De Franco A, Monteforte M G, Maresca G, De Gaetano A M, Manfredi R, Marano P
Istituto di Radiologia, Università Cattolica del Sacro Cuore, Roma.
Radiol Med. 1997 Jan-Feb;93(1-2):87-94.
Hemangiomas are the most frequent benign tumors of the liver which are diagnosed at clinics only when they are very big. Their differentiation from primary or secondary liver tumors is a very difficult step. To investigate the capabilities of color Doppler US, CT and MRI in characterizing liver hemangiomas, we retrospectively examined 27 patients with 35 hemangiomas of the liver, all identified with US from January, 1995, to March, 1996, plus 26 patients with color Doppler US, 23 with dynamic CT and 19 with MR findings; 16 patients had been examined with all imaging techniques. Our gold standard was red blood cell SPECT in 16 patients, clinical-diagnostic follow-up in 8 and surgical specimens in 3 patients. Morphological data, the number, size and segmental site were investigated for all lesions. Doppler US was used to study the qualitative and quantitative data of intralesional color patterns, according to a modified Tanaka's classification, and of intralesional max, systolic velocity and pulsatility index (PI). Dynamic CT was performed with the polyphasic technique, that is unenhanced scans followed by selective, dynamic sequential and delayed scans. Morphological data and dynamic time/density curves of lesion/parenchyma were studied. MRI was performed at 0.5 T with T1-weighted SE (TR/TE 280/18), T2-weighted SE (TR/TE 2000/45-90-120) and T2-weighted fast SE (TR/TE 5000/160) sequences. Dynamic T1-weighted GE sequences were performed after Gd-DTPA bolus injection, using the same times as those of dynamic CT. US showed 1.3 lesion/patient, while CT and MRI showed 1.5 and 1.7 lesion/patient, respectively. No differences were observed in lesion size with the three imaging techniques and there was 78% agreement on segmental site. As for morphological patterns, US had 46% sensitivity, considering medium-small lesions only. Intralesional color signals, with spot and branch patterns, were seen in 6/20 lesions (mean phi; 8.4 cm), with a mean intralesional systolic velocity of 0.30 m/s (1.46 KHz); mean intralesional PI value was 0.83. Color Doppler sensitivity was 23%. Combined B-mode and color Doppler US had 69% sensitivity. The typical CT pattern (ring or globular enhancement with centripetal progression and late hyperdensity) had 66% sensitivity. Hyperintensity on T2-weighted images had 96% sensitivity; Gd-DTPA MRI combined with dynamic CT had 100% sensitivity. In conclusion, the US findings of a suspected hemangioma should be diagnostically integrated with MR studies to make the correct diagnosis.
肝血管瘤是肝脏最常见的良性肿瘤，通常只有在瘤体非常大时才会在临床上被诊断出来。将其与原发性或继发性肝肿瘤区分开来是非常困难的一步。为了研究彩色多普勒超声（US）、CT和MRI在肝血管瘤特征性诊断方面的能力，我们回顾性地检查了27例患有35个肝血管瘤的患者，所有患者的肝血管瘤均于1995年1月至1996年3月期间通过超声检查确诊，另外还有26例患者进行了彩色多普勒超声检查、23例进行了动态CT检查以及19例进行了MRI检查；16例患者接受了所有影像学检查。我们的金标准是16例患者采用红细胞SPECT检查、8例采用临床诊断随访以及3例采用手术标本检查。对所有病变的形态学数据、数量、大小和节段部位进行了研究。根据改良的田中分类法，使用多普勒超声研究瘤体内彩色模式的定性和定量数据，以及瘤体内最大收缩期速度和搏动指数（PI）。动态CT采用多期技术进行，即先进行平扫，然后进行选择性动态序列扫描和延迟扫描。研究病变/实质的形态学数据和动态时间/密度曲线。MRI在0.5T磁场下采用T1加权SE序列（TR/TE 280/18）、T2加权SE序列（TR/TE 2000/45 - 90 - 120）和T2加权快速SE序列（TR/TE 5000/160）进行检查。在静脉注射钆喷酸葡胺（Gd - DTPA）后进行动态T1加权GE序列检查，时间与动态CT相同。超声显示每位患者有病变1.3个，而CT和MRI分别显示每位患者有病变1.5个和1.（此处原文MRI显示的数据有误，推测应为1.7）个。三种影像学检查在病变大小方面未观察到差异，在节段部位方面有78%的一致性。至于形态学模式，仅考虑中小病变时，超声的敏感性为46%。在20个病变中的6个（平均直径8.4cm）观察到瘤体内有斑点状和分支状彩色信号，瘤体内平均收缩期速度为0.30m/s（1.46KHz）；瘤体内平均PI值为0.83。彩色多普勒敏感性为23%。B超与彩色多普勒超声联合使用时敏感性为69%。典型的CT模式（环形或球形强化，向心性进展且延迟期高密度）敏感性为66%。T2加权图像上的高信号敏感性为96%；Gd - DTPA MRI联合动态CT敏感性为100%。总之，对于疑似肝血管瘤的超声检查结果，应与MR研究进行诊断性整合以做出正确诊断。
