文献速递心血管MR

心梗后CMR疤痕空间特征对MACE的预后价值

Prognostic Value of Scar Spatial Properties on CMR for Major Adverse Cardiovascular Events After Myocardial Infarction.
PubMed 文献 · Acad Radiol · 2026-07-30

该回顾性研究纳入心肌梗死后经ADAS 3D后处理确认存在边缘区通道的病例,基于LGE图像量化疤痕空间特征。中位随访34个月期间,复合终点事件共69例。结果显示ADAS 3D衍生的疤痕空间特征对心梗后主要不良心血管事件具有独立预测价值,其判别能力优于常规CMR疤痕参数。

要点

  • 总边缘区通道质量是唯一独立预测主要和次要终点的影像指标,适合在报告中重点描述。
  • LGE后处理量化疤痕空间特征可提供超越常规疤痕参数的预后信息。
  • 出现终点事件者整体纵向应变更差、疤痕质量更大、边缘区通道空间特征更高。

核心结论

  • ADAS 3D衍生的疤痕空间特征对心梗后主要不良心血管事件具有独立预测价值。
  • 总边缘区通道质量的判别能力优于常规CMR疤痕参数。

研究设计与样本

  • 回顾性研究,纳入心肌梗死后经ADAS 3D后处理分析确认存在边缘区通道的病例。
  • 基于钆延迟增强图像,采用ADAS 3D软件定量获取疤痕空间特征。
  • 变量先用单因素Cox回归筛选(p < 0.1),再用LASSO-Cox回归以λ1se惩罚准则筛出保留特征。
  • 模型性能评估采用时间依赖性Harrell C指数、经1000次bootstrap验证的校准曲线及决策曲线分析。

关键影像表现/征象

  • 终点事件患者整体纵向应变和左房室耦连指数受损更明显。
  • 终点事件患者疤痕质量更大,边缘区通道空间特征更高(均 p < 0.05)。
  • 边缘区通道空间特征具体包括通道数量、质量、长度、透壁程度和复杂度评分。

诊断标准与关键数值

  • 经LASSO-Cox回归以λ1se惩罚准则筛选后,仅总边缘区通道质量与主要终点和次要终点独立相关。
  • 时间依赖性C指数曲线、校准曲线和决策曲线分析均显示总边缘区通道质量对两类终点事件具有良好判别能力。
  • 主要终点事件为复合心律失常结局,中位随访34个月期间共36例。
  • 次要终点在主要终点基础上加入全因死亡和心衰再住院,共69例。

临床应用与阅片要点

  • 心梗后CMR报告中可加用ADAS 3D后处理量化边缘区通道质量辅助预后分层。
  • 影像科医生应关注边缘区通道的透壁程度和复杂度评分,而不仅是常规疤痕大小。
  • 对于疤痕质量大且边缘区通道特征高的患者,应在报告中提示主要不良心血管事件风险增加。

局限性

  • 本研究为回顾性设计,存在选择偏倚和信息偏倚的潜在可能。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

英文原文(摘要)· 点击展开

RATIONALE AND OBJECTIVES: The border zone (BZ) assessed by cardiac magnetic resonance (CMR) is the critical substrate for major adverse cardiac events (MACEs) in post-myocardial infarction (MI) patients. However, conventional BZ characterization fails to capture the spatial properties of abnormal electrical conduction BZC. The study aims to evaluate ADAS 3D-quantified BZ channel (BZC) spatial features as novel MACEs predictors.

MATERIALS AND METHODS: This retrospective study included post-MI patients with ADAS 3D post-processing analysis confirming the presence of BZC. Primary endpoint events include composite arrhythmic outcomes, while secondary endpoints incorporated these events plus all-cause mortality and heart failure (HF) rehospitalization. The scar spatial properties (BZC number, mass, length, transmurality, and complexity score) were quantitatively obtained based on the late gadolinium enhancement (LGE) images using the ADAS 3D software. To establish robust prognostic models, candidate variables were first screened via univariate Cox regression (p < 0.1) followed by Least Absolute Shrinkage and Selection Operator (LASSO)-Cox regression with the λ1se penalty criterion to minimize overfitting. The optimal predictive model was constructed based on the retained features. Model performance was comprehensively evaluated using time-dependent Harrell's C-index, calibration curves (validated by 1000 bootstrap) and decision curve analysis (DCA) to assess its discriminative ability and clinical utility.

RESULTS: During a median follow-up of 34 months, 69 patients experienced secondary endpoint events, including 36 with primary endpoint events. Patients with endpoint events demonstrated significantly impaired global longitudinal strain(GLS) and left atrioventricular coupling index (LACI), greater scar mass, and higher BZC spatial characteristics(all p < 0.05). Via LASSO-Cox regression with the λ1se penalty criterion, only total BZC mass was independently associated with primary and secondary endpoints. Time-dependent C-index curves, calibration curves and DCA indicated that the total BZC mass provided great discriminative ability for predicting both endpoint events.

CONCLUSION: ADAS 3D-derived scar spatial characteristics were independently predictive of post-MI MACE and demonstrated discriminatory performance beyond conventional CMR scar parameters.

来源:Acad Radiol,查看原文。版权归原作者所有。

引用

  1. Zhao, Wang, Chen, et al. Prognostic Value of Scar Spatial Properties on CMR for Major Adverse Cardiovascular Events After Myocardial Infarction. Acad Radiol. 2026. PMID: 42532715. doi:10.1016/j.acra.2026.07.018

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。