文献速递腹部技术与安全US

Sonazoid超声造影预测肝细胞癌微血管侵犯风险的前瞻性多中心研究分析

Analysis of Sonazoid contrast-enhanced ultrasound for predicting the risk of microvascular invasion in hepatocellular carcinoma: a prospective multicenter study.
PubMed 文献 · Eur Radiol · 2023-04-28

这项前瞻性多中心研究纳入211例肝细胞癌患者,基于Sonazoid超声造影及临床变量构建微血管侵犯预测模型。联合模型在推导队列和外部验证队列的AUROC分别为0.859和0.812,结论认为该模型可在术前较准确预测微血管侵犯风险。

要点

  • 肿瘤大小>49.2 mm、动脉期不均匀增强、非单结节形态、廓清时间<90 s及灰度值比≤0.50与微血管侵犯显著相关
  • 联合模型术前预测微血管侵犯的AUROC在推导队列为0.859、外部验证队列为0.812
  • Sonazoid超声造影在肝内可形成独特的Kupffer期,有助于术前无创预测微血管侵犯

核心结论

  • 联合临床与Sonazoid超声造影影像特征的模型可在术前较准确预测肝细胞癌患者的微血管侵犯风险。
  • 基于Sonazoid的术前无创微血管侵犯预测模型可能有助于临床医生制定个体化治疗决策。
  • 研究结论认为该模型预测微血管侵犯具有较高准确性。

研究设计与样本

  • 前瞻性多中心研究,分析Sonazoid超声造影在肝肿瘤中的临床应用。
  • 2020年8月至2021年3月共评估211例患者,分为推导队列170例和外部验证队列41例。
  • 采用多变量logistic回归建立微血管侵犯预测模型,分别构建临床模型、Sonazoid超声造影模型和联合模型,并进行外部验证。
  • 通过亚组分析探讨Sonazoid超声造影模型在无创预测微血管侵犯中的表现。

关键影像表现/征象

  • 多变量分析显示肿瘤大小>49.2 mm、病理分化、动脉期不均匀增强模式、非单结节大体形态、廓清时间<90 s及灰度值比≤0.50与微血管侵犯显著相关。
  • Sonazoid可在内皮网络中蓄积并在肝脏成像中形成独特的Kupffer期。

诊断标准与关键数值

  • 微血管侵犯预测模型纳入的显著变量为:肿瘤大小>49.2 mm、病理分化、动脉期不均匀增强模式、非单结节大体形态、廓清时间<90 s、灰度值比≤0.50。
  • 联合模型在推导队列的AUROC为0.859(95%CI:0.803-0.914),在外部验证队列为0.812(95%CI:0.691-0.915)。
  • Sonazoid超声造影模型在直径≤30 mm亚组的AUROC为0.819(95%CI:0.698-0.941),在直径>30 mm亚组为0.747(95%CI:0.670-0.824)。
  • 211例患者中发生微血管侵犯者89例(42.2%)。

临床应用与阅片要点

  • 作者认为,基于Sonazoid的术前无创微血管侵犯预测模型有助于临床医生制定个体化治疗决策。
  • 研究要点指出,该发现可帮助临床医生在术前预测肝细胞癌患者的微血管侵犯,并为优化手术管理和监测策略提供依据。

局限性

  • 原文未报告局限性。

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

OBJECTIVES: The aim of this study was to evaluate the potential of Sonazoid contrast-enhanced ultrasound (SNZ-CEUS) as an imaging biomarker for preoperative prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC).

METHODS: From August 2020 to March 2021, we conducted a prospective multicenter study on the clinical application of Sonazoid in liver tumor; a MVI prediction model was developed and validated by integrating clinical and imaging variables. Multivariate logistic regression analysis was used to establish the MVI prediction model; three models were developed: a clinical model, a SNZ-CEUS model, and a combined model and conduct external validation. We conducted subgroup analysis to investigate the performance of the SNZ-CEUS model in non-invasive prediction of MVI.

RESULTS: Overall, 211 patients were evaluated. All patients were split into derivation (n = 170) and external validation (n = 41) cohorts. Patients who had MVI accounted for 89 of 211 (42.2%) patients. Multivariate analysis revealed that tumor size (> 49.2 mm), pathology differentiation, arterial phase heterogeneous enhancement pattern, non-single nodular gross morphology, washout time (< 90 s), and gray value ratio (≤ 0.50) were significantly associated with MVI. Combining these factors, the area under the receiver operating characteristic (AUROC) of the combined model in the derivation and external validation cohorts was 0.859 (95% confidence interval (CI): 0.803-0.914) and 0.812 (95% CI: 0.691-0.915), respectively. In subgroup analysis, the AUROC of the SNZ-CEUS model in diameter ≤ 30 mm and ˃ 30 mm cohorts were 0.819 (95% CI: 0.698-0.941) and 0.747 (95% CI: 0.670-0.824).

CONCLUSIONS: Our model predicted the risk of MVI in HCC patients with high accuracy preoperatively.

CLINICAL RELEVANCE STATEMENT: Sonazoid, a novel second-generation ultrasound contrast agent, can accumulate in the endothelial network and form a unique Kupffer phase in liver imaging. The preoperative non-invasive prediction model based on Sonazoid for MVI is helpful for clinicians to make individualized treatment decisions.

KEY POINTS: • This is the first prospective multicenter study to analyze the possibility of SNZ-CEUS preoperatively predicting MVI. • The model established by combining SNZ-CEUS image features and clinical features has high predictive performance in both derivation cohort and external validation cohort. • The findings can help clinicians predict MVI in HCC patients before surgery and provide a basis for optimizing surgical management and monitoring strategies for HCC patients.

来源:Eur Radiol,查看原文。版权归原作者所有。

引用

  1. Yao J, Li K, Yang H, et al. Analysis of Sonazoid contrast-enhanced ultrasound for predicting the risk of microvascular invasion in hepatocellular carcinoma: a prospective multicenter study. Eur Radiol. 2023. PMID: 37115213. doi:10.1007/s00330-023-09656-3

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。