Project 1:The Carolina Breast Cancer Study

项目1摘要 黑人女性预后较差的乳腺癌亚型发生率较高，分期较差 死亡率。我们之前基于卡罗莱纳乳腺癌研究(CBCS)的孢子发现整合了 基于人群的流行病学和肿瘤生物学证明年轻的黑人女性更有可能 发展为基底样肿瘤。最近，我们发现黑人女性也有更高频率的Lumina B 和富含HER2的乳腺肿瘤。此外，即使是在雌激素的临床亚组中预后最好的 受体(ER)阳性/HER2阴性的癌症中，黑人女性复发风险(ROR)得分较高。 这些发现表明，肿瘤生物学是导致种族死亡率差异的关键因素。一些肿瘤因子 (即更高频率的HER2富集亚型)可能是有针对性的，说明了基于人口的 基因组学可以推动精准医学的发展。在这个项目中，我们建议加深对种族问题的理解 肿瘤生物学、病因和进展的差异，重点是肿瘤突变特征和 免疫反应。特定突变(即TP53、PIK3CA)的频率和总的模式 体细胞突变(突变特征)可以提供有关潜在生物过程的信息， 从癌症发生的最早阶段开始就出了问题。此外，新的发现强调了 免疫系统在乳腺癌病因和进展中的重要性。凭借我们丰富的肿瘤资源 生物标本(例外96%的肿瘤取自3000例)，我们将对肿瘤DNA和 识别黑人和白人女性的突变特征(目标1)，我们将确定肿瘤免疫的特征 使用6个免疫组织化学标记物和50个基因的RNA图谱分析微环境(目标2)。而当 癌症基因组图谱(TCGA)已经用数十万数据评估了肿瘤的种族差异 每名患者积分，限制包括少数黑人参与者(CBCs比 (TCGA)，这是一种不以总体为基础的抽样方案，缺乏详细的协变量和后续数据。 TCGA的进步使我们能够提高基于人口的工作的测序效率，并已 帮助开发了用于免疫分析的高效RNA小组，从而制定了具有成本效益的收集计划 2000名女性(1000名黑人，1000名白人，来自 全部CBCs)。在CBCS的背景下利用这些进展，我们假设突变 签名和免疫反应与种族、年龄和乳腺癌亚型有不同的关联。我们 还希望识别一些突变特征，这些特征会随着免疫特征的不同而变化。肿瘤 在AIMS 1和AIMS 2中收集的生物变量将被评估为乳腺癌复发和 生死存亡。此外，肿瘤生物学的影响可能会被其他患者层面的因素所混淆，例如 社会经济地位、获得护理的机会、治疗延迟和治疗依从性。多变量建模(AIM 3)将确定有助于生存的肿瘤生物学和患者水平特征的相对优势，从而 从而实现有效的精准医疗策略。 1

Project 1 Abstract Black women suffer higher incidence of poor-prognosis breast cancer subtypes and worse stage-specific mortality. Our previous Carolina Breast Cancer Study(CBCS)-based SPORE discoveries have integrated population-based epidemiology with tumor biology to demonstrate that young black women are more likely to develop Basal-like tumors. More recently, we found that black women also have higher frequency of Luminal B and HER2-enriched breast tumors. Furthermore, even among the best-prognosis clinical subset of estrogen receptor (ER)-positive/HER2-negative cancers, black women have higher risk of recurrence (ROR) scores. These findings implicate tumor biology as a key contributor to racial mortality disparities. Some tumor factors (i.e., higher frequency of HER2-enriched subtype) may be targetable, illustrating how population-based genomics can advance precision medicine. In this Project, we propose to deepen our understanding of racial differences in tumor biology, etiology and progression, with a focus on tumor mutational signatures and immune responses. The frequency of specific mutations (i.e. TP53, PIK3CA), and the overall patterns of somatic mutations (mutational signatures) can be informative about the underlying biological processes that have gone awry from the earliest stages of carcinogenesis. In addition, new discoveries have emphasized the importance of the immune system in breast cancer etiology and progression. With our rich resource of tumor biospecimens (exceptional 96% tumor procurement from 3000 cases), we will sequence tumor DNA and identify mutational signatures in black and white women (Aim 1) and we will characterize tumor immune microenvironments using six immunohistochemical markers and RNA profiling for 50 genes (Aim 2). While The Cancer Genome Atlas (TCGA) has evaluated racial differences in tumors with hundreds of thousands of data points per patient, limitations included small numbers of black participants (CBCS is ten-fold larger than TCGA), a sampling scheme that was not population-based, and lack of detailed covariate and follow-up data. TCGA advances have allowed us to increase sequencing efficiency for population-based work and have helped in developing an efficient RNA panel for immune profiling, resulting in a plan for cost-effective collection of mutational signatures and expression profiles in 2000 women (1000 black, 1000 white, sampling from the entirety of CBCS). Utilizing these advances in the context of the CBCS, we hypothesize that mutational signatures and immune responses are differentially associated with race, age, and breast cancer subtype. We also expect to identify some mutational signatures that vary as a function of immune profiles. The tumor biological variables collected in Aims 1 and 2 will be assessed as predictors of breast cancer recurrence and survival. In addition, impact of tumor biology may be confounded by other patient-level factors such as socioeconomic status, access to care, treatment delay and treatment adherence. Multivariable modeling (Aim 3) will identify the relative strengths of tumor biology and patient-level features that contribute to survival, thus allowing for impactful precision medicine strategies. 1