Division of Translational Research & Applied Statistics, Department of Public Health Sciences, University of Virginia, P.O. Box 800717, Charlottesville, VA, USA.
BMC Cancer. 2018 Feb 5;18(1):133. doi: 10.1186/s12885-018-4038-x.
Broad implementation of model-based dose-finding methods, such as the continual reassessment method (CRM), has been limited, with traditional or modified 3 + 3 designs remaining in frequent use. Part of the reason is the lack of reliable, easy-to-use, and robust software tools for designing and implementing more efficient designs.
With the aim of augmenting broader implementation of model-guided methods, we have developed a web application for the Bayesian CRM in the R programming language using the Shiny package. The application has two components, simulation and implementation. Within the application, one has the ability to generate simulated operating characteristics for the study design phase, and to sequentially provide the next dose recommendation for each new accrual or cohort based on the current data for the study implementation phase. At the conclusion of the study, it can be used to estimate the maximum tolerated dose (MTD). The web tool requires no programming knowledge, and it is free to access on any device with an internet browser.
The application provides the type of simulation information that aid clinicians and reviewers in understanding operating characteristics for the accuracy and safety of the CRM, which we hope will augment phase I trial design. We believe that the development of this software will facilitate more efficient collaborations within study teams conducting single-agent dose-finding trials.
基于模型的剂量发现方法(如连续评估法(CRM))的广泛应用受到限制,传统或改良的 3+3 设计仍在频繁使用。部分原因是缺乏可靠、易用且稳健的软件工具来设计和实施更有效的设计。
为了增强对模型指导方法的更广泛应用,我们使用 R 编程语言中的 Shiny 包为贝叶斯 CRM 开发了一个网络应用程序。该应用程序有两个组件,模拟和实现。在应用程序中,用户可以在研究设计阶段生成模拟操作特征,并在研究实施阶段根据当前数据为每个新入组或队列提供下一个剂量推荐。在研究结束时,它可用于估计最大耐受剂量(MTD)。该网络工具不需要编程知识,并且可以在任何带有互联网浏览器的设备上免费访问。
该应用程序提供了有助于临床医生和审查者理解 CRM 的准确性和安全性的操作特征的模拟信息类型,我们希望这将增强 I 期试验设计。我们相信,该软件的开发将促进进行单药剂量发现试验的研究团队之间更有效的合作。