Endogenous hormones and postmenopausal breast cancer: Etiologic insights and improving risk prediction

我们建议继续我们研究小组的工作，以确定和验证预测风险的激素标志物 绝经后妇女浸润性乳腺癌的风险。尽管雌激素途径在 乳腺癌的发生，越来越多的证据表明只测量经典的雌激素，如在 以前的流行病学研究(包括我们自己的研究)可能会遗漏雌激素活性的一个相当大的组成部分，即 病因学上与风险相关的。此外，胰岛素途径是另一个关键，但相对较少的研究， 可能影响乳腺癌风险的荷尔蒙途径。采用前瞻性嵌套病例对照设计，我们 计划分析从护士健康研究(NHS)的32,826名参与者那里采集的血液样本 提供了-90的血液样本，并在1999-2000年为其中18,743名妇女提供了第二次样本。 具体地说，我们建议评估27-羟基胆固醇(27HC；最近的一个 已确定的内源性选择性雌激素受体调节剂)以及雌激素生物测定(评估 雌激素途径激活)与乳腺癌风险的关系。我们还将评估几个新的方面 血浆C-肽(反映胰岛素分泌)与乳腺癌风险的关系。用26年的时间 在随访和两次采血的一组女性中，我们将评估C-肽时机的重要性 在乳腺癌发生中的暴露。此外，我们将评估c-肽与乳腺癌的关系 不同乳腺肿瘤胰岛素受体的表达或PI3K/Akt/mTOR通路激活的几个标志物不同。 以前的工作，由于没有考虑到这些肿瘤的分子特征，可能低估了 胰岛素在乳腺癌风险和预防机会中的病因学重要性。我们也会 评价C-肽(以及新的雌激素相关标记物)是否能改善Gail和Rosner-Colditz 风险预测模型，并验证多个生物标志物的添加(即血浆激素、基因 风险分数和乳房X光密度)对这些模型进行前瞻性的Mayo乳房X光检查健康研究 (MMHS)和墨尔本协作队列研究(MCCS)。MMHS和MCCS有风险因素和 生物标志物信息，可与NHS中的信息相媲美。除了提供一两个血样外， NHS队列成员每两年提供详细的暴露和疾病信息(包括 最已知或最可能的乳腺癌危险因素)，自1976年以来，对于大约70%的乳腺癌病例，我们 收集了福尔马林固定的石蜡包埋的肿瘤组织，从而可以对肿瘤进行分子分析。 最近资助的NHS UM1赠款将提供队列随访、乳腺癌文献和重要的 协变量数据，增加了拟议项目的成本效益。这个项目，通过在 从新的激素生物标志物的发现到油井的实际应用 在风险预测模型中确认的生物标记物，应该会增加对乳腺癌病因和 如何更好地确定可能受益于降低风险努力(例如，化学预防)的高危妇女。

We propose to continue our research group’s work to identify and validate hormonal markers that predict risk of invasive breast cancer in postmenopausal women. Although the estrogen pathway plays a critical role in breast carcinogenesis, accruing evidence suggests measurement of the classic estrogens only, as done in prior epidemiologic studies (including our own) may miss a sizeable component of estrogenic activity that is etiologically relevant to risk. Further, the insulin pathway is another critical, but relatively understudied, hormonal pathway that may impact breast cancer risk. Using a prospective nested case-control design, we plan to analyze blood samples collected from the 32,826 participants in the Nurses' Health Study (NHS) who provided a blood sample in 1989-90 and, for 18,743 of these women, a second sample in 1999-2000. Specifically, we propose to evaluate the independent associations of 27-hydroxycholesterol (27HC; a recently identified endogenous selective estrogen receptor modulator) as well as an estrogen bioassay (that assesses estrogen pathway activation) in relation to breast cancer risk. We also will evaluate several new aspects of the relationship between plasma c-peptide (reflecting insulin secretion) and breast cancer risk. With 26 years of follow-up and two blood samples in a subset of women, we will evaluate the importance of timing of c-peptide exposure in breast carcinogenesis. Further, we will assess whether the c-peptide/breast cancer relationship varies by breast tumor expression of insulin receptor or several markers of PI3k/Akt/mTOR pathway activation. Previous work, by not accounting for these tumor molecular characteristics, may have underestimated the etiologic importance of insulin in breast cancer risk and hence opportunities for prevention. We also will evaluate if c-peptide (and secondarily the novel estrogen related markers) improve the Gail and Rosner-Colditz risk prediction models, and validate the addition of multiple biomarkers (i.e., the plasma hormones, a genetic risk score and mammographic density) to these models in the prospective Mayo Mammography Health Study (MMHS) and Melbourne Collaborative Cohort Study (MCCS). The MMHS and MCCS have risk factor and biomarker information that is comparable to that in the NHS. In addition to providing one or two blood samples, NHS cohort members have provided detailed exposure and disease information biennially (including data on most known or probable breast cancer risk factors) since 1976, and, for about 70% of breast cancer cases, we have collected formalin-fixed paraffin embedded tumor tissue, thus allowing molecular analyses of the tumor. The recently funded NHS UM1 grant will provide cohort follow-up, breast cancer documentation, and important covariate data, increasing the cost-effectiveness of the proposed project. This project, by working across the translational research spectrum from discovery of novel hormonal biomarkers to practical application of well confirmed biomarkers in risk prediction models, should add considerable insight into breast cancer etiology and ways to better identify high risk women who may benefit from risk reduction efforts (e.g., chemoprevention).