Precision Medicine in Crohn's Disease and Ulcerative Colitis: Predicting Disease Flare and Disease progression

克罗恩病和溃疡性结肠炎是炎症性肠病的常见形式，在英国，每100人中就有1人受到影响。在城市化和采用西方生活方式的推动下，以前不发达的世界的发病率正在急剧上升。IBD通常表现在青春期/成年期早期，肠道功能紊乱，强烈的炎症反应，全身不适，心理社会障碍和严重的健康经济负担。近年来，在IBD方面进行了大规模的生物技术/制药投资，目前已批准使用多种不同的药物形式。然而，缓解率正在触及一年30-40%的上限，而且没有可靠的生物标记物来帮助药物定位。由于我们无法预测治疗反应、疾病爆发或疾病进展，IBD的治疗进一步复杂化。由Charlie Lees博士领导的PrediCCt Study www.predicct.co.uk目前正在从英国各地招募3100名IBD患者。在临床缓解期招募患者，并在两年内进行随访。收集所有患者的详细基线数据。其中包括NHS持有的电子健康记录、全基因组和代谢组测序、通过移动应用程序记录的饮食/生活方式数据以及完整的血样分析。在整个随访过程中，移动应用程序还被用于定期收集患者自我报告的疾病结果。该项目的总体目标是开发和应用计算策略，以改善IBD患者的分层和后续治疗。高维贝叶斯分层模型将被用作一个自然而强大的框架，将来自多个数据源的信息结合在一起，适当地量化统计的不确定性。特别是，我们试图解决以下研究问题：1.疾病表型、饮食、生活方式、遗传学和肠道微生物区系的哪些方面对IBD的疾病发作、b)疾病进展和c)治疗反应有贡献？2.我们如何根据疾病生物学和进展风险对患者进行分层？3.基于此，我们如何干预以改善患者的预后？

Crohn's disease and ulcerative colitis are the common forms of inflammatory bowel disease affecting up to 1 in 100 people in the UK. Incidence rates are increasing sharply in the previously undeveloped world, driven by urbanisation and adoption of a Western lifestyle. IBD typically manifests in adolescence / early adulthood with disturbed bowel function, a robust inflammatory response, systemic upset, psycho-social disturbance and substantial health-economic burden. Recent years have seen massive biotech / pharma investment in IBD with multiple different drug modalities now approved for use. However, remission rates are hitting a ceiling at 1 year of 30-40%, and no reliable biomarkers exist to aid with drug positioning. Management of IBD is further complicated by our inability to prognosticate on treatment response, disease flare or disease progression. The PREdiCCt study www.predicct.co.uk, led by Dr Charlie Lees, is presently recruiting 3100 people with IBD from across the UK. Patients are recruited in clinical remission, and followed-up during two years. Detailed baseline data is collected for all patients. Among others, this includes electronic health records held by NHS, whole-genome and metabolomic sequencing, diet/lifestyle data recorded through a mobile app as well as a full blood panel. Throughout the follow-up, the mobile app is also used to regularly collect patient self-reported disease outcomes. The overarching goal of this project is to develop and apply computational strategies which can improve the stratification and subsequent treatment of IBD patients. High-dimensional Bayesian hierarchical models will be used as a natural and powerful framework to combine information from multiple data sources, whist appropriately quantifying statistical uncertainty. In particular, we seek to address the following research questions: 1. What aspects of disease phenotype, diet, lifestyle, genetics and the gut microbiota contribute to a) disease flare, b) disease progression & c) treatment response in IBD? 2. How can we stratify patients based on disease biology and risk of progression? 3. Based on this, how can we intervene to improve patient outcomes?