比较MR弹性成像及动态增强成像评价肝硬化食管胃底静脉曲张

目的 比较MR弹性成像（MRE）及动态增强成像（DCE-MRI）诊断肝硬化食管胃底静脉曲张（GEV）的价值。方法 收集接受MRE及DCE-MRI检查的肝硬化患者59例，记录血小板计数（PLT），测量肝弹性值（HS）、脾弹性值（SS）和MR增强视觉分级相关指标；以内镜结果为金标准，采用ROC曲线下面积（AUC）比较相关指标诊断GEV的价值。结果 PLT、HS、SS及MR增强视觉分级与肝硬化GEV分级具有相关性（r_s=-0.317、0.436、0.682、0.703，P均<0.05）。诊断有无GEV时，SS的AUC略高于MR增强视觉分级、HS、PLT （AUC分别为0.880、0.795、0.744、0.635），其中SS与PLT的AUC差异有统计学意义（P=0.002）；诊断中重度GEV时，MR增强视觉分级的AUC略高于SS、HS、PLT （AUC分别为0.893、0.816、0.713、0.665），其中MR增强视觉分级与HS、PLT的AUC差异有统计学意义（P=0.018、0.002）。联合SS及MR增强视觉分级，鉴别诊断有无GEV及中重度GEV的敏感度分别为94.16%、96.83%。结论 MRE可有效预测GEV及其严重程度，与DCE-MRI效果相当。

Comparison of MR elastography and dynamic contrast-enhanced imaging in evaluation on gastroesophageal varices with liver cirrhosis

Objective To explore the value of MR elastography (MRE) and dynamic contrast-enhanced imaging (DCE-MRI) in diagnosis of gastroesophageal varices (GEV) in patients with liver cirrhosis.Methods Totally 59 patients with liver cirrhosis underwent MRE and DCE-MRI. Taking endoscopic examination as the standard, platelet (PLT) count, hepatic stiffness (HS), spleen stiffness (SS) and MR visual score (MR VS) were measured. Values of related parameters in diagnosis of liver cirrhosis GEV were compared with area under the curve (AUC).Results PLT, HS, SS and MR VS were significantly correlated with the grades of GEV with liver cirrhosis (r_s=-0.317, 0.436, 0.682, 0.703, all P<0.05). In the diagnosis of with or without GEV, AUC of SS was higher than that of MR VS, HS, and PLT (AUC=0.880, 0.795, 0.744, 0.635, respectively), and the AUC of SS and PLT had statistically difference (P=0.002). In diagnois of moderate or severe GEV, the AUC of MR VS was higher than that of SS, HS and PLT (AUC=0.893, 0.816, 0.713, 0.665, respectively), and statistical differences of AUC were found between MR VS and HS, as well as between MR VS and PLT (P=0.018, 0.002). Sensitivities of MR VS combined with SS in differential diagnosis of liver cirrhosis with or without GEV, liver cirrhosis with moderate or severe GEV were 94.16% and 96.83%, respectively.Conclusion MRE is effective in the prediction of GEV with severity and diagnostic value equivalent to DCE-MR.