Calcification- or stent-related artifacts limit interpretability and accuracy of coronary computed tomography angiography (CCTA). These limitations could theoretically be limited by using subtraction CCTA (sCCTA). Therefore, the purpose of this study was to evaluate the diagnostic accuracy of sCCTA and the effect of patient-related factors on sCCTA diagnostic confidencein calcified or stented coronary arteries, using invasive angiography (ICA) as the reference.

We conducted a retrospective study including all consecutive patients who underwent both CCTA and ICA within a 3 months period, from June to December 2016. From the 24 patients (17 males,mean age 62±7 years) who matched these criteria, all segments with both a diameter >1.5mm and either calcifications (n =63) or stents (n =11) on CCTA were selected. Each segment was evaluated for the following image quality markers using Likert scale (from 1 representing no artifact to 4 standing for severe artifacts): motion, beam-hardening, blooming...

On CCTA, 7/74 segments were non diagnostic because of blooming (n = 7) and motion (n =1) artifacts. On sCCTA, 11/74 segments were non diagnostic, all because of misregistration. One stent was non-diagnostic on CCTA, while 5 stents were non-diagnostic on sCCTA. On regression analysis, diagnostic confidence wasdetermined by the patient’s heart rate, the segment’s diameter and the circumferential extent of calcification (all p

Compared to CCTA, sCCTA tend to increase the diagnostic accuracy forstenoses in calcified and stentedcoronarysegments, but the number of non-diagnostic segments increases as well, owing to misregistration artifacts.