CANTRANce: A tool to translate intermediate endpoints to mortality in CE studies

描述（由申请人提供）：癌症比较有效性（CE）研究的主要挑战之一是随访数据的局限性通常要求研究者使用中间终点，而不是最感兴趣的终点，即疾病特异性死亡率。该提案旨在开发一个灵活、便携、高效的框架，将癌症CE研究的中间终点结果转化为疾病特异性死亡的推断。CANTRANce（CANCer有效性比较翻译）框架将由一套计算机微观模拟模型组成，这些模型对应于癌症控制研究的主要类型，即癌症预防、早期检测的生物标志物开发、筛查、治疗选择的诊断检测和主要治疗研究。该框架将通过用户友好的界面公开提供，使研究者能够输入其CE研究结果，提供翻译过程所需的参数，并指定感兴趣的输出。我们的目标如下：（1）开发适用于各种癌症的五个原型模型，将主要类型癌症控制研究中常见的干预措施对中间终点的影响转化为对特定疾病死亡影响的预测。CANTRANce处理的中间终点将包括：（1）疾病发生率，（2）生物标志物检测潜伏性疾病的灵敏度，（3）诊断时的分期分布，（4）主要治疗频率，和（5）主要治疗失败时间。(2)使用模型（e）作为基于斯堪的纳维亚前列腺癌第4组临床试验的疾病复发结果的原理证明来证明这种方法，以预测随机接受根治性前列腺切除术或观察等待组的男性之间的死亡率差异。使用疾病特异性死亡率的试验结果来验证模型，并探索不同关键用户假设的影响。(3)设计、实现、测试和发布一个Web接口，以促进建模框架的直接利用，并使底层技术公开可用。作为这一目标的一部分，我们将为使用目标1中列出的中间终点进行癌症CE研究的研究者提供一个界面版本，并将征求他们的反馈意见，以改进模型和界面。建议的框架代表了一种方法，用于解决一个普遍存在的问题，在CE研究和技术的方法的实施。考虑到大量计划或正在进行的癌症CE研究，以及其中许多可能使用中间终点，我们预计CANTRANce将成为癌症CE研究方法工具箱中非常有用和广泛适用的组成部分。 公共卫生相关性：许多比较预防、治疗或治愈癌症方法的研究没有时间或信息来评估正在研究的方法如何影响癌症死亡。该提案将开发一个软件系统，将这些研究的结果转化为对所比较的方法对疾病所致死亡的影响的预测。

DESCRIPTION (provided by applicant): One of the main challenges in Comparative Effectiveness (CE) research in cancer is that limitations in follow- up data often require investigators to use intermediate endpoints rather than the endpoint of most interest, namely, disease-specific mortality. This proposal aims to develop a flexible, portable, and efficient framework to translate cancer CE study results about intermediate endpoints into inferences about disease-specific deaths. The framework, CANTRANce (CANcer TRANslation for Comparative Effectiveness), will consist of a suite of computer micro simulation models corresponding to the main types of cancer control studies, namely, cancer prevention, biomarker development for early detection, screening, diagnostic testing for treatment selection, and primary treatment studies. The framework will be made publicly available via a user-friendly interface that will enable study investigators to input their CE study results, provide parameters necessary for the translation process, and specify outputs of interest. Our aims are as follows: (1) Develop five prototype models applicable across a broad range of cancers to translate the effects of interventions on intermediate endpoints commonly encountered in the main types of cancer control studies into projections of impacts on disease-specific deaths. The intermediate endpoints addressed by CANTRANce will consist of: (1) disease incidence, (2) sensitivity of biomarkers to detect latent disease, (3) stage distributions at diagnosis, (4) primary treatment frequencies, and (5) primary treatment failure times. (2) Demonstrate this methodology using Model (e) as a proof-of-principle based on disease recurrence results from the Scandinavian Prostate Cancer Group 4 clinical trial to project mortality differences between men randomized to radical prostatectomy or watchful waiting arms. Use the trial results on disease-specific mortality to validate the model and explore the implications of varying key user assumptions. (3) Design, implement, test, and release a web interface that will facilitate direct utilization of the modeling framework and make the underlying technology publicly available. As part of this aim we will provide a version of the interface to investigators doing CE research in cancer using the intermediate endpoints listed in Aim 1 and will solicit their feedback to improve the models and interface. The proposed framework represents both a method for addressing a pervasive problem in CE studies and a technology for implementation of the method. Given the large number of cancer CE studies that are either planned or ongoing, and given that many of these are likely to use intermediate endpoints, we expect that CANTRANce will become a highly useful and broadly applicable component of the methods toolbox for CE research in cancer. PUBLIC HEALTH RELEVANCE: Many studies comparing methods to prevent treat, or cure cancer do not have the time or the information to evaluate how the approaches being studied affect cancer deaths. This proposal will develop a software system to translate the results of these studies into projections of the effects of the methods being compared on deaths due to the disease.