Improving Dose-Finding for New Agents as Monotherapy and Add-On Therapy

With the recent advancement in immuno-oncology, next-stage early oncology development is focusing on identifying the best combinations of established immunotherapies with new agents to either overcome drug resistance or achieve synergic effects. Although the combination is the focus, safety profile of the new agent alone must be explored. Therefore, many trials have both monotherapy and add-on combination therapy arms. Finding the maximum tolerable dose (MTD) for the new agent in both arms is critical. Traditional oncology dose-finding methods and MTD estimation algorithm do not handle the correlation and interplay between the two arms and the selected MTDs may contradict with each other. To overcome these issues, we applied a two-dimensional pool-adjacent-violators algorithm to MTD estimation and modified the standard Bayesian optimal interval design (BOIN) to allow for information flow between arms during dose-finding. We also showed that a naïve adaptation of standard BOIN that is much simpler to implement demonstrated empirically similar performance. These new approaches were assessed with simulations and demonstrated improvement for trials with both monotherapy and add-on combination therapy arms. Albeit proposed in the context with immunotherapy as the backbone drug, our approaches can be applied to any new agent in combination with a fixed dose of another drug. Supplementary materials for this article are available online.

随着肿瘤免疫治疗（immuno-oncology）领域的近期进展，下一代早期肿瘤学研发正聚焦于探索已获批免疫疗法与新型药剂的最优联合方案，以克服耐药性或实现协同效应。尽管联合治疗是当前研发的核心方向，但仍需单独评估新型药剂的安全性特征。因此，诸多临床试验同时设置了单药治疗组与联合治疗追加组。在这两类试验组中确定新型药剂的最大耐受剂量（maximum tolerable dose, MTD）是关键环节。传统肿瘤学剂量探索方法与MTD估计算法无法处理两组间的相关性与相互作用，由此得到的MTD选择可能相互矛盾。为解决上述问题，我们将二维相邻违反合并算法（two-dimensional pool-adjacent-violators algorithm）应用于MTD估计，并对标准贝叶斯最优区间设计（Bayesian optimal interval design, BOIN）进行改良，以实现剂量探索阶段两组间的信息流转。此外，我们还证实，一种实现难度更低的标准BOIN朴素适配方案在实证中表现出相近的性能。我们通过模拟试验对上述新方法进行了评估，结果显示其可有效改善同时包含单药治疗组与联合治疗追加组的临床试验表现。尽管本方法是基于以免疫疗法为基础用药的场景提出的，但同样可应用于其他固定剂量药物与新型药剂的联合研究。本文的补充材料可在线获取。