简化US-Node-RADS联合血管后期全氟丁烷CEUS对不确定颈部淋巴结的分步诊断策略
双中心回顾性研究开发了基于常规超声的5分法US-Node-RADS，并在训练集（n=356）及内外部验证集中评估。对判为Node-RADS 3的不确定淋巴结加做血管后期全氟丁烷CEUS，等/高增强降为2级随访、无/低增强升为4级活检，可安全避免14.7%–40.0%的不必要活检，同时恶性检出率提升至79.3%–100%。
要点
- 常规超声Node-RADS判为3级的颈部淋巴结占比达40.2%，属诊断不确定区。
- 对Node-RADS 3加做血管后期全氟丁烷CEUS可重新分类并减少不必要的活检。
- 再分类后恶性检出率79.3%–100%，高于各队列Node-RADS 3的基线恶性率。
核心结论
- 对Node-RADS 3的颈部淋巴结加做血管后期全氟丁烷CEUS可有效解决诊断不确定性。
- 该两步策略在保持高恶性检出率的同时显著减少不必要的活检。
研究设计与样本
- 双中心回顾性研究，影像按标准化方案前瞻性采集，参考标准为组织病理学和随访。
- 训练集n=356开发5分法US-Node-RADS，内部验证n=152、外部验证n=45。
关键影像表现/征象
- 常规超声Node-RADS诊断效能中等，训练集AUC为0.78。
- Node-RADS 3淋巴结中40.2%处于不确定类别。
- 血管后期CEUS等/高增强者降为Node-RADS 2（随访），无/低增强者升为Node-RADS 4（活检）。
诊断标准与关键数值
- US-Node-RADS为5分法，Node-RADS 3定义为不确定类别。
- 训练集AUC 0.78；Node-RADS 3基线恶性率训练集67.8%、内部验证69.0%、外部验证60.0%。
- 再分类后恶性检出率79.3%–100%，各队列避免不必要活检14.7%–40.0%。
临床应用与阅片要点
- 对常规超声判为Node-RADS 3的颈部淋巴结，可加做血管后期全氟丁烷CEUS再分类。
- 等/高增强提示良性可能，降级随访；无/低增强提示恶性可能，升级活检。
- 该路径有助于术前决策，但需在适当选择的患者中应用。
局限性
- 回顾性设计，外部验证样本量仅45例。
- 原文未报告其他局限。
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
OBJECTIVE: To develop a conventional ultrasound (US)-based Node-reporting and data systems (US-Node-RADS) for cervical lymph nodes (CLNs) and to evaluate whether the subsequent application of postvascular phase perfluorobutane-contrast-enhanced ultrasound (CEUS) can reclassify its indeterminate (Node-RADS 3) category and reduce unnecessary biopsies.
METHODS: In this dual-center retrospective study, data were analyzed from patients whose imaging had been prospectively acquired under a standardized protocol. A 5-point US-Node-RADS was developed in a training set (n = 356) and validated internally (n = 152) and externally (n = 45). For nodes classified as indeterminate Node-RADS 3, PVP-CEUS was applied: iso-/hyperenhancement prompted downgrading to Node-RADS 2 (follow-up), while no-/hypoenhancement prompted upgrading to Node-RADS 4 (biopsy). Histopathology and follow-up served as the reference standard. Diagnostic performance metrics were statistically evaluated.
RESULTS: The US-based Node-RADS showed moderate diagnostic performance (AUC 0.78 in training) but left a substantial proportion (40.2%) of nodes in the indeterminate Node-RADS 3 category. Applying PVP-CEUS to these nodes successfully reclassified them, safely avoiding 14.7% to 40.0% of unnecessary biopsies across the cohorts. This was achieved while maintaining high malignancy identification rates, which increased to 79.3%-100% after reclassification, exceeding the baseline malignancy rates for Node-RADS 3 (training set: 67.8%; internal validation: 69.0%; external validation: 60.0%).
CONCLUSION: Applying PVP-CEUS to Node-RADS 3 nodes effectively resolves diagnostic ambiguity for indeterminate CLNs. This two-step strategy significantly reduces unnecessary biopsy procedures while preserving high sensitivity for malignancy identification, can serve as a potential standardized and actionable pathway to aid preoperative decision-making in appropriately selected patients.
引用
- Liu N, Lan L, Lin W, et al. A Stepwise Diagnostic Strategy Combining a Simplified US-Based Node-RADS with Postvascular Phase Perfluorobutane-CEUS for Indeterminate Cervical Lymph Nodes: A Dual-Center Retrospective Study. Ultrasound Med Biol. 2026. PMID: 42069486. doi:10.1016/j.ultrasmedbio.2026.03.015
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。