Department of Breast and Endocrine Surgery, Okayama University Hospital, Okayama, Japan.
Department of Breast Oncology, Miyake Ofuku Clinic, Okayama, Japan.
Chin Clin Oncol. 2020 Jun;9(3):27. doi: 10.21037/cco.2020.01.06. Epub 2020 Mar 13.
The improvement of tumor biomarkers prepared for clinical use is a long process. A good biomarker should predict not only prognosis but also the response to therapies. In this review, we describe the biomarkers of neoadjuvant/adjuvant chemotherapy for breast cancer, considering different breast cancer subtypes. In hormone receptor (HR)-positive/human epidermal growth factor 2 (HER2)-negative breast cancers, various genomic markers highly associated with proliferation have been tested. Among them, only two genomic signatures, the 21-gene recurrence score and 70-gene signature, have been reported in prospective randomized clinical trials and met the primary endpoint. However, these genomic markers did not suffice in HER2-positive and triple-negative (TN) breast cancers, which present only classical clinical and pathological information (tumor size, nodal or distant metastatic status) for decision making in the adjuvant setting in daily clinic. Recently, patients with residual invasive cancer after neoadjuvant chemotherapy are at a high-risk of recurrence for metastasis, which, in turn, make these patients best applicants for clinical trials. Two clinical trials have shown improved outcomes with post-operative capecitabine and ado-trastuzumab emtansine treatment in patients with either TN or HER2-positive breast cancer, respectively, who had residual disease after neoadjuvant chemotherapy. Furthermore, tumor-infiltrating lymphocytes (TILs) have been reported to have a predictive value for prognosis and response to chemotherapy from the retrospective analyses. So far, TILs have to not be used to either withhold or prescribe chemotherapy based on the absence of standardized evaluation guidelines and confirmed information. To overcome the low reproducibility of evaluations of TILs, gene signatures or digital image analysis and machine learning algorithms with artificial intelligence may be useful for standardization of assessment for TILs in the future.
用于临床的肿瘤标志物的改进是一个漫长的过程。一个好的标志物不仅应该预测预后，还应该预测对治疗的反应。在这篇综述中，我们描述了新辅助/辅助化疗的乳腺癌标志物，考虑了不同的乳腺癌亚型。在激素受体（HR）阳性/人表皮生长因子 2（HER2）阴性的乳腺癌中，已经测试了各种与增殖高度相关的基因组标志物。其中，只有两个基因组特征，即 21 基因复发评分和 70 基因特征，在前瞻性随机临床试验中得到了报道，并达到了主要终点。然而，这些基因组标志物在 HER2 阳性和三阴性（TN）乳腺癌中并不足够，这些乳腺癌仅提供经典的临床和病理信息（肿瘤大小、淋巴结或远处转移状态），以便在辅助治疗中做出决策日常诊所。最近，新辅助化疗后仍有浸润性癌残留的患者发生转移的风险很高，这反过来使这些患者成为临床试验的最佳申请者。两项临床试验表明，在新辅助化疗后仍有疾病残留的 TN 或 HER2 阳性乳腺癌患者中，术后卡培他滨和 ado-trastuzumab emtansine 治疗可改善结局。此外，从回顾性分析中报告了肿瘤浸润淋巴细胞（TILs）对预后和化疗反应具有预测价值。到目前为止，TILs 还没有被用于根据缺乏标准化评估指南和确认信息来保留或规定化疗。为了克服 TILs 评估的低重现性，基因特征或数字图像分析和人工智能的机器学习算法可能有助于未来 TILs 评估的标准化。
