Author + information
Syntax Score has been developed to determine the complexity of coronary artery disease and to identify patients at risk for major adverse events following coronary interventions. We aimed to study the feasibility of non invasive Syntax Score calculation by Multislice Computed Tomographic Angiography (MSCTA) and compare it with that obtained from Invasive coronary Angiograph (ICA).
Syntax Score was calculated and compared for 91 consecutive patients (55.8 ± 7.8 years; 76 males) who underwent 64 MSCTA and ICA for evaluation of chest pain. The duration between the two examinations was 10 ± 5 days. Post CABG patients and those with bad image quality in MSCTA were excluded.
MSCTA compared to ICA for diagnosis of >50 % stenosis per coronary artery (left main, left anterior descending, left circumflex and right coronary arteries), showed that MSCTA had a specificity of 92.2% (217/235), sensitivity of 97.1 % (125/129), negative predictive value of 98.2% (217/221) and accuracy of 94.3 (364/386). Agreement between modalities was high with a kappa of 0.74. There was a positive correlation between MSCTA and ICA Syntax scores (r 0.73, p = 0.000). The mean Syntax score was 15.8 ± 7.16 for ICA versus 16.3 ± 7.6 For MSCTA (Kappa of Cohen 0.66), p 0.000). Bland Altman plot revealed that: the estimated bias was 1.9 ± 3.4 and the most bias occurred with higher syntax score. Lesions per patient were more identified with MSCTA than ICA (2.5 ±1.4 vs. 1.9 ±1.1, p<0.001), with a good degree of agreement (kappa=0.65). Syntax score per lesion was similar with higher degree of agreement (6.3±5.8 vs. 6.0±4.8, kappa = 0.74 (p<0.001). Calcified lesions were identified to a similar extent (28 vs. 26 cases) with fair degree of agreement (kappa=0.40). Both Bifurcation lesions and total occlusions were identified to a similar degree by MSCTA and ICA, (kappa=0.38 for bifurcation lesions and =0.64 for total occlusion). For lesions identified with both techniques the degree of agreement was higher than total score (6.5±4.8 for ICA vs. 6.9±6.3 for MSCTA, p<0.05), kappa = 0.76).
We found a good degree of agreement between 64 MSCTA and ICA in syntax score calculation. Larger studies with new scanners of MSCT are needed to confirm our results.