Steger-Hartmann Thomas, Sanz Ferran, Bringezu Frank, Soininen Inari
Bayer AG, Pharmaceuticals, Berlin, Germany.
Universitat Pompeu Fabra, Hospital del Mar Research Institue, Barcelona, Spain.
Toxicol Pathol. 2025 Feb;53(2):230-233. doi: 10.1177/01926233241303906. Epub 2024 Dec 12.
The virtual control group (VCG) concept was originally developed in the IMI2 project eTRANSAFE, using data of control animals which pharmaceutical companies have accrued over decades from animal toxicity studies. This control data could be repurposed to create virtual control animals to reduce or replace concurrent controls in animal studies. Initial work demonstrated the general feasibility of the VCG concept, but implementation requires significant further collaborative efforts. The new Innovative Health Initiative (IHI) project VICT3R aims to address these challenges and to obtain regulatory acceptance for the VCG concept. To achieve these goals, VICT3R will build a database comprising high-quality, standardized, and duly annotated control animal data from past and forthcoming toxicity studies. The VICT3R project will create workflows and computational tools to generate adequate VCGs based on statistical and artificial intelligence (AI) approaches. The validity, reproducibility, and robustness of the resulting VCGs will be assessed by comparing the performance of their use with that of real control groups.
虚拟对照组(VCG)概念最初是在IMI2项目eTRANSAFE中提出的,它利用了制药公司几十年来从动物毒性研究中积累的对照动物数据。这些对照数据可被重新利用,以创建虚拟对照动物,从而减少或取代动物研究中的同期对照。初步工作证明了VCG概念的总体可行性,但实施需要进一步开展大量合作。新的创新健康计划(IHI)项目VICT3R旨在应对这些挑战,并使VCG概念获得监管认可。为实现这些目标,VICT3R将建立一个数据库,该数据库包含来自过去和即将进行的毒性研究的高质量、标准化且经过适当注释的对照动物数据。VICT3R项目将创建工作流程和计算工具,以基于统计和人工智能(AI)方法生成合适的虚拟对照组。通过比较虚拟对照组与真实对照组的使用性能,评估所得虚拟对照组的有效性、可重复性和稳健性。