全身免疫炎症指数联合左心室射血分数对急性心肌梗死患者冠状动脉病变严重程度的评估价值

Evaluation value of systemic immune-inflammation index combined with left ventricular ejection fraction for the severity of coronary artery lesions in patients with acute myocardial infarction

摘要:

目的]探讨全身免疫炎症指数(SII)联合左心室射血分数(LVEF)对急性心肌梗死(AMI)患者冠状动脉病变严重程度的评估价值。 [方法]回顾性纳入1 963例冠心病患者,其中AMI患者873例,非AMI患者1 090例。AMI组按照Gensini评分三分位数分为轻度、中度和重度病变组。采用Spearman相关分析、限制性立方样条(RCS)、有序Logistic回归(包含部分比例优势模型)、ROC曲线及DeLong检验分析SII与LVEF对冠状动脉病变严重程度的评估效能；构建列线图模型,并通过Bootstrap内部验证与决策曲线分析(DCA)评估模型性能。 [结果]AMI组SII水平高于对照组,LVEF水平低于对照组,且SII随Gensini分级递增而升高,LVEF随Gensini分级递增而降低,差异均具有统计学意义(均P＜0.001)。由于比例优势假设不成立,多因素采用部分比例优势模型,结果显示,SII每增加100个单位与病变严重程度呈正向关联且效应随切点而异(中重度病变对比轻度病变:OR＝1.139,95%CI:1.097～1.181,P＜0.001；重度病变对比轻中度病变:OR＝1.045,95%CI:1.024～1.065,P＜0.001),LVEF与病变严重程度呈负向关联(OR＝0.937,95%CI:0.919～0.956,P＜0.001)。RCS分析结果显示,SII与冠状动脉重度病变呈非线性关联(非线性检验P＜0.001)。在评估冠状动脉重度病变时,SII、LVEF以及二者联合模型的AUC分别为0.674、0.664、0.696,联合模型相较于单独使用SII未带来显著增量(P＝0.225),区分能力有限；纳入四变量的列线图经500次Bootstrap内部验证校正后,AUC为0.707。在鉴别AMI与非AMI冠心病的独立探索性分析中,联合模型的表观AUC为0.910。 [结论]在AMI患者中,SII与冠状动脉病变严重程度呈正相关,且与重度病变呈非线性关联,LVEF与冠状动脉病变严重程度呈负相关,二者联合应用对冠状动脉病变严重程度评估的区分能力未得到显著提升。

Abstract:

Aim To explore the evaluation value of systemic immune-inflammatory index (SII) combined with left ventricular ejection fraction (LVEF) for the severity of coronary artery lesions in patients with acute myocardial infarction (AMI). Methods 1 963 patients with coronary heart disease were retrospectively enrolled, including 873 patients with AMI and 1 090 patients without AMI. The AMI group was divided into mild, moderate, and severe lesion groups according to tertiles of Gensini score. Spearman correlation analysis, restricted cubic spline (RCS), ordered Logistic regression (including partial proportional odds models), ROC curve and DeLong test were used to analyze the efficacy of SII and LVEF in assessing the severity of coronary artery lesion; a nomogram was constructed and its performance was evaluated through Bootstrap internal validation and decision curve analysis (DCA). Results The SII level in the AMI group was higher than that in the control group, and the LVEF level was lower than that in the control group. SII increased with increasing Gensini grading, while LVEF decreased with increasing Gensini grading, and the differences were statistically significant (both P＜0.001). Since the proportional odds assumption was violated, a partial proportional odds model was adopted for multiple factors. The results showed that for every 100 units increase in SII, there was a positive correlation with the lesion severity, and the effect varied with the cut-off point (moderate-to-severe vs mild lesions:OR＝1.9,5%CI:1.097～1.181, P＜0.001; severe lesions vs mild-to-moderate lesions:OR＝1.5,5%CI:1.024～1.065, P＜0.001), LVEF was negatively correlated with lesion severity (OR＝0.7,5%CI:0.919～0.956, P＜0.001). The RCS analysis results showed a non-linear association between SII and severe coronary artery lesions (P for nonlinearity＜0.001). When evaluating severe coronary artery lesions, the AUC of SII, LVEF and their combined models were 0.4,0.664 and 0.696, respectively; the combined model did not yield a significant improvement compared to using SII alone (P＝0.225), indicating limited discriminative ability; after 500 bootstrap internal validation corrections, the AUC of nomogram with four variables included was 0.707. In the independent exploratory analysis of distinguishing AMI from non-AMI coronary heart disease, the apparent AUC of the combined model was 0.910. Conclusions In AMI patients, SII was positively associated with the severity of coronary artery lesions and nonlinearly associated with severe lesions, while LVEF was negatively associated with the severity of coronary artery lesions. The combined application of the two indices did not significantly improve the ability to distinguish the severity of coronary artery lesions.