【作者】 王世江；

【导师】 王平； 袁智勇； 朱莉； 王佩国；

【作者基本信息】 天津医科大学 ， 肿瘤学， 2008， 硕士

【摘要】 目的:建立成功的兔VX2肺癌模型,进行兔肺癌的多层螺旋CT普通平扫和灌注扫描,对肺癌标本进行免疫组织化学检查,分析兔VX2肺癌模型多层螺旋CT灌注扫描图像各灌注参数和肿瘤乏氧标记物表达水平的相对关系,探讨多层螺旋CT灌注扫描图像和肿瘤乏氧区间的对应关系。方法:17只新西兰大白兔,在CT引导下行肺部原位成瘤,肿瘤生长至1.5～2cm行动态CT灌注扫描,应用Perfusion 3(GE Medical Systems)灌注软件的体部肿瘤程序分析,运用高空间分辨率逐像素计算并显示能够反映肿瘤组织血流灌注状态的伪彩灌注参数图像,包括血流量(blood flow,BF)、血容量(blood volume,BV)、平均通过时间(mean transit time,MTT)和表面通透性(permeability surface,PS)等相关灌注图像,流入动脉选择降主动脉,以血容量三色图(红色,绿色,蓝色)为基准,范围为0-10000Hu,定义红色、绿色区域为高灌注区,蓝色及低于蓝色区域均为低灌注区。在肿瘤高灌注区、低灌注区各选取3个感兴趣区(region of interest ROI),取平均值为相关灌注参数绝对值,测量时避开灌注BV图像显示数值为0的区域。取肿瘤病理标本行免疫组化检查(链霉菌抗生物素蛋白-过氧化酶免疫组化方法,SP法),检测HIF-1α、VEGF、MVD的表达情况。结果:成功完成了兔VX2肺癌模型的建立,成功进行了兔肺癌的多层螺旋CT灌注扫描,以血流量三色图(红色,绿色,蓝色)为基准的高灌注区与低灌注区比较:BV、BF、PS两组差异在统计学上均有意义(F值分别为18.2、1、1,P值均小于0.05),MTT两组在统计学上无明显差异(F值为1,P=0.0978),HIF-1、VEGF两组差异在统计学上均有意义(F值分别为28、24.7,P值均小于0.05),MVD两组差异在统计学上无明显意义(F值为32,P=0.5654)。行组内Spearman相关性分析,高灌注区BV与BF、MTT呈正相关(r=0.60539、0.54881,P=0.0100、0.0225),与PS无明显相关(r=-0.06618,P=0.8008),BF与MTT、PS无明显相关(r=-0.05488、-0.17892,P=0.8343、0.4920),MTT与PS无明显相关(r=0.09467,P=0.7178);低灌注区中BV与BF、PS呈正相关(r=0.77206、0.77206,P=0.0003、0.0003),与MTT无明显相关(r=-0.05762,P=0.8261),BF与MTT、PS亦呈明显相关(r=-0.58373、0.62990,P=0.0139、0.0067),MTT与PS无明显相关(r=0.00752,P=0.9772)。高灌注区HIF-1α和BV、BF在统计学上均有明显相关(r值分别为-0.74510、-0.50000,P值分别为0.0006、0.0410),低灌注区HIF-1α和BV、BF、PS在统计学上均有明显相关(r值分别为-0.5028、-0.4880、-0.5261,P值分别为0.0397、0.0469、0.0301),MVD值与PS在统计学上明显相关(r值为0.64490,P值为0.0052)。在高低灌注区HIF-1和BV均在统计学上有明显的相关(P=0.0006、0.0397,r=-0.7451、-0.5028)。结论:在高低灌注区,HIF-1和BV有明显的统计学相关,血容量的改变,会影响肿瘤的供氧情况,提示CT灌注图像高低灌注区和乏氧区有明显的相关性,运用多层螺旋CT灌注图像低灌注区域可初步标记肿瘤乏氧区的范围。更多还原

【Abstract】 Objective:To establish the model of rabbit VX2 Lung Cancer successfully,perform conventional of multi-slice CT(MSCT)plain scan and perfusion scan,immunohistochemical staining for the tumor specimen.To analyze the correspondence relationship between every parameter in MSCT perfusion imaging and the express level of tumor hypoxia marker,investigate the correlation between the CT perfusion and hypoxia imaging in Rabbit VX2 Lung Cancer.Methods:17 female New Zealand white rabbits were implanted with the VX2 tumor in lung.MSCT plain scan and perfusion scan were performed when tumor was big in the scale 1.5～2cm.With the help of perfusion 3 software(GE Medical Systems), applying the method of high space resolution,we calculated and displayed the falsecolor perfusion parameter image,which can reflect the condition of tumor tissue perfusion, blood flow(BF),blood volume(BV),mean transit time(MTT)and permeability surface(PS)mapping images were acquired.The descending aorta as inflow artery,using the blood flow three-color-image(red,green,blue)as the discriminating standard,its scale was 0-10000Hu.we used red and green scale reflecting high perfusion region,blue and other regions which lower than blue reflecting low perfusion region. In the high and low perfusion image,we select three region of interest(ROI) respectively,using their mean value as the last value of every correlation perfusion parameter.during measurement,we should avoid the regions whose blood flow were zero in perfusion image.After CT perfusion,the rabbit was sacrificed and the corresponding tumor specimen was harvested.H E staining,hypoxia,VEGF and MVD immunohistochemical staining(streptavidin peroxidase method)were carried out.Results:Established the model of rabbit VX2 Lung Cancer successfully.Performed conventional of multi-slice CT(MSCT)plain scan and perfusion scan to rabbit VX2 Lung Cancer.To compare high perfusion region and low perfusion region,using the blood flow three-color-image(red,green,blue)as the discriminating standard,The mean values of BF、BV and PS had obviously difference in statistic between high and low perfusion(F=18.2,1,1,all the P value less than 0.05).HIF-1,VEGF had obviously difference in statistic between high and low perfusion(F=28,24.7,all the P value less than 0.05).MVD had no difference between two groups(F=32,P=0.57).The result of Spearman correlation analyze in group:in high perfusion region,BV and BF、MTT have positive correlation(r=0.60539,0.54881,p=0.0100,0.0225).BF and MTT,PS have no obviously correlation(r=-0.05488,-0.17892,p=0.8343,0.4920),MTT and PS have no obviously correlation(r=0.09467,p=0.7178);in low perfusion region, BV and BF、PS have positive correlation(r=0.77206,0.77206,p=0.0003,0.0003),but have no obviously correlation with MTT(r=-0.05762,p=0.8261),BF and MTT,PS also have obviously correlation(r=-0.58373,0.62990,p=0.0139,0.0067),MTT and PS have no obviously correlation(r=0.00752,p=0.9772).HIF-1αand BV,BF all have negative correlation in statistics(r=-0.74510,-0.50000,P=0.0006,0.0410)in high perfusion region,HIF-1 and BV,BF,PS all have negative correlation in statistics in low perfusion region(r=-0.50276,-0.48804,-0.52606,P=0.0397,0.0469,0.0301),MVD and PS have positive correlation in statistics(r=0.64490,P=0.0052).the HIF-1 had negative correlation with BV both in two groups(P=0.0006,0.0397,r=-0.74510, -0.50276).Conclusion:There were obviously difference between HIF-1 and BV in statistic in high and low perfusion region.the change of blood flow would influence tumor’s oxygen supply,which is a hint that high and low perfusion region of CT perfusion image have obviously correlation with hypoxia image in rabbit’s lung cancer.Low perfusion region of CT perfusion image could reflect the tumor hypoxia region initially.更多还原