【作者】 李洁； 张晓鹏； 孙应实； 曹崑； 唐磊； 单军； 汪宁；

【Author】 Li Jie Zhang Xiao-Peng Sun Ying-Shi Cao Kun Tang Lei Shan Jun Wang Ning Beijing Cancer Hospital & Clinical oncology Institute, Peking University, Radiology department, Beijing, 100036, Beijing

【机构】 北京肿瘤医院暨北京大学临床肿瘤学院放射科；

【摘要】 目的:评价动态增强MRI对乳腺癌新辅助化疗后残存病变检出及病理反应性的术前诊断价值。方法:45例浸润性乳腺癌新辅助化疗结束后行动态增强乳腺MRI检查, 而后行乳腺癌手术治疗。由一位高年资病理医生对所有病理切片进行复阅,根据手术标本的病理检查结果,参照miller &payne病理反应性分级标准对化疗后乳腺癌残存病变的病理反应性进行评价,将5级(无镜下可见的浸润癌组织) 判为病理完全缓解(Pathological complete response,pCR), 4级和5级作为组织学显著反应(major histological response,MHR)。由2名具有乳腺MRI诊断经验的放射医生对动态增强M R I图像分别进行盲法独立阅片,在 Functool诊断工作站采用感兴趣区法对强化病变的曲线类型进行观察。原肿瘤区域任何非血管性强化提示肿瘤病变残存,无强化提示病理理完全缓解,残余非肿块性强化及流入型时间信号强度曲线提示MHR,以此作为诊断标准,评价动态增强MRI对残存病变检出及病理反应性术前评价的真实性和可靠性,用Kappa分析对观察者间进行一致性检验。结果:45例中23例为组织学显著反应,其中7例为病理学完全缓解,共有38例浸润癌残余。DCE—MRI对化疗后浸润癌残余的诊断有较高的敏感性,两医生敏感性分别为94．7％和97．4％,特异性分别为42．8％和57．1％,准确性分别为 86．6％和91．1％。DCE—MRI对化疗后组织学显著反应的判断,两医生诊断的敏感性分别为95．5％和81．8％,特异性分别为73．9％和82．6％,准确性分别为84．4％和82．2％。,两医生对残余病变检出和组织学显著反应判断的kappa值分别为0．728和0．778,说明观察者间有中高度一致性。对残存病变的低估见于组织学散在微灶残存肿瘤细胞变性明显,对残存病变的假阳性高估的组织学表现见于导管内癌残存和反应性炎症。结论:动态增强MRI病变的检出具有很高的敏感性,对病理反应性的术前诊断有潜在的应用价值。更多还原

【Abstract】 Objective: To investigate the value of dynamic contrast-enhanced MRI in detection of residual disease and assessment of pathological response in breast car- cinoma treated with neoadjuvant chemotherapy. Methods: Forty-five patients with pathologically confirmed breast carcinoma who finished last course of neoadjuvant chemotherapy had breast MRI examination prior to operation. All the specimen slices were reviewed by a pathologist without known the MRI results. Pathological response was assessed according Miller & Payne five points classification, of which grade 5 defined as pCR, grade 5 or 4 defined as MHR (Major histological response ). Dynamic contrast-enhanced MRI images were blindly reviewed by two radiologists retrospectively on workstation with Functool software. Any non-vassal enhancement in previous tumor bed in any phase of postcontrast acquisition was determined as residual disease, and absence of enhancement in tumor bed was regard as pathological complete response. Non mass residual enhancement and wash in time signal intensity curve indicate MHR. The diagnostic results of two radiologists were correlated to pathological gold standard. Validity and reliability of dynamic contrast enhanced MRI in residual disease detection and pathological response assessment were calculated. Interobserver agreement was measured by Kappa statistics. Results: Seven out of 45 cases were demonstrated pathological complete response, 23 cases were demonstrated MHR, and 38 cases were pathological invasive residual disease histologically. DCE-MRI for pathological invasive residual disease detection in two radiologists had sensitivities of 94.7% (36/38) and 97.4% (37/38), specificities of 42.8% and 57.1%, and accuracy of 86.6% and 91.1% respectively. For MHR evaluation, the sensitivities were 95.5% and 81.8%, specificities were 73.9% and 82.6%, and accuracy of 84.4% and 82.2% in two radiologists respectively. K values of two radiologists for detection and response assessment was 0.728 and 0.778 respectively, showing a good agreement between observers. Underestimation for detection occurred in a case of scattered microfocal residual disease with obvious degeneration of tumor cell. The false positive cases were seen in the cases with pure DCIS residua or responsive inflammation in the tumor bed. Conclusion: DCE-MRI has a high sensitivity for invasive residual disease detection, and had showed potential value in preoperatively assessment of pathological response in breast cancer after neoadjuvant chemotherapy.更多还原