Coronary evaluation as well as radiographic visualization of the aorta and aortic annulus are crucial in patients that are evaluated for a TAVI procedure. Today, this is routinely accomplished by CT and invasive coronary angiography. 

To assess the feasibility of exclusive CT imaging for aortic and coronary evaluation in a routine TAVI CT vs. default combination of CT and invasive coronary angiogram.

In this single center study 458 patients that underwent TAVI from 2018 to 2019 were analyzed. All patients received a standard pre-TAVI CT designed for annulus and aortic visualization as well as coronary angiogram via cardiac catheterization. CT was performed in absence of specific heart rate control. All scans were performed using the Aquilion One Genesis Edition volumetric CT (Canon Medical Systems). The coronary artery disease reporting and data system (CAD-RADS) was used to rate coronary alteration. Evaluation of the coronary tree was limited to the left main (one segment), RCA (three segments), LAD (three segments) and LCX (two segments). Segments with previous stent implantation were marked as not assessable in the CT. Weighted Kappa was used to compare coronary evaluation in CT and invasive coronary angiography. Invasive coronary angiography was used as the common reference.

Results: 

Of 458 patients, 181 (39.5%) presented with no coronary artery disease (CAD), 93 (20.3%) with one vessel CAD, 73 (15.9%) with two vessel CAD and 111 (24.2%) with three vessel CAD. 199 (43.5%) patients had previous diagnosis of atrial fibrillation. 148 (32.3%) of the patients received beta-blockers, 74 (16.2%) received calcium channel blockade and none ivabradine. Mean heart rate was 70.7 bpm (CI 95%: 65.5 – 76). A total of 4122 segments were evaluated, with 27.8% deemed not assessable via CT. 328 (71.6%) of all patients had at least one segment that was not evaluable because of previous stent implantation, low visualization quality or both (51, 199, and 78 patients respectively). 49 of 181 (27.1%) patients were correctly identified free of CAD by CT. 81 of 277 (29.2%) patients had correct diagnosis of CAD. 27 patients presented with significant stenosis, of these 10 patients were not assessable by CT because of previous stent implantation (5 patients) or because of motion artefacts (5 patients). 17 patients with significant stenosis were correctly detected by CT. In assessable CT scans overall sensitivity for significant stenosis was 100% and specificity for CAD 27.1%. In all segments weighted Kappa showed lowest concordance between CT and invasive coronary angiography for non significant stenosis (CAD RADS 1-3) in distal segments, while best concordance was found for significant stenosis in proximal segments.

Due to its low specificity of 27.1% for CAD, the routine TAVI CT scan had limited potential to dispense with invasive coronary angiography. However, due to its high sensitivity, this seems safe as long as the CT scan is completely assessable.