Genomic and Morphologic Predictor of High-Risk DCIS

 描述(申请人提供)：最近的大规模基因组研究已经确定并证实了浸润性乳腺癌(IBC)中许多反复出现的单核苷酸变异和非整倍体。与IBC相比，从正常组织到早期肿瘤再到导管原位癌(DCIS)到IBC，与乳腺癌进展相关的基因组变化知之甚少。在筛查乳腺活检中发现的DCIS的临床评估完全依赖于几十年前开发的形态标准和光学显微镜。这些形态标准在识别与IBC相关或注定进展为IBC的DCIS病变方面表现不佳。在美国新诊断的乳腺癌病例中，20%将是DCIS。随着筛查乳房X光检查的出现，无症状女性中DCIS的诊断率显著上升，而乳腺癌死亡率却没有预期的下降，这导致了人们对过度诊断和过度治疗的担忧。临床试验正在涌现，以分析主动监测作为筛查出低级别DCIS患者的临床策略。许多筛查出DCI的患者现在陷入了进退两难的境地，他们想知道如果他们的病变在最初检测时没有得到治疗，他们进展到IBC的风险是什么。由于IBC代表重复的、常见的基因变化的积累，我们假设这些突变的识别、它们在DCIS中的积累程度以及相关的核变化将有助于我们识别极有可能进展为IBC的病例。这将使我们能够对这些病变进行分层，以确定发生IBC的风险。我们已经用三项独立研究的初步数据证明了这种方法的可行性：一项涉及进展为IBC的肿瘤的全基因组测序，第二项涉及DCIS队列中的靶向DNA拷贝数测量，第三项检查DCIS和增生性疾病之间的核形态计量学差异。我们提出了一个病例对照研究设计，使用三个不同的纵向队列。我们将使用病例对照研究设计对两个大型独立队列进行有针对性的基因组和核表型特征分析，包括来自护士健康研究(NHS)、华盛顿大学(WASHU)的队列和来自斯坦福大学(SU)的较小队列，以创建发现、培训和交叉验证以及测试集。在我们的发现队列中，我们将使用整个外显子组测序、FISH和核形态计量学分析来识别发生IBC的DCIS与不发生IBC的DCIS病例的基因组和表型变化。从这些研究中确定的生物标记物将被用于构建DCIS乳腺癌风险的基因组预测因子。预测模型将使用护士健康研究中病例和对照的DCIS样本来建立。NHS中产生的未来IBC风险的基因组预测因子将在华盛顿大学进行长期临床随访的独立大DCIS样本队列中得到验证。通过提高我们将DCIS患者分成分子定义的高风险和低风险患者亚组的能力，对DCIS风险分层的了解将有助于降低IBC的发生率和死亡率。高危患者可能受益于更积极的临床治疗，如完全手术切除、化学预防和/或通过乳腺MRI等技术进行强化监测，而低风险患者可能不需要或受益于这些措施。

 DESCRIPTION (provided by applicant): Recent large scale genomic studies have identified and confirmed numerous recurrent single nucleotide variations and aneuploidies in invasive breast cancer (IBC). In contrast to IBC, little is understood about the genomic changes associated with progression to breast cancer, from normal tissue to early neoplasias to ductal carcinoma in situ (DCIS) to IBC. The clinical evaluation of DCIS found in screening breast biopsies relies solely on morphologic criteria and the light microscope that were developed decades ago. These morphologic criteria do not perform well in identifying DCIS lesions that are associated with or are destined to progress to IBC. Among newly diagnosed breast cancer cases in the United States, 20% will be DCIS. With the advent of screening mammography, there has been a remarkable rise in the diagnosis of DCIS among asymptomatic women without the expected reduction in breast cancer mortality, leading to concerns about both over- diagnosis and over-treatment. Clinical trials are emerging to analyze active surveillance as a clinical strategy for patients screen detected with low grade DCIS. Many patients with screen detected DCIS are now left in a quandary, wondering what is their risk of progression to IBC if their lesion were left untreated at initial detection. Because IBC represents the accumulation of recurrent, common genetic changes, we hypothesize that the identification of these mutations, the degree of their accumulation in DCIS, and associated nuclear changes will help us identify cases that have a high likelihood of progressing to IBC. This would allow us to stratify these lesions for risk of developing IBC. We have demonstrated the feasibility of this approach with preliminary data from three independent studies: one involving whole genome sequencing of neoplasms in the progression to IBC, a second involving targeted DNA copy number measurements in a cohort of DCIS, and a third examining the nuclear morphometric differences between DCIS and hyperplasias. We propose a case-control study design using three distinct longitudinal cohorts. We will perform targeted analysis of genomic and nuclear phenotypic features with a case-control study design on two large independent cohorts, including cohorts from the Nurses' Health Study (NHS), Washington University (WashU), and a smaller cohort from Stanford University (SU) to create Discovery, Training and Cross-validation, and Test sets. In our discovery cohort, we will use whole exome sequencing, FISH, and nuclear morphometric analysis to identify genomic and phenotypic changes in DCIS that develop IBC versus cases of DCIS that do not develop IBC. Biomarkers identified from these studies will be used to construct a genomic predictor of breast cancer risk in DCIS. The predictive model will be built using DCIS samples from cases and controls in the Nurses' Health Study. The genomic predictor of future IBC risk generated in the NHS will then be validated on an independent large cohort of DCIS samples with long-term clinical follow-up from Washington University. Improved knowledge of risk stratification for DCIS will help reduce IBC incidence and mortality by improving our ability o stratify patients with DCIS into molecularly defined subgroups of high- and low-risk patients. The high-risk patients may benefit from more aggressive clinical treatment, like complete surgical excision, chemoprophylaxis, and/or intensive surveillance with techniques such as breast MRI, while the low-risk patients may not require or benefit from these measures.