of the aortic annulus. Therefore, valve sizing may not be reliably performed on the basis of CACT measurements alone.