Before early-phase oncology studies can answer whether a treatment should move forward, they have to answer a more immediate question: how should the next dose be chosen?
That question is clinical, but it is also operational. A study has to define when escalation is allowed, when a dose should pause, when more patients are needed, and when toxicity prevents further movement. Without structure, dose decisions can become inconsistent, delayed, or difficult to defend.
The traditional 3+3 dose-escalation design became familiar because it gave early-phase studies a simple framework for moving through that uncertainty (Le Tourneau et al., 2009; Storer, 1989).
The 3+3 model
The 3+3 design is one of the best-known approaches in Phase I oncology dose escalation. A small group of patients is treated at a starting dose. If no dose-limiting toxicities are observed, the study can move to the next dose level. If a toxicity is observed, additional patients may be enrolled at the same dose. If too many toxicities are observed, escalation stops or the previous dose is considered (Le Tourneau et al., 2009; Storer, 1989).
The model is not sophisticated by modern standards, and it has been widely debated. But its historical value is clear. It created a defined process for making dose decisions in a setting where uncertainty is high and patient safety is central (Le Tourneau et al., 2009; Paoletti et al., 2015).
The important point is not that 3+3 is the ideal model for modern development. It is that this model gave dose movement structure.
What the structure protected
Dose escalation is not simply movement from a lower number to a higher number. Each decision affects which patients can enter the study, which dose level is active, what safety data must be reviewed, and what treatment can be assigned next.
The 3+3 model made those decision points explicit. The study did not move because someone felt ready to move. It moved because predefined conditions had been met.
That distinction matters. Early-phase studies are designed to learn, but they also need to protect patients while learning. A structured escalation model creates discipline around when the study can proceed and when it must stop, wait, or gather more information.
This is where the historical lesson still holds. Dose movement needs rules because dose movement carries risk.
What dose movement affects now
Modern dose escalation and expansion studies often carry far more operational complexity than the traditional 3+3 model was built to describe.
A study may include escalation cohorts, backfill cohorts, expansion cohorts, de-escalation, intra-patient escalation, titration, multiple schedules, combination therapy, or population-specific dose rules. A dose level may open for one cohort and remain closed for another. A cohort may pause while safety data are reviewed. An expansion arm may open only after a decision is made elsewhere in the study. Later literature on Phase I oncology design has also highlighted the challenges that expansion cohorts, targeted agents, combinations, and newer definitions of dose-limiting toxicity create for traditional 3+3 approaches (Paoletti et al., 2015).
The dose decision is still central, but it no longer sits alone.
It can affect cohort status, eligibility rules, treatment assignment, visit schedules, dispensing logic, supply demand, reporting, and amendment planning. One decision may change what sites can do, what patients can receive, and what inventory needs to be available.
That is why modern early-phase execution requires more than general “flexibility.” It requires a system design that understands where movement is likely and which parts of the study are connected to that movement.
RTSM impact
In a dose escalation and expansion study, dose status has to be reflected correctly in the system.
If a dose level is open, the right patients need access to the right pathway. If a dose level is closed, paused, or no longer available, the system needs to prevent assignment. If a cohort expands, the system needs to reflect the approved cohort status, treatment options, dispensing rules, and reporting requirements. If backfill is allowed, the logic needs to distinguish between patients entering the active escalation path and patients entering a previously evaluated dose.
These are not cosmetic updates. They affect what sites can do, what patients can receive, what supply is required, and what the study team can see.
This is why early RTSM planning matters. The study does not need every theoretical scenario built in from day one. It does need the likely movement mapped early enough that change can be handled without unnecessary rebuilds, avoidable delays, or uncontrolled workarounds.
The question is not simply, “Can the system support dose escalation?”
The better question is, “What happens operationally when the dose decision changes?”
The structure still matters
The 3+3 design is useful because it shows the original need clearly: dose movement required rules. Modern early-phase oncology has moved beyond that simple model, but not beyond the need for structure.
The difference is scale.
A dose decision may now affect cohort status, eligibility logic, titration rules, expansion pathways, dispensing, inventory, forecasting, reporting, and amendment planning. One decision can create movement across multiple parts of the study.
That is where RTSM becomes critical. It is the system layer that helps translate approved dose decisions into controlled study actions. It keeps the operational model aligned with what the protocol allows, what the study team has approved, and what sites can safely execute.
Dose movement needed structure in traditional early-phase oncology.
In modern escalation and expansion studies, that structure has to extend across the live study environment.
References
Le Tourneau, C., Lee, J. J., & Siu, L. L. (2009). Dose escalation methods in phase I cancer clinical trials. Journal of the National Cancer Institute, 101(10), 708-720. https://doi.org/10.1093/jnci/djp079
Paoletti, X., Ezzalfani, M., & Le Tourneau, C. (2015). A requiem for the 3 + 3 design for phase I trials. Annals of Oncology, 26(9), 1808-1812. https://doi.org/10.1093/annonc/mdv266
Storer, B. E. (1989). Design and analysis of phase I clinical trials. Biometrics, 45(3), 925-937. https://doi.org/10.2307/2531693