Data-driven cancer genome interpretation for personalised cancer treatment

CGI-Clinics旨在通过优化基因组数据解释（测序后和建议兼容的靶向治疗前）来改善肿瘤学的个性化医疗。解释是下一代测序（NGS）在癌症管理中全面部署和广泛使用的瓶颈。该项目解决了解释癌症突变的3个主要障碍：它不是系统性的;它处理了大多数未知意义的变异，它未能赋予患者权力。目前，肿瘤基因组数据的解释依赖于专家审查分散的数据库和资源的工作，这是一个耗时的过程，可能导致次优的临床决策。CGI-Clinics将通过将相关的公共和私人数据库整合到一站式工具中，使解释过程系统化，并有可能组织由参考医院共同促进的虚拟分子肿瘤委员会。该项目将分为三个阶段：设置（评估需求），验证（与9个临床合作伙伴的试点）和复制（在欧盟的30家医院）。它将实现基因组数据解释的民主化（独立于医院的规模，资源和分析技术），并提供卫生经济学验证。CGI-Clinics将依靠系统的自动学习平台，增加肿瘤中可解释变异的份额（从目前的9-12%增加到至少50%），以及构成药物反应生物标志物的特征。对于大多数癌症患者来说，解释过程是复杂的，使他们远离自己的疾病知识。CGI-Clinics将建立EduCGI，这是一种基于网络的工具，可帮助患者了解通过解释其肿瘤获得的信息，促进与临床医生的知情讨论，并分享他们的研究数据。最终，该项目旨在为决策者提供有关癌症管理的信息，并增强患者的权能。

CGI-Clinics aims at improving personalized medicine in oncology by optimizing genomic data interpretation (after sequencing and before advising on compatible targeted therapies). Interpretation is a bottleneck for the full deployment and broad accessibility of Next Generation Sequencing (NGS) in cancer management. The project tackles the 3 main hurdles in the interpretation of cancer mutations: it is not systematic; it deals with a majority of variants of unknown significance, and it fails to empower patients. Currently the interpretation of tumour genomic data relies on the work of experts reviewing scattered databases and resources, in a time-consuming process that may lead to sub-optimal clinical decisions. CGI-Clinics will systematize the interpretation process by integrating relevant public and private databases in a one-stop shop tool, with the possibility to organize virtual molecular tumour boards co-facilitated by reference hospitals. The project will have three phases: setup (to assess needs), validation (a pilot with the 9 clinical partners) and replication (in 30 hospitals across the EU). It will enable democratization of genomic data interpretation (independent of a hospital’s size, resources, and profiling technology) and provide health economics validation. Relying on a systematic automatic learning platform, CGI-Clinics will increase the share of interpretable variants in tumours (from the current 9-12% to at least 50%), and features that constitute biomarkers of drug response. The interpretation process is complex for most cancer patients, alienating them from knowledge of their illness. CGI-Clinics will build EduCGI, a web-based tool to help patients understand the information gained through interpretation of their tumours, facilitating informed discussions with clinicians, and sharing their data for research. Ultimately, the project is built to inform policymakers on cancer management and to empower patients.