Overview of prognostic models for major adverse cardiovascular events in patients with acute coronary syndrome

Tóm tắt

Major adverse cardiovascular events (MACE) represent a key outcome in the management of acute coronary syndrome (ACS), requiring prognostic models with strong performance and standardized reporting. This review summarizes studies published from 2000 to 2025 investigating models predicting MACE in ACS, including classical clinical scores, biomarker-based models, nomograms, and machine learning approaches. Traditional models reported moderate discrimination for the TIMI score (C-statistic = 0.65) and higher accuracy for GRACE 2.0 (C = 0.75-0.81). In post-intervention cohorts, the CADILLAC, ACTION, and CAMI scores achieved C = 0.80-0.90. Biomarker-integrated models demonstrated improved performance, with BIPass showing C = 0.79 (95% CI 0.73-0.85). Nomogram-based tools predicting 6-24-month or post-PCI MACE reported AUC = 0.79-0.89. Machine learning models, including deep neural networks and long short-term memory algorithms using dynamic hs-troponin and ECG data, achieved AUC > 0.90 for short-term MACE prediction in emergency department chest-pain cohorts. The findings provide a quantitative overview of the prognostic performance of MACE prediction models in patients with ACS. Overall, the GRACE score remains the cornerstone of risk stratification in NSTEMI-ACS, while nomogram-based and machine learning models show substantial potential for individualized prognostication, and still require local recalibration and external validation to establish their clinical applicability.