In June 2026, the U.S. Food and Drug Administration (FDA) issued a revised draft guidance, Master Protocols for Drug and Biological Product Development, replacing the draft published in December 2023. The revision incorporates stakeholder feedback and provides more detailed recommendations on the design, analysis, conduct, and regulatory submission of trials conducted under master protocols.

The overall direction has not changed: the FDA continues to view well-designed master protocols as an important way to generate reliable evidence more efficiently. What has changed is the level of detail. The 2026 revision provides greater clarity on basket trials, shared controls, randomization, blinding, and the regulatory infrastructure needed to support complex master protocols.

Master protocols: One framework, multiple questions

Master protocols allow multiple research questions to be addressed within a single overarching trial infrastructure. Depending on the design, sponsors may be able to share protocol elements, sites, operational procedures, oversight, data collection processes, and control groups.

The FDA recognizes three important types:

• Umbrella trials evaluate multiple interventions within a single indication.
• Basket trials evaluate an intervention across multiple indications.
• Platform trials evaluate multiple interventions within an ongoing framework in which treatment arms can enter or leave over time.

These designs can create substantial efficiencies compared with conducting a series of independent trials. However, those efficiencies introduce statistical, operational, and regulatory complexities that must be addressed prospectively. The 2026 revision provides more direction on how the FDA expects sponsors to manage those tradeoffs.

Basket trials move into the spotlight

One of the most notable differences between the 2023 and 2026 drafts is the FDA’s expanded treatment of basket trials.  The 2023 draft concentrated primarily on randomized umbrella and platform trials. The revised guidance more clearly incorporates basket trials and provides additional recommendations for their design and analysis.

This is particularly relevant to biomarker-driven and precision medicine development, where a therapy may provide benefit across diseases or disease subtypes linked by a common molecular or biological characteristic. For development teams, the implication is important: the scientific rationale for combining populations and the statistical strategy for evaluating treatment effects across them should be considered early in development rather than treated as an analysis stage question. For statisticians, the important change is not simply that basket trials are now explicitly covered. The revised guidance puts greater regulatory attention on the assumptions that allow information to be combined or borrowed across populations and on demonstrating that the resulting design has acceptable operating characteristics.

The revised guidance also addresses the use of statistical methods that borrow information across diseases or disease subtypes within a basket trial. Such approaches can improve efficiency, particularly when individual populations are small, but their validity depends on the degree to which treatment effects can reasonably be assumed to be similar across populations. Sponsors should therefore justify the assumptions underlying any borrowing strategy and evaluate how the approach performs when those assumptions are not met. This makes the assessment of heterogeneity across populations an important part of both trial design and statistical planning.

Shared controls: Efficiency comes with conditions

Shared control groups are one of the major potential advantages of umbrella and platform trials. They can reduce the number of participants assigned to control and improve trial efficiency. However, the FDA’s revised guidance reinforces an important principle: not all control participants are equally informative for every treatment comparison.

For primary treatment comparisons, the FDA emphasizes the importance of concurrently eligible controls, that is, control participants who could have been randomized to the investigational treatment during the same period. This matters particularly in long-running platform trials. Standards of care, patient characteristics, site mix, diagnostic practices, and background therapies may change over time. Including nonconcurrent controls may improve statistical precision, but it can also introduce temporal bias.

The practical message is significant: a larger shared control dataset does not automatically produce stronger evidence. The relevance and comparability of those controls are just as important as their number. For registrational trials, control arm strategy therefore needs to be considered not simply as a statistical efficiency question but as a core component of trial interpretability.

More detailed direction on randomization

Master protocols create choices that conventional two-arm studies generally do not. How many participants should be assigned to a shared control? How should allocation account for multiple treatment comparisons? Should allocation change as treatment arms enter or leave a platform? The revised guidance discusses allocation approaches for shared control designs, including considerations around variance minimizing allocation.

These decisions affect more than sample size. Randomization can influence the precision of treatment comparisons, the efficiency of the shared control, recruitment, operational feasibility, and ultimately the credibility of the evidence submitted to the FDA. As with any clinical trial design, sponsors should therefore treat randomization as an integral part of the master protocol’s statistical architecture rather than a downstream implementation decision.

Blinding: Managing bias within complex designs

Blinding can be challenging in master protocols because multiple therapies may have different routes of administration, schedules, formulations, or monitoring requirements. The revised guidance provides additional discussion of blinding approaches and circumstances in which open label designs may be appropriate. The underlying regulatory principle, however, remains familiar: operational convenience should not come at the expense of reliable endpoint assessment.

Where open label treatment is necessary, sponsors should consider how treatment awareness could influence participant behavior, investigator decisions, treatment discontinuation, concomitant therapy, or endpoint assessment and build appropriate protections into the design. The key question is therefore not simply “Can this master protocol be blinded?” but “Where could lack of blinding introduce bias, and how will that risk be controlled?”

Regulatory strategy should start early

The revised guidance also gives greater attention to the regulatory mechanics behind master protocols. A platform involving multiple drugs, substudies, and sponsors creates a different regulatory environment from a conventional standalone trial. The FDA addresses topics including IND structure, cross referencing, protocol amendments, safety reporting, and communication among participating parties.

The FDA recommends that a master protocol generally be submitted under a new IND and that sponsors request a pre-IND meeting to discuss the protocol and submission strategy. For multi-sponsor platforms, governance is especially important. Responsibilities for safety information, protocol amendments, cross referencing, and FDA interactions should be clearly defined before the platform becomes more complex.

What has not changed since 2023

Despite the additional detail, the 2026 revision does not represent a fundamental shift in the FDA’s position. The FDA continues to recognize that well-designed master protocols can accelerate development while generating reliable evidence of safety and effectiveness. Core regulatory principles remain unchanged: protecting participants, minimizing bias, maintaining data integrity, preserving interpretability, and generating evidence sufficiently robust for regulatory decision-making.

From 2023 to 2026: A more operational framework

The 2023 draft established a broader framework for the FDA’s expectations around master protocols, building in part on experience with large platform trials during the COVID-19 pandemic. The 2026 revision makes that framework more operational, incorporating stakeholder feedback and expanding the FDA’s recommendations, particularly for basket trials. The regulatory conversation is moving beyond whether master protocols are acceptable toward a more practical question: what is required to make them sufficiently rigorous for regulatory decision-making? For sponsors, that places greater emphasis on prospective, cross functional planning. Statistical design, clinical strategy, operations, data management, safety surveillance, governance, and regulatory strategy need to work as an integrated system.

Key takeaways for sponsors

The June 2026 revised draft guidance is best viewed as an evolution rather than a change in direction. The FDA remains supportive of master protocols and recognizes their potential to reduce duplicative infrastructure and accelerate evidence generation. At the same time, efficiency does not lower the evidentiary standard.

Teams planning umbrella, basket, or platform trials should address several questions early:

• Are the populations grouped within the protocol scientifically justified?
• Are controls comparable and concurrently eligible for the relevant treatment comparison?
• Does the randomization strategy support each key research question?
• Could treatment awareness introduce bias?
• Is the regulatory and governance structure capable of adapting as the protocol evolves?

Master protocols should answer these questions prospectively.

The message from the FDA’s 2026 revision is therefore not simply that master protocols can make development more efficient. It is that those efficiencies must be achieved without sacrificing the quality, interpretability, and regulatory credibility of the evidence they produce.

The June 2026 document remains a revised draft guidance and contains nonbinding recommendations.

Share Via