PS41: Revolutionizing Clinical Trials with Digital Twins, In Silico control, RWD, or Synthetic Patient Data

OVERVIEW This session will explore the transformative impact of digital twins (DTs), in silico trials (ISTs), real world data (RWD) or synthetic patient data on clinical research, drug and device development, and patient care. By harnessing computational modeling, simulation, and advanced data generation techniques, these approaches are revolutionizing how medical interventions are evaluated, optimized, and personalized. The session will feature expert presentations and real-world case studies, with special emphasis on regulatory and ethical considerations, as well as strategies for overcoming current challenges. BACKGROUND & MOTIVATION Over the past decade, major initiatives such as the Avicenna Alliance, FDA, EMA, and the Virtual Physiological Human Initiative have laid the foundation for the adoption of DTs, ISTs, RWD or synthetic patient data in biomedical research. These innovations are particularly valuable in pediatric and rare disease randomized controlled trials, which often lack statistical power due to small sample sizes and recruitment challenges. They offer promising solutions to issues that traditional trials may face, such as early discontinuation, missing data, high costs for long-term follow-up, and the exposure of vulnerable populations to unproven therapies. The rise of DTs, ISTs, RWD or synthetic patient data has been propelled by advances in computational modeling, simulation, and the integration of real-world data-including clinical, imaging, and genomic information. These methods aim to accelerate the evaluation of medical interventions, reduce reliance on traditional clinical trials, and lower costs. By minimizing the need to expose real patients to experimental treatments, especially in vulnerable populations, these technologies enable more tailored predictions and treatment strategies by modeling individual or group responses. Regulatory agencies such as the FDA and EMA are increasingly recognizing these tools as valuable for evidence generation and decision-making. VIRTUAL REPRESENTATIONS Digital Twins (DTs) are highly detailed, dynamic virtual replicas of individual patients, organs, or biological systems, continuously updated with real patient data. They enable personalized medicine by simulating and predicting individual patient responses to interventions in real time. Traditionally, DTs have been used to simulate control arms in clinical trials, reducing the need for real patients in these groups. This is especially valuable in pediatric studies, where finding suitable controls is difficult. Early studies, such as microbiome modeling in preterm infants using generative AI, demonstrate the potential of DTs to capture dynamic biological interactions and predict outcomes [1]. In Silico Trials (ISTs) are clinical trials conducted entirely via computer simulation and modeling, often using virtual populations. They allow for the testing and optimization of medical interventions, devices, or drugs before or alongside real-world trials, reducing the need for human or animal subjects. ISTs can simulate both control and intervention arms, bypassing recruitment and ethical challenges of traditional trials. While they complement real-world clinical trials, they do not fully replace them. Notable examples include virtual imaging trials and interventions for rare diseases. Synthetic Patient Data is artificially generated using AI or mathematical models to mimic real patient data without exposing personal information. It supplements small datasets, particularly in rare diseases, or when privacy concerns prevent the use of real patient data. Adult studies in Alzheimer's disease and myelodysplasia have shown the feasibility of this approach [2]. However, for implementation, quality, diversity, and generalizability of synthetic data must be ensured. While RWD has been extensively explored and discussed in literature, the further adoption of DTs, ISTs, and synthetic patient data brings significant benefits to medical research and healthcare. These technologies dramatically reduce trial time and costs, streamline the design and validation of medical products, and decrease reliance on animal testing. They also provide more definitive quantitative endpoints for studies and support personalized, optimized patient care by allowing for tailored predictions and treatment strategies. The FDA and EU are actively developing frameworks for digital health and AI in clinical research [3, 4]. Key regulatory concerns include data privacy, security, validation, and the inability to predict unknown adverse events. Ethical challenges encompass consent, data ownership, risk of dehumanization, and maintaining trust. Recent milestones include the establishment of the FDA Digital Health Center of Excellence and the passage of the EU Artificial Intelligence Act in 2024. CHALLENGES & PRIORITY ACTIONS Despite the significant promise of DTs, ISTs, RWD or synthetic patient data, several challenges must be addressed to realize their full potential. Key obstacles include the need for a unified vision and collaboration among diverse stakeholders, advancements in modeling, validation, and integration of varied data types, and the development of workforce training, user engagement, and standardized protocols. Ethical considerations such as ensuring data privacy and patient ownership are critical. To advance the field, it is essential to advocate for multidisciplinary collaboration, develop standards and harmonize efforts across regulators and academia, foster industry uptake and professional development, and actively engage patients and the public to build trust. Addressing these priorities will help overcome current barriers and accelerate the responsible adoption of these transformative technologies in healthcare. WHY THIS SESSION? Given these challenges and the urgent need for practical solutions, we propose the session "Revolutionizing Clinical Trials with In Digital Twins, Silico control, RWD or Synthetic Patient Data" to bring together leading experts and stakeholders. This session addresses a rapidly emerging frontier in biomedical research -- one that is fundamentally changing how clinical trials and patient care are approached. By featuring expert presentations and real-world case studies, this session will provide a platform to share the latest advances, best practices, and lessons learned in this new and evolving field. Special emphasis on regulatory and ethical considerations will help attendees navigate the shifting landscape and ensure responsible adoption. Ultimately, this session aims to foster collaboration and discussion, inspire innovation, and equip participants with actionable strategies to overcome current barriers and accelerate progress in clinical research. REFERENCES [1] Sizemore N, Oliphant K, Zheng R, Martin CR, Claud EC, Chattopadhyay I. A digital twin of the infant microbiome to predict neurodevelopmental deficits. Sci Adv 2024; 10: eadj0400. [2] Bertolini D, Loukianov AD, Smith A, et al. Forecasting progression of mild cognitive impairment (MCI) and Alzheimer's disease (AD) with digital twins. Alzheimers Dement 2021; 17 (suppl 9): e054414. [3] US Food and Drug Administration. Digital Health Center of Excellence. 2024. https://www.fda.gov/medical-devices/digital-health-center-excellence [4] US Food and Drug Administration. CFDARH. Assessing the credibility of computational modeling and simulation in medical device submissions. Docket number: FDA-2021-D-0980. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/assessing-credibility-computational-modeling-and-simulation-medical-device-submissions