超声衍生脂肪分数与MRI质子密度脂肪分数定量肝脂肪变性的比较:一项真实世界前瞻性研究
这项前瞻性研究纳入176例患者,以MRI-PDFF为参考标准,评估超声衍生脂肪分数(UDFF)定量肝脂肪变性的可靠性和诊断性能。UDFF与MRI-PDFF一致性高(ICC=0.899),观察者间一致性极好(ICC=0.951),UDFF诊断轻度、中度、重度肝脂肪变性的最佳阈值分别为5.5%、15.5%、17.5%。
要点
- UDFF与MRI-PDFF一致性高(ICC=0.899),不同超声医师间重复性极好(ICC=0.951)。
- UDFF诊断轻度、中度、重度肝脂肪变性的阈值分别为5.5%、15.5%、17.5%。
- 对应阈值下敏感度分别为79.2%、87.5%、88.9%,特异度分别为81.5%、90.6%、90.0%。
核心结论
- UDFF是定量和分级肝脂肪变性的可靠、准确方法,与MRI-PDFF高度一致;UDFF阈值5.5%、15.5%、17.5%对检测轻度、中度、重度肝脂肪变性具有高敏感度和特异度。
研究设计与样本
- 前瞻性研究,最终纳入176例患者(中位年龄36.0岁;男82例,女94例),于2023年2月至2024年2月接受MRI-PDFF和UDFF检查;MRI-PDFF取三个肝段测量值的中位数作为参考标准,UDFF由两名超声医师各测量六次取中位数。
关键影像表现/征象
- MRI-PDFF中位值为11.3%(IQR 7.5-18.9),84.7%患者MRI-PDFF中位值≥6.4%。
- 两名超声医师测量的UDFF中位值分别为9.5%(IQR 5.0-18.0)和9.0%(IQR 5.0-18.0)。
- UDFF与MRI-PDFF的Bland-Altman分析显示平均偏差1.7%(95% LOA -8.7至12.2%)。
诊断标准与关键数值
- UDFF诊断MRI-PDFF阈值6.4%(轻度)的最佳截断值为5.5%,AUC 0.851,敏感度79.2%,特异度81.5%。
- UDFF诊断MRI-PDFF阈值17.4%(中度)的最佳截断值为15.5%,AUC 0.952,敏感度87.5%,特异度90.6%。
- UDFF诊断MRI-PDFF阈值22.1%(重度)的最佳截断值为17.5%,AUC 0.948,敏感度88.9%,特异度90.0%。
临床应用与阅片要点
- UDFF可作为MRI-PDFF的可靠替代方法用于肝脂肪变性的定量和分级。
- 不同超声医师间UDFF测量重复性极好(ICC=0.951),支持其在实际工作中的可重复性。
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
OBJECTIVE: To compare the agreement between ultrasound-derived fat fraction (UDFF) with magnetic resonance proton density fat fraction (MRI-PDFF) for quantification of hepatic steatosis and verify its reliability and diagnostic performance by comparing with MRI-PDFF as the reference standard.
METHODS: This prospective study included a primary analysis of 191 patients who underwent MRI-PDFF and UDFF from February 2023 to February 2024. MRI-PDFF were derived from three liver segment measurements with calculation of an overall median PDFF. UDFF was performed by two different sonographers for each of the six measurements, and the median was taken. Intraclass correlation coefficient (ICC) and Bland-Altman analysis were used to assess agreement. Receiver operating characteristics (ROC) curves were used to evaluate the diagnostic performance of UDFF in detecting different degrees of hepatic steatosis.
RESULTS: A total of 176 participants were enrolled in the final cohort of this study (median age, 36.0 years; 82 men, 94 women). The median MRI-PDFF value was 11.3% (interquartile range (IQR) 7.5-18.9); 84.7% patients had a median MRI-PDFF value ≥ 6.4%. The median UDFF measured by different sonographers were 9.5% (IQR: 5.0-18.0) and 9.0% (IQR: 5.0-18.0), respectively. The interobserver agreement of UDFF measurement was excellent agreement (ICC = 0.951 [95% CI: 0.934-0.964], p < 0.001). UDFF was positively strongly correlated with MRI-PDFF with ICC of 0.899 (95% CI: 0.852-0.930). The Bland-Altman analysis showed high agreement between UDFF and MRI-PDFF measurements, with a mean bias of 1.7% (95% LOA, -8.7 to 12.2%). The optimal UDFF cutoff values were 5.5%, 15.5% and 17.5% for detecting MRI-PDFF at historic thresholds of 6.4%, 17.4%, and 22.1%, with AUC of 0.851, 0.952, and 0.948, respectively. The sensitivity was 79.2%, 87.5%, 88.9%, and specificity was 81.5%, 90.6%, 90.0%, respectively.
CONCLUSIONS: UDFF is a reliable and accurate method for quantification and classification of hepatic steatosis, with strong agreement to MRI-PDFF. The UDFF cutoff values of 5.5%, 15.5%, and 17.5% provide high sensitivity and specificity for the detection of mild, moderate, and severe hepatic steatosis, respectively.
KEY POINTS: Question Is ultrasound-derived fat fraction (UDFF) reliable for the quantitative detection of hepatic steatosis compared to MRI proton density fat fraction (MRI-PDFF)? Findings UDFF cutoff values of 5.5%, 15.5%, and 17.5% provided high sensitivity and specificity for the detection of mild, moderate, and severe hepatic steatosis, respectively. Clinical relevance UDFF is a reliable and accurate method for quantification and classification of hepatic steatosis, with strong agreement to MRI-PDFF and high reproducibility of liver fat content by different sonographers.
引用
- Qi R, Lu L, He T, et al. Comparing ultrasound-derived fat fraction and MRI-PDFF for quantifying hepatic steatosis: a real-world prospective study. Eur Radiol. 2024. PMID: 39414658. doi:10.1007/s00330-024-11119-2
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。