Objective: To evaluate the predictive value of the 2025 appropriate use criteria (AUV) for implantable cardioverter defibrillators (ICD) jointly issued by the American College of Cardiology/American Heart Association and other societies, for sudden cardiac death (SCD) in patients with dilated cardiomyopathy (DCM), and to explore the incremental value of cardiac magnetic resonance (CMR) tissue characterization in SCD risk stratification. Methods: This was a single-center, retrospective study that consecutively enrolled DCM patients who underwent CMR with late gadolinium enhancement (LGE) and T1 mapping at Fuwai Hospital, Chinese Academy of Medical Sciences from February 2012 to September 2021. According to the 2025 ICD appropriate use criteria and LGE extent, patients were divided into three groups: AUC-appropriate with LGE<7.2% (AUC-A+LGE<7.2%) group, AUC-appropriate with LGE≥7.2% (AUC-A+LGE≥7.2%) group, and AUC-maybe appropriate (AUC-M) group. Baseline data and CMR parameters were collected, and all patients were followed up. Predictors of the SCD composite endpoint were identified using Cox proportional hazards regression, and the clinical efficacy of LGE and extracellular volume fraction (ECV<31.8% or ≥31.8%) for SCD risk stratification was assessed using Kaplan-Meier survival analysis. Results: A total of 741 DCM patients were enrolled, aged (47.0±14.3) years, with 581 (78.4%) males. There were 447 patients in the AUC-A+LGE<7.2% group, 162 in the AUC-A+LGE≥7.2% group, and 132 in the AUC-M group. Over a follow-up of 53.4 (32.9, 74.8) months, the SCD composite endpoint occurred in 78 patients (10.5%). Multivariable Cox regression analysis identified left atrial volume index (HR=1.005, 95%CI 1.001-1.009, P=0.015), right ventricular ejection fraction (HR=0.976, 95%CI 0.963-0.990, P<0.001), LGE≥7.2% (HR=4.308, 95%CI 2.633-7.050, P<0.001), and ECV≥31.8% (HR=2.719, 95%CI 1.515-4.878, P<0.001) were predictors of the SCD composite endpoint. Kaplan-Meier analysis showed no significant difference in SCD risk between the AUC-A+LGE<7.2% group and AUC-M group (log-rank P=0.269). Further stratification revealed that patients with ECV<31.8% in the AUC-A+LGE<7.2% subgroup had an extremely low annual SCD event rate of 0.6%, representing a truly low-risk population accounting for 35.8% (265/741) of the total cohort. In contrast, the annual SCD event rate in the AUC-A+LGE≥7.2% group was as high as 5.5%. Conclusions: CMR myocardial tissue characterization parameters (LGE and ECV) provide significant incremental prognostic value to the 2025 ICD appropriate use criteria, enabling precise re-stratification of SCD risk in DCM patients and facilitating optimized clinical decision-making for primary prevention with ICD implantation. 目的： 评估2025年美国心脏病学会/美国心脏协会等联合发布的植入式心律转复除颤器（ICD）适当使用标准对扩张型心肌病（DCM）患者心脏性猝死（SCD）的预测价值，并探讨心脏磁共振（CMR）组织特征成像在SCD风险分层中的增量作用。 方法： 本研究为单中心回顾性研究，连续纳入2012年2月至2021年9月于中国医学科学院阜外医院接受CMR钆对比剂延迟强化（LGE）和T1 mapping检查的DCM患者。根据2025年ICD适当使用标准定义和LGE范围将患者分为标准适用且LGE<7.2%（AUC-A+LGE<7.2%）组、标准适用且LGE≥7.2%（AUC-A+LGE≥7.2%）组和标准可能适用（AUC-M）组。收集各组患者的基线资料及CMR指标，并对患者进行随访。采用Cox比例风险回归模型分析DCM患者发生SCD复合终点事件的预测因子，并通过Kaplan-Meier生存曲线评估LGE和细胞外间质容积分数（ECV<31.8%或≥31.8%）对DCM患者SCD风险分层的临床应用效能。 结果： 共纳入741例DCM患者，年龄（47.0±14.3）岁，男性581例（78.4%）。AUC-A+LGE<7.2%组447例，AUC-A+LGE≥7.2%组162例，AUC-M组132例。随访53.4（32.9，74.8）个月，共78例（10.5%）患者发生SCD复合终点事件。多因素Cox比例风险回归分析显示，左心房容积指数（HR=1.005，95%CI 1.001~1.009，P=0.015）、右心室射血分数（HR=0.976，95%CI 0.963~0.990，P<0.001）、LGE≥7.2%（HR=4.308，95%CI 2.633~7.050，P<0.001）和ECV≥31.8%（HR=2.719，95%CI 1.515~4.878，P<0.001）是DCM患者SCD复合终点事件的预测因子。Kaplan-Meier生存分析显示，AUC-A+LGE<7.2%组患者SCD风险与AUC-M患者差异无统计学意义（log-rank P=0.269）。进一步分层分析显示，AUC-A+LGE<7.2%组患者中，ECV<31.8%的患者SCD年均事件发生率仅为0.6%，属于真正低危人群，占研究队列的35.8%（265/741）；而AUC-A+LGE≥7.2%组患者SCD年均事件发生率高达5.5%。 结论： CMR心肌组织特征参数（LGE与ECV）可为2025年ICD适当使用标准提供重要的增量预后价值，实现DCM患者SCD风险的精准再分层，有助于优化ICD一级预防的临床决策。.