Validation of coronary CTA based-British cardiovascular intervention society Jeopardy score (CT-BCIS-JS) against invasive BCIS-JS - a feasibility and reproducibility study

Abstract

Abstract Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): Promedica Stiftung and the Iten-Kohaut Foundation in collaboration with the USZ Foundation Background/Introduction The invasive British Cardiovascular Intervention Society Jeopardy Score (iBCIS-JS) is a simple angiographic scoring system, enabling quantification of the extent of jeopardized myocardium related to clinically significant coronary artery disease (CAD). Ranging from 0 (no significant CAD) to 12 (CAD jeopardizing the whole LV- myocardium), the score is used for the entire spectrum of CAD including left main coronary artery and coronary artery bypass graft disease. A score of ≥6 identifies patients with anatomically high-risk CAD. Non-invasive CAD assessment with coronary CTA is hampered by modest specificity, leading potentially to increased utilization of downstream diagnostic tests. Purpose To validate coronary CTA-based assessment of the BCIS-JS (CT-BCIS-JS) against the iBCIS-JS and to assess its feasibility and reproducibility. Methods A total of 137 patients who underwent clinically indicated coronary CTA and subsequent invasive coronary angiography within 90 days were included in the analysis. Eleven patients were excluded due to suboptimal CTA image quality. CT-BCIS-JS and iBCIS-JS were calculated. Correlations were assessed with the Spearman’s rank correlation coefficient and agreement between scores with Bland-Altman (BA) analysis and with weighted kappa statistics. Inter-observer variability was investigated in a subset of 51 randomly selected patients. Results Median and interquartile range (IQR) of CT-BCIS-JS was 4 (IQR=8) and of iBCIS-JS 4 (IQR=8) (Figure 1). A CT-BCIS-JS of ≥6 and was observed in 56 (44%) patients and an iBCIS-JS of ≥6 in 54 patients (43%). An excellent correlation between the iBCIS-JS and the CT-BCIS-JS was observed (rho=0.97, p<0.0001; Figure 2A) with high agreement (BA bias 0.1 and limits of agreement −2.0 to 2.2; Figure 2B). Kappa analysis demonstrated excellent aagreement for identification of a iBCIS-JS of ≥6 (Kappa Weighted= 0.94, 95% CI 0.87–0.99). Inter-observer analysis revealed excellent correlation (rho=0.98, p<0.0001) and agreement (BA bias 0.12 and limits of agreement −1.33 to 1.57). Conclusions The CT-BCIS-JC represents a feasible and highly reproducible tool for accurate, non-invasive identification of patients with anatomically high-risk CAD. Hence, the CT-BCIS-JC may serve as reliable gatekeeper for invasive testing in the management of patients with CAD.