[EBCC 2016]李晔雄教授:MRI在乳腺癌腋窝淋巴结转移中的诊断价值

作者:  李晔雄   日期:2016/3/11 17:53:09  浏览量:29141

肿瘤瞭望版权所有,谢绝任何形式转载,侵犯版权者必予法律追究。

第10届欧洲乳腺癌大会(EBCC)已经开始了第3天的日程,《肿瘤瞭望》撷取会议重点摘要,邀请中国医学科学院肿瘤医院放疗科李晔雄教授、唐玉医生给予点评,以加深读者对研究的理解。

  Z0011研究结果提示,对于接受保乳术的早期乳腺癌(cT1-2N0)患者,即使存在1-2个前哨淋巴结转移,不进行全腋窝淋巴结清扫也不会降低患者的生存率和局部控制率。在这一背景下,在术前筛选出存在腋窝淋巴结广泛转移(pN2-3)的病例,就变得尤为重要。随着MRI检查在乳腺癌检查中的广泛应用,利用MRI进行腋窝淋巴结分期成为了近年的一个研究热点。

 

  来自荷兰的这项单中心回顾性研究探讨了术前MRI在乳腺癌腋窝淋巴结转移中的诊断价值。研究纳入了200名于2009-2014年间在研究中心接受术前标准MRI检查的浸润性乳腺癌患者,所有患者均未接受过新辅助治疗或腋窝手术/放疗。由两名资深放射科医师根据术前MRI对腋窝淋巴结进行分期,并将其分期结果与最终的手术病理分期进行对比,来计算腋窝淋巴结为pN0和pN1时的假阴性率和阴性预测值(NPV)。研究发现,当MRI评价为N0时,实际腋窝淋巴结病理分期为N2-3的比例低于1.5%,相应NPV在两名医生中分别达到了98.8%(95.3%~99.8%)和98.7%(94.7%~99.8%)。而当MRI评价为N1时,实际病理分期为N2-3的比例增高到了10%以上,NPV下降至不足90%。

 

  该结果与近期发表在Eur Radiol的一项韩国研究的结果类似。在后一项纳入257名患者的研究中,对于未接受新辅助化疗的早期乳腺癌患者,当MRI未见腋窝淋巴结转移(cN0)时,仅有0.4%的患者实际腋窝淋巴结病理分期为N2-3,相应的NPV高达99.6%。从这两个研究的结果来看,如果术前MRI未发现腋窝转移淋巴结,则基本可以排除存在广泛腋窝淋巴结转移(pN2-3)的可能。

 

  既往有关MRI分期的系统性回顾显示,采用不同扫描序列时,MRI对腋窝淋巴结检测的敏感性为60%~88.4%,NPV在80%~95.9%之间。这一结果优于现有其他无创性检查手段,但如果想要依据MRI来修改治疗决策,现有的诊断准确性仍然不够。而此项荷兰研究的结果提示,对未接受新辅助化疗的临床N0乳腺癌患者,如果MRI检查未发现腋窝转移淋巴结,患者实际病理分期为pN0-1的概率高达98%以上。结合Z0011研究的结果,对于MRI检查未见腋窝淋巴结转移的这部分患者,可以考虑不做腋窝前哨淋巴结活检及清扫,从而进一步降低腋窝手术操作可能带来的副反应。

 

  此项研究的样本量偏小,为回顾性研究,仍需等待更大样本的前瞻性研究来确认其结果的可靠性。同时,MRI扫描的诊断价值与所选择的扫描序列、扫描范围及转移淋巴结的判断标准等密切相关,而摘要中基本没有描述上述信息,扫描技术会影响诊断的准确性。期待能够早日看到研究全文,对相关的技术细节获得更全面的了解。

 

  李晔雄,博士,教授,主任医师,博士研究生导师。现任中国医学科学院肿瘤医院放疗科主任。中华医学会放射肿瘤治疗学分会第七届主任委员,中华医学会肿瘤学分会委员,北京医学会肿瘤学分会副主任委员,临床肿瘤研究协会(CSCO)常务委员,国际淋巴瘤放射肿瘤研究组(ILROG)专家指导委员,国际结外淋巴瘤研究组(IELSG)委员,《中华放射肿瘤学杂志》主编,等职。获新世纪百千万人才工程国家级人选,卫计委突出贡献中青年专家,政府特殊津贴和吴阶平-保罗·杨森医学药学奖。

 

Diagnostic performance of standard breast MRI to exclude extensive nodal disease in breast cancer patients

Poster Spotlight: T. van Nijnatten (Netherlands)

Background: According to recent studies, limited axillary nodal disease (pN1) in breast cancer patients does not affect prognosis compared to node negative patients (pN0). As a result, excluding extensive axilary nodal disease (pN2–3) at initial diagnosis is becoming more and more important. Therefore, the purpose of this study was to evaluate the diagnostic performance of standard breast MRI in breast cancer patients to exclude pN2–3, in case pN0 or pN1 is predicted.

Material and Methods: All patients diagnosed with primary invasive breast cancer who underwent standard breast MRI prior to surgery in our hospital between 2009 and 2014 were included. Exclusion criteria were neoadjuvant systemic therapy and previous axillary surgery or radiotherapy. Two dedicated breast radiologists independently reassessed all breast MRIs and scored each axillary lymph node on a confidence level scale of 0 to 4. Results were compared to the gold standard of histopathology. Diagnostic performance was analysed by calculating false negative percentages and negative predictive values (NPV) for respectively pN0 and pN1. Quadratic weighted kappa measured interobserver agreement.

Results: A total of 200 patients were included. In case pN0 was predicted by the radiologists, pathology showed pN2–3 in 1.2% and 1.4%, with a NPV of 98.8% (95.3–99.8%) and 98.7% (94.7–99.8%) respectively. When pN1 was predicted by the radiologists, pathology showed pN2–3 in 13.0% and 10.2%, with a NPV of 87.0% (65.3–96.6%) and 89.8% (77.0–96.2%) respectively (Table 1). Interobserver agreement between both radiologists was considered good (weighted kappa = 0.624).

Table 1. Diagnostic performance of nodal staging on standard breast MRI 

 

Reader 1

Reader 2 

 

pN0

pN1

pN23

pN0

pN1

pN2

pN0 (n=153)

140

9

4

127

26

pN1 (n=38)

24

11

3

18

18

pN23 (n=9)

2

3

4

2

5

Conclusion: A negative preoperative breast MRI can exclude extensive nodal disease in breast cancer patients. Furthermore, breast MRI differentiates more accurately between limited and extensive axillary nodal disease, compared to current conventional imaging.

版面编辑:张楠  责任编辑:张彩琴

本内容仅供医学专业人士参考


MRI腋窝淋巴结转移

分享到: 更多