AI, Integrated Models and Decentralized Designs Reshape Clinical Trial Execution

Clinical trial sponsors are increasingly turning to AI-enabled design tools, integrated operational models, and decentralized trial elements to accelerate drug development while controlling costs. FDA and EMA guidance confirms that decentralized activities are accepted when carefully planned, but sponsors remain accountable for safety, data quality, and compliance. New modeling also shows that merging clinical and manufacturing operations can raise program value in cardiovascular outcomes trials.

New lifecycle modeling from the Tufts Center for the Study of Drug Development examined a large set of Phase III cardiovascular outcomes trials (CVOTs) to compare traditional, multi-vendor execution against a single, integrated operating model spanning clinical research and manufacturing. The findings showed that integration was consistently linked to meaningfully higher expected program value and strong returns relative to implementation costs, holding up even under more conservative assumptions. Closing coordination gaps between clinical execution and supply functions, and proactively managing event accrual, can shorten timelines, protect budgets, and preserve program value.

Amgen is embedding artificial intelligence into every stage of its drug development continuum. Its Center for Design and Analysis uses digital twins—AI-enabled, patient-level predictive models—that generate personalized synthetic controls for each enrolled patient using historical trial and real-world data. In some settings, these methods may help reduce reliance on traditional enrollment, subject to study design and regulatory considerations. Amgen's Global Development Operations is also integrating AI-powered platforms to improve study start-up, data oversight, and real-time insights, while proactively identifying and managing risk earlier. The same AI-enabled design approaches are applied to drug discovery, using protein and chemistry language models to design molecules with intended therapeutic properties built in from the start.

The FDA and EMA have each issued or updated guidance on decentralized clinical trials. The FDA's final guidance confirms that decentralized elements can be incorporated into appropriate trials but emphasizes that decentralization does not change the sponsor's core obligations. The EMA's guidance is broadly aligned but places greater emphasis on GDPR and Member State law, meaning DCT models may need to be tailored jurisdiction by jurisdiction for multinational EU studies. Sponsors should review contracts and informed consent forms, map decentralized activities early, and conduct targeted risk assessments.

In cardiometabolic trials, hard outcomes such as cardiovascular events, strokes, and mortality remain the gold standard, but they require lengthy schedules and large patient populations. Surrogate endpoints can provide earlier signals, though regulators' acceptability differs. Adaptive designs allow interim analyses and changes, and active comparator trials are gaining popularity as placebo arms become less acceptable ethically and operationally. Statistical innovation now allows indirect placebo comparisons by integrating new active-comparator data with older placebo data. Assay planning should be incorporated into trial design from the start to provide early proof-of-concept signals.

Small clinical operations teams are also executing complex multi-trial programs. Step Pharma's three-person clinical operations group is simultaneously running three clinical programs for dencatistat, a first-in-class CTPS1 inhibitor, in lymphoma, solid tumors, and essential thrombocythemia. Emerging data from the lymphoma study identified a dose-dependent reduction in platelets, enabling the team to quickly expand into essential thrombocythemia, a rare clonal blood disorder where platelet levels are raised.

Broader infrastructure remains a challenge. Most clinical research in the U.S. remains fragmented and reliant on separate data, measurement, and process requirements that do not fit well into routine care, limiting patient access and slowing evidence generation. A draft framing report proposes policy initiatives to advance aligned, modernized standards for clinical development across regulatory agencies. Meanwhile, intuitive technology design can reduce errors and improve site compliance, and vaccine developers can extend follow-up timelines by combining traditional site visits with real-world data collection through clinical trial tokenization.