文章快速检索     高级检索
   复旦学报(医学版)  2020, Vol. 47 Issue (5): 660-668      DOI: 10.3969/j.issn.1672-8467.2020.05.004
0
Contents            PDF            Abstract             Full text             Fig/Tab
133例肝血管平滑肌脂肪瘤的MRI特征及分析
于洋力1 , 曾蒙苏1,2 , 杨春1,2 , 盛若凡1,2 , 种欢欢1,2     
1. 复旦大学附属中山医院放射科 上海 200032;
2. 上海市影像医学研究所 上海 200032
摘要目的 总结上皮样型及经典型肝血管平滑肌脂肪瘤(epithelioid and non-epithelioid hepatic angiomyolipoma,Epi-HAML和non-Epi-HAML)的临床特点及磁共振(magnetic resonance imaging,MRI)特征。方法 回顾性收集并分析2011-2018年于复旦大学附属中山医院行肝脏手术且术后病理证实为HAML的133例病例资料,根据病理类型分为Epi-HAML组及non-Epi-HAML组,根据肿瘤直径分为 < 3.0 cm、3.0~5.0 cm及>5.0 cm组,通过对比其MRI影像资料、肿瘤标志物等血清学指标比较各组差异。结果 133例患者中,大多数患者(83.6%)无明显临床症状。相比于non-Epi-HAML组的患者,Epi-HAML组多因出现临床症状而就诊(P < 0.05),两者在发病年龄、性别、慢性病史及血清肿瘤标志物等方面差异无统计学意义。共检出141枚HAML病灶,其中Epi-HAML组87枚,non-Epi-HAML组64枚,后者影像学诊断准确率较高(P < 0.05);对比两者的MRI特征,T2WI、扩散加权成像(diffusion weighted imaging,DWI)、强化方式及有无脂肪信号差异有统计学意义(P < 0.05)。而以肿瘤直径作为分组时(0.7~2.9 cm、3.0~5.0 cm、5.1~20.5 cm),其影像学上出现坏死、出血、肿瘤供血动脉、引流静脉及异常灌注等特征的构成比差异存在统计学意义(P < 0.05)。结果 T2WI、DWI、强化方式及有无脂肪信号在两组HAML中有差别;肿瘤直径>5 cm时,其供血动脉、引流静脉等影像学特征更突出。
关键词肝血管平滑肌脂肪瘤(HAML)    磁共振(MRI)    组织亚型    
MRI features and analysis of 133 cases of hepatic angiomyolipoma
YU Yang-li1 , ZENG Meng-su1,2 , YANG Chun1,2 , SHENG Ruo-fan1,2 , CHONG Huan-huan1,2     
1. Department of Radiology, Zhongshan Hospital, Fudan University, Shanghai 200032, China;
2. Shanghai Institute of Medical Imaging, Shanghai 200032, China
Abstract: Objective To summarize the clinical and MRI features of epithelioid and non-epithelioid hepatic angiomyolipoma (Epi-HAML and non-Epi-HAML). Methods We retrospectively collected and analyzed 133 cases of liver surgery performed at Zhongshan Hospital of Fudan University and postoperative pathology confirmed as HAML during 2011-2018.According to pathological type, they were divided into Epi-HAML group and non-Epi-HAML group; and were divided into three sub groups:smaller than 3.0 cm, 3.0-5.0 cm and larger than 5.0 cm according to tumor diameter.We compared these groups by studying their MRI images, tumor markers and other blood tests. Results Most of the patients were diagnosed without specific symptoms (83.6%).Epi-HAML patients tended to show more symptoms compared with non-Epi-HAML patients (P < 0.05).There was little difference between these two groups in age, gender, history of disease, or serum tumor makers.A total of 141 HAML lesions were discovered after surgery, 87 were Epi-HAML and 64 were non-Epi-HAML.The differences of diagnosis accurateness, T2WI, diffusion weighted imaging (DWI), enhancement mode and fat component between the two groups were significant (P < 0.05).When we grouped these lesions by tumor diameter (0.7-2.9 cm, 3.0-5.0 cm, 5.1-20.5 cm), the appearance of necrosis, cysts, hemorrhage, tumor vessels, early draining veins and perfusion disorders showed significant difference (P < 0.05). Conclusion There are significant difference between the two groups of HAML in T2WI, DWI, enhancement mode and fat signal.The imaging features such as tumor vessels and early drainage veins are more prominent when the tumor diameter is larger than 5 cm.
Key words: hepatic angiomyolipoma (HAML)    magnetic resonance imaging (MRI)    histologic subtype    

血管平滑肌脂肪瘤(angiomyolipoma,AML)是一种少见的间质来源的肿瘤,以肾脏多发,其他部位多见于肝脏(5%~15%)[1],在肝脏占位性病变的比例约为0.4%[2]。2002年WHO将AML归为血管周上皮样肿瘤(perivascular epithelioid cell tumor,PEComa)的一种,主要因为其具有典型的组织学结构,即厚壁血管、平滑肌、脂肪组织伴或不伴有一定比例的上皮样细胞,而且有典型的免疫学标记-黑色素及平滑肌肌动蛋白染色阳性(HMB-45+和SMA+)[3]。肝血管平滑肌脂肪瘤(hepatic angiomyolipoma,HAML)好发于女性,男女比例约为1:3.0~1:4.8[1, 4],相比于肾血管平滑肌脂肪瘤(renal angiomyolipoma,RAML)的患者多伴有TSC染色体异常[5- 6]。HAML发病原因目前争论较多,以往HAML一直被认为是一种良性病变,其主要处理手段为保守治疗[7]。2000年Dalle等[8]报道了第一例恶性HAML,随后越来越多的瘤体过大压迫周围组织、破裂出血、复发的病例出现,部分学者主张以手术治疗为主[9]。典型的HAML影像学表现包括较多的脂肪成分、丰富的血管等,但典型病例据统计不足50%,乏脂HAML多与肝细胞癌表现相似,误诊率较高[10]。MRI是研究HAML的重要影像学检查,相比于CT,其多序列的扫描可提供更多的诊断信息,并且对肿瘤内部成分有更好的分辨力,有助于对HAML的诊断及鉴别诊断。

资料和方法

一般资料  收集复旦大学附属中山医院2011年1月至2018年12月经手术病理证实为肝血管平滑肌脂肪瘤的患者资料,所有患者均在初诊时于我院行MRI检查及完整的血液学检查,最终共计133例进入本项研究。本研究涉及多项指标,包括患者年龄、性别、起病方式、肝炎等慢病病史、血清标志物等,所有病例均已随访。本研究已通过复旦大学附属中山医院伦理委员会审批(编号:B2018-236),免除患者知情同意。

检查方法  检查使用1.5 T MR扫描仪(Verio,德国Siemens公司)及3.0T MR扫描仪(上海联影医疗科技有限公司)。被检查者取仰卧位,脚先进,扫描范围从膈顶至肝下缘。检查序列包括:正反相位梯度回波序列T1WI,常规T2WI序列、脂肪抑制T2WI序列、平扫脂肪抑制T1WI序列。静脉注射马根维显(德国Bayer Healthcare公司,2 mL/s,0.1 mmol/kg);屏气状态下采集脂肪抑制T1WI序列动脉期(自动检测技术或延迟25~30 s)、门静脉期(延迟60~80 s)、延迟期横断位及冠状位(120~180 s)扫描。DWI序列在增强前扫描,具体参数(b值)为0、50、500 s/mm2

图像分析  所有图像均在我院影像归档和通信系统(Picture Archiving and Communication Systtems,PACS)上浏览及分析。用平均面积约9.8 cm2(7.1~12.6 cm2,直径1.5~2 cm)的圆形或椭圆形感兴趣区(region of interest,ROI),分别测量增强前后的AML病变、肝脏实质、腹主动脉和门静脉信号强度值。在测量时将ROI放置于病灶中心非坏死或出血区,避开对图像质量影响较大的区域,如大血管、肝内胆管和肝硬化结节区域。分析病灶分布、大小、形状、边界、信号、强化方式和其他征象,如病灶内是否血管穿行、引流静脉等。强化方式分为渐进性强化(即动脉期病灶轻度强化,门脉期及延迟期信号持续升高)、持续强化(即病灶在动脉期轻度或明显强化,门脉期及延迟期强化程度与其在动脉期相似)、递减性强化(动脉期病灶明显强化,门脉期及延迟期信号减低,但仍高于肝实质)以及“快进快出”(动脉期明显强化,门脉期及延迟期信号减低,且低于周围肝实质)。

统计学分析  采用SPSS 20.0统计分析软件。计量资料以x±s表示,符合正态分布的资料采用t检验,偏态分布的资料采用Mann-Whitney U检验,普通构成比四格表数据采用χ2检验,部分表格频数小于5时采用Fisher检验,配对四格表数据比较两者相关性采用Kappa检验,有序或等级行列表数据采用Kruskal-Walis H或Spearman法比较其差异性。P < 0.05为差异有统计学意义。

结果

一般资料  133名患者中,男性26名(19.5%),女性107名(80.5%),男女比为1:4.1;发病年龄21~74岁;平均48.1岁,男性平均年龄49.0岁,女性平均年龄47.9岁;其中24名患者(18.0%)因上腹部疼痛、腰部疼痛等症状就诊,另外109名患者(82.0%)因查体发现肝占位入院;甲肝病史者1例(0.8%),乙肝病史者19例(14.3%);糖尿病患者5例(3.8%);肝硬化患者5例(3.8%);脂肪肝患者10例(7.5%);复发4例(3.0%)。肿瘤血清标志物:AFP(甲胎蛋白)中位数2.5 ng/mL(0.6~13 ng/mL),均于正常范围(0~20 ng/mL)内;CEA(癌胚抗原)中位数1.3 ng/mL(0.27~111.5 ng/mL),除1例偏高(111.5 ng/mL),余均于正常范围(0~5 ng/mL)内;CA19-9(糖类抗原19-9)中位数7.75 U/mL(0.6~653 U/mL),其中2例偏高(51.4、653 U/mL),余均于正常范围(0~37 U/mL)内。所有HAML病例中,单发者126例(94.7%),2处病灶者6例(4.5%)(其中1例2枚分别为Epi-HAML及non-Epi-HAML),3处者1例(0.8%)。上皮样型HAML(Epi-HAML)与经典型HAML(non-Epi-HAML)的对比见表 1

表 1 上皮样型HAML与经典型HAML一般资料比较 Tab 1 Comparison of clinical data of Epi-HAML and non-Epi-HAML  
[n(%)]
Variable Epi-HAML (n=81a) Non-Epi-HAMl (n=53a) P
Age(y) 47.19±10.81 49.36±11.64 0.272
Gender(F/M) 67/14 41/12 0.397
Syndrome for hospitalized 18 (22.2) 4 (7.5) 0.025b
History of hepatitis 14 (17.3) 6 (11.3) 0.217
History of liver cirrhosis 4 (4.9) 1 (1.9) 0.648b
History of fatty liver 5 (6.2) 5 (9.4) 0.516
History of DM
  Tumor maker 3 (3.7) 4 (7.5) 0.434b
  AFP (ng/mL) 2.5 2.5 0.672c
  CEA (ng/mL) 1.3 1.4 0.684c
  CA19-9 (U/mL) 7.3 8.1 0.387c
  Tumor number (1/2/3) 76/4/1 49/4/0 0.599b
  Tumor recurrence 3 (3.7) 1 (1.9) 1.000b
a One patient had two lesions, Epi-HAML and non-Epi-HAML, which was considered as different case in two groups.b Fisher’s test;c Median and the rank test.

影像学资料结果  133例病例资料共计141枚AML病灶,Epi-HAML 87枚(61.7%),non-Epi-HAML 54枚(38.3%);整体病灶直径为(4.35±3.05)cm(0.7~20.5 cm),直径 < 3 cm的病灶51枚(36.2%),直径在3.0~5.0 cm的病灶43枚(30.5%),直径 > 5.0 cm的病灶47枚(33.3%);病灶位置分布参见表 2。大部分病灶(133枚,94.3%)形态规则,呈类圆形,边界清,其余8枚形态不规则,边界欠清;术前影像学诊断准确者61枚(43.2%),其余大多诊断为HCC(58/141,41.3%)或良性结节(22/141,15.6%)。HAML特征性的征象为脂肪信号,Epi-HAML组出现脂肪信号者33枚(37.9%),non-Epi-HAML组为37枚(68.5%),两组存在统计学差异(P=0.002)(图 1 E-G为non-Epi-HAML病例,肝左外叶病灶呈现明显脂肪信号)。此外,强化方式上两组同样存在差异(P=0.013),大多数Epi-HAML(60/87,68.9%)呈现“递减性”持续强化方式,而non-Epi-HAML为31枚(57.4%);前者呈“快进快出”的结节为17枚(19.5%),后者为5枚(9.1%)。门脉期及延迟期呈现“包膜样”强化易误诊为HCC,Epi-HAML组中可见19枚(21.8%),non-Epi-HAML组为10枚(18.5%),两组差异无统计学意义(P=0.675)(图 1A-DH-K术后病理均为non-Epi-HAML,但前者强化方式为“快进快出”,后期甚至出现包膜样强化,术前误诊为HCC;后者强化方式呈现“持续强化”,冠状位重建后可见“引流静脉”)。“供血动脉”在Epi-HAML组中出现率为10.3%(9/87),non-Epi-HAML组为11.1%(6/54);“引流静脉”在前者出现率为32.2%(28/87),后者为40.7%(22/54),构成比差异均无统计学意义(P=0.886;P=0.366)[图 2A-C为Epi-HMAL病例,增强扫描动脉晚期可见供血动脉(黑箭),同期可见汇入腔静脉的引流静脉(白箭)]。其余MRI信号特征见表 3。在读片过程中,注意到较大直径的HAML其出现典型影像学特征的频率有所上升,如肿瘤供血动脉、引流静脉及肿瘤内坏死出血(图 2D-F 为巨大Epi-HAML病例,内见明显坏死、囊变及慢性出血信号,同时肿瘤内见扭曲增粗的供血动脉)。因此,比较Epi-HAML及non-Epi-HAML两组的同时,另以肿瘤直径(0.7~2.9/3.0~5.0/5.1~20.5 cm)为分组进一步分析,详见表 4

表 2 上皮样型HAML与经典型HAML影像学一般特征 Tab 2 General imaging features of Epi-HAML and non-Epi-HAML
Variable Epi-HAML (n=87) Non-Epi-HAMl (n=54) P
Diagnosis accurateness [n(%)] 30 (34.5) 31 (57.4) 0.008
Tumor diameters (cm) 4.3±3.1 4.4±3.0 0.826
Diameters range
0-3/3-5/5-20.5 (cm) 30/30/27 21/13/20 0.423
Shape (round/irregular) 83/4 50/4 0.364b
Boundary (smooth/ill-defined) 83/4 50/4 0.364b
Tumor location (segement) 119a 76a 0.596
Caudate lobe[n(%)] 4 (3.5) 1 (2.6)
  Ⅰ 4 1
Left lobe[n(%)] 48 (41.7) 30 (39.5)
  Ⅱ/Ⅲ/Ⅳ 15/22/11 6/9/15
Right lobe[n(%)] 63 (53.8) 45 (59.2)
  Ⅴ/Ⅵ/Ⅶ/Ⅷ 19/21/9/14 8/10/12/15
a Some tumors involved multiple liver segments, we took these liver segments into concern; b Fisher’s test.
A-D:Patient 1.Diagnosed as liver lesion during physical examination; low signal intense on T1-FS images at junction of left lobe and right lobe; hepatic arterial phase shows significant arterial enhancement; portal venous phase shows wash out (black arrow), delayed phase shows capsularenhancement (white arrow). Pathology:non-Epi-HAML.E-G: Patient 2.Diagnosed as liver lesion during physical examination; lesion in left lobe, opposed phase and T1-FS shows diffuse low signal intense and Indian ink sign (white arrow).Pathology:non-Epi-HAML.H-K:Patient 3.Diagnosed as liver lesion during physical examination; lesion in right lobe, hepatic arterial phase shows tumor vessel originating from portal vein (black arrow), early drainage vein inflow into the right hepatic vein (white arrow) on coronal position.Pathology: non-Epi-HAML. 图 1 3例经典型血管平滑肌脂肪瘤术前MRI图像 Fig 1 Preoperative MRI of 3 non-Epi-HAML cases
表 3 上皮样型HAML与经典型HAML磁共振影像学特征比较 Tab 3 MRI features of Epi-HAML and non-Epi-HAML foci
Variable Epi-HAML (n=87) Non-Epi-HAML (n=54) P
T1WI(low/equal/high/mixed) 85/0/0/2 53/1/0/0 0.524a
T2WI(low/equal/high/mixed) 1/0/84/2 4/2/47/1 0.038a
T2WI(liver-spleen, hyper or equal/hypo) 16/71 14/40 0.298
T2WI homogeneous[n(%)] 5 (5.7) 6 (11.1) 0.334
Arterial phase enhancement
  Entire/partial/circle/none 6/74/7/0 2/43/7/2 0.206a
Enhancement degree (AML-portal vein)
  Hypo-/iso-/hyper- 59/25/3 45/9/0 0.091a
Enhancement mode: progressive/persistent/degressive/wash out 2/8/60/17 4/14/31/5 0.013a
DWI:Entire hyperintensity/circular hyper-/iso-/hypo-/heterogeneous 80/0/3/1/3 37/3/1/10/3 0.000a
ADC map:Entire hypointensity/circular hypo-/iso-/hyper/heterogeneous 38/0/10/37/2 24/2/2/25/1 0.227a
Fat (none/diffuse/focal) 54/20/13 17/25/12 0.002
Capsules (none/contain) 68/19 44/10 0.675
Necrosis or cysts(none/contain) 83/4 51/3 1.000a
Hemorrhage(none/contain) 84/3 50/4 0.428a
Tumor vessels(none/contain) 9/78 6/48 0.886
Early draining veins(none/contain) 28/59 22/32 0.366
Perfusion disorders(none/contain) 39/48 33/21 0.083
Intravascular tumor embolis(none/contain) 86/1 53/1 1.000a
a Fisher’s test.
A-C:Patient 1.Diagnosed as liver lesion during physical examination; lesion in left lobe, hepatic arterial phase shows tumor vessel originating from hepatic artery (black arrow); also shows early drainage veininflow into vena cava. Pathology:Epi-HAML.D-F:Patient 2.Hospitalized because of abdominal discomfort; lesion in right lobe, both T1-FS and T2-FS show heterogeneous signal intense without enhancement (white arrow). Pathology:Epi-HAML, hemorrhage with necrotic and cystic change inside. 图 2 2例上皮样型血管平滑肌脂肪瘤术前MRI图像 Fig 2 Preoperative MRI of 2 Epi-HAML cases
表 4 不同直径肿瘤的MRI特征 Tab 4 MRI features of HAML grouped by diameter  
(n)
Variable 0.7-2.9 cm (n=51) 3.0-5.0 cm (n=43) 5.1-20.5 cm (n=47) P
Pathology(Epi/non-Epi) 30/21 30/13 27/20 0.423
Diagnosis accuratenes[n(%)] 16(31.4) 17(39.5) 28(59.6) 0.016
Fat (none/diffuse/focal) 29/15/7 24/11/8 18/19/10 0.335
Capsules (none/contain) 43/8 38/5 31/16 0.018a
Necrosis or cysts (none/contain) 51/0 43/0 40/7 0.001a
Hemorrhage (none/contain) 51/0 43/0 40/7 0.001a
Tumor vessels (none/contain) 11/40 2/41 2/45 0.013a
Early draining veins (none/contain) 34/17 11/32 5/42 0.000
Perfusion disorders (none/contain) 42/9 18/25 12/35 0.000
Intravascular tumor embolis (none/contain) 51/0 43/0 45/2 0.131a
Tumor recurrence[n(%)] 1 (1.9) 0 3 (6.3) 0.215a
a Fisher’s test.
讨论

临床资料特点  HAML以中年女性多见,本研究男女之比1:4.1,与文献报道一致。Epi-HAML男女比例1:4.8,non-Epi-HAML为1:3.4,前者女性比例稍高,但差异无统计学意义。两组肝炎、肝硬化、脂肪肝、糖尿病等既往史虽有不同,但差异均无统计学意义。就诊者大多无症状,但Epi-HAML因腹部不适就诊为22.2%,而non-Epi-HAML仅为7.5%(P < 0.05),虽两组肿瘤平均直径相仿[(4.3±3.1)vs.(4.4±3.0)cm],且大于5 cm的肿瘤比例(31% vs. 37%)后者更高,推测是因为Epi-HAML生长速度更快,引起较强的占位效应,而non-Epi-HAML生长较缓慢,机体逐渐耐受。Pleniceanu等[11]在人源AML肿瘤接种实验中发现,上皮样型AML生长速度相对经典型AML更快,并且肿瘤含脂肪组织更少。

本组患者血清肿瘤标志物基本处于正常范围,仅有2例高于参考值,其中1例CA19-9为51.4 U/mL(0~37 U/mL),另1例为653 U/mL,CEA 111.5 ng/mL(0~5 ng/mL),且均为Epi-HAML,但差异无统计学意义。本研究共有4例术后复发(3.0%),与既往研究相似[4];其中,Epi-HAML组有3例(3.7%)复发,non-Epi-HAML组1例(31.9%)复发,两者差异无统计学意义。

影像学资料特点  133例共计141枚HAML,其中87枚为Epi-HAML,54枚为non-Epi-HAML,病灶多累及数个肝段。本研究两组HAML无论各肝段构成比及肝分叶构成比,差异均无统计学意义。141枚HAML影像学诊断准确率为43.3%(61/141),两组分别为34.5%(30/87)与57.4%(31/54),差异有统计学意义(χ2=7.134,P < 0.05),前者误诊为恶性肿瘤如原发性肝癌(HCC,47.1%)等恶性占位比率较高,后者误诊为腺瘤、局灶性结节增生(focal nodular hyperplasia,FNH)、脂肪瘤或良性结节(18.5%)等良性占位较高。目前HAML的诊断准确率约为25%~52%[12],部分临床中心甚至更低。相对而言,Epi-HAML大多缺乏典型的HAML特征,部分病灶增强扫描表现为“快进快出”,与无肝炎、肝硬化病史的HCC影像学表现相似[13]。有学者认为主要是因为部分HAML病灶内布满窦隙状血管网,血流速度较快,门静脉期及延迟期其对比剂快速流出,从而呈现“快进快出”的征象,尤其在Epi-HAML中更明显[14]

本组Epi-HAML在T1WI大都呈低信号(97.7%),极少呈混杂信号(2.3%),non-Epi-HAML也几乎呈低信号(98.1%),极少呈混杂信号(1.9%),构成比差异无统计学意义。Epi-HAML在T2WI大都呈高信号(95.4%),极少呈低或等信号(2.3%),而non-Epi-HAML部分呈高信号(81.5%),小部分呈稍低信号(11.1%),构成比存在统计学差异(Fisher’s检验,P < 0.05)。有研究认为,non-Epi-HAML因含一定比例平滑肌细胞,T2WI可以呈现等或低信号[15]。无论是Epi-HAML还是non-Epi-HAML,两者在组织学上均含有不同比例、不同种类的细胞,因此两者信号几乎都不均匀(94.3%/88.9%),且两组在T2WI均低于脾脏信号。

在增强扫描中可以观察到两组HAML病灶大部分呈不均匀强化(85.1%/79.6%),均匀强化者较少(6.9%/3.7%),另少量病灶呈环形强化(8.6%/13.0%),其强化方式差异无统计学意义。大部分病灶在强化程度上稍低于或等于门静脉信号,极少数Epi-HAML(3.4%)信号强度高于门静脉甚至与主动脉相当,但差异无统计学意义。增强扫描的强化模式,两组病灶大多呈持续减低(70.1%/57.4%),部分呈持续强化(9.2%/25.9%),18.4%的Epi-HAML呈快进快出,而non-Epi-HAML呈快进快出的比例为9.3%,差异有统计学意义(P < 0.05),这一点与Liu等[16]统计的Epi-HAML病灶出现“快进快出”频率为23.3%相似,且这部分病灶几乎均诊断为HCC。另外,有15.6%的病灶诊断为肝腺瘤或FNH,并且其强化方式大都呈现持续强化(图 1A-DH-K)。FNH一般在平扫与肝实质信号接近,典型的FNH可有中心瘢痕延迟强化,但部分FNH表现不典型,与HAML鉴别难度较大,需借助肝特异性对比剂的肝胆特异期来对比。乏脂的HAML与肝腺瘤表现相似,临床特点鉴别价值有限,目前仍是鉴别的难点。

两组病灶在DWI上大多呈高信号(92.0%/68.5%),但18.5%的non-Epi-HAML呈低信号,仅1.1% Epi-HAML呈低信号,其构成比存在统计学差异(P < 0.05)。DWI信号高低反映了水分子运动的受限程度,Epi-HAML含上皮样细胞的比例更高,而non-Epi-HAML很少或几乎不含有上皮样细胞,上皮样细胞增殖速度更快,排列更紧密,因此水分子扩散受限更明显[17]。而non-Epi-HAML尤其是富含脂肪的病灶,其细胞结构疏松,水分子受限不明显,因此出现部分富脂病灶表现为DWI低信号。

几乎所有病灶的表观扩散系数(apparent diffusion coefficient,ADC)图上呈现高低混杂信号,仅能以整体病灶同周围肝实质对比其信号数值的高低。总体上,两组ADC图上信号基本相似,整体高信号与整体低信号频率相似(43.7%/42.5%;44.4%/46.3%),总体构成比差异无统计学意义。从ADC的计算公式[ln(S2/S1)(b1-b2)]来看,不同的扩散敏感系数b值对其ADC变化也不尽相同,另外,相同的肿瘤存在异质性,其成分、细胞密度、血供以及肿瘤内部坏死、囊变、纤维化均会导致其信号的差异。不可否认,ADC对诊断病灶恶性程度有辅助作用,但有待深入研究[18]

HAML的成熟脂肪组织在T1反相位可见信号明显减低,Epi-HAML无明显脂肪信号者62.1%,而non-Epi-HAML大部分含有脂肪信号(68.5%,图 1E-G),少部分为局灶脂肪信号,其余为弥漫脂肪信号,两组病灶在有无脂肪信号上差异存在统计学意义(P < 0.05)。另外,两组病灶无论在影像学或者病理组织学标本上,均有大部分不含包膜(78.2%/77.2%)、病灶内坏死或囊变少见(4.6%/5.6%)及出血较少见(3.5%/7.4%)等特征,且两组差异均无统计学意义。

HAML的供血动脉一般在平扫见血管流空影,增强扫描动脉期见病灶内点状强化或条索状穿行强化影,门静脉期及延迟期信号减低,变化与肝动脉一致。两组病灶大多可见供血动脉影(89.7%/88.9%),差异无统计学意义。既往研究[19]发现增强扫描动脉晚期出现的引流静脉在HAML中有鉴别诊断价值,即增强扫描动脉晚期可见病灶引流向肝静脉或腔静脉的静脉影,本研究也类似(图 2A-C)。本研究HAML中引流静脉出现频率为64.5%(91/141),与文献统计的61.3%~80%相当,但两组出现引流静脉构成比(67.8%/59.2%)差异无统计学意义。肿瘤周围肝实质出现动脉期一过性高信号,其余序列与肝实质信号一致,即存在异常灌注,两组病灶出现异常灌注的构成比略有差异(44.8%/61.1%),但无统计学意义。两组影像学见脉管内充盈缺损或病理证实有脉管癌栓者各1枚,差异无统计学意义。

在读片过程中,我们发现不同直径的病灶(无论何种病理类型)其影像学特征出现的频率各有差别,因此,我们进一步以肿瘤直径(0.7~2.9 cm/3.0~5.0 cm/5.1~20.5 cm)作为分组(表 4)进行分析。当肿瘤直径较大(> 5 cm)时,有34%的病灶出现增强扫描后期的“包膜样”强化(Fisher’s检验,P < 0.05),而与病理类型关系不明显(P=0.675),其原因可能为当肿瘤生长过快或体积过大时,会刺激并推挤肝实质,导致其周围血管及纤维组织增生,形成假包膜,并非病灶本身形成,以往文献从病理组织的研究中也证实了这一点[20]。同时,出现坏死、囊变或出血的病灶直径均大于5 cm(图 2D-F),占本组的14.8%,构成比存在统计学差异(Fisher’s检验,P < 0.05)。此外,当以肿瘤直径作为分组,其供血动脉出现的比例也有所不同,3组分别为78.4%、95.3%、95.7%,< 3 cm组分别与另外两组的构成比相比存在统计学差异(Fisher’s检验,P < 0.05)。同样地,早期引流静脉出现的比例也存在差异,3组比例分别为33.3%、74.4%、89.4%(P < 0.05)。因此,引流静脉可能是HAML相对特异性的影像学特征,但多在肿瘤体积较大时(直径≥ 3 cm)明显可见,而与不同病理分类的HAML(即Epi-HAML或non-Epi-HAML)关系不明显。出现异常灌注的比例各组也均有不同,分别为17.6%、58.1%、74.5%(P < 0.05)。

此外,在复发病例中,分析其初诊影像资料,发现1枚(1.9%)直径为1 cm,3枚(6.3%)肿瘤直径大于5 cm,3组构成比差异无统计学意义。

总之,Epi-HAML缺乏典型的影像学特征,部分病例在随访过程中生长较快,与HCC有重叠征象,常引起误诊,尤其在合并乙肝病史的病例中,放射科医师更倾向于将其诊断为恶性肿瘤以提示外科行根治性切除。而non-Epi-HAML及直径较大的病灶有更显著的MRI特征,如坏死、囊变、出血、肿瘤供血动脉及引流静脉,因此有更高的检出率。此外,病灶的间接征象同样重要,当HAML病灶紧邻、包绕血管时无血管侵犯并且边界清晰,是与HCC鉴别的重要征象。大部分学者认为HAML为良性病变,但部分病灶体积过大易压迫周围器官,引起腹部不适等症状,甚至存在破裂出血的风险,所以针对直径大于3 cm的HAML,外科也主张以手术切除为主[10]

目前,乏脂型HAML与无肝炎、肝硬化病史的HCC鉴别难度较大,术前针对两者制定的切除方案也截然不同,相对准确的诊断对患者术后最大肝功能的保留意义非凡[21]。而我院肝脏肿瘤的病例丰富,如果借助影像组学进一步分析两者影像学特征,可以开展更深入的研究。

参考文献
[1]
MAO JX, TENG F, LIU C, et al. Two case reports and literature review for hepatic epithelioid angiomyolipoma:Pitfall of misdiagnosis[J]. World J Clin Cases, 2019, 7(8): 972-983. [DOI]
[2]
JUNG DH, HWANG S, HONG SM, et al. Clinico-pathological correlation of hepatic angiomyolipoma:a series of 23 resection cases[J]. ANZ J Surg, 2018, 88(1-2): E60-E65. [DOI]
[3]
MARTIGNONI G, PEA M, REGHELLIN D, et al. PEComas:the past, the present and the future[J]. Virchows Arch, 2008, 452(2): 119-132. [DOI]
[4]
KLOMPENHOUWER AJ, VERVER D, JANKI S, et al. Management of hepatic angiomyolipoma:A systematic review[J]. Liver Int, 2017, 37(9): 1272-1280. [DOI]
[5]
ZHAO DD, YUAN J, CHENG Q, et al. Evidence for a role of geranylgeranylation in renal angiomyolipoma and renal epithelioid angiomyolipoma[J]. Oncol Lett, 2019, 17(2): 1523-1530. [URI]
[6]
DAVIES M, SAXENA A, KINGSWOOD JC. Management of everolimus-associated adverse events in patients with tuberous sclerosis complex:a practical guide[J]. Orphanet J Rare Dis, 2017, 12(1): 35. [DOI]
[7]
DAMASKOS C, GARMPIS N, GARMPI A, et al. Angiomyolipoma of the liver:a rare benign tumor treated with a laparoscopic approach for the first time[J]. In Vivo, 2017, 31(6): 1169-1173. [PubMed]
[8]
DALLE I, SCIOT R, VOS R, et al. Malignant angiomyolipoma of the liver:a hitherto unreported variant[J]. Histopathology, 2000, 36(5): 443-450. [DOI]
[9]
KIM SH, KANG TW, LIM K, et al. A case of ruptured hepatic angiomyolipoma in a young male[J]. Clin Mol Hepatol, 2017, 23(2): 179-183. [PubMed]
[10]
LIU J, ZHANG CW, HONG DF, et al. Primary hepatic epithelioid angiomyolipoma:A malignant potential tumor which should be recognized[J]. World J Gastroenterol, 2016, 22(20): 4908-4917. [DOI]
[11]
PLENICEANU O, SHUKRUN R, OMER D, et al. Peroxisome proliferator-activated receptor gamma (PPARγ) is central to the initiation and propagation of human angiomyolipoma, suggesting its potential as a therapeutic target[J]. EMBO Mol Med, 2017, 9(4): 508-530. [DOI]
[12]
YANG X, LEI C, QIU Y, et al. Selecting a suitable surgical treatment for hepatic angiomyolipoma:a retrospective analysis of 92 cases[J]. ANZ J Surg, 2018, 88(9): E664-E669. [DOI]
[13]
WANG SY, KUAI XP, MENG XX, et al. Comparison of MRI features for the differentiation of hepatic angiomyolipoma from fat-containing hepatocellular carcinoma[J]. Abdom Imaging, 2014, 39(2): 323-333. [DOI]
[14]
ZHAO Y, OUYANG H, WANG X, et al. MRI manifestations of liver epithelioid and non-epithelioid angiomyolipoma[J]. J Magn Reson Imaging, 2014, 39(6): 1502-1508. [DOI]
[15]
程劲松, 韩津梁. 应用MRI鉴别小肾癌与乏脂型血管平滑肌脂肪瘤的研究[J]. 中国实验诊断学, 2018, 22(3): 478-480. [URI]
[16]
LIU W, WANG J, HUANG Q, et al. Comparison of MRI features of epithelioid hepatic angiomyolipoma and hepatocellular carcinoma:imaging data from two centers[J]. Front Oncol, 2018, 8: 600. [DOI]
[17]
KUBO H, YAMAZAKI H, OKADA T, et al. Primary hepatic angiomyolipoma:immunohistochemistry and electron microscopic observations:a case report[J]. J Med Case Rep, 2017, 11(1): 76. [DOI]
[18]
KURAMOTO K, BEPPU T, NAMIMOTO T, et al. Hepatic angiomyolipoma with special attention to radiologic imaging[J]. Surg Case Rep, 2015, 1(1): 38. [DOI]
[19]
YOSHIOKA M, WATANABE G, UCHINAMI H, et al. Hepatic angiomyolipoma:differential diagnosis from other liver tumors in a special reference to vascular imaging-importance of early drainage vein[J]. Surg Case Rep, 2015, 1(1): 11. [DOI]
[20]
刘孝臣, 徐鹏举, 胡国祥, 等. 肝脏上皮样血管平滑肌脂肪瘤的MRI表现及误诊分析[J]. 中国医学影像学杂志, 2018, 26(4): 285-287, 289. [URI]
[21]
杨新官, 刘光俊, 何敏丽, 等. 肝脏上皮样血管平滑肌脂肪瘤的CT和MRI表现特征[J]. 实用放射学杂志, 2017, 33(10): 1545-1548. [URI]

文章信息

于洋力, 曾蒙苏, 杨春, 盛若凡, 种欢欢
YU Yang-li, ZENG Meng-su, YANG Chun, SHENG Ruo-fan, CHONG Huan-huan
133例肝血管平滑肌脂肪瘤的MRI特征及分析
MRI features and analysis of 133 cases of hepatic angiomyolipoma
复旦学报医学版, 2020, 47(5): 660-668.
Fudan University Journal of Medical Sciences, 2020, 47(5): 660-668.
Corresponding author
ZENG Meng-su, E-mail: zeng.mengsu@zs-hospital.sh.cn.
基金项目
国家自然科学基金重大研究专项(培育项目)(91859107);上海市科学技术委员会“科技创新行动计划”产学研医合作领域项目(18DZ1930102);上海市科学技术委员会医学引导类(中、西医)科技支撑项目(19411965500)
Foundation item
This work was supported by the Major Research Project of National Natural Science Foundation of China (Cultivation Project) (91859107), Industry-University-Research-Medical Cooperation Project of"Science and Technology Innovation Action Plan"of Shanghai Science and Technology Commission (18DZ1930102), and Science and Technology Support Project of Medical Guidance (Chinese and Western Medicine) of Shanghai Science and Technology Commission (19411965500)

工作空间