【作者】 沈亚琪；

【导师】 胡道予；

【作者基本信息】 华中科技大学 ， 影像医学与核医学， 2010， 博士

【摘要】 第一部分肝纤维化实验模型的建立目的构建稳定、可靠、肝纤维化实验动物模型,研究模型肝脏病变的进展情况和常规影像学表现,为影像学功能成像定量评价做准备。方法50只雄性清洁级Sprague-Dawley鼠,30只普通级雄性新西兰大白兔,随机分为实验组(新西兰大白兔n=20；SD大鼠n=40)和对照组(新西兰大白兔n=10;SD大鼠n=10)。采用CCL4(四氯化碳)腹腔注射法诱导肝纤维化鼠模型,血吸虫尾蚴腹贴法诱导肝纤维化兔模型。肝纤维化鼠模型在20W时,肝纤维化兔模型在8W、10W、12W分别行MDCT和MRI检查。以HE染色、Masson三色染色、透射电镜观察两种肝纤维模型肝脏病变,并对其纤维化程度和炎症进行分级分期。结果肝纤维化兔模型4W时仅肝脏观察到炎性浸润,6W时出现急性虫卵结节,8W时观察到肝纤维化胶原沉积,10W及12W胶原沉积明显加重。病理学证实两种肝纤维化模型成功建立,两种模型纤维化程度相似,但是炎症评分肝纤维化鼠明显高于肝纤维化兔,具有统计学意义。肝纤维化鼠镜下脂肪变较肝纤维化兔明显。MDCT和MRI扫描可以显示两种肝纤维化模型肝体积缩小、腹水等间接征象。MDCT后处理重建计算肝脏体积,两种模型组体积小于对照组,并具有统计学差异。肝纤维化兔模型MRI扫描多次观察发现门脉分支扩张、门脉压力增高表现,并随时间而进展加重。结论结合两种肝纤维化模型,有助于了解复杂的人类自然病程。MDCT和MRI扫描可以观察到肝纤维化模型肝体积缩小、腹水、门脉压力增高等间接表现,但无法提供定量评价指标。第二部分两种肝纤维化动物模型DTI定量研究目的两种肝纤维模型行DTI扫描并与病理分级行相关分析,初步探讨了DTI在肝纤维化定量研究的可行性及应用价值。方法CC14诱导肝纤维化SD大鼠模型10只,同期对照5只及血吸虫腹贴法诱导新西兰大白兔肝纤维化模型18只,同期对照10只。肝纤维化鼠模型在造模后20W及断头处死后4h内离体肝组织,肝纤维化兔模型在造模后8W、10W、12W分别行DTI扫描(EPI, TR/TE=1600/68.1,6个方向,b=500s／mm2),完成扫描后以HE染色、Masson三色染色、透射电镜观察两种肝纤维模型肝脏病变,并对其纤维化程度和炎症进行分级分期。扫描图像经AW4.2工作站处理,计算并分析肝纤维化鼠模型和对照的在体和离体ADC值和各向异性值(FA),肝纤维化兔模型各个时期的ADC值和各向异性值(FA),并与病理分级行相关性分析,采用SSPS13.0进行统计分析,显著水平取0.05。结果病理学证实两种肝纤维化模型均发生肝纤维化改变,两种模型纤维化程度分布相似,但是炎症评分肝纤维化鼠明显高于肝纤维化兔,具有统计学意义,肝纤维化鼠镜下脂肪变较肝纤维化兔明显。两种肝纤维化模型中ADC值和FA值改变一致,均表现为ADC值较对照组降低,FA值较对照组升高。具有统计学意义指标为肝纤维化鼠在体ADC值,肝纤维化兔ADC值。肝纤维化兔模型ADC值随着模型时间进展而减低,并且不同病理分级之间差别具有统计学意义。两种肝纤维化模型ADC值与肝纤维化分级之间存着负相关,并具有统计学意义。(肝纤维化鼠在体ADCr=-0.607P=0.016；肝纤维化兔,r=-0.765 P=<0.01)结论两种肝纤维化模型病理分期与ADC值均有相关性,ADC值可成为定量评价肝纤维化指标。第三部分两种肝纤维化动物模型1H-MRS定量研究目的两种肝纤维模型行1H-MRS扫描并与病理分级行相关分析,初步探讨了1H-MRS在肝纤维化定量研究的可行性及应用价值。方法CC14诱导肝纤维化SD大鼠模型10只,同期对照5只及血吸虫腹贴法诱导新西兰大白兔肝纤维化模型18只,同期对照10只。肝纤维化鼠模型在造模后20W,肝纤维化兔模型在造模后8W、10W、12W分别行1H-MRS扫描(PRESS, TR/TE=1500/35, FOV=16×16cm, NEX8,不抑水,时间2分12秒),完成扫描后以HE染色、Masson三色染色、透射电镜观察两种肝纤维模型肝脏病变,并对其纤维化程度和炎症进行分级分期。扫描图像经sage后处理,分别测量两种肝纤维化模型糖原和葡萄糖复合物(Glyu)、胆碱(Cho)、谷氨酰胺和谷氨酸复合物(Glx)、甘油三酯及其化合物(Lip)、水(Water)的平均峰高,并计算Fat%百分比[FAT=Lip/(lip+water)×100%],并与病理分级行相关性分析,采用SSPS13.0进行统计分析,显著水平取0.05。结果病理学证实两种肝纤维化模型均发生肝纤维化改变,两种模型纤维化程度分布相似,但是炎症评分肝纤维化鼠明显高于肝纤维化兔,具有统计学意义,一肝纤维化鼠镜下脂肪变较肝纤维化兔明显。两种肝纤维化模型MRS均可以观察到Cho峰随着纤维化分期升高表现,Cho峰高在两种模型中均有统计学差异,并与肝纤维化分级呈正相关(肝纤维化鼠r=0.642,P<0.01；肝纤维化兔,r=0.643,P<0.01)。肝纤维化鼠模型Fat高于肝纤维化兔。结论两种肝纤维化模型病理分期与Cho峰高均有相关性,Cho峰可成为定量评价肝纤维化指标。第四部分两种肝纤维化动物模型T2*MAP定量研究目的两种肝纤维模型行T2*Map扫描并火焰原子光谱法测定肝铁含量相对照,并与病理分级行相关分析,初步探讨了T2*Map在肝纤维化定量研究的可行性及应用价值。方法CC14诱导肝纤维化SD大鼠模型10只,同期对照5只及血吸虫腹贴法诱导新西兰大白兔肝纤维化模型18只,同期对照10只。肝纤维化鼠模型在造模后20W,肝纤维化兔模型在造模后8W、10W、12W分别行T2*Map扫描(GRE,TR=100,FA=40,FOV=10×10cm,矩阵160×128,NEX6),完成扫描后以HE染色、Masson三色染色、透射电镜观察两种肝纤维模型肝脏病变,并对其纤维化程度和炎症进行分级分期。扫描图像经AW4.2工作站后处理,分别测量两种肝纤维化R2*值并与火焰原子光谱法测定肝铁含量相对照,并与病理分级行相关性分析,采用SSPS13.0进行统计分析,显著水平取0.05。结果病理学证实两种肝纤维化模型均发生肝纤维化改变,两种模型纤维化程度分布相似,但是炎症评分肝纤维化鼠明显高于肝纤维化兔,具有统计学意义,肝纤维化鼠镜下脂肪变较肝纤维化兔明显。采用Spearman’s法对肝铁铁含量和R2*值行相关性分析,两者呈明显正相关,r=0.966,P<0.01,肝纤维化鼠模型肝铁含量明显高于肝纤维化兔模型,而两者的对照组差别不大。两种肝纤维化模型R2*值较对照升高,并具有统计学意义。肝纤维化兔R2*值随模型进展时间升高,并与肝纤维化分级呈正相关(r=0.807,P<0.01)。结论两种肝纤维化模型R2*值较对照升高,R2*值可成为定量评价肝纤维化模型铁沉积指标。肝纤维化鼠模型肝铁含量较肝纤维化兔模型高,可能与其炎症和脂肪变程度较重相关。第五部分T2*MAP定量研究的临床初步应用目的利用T2*Map扫描研究肝铁含量并与火焰原子光谱法测定肝铁含量相对照,初步探讨了T2*Map对肝铁含量进行定量研究的可行性及应用价值。方法肝内占位26例患者术前三天内行T2*Map扫描(FGRE, TR=100, FA=40,FOV=40×40cm,层厚7mm,层间距lmm,矩阵160×128,NEX2)。扫描图像经AW4.2工作站后处理,分别测量肝内肿瘤及肿瘤旁肝组织R2*值并与火焰原子光谱法测定肿瘤旁肝组织铁含量相对照。采用SSPS13.0进行统计分析,显著水平取0.05。结果术后病检证实原发性肝癌23例(中低分化11例,中分化8例,中高分化4例),血管瘤1例,局灶性结节增生2例。采用Spearman法对组织铁含量和R2*值行相关性分析,两者呈正相关(r=0.567,P=0.003)。肝癌患者癌旁肝组织的肝铁含量明显高于良性病变患者,但无统计学意义。肝癌组织的R2*低于癌旁组织,具有统计学意义。原发性肝癌病理分级之间的R2*值的差异,无统计学意义。结论R2*值可成为定量评价肝铁沉积指标。肝癌组织的R2*值低于癌旁组织,可作为鉴别肝内占位的量化指标之一。更多还原

【Abstract】 Part 1 The construction of experimental animal models with liver fibrosisPurpose To construct a stable, reliable, and consistently animal models with liver fibrosis; to study liver lesions on MRI and CT, and to prepare for the further quantitative analysis and assessment for the functional MRI study.Methods Fifty male Sprague-Dawley rats and thirty male New Zealand rabbits were used in this study,which were randomly assigned to the experimental group (rabbits n=20;rats n=40) and the control group (rabbits n=10;rats n=10). To duplicate model of liver fibrosis, the rats were intraperitoneal injected with CCL4,and the rabbits were transcutaneously infected with schistosomiasis cercariae. The rats were performed with MRI and MDCT after 20 weeks time, and the rabbits were performed with MRI and MDCT in the 8th,10th and 12th week. The observation on liver lesion was conducted by HE staining, Masson trichrome staining, transmission electron microscopy, and meanwhile the degree of fibrosis and inflammation was pathologically graded.Results Only inflammatory infiltration were observed in the 4th week in the rabbits infected with schistosomiasis cercariae.The acute foreign granuloma node formed by worm eggs were observed in the 6th week,and then collagen deposition were observed in the 8th week. After that, the degree of collagen deposition was increased on the 10th and 12th week.All the experimental models were successfully duplicated with liver fibrosis,the score of liver fibrosis were similar in two kinds of models,but the score of inflammation between two kinds of models showed statistically significant difference,and the rats gain higher score than the rabbit did.MDCT and MR imaging showed indirect signs,such as liver size shrunk,and ascites. The liver volume were calculated with post-processed method on MDCT,and the volume of model groups were statistically significantly lower than that in the control groups.And the multiple observation on MRI showed that the branch of portal venous were extended,and the degree were aggravated over times.Conclusion These two kinds of experimental animal models with liver fibrosis can help to understand the complex condition of human natural evolution. MDCT and MRI imaging can show the indirect sign of liver size reduction and ascites.Part 2 Quantitative study of DTI in experimental animal models with liver fibrosisPurpose Two kinds of experimental animal models were performed with DTI,and the results were correlated with the pathology results,to explore the feasibility and value of DTI in quantitative study of liver fibrosis.Methods Two kinds of experimental animal models of liver fibrosis (10 rats were intraperitoneal injected with CCL4 and 18 New Zealand rabbits were transcutaneously infected with schistosomiasis cercariae) and the control ones (5 rats and 10 rabbits) were used in this study. The rats were performed with DTI imaging 20 weeks later and in vitro,and the rabbits were performed with DTI imaging in the 8th,10th and 12th week (EPI, TR/TE=1600/68.1, b=500s/mm2).The observation on liver lesion was conducted by HE staining, Masson trichrome staining, transmission electron microscopy, and meanwhile the degree of fibrosis and inflammation was pathologically graded. The apparent diffusion coefficient (ADC) and the fractional anisotropy(FA)of liver were measured on rats(in vivo and in vitro) and rabbits (8th week,10th week,12th week),the results were correlated with the pathology results. Statistical analysis was using SPSS13.0, take significant level of 0.05.Results Pathology results confirmed two kinds of experimental animal models were successfully estabished. There was no statistically significant difference between the two model groups for the degree of liver fibrosis.And degree of the inflammation in the rat model group was statistically significant difference were higher than that in the rabbit model group.Meanwhile, steatosis of liver was widely distributed in the rat model group.The change of ADC values and FA value in the two model groups was similar, compared with the controls, the ADC values was decrease and FA value was increase.The ADC values of rats (in vivo)and ADC values of rabbits were both had statistically significant different from the control groups.The ADC values of rabbit model group were decrease as time pass by,and the ADC values between different pathological grading were statistically significant difference.There was significantly negative correlation between ADC values and pathological grading of fibrosis(in vivo ADC of rats r=-0.607 P=0.016; rabbits r=-0.765 P=<0.01).Conclusion There was significantly negative correlation between ADC values and pathological grading of fibrosis in the two kinds of experimental animal models with liver fibrosis.ADC values can be quantitative evaluated with liver fibrosis.Part 3 Quantitative study of’H-MRS in experimental animal models with liver fibrosisPurpose Two kinds of experimental animal models were performed with 1H-MRS,and the results were correlated with the pathology results.To explore the feasibility and value of ’H-MRS in quantitative study of liver fibrosis.Methods Two kinds of experimental animal models of liver fibrosis (10 rats were intraperitoneal injected with CCL4 and 18 New Zealand rabbits were transcutaneously infected with schistosomiasis cercariae) and the control ones(5 rats and 10 rabbits) were used in this study. The rats were performed with 1H-MRS imaging after 20 weeks,and the rabbits were performed with 1H-MRS in the 8th,10th and 12th week (PRESS, TR/TE= 1500/35, NEX8, none water suppress).The observation on liver lesion was conducted by HE staining, Masson trichrome staining, transmission electron microscopy, and meanwhile the degree of fibrosis and inflammation was pathologically graded. The 1H-MRS analysis was performed on workstation, and the results were correalted with the pathology results. Statistical analysis were using SPSS 13.0, take significant level of 0.05Results Pathology results confirmed two kinds of experimental animal models were successfully estabished. There was no statistically significant difference between the two model groups for the degree of liver fibrosis.And degree of the inflammation in the rat model group was statistically significant difference were higher than that in the rabbit model group.Meanwhile, steatosis of liver was widely distributed in the rat model group. A significant statistical difference was observed between the control group vs.the liver fibrosis groups in choline (Cho). There is significantly positive correlation between Cho and pathological grading of fibrosis(rats r=0.642, P<0.01; rabbits r=0.643, P<0.01)Conclusion There was significantly positive correlation between Cho and pathological grading of fibrosis in the two kinds of experimental animal models with liver fibrosis.Cho can be quantitative evaluated with liver fibrosis.Part 4 Quantitative study of T2*map in experimental animal models with liver fibrosis Purpose Two kinds of experimental animal models were performed with T2*Map, and the results were correlated with the pathology results.To explore the feasibility and value of T2*Map in quantitative study of liver fibrosis.Methods Two kinds of experimental animal models of liver fibrosis (10 rats were intraperitoneal injected with CCL4 and 18 New Zealand rabbits were transcutaneously infected with schistosomiasis cercariae) and the control ones (5 rats and 10 rabbits) were used in this study. The rats were performed with T2*Map after 20 weeks,and the rabbits were performed with T2*Map in the 8th,10th and 12th week (GRE, TR=100, FA=40, NEX 6). The observation on liver lesion was conducted by HE staining, Masson trichrome staining, transmission electron microscopy, and meanwhile the degree of fibrosis and inflammation was pathologically graded.The R2* value were measured on workstation, and the results were correalted with the pathology results. Statistical analysis were using SPSS13.0, taken significant level of 0.05Results Pathology results confirmed two kinds of experimental animal models were successfully estabished. There was no statistically significant difference between the two model groups for the degree of liver fibrosis.And degree of the inflammation in the rat model group was statistically significant difference were higher than that in the rabbit model group.Meanwhile, steatosis of liver was widely distributed in the rat model group. There is significantly positive correlation between R2* values and liver iron content (r=0.966, P<0.01).And the liver iron content in liver fibrosis rats was higher than that in liver fibrosis rabbits. The R2* values of liver fibrosis rabbit were increase as time pass by. There was significantly positive correlation between R2* values of liver fibrosis rabbits and pathological grading of fibrosis(r=0.807, P<0.01)Conclusion There was significantly positive correlation between R2* values and pathological grading of fibrosis. R2* values can be quantitative evaluated with liver fibrosis.Part 5 Quantitative study of T2*map in Clinical application:Preliminary studiesPurpose To explore the feasibility and value of T2*Map in quantitative study of liver iron content in a clinic study.Methods Twenty-six patients with focal liver lesion were performed with T2*Map (FGRE, TR=100, FA=40, NEX 2). The R2* value of liver lesion and liver tissue around the lesion were measured on workstation, and the results were correalted with the pathology results. Statistical analysis was using SPSS13.0, take significant level of 0.05.Results All the patients were verified by postoperative pathological examination,there were 23 cases of primary liver carcinoma(11 were poorly-differentiated,8 were intermediately-differentiated and 4 were well-differentiated),1 case of hemangioma and 2 cases of focal nodular hyperplasia (FNH).There is significantly positive correlation between R2* values and liver iron content (r=0.567,P=0.003). R2* values of liver in patiens with malignant tumor were higher than the one with benign tumor, whereas no significant statistical difference was observed. A significant statistical difference was observed between liver tissue vs. liver tumor in R2* values.There is no significant statistical difference among different pathological grading of liver tumor.Conclusion R2* values can be quantitative evaluated with liver iron content.The R2* values were lower in liver tumor than the normal liver,which can be used as a feature to identify liver lesion.更多还原