Department of Biomedical Informatics, The Ohio State University College of Medicine, Columbus, Ohio.
Division of Surgical Oncology, The Arthur G. James Cancer Hospital and Richard J. Solove Research Institute, Comprehensive Cancer Center, The Ohio State University, Columbus, Ohio.
Cancer Res Commun. 2022 Sep 6;2(9):929-936. doi: 10.1158/2767-9764.CRC-22-0160. eCollection 2022 Sep.
In this study, we summarized critical databases of drug combination toxicity and pharmacokinetics. We further conducted a feasibility and utility study that demonstrates how different data sources can contribute to and assist phase I trial designs. Single-drug and drug combination toxicity and pharmacokinetic data were primarily reviewed from several databases. We focused on the MTD, dose-limiting toxicity (DLT), toxicity, and pharmacokinetic profiles. To demonstrate the feasibility and utility of these data sources in improving trial designs, phase I studies reported in ClincalTrials.gov from January 1, 2018 to December 31, 2018 were used as examples. We evaluated whether and how these studies could have been designed differently given toxicity and pharmacokinetic data. None of the existing pharmacokinetic and toxicity databases contain either MTD or DLT. Among 268 candidate trials, four drug combinations were studied in other phase I trials before 2018; 185 combinations had complete or partial information on drug interactions or overlapping toxicity, and 79 combinations did not have available information. Two drug combination trials were selected as case studies. The nivolumab-axitinib trial could have been designed as a dose deescalating study, and the vinorelbine-trastuzumab emtansine trial could have been designed with a lower dose of either drug. Public data sources contain significant knowledge of the drug combination phase I trial design. Some important data (MTD and DLT) are not available in existing databases but in the literature. Some phase I studies could have been designed more efficiently with additional preliminary data.
Prior preclinical and clinical knowledge is critical for designing effective and efficient cancer drug combinatory trials. We reported results on the feasibility and utility of different informatics resources for contributing to and assisting phase I trial designs based on our designed classification approach. We also found that public data sources contained significant knowledge for drug combination phase I trial design, but some critical data elements (MTD and DLT) were missing.
在这项研究中，我们总结了药物组合毒性和药代动力学的关键数据库。我们进一步进行了一项可行性和实用性研究，展示了不同数据源如何为 I 期临床试验设计做出贡献和提供帮助。单药和药物组合的毒性和药代动力学数据主要从几个数据库中进行了审查。我们重点关注最大耐受剂量（MTD）、剂量限制毒性（DLT）、毒性和药代动力学特征。为了展示这些数据源在改善试验设计方面的可行性和实用性，我们以 2018 年 1 月 1 日至 2018 年 12 月 31 日在 ClinicalTrials.gov 上报告的 I 期研究为例。我们评估了在考虑毒性和药代动力学数据的情况下，这些研究是否以及如何可以进行不同的设计。现有的药代动力学和毒性数据库都没有包含 MTD 或 DLT。在 268 项候选试验中，有 4 种药物组合在 2018 年之前已在其他 I 期试验中进行了研究；185 种组合具有药物相互作用或重叠毒性的完整或部分信息，而 79 种组合没有可用信息。选择了两个药物组合试验作为案例研究。nivolumab-axitinib 试验本可以设计为剂量递增研究，vinorelbine-trastuzumab emtansine 试验本可以设计为降低任何一种药物的剂量。公共数据源包含关于药物组合 I 期临床试验设计的重要知识。一些重要数据（MTD 和 DLT）在现有数据库中不可用，但在文献中可用。有了额外的初步数据，一些 I 期研究本可以更有效地进行设计。
在设计有效的癌症药物组合试验时，预先的临床前和临床知识是至关重要的。我们根据设计的分类方法，报告了不同信息学资源对 I 期临床试验设计做出贡献和提供帮助的可行性和实用性研究结果。我们还发现，公共数据源包含了药物组合 I 期临床试验设计的重要知识，但一些关键数据元素（MTD 和 DLT）缺失。
