TCT-421 Predicting 2 Years All-Cause Mortality After Contemporary PCI: Revising and Updating the Logistic Clinical SYNTAX Score in the GLOBAL LEADERS Study

Author(s) Chichareon P.; Modolo R.; Kogame N.; Takahashi K.; van Klaveren D. et al.
Source Journal of the American College of Cardiology; Oct 2019; vol. 74 (no. 13)
DOI 10.1016/j.jacc.2019.08.512
Background: The logistic clinical SYNTAX score combined anatomic SYNTAX score and 7 variables to predict 2-year all-cause mortality after percutaneous coronary intervention (PCI). The score was derived from 7 stent trials in which new-generation drug-eluting stent and potent antiplatelet inhibitors were not widely utilized. The original logistic clinical SYNTAX score has been shown to overestimate 2-year all-cause mortality in the GLOBAL LEADERS study. This study aimed to revise and update the logistic clinical SYNTAX score to predict 2-year all-cause mortality after contemporary PCI using the population in the GLOBAL LEADERS study.

Method(s): The authors analyzed 15,883 patients in the GLOBAL LEADERS study who underwent PCI. The original model of the logistic clinical SYNTAX score was revised by re-estimating the predictor effects in Cox regression model (revised model). The authors updated the score by combining the original model with additional independent predictors of 2-year mortality identified in the GLOBAL LEADERS study population (updated model). Internal validation with bootstrap method was used to quantify and correct for overfitting and optimism.

Result(s): This study has identified 5 additional independent predictors of 2-year all-cause mortality: hemoglobin level; white blood cell count; chronic obstructive pulmonary disease; prior stroke; and current smoker. The updated model (C-index: 0.78; 95% confidence interval [CI]: 0.76 to 0.80) better discriminated 2-year all-cause mortality than the previous version (C-index: 0.75; 95% CI: 0.73 to 0.77), in an internal validation using bootstrap method. The updated model performed better in differentiating mortality among non-high-risk patients. The calibration of both models was appropriate. Currently the updated model is tested in an external validation cohort of 10,724 patients.

​Conclusion(s): By adding 5 predictors to the logistic clinical SYNTAX score, the score better predicts 2-year all-cause mortality after PCI. The score can be used for risk stratification and prediction in patients undergoing PCI in the contemporary practice. Categories: CORONARY: Stents: Drug-ElutingCopyright © 2019

• Available in full text at Journal of the American College of Cardiology from Unpaywall