Castoldi M C, Fauda V, Scaramuzza D, Vergnaghi D
Servizio di Radiologia, Ospedale S. Carlo Borromeo, Via Pio II 3, 20153 Milano MI.
Radiol Med. 2000 Sep;100(3):160-7.
To compare prospectively dynamic gadolinium (Gd)-enhanced with superparamagnetic iron oxide (SPIO)-enhanced MRI for the detection of hepatocellular carcinoma (HCC).
Twenty-five patients with histologically proven HCC and liver cirrhosis (28% of them in B or C Child class) underwent dynamic Gd-enhanced MRI and, a few days later, (mean interval: three days) SPIO-enhanced MRI. Only patients with availability of clinical and imaging follow-up for at least seven months were enrolled in this prospective study. Axial dynamic Gd-enhanced imaging was performed with T1 gradient-recalled echo (GRE) sequences. Both axial and sagittal SPIO-enhanced imaging were performed with respiratory triggered T2-weighted turbo spin-echo (TSE) and T1-T2*-weighted GRE sequences. MR images were reviewed by two independent radiologists. The readers scored each lesion for the presence of HCC and assigned confidence levels based on a five-grade scale: 1, definitely or almost definitely absent; 2, possibly present; 3, probably present; 4, definitely present; 5, definitely present with optimal liver/lesion contrast or good liver/lesion contrast and morphological signs (intact capsule, intranodular septa, extra-capsular infiltration), useful for locoregional treatment planning. A positive diagnostic value was assessed for scores of 3 or higher.
Gd-enhanced and SPIO-enhanced MRI found 44 lesions. The combined use of TSE and GRE SPIO-enhanced sequences detected 11 more lesions (25% improvement in sensitivity) than Gd-enhanced MRI. One lesion (2.27%) was detected only with Gd-enhanced MRI. Eight of twelve lesions visible with a single contrast agent measured less than 1 cm in diameter. HCC detectability was 75% with Gd-enhanced MRI and 97.7% with SPIO-enhanced MRI. SPIO-enhanced T2-weighted TSE images showed significantly higher diagnostic value than SPIO-enhanced T1-T2*GRE images only in three cases, while nodule morphological characteristics (capsule, septa, different cell differentiation components) were better depicted by TSE images.
In our study the combined use of SPIO-enhanced T2-weighted TSE and T1-T2*-weighted GRE sequences showed higher sensitivity than gadolinium-enhanced GRE dynamic imaging (97.7% versus 75%). These results are at least partly related to our study conditions, that is: 1) MRI was performed with a 1T system, 2) both axial and sagittal SPIO-enhanced imaging were performed with respiratory triggered T2-weighted TSE and T1-T2*-weighted GRE, 3) there was a low freaquency of severe cirrhosis.
前瞻性比较动态钆(Gd)增强磁共振成像(MRI)与超顺磁性氧化铁(SPIO)增强MRI检测肝细胞癌(HCC)的效果。
25例经组织学证实为HCC且合并肝硬化的患者(其中28%为Child B或C级)接受了动态Gd增强MRI检查,几天后(平均间隔:3天)又接受了SPIO增强MRI检查。本前瞻性研究仅纳入了有至少7个月临床及影像随访资料的患者。采用T1梯度回波(GRE)序列进行轴位动态Gd增强成像。采用呼吸触发的T2加权快速自旋回波(TSE)序列和T1-T2*加权GRE序列进行轴位和矢状位SPIO增强成像。由两名独立的放射科医生对MR图像进行评估。阅片者对每个病变进行HCC存在情况评分,并根据五级评分标准给出置信度:1,肯定或几乎肯定不存在;2,可能存在;3,很可能存在;4,肯定存在;5,肯定存在且肝脏/病变对比度最佳或肝脏/病变对比度良好并有形态学征象(完整包膜、结节内分隔、包膜外浸润),有助于局部治疗方案的制定。评分为3或更高时评估为阳性诊断价值。
Gd增强和SPIO增强MRI共发现44个病变。与Gd增强MRI相比,联合使用TSE和GRE SPIO增强序列多检测出11个病变(敏感性提高25%)。仅Gd增强MRI检测到1个病变(2.27%)。12个仅用一种对比剂可见的病变中,8个直径小于1 cm。Gd增强MRI检测HCC的能力为75%,SPIO增强MRI为97.7%。仅在3例中,SPIO增强的T2加权TSE图像显示出明显高于SPIO增强的T1-T2*GRE图像的诊断价值,而TSE图像能更好地显示结节的形态学特征(包膜、分隔、不同细胞分化成分)。
在我们的研究中,联合使用SPIO增强的T2加权TSE序列和T1-T2加权GRE序列显示出比钆增强GRE动态成像更高的敏感性(97.7%对75%)。这些结果至少部分与我们的研究条件有关,即:1)使用1T系统进行MRI检查;2)采用呼吸触发的T2加权TSE序列和T1-T2加权GRE序列进行轴位和矢状位SPIO增强成像;3)严重肝硬化的发生率较低。