entimICE FastTrack
entimICE ValoRpacks
entimICE EverKeep
Real-time clinical trials
The FDA announced plans for speeding up the clinical trial process by enabling real-time transmission of data.
In late April 2026 the FDA unveiled a pilot for real-time clinical trials (RTCTs). Two early studies – AstraZeneca’s TRAVERSE (Phase 2 lymphoma) and Amgen’s STREAM-SCLC (Phase 1b lung cancer) – are already streaming pre-specified efficacy and safety endpoint data directly to FDA reviewers “in real time” as data accrue. In these proof-of-concept trials, the FDA has established cloud-based data feeds via Paradigm Health and has “received and validated” ongoing signals from the sites. The goal is to collapse the long lag time between phases and move toward continuous trials. As former FDA Commissioner Marty Makary noted, roughly 45% of the time between Phase I and submission is currently “dead time” waiting on analysis and paperwork. Real-time trials aim to cut that dramatically by letting FDA scientists see, e.g., tumor responses or fevers as they happen. “When a patient develops a fever, or a tumor shrinks, FDA regulators can see in the cloud, in real time, exactly what is happening,” said Makary. An FDA Request for Information (RFI) has just opened to design a broader AI-enabled pilot this summer; comments are due by May 29.
- Goals: Eliminate the hiatus between phases by sharing endpoints continuously. Shorten review timelines (“shave years off approval,” according to one FDA official) without compromising safety.
- How: Use AI and cloud tools to stream key data. The FDA’s Chief AI Officer Jeremy Walsh emphasizes partnership: “real-time trials are not just technically possible; they are transformative for the entire clinical ecosystem”.
- Status: Two current cancer trials to show feasibility. The FDA plans to invite more sponsors into a pilot by late summer.
Observers caution that this shift raises new statistical and governance questions. Rolling data feeds mean multiple looks at outcomes. In particular, continuous AI models can “drift” an endpoint definition mid-trial, undermining the statistical plan. Every AI update needs formal versioning: log each model change, document its effect on the primary endpoint, and adjust the analysis plan accordingly. Otherwise the data FDA sees today may be based on a different algorithm than tomorrow’s. Legally, the FDA assumes a quasi-DSMB role here. If the FDA catches an early safety signal it could intervene, so the roles of sponsor, CRO, FDA and (if present) DSMB must be clearly defined. In short, novel SOPs will be needed for data sharing, and SCEs must be able to audit every streamed update against the trial’s master protocol.
UK Trial Reforms: Meanwhile, in the UK new CT laws (April 2026) enact the most sweeping change in more than 20 years. Highlights: introducing “notifiable trials” for lower-risk studies. Such trials get automatic MHRA approval as long as ethics clearance is in place, greatly speeding startup. Safety reporting rules were relaxed: sponsors are no longer required to report SUSARs directly to sites and IRBs, and urgent safety measures can be notified within seven days (vs. the old three-day deadline). MHRA has issued draft GCP guidance in parallel. Altogether the UK changes aim to cut red tape and boost research competitiveness. (MHRA grew staff by 20% and cleared its Brexit backlog in preparation.) These reforms, like the FDA’s, push toward a more adaptive, risk-based framework – for example, by allowing earlier, streamlined trial starts for simpler studies.
Implications & recommendations for SCEs
Entimo AG has provided GxP-compliant Clinical Data Repositories (CDR) and Statistical Computing Environments (SCE) to the pharmaceutical industry for more than two decades, enabling sponsors to excel in submissions to regulatory authorities towards timely market approvals. During the COVID pandemic many vaccine trials where processed with entimICE solutions, exercising a much more frequent data exchange with the FDA than ever before. From this experience, the authorities and industry derived what can be done to fast-track submissions to bring medicines to patients faster, and entimo has a good understanding of what is required for software vendors to support this acceleration.
These shifts will have direct impact on statistical computing environments (SCEs) in industry and CROs. To handle real-time/continuous data and more adaptive designs, modern SCEs like entimICE FastTrack should be prepared as follows:
- Cloud-based, automated pipelines: Ensure the SCE can ingest streaming data feeds. Modern SCE platforms already allow scheduling of analyses to run automatically when new data arrive. Partner with IT/EDC teams to connect site databases (or a hub) into the SCE data repository in near-real-time.
- Robust version control and audit trails: Use strict versioning for all analysis code, datasets, and AI models. Every change, whether a new data lock or an AI model update, must be logged with an identifier and timestamp. A compliant SCE already offers unified version tracking of programs and data; extend this to cover real-time data events (e.g., log which data snapshot corresponds to each analysis run).
- Dynamic analysis planning: Be ready to pre-specify adaptive rules. For example, if conducting interim looks on a live endpoint, pre-define how often analyses occur. If using AI-derived endpoints, plan how to validate or “freeze” those models during the trial. The SCE should facilitate quick sensitivity analyses: e.g., re-run results under different definitions or excluding recent updates, to check robustness.
- Enhanced quality controls: Real-time monitoring requires continuous QC. Build automated checks for data quality and consistency (missing data rates, outliers) as data stream in. For example, modern SCE tools can attach checkpoints to alert statisticians if an unusual data shift occurs. Also ensure all data transformations are reproducible: metadata and derivations should accompany raw values.
- Regulatory compliance and security: Confirm that the SCE meets 21 CFR 11 and other GxP requirements. Real-time feeds mean the SCE will partially parallel FDA’s oversight. Ensure roles/permissions are clear (which user can see what data and actions). Encrypt data in motion to the FDA and maintain logs of what was transmitted. Update standard operating procedures for audits. Inspectors will likely want to see how real-time data were handled.
- Team training and collaboration: Prepare biostatisticians, data scientists, and programmers for this model. They may need new skills in cloud analytics, streaming data handling, and adaptive methods. Encourage close collaboration among clinical data managers, biostats, and IT. Undertake dry runs or mock real-time scenarios (e.g., feed “live” data into a test SCE) to uncover gaps.
- Leverage modern SCE features: Consider next-generation SCE solutions built for continuous workflows. For example, some SCEs integrate an analysis workflow that re-executes programs on any data update. Others allow role-based dashboards showing key endpoints evolving over time. Tools that provide real-time visualization of accrued data or interim results can help both statisticians and clinicians stay aligned.
By selecting an SCE ready for these challenges, sponsors and CROs can turn these regulatory changes into an advantage. The underlying message from FDA leaders is clear: clinical development should be “unbroken” and data-driven. SCEs that enable live data sharing, continuous QA, and flexible analysis will be well positioned to meet the new paradigm. In summary, treat the FDA’s pilot and UK reforms as a call to modernize your data infrastructure and processes. Embrace cloud platforms, reinforce versioning/audit mechanisms, and revise SOPs for streaming endpoints. In doing so, your SCE will be ready to deliver timely insights in this emerging, adaptive trial era.