Validation and extension of the PREMM Model for mismatch repair gene mutations

描述(由申请人提供)：我们最近开发了MLH1和MSH2突变预测模型(PREMM1，2)，这是一种临床预测规则，旨在由医疗保健提供者使用来估计个人携带MLH1或MSH2错配修复基因突变的概率(Balmana等人)。JAMA 2006)。PREMM1，2被开发出来，随后在2000多名有林奇综合症高风险并接受遗传评估的患者中得到验证。Lynch综合征，又称遗传性非息肉病性结直肠癌(HNPCC)，是最常见的遗传性结肠癌综合征，也增加了几种结肠癌的风险。MLH1和MSH2突变是最常见的潜在原因。在一个家族子集中，在第三个错配修复基因MSH6中发现了突变，并导致了较温和的表型。Lynch综合征患者的肿瘤表现为微卫星不稳定性(MSI)，并可表现出受影响的错配修复基因的表达缺失，这可通过免疫组织化学分析(IHC)来检测。除了PREMM1，2，最近还开发了另外两个预测模型，临床医生也可以使用各种临床指南作为风险评估工具。现有模型和临床指南的性能特点尚未与另一种进行比较，也不知道哪种方法或方法的组合是识别和管理Lynch综合征患者的最具成本效益的方法。作为我们先前工作的扩展，该项目的目标是(1)在几个现有的美国和国际结肠癌登记中验证PREMM1，2，包括结肠癌家族登记(CCFR)、EPICOLON和俄亥俄州立大学队列；(2)扩展PREMM1，2模型，以包括MSH6突变携带者的预测；(3)开发一个版本的PREMM1，2，6模型，该模型(I)允许纳入关于MSI和IHC的信息，以进一步精炼胚系突变的预测概率，以及(Ii)基于临床病史、MSI和IHC结果进行特定基因的预测；(4)比较PREMMM1、2，6、Barnetson、MMRPro模型在(I)区分突变携带者和非携带者的能力；(Ii)不同环境(人群或临床病例)；(Iii)基于人群的一系列子宫内膜癌病例；以及(5)进行成本-效果分析，以确定PREMM1、2，6模型和其他临床和分子策略在确定Lynch综合征风险个体方面的作用。公共卫生相关性：这一努力的结果将实现PA-07-021的几个目标。它将(I)导致采用全面的临床和公共卫生方法来识别和管理遗传性结肠癌患者；(Ii)允许更准确地估计与每个MMR基因相关的特定肿瘤，从而更有针对性地进行癌症筛查和预防方法；以及(Iii)将来自不同背景的研究人员聚集在一起，包括遗传学家、分子流行病学家、统计学家和临床医生，他们将合作，分享现有数据和专业知识，以解决无法单独回答的研究问题。

DESCRIPTION (provided by applicant): We have recently developed the Prediction of MLH1 and MSH2 mutations model (PREMM1,2), a clinical prediction rule designed to be used by healthcare providers to estimate the probability that an individual carries a mutation in the MLH1 or MSH2 mismatch repair genes (Balmana et al. JAMA 2006). PREMM1,2 was developed and subsequently validated in over 2000 patients who were at elevated risk of Lynch Syndrome and undergoing genetic evaluation. Lynch Syndrome, also known as hereditary non- polyposis colorectal cancer (HNPCC), is the most common hereditary colon cancer syndrome, and also increases the risk for several extracolonic cancers. MLH1 and MSH2 mutations are the most common underlying cause. In a subset of families, mutations are found in a third mismatch repair gene, MSH6, and lead to a milder phenotype. Tumors in patients with Lynch syndrome demonstrate microsatellite instability (MSI) and can show loss of expression of the affected mismatch repair gene, which may be detected by immunohistochemical analysis (IHC). In addition to PREMM1,2, two other prediction models have recently been developed, and a variety of clinical guidelines are also available for the clinician as risk assessment tools. The performance characteristics of the available models and clinical guidelines have not been compared with another, nor is it known which approach, or combination of approaches, is the most cost-effective method of identifying and managing patients with Lynch Syndrome. As an expansion of our prior work, the aims of this project are (1) To validate PREMM1,2 in several existing U.S. and international colon cancer registries, including the Colon Cancer Family Registries (CCFR), EPICOLON, and Ohio State cohorts; (2)To expand the PREMM1, 2 model to include prediction of MSH6 mutation carriers; (3) To develop a version of the PREMMM1,2,6 model that (i) allows for the incorporation of information on MSI and IHC for further refinement of the predicted probabilities of a germline mutation, and (ii) makes gene-specific predictions based on clinical history, MSI and IHC results; (4)To compare the PREMMM1,2,6 , Barnetson, MMRPro models in (i) their ability to distinguish mutation carriers from non-carriers; (ii) different settings (population or clinic-based cases); (iii) a population-based series of endometrial cancer cases; and (5) To perform a cost-effectiveness analysis to define the role of the PREMM1,2,6 model and other clinical and molecular strategies for identification of individuals at risk of Lynch syndrome. PUBLIC HEALTH RELEVANCE: The results of this effort will achieve several goals of PA-07-021. It will (i) lead to a comprehensive clinical and public health approach to the identification and management of patients with hereditary colon cancer; (ii) allow for more precise estimation of the specific tumors associated with each MMR gene and therefore more targeted cancer screening and prevention approaches; and (iii) bring together investigators from diverse backgrounds, including geneticists, molecular epidemiologists, statisticians and clinicians, who will work together, share existing data and expertise to address research questions that could not be answered in isolation.