Bioinformatics Centre, School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu, China.
Breast Cancer Res Treat. 2013 Dec;142(3):505-14. doi: 10.1007/s10549-013-2767-8. Epub 2013 Nov 20.
Current predictors for estrogen receptor-positive (ER-positive) breast cancer patients receiving tamoxifen are often invalid in inter-laboratory validation. We aim to develop a robust predictor based on the relative ordering of expression measurement (ROE) in gene pairs. Using a large integrated dataset of 420 normal controls and 1,129 ER-positive breast tumor samples, we identified the gene pairs with stable ROEs in normal control and significantly reversed ROEs in ER-positive tumor. Using these gene pairs, we characterized each sample of a cohort of 292 ER-positive patients who received tamoxifen monotherapy for 5 years and then identified relapse risk-associated gene pairs. We extracted a gene pair subset that resulted in the largest positive and negative predictive values for predicting 10-year relapse-free survival (RFS) using a genetic algorithm. A predictor was developed based on the gene pair subset and was validated in 2 large multi-laboratory cohorts (N = 250 and 248, respectively) of ER-positive patients who received 5-year tamoxifen alone. In the first validation cohort, the patients predicted to be tamoxifen sensitive had a 10-year RFS of 91 % (95 % confidence interval [CI] 85-97 %) with an absolute risk reduction of 34 % (95 % CI 17-51 %). The patients predicted to be tamoxifen insensitive had a significantly higher relapse risk than the patients predicted to be tamoxifen sensitive (hazard ratio = 4.99, 95 % CI 2.45-10.17, P = 9.13 × 10(-7)). Similar performance was achieved for the second validation cohort. The predictor performed well in both node-negative and node-positive subsets and added significant predictive power to the clinical parameters. In contrast, 2 previously proposed predictors did not achieve significantly better performances than the baselines of the validation cohorts. In summary, the proposed predictor can accurately and robustly predict tamoxifen sensitivity of ER-positive breast cancer patients and identified patients with a high probability of 10-year RFS following tamoxifen monotherapy.
目前用于预测接受他莫昔芬治疗的雌激素受体阳性(ER 阳性)乳腺癌患者的预测因子在实验室间验证中往往并不有效。我们旨在开发一种基于基因对表达测量相对排序(ROE)的稳健预测因子。使用一个包含 420 名正常对照和 1129 名 ER 阳性乳腺癌肿瘤样本的大型综合数据集,我们鉴定了在正常对照中具有稳定 ROE 且在 ER 阳性肿瘤中具有显著逆转 ROE 的基因对。使用这些基因对,我们对接受他莫昔芬单药治疗 5 年的 292 名 ER 阳性患者队列中的每个样本进行了特征分析,然后鉴定了与复发风险相关的基因对。我们使用遗传算法提取了一个基因对亚组,该亚组对预测 10 年无复发生存率(RFS)的阳性和阴性预测值最大。基于该基因对亚组开发了一个预测因子,并在接受 5 年他莫昔芬单药治疗的 2 个大型多实验室队列(分别为 N = 250 和 N = 248)中进行了验证。在第一个验证队列中,预测为对他莫昔芬敏感的患者 10 年 RFS 为 91%(95%CI 85-97%),绝对风险降低 34%(95%CI 17-51%)。预测为对他莫昔芬不敏感的患者比预测为对他莫昔芬敏感的患者具有更高的复发风险(危险比=4.99,95%CI 2.45-10.17,P=9.13×10(-7))。第二个验证队列的结果相似。该预测因子在淋巴结阴性和淋巴结阳性亚组中表现良好,并为临床参数提供了显著的预测能力。相比之下,另外两个先前提出的预测因子在验证队列的基线水平上并没有显著更好的表现。总之,该预测因子可以准确、稳健地预测 ER 阳性乳腺癌患者对他莫昔芬的敏感性,并确定了接受他莫昔芬单药治疗后 10 年 RFS 概率较高的患者。