Personalized Risk Prediction to Reduce Cardiovascular Disease in Childhood Cancer Survivors

项目概要/摘要 在今天生活在美国的50万儿童癌症幸存者中，最常报告的非- 癌症严重、危及生命或致命的慢性疾病是心血管疾病（CVD）。它是领先的非- 癌症是这个人群中过早死亡的原因。心脏辐射和蒽环类药物暴露 与各种CVD结局相关，包括心肌病、冠状动脉疾病（CAD）和 心脏瓣膜病放射治疗（RT）相关CVD的研究通常已建立 仅基于整个心脏剂量度量的关联;因此，忽略了器官及其 子结构我们的团队是第一个报告显示CVD亚结构水平剂量反应的数据的团队。 儿童癌症幸存者的风险。尽管建立了不同的放射性，心脏亚结构剂量 由于缺乏经验证的风险，RT治疗计划中通常不包含约束条件 因此，预测模型错过了前瞻性优化RT计划和回顾性优化RT计划的机会。 个性化的风险咨询和长期心血管监测在当前和未来的癌症幸存者。 拟议项目的目标是开发和验证新的CVD风险预测模型， 心脏亚结构剂量。此外，我们建议开发工具，将这些模型临床转化为 有效的个性化治疗范例，可用于护理提供者的前瞻性和回顾性应用 降低CVD风险。我们将：（1）开发和验证心肌病、CAD的风险预测模型， 心脏瓣膜疾病合并心脏RT子结构剂量，调整人口统计学和 化疗暴露;（2）将CVD风险预测模型整合到商业RT治疗计划中 系统和基于Web的应用程序，并通过对当代患者的计算机模拟研究建立其使用 用RT治疗。 这将是第一次研究使用不同心脏亚结构的独特放射敏感性作为 建立模型，预测新诊断患有癌症的儿童患特定类型CVD的风险 以及长期幸存者的情况。将子结构剂量转化为预测模型将 前瞻性RT治疗计划和回顾性风险临床护理显著进步 评估。通过优化分娩， 胸部定向RT与心脏子结构剂量约束，并选择计划，赋予最低 风险，同时保持最佳的临床目标体积覆盖。治疗后，临床团队 可以提供基于证据的个性化风险缓解咨询，基于个性化的风险概况 根据针对化疗暴露调整的输送心脏子结构剂量确定， 人口统计学拟议项目的成功执行有可能改变临床实践， 治疗儿童和青少年癌症患者。

PROJECT SUMMARY/ABSTRACT Among the half a million childhood cancer survivors alive in the US today, the most commonly reported non- cancer severe, life-threatening, or fatal chronic condition is cardiovascular disease (CVD) . It is the leading non- cancer cause of premature death in this population. Heart radiation and anthracycline exposure have been associated with a variety of CVD outcomes including cardiomyopathy, coronary artery disease (CAD), and heart valve disease. Investigations of radiation therapy (RT)-related CVD have typically established associations based solely on whole heart dose metrics; thus, overlooking the heterogeneity of the organ and its substructures. Our team was the first to report data demonstrating substructure-level dose response of CVD risk in childhood cancer survivors. Despite establishing distinct radiosentivities, cardiac substructure dose constraints are not commonly incorporated into RT treatment planning due to the lack o f validated risk prediction models, thus, missing opportunities to prospectively optimize RT planning and retrospectively personalize risk-counseling and long-term cardiovascular surveillance in current and future cancer survivors. The goal of the proposed project is to develop and validate novel CVD risk prediction models that incorporate cardiac substructure doses. Further, we propose to develop tools to clinically translate these models into effective personalized treatment paradigms with prospective and retrospective applications for care providers to reduce CVD risk. We will: (1) develop and validate risk prediction models for cardiomyopathy, CAD, and heart valve disease incorporating cardiac RT substructure doses, adjusting for demographics and chemotherapy exposures; and (2) integrate CVD risk prediction models into commercial RT treatment planning systems and web-based applications, and establish their use via in-silico studies of contemporary patients treated with RT. This will be the first investigation to use the unique radiosensitivity of different cardiac substructures as the foundation for models that can predict the risk of specific types of CVD in children newly diagnosed with cancer as well as among long-term survivors. Incorporating the substructure doses into prediction models will significantly advance clinical care for both prospective RT treatment planning and retrospect ive risk assessments. Prospectively, late CVD risk could be decreased in future survivors by optimizing delivery of chest-directed RT with cardiac substructure dose constraints and selecting the plan that confers the lowest risk, while maintaining optimal clinical target volume coverage. Retrospectively post treatment, the clinical team can provide evidence-based personalized risk mitigation counseling, based on individualized risk profiles determined from delivered cardiac substructure doses adjusted for chemotherapy exposures and demographics. Successful execution of the proposed project has the potential to transform clinical practice for treatment of childhood and adolescent patients with cancer.