Department of Radiology, NYU Langone Health, 660 First Ave, New York, NY, 10016, USA.
Siemens Healthcare GmbH, Erlangen, Germany.
Abdom Radiol (NY). 2020 Sep;45(9):2680-2687. doi: 10.1007/s00261-020-02432-7.
To evaluate the correlation between liver stiffness as measured on MR elastography and T1 and T2 relaxation times from T1 and T2 mapping with clinical parameters of liver disease, including the MELD score, MELD-Na and ALBI grade, and endoscopically visible esophageal varices.
223 patients with known or suspected liver disease underwent MRI of the liver with T1 mapping (Look-Locker sequence) and 2D SE-EPI MR elastography (MRE) sequences. 139 of these patients also underwent T2 mapping with radial T2 FS sequence. Two readers measured liver stiffness, T1 relaxation times and T2 relaxation times, and assessed qualitative features such as presence or absence of cirrhosis, ascites, spleen length, and varices on conventional MRI images. A third reader collected the clinical data (MELD score, MELD-Na Score, ALBI grade, and results of endoscopy in 78 patients).
Significant moderate correlation was found between MELD score and all three imaging techniques for both readers (MRE, r = 0.35 and 0.28; T1 relaxometry, r = 0.30 and 0.29; T2 relaxometry, r = 0.45, and 0.37 for reader 1 and reader 2 respectively). Correlation with MELD-Na score was even higher (MRE, r = 0.49 and 0.40; T1, r = 0.45 and 0.41; T2, r = 0.47 and 0.35 for reader 1 and reader 2 respectively). Correlations between MRE and ALBI grade was significant and moderate for both readers: r = 0.39 and 0.37, higher than T1 relaxometry (r = 0.22 and 0.20) and T2 relaxometry (r = 0.17, and r = 0.24). Significant moderate correlations were found for both readers between MRE and the presence of varices on endoscopy (r = 0.28 and 0.30). MRE and T1 relaxometry were significant predictors of varices at endoscopy for both readers (MRE AUC 0.923 and 0.873; T1 relaxometry AUC = 0.711 and 0.675 for reader 1 and reader 2 respectively). Cirrhotic morphology (AUC = 0.654), spleen length (AUC = 0.610) and presence of varices in the upper abdomen on MRI (AUC of 0.693 and 0.595) were all significant predictors of endoscopic varices. Multivariable logistic regression model identified that spleen length and liver MRE were significant independent predictors of endoscopic varices for both readers.
MR elastography, T1 and T2 relaxometry demonstrated moderate positive correlation with the MELD score and MELD-Na Score. Correlation between MRE and ALBI grade was superior to T1 and T2 relaxometry methods. MRE performed better than T1 and T2 relaxometry to predict the presence of varices at endoscopy. On multivariate analyses, spleen length and MRE were the only two significant independent predictors of endoscopic varices.
评估磁共振弹性成像(MRE)测量的肝脏硬度与 T1 和 T2 弛豫时间之间的相关性,这些弛豫时间来自 T1 和 T2 映射,与包括 MELD 评分、MELD-Na 和 ALBI 分级以及内镜可见食管静脉曲张在内的肝脏疾病的临床参数相关。
223 例已知或疑似肝脏疾病的患者接受了肝脏 T1 映射(Look-Locker 序列)和二维 SE-EPI MRE(MRE)序列的 MRI 检查。其中 139 例患者还接受了径向 T2 FS 序列的 T2 映射。两位读者测量了肝脏硬度、T1 弛豫时间和 T2 弛豫时间,并在常规 MRI 图像上评估了肝硬化、腹水、脾脏长度和静脉曲张等定性特征。第三位读者收集了临床数据(MELD 评分、MELD-Na 评分、ALBI 分级和 78 例患者的内镜结果)。
两位读者均发现 MELD 评分与三种影像学技术之间存在显著中度相关性(MRE,r=0.35 和 0.28;T1 弛豫率,r=0.30 和 0.29;T2 弛豫率,r=0.45 和 0.37,分别为读者 1 和读者 2)。与 MELD-Na 评分的相关性甚至更高(MRE,r=0.49 和 0.40;T1,r=0.45 和 0.41;T2,r=0.47 和 0.35,分别为读者 1 和读者 2)。MRE 与 ALBI 分级之间的相关性对于两位读者都是显著的中度相关性:r=0.39 和 0.37,高于 T1 弛豫率(r=0.22 和 0.20)和 T2 弛豫率(r=0.17 和 r=0.24)。两位读者均发现 MRE 与内镜下静脉曲张的存在之间存在显著中度相关性(r=0.28 和 0.30)。MRE 和 T1 弛豫率是两位读者内镜下静脉曲张的显著预测因子(MRE AUC 0.923 和 0.873;T1 弛豫率 AUC 分别为 0.711 和 0.675,读者 1 和读者 2)。肝硬化形态(AUC=0.654)、脾脏长度(AUC=0.610)和上腹部 MRI 中静脉曲张的存在(AUC 分别为 0.693 和 0.595)均为内镜静脉曲张的显著预测因子。多变量逻辑回归模型确定,脾脏长度和肝脏 MRE 是两位读者内镜静脉曲张的独立显著预测因子。
MRE、T1 和 T2 弛豫率与 MELD 评分和 MELD-Na 评分呈中度正相关。MRE 与 ALBI 分级之间的相关性优于 T1 和 T2 弛豫率方法。MRE 比 T1 和 T2 弛豫率更能准确预测内镜下静脉曲张的存在。在多变量分析中,脾脏长度和 MRE 是内镜下静脉曲张的唯一两个显著独立预测因子。