Lombardi Comprehensive Cancer Center, Georgetown University Medical Center, MedStar Health, Washington, DC.
NSABP Foundation, Pittsburgh, Pennsylvania.
Clin Cancer Res. 2020 Aug 15;26(16):4233-4241. doi: 10.1158/1078-0432.CCR-20-0152. Epub 2020 May 5.
In NSABP B-41, pathologic complete response (pCR) was associated with prolonged survival among women with HER2-positive operable breast cancer treated with neoadjuvant chemotherapy and lapatinib, trastuzumab, or the combination. We used a large human breast cancer gene expression panel to select candidate prognostic biomarkers for pCR among women treated with trastuzumab in NSABP B-41.
Eligible patients had a baseline preadjuvant treatment core biopsy sample, known pCR status, and no withdrawal of consent. We analyzed extracted RNA using the human nCounter Breast Cancer 360 gene expression panel. Gene counts were normalized to housekeeping genes and transformed into logarithmic scale with base 2. To screen for candidate genes and metagene signatures prognostic of pCR, we used univariate logistic regression. Variable selection was done by multivariable logistic regression with lasso regularization.
Analyses of data from 130 patients revealed that a composite of gene expression from 19 genes and one gene signature appeared to predict pCR in women with HER2-positive early-stage breast cancer undergoing neoadjuvant chemotherapy with trastuzumab-containing regimens. The identified genes are involved in important pathways such as epithelial-mesenchymal transition, adhesion and migration, estrogen receptor signaling, DNA damage and repair, apoptosis, and proliferation. The AUC from a 10-fold cross-validation on predicting pCR, with these 20 genomic markers in a logistic regression model, was 0.73.
The expression level of , and a few other genomic markers was highly predictive of pCR after trastuzumab-containing regimens. These findings need to be validated and calibrated in future studies.
在 NSABP B-41 试验中,接受新辅助化疗和拉帕替尼、曲妥珠单抗或联合治疗的 HER2 阳性可手术乳腺癌患者中,病理完全缓解(pCR)与生存延长相关。我们使用大型人类乳腺癌基因表达谱来选择 NSABP B-41 中接受曲妥珠单抗治疗的患者中 pCR 的候选预后生物标志物。
合格患者具有基线新辅助治疗核心活检样本、已知 pCR 状态且未撤回同意书。我们使用人类 nCounter Breast Cancer 360 基因表达谱分析提取的 RNA。基因计数通过管家基因进行标准化,并以 2 为底转换为对数尺度。为了筛选与 pCR 相关的候选基因和基因表达谱,我们使用单变量逻辑回归。通过具有拉索正则化的多变量逻辑回归进行变量选择。
对 130 例患者数据的分析显示,来自 19 个基因和一个基因特征的基因表达组合似乎可以预测接受曲妥珠单抗联合新辅助化疗的 HER2 阳性早期乳腺癌患者的 pCR。鉴定的基因涉及重要途径,如上皮-间充质转化、黏附和迁移、雌激素受体信号转导、DNA 损伤和修复、细胞凋亡和增殖。在使用这些 20 个基因组标记的逻辑回归模型中,对 pCR 进行 10 倍交叉验证的 AUC 为 0.73。
这些基因的表达水平和其他一些基因组标记高度预测了曲妥珠单抗联合治疗方案后的 pCR。这些发现需要在未来的研究中进行验证和校准。