【作者】 顾雅佳； 冯晓源； 肖勤； 吴斌； 彭卫军； 杨文涛； 唐峰； 毛健； 钱敏； 邵志敏；

【Author】 GU Ya-Jia Feng Xiao-Yuan Xiao Qin Wu Bin Peng Wei-Jun Yang Wen-Tao Tang Feng Mao Jian Qian Min Shao Zhi-Min Department of Radiology, the Affiliated Cancer Hospital of Fu Dan University, Department of Oncology, Shanghai Medical College. Fudan University, Shanghai 200032, China.

【机构】 复旦大学附属肿瘤医院；

【摘要】 目的:探讨利用磁共振弥散成像(DWI)和表观弥散系数(ADC)测定对乳腺癌范围确定的可行性。方法:测定57例59个乳腺癌在b分别取500 s／mm2和1000 s／mm2 时的ADC值,比较各种病理类型乳腺癌的平均ADC值,根据设定的ADC值测量不同b值时的肿瘤范围;比较动态增强和DWI测量病灶大小的异同。以两条径线作为比较参照:病灶最大径、过最大径线中点并与之垂直的径线。所有测量结果与病理对照。结果:1)59个病灶包括浸润性导管癌48个, 导管原位癌伴微浸润6个,黏液腺癌3个,髓样癌2个。在b 为500 s／mm2和b为1000 s／mm2时的平均ADC值依次为: 浸润性导管癌1．305±0．358×10-3 mm2／s和1．099±0． 300×10-3 mm2／s;髓样癌1．340±0．057×10-3 mm2／ s和1．140±0．311×10-3 mm2／s;导管原位癌伴微浸润1． 557±0．354×10-3 mm2／s和1．305±0．208×10-3 mm2／ s;黏液腺癌1．873±0．365×10-3 mm2／s和1．623±0．327 ×10-3 mm2／s。LSD法两两比较显示浸润性导管癌和黏液腺癌间差异有显著性,b=500 s／mm2时P=0．010,b=1000 s／mm2时P=0．004,余两两比较差异均无显著性。2)设定b =500 s／mm2时ADC值1．5×10-3 mm2／s,b=1000 s／ mm2时ADC值1．3×10-3 mm2／s作为测定值的上限,小于此值的异常区域作为弥散所测病灶大小,两组结果与病理诊断的肿瘤范围比较。范围一致b=500 s／mm2组略高于b= 1000 s／mm2组,但无统计性差异(X2=0．160,P=0．689), 过度诊断(小于ADC设定值区域大干实际肿瘤20％)两组一致(2个),假阴性(所测ADC值大干所设定的值)b=500 s／mm2组略低于b=1000 s／mm2组(X2=0．172,P=0．679)。两组均诊断错误14个,分别是过度诊断2个(病理均为浸润性导管癌伴周围较明显的不典型增生改变),假阴性12个,分别为导管原位癌伴微浸润2个、浸润性导管癌伴比较明显坏死或出血胶元化改变4个、黏液腺癌3个和浸润性导管癌3个。 8个病灶2组表现不一致,5个在b为500 s／mm2时诊断正确的病灶中3个是导管原位癌伴微浸润。3)以4分钟时测定的大小作为动态增强显示病灶的测定点,与同一层面DWI图上显示的异常区域大小进行比较。两者符合47个(80％),增强径线测定较小而DWI测定符合8个,其中3个为黏液腺癌 (100％),5个为浸润性导管癌3级(83％)。结论:磁共振弥散成像(DWI)和表观弥散系数(ADC)测定可以对乳腺癌范围进行评价。对导管原位癌(伴微浸润)来说b取500 s／ mm2对测量病灶大小更为准确。伴有出血坏死胶元化的浸润性导管癌、黏液腺癌和较小的导管原位癌伴微浸润容易表现为ADC测量假阴性。相比较动态增强,对黏液腺癌和浸润性导管癌3级的范围评价,弥散成像有其优势。更多还原

【Abstract】 Objective: To investigate the feasibility with diffusion-weighted imaging (DWI) and the apparent diffusion coefficient (ADC) value in detecting accuracy of the cancer extension. Methods: We used DWI to obtain images of 59 lesions (57 patients) before surgical excision. The ADC values of vary breast cancer type were compared. The cancer extension was invested in the different b value ADC map, which represents the distribution of ADC values, according to the threshold values we discussed before. The lesion extension confirmed in enhanced image and in DWI map was also compared. The tumor extension was determined by calculating two lines. Line one: the maximum diameter of lesion. Line two: perpendicular crossed the midpoint at line one. All measurement was compared with the pathologic specimen. Results: 1) The mean ADC value of each cancer type in b at 500 and 1000 s/mm2 was as follows respectively: invasive ductal carcinoma (n=48), 1.305 ± 0.358 × 10-3 mmVs and 1.099 ± 0.300 × 10-3 mm2/s, medullary carcinoma (n=2), 1.340 ± 0.057 × 10 -3 mmVs and 1.140 ± 0.311 × 10 -3 mm2/s, ductal carcinoma in situ with small invasive foci (n=6), 1.557 ± 0.354 × 10-3 mm2/s and 1.305 ± 0.208 × 10 -3 mm2/s, muci-nous carcinoma (n=3), 1.873 ± 0.365 × 10 -3 mm2/s and 1.623 ± 0.327 × 10 -3 mm2/s. No signiflcant differences were found between each type of breast cancer but between invasive ductal carcinoma and mucinous carcinoma at b was 500 s/mm2 (P=0.01) and 1000 s/mm2 (P=0.004). 2) Based on the upper threshold 1.5 × 10-3 mm2/s, 1.3 × 10 -3 mm2/s at b was 500 s/mm2 and 1000 s/mm2 respectively, a near precise distribution of low ADC value on ADC maps was described as cancer extension. The measure results were compared to pathologic figures. We categorized the pattern of correlation into 3 groups: Group 1, where the area of low ADC values was almost the same the pathological tumor extension; Group 2 (overdiagnose), where the area of low ADC values was wider and more than 20% the area of tumor extension; Group 3 (false negative), where no low ADC value was observed. There were all no significant difference in Groups 1 and Group 3 between DWI at b was 500 s/mm2 and 1000 s/mm2 (X2 = 0.160, P=0.689; X2 = 0.172, P=0.679). There were 2 lesions in Group 2, which were consistent in DWI of 500 s/mm2 and 1000 s/ mm2. There were 14 lesions misdiagnose, including overdiagnose 2 lesions and false negative 12 lesions. The former were invasive ductal carcinoma with notable atypical ductal hyperplasia. The latter were ductal carcinoma in situ with small invasive foci (2), invasive ductal carcinoma with marked bleeding necrosis and/or collagen (4), mucinous carcinoma (3) and invasive ductal carcinoma (3). Eight lesions measured at DWI of 500 s/mm2 and 1000 s/ mm2 were not consistent Five lesions were diagnosed correctly at DWI of 500 s/mm2, three of them were ductal carcinoma in situ with small invasive foci. 3) The extension of lesion on dynamic enhanced imagi更多还原