Global metabolomics profiling, dietary factors, and colorectal cancer risk in the NIH-Consortium of Metabolomics Studies (COMETS)

总结 结直肠癌（CRC）的患病率在全球范围内持续增加，并且仍然是第三大癌症。 在美国被诊断为癌症。加强对CRC病因学的了解对于 开发量身定制的风险预测方法。代谢组学是对小分子代谢物的综合研究， 发现CRC的病因学生物标志物的有前途的方法。代谢组学分析结合 饮食摄入信息可用于确定CRC的客观饮食生物标志物。 然而，缺乏精确的代谢生物标志物来预测CRC。到目前为止，没有客观的饮食 CRC风险的生物标志物和饮食与CRC之间关系的生物学基础仍然很差 明白 拟议工作的目标是调查代谢物与CRC风险的关系，并加强 我们对饮食与CRC关系的基础生物学的理解。为了实现这一目标，我们将使用 现有的全球代谢组学数据在三个最先进的实验室中使用预诊断生成 来自美国、欧洲和亚洲的8个独立前瞻性队列的生物标本。这些好- 注释和独特的数据集包括来自不同研究的n= 3，085个匹配病例对照对的数据。 因此，确保了所鉴定的生物标志物的广泛的普遍性和临床适用性。数据 整合在NIH资助的代谢组学研究联盟（COMETS）中， 验证的食物频率问卷（FFQ），流行病学和临床数据。我们会发现， 在n= 1，900的发现集中验证新的基于血液的CRC风险代谢生物标志物（目标1） 配对病例对照组。将在独立验证集中确认结果，包括n= 1，185 配对病例对照组。此外，我们将按性别、肿瘤位置和年龄进行分层分析， 诊断，以促进我们对不同亚型CRC病因的理解。使用所描述的 数据集，我们将发现和验证食物组与代谢物的相关性。我们将执行 对表现最好的饮食-代谢物相关性进行中介分析，以研究间接影响 饮食通过代谢物对CRC风险的影响（目标2）。这项研究具有很高的创新性，因为它使用 一个严格的多步骤设计，首次采用了广泛的代谢生物标志物（n~381）和饮食 信息，来自具有癌症前收集的生物样本和问卷的高度特征化队列 诊断，从而防止反向因果关系。我们的跨学科团队在以下方面拥有丰富的经验： 在癌症研究中使用代谢组学，并利用大量的初步数据。我们希望我们 研究将发现和验证新的CRC生物标志物，并提高我们对CRC的理解。 CRC病因学中的饮食基础生物学，从而解决明确定义的临床和公共卫生需求。

Summary The prevalence of colorectal cancer (CRC) continues to increase worldwide and it remains the third most commonly diagnosed cancer in the United States. An enhanced understanding of CRC etiology is essential to develop tailored risk prediction methods. Metabolomics, the comprehensive study of small metabolites, is a promising approach to discover etiological biomarkers for CRC. Metabolomics analysis in combination with information on dietary intake could be used for the identification of objective dietary biomarkers for CRC. However, precise metabolic biomarkers to predict CRC are missing. To date, there are no objective dietary biomarkers for CRC risk and the biology underlying the relationship between diet and CRC remains poorly understood. The goal of the proposed work is to investigate associations of metabolites with CRC risk and to enhance our understanding of the underlying biology of the diet-CRC relationship. To achieve this goal, we will use existing global metabolomics data generated in three state-of-the art laboratories using pre-diagnosis biospecimens from eight independent, prospective cohorts from the US, Europe, and Asia. These well- annotated and unique datasets include data from n=3,085 matched case-control pairs from diverse populations, and thus ensure broad generalizability and clinical applicability of identified biomarkers. Data are integrated in the NIH-funded Consortium of Metabolomics Studies (COMETS) together with standardized and validated food frequency questionnaires (FFQ), and epidemiologic and clinical data. We will discover and validate novel blood-based metabolic biomarkers for CRC risk (Aim 1) in a discovery set of n=1,900 matched case-control pairs. Findings will be confirmed in an independent validation set including n=1,185 matched case-control pairs. Additionally, we will perform stratified analyses by sex, tumor location, and age at diagnosis to advance our understanding of CRC etiology across distinct subtypes. Using the described datasets, we will discover and validate correlations of food groups with metabolites. We will perform mediation analysis for the top performing diet-metabolite correlations to investigate the indirect effect of diet on CRC risk through metabolites (Aim 2). The proposed research is highly innovative in that it uses a rigorous multi-step design, employs for the first time a broad set of metabolic biomarkers (n~381) and dietary information, from highly characterized cohorts with biospecimens and questionnaires collected before cancer diagnosis, thus, protecting against reverse causation. Our interdisciplinary team has extensive experience in using metabolomics in cancer research and leverages substantial preliminary data. We expect that our investigation will discover and validate novel CRC biomarkers and enhance our understanding of the underlying biology of diet in CRC etiology, thus addressing a clearly defined clinical and public health need.