Identification of a robust gene signature that predicts breast cancer outcome in independent data sets [UCSF]

Background: Numerous studies have focused on the PI3K/AKT/mTOR pathway in estrogen receptor positive (ER) breast cancer (BC), as a linear signal transduction pathway and reported its association with worse clinical outcomes. To gain better insight into the relative contribution of each of the signalling pathways which lie downstream to PI3K (namely AKT and mTOR) to BC outcomes, we have developed a novel in silico approach which assessed the activation of each of these signalling pathways separately. Methods: By analysing gene expression and matched proteomic data in ER-positive patients from the TCGA repository, we developed gene signatures that reflect the level of expression of phosphorylated-Serine473-AKT (pAKT) and phosphorylated-Serine2448-mTOR (p-mTOR) separately, capturing their corresponding level of pathway activation. Using pooled analysis of gene-expression data from over 7,000 patients with ER-positive early BC, we then investigated the association between pathway activation and outcomes. Results: Our analysis revealed that the pAKT pathway activation (as measured by the developed signature) was associated with luminal A BC while the p-mTOR pathway activation was more associated with luminal B BC (Kruskal-Wallis test p<10-10). pAKT pathway activation was significantly associated with better outcomes (multivariable HR, 0.79; 95% CI, 0.74 to 0.85; p=2.5x10-10) whereas p-mTOR pathway activation showed worse outcomes (multivariable HR, 1.1; 95% CI, 1.1 to 1.2; p=9.9x10-4). Similar associations between activation and outcomes were obtained when the analysis was performed in patients who were treated with hormonal therapy. Additionally, pAKT pathway activation predicted better outcomes irrespective of the BC molecular subtypes (luminal A multivariable HR, 0.85; 95% CI, 0.75 to 0.96; p=0.01; luminal B HR, 0.91; 95% CI, 0.83 to 0.99; p=0.033). pAKT pathway activation was associated with PIK3CA mutations (p=0.0001) while p-mTOR pathway activation was associated with p53 mutations (p=0.04). Finally, we show that pAKT pathway activation was associated with less response to Everolimus. Conclusions: Our data show that pAKT and p-mTOR pathway activation have differing impact on prognosis and suggest that they are not linearly connected in luminal breast cancers. These findings add to our understanding of signalling through the PI3K pathway and could have important implications for the treatment of luminal BC. Data are from Korkola et al., BMC Cancer 2007 (PMID 17428335) and are no longer available on the web site. Version published here is as formatted for Haibe-Kains et al., JNCI 2007 (PMID 22262870), reused in Sonnenblick et al., NPJ Breast, in press. Data were downloaded from the authors web site by Haibe-Kains before 2007. Fresh frozen breast tumors were collected from patients treated at UCSF and California Pacific Medical Centers after institutional review board approval. Randomly selected frozen tumor blocks were trimmed to ensure that a minimum of 70% of the remaining cells present were tumor. A total of 140 invasive ductal tumors, 17 invasive lobular tumors, 4 DCIS, 1 inflammatory, and 8 normal breast samples from reduction mammoplasties were analyzed. There were 4 DCIS, 35 Stage I, 80 Stage II, 16 Stage III, 6 Stage IV tumors, and 21 with unknown stage. ER negative tumors comprised 29% of the samples, and 45% of the samples were node negative.

研究背景：既往多项研究已将PI3K/AKT/mTOR通路作为一条线性信号转导通路，聚焦于雌激素受体阳性（estrogen receptor positive, ER）乳腺癌（breast cancer, BC），并报道了该通路与不良临床结局的关联。为更深入解析PI3K下游各信号通路（即AKT与mTOR）各自对乳腺癌结局的相对贡献，本研究开发了一种全新的虚拟（in silico）分析方法，可分别评估上述每条信号通路的激活状态。
研究方法：通过分析TCGA数据库中雌激素受体阳性患者的基因表达数据与匹配的蛋白质组学数据，本研究分别构建了可反映磷酸化丝氨酸473位点AKT（pAKT）与磷酸化丝氨酸2448位点mTOR（p-mTOR）表达水平的基因特征标签，以此捕获两条通路各自的激活水平。随后，针对7000余例雌激素受体阳性早期乳腺癌患者的基因表达数据开展合并分析，探究了通路激活状态与临床结局之间的关联。
研究结果：本研究分析显示，通过所构建基因特征标签检测的pAKT通路激活与腔型A乳腺癌相关，而p-mTOR通路激活则更多见于腔型B乳腺癌（克鲁斯卡尔-沃利斯检验，p<10^-10）。pAKT通路激活与更佳的临床结局显著相关（多变量风险比（HR）=0.79；95%置信区间（CI）：0.74~0.85；p=2.5×10^-10），而p-mTOR通路激活则与不良结局相关（多变量HR=1.1；95%CI：1.1~1.2；p=9.9×10^-4）。在接受内分泌治疗的患者亚组中分析也得到了相似的通路激活与临床结局关联结果。此外，无论乳腺癌分子亚型如何，pAKT通路激活均可预测更佳的临床结局（腔型A：多变量HR=0.85；95%CI：0.75~0.96；p=0.01；腔型B：HR=0.91；95%CI：0.83~0.99；p=0.033）。pAKT通路激活与PIK3CA突变相关（p=0.0001），而p-mTOR通路激活则与p53突变相关（p=0.04）。最后，本研究发现pAKT通路激活与依维莫司（Everolimus）应答率降低相关。
研究结论：本研究数据表明，pAKT与p-mTOR通路激活对患者预后的影响存在差异，提示在腔型乳腺癌中两条通路并非呈线性关联。上述发现加深了我们对PI3K通路信号转导的理解，或对腔型乳腺癌的临床治疗具有重要指导意义。本数据集源自Korkola等人2007年发表于《BMC癌症》的研究（PubMed ID: 17428335），目前该数据集已无法从原网站获取。本次发布的版本按照Haibe-Kains等人2007年发表于《美国国家癌症研究所杂志》（JNCI）的格式进行整理（PubMed ID: 22262870），该数据集也曾被Sonnenblick等人复用至《NPJ乳腺癌》（NPJ Breast）的待发表研究中。本数据集由Haibe-Kains于2007年前从原作者网站下载。本研究中的新鲜冷冻乳腺肿瘤样本采集自加州大学旧金山分校（UCSF）与加利福尼亚太平洋医疗中心接受治疗的患者，所有样本采集均已获得伦理审查委员会批准。研究人员对随机选取的冷冻肿瘤组织块进行了修剪，以确保剩余组织中肿瘤细胞占比不低于70%。本次分析共纳入140例浸润性导管癌、17例浸润性小叶癌、4例导管原位癌（DCIS）、1例炎性乳腺癌样本，以及8例来自乳房缩小整形术的正常乳腺组织样本。其中，肿瘤样本的分期情况为：35例Ⅰ期、80例Ⅱ期、16例Ⅲ期、6例Ⅳ期乳腺癌，另有21例样本的肿瘤分期未知；此外还包含4例导管原位癌。雌激素受体阴性肿瘤占总样本的29%，45%的样本为淋巴结阴性。