【作者】 汤光宇；

【导师】 肖湘生；

【作者基本信息】 第二军医大学 ， 影像医学与核医学， 2008， 博士

【摘要】 第一部分MR影像报告数据系统鉴别良恶性乳腺肿块的价值初探目的评价美国放射学会的乳腺影像报告数据系统（BI-RADS）在动态增强MRI诊断乳腺癌中的价值,并分析动态增强参数最大增强线性斜率(S_max)、最大流入曲线斜率（K₁）的意义。方法81例疑似有乳腺肿块的患者术前做动态增强MRI检查（DCE-MRI）,参照BI-RADS评分标准进行评分、分级;并计算动态增强参数S_max和K₁。结果以手术病理结果为金标准,根据BI—RADS MR评分法:42例乳腺癌评为Ⅰ级1例,Ⅲ级6例,Ⅳ级12例,Ⅴ级的23例。39例良性占位评为Ⅰ级18例、Ⅱ级6例、Ⅲ级12例,Ⅳ级3例,Ⅴ级0例。两组间分级差异有统计学意义（x²=50.649,P＜0.0001）。BI—RADS评分法（将Ⅳ级以上的病例定为恶性）诊断乳腺癌的敏感性、特异性、准确性、阳性预测值、阴性预测值分别为83.3%、92.9%、91.4%、92.1%、84.8%。结合BI—RADS评分法和动态增强参数的阈值诊断乳腺癌,与单纯BI—RADS法比较,两者差别无统计学意义（u=0.907,P=0.330）结论BI—RADS评分法为乳腺癌的早期诊断制定了较规范、客观的诊断步骤和标准,结合动态增强参数S_max、K₁并不能提高良、恶性病变的鉴别诊断。第二部分乳腺病变MRI动态增强参数与血管生成的相关性研究目的探讨乳腺良、恶性病变MR动态增强（DCE）参数与肿瘤血管生成的关系。方法对51例乳腺良、恶性病变患者行前瞻性DCE-MR检查,计算动态增强参数:最大增强线性斜率(S_max)、增强峰值（PH）、峰值时间(T_peak)。所有病例手术病理标本行免疫组化染色,测定微血管密度（MVD）计数和血管内皮生长因子（VEGF）表达,分析各动态增强参数与乳腺癌MVD和VEGF表达的相关性。并比较29例乳腺癌与12例乳腺纤维腺瘤、10例乳腺病、10例癌旁正常组织MVD计数和VEGF表达的差别。结果乳腺癌S_max（2.04±1.8）（r=0.807,P＜0.001）、PH（678±260）（r=0.697,p＜0.001）与MVD呈正相关,T_peak（69±38）与MVD呈负相关（r=-0.425,p＜0.05）。乳腺癌组MVD计数[（65.09±15.81）个/200倍视野]明显高于纤维腺瘤组、乳腺腺病组及癌旁正常组织组（p值分别为0.043、0.018、0.002）。69%（20/29）乳腺癌VEGF表达阳性,也明显高于其余各组（p值分别为0.035、0.007、0.001）。乳腺癌边缘区域的MVD（60.38±24.14）高于中央区域（37.64±16.52）（t=2.635,P=0.016）。腋窝淋巴结转移组的MVD（73.23±23.02）高于无转移组（59.34±18.03）（t=2.303,p=0.031）。结论乳腺癌部分DCE-MR参数与MVD、VEGF相关,能较客观地反映乳腺癌血管生成状况,其MVD计数和VEGF表达明显高于乳腺良性病变及癌旁正常组织。第三部分乳腺癌MR动态增强参数与预后因子的相关性研究目的通过分析乳腺癌MR动态增强（DCE-MRI）形态学特征、动态增强参数与分子预后因子的相关性,探讨DCE-MRI活体评估乳腺癌预后的可行性。方法对34例乳腺癌患者术前行DCE-MRI检查,病理组织免疫组化染色记录分子预后因子ER、Ki-67、c-erbB-2、uPA/PAI-1的表达情况,用Spearman等级相关分析法探讨乳腺癌DCE-MRI形态学特征、时间信号强度曲线、动态增强参数（强化率、峰值时间、初始斜率、流出率）与预后因子的相关性。结果肿瘤毛刺状边缘与uPA/PAI-1阳性表达呈正相关（r=0.510）,内部强化均匀性与ER表达情况呈负相关（r=-0.332）,环形强化与淋巴结转移（r=0.560）、肿瘤病理学分级（r=0.354）、uPA/PAI-1阳性表达（r=0.314）呈正相关,与ER阳性表达呈负相关（r=-0.407）,内部间隔情况与病理学分型（r=-0.684）、uPA/PAI-1阳性表达（r=-0.370）呈负相关。廓清型TIC曲线与uPA/PAI-1阳性表达呈正相关（r=0.507）;峰值时间与ER阳性表达呈正相关（r=0.532）,与Ki-67（r=-0.542）、c-erbB-2（r=-0.535）、uPA/PAI-1（r=-0.640）阳性表达呈负相关。结论DCE-MRI不仅在乳腺肿瘤的定性诊断中起着重要作用,还可以间接预测肿瘤的预后,有潜在指导制定临床治疗方案的价值。其形态学特征和部分动态增强参数还与分子预后因子间存在相关性,肿瘤不均匀强化、环状强化、毛刺状边缘和动态增强短峰值时间、廓清型TIC常提示预后不良。更多还原

【Abstract】 partⅠEvaluation of BI-RADS category in the diagnosis of breast masses with dynamic contrast enhanced MRI.Objective:To evaluate the role of the breast imaging reporting and dada system （BI-RADS） categories developed by American College of Radiology and some parameters such as S_mak、K₁ in the diagnosis of breast masses with dynamic contrast enhanced magnetic resonance imaging（DCE-MRI）Methods:Eighty-one patients with suspected breast mass underwent preoperative DCE-MRI of the breast.The MR images were divided into five categories according to BI-RADS criteria（categoryⅠ,negative;categoryⅡ,benign;categoryⅢ,probably benign; categoryⅣ,suspicious for malignancy;categoryⅤ,highly suggestive of malignancy）. The parameters such as steepest slope(S_max);wash-in slope（K₁） were calculated based on time-signal intensity curves of DCE-MRI.Results:Pathologic findings were conducted as golden standard.Of 42 breast cancers,1, 0,6,12,23 cases were assigned to categoryⅠ,categoryⅡ,categoryⅢ,categoryⅣ, categoryⅤ,respectively.Of 39 benign lesions,18,6,12,3,0 cases were assigned to categoryⅠ,categoryⅡ,categoryⅢ,categoryⅣ,categoryⅤ,respectively.A significant difference was found between malignant and benign breast masses（x²= 50.649,P＜0.0001）.The sensitivity,specificity,accuracy rate,positive and negative predictive values of DCE-MRI according to BI-RADS（if category aboveⅣwere classified as malignant） were 83.3%、92.9%、91.4%、92.1%,84.8%,respectively.No significant difference between BI-RADS and BI-RADS combined with dynamic parameters was found in the diagnosis of breast cancer（u=0.907,P=0.330）.Conclusion:BI-RADS category is a standard and objective criterion for DCE-MRI in the diagnosis of breast cancer.However,dynamic parameters such as S_max,K₁ aren’t more helpful to the BI-RADS category in the differential diagnosis. PartⅡCorrelative study of the parameters of dynamic contrast- enbanced MRI and angiogenesis in breast lesionsObjective To evaluate the relationship between dynamic contrast-enhanced MRI （DCE-MRI）-derived parameters and tumor angiogenesis in malignant and benign breast lesions.Methods Fifty-one patients with malignant and benign breast diseases underwent DCE-MRI using Philips Intera 1.5 T MR system and dedicated breast coil prospectively before operation.DCE-MRI derived parameters such as steepest slope(S_max),peak height（PH）,time-to-peak(T_peak) were calculated based on time-signal intensity curve.The micro-vessel density（MVD） was counted and vascular endothelial growth factor（VEGF） expression was assessed in these patients after operation with immunohistochemical staining method.The parameters were correlated statistically with MVD counts and VEGF expression in breast cancer.The MVD counts and VEGF expression were also compared among the patients with breast cancer（29 case）,with fibroadenoma（12 case）,mastopathy （10 case） and the normal tissue（10 case） beside the cancer.Results The enhancement parameters S_max（r=0.807,p＜0.001）、PH（r=0.697,p＜0.001） correlated positively with MVD respectively.T_peak（69±38） correlated negatively with MVD counts（r=-0.425,P＜0.05）.The mean value of MVD（65.09±15.81/200 times field） in patients with breast cancer were significantly more than those with fibroadenoma,mastopathy and the normal tissue beside the cancer,respectively（p= 0.043,p=0.018,P=0.002）.69%（20/29）cases of breast cancers demonstrated positive VEGF expression,which were significantly more than the cases of fibroadenoma, mastopathy and the normal tissue beside the cancer（p=0.035,p=0.007,p= 0.001）,respectively.Moreover,the MVD counts（60.38±24.14） at the peripheral region of breast cancer were more than those at central region（37.64±16.52）（t=2.635,p=0.016）. There was a significant difference in MVD counts between breast cancers（73.23±23.02） with metastasis to axillary lymph nodes and those without metastasis（59.34±18.03）（t= 2.303,p=0.031）.Conclusion Some parameters derived from DCE-MRI correlated positively with MVD counts and VEGF expression in patients with breast cancer,which can reflect tumor angiogenesis objectively.The MVD counts and VEGF expression in patients with breast cancer were significantly more than those in patients with benign lesions. PartⅢCorrelation of the parameters of dynamic contrast- enhanced MRI with prognostic factors in breast cancerObjective To explore the possibility of dynamic contrast-enhanced MR imaging （DCE-MRI） in predicting the prognosis of breast cancer in vivo by correlating its morphological characteristics,the kinetic MR parameters on DCE-MRI with classical and molecular prognostic factors.Methods Thirty-four patients with breast cancer underwent dynamic contrast-enhanced MR imaging before operation.The morphological characteristics on DCE-MRI,the type of time signal-intensity curve（TIC） and the kinetic enhancing parameters(enhancement ratio, E_1-6;time to peak,T_peak;the initial slope,Slope_i and the washout ratio,W_peak-6) were recorded.The correlation between them and detected biomarkers（ER,c-erbB-2,Ki-67, uPA/PAI-1）,were statistically analyzed with the Spearman Rank Correlation.Results The spicular enhancing periphery of the tumor correlated positively with the the expression of uPA/PAI-1（r=0.510）.The internal homogenous enhancemant correlated negatively with the expression of ER（r=-0.332）.The ring-like enhancement correlated positively with the metastases to lymph node（r=0.560）,the pathological grade（r=0.354）, the expression of uPA/PAI-1（r=0.314） and negatively with the expression of ER（r= -0.407）.The wash-out type of TIC correlated positively with the expression of uPA/PAI-1 （r=0.507）.The T_peak correlated negatively with the expression of Ki-67（r=-0.542）, c-erbB-2（r=-0.535） and uPA/PAI-1（r=-0.640）.It also correlated positively with the expression of ER（r=0.532）.Conclusion DCE-MRI play not only a important role in the differential diagnosis of breast cancer,but also is indirectly helpful in vivo both in evaluating the prognosis and in drawing up the treatment plan of the breast carcinoma.The statistical significant correlation between morphological characteristics,some kinetic enhancing parameters on DCE-MRI and detected biomarkers was found.The internal inhomogenous enhancemant, ring-like enhancement,spicular enhancing periphery of the tumor and short time to peak, washing out TIC are suggestive of unhealthy prognostic factors.更多还原