Risk stratification in unprotected left main coronary disease: do we have the tools?

Abstract

Abstract Introduction The evolution of percutaneous coronary intervention (PCI) and coronary artery bypass grafting (CABG) techniques made the choice of the optimal revascularization strategy of unprotected left main coronary disease (ULMD) challenging. Scoring systems are useful tools for the decision-making process and for risk stratification. Purpose To evaluate 1) the performance of the SYNTAX score I (SSI) and II (SSII) and Euroscore II (EII) in risk stratification and 2) the outcome predictors of patients (pts) with ULMD, according to the treatment strategy chosen (PCI or CABG). Methods Retrospective single centre cohort study of 440 consecutive pts (age 68±11 years; 76.6% male) with significant ULMD (defined as left main coronary artery stenosis >50%, with no patent arterial or venous bypass graft to left anterior descending artery), who were submitted to PCI (n=135) or CABG (n=307), between January 2006 and December 2018. Median follow-up (FU) was 4.0±1.8 years. The primary outcome was a composite of cardiovascular (CV) death, non-fatal myocardial infarction (MI) and target lesion revascularization (TLR). Results During the FU period, there were 112 (25.5%) CV deaths, 26 (5,9%) non-fatal MI and 53 (12.0%) TLR. Multivariate analysis of pts submitted to PCI showed that SSII and anatomical complete revascularization were independent predictors of the primary outcome (HR 1.045, CI 1.015–1.075, p=0.003 and HR 3,014, CI 1.655–5.489, p<0.0001, respectively). The 63 pts submitted to PCI, who had a SSII favoring CABG, had slightly more adverse events (42.9% vs 41.7%, p=0.889). In the CABG cohort, only SSII was an independent predictor of the outcome by multivariate analysis (HR 1.061, CI 1.035–1.086, p<0.0001). The ROC curve analysis for all cohort presented no discriminative capacity for SSI (AUC 0.538, CI 0.482–0.593, p=0.186) and a weak discrimination for SSII (AUC 0.659, CI 0.605–0.713, p<0.0001) and EII (AUC 0.653, CI 0.599–0.707, p<0.0001; Figure 1). The difference between SSII and EII was not statistically significant (DeLong test p=0.828). Similar results were found when analysing the CAGB group, however, in PCI cohort, SII and EII showed an acceptable discriminative capacity (AUC 0.722, CI 0.636–0.809, p<0.0001 and AUC 0.700, CI 0.610–0.791, p<0.0001, respectively). Conclusion In a real-world ULMD population, the most common risk scores, mainly those integrating anatomical and clinical features, presented a very modest role in the risk stratification, both in chronic and acute coronary syndromes. However, in pts with ULMD submitted to PCI, those risk scores had a more significant role in the risk stratification of these pts. Figure 1. SSI, SSII, EII ROC curves for all cohort Funding Acknowledgement Type of funding source: None