Genetic testing to guide pediatric cancer care and follow up: using anthracycline-associated cardiac toxicity as a model for the future

项目总结/摘要 最近的基因组学进展已经确定了生殖系风险标志物预测治疗风险的潜力- 治疗癌症的儿童的相关毒性。尽管儿科癌症治疗的进步已经导致 存活率、长期发病率和早期死亡风险的显著增加，强调需要 在保持疾病控制的同时最大限度地减少晚期毒性的新方法;使用遗传标记， 预测风险和改变治疗是减少不良结局的潜在策略。遗传风险 评估使人们对适应患者的个性化治疗的可能性感到非常兴奋 风险增加或减少的人。例如，对癌症幸存者群体的研究表明， 充血性心力衰竭的风险增加五倍，心脏性早死的风险增加七倍。 最近发表了临床实践建议，用于治疗的个体化适应， 基于蒽环类药物治疗儿童癌症患者的遗传心脏风险特征的监测 疗法然而，相当大的不确定性围绕着遗传信息如何成功地转化为 改善癌症儿童的护理和结果。此外，儿童癌症的罕见性和长期 观察晚期结果所需的延迟限制了前瞻性试验评估精确性的可行性， 医学方法进行护理和随访。我们建议采用决策分析建模方法， 确定基因检测如何为临床护理提供信息，包括初始癌症治疗和治疗后护理 基于个体对心脏毒性的敏感性。我们将开发一种新颖的，灵活的微观模拟模型， 儿童癌症的临床过程，以预测与儿童相关的全方位健康结果， 癌症，包括最初的疾病控制和治疗相关的晚期毒性，然后将遗传数据 评估对这些结果的影响。该建模框架将集成来自多个 资源，包括儿童癌症幸存者研究（CCSS），以（1）预测长期结果， （2）确定利用遗传变异检测的临床影响， 指导癌症护理的心脏毒性;和（3）评估遗传标记的考虑如何改善随访- 为高危幸存者提供心脏筛查建议。我们的目标是描绘的范围和性质， 围绕模型参数的不确定性及其对建模结果的影响， 自举方法，外推心脏毒性风险的统计方法，以及 不确定性分析通过独特地利用CCSS数据来表征终身治疗相关的 心脏毒性风险，我们提出的研究将建立一个新的分析框架，以评估 在儿科肿瘤学中使用基因检测的不确定性和权衡，并形成了基础， 了解遗传风险测试对其他毒性的影响。

PROJECT SUMMARY/ABSTRACT Recent genomic advances have identified a potential for germline risk markers to predict risk of treatment- related toxicity in children treated for cancer. Although advances in pediatric cancer treatment have led to remarkable increases in survival rates, long-term morbidity and early mortality risks underscore the need for new approaches that minimize late toxicities while maintaining disease control; using genetic markers to predict risk and change therapy represents a potential strategy to reduce adverse outcomes. Genetic risk assessment has created great excitement about the possibility of individualized therapy adapted for patients who carry increased or decreased risk. For example, research in cancer survivor cohorts have demonstrated a five-fold increased risk of congestive heart failure and a seven-fold increased risk of premature cardiac death. Clinical practice recommendations have recently been published for individualized adaptation of treatment and surveillance based upon genetic cardiac-risk profile in childhood cancer patients receiving anthracycline therapy. However, considerable uncertainty surrounds how successfully genetic information may translate into improved care and outcomes for children with cancer. Further, the rarity of childhood cancer and the long latency needed to observe late outcomes limit the feasibility of prospective trials to evaluate a precision- medicine approach to care and follow-up. We propose to employ a decision-analytic modeling approach to determine how genetic testing may inform clinical care, both of initial cancer therapy and post-treatment care based on individual susceptibility for cardiotoxicity. We will develop a novel, flexible microsimulation model of the clinical course of childhood cancer to project the full spectrum of health outcomes relevant to the childhood cancer, including initial disease control and treatment-related late toxicities, and then incorporate genetic data to assess the impact upon these outcomes. This modeling framework will integrate data from multiple resources, including the Childhood Cancer Survivors Study (CCSS) to (1) project long-term outcomes for children diagnosed with cancer; (2) determine the clinical impact of utilizing genetic variant testing for cardiotoxicity in guiding cancer care; and (3) assess how consideration of genetic markers can improve follow- up cardiac screening recommendations for at-risk survivors. We aim to portray the scope and nature of the uncertainties that surround model parameters and their impact on modeled outcomes by employing bootstrapping methods, statistical methods to extrapolate cardiotoxicity risks, and rigorous approaches to uncertainty analysis. By uniquely leveraging the CCSS data to characterize the lifelong treatment-related cardiotoxicity risks, our proposed research will establish a novel analytic framework for evaluating the uncertainties and tradeoffs surrounding the use of genetic testing in pediatric oncology, and form the basis for understanding the impact of genetic risk testing for other toxicities.