Polygenic Risk Scores for Diverse Populations - Bridging Research and Clinical Care

摘要 心血管疾病(CVD)及其危险因素造成重大社会负担，是 发病率、死亡率和残疾。精准医疗在解决心血管疾病及其风险因素方面具有独特的地位， 经过数十年的调查和数十亿美元的投资，建立了强大的 潜在的遗传基础。多基因风险评分(PR)，将风险变量聚合为单一评分， 提供了一个这样的例子。心脑血管疾病的PRS研究已从估算过渡到临床检验 公用事业；即确定何时以及如何采用PR，以及如何以类似方式构思 环境/生活方式风险评分(ERS)可与PR一起用于临床。然而，大多数人 包括在大规模心血管疾病研究中的参与者都是欧洲血统(EA)，限制了全球 将遗传关联转化为与所有人口相关的临床和公共卫生应用。 PAGE联盟和其他人已经证明，EA派生的PR不能直接翻译到 种族/民族多元化的人口。用于估计和解释比率的统计工具建立在 在人口结构的背景下，被违反并造成偏见的强烈假设 种族/民族多元化的人口。这些研究差距将加剧长期存在的种族/民族问题 心血管疾病及其风险因素的差异，强调需要进行研究，使所有群体都能获得 支持PRS的个性化防护的优势。 在此修订的应用程序中，我们通过利用以下功能解决了以前在原始应用程序中确定的限制 来自群组和生物库网络的高质量、协调和集中可用的数据， 电子健康记录，捕捉心血管疾病及其风险因素。通过这一努力，我们将包括150多万非 欧洲血统参与者开发和验证人种/人种中心血管疾病相关特征的PR 不同的人群。我们将为临床和公共卫生创建方法、资源和最佳实践 社区。这项研究将允许采用和应用PRS进行检测、干预和 心血管疾病危险因素的处理。我们的最终目标是减少和预防所有人群的心血管疾病负担。 我们的具体目标是(1)创建不偏不倚的方案：结合ERS开发和评估CVD方案 在人种/人种多样化的大页研究中；和(2)验证、校准和传播： 对生物库中的风险评分模型进行外部验证和改进，并将风险评分模型转换为改进的 医学界的接触和理解。 我们将构建下一代方法、资源和最佳实践，以增强适当的 脑血管病患者认知功能障碍的发展及后续预测和临床询问。有意将重点放在非EA上 人口将确保他们不会是基因组医学新时代的最后受益者。

Abstract Cardiovascular disease (CVD) and its risk factors impose major societal burdens and are leading causes of morbidity, mortality, and disability. Precision medicine is uniquely positioned to address CVD and its risk factors, enabled by decades of investigation and billions of dollars of investment that have established their strong underlying genetic basis. Polygenic risk scores (PRS), the aggregation of risk variants into a single score, provides one such example. Research on PRS in CVD has transitioned from estimation to examining the clinical utility; i.e., determining when and how PRS adoption will occur and how similarly conceived environmental/lifestyle risk scores (ERS) can be used clinically in concert with PRS. However, the majority of participants included in large-scale CVD research have been of European ancestry (EA), limiting the global translation of genetic associations into clinical and public health applications relevant for all populations. The PAGE consortium and others have demonstrated that EA-derived PRS are not directly translatable to racially/ethnically diverse populations. Statistical tools for PRS estimation and interpretation are founded on strong assumptions that are violated, and create bias, in the context of population structure that characterizes racially/ethnically diverse populations. These research gaps will exacerbate long-standing racial/ethnic disparities in CVD and its risk factors, underscoring the need for research that enables all groups to reap the benefits of PRS-enabled personalized prevention. In this revised application, we address the limitations previously identified in our original application by leveraging high-quality, harmonized, and centrally available data from a network of cohorts and biobanks with linked electronic health records, capturing CVD and its risk factors. Through this effort, we will include over 1.5M non- European ancestry participants to develop and validate PRS for CVD-associated traits in racially/ethnically diverse populations. We will create the methods, resources, and best practices for the clinical and public health communities. This research will permit adoption and application of PRS for the detection, intervention, and treatment of CVD risk factors. Our ultimate goal is to reduce and prevent the burden of CVD in all populations. Our Specific Aims are (1) Creation of unbiased PRS: Develop and evaluate CVD PRS in combination with ERS in the large and racially/ethnically diverse PAGE study; and (2) Validation, calibration and dissemination: Externally validate and improve upon risk score models in biobanks and translate risk score models for improved access and understanding for the medical community. We will build the next generation of methods, resources, and best-practices to empower appropriate development of PRS and subsequent prediction and clinical interrogation in CVD. Deliberate focus on non-EA populations will ensure that they are not the last to benefit in the new era of genomic medicine.