Validation of the GRACE 3.0 score and redefinition of the risk threshold for early invasive treatment in non-ST-segment elevation acute coronary syndromes: a modelling study from five countries

Abstract

Abstract Background Clinical use of the GRACE scoring system is recommended across international guidelines to guide early treatment stratification in patients with non-ST-segment elevation acute coronary syndromes (NSTE-ACS). Recently, the machine learning-based GRACE 3.0 score was derived from patients with NSTE-ACS undergoing contemporary treatment approaches. External validation studies and the reassessment of clinical risk categories are lacking. Purpose We aimed to evaluate the predictive performance of the GRACE 3.0 score and to explore clinically meaningful risk groups in contemporary patients with NSTE-ACS. Methods We studied the GRACE 3.0 score in 8070 patients with NSTE-ACS in contemporary ACS cohorts from Denmark (VERDICT, n=2147), Germany (Heidelberg-ACS, n=2034), Italy and Spain (CORALYS, n=1650), and Switzerland (SPUM-ACS, n=2239). Heterogeneity in the treatment effect of very early invasive management (within 12 hours) or standard invasive management (within 48 to 72 hours) on the primary composite outcome of all-cause death, nonfatal recurrent myocardial infarction, hospital admission for refractory myocardial ischemia, or hospital admission for heart failure in relation to baseline mortality risk was assessed in 2147 patients enrolled in the VERDICT trial who were randomized to different timing of invasive treatment. Results The GRACE 3.0 score showed excellent discriminatory properties with an area under the receiver operating characteristic curve (AUC) for in-hospital death of 0.89 (95% CI, 0.85–0.93) in the total study cohort. Similar results were observed in Denmark (AUC 0.93, 95% CI, 0.88–0.98), Germany (AUC 0.90, 95% CI, 0.85–0.94), and Italy and Spain (AUC 0.85, 95% CI, 0.77–0.94), and Switzerland (AUC 0.93, 95% CI, 0.86–1.00). Early invasive treatment reduced the composite endpoint (absolute risk reduction: 0.12, 95% CI, 0.03–0.21, P=0.011) in patients at high risk (≥1.7%), but not in those with lower risk profiles (P interaction=0.033). Based on this new high-risk threshold, GRACE 3.0 stratified 475 (42.2%) more patients into the high-risk group, as compared to traditional GRACE risk categories (P<0.001). Conclusion GRACE 3.0 shows unprecedented predictive performance in patients with NSTE-ACS. In the context of a comprehensive clinical evaluation, GRACE 3.0 can support clinical decision making for the timing of invasive treatment. The high-risk group of patients with NSTE-ACS who benefit from early invasive treatment is substantially larger than previously described.ROC curve and risk groupsRestratfication towards high risk group