Predictive Value of the Logistic Clinical SYNTAX Score.

Abstract

The SYNTAX score was developed for use in the SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery (SYNTAX) trial as an objective and comprehensive angiographic tool to grade the complexity of coronary artery disease (CAD) before randomizing patients to coronary artery bypass graft (CABG) surgery or percutaneous coronary intervention (PCI). The primary objective of the scoring system was to define technical complexity and feasibility of PCI. At the conclusion of the trial, additional advantages of the SYNTAX score were appreciated, that is, it was a strong predictor of adverse outcomes in the PCI group and it defined subsets of patients in whom PCI resulted in similar or inferior outcomes compared with CABG. Nevertheless, the SYNTAX scoring system has many limitations, the most important of which is its pure angiographic nature, that is, it does not incorporate clinical variables that influence procedural and patient outcomes. Additionally, it may include lesions that are not hemodynamically significant, and it does not predict outcomes of patients undergoing CABG in the original SYNTAX trial population. Since the development of the SYNTAX score, a number of risk models have been developed to enhance prognostic stratification in complex PCI. While employing the basic angiographic anatomical variables of the original score, the modified risk models incorporate major clinical variables known to influence outcomes of PCI patients in particular and CAD patients in general. The combination of the angiographic SYNTAX score with the clinical variables comprising the European System for Cardiac Risk Evaluation (EuroSCORE) (typically used for patients with CABG) led to the Global Risk Classification (GRC) model. The multiplication of the age, creatinine level, and ejection fraction score and the SYNTAX score resulted in Clinical SYNTAX score (CSS) and its logistic derivative, log CSS. More recently, the proposed SYNTAX score II includes the original anatomic score in addition to 7 clinical variables. These, as well as a few other, combined risk models were designed to optimize the selection of the revascularization approach, provide better prognostic information, and can be individualized rather than placing single patients into risk categories. In this issue of Angiology, Ozturk et al attempt to use log CSS, a combined angiographic/clinical scoring system developed for and validated in PCI patients, for the prediction of saphenous vein graft (SVG) failure plus major adverse cardiac and cerebrovascular events (MACCEs) in patients undergoing CABG surgery. They included 267 patients who were followed up clinically with indicated coronary angiography at varying time points following CABG surgery. In addition to low ejection fraction, a high log CSS was found to be predictive of SVG failure (odds ratio 2.21, 95% confidence interval 1.02-4.75, P 1⁄4 .04). Semiquantitatively, the incidence of graft failure increased in a stepwise fashion in intermediate and high tertiles of log CSS. Both the SYNTAX score and the log CSS were associated with an increased risk of MACCE (P 1⁄4 .001 and P < .001, respectively). Although the sample is small and selected based on clinical events, the use of the log CSS in predicting outcomes of patients with CABG has not been previously reported. Traditionally, risk models of PCI and CABG have been distinct entities that are not used interchangeably. Given the fundamental differences between the 2 revascularization approaches, this is not surprising. Percutaneous coronary intervention is a segmental therapeutic approach in which the feasibility and procedural outcomes are directly linked to the nature and complexity of the target lesion and the target segment. Coronary artery bypass graft surgery involves the creation of a parallel conduit in which the target lesion and segment are not dealt with directly, hence the outcome is more related to the quality of the runoff distal vessel than it is related to the proximal diseased segment. As such, risk models that focus on the lesion characteristics, such as the SYNTAX score, are not expected to provide insight into the outcome of CABG surgery. Despite these