Real-world insights into coronary CTA prognostication: value of semiquantitative scores

Purpose Several semiquantitative coronary computed tomography angiography (CCTA) scores including different parameters describing stenosis degree, plaque burden and plaque features have been developed for diagnostic and prognostic purposes. However, their clinical application is still limited. The aim of the study is to assess the prognostic implication of semiquantitative coronary CCTA scores in clinical practice. Material and methods In this retrospective, single-center study, 6818 adults who underwent elective CCTA between 2016 and 2020 were screened. A total of 1878 patients were enrolled based on low-to-intermediate pretest risk, good image quality (Likert score >= 4), and >= 3 years of follow-up. Clinical data were collected, and the following CCTA scores were calculated: CACS, CAD-RADS, Leiden, CT Leaman, SSS and SIS. Prognostic performance for five-point and two-point MACE composite outcomes was evaluated using survival analysis and multivariable models. Results Over the follow-up period, 10% (187/1878) experienced a five-point MACE and 5.4% (102/1878) a two-point MACE. All CCTA scores stratify the risk of MACE, and all CCTA scores were predictors of outcome with HR increasing as the score increases, with the highest HR in case of severe CAD-RADS (five-point MACE composite outcome: HR 16.35, 95%CI [7.5-35.61]; p < .001 and two-point MACE composite outcome: HR 19.49, 95%CI [6.01-63.2]; p < .001). CAD-RADS outperformed other scores in multivariable models including age, sex and cardiovascular risk factors with a C-index of 0.75 for five-point MACE and of 0.78 for two-point MACE, always p < 0.001. High-risk features were not predictors of outcome. Conclusions CAD-RADS is the most effective CCTA-derived score for MACE prediction in real-world application.