数字孪生技术在药物临床研究中的应用进展与监管策略分析

Digital Twin Technology in Clinical Drug Research: Application Progress and Regulatory Strategy Analysis

摘要

人工智能（artificial intelligence，AI）赋能的数字孪生（digital twin，DT）技术作为新兴技术工具，通过创建能够精准模拟真实患者特征的虚拟副本，为药物临床研究提供了新思路。综述系统介绍了 DT 在药物临床研究中的应用进展，DT 技术是以机制驱动的定量药理学建模、数据驱动的 AI 建模为核心实现方法，在临床研究中可应用于试验方案优化、治疗反应预测、虚拟对照组构建及受试者招募辅助等核心场景。通过分析和比较全球主要监管机构的相关监管政策，指出全球监管呈现“方向趋同、细节本土化”的特点，倡导基于风险的监管框架。同时指出当前 DT 技术应用于药物临床试验仍面临数据质量管控、模型可信度评估、监管规范制定等多重挑战，并提出了实际应用中的可行建议，对推动 DT 在药物临床研究中的规范化应用与监管策略的制定具有重要参考意义。总体而言，DT 技术在提升临床研发效率、支持罕见病与儿科用药等特殊领域研究方面具有显著潜力，但现阶段仍仅适合作为探索性研究中辅助试验设计与决策的工具，尚无法完全替代确证性临床研究。

Abstract

Artificial intelligence (AI)-enabled digital twin (DT) technology, as an emerging technological tool, creates virtual replicas of real patients by accurately simulating their characteristics, offering new approaches to clinical drug research. This review systematically describes advances in the application of DT in clinical drug research: DT technology uses mechanism-driven quantitative pharmacology modeling and data-driven AI modeling as its core implementation methods, and can be applied to key scenarios such as trial protocol optimization, treatment response prediction, construction of virtual control groups, and assistance in participant recruitment. This review further analyzes and compares the relevant regulatory policies of major global regulatory agencies, points out that global regulation is characterized by "convergence in direction and localization in detail," and advocates a risk-based regulatory framework. It also notes that the current application of DT technology in clinical drug trials still faces multiple challenges, including data quality control, model credibility assessment, and the formulation of regulatory standards, and it proposes feasible recommendations for practical application. This review provides important references for promoting the standardized application of DT in drug clinical research and informing the formulation of regulatory strategies. Overall, DT technology holds significant potential for improving clinical development efficiency and supporting research in special fields such as rare diseases and pediatric medicines; however, at the current stage, it remains only suitable as a tool to assist in trial design and decision-making in exploratory research and cannot yet fully replace confirmatory clinical trials.

关键词

数字孪生 / 人工智能 / 药物研发 / 临床研究 / 监管策略

Key words

digital twin / artificial intelligence / drug development / clinical research / regulatory strategy

参考文献

[1] Zhang K, Zhou H Y, Baptista-Hon D T,et al. Concepts and applications of digital twins in healthcare and medicine[J].Patterns, 2024, 5(8): 101028.

[2] Katsoulakis E, Wang Q, Wu H,et al. Digital twins for health: a scoping review[J].npj Digit Med, 2024, 7: 77.

[3] Elias M, König M. A digital twin of glimepiride for personalized and stratified diabetes treatment[J].Front Pharmacol, 2025, 16: 1686415.

[4] Gonnabathula P, Li M, Nagumalli S K,et al. Applications of PBPK modeling to estimate drug metabolism and related ADME processes in specific populations[J].Pharmaceutics, 2025, 17(9): 1207.

[5] Ezuruike U, Zhang M, Pansari A,et al. Guide to development of compound files for PBPK modeling in the Simcyp population-based simulator[J].CPT Pharmacometrics Syst Pharmacol, 2022, 11(7): 805-821.

[6] Dette C M U, Alberg V, Rüdesheim S,et al. Advancing rare disease therapeutics through digital twins: opportunities in drug development and precision dosing[J].Comput Struct Biotechnol J, 2025, 28: 592-608.

[7] Kaddi C, Tao M, Bergeler S,et al. Quantitative systems pharmacology-based digital twins approach supplements clinical trial data for enzyme replacement therapies in pompe disease[J].Clin Pharmacol Ther, 2025, 117(2): 579-588.

[8] 赵宸, 李改玲, 王亚宁. 新药研发中的定量系统药理学(QSP)模型与虚拟临床试验: 发展及前沿应用[J].药学学报, 2023, 58(11): 3296-3310.

[9] McComb M, Bies R, Ramanathan M. Machine learning in pharmacometrics: opportunities and challenges[J].Br J Clin Pharmacol, 2022, 88(4): 1482-1499.

[10] Richardson S, Irurzun Arana I, Nowojewski A,et al. A machine learning approach to population pharmacokinetic modelling automation[J].Commun Med, 2025, 5: 327.

[11] Unlearn. AI, Inc. PROCOVA™ handbook for the target trial Statistician: version 1.0[EB/OL]. (2021-12-29) [2026-06-01].https://www.ema.europa.eu/documents/other/procovatm-handbook_en.pdf.

[12] European Medicines Agency. Qualification opinion for Prognostic Covariate Adjustment (PROCOVA™) [EB/OL]. (2022-09-20) [2026-08-01].https://www.ema.europa.eu/en/documents/regulatory-procedural-guideline/qualification-opinion-prognostic-covariate-adjustment-procovatm_en.pdf.

[13] European Medicines Agency. Overview of comments received on the draft qualification opinion for Prognostic Covariate Adjustment (PROCOVA™): with responses[EB/OL]. (2022-09-20) [2026-06-01].https://www.ema.europa.eu/en/documents/comments/overview-comments-received-prognostic-covariate-adjustment-procovatm-responses_en.pdf.

[14] Pellizzari E, Cappon G, Sparacino G,et al. Digital-twin based optimization of bolus insulin dosing in pediatric type 1 diabetes: an in silico feasibility study^* [C]//2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC), 2025: 1-5.

[15] Deiteren A, Krupka E, Bontinck L,et al. A proof-of-mechanism trial in asthma with lunsekimig, a bispecific NANOBODY molecule[J].Eur Respir J, 2025, 65(4): 2401461.

[16] Li G, Chen Y B, Peachey J. Construction of a digital twin of chronic graft vs. host disease patients with standard of care[J].Bone Marrow Transplant, 2024, 59(9): 1280-1285.

[17] Bayer M. Digital twin technology may serve as SOC control in GVHD trial[J].Targeted Ther Oncol, 2024, 13(11): 54.

[18] U. S. Food and Drug Administration. Digital health technologies for remote data acquisition in clinical investigations[EB/OL]. (2023-12-22) [2026-06-01]. https://www. fda. gov/regulatory-information/search-fda-guidance-documents/digital-health-technologies-remote-data-acquisition-clinical-investigations.

[19] U.S. Food and Drug Administration. Assessing the credibility of computational modeling and simulation in medical device submissions: guidance for Industry and Food and Drug Administration Staff[EB/OL]. (2023-11-16) [2026-06-01].https://www.fda.gov/regulatory-information/search-fda-guidance-documents/assessing-credibility-computational-modeling-and-simulation-medical-device-submissions.

[20] U.S. Food and Drug Administration. Considerations for the use of artificial intelligence to support regulatory decision-making for drug and biological products[EB/OL]. (2025-01-06)[2026-06-01].https://www.fda.gov/regulatory-information/search-fda-guidance-documents/considerations-use-artificial-intelligence-support-regulatory-decision-making-drug-and-biological.

[21] U. S. Food and Drug Administration. Using artificial intelligence & machine learning in the development of drug & biological products: discussion paper and request for feedback[EB/OL]. (2023-05-11) [2026-06-01].https://www.fda.gov/media/167973/download.

[22] European Medicines Agency. Reflection paper on the use of artificial intelligence (AI) in the medicinal product lifecycle[EB/OL]. (2024-09-09)[2026-06-01].https://www.ema.europa.eu/system/files/documents/scientific-guideline/reflection-paper-use-artificial-intelligence-ai-medicinal-product-lifecycle-en.pdf.

[23] 中华人民共和国工业和信息化部, 商务部, 国家卫生健康委员会, 等. 工业和信息化部等七部门关于印发《医药工业数智化转型实施方案（2025-2030年）》的通知[EB/OL]. (2025-04-03) [2026-06-01].https://www.gov.cn/zhengce/zhengceku/202504/content_7020857.htm.

[24] 国家药品监督管理局医疗器械技术审评中心. 人工智能医疗器械注册审查指导原则[EB/OL]. (2022-03-09)[2026-06-01].https://www.cmde.org.cn/flfg/zdyz/zdyzwbk/20220309091014461.htm.

[25] 国家药品监督管理局药品审评中心. 真实世界证据支持药物研发与审评的指导原则（试行）[EB/OL]. (2020-01-07) [2026-06-01].https://www.cde.org.cn/zdyz/domesticinfopage?zdyzIdCODE=db4376287cb678882a3f6c8906069582.

[26] 国家药品监督管理局药品审评中心. 用于产生真实世界证据的真实世界数据指导原则（试行）[EB/OL]. (2021-04-15) [2026-06-01].https://www.cde.org.cn/zdyz/domesticinfopage?zdyzIdCODE=7d2e46cea0e459358257760383526e9d.

[27] 国家药品监督管理局. 国家药监局综合司关于印发药品监管人工智能典型应用场景清单的通知[EB/OL]. (2024-06-18) [2026-06-01].https://www.nmpa.gov.cn/xxgk/fgwj/gzwj/gzwjzh/20240618144318144.html.

[28] U. S. Food and Drug Administration, European Medicines Agency. Guiding Principles of Good AI Practice in Drug Development[EB/OL]. (2026-01-14) [2026-06-01].https://www.fda.gov/media/189581/download.

[29] Pammi M, Shah P S, Yang L K,et al. Digital twins, synthetic patient data, andin-silico trials: can they empower paediatric clinical trials?[J].Lancet Digit Health, 2025, 7(5): 100851.

基金资助

药品监管科学全国重点实验室课题(2024SKLDRS0230)