Diagnostic accuracy of coronary ct for the quantification of the syntax score in patients with left main and/or 3-vessel coronary disease. Comparison with invasive angiography.

Background: The SYNTAX score helps in the treatment decision in multivessel coronary disease. Coronary computed tomography angiography (CTA) can measure the SYNTAX score but has been used in few patients with multivessel disease. Our aim was to assess the feasibility, accuracy and reproducibility of SYNTAX score with CCTA compared with invasive coronary angiography (ICA) in de novo left main and/or 3-vessel disease.

Methods: 57 patients with new left main and/or 3-vessel disease on ICA and a CCTA performed within the previous month were included. The SYNTAX score was calculated retrospectively for both modalities. Agreement for the global score, vessel score, different components and inter-readers was evaluated with intraclass correlation coefficient (ICC). The ability to classify SYNTAX score categories (low, intermediate and high) was assessed using weighted kappa (K) coefficient.

Results: CCTA-based SYNTAX score showed an acceptable concordance (ICC=0.64) and good correlation (r=0.65, p<0.001) with ICA. ICC per vessel and component ranged from 0 to 0.73. There was good agreement classifying the SYNTAX score categories (K=0.53) and interobserver reproducibility (ICC=0.85). CCTA demonstrated high diagnostic accuracy (0.84) for detecting patients in the high score group. No patient with a high CCTA SYNTAX score had a low risk score by ICA that would suggest benefit from percutaneous revascularization.

Conclusions: CCTA showed good correlation, acceptable concordance, and high reproducibility for the quantification of the SYNTAX score in de novo left main and/or 3-vessel coronary disease. A high CCTA SYNTAX score identified a group of patients less likely to benefit from percutaneous coronary intervention.