Traditional clinical trials are often easier to picture as fixed designs. The protocol defines the population, treatment arms, endpoints, and analysis plan before the study begins. Patients are enrolled into that structure, data are collected, and the trial continues until the planned analysis is complete.
That model creates clarity. It also has limits. When each treatment question requires a separate trial, the research process can become slow, repetitive, and difficult to adapt as evidence changes.
Adaptive and platform trials changed that expectation. They introduced the idea that a study could be designed to learn while it was still running. Treatment arms could be assessed at planned points. Ineffective approaches could be stopped. Promising comparisons could continue. New research questions could be added without starting again from zero (Sydes et al., 2012).
The important point is not that adaptive trials are less structured. They require more structure, because the study is expected to change without losing scientific or operational control.
The STAMPEDE trial
A strong example is STAMPEDE, a multi-arm, multi-stage platform trial in prostate cancer that opened in 2005 (James et al., 2008; National Institute for Health and Care Research, 2021).
STAMPEDE was designed to test several treatment approaches within one ongoing trial structure. It began with one control arm and multiple research arms, creating a way to evaluate more than one comparison over time. Research arms could be stopped if they did not show enough benefit. Other arms could continue. New arms could be added as the treatment landscape changed (National Institute for Health and Care Research, 2021; Sydes et al., 2012).
That made STAMPEDE an important example of a trial designed to move.
The movement was not casual. It was built into the design. The trial had rules for when arms could continue, stop, or be added. That is what makes it useful for this series. It shows that complexity is not always something that happens to a study after launch. Sometimes complexity is part of the model from the beginning.
What the design changed
STAMPEDE changed the way we can think about trial execution.
In a fixed trial, the operational model is built around a stable set of arms, pathways, and assumptions. In a platform trial, the structure has to support change while the study continues. The trial may still have strict rules, but those rules have to account for movement.
That changes the questions study teams have to answer.
What happens when an arm closes? What happens when a new arm opens? Which patients are eligible for which comparison? What treatment options should a site be able to assign today? What supply is required now, not six months ago? What should be visible in reporting? What needs to be documented so the study remains traceable?
These questions are not administrative details. They are part of how the trial protects its integrity while it adapts.
The historical lesson is practical: adaptive movement only works when the movement is governed.
What changes when the trial is designed to move
Not every modern trial is a platform trial, but many borrow from the same logic. A study may include adaptive cohorts, expansion arms, dose decisions, biomarker-defined groups, changing randomization ratios, country-specific requirements, or amendment-driven updates.
Each feature can help the science move more efficiently. Each also adds operational dependencies.
When an arm closes, sites need to know what can still be assigned. When a new arm opens, eligibility and treatment pathways may need to change. When a comparison pauses, the system needs to prevent the wrong action at enrollment or dispensing. When randomization logic changes, the current rules need to be clear, controlled, and traceable.
The study may be designed to move, but the controls still have to hold.
This is where adaptive execution becomes difficult. The challenge is not only deciding what should change. It is making sure the approved change is reflected consistently across the systems, sites, supply processes, and reports that keep the study running.
RTSM impact
In an adaptive or platform trial, RTSM has to keep the live study aligned to the approved version of the design.
When an arm stops, that change affects more than randomisation. It can affect what sites can assign, what treatment is available, what inventory is expected to move, and what reports should show. When a new comparison is added, the same logic applies in reverse. Eligibility, assignment, dispensing, supply, reporting, and audit trail may all need to change together.
The risk is version drift. Sites, supply teams, reports, and system logic cannot operate from different versions of the study.
For RTSM, the requirement is controlled change across connected dependencies: assignment logic, arm or cohort status, treatment availability, dispensing rules, inventory visibility, reporting, and traceability.
Movement still needs rules
STAMPEDE gives us a useful historical marker because it shows a trial designed around controlled movement. The model was not flexible in a loose sense. It was adaptive because the structure defined how and when movement could occur (Sydes et al., 2012).
That distinction matters for modern RTSM.
Flexibility alone is not enough. A system that supports adaptive execution has to manage change with discipline, so the study can evolve without creating confusion at sites, gaps in supply, or uncertainty about what was allowed at a given point in time.
Adaptive trials made movement part of the model. In modern study execution, that movement has to be governed, visible, and controlled.
References
James, N. D., Sydes, M. R., Clarke, N. W., Mason, M. D., Dearnaley, D. P., Anderson, J., Popert, R. J., Sanders, K., Morgan, R. C., Stansfeld, J., Dwyer, J., Masters, J., & Parmar, M. K. B. (2008). STAMPEDE: Systemic therapy for advancing or metastatic prostate cancer: A multi-arm multi-stage randomised controlled trial. Clinical Oncology, 20(8), 577-581. https://doi.org/10.1016/j.clon.2008.07.002
National Institute for Health and Care Research. (2021). STAMPEDE: A turning point in trial design.
Sydes, M. R., Parmar, M. K. B., Mason, M. D., Clarke, N. W., Amos, C., Anderson, J., de Bono, J., Dearnaley, D. P., Dwyer, J., Green, C., Jovic, G., Ritchie, A. W. S., Russell, J. M., Sanders, K., Thalmann, G., & James, N. D. (2012). Flexible trial design in practice: Stopping arms for lack-of-benefit and adding research arms mid-trial in STAMPEDE: A multi-arm multi-stage randomised controlled trial. Trials, 13, 168. https://doi.org/10.1186/1745-6215-13-168