Exploratory Development of a Prognostic Model for Coronary Artery Disease Utilizing CT-FFR Derived Functional Duke Jeopardy Score.

Rationale And Objectives: To explore the prognostic value of the functional Duke Jeopardy Score based on CT-FFR(fDJS) in assessing major adverse cardiovascular events (MACE) in patients with coronary artery disease (CAD).

Materials And Methods: A total of 894 patients with stable CAD with stenosis ranging from 30% to 90%, who underwent CCTA were included in the study. Follow-up was performed to record MACE. The patients were randomly divided into training and validation sets in a 7:3 ratio. In the training set, prognostic analysis was performed and predictive model was constructed using univariable and multivariable Cox regressions and compared the area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI) and integrated discrimination improvement (IDI) of different indicators. The receiver operating characteristic curve, calibration curve and clinical decision curve were used to evaluate the model's discrimination, calibration and clinical efficacy.

Results: The median follow-up period was 33 (16-36) months, during which 167 cases (18.68%) of MACE occurred. Males accounted for 61.52% (550/894) of the cohort, with a median age of 61.92 years. The multivariate Cox regression analysis indicated that DJS (HR: 2.07, 95% CI: 1.17 ∼ 3.68) and fDJS (HR: 4.68, 95% CI: 2.97 ∼ 7.38) were independent predictors of MACE. Using MACE as a standard, fDJS improved the risk re-stratification ability of CT-FFR (NRI:0.993, P < 0.001) and the predictive ability of CT-FFR (IDI:0.101, P < 0.001) and DJS (IDI:0.079, P < 0.001). The prediction model demonstrated high discrimination (training AUC: 0.84 [0.80-0.89]; validation AUC: 0.82 [0.75-0.89]), good calibration and clinical efficacy.

Conclusion: The fDJS was the strongest predictor of MACE. The model constructed based on fDJS has certain clinical utility in prognostic evaluation for CAD.