Improving Prediction of Medical Responsiveness and Clinical Outcomes in Crohn's Disease

 描述(由申请人提供)：这是密歇根大学胃肠病专家瑞安·斯蒂德汉姆博士重新提交的K23职业发展奖。斯蒂德姆博士的长期职业目标是开发客观的预后，使炎症性肠病患者的治疗决策个人化。该提案的近期目标是评估分析形态组学和血清糖蛋白组学，前者是量化CT扫描结果的图像分析平台，后者用于预测克罗恩病患者的治疗反应。Stidham博士的综合研究和职业发展计划与以下目标相一致：将自己发展成为一名协作式翻译调查员，擅长大型数据集的统计分析，应用新的预后方法来改善克罗恩病的临床结果。克罗恩病(CD)在美国影响着70多万名患者。近60%的CD患者在确诊后10年内因药物反应不佳的疾病需要手术切除。通常，患者的深肠损伤包括药物反应性炎性损伤和无反应性纤维化损伤。尽管存在炎症靶点，内科治疗可能是徒劳的，应该及时进行手术治疗。此外，使用早期高强度免疫抑制来预防未来的纤维狭窄疾病的策略可能会过度治疗部分CD人群。因此，治疗决定是基于(1)现有疾病活动对药物治疗有反应的概率和(2)未来发生纤维狭窄并发症的可能性。使用结肠镜和炎症替代物的评估，包括成像、血液和大便生物标志物，无法提供包括纤维化在内的肠道损伤的全面和定量评估。尽管它很重要，但目前还没有措施来解释深层肠壁损伤和肠道纤维化。这项研究计划将利用深肠损伤的定量成像结果以及反映疾病表型的基于血液的生物标记物来预测医疗失败，定义为需要未来的肠切除、住院和长期使用类固醇。为了实现我们的成像目标(目标1)，我们将使用分析形态组学，一种计算机图像分析方法，从纵向1200名CD患者队列中的4000多张CT-肠道造影术中量化身体和肠道成分。在目标2中，血清糖蛋白组的变化将在仔细分型的炎症和狭窄的受试者中进行表征，以提炼我们先前存在的纤维化的生物标志物。我们已经证明，血清糖蛋白质组的变化反映了肠纤维化的程度，并期望血清糖蛋白质组学能够区分CD的炎性和纤维化表型。在一项儿科前瞻性队列研究中，优化的血清糖蛋白组谱将用于预测未来从B1炎症性疾病转变为B2/3纤维狭窄疾病(风险研究)。目的3进行一项结合分析形态组学和血清糖蛋白组学以预测治疗反应和临床结果的前瞻性研究；为未来的R03/R01研究提供初步数据。斯蒂德姆博士是密歇根大学胃肠病学部的医学讲师。他在弗吉尼亚大学(AOA)获得医学学位，在宾夕法尼亚大学完成内科实习，并于2011年成为密歇根大学的T-32研究员。他的研究背景结合了Peter Higgins博士(主要导师)的实验室和临床经验，专注于新型生物标记物的开发。Stidham博士之前曾因试验工作获得过几项补助金，包括MICHR-CTSA T32蛋白质组学试验补助金和克罗恩和结肠炎基金会超声成像研究职业发展奖。他正在密歇根大学公共卫生学院完成临床研究设计和统计分析硕士课程(2015年4月入学)。他在几个胃肠病学杂志上发表了文章，在国际上发表了研究成果，并建立了本提案中的独立合作。中心到 在这个职业发展奖中，候选人将利用几位导师和合作者的专业知识来发展分析成像、翻译蛋白质组学、 生物信息学和机器学习方法。密歇根大学拥有全国公认的炎症性肠病计划、国际公认的蛋白质组学专业知识，以及专注于器官计算图像分析的专门分析形态组学小组。辅导性研究培训将通过密歇根大学计算医学和生物信息学中心和公共卫生学院提供生物信息学、医学图像分析、机器学习方法和决策支持系统方面的重点研究生水平课程。胃肠病科和医学部一直以来都为斯蒂德姆博士提供强有力的支持，并将提供必要的时间、资源和指导，以实现他的专业和研究目标。通过该职业发展奖提供的培训将对候选人作为专注于个性化治疗管理的独立翻译调查员的发展至关重要。这项研究将为未来的合作研究进一步开发这些和其他炎症性肠病预后工具提供基础。

 DESCRIPTION (provided by applicant): This is a K23 career development award resubmission for Dr. Ryan Stidham, a gastroenterologist at the University of Michigan. Dr. Stidham's long-term career goal is to develop objective prognostics to personalize treatment decisions for patients with inflammatory bowel disease. The immediate goals of the proposal are to evaluate both analytic morphomics, an image analysis platform to quantify CT scan findings, and serum glycoproteomic profiles to predict therapeutic response in patients with Crohn's disease. Dr. Stidham's integrated research and career development plan align with goals of developing himself as a collaborative translational investigator, skilled in statistical analysis o large datasets, applying novel prognostics to improve clinical outcomes in Crohn's disease. Crohn's disease (CD) affects over 700,000 patients in the United States. Nearly 60% of CD patients require surgical resection for medically-unresponsive disease within 10 years of diagnosis. Frequently, patients present with deep bowel injury composed of both medically-responsive inflammatory and non- responsive fibrotic injury. Despite the presence of an inflammatory target, medical therapy may be futile and timely surgical management should be pursued. Further, strategies using early high-intensity immunosuppression to prevent future fibrostenotic disease are likely to over treat portions of the CD population. Therefore, therapeutic decisions are predicated on (1) the probability that existing disease activity will respond to medical therapy and (2) the probability of developing future fibrostenotic complications. Assessment using colonoscopy and surrogates of inflammation including imaging, blood, and stool-based biomarkers fail to provide comprehensive and quantitative assessments of bowel injury, including fibrosis. Despite its importance, there are no measures to account for deep bowel wall injury and intestinal fibrosis. This research proposal will utilize quantitative imaging findings of deep bowel injury in conjunction with blood-based biomarkers reflecting disease phenotypes to predict medical failure, defined as requiring future bowel resection, hospitalization, and prolonged steroid use. To achieve our imaging aims (Aim 1) we will use analytic morphomics, a computer image analysis method, to quantify body and bowel composition from over 4,000 CT-enterographies in a longitudinal 1200 patient CD cohort. In Aim 2, serum glycoproteome variance will be characterized in carefully phenotyped inflammatory and stricturing subjects to refine our preexisting biomarkers of fibrosis. We have demonstrated that variations of the serum glycoproteome reflect the degree of intestinal fibrosis present and expect that serum glycoproteomics can differentiate inflammatory from fibrotic phenotypes in CD. Optimized serum glycoproteome profiles will be used to predict future conversion from B1- inflammatory disease to B2/3 fibrostenotic disease in a pediatric prospective-inception cohort (RISK study). Aim 3 entails a prospective study combining analytic morphomics and serum glycoproteome profiles to predict therapeutic response and clinical outcomes; providing preliminary data for future R03/R01 studies. Dr. Stidham is a Lecturer of Medicine at the University of Michigan Division of Gastroenterology. He earned his medical degree from the University of Virginia (AOA), completed an internal medicine residency at the University of Pennsylvania, and was a T-32 research fellow at the University of Michigan (2011). His research background combines laboratory and clinical experience with Dr. Peter Higgins (primary mentor), focusing on novel biomarker development. Dr. Stidham has been awarded several prior grants for pilot work, including a MICHR-CTSA T32 Pilot Grant for Proteomics and a Crohn's and Colitis Foundation Career Development Award for ultrasound imaging research. He is in the process of completing a Masters program in Clinical Research Design and Statistical Analysis at the University of Michigan School of Public Health (matriculation in April 2015). He has published in several gastroenterology journals, presented research internationally, and has built the independent collaborations featured in this proposal. Central to this career development award, the candidate will leverage the expertise of several mentors and collaborators to develop deep foundational skillsets in analytic imaging, translational proteomics, bioinformatics, and machine learning methods. The University of Michigan houses a nationally recognized Inflammatory Bowel Disease Program, internationally recognized proteomics expertise, and a dedicated Analytic Morphomics Group focused on computational image analysis of organs. The mentored research training will be supplemented with focused graduate-level coursework in bioinformatics, medical image analysis, machine learning methodologies, and decision support systems provided through the University of Michigan Center for Computational Medicine and Bioinformatics and the School of Public Health. The Division of Gastroenterology and Department of Medicine have a history of strong support for Dr. Stidham and will provide the time, resources, and mentorship necessary to achieve his professional and research goals. The training provided through this career development award will be pivotal for the candidate's development as an independent translational investigator focused on individualizing therapeutic management. This research will provide the foundation for future collaborative studies to further develop these and other prognostic tools for the inflammatory bowel diseases.