当前产品

免费用户可以在一个方案场景上运行最多10次模拟

一期临床

剂量探索 - 单药剂量试验设计

剂量探索 - 双药联合剂量试验设计

队列扩展

二期临床

监测

2/3期无缝设计

样本量计算

即将上线产品

我们不懈努力,只为给您带来更多更好的产品,敬请关注!

CAR-T一期试验设计

一款简单高效的CAR-T一期剂量探索试验设计

亚组富集试验设计

用于精准临床试验的亚组富集设计和方法

适应性剂量嵌入

允许在试验期间插入新剂量以增加成功可能性的适应性剂量插入设计(用于寻找最佳剂量)

PoD-TPI

Time-to-event toxicity probability interval design based on probabilities of decisions

MUCE (多队列扩展)

MUCE 咨询服务请联系我们

背景
现代的早期临床试验通常会在多个适应症中测试新药的多个剂量,以此确定哪些适合的剂量和臂(arm)可以推进到II期或III期试验。传统做法一般是对于每个剂量或适应症都会在一个试验中进行单独测试,从而产生多种方案和多个试验。而MUCE是一种全新的贝叶斯统计解决方案,它可以允许在一次试验中对多个臂进行研究,并通过使用贝叶斯分层模型来借用臂之间的信息,进而极大地提高决策效率。

什么是 MUCE?
MUCE是一种全方位的统计解决方案,包括对多臂临床试验的自适应设计和数据分析。它采用了贝叶斯分层模型,实现了三个主要目标:
  1. 提高了药物开发的统计效力(成功概率)
  2. 减少样本量
  3. 有效地控制I类错误(选择错误药物进入下一阶段的概率)
MUCE 可用于任何具有3个或更多臂的临床试验,包括:
  • 多个队列扩展的临床1b期试验
  • 多臂的临床II期试验
  • 包括篮式,伞式和平台试验的主协议

MUCE 可以做什么?
MUCE涵盖了从方案撰写到最终决策制定的整个临床试验过程。所提供的服务总结在以下流程图中。

MUCE 实战案例
MUCE 已在中国和美国的多个临床试验上实施。
  1. 2018年,MUCE应用于中国一家制药公司的1a-1b期无缝试验。其中MUCE可对3个剂量和4个适应症进行扩展,也就是说试验可以同时对12个臂进行扩展。 该临床试验的IND申请已获得中国药监局审评中心(CDE)临床试验默示许可,并已实际启动开始入组第一位患者。MUCE已成为唯一可以在有限样本量下同时处理12个臂的临床试验设计。
  2. 2019年,MUCE应用于美国的一个1a /1b期试验,以帮助设计多个队列的扩展。该试验在剂量爬坡过后有3个队列进入到了下一阶段以确定RP2D是否可以在3种不同适应症中表现出抗肿瘤活性。该试验已在美国提交IND批准。
  3. 2019年,MUCE应用于一个3臂的II期试验。该实验在每个臂上都对两种免疫疗法联合治疗的效果在一二三线癌症治疗中进行了研究。与Simon的两阶段设计相比,MUCE为该试验节约了一半的样本量。
  4. 2019年,MUCE应用于一个两臂的1b / II期试验以指导两臂的队列扩展并对每个臂进行Go/No Go决策。在试验方案研究当中,我们把MUCE和Berry等人的贝叶斯方法(Berry et al. 2013)进行了比较,结果显示MUCE对于I类错误率的控制有了极大的提高。

Laiya-One: A 21st Century Phase 1 Trial Design Suite

Laiya-ONE 咨询服务请联系我们

A proposal to transform phase 1 oncology trials at company xyz
Laiya Consulting, Inc.
September 28, 2019

Purpose:

to promote the use of a novel phase 1 trial design suite that tailors to modern needs of phase 1 clinical trials in oncology.

Introduction:

Based on a set of innovative statistical designs and approaches, Laiya proposes a suite of solutions to transform phase 1 trials for new drug development (Figure 1). The solutions aim to short the trial duration, reduce sample size and cost, but still maintain safety and sufficient power. Innovatively, we also provide decision tools that help control the type I error of critical decision making to reduce the failure rates of phase 2 and 3 trials.

Figure 1: Laiya One -- A 21st Century Phase 1 Trial Design Suit

How/Where can Laiya-One Help During Phase 1 Development

Laiya-One provides end-to-end services on the entire phase 1 trial and drug develop (Figure 2)

Figure 2 Laiya-One main functions for the phase 1 trial and drug development

Available Designs and Benefits:

The suite includes four novel designs.

  1. The i3+3 design (and extensions). It is a cohort-based dose-finding design suitable for standard DLT-based phase 1a trials. Below are some notable features and benefits.
    • i3+3 has much higher chance of identifying the true MTD
    • i3+3 has proven success at US and Chinese IND applications
    • i3+3 is operationally as simple as the 3+3 design
    • i3+3 variations (mi3+3 and CI3+3) allows biomarker subgroups and combo dose finding
  2. The PoD-TPI design. It is a break-through type of design allowing faster patient enrollment for standard DLT-based phase 1a trials. Below are some notable features and benefits. It is the first type of dose-finding designs that offers the following features and benefits.
    • PoD-TPI quantifies the risk of dose-finding decisions – this is critical if we want to speed up enrollment
    • PoD-TPI accelerates patient enrollment while maintaining safety
    • PoD-TPI assesses existing designs and their risks – important to benchmark future designs
  3. The MUCE design. It is a novel design and decision-making tool allowing more efficient sample size/power/decision making process for phase 1b multiple expansion cohorts studies.
    • MUCE can accommodate multiple doses, schedules, combinations, and indications to be included in a single study.
    • MUCE results in a smaller sample size (as much as half) than the standard designs, such as the Simon’s two-stage design
    • MUCE has a build-in function to control the overall type I error rate – to reduce false selection of indication and dose for phase 2 trials

Examples:

The i3+3 design and its alternatives

  1. Laiya has helped a US company to secure an IND approval including an mi3+3 design is used to help guide all comers and a biomarker-positive subpopulation go through dose finding states simultaneously. This greatly reduced the time and cost since without the mi3+3 design, two sequential dose-finding trials will be needed, one for the all comers and one for the subpopulation.
  2. Laiya has helped multiple Chinese and US companies design seamless phase 1a dose-escalation and multiple expansion cohort trials. The dose-escalation part is based on the i3+3 design and the expansion cohort part is based on the MUCE design. This helps speed up the drug development with a single protocol, and the MUCE design results in 10-40% sample size reduction.
  3. Laiya has helped with investigator initiated trials (IITs) in which the MUCE design is applied to multi-arm phase 2 basket trials. In a three-arm trial targeting different lines of therapy, the MUCE design results in a 50% sample size reduction when compared to the Simon’s two-stage design.

参考文献

  1. Guo, W., Wang, S. J., Yang, S., Lynn, H., & Ji, Y. (2017a). A Bayesian interval dose-finding design addressing Ockham's razor: mTPI-2. Contemporary clinical trials, 58, 23-33.
  2. Guo, W., Ji, Y., & Li, D. (2018). R-TPI: Rolling Toxicity Probability Interval Design to Shorten the Duration and Maintain Safety of Phase I Trials. Journal of Biopharmaceutical Statistics.
  3. Lyu, J., Ji, Y., Zhao, N., & Catenacci, D. V. (2018). AAA: triple adaptive Bayesian designs for the identification of optimal dose combinations in dual‐agent dose finding trials. Journal of the Royal Statistical Society: Series C (Applied Statistics).
  4. Guo, W., Ji, Y., & Catenacci, D. V. (2017b). A subgroup cluster‐based Bayesian adaptive design for precision medicine. Biometrics, 73(2), 367-377.
  5. Berry, S. M., Broglio, K. R., Groshen, S., & Berry, D. A. (2013). Bayesian hierarchical modeling of patient subpopulations: efficient designs of phase II oncology clinical trials. Clinical Trials, 10(5), 720-734.
  6. Simon R (1989). Optimal two-stage designs for phase II clinical trials, Controlled Clinical Trials 10: 1-10.
  7. Chu, Y., & Yuan, Y. (2018). A Bayesian basket trial design using a calibrated Bayesian hierarchical model. Clinical Trials, 15(2), 149-158.
  8. Neuenschwander, B., Wandel, S., Roychoudhury, S., & Bailey, S. (2016). Robust exchangeability designs for early phase clinical trials with multiple strata. Pharmaceutical statistics, 15(2), 123-134.