A 13-gene signature predicting rapid development of brain metastases in breast cancer. Homo sapiens

Brain metastases from breast and other cancers constitute an important part of therapeutic failures and are associated with severe morbidity and mortality. Here, we have examined histopathological data and generated gene expression data in two independent cohorts of primary tumors from HER2-positive advanced breast cancer patients. We report that the combination of estrogen receptor (ER) negativity and expression of a novel 13-gene signature identify a subset of patients with rapid (median, 31 and 41 months in discovery and validation cohorts, respectively) versus slower (median, 66 months and 77 months in discovery and validation cohorts, respectively) development of brain metastases (P<0.0001). The 13-gene signature also predicted rapid brain metastasis formation within the ER-negative subset of patients (P=0.014). Interestingly, three of the genes in the signature (RAD51, BARD1, FANCG) function in DNA double strand break repair. Overexpression of RAD51 in immortal MCF-10A breast epithelial cells altered their three-dimensional acinar morphology to increase the percentage of invasive structures by 6.5 fold, in the presence or absence of HER2 overexpression. In summary, ER negativity and a novel 13-gene signature may have the potential to identify subpopulations at highest immediate risk for the development of brain metastases in HER2-positive advanced breast cancer. Our results also suggest that RAD51, found in the 13-gene signature, may promote aggressiveness in breast epithelial cells. These data may be useful in the design of brain metastasis preventive trials and may prompt new treatment strategies Median normalized data provided Overall design: Using Illumina DASL Cancer gene panel we analyzed gene expression of 502 oncogenes in 87 breast cancer patients. The data was divided into a training (n=62) and an internal validation set (n=25). Clincal parameters were noted for each patient. The data was analyzed in order to determine whether the 13-gene signature is associated with the brain metastasis-free survival (BMFS) when it was simply classified as being lower or higher than the median.

乳腺癌及其他癌症引发的脑转移是治疗失败的重要诱因之一，且与极高的发病率和死亡率紧密相关。本研究针对两批独立的人类表皮生长因子受体2（HER2）阳性晚期乳腺癌患者的原发肿瘤样本，开展了组织病理学数据检测与基因表达谱分析。本研究发现，雌激素受体（ER）阴性状态与全新13基因特征的表达联合，可有效区分两类患者亚群：一类脑转移发生较快（发现队列与验证队列的中位时间分别为31个月与41个月），另一类则进展较慢（对应中位时间分别为66个月与77个月），该差异具有显著统计学意义（P<0.0001）。该13基因特征还可在ER阴性患者亚群中有效预测脑转移的快速发生（P=0.014）。值得注意的是，该基因特征中的3个基因（RAD51、BARD1、FANCG）均参与DNA双链断裂修复通路。在永生化MCF-10A乳腺上皮细胞中过表达RAD51，无论是否伴随HER2过表达，均可改变其三维腺泡形态，使侵袭性结构的占比提升6.5倍。综上，雌激素受体阴性状态与全新13基因特征，有望在HER2阳性晚期乳腺癌患者中筛选出脑转移发生风险最高的亚群。本研究结果同时提示，13基因特征中的RAD51可能促进乳腺上皮细胞的侵袭性表型。本数据集可为脑转移预防临床试验的设计提供重要参考，并有望催生全新的治疗策略。本研究提供了经中位数归一化处理的基因表达数据。整体实验设计如下：采用Illumina DASL癌症基因面板（Illumina DASL Cancer Gene Panel），本研究对87名乳腺癌患者体内502个癌基因的表达水平进行了检测。将所有样本划分为训练集（n=62）与内部验证集（n=25）两组，并记录了每一位患者的临床参数。本研究通过将13基因特征的表达量按照中位数划分为高低两组，分析其与脑转移无进展生存期（BMFS）的相关性。