Scholarlink

<p>Combination therapies are increasingly prevalent in oncology drug development, yet demonstrating the contribution of each component (CoC) to the overall treatment effect poses unique challenges requiring multidisciplinary expertise including biological understanding of treatment mechanisms, clinical knowledge of disease landscapes, regulatory considerations, and statistical methods for trial design and analysis. This paper explores the current landscape of oncology drug combination development, focusing on the statistical approaches that satisfy regulatory requirements for designing and analyzing late-phase randomized trials to demonstrate CoC. It further considers contribution of phases (CoP), which is related to CoC, with special attention to design and analysis options for perioperative settings.</p><p>While traditional factorial designs provide the strongest evidence for CoC, they may not be practical. In such cases, alternative design strategies including semi-factorial designs, adaptive arm-dropping strategies, or incorporation of external data can be employed. Analytical alternatives such as positive trend analysis, relaxed alpha testing, and composite endpoints may also be worth exploring when a fully powered factorial design is not feasible. For CoP assessment across multiple treatment periods, Sequential Multiple Assignment Randomized Treatment (SMART) designs present a promising alternative to factorial designs. Ultimately, success in CoC and CoP assessment relies on early regulatory engagement, appropriate design and analysis. Continued statistical innovation will be essential to balance thorough CoC and CoP assessment and timely therapy delivery.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessing the Contribution of Components in Late-Phase Oncology Trials: A Roadmap of Key Approaches

  • 期刊论文

  • 发表日期：

    2026年3月7日

查看全文
  • Xiaowen Tian,
  • Kristine Broglio,
  • Di Ran,
  • Jianliang Zhang,
  • Xia Li

摘要

Combination therapies are increasingly prevalent in oncology drug development, yet demonstrating the contribution of each component (CoC) to the overall treatment effect poses unique challenges requiring multidisciplinary expertise including biological understanding of treatment mechanisms, clinical knowledge of disease landscapes, regulatory considerations, and statistical methods for trial design and analysis. This paper explores the current landscape of oncology drug combination development, focusing on the statistical approaches that satisfy regulatory requirements for designing and analyzing late-phase randomized trials to demonstrate CoC. It further considers contribution of phases (CoP), which is related to CoC, with special attention to design and analysis options for perioperative settings.

While traditional factorial designs provide the strongest evidence for CoC, they may not be practical. In such cases, alternative design strategies including semi-factorial designs, adaptive arm-dropping strategies, or incorporation of external data can be employed. Analytical alternatives such as positive trend analysis, relaxed alpha testing, and composite endpoints may also be worth exploring when a fully powered factorial design is not feasible. For CoP assessment across multiple treatment periods, Sequential Multiple Assignment Randomized Treatment (SMART) designs present a promising alternative to factorial designs. Ultimately, success in CoC and CoP assessment relies on early regulatory engagement, appropriate design and analysis. Continued statistical innovation will be essential to balance thorough CoC and CoP assessment and timely therapy delivery.