Methods for Statistical Analysis of Cancer Genetic Network Studies

项目摘要/摘要 美国国家癌症研究所(NCI)建立了癌症遗传学网络(CGN) 1998年作为一个多中心的国家项目，旨在支持 肿瘤易感性的遗传基础研究及机制探讨 将这一新知识融入到医疗实践中。这其中的一个根本问题 高危人群的问题是如何进行最佳筛查，以便及早发现疾病。近期 随着疾病早期阶段的变化而变化的新生物标志物的发现提供了希望 非侵入性的筛查策略。然而，有两个方面是 确定最优策略是：对于什么癌症，风险较高的个体 生物标记物上的纵向测量如何用来预测 疾病。在癌症遗传学网络的背景下关注这些问题 项目，这项提议有两个目的。第一个目标是开发和应用 分析家族史数据以发现癌症倾向的方法 聚集在个人和家庭中。我们将开发出一种可以 适用于CGN注册处(家族史)数据，并适当说明年龄 以及参与者的确定方式。我们将应用这些方法来识别 家族中一起发生的癌症部位的聚集性。这样做的第二个目的是 提案的重点是开发和应用一种方法来评估 纵向采集的生物标志物与早期证据的关系 疾病发作。该方法将使我们能够分析来自CGN卵巢的数据 癌症(CA125)生物标志物筛查研究，因为它将适当地处理这一问题 纵向和事件时间数据，即使这些妇女的筛查时间表 由于错过或延误访问而变化。此提案的PI是 癌症遗传学网络统计协调中心，负责 监督CGN研究的数据收集和分析，但不提供资金 统计方法研究。这笔赠款将提供必要的支持，以 完成所需统计方法的研究和应用。这项研究 有可能对弱势群体产生重大影响 癌症家族史，因为这将使筛查战略更专注于 个人患癌症的风险增加，并有可能增加患癌症的几率 这种疾病将在可治疗的阶段被检测到。

Project Summary/Abstract The National Cancer Institute (NCI) established the Cancer Genetics Network (CGN) in 1998 as a multi-centered national project designed to support collaborative investigations on the genetic basis of cancer susceptibility and to explore mechanisms to integrate this new knowledge into medical practice. A fundamental issue in this high-risk population is how to optimally screen for early detection of disease. Recent discovery of new biomarkers that change with early stages of disease offer promising non-invasive strategies for screening. However, two aspects that are fundamental for determining an optimal strategy are: for what cancers is the individual at elevated risk and how can longitudinal measurements on a biomarker be used to predict risk of disease. To focus on these issues in the context of the Cancer Genetics Network project, there are two aims in this proposal. The first aim will develop and apply methods for analyzing data on family history to discover what cancers tend to aggregate in individuals and families. We will develop methodology that can be applied to the CGN Registry (family history) data and appropriately accounts for age and mode of ascertainment of participants. We will apply these methods to identify clusters of cancer sites that occur together in families. The second aim of this proposal focuses on the development and application of a method to assess the relationship between longitudinally collected biomarkers and early evidence of disease onset. The method will allow us to analyze data from the CGN Ovarian Cancer (CA125) Biomarker Screening study, as it will appropriately handle this longitudinal and event time data even though the screening schedule for these women varied due to missed or delayed visits. The PI of this proposal is the PI on the Statistical Coordinating Center of the Cancer Genetics Network, and is responsible for overseeing data acquisition and analysis of CGN studies, but is not funded for statistical methods research. This grant would provide the support needed to complete the research for and apply the required statistical methods. This research has the potential of a major impact on the vulnerable population of individuals with family history of cancer, as it will allow the screening strategy to be more focused on cancers to which the individual is at elevated risk, and potentially improve the chances that the disease will be detected in a treatable stage.