A master protocol is an overarching study plan that simultaneously investigates multiple treatment arms, indications, or patient subgroups under a unified regulatory and operational framework. Instead of initiating a separate clinical trial for each research question—with its own application, infrastructure, and data management—the master protocol consolidates all components into a single shared platform. This concept has proven particularly efficient, especially in oncology and in the development of therapies for rare diseases.
Types of master protocols
There are three established main types, which differ in study structure and research question. Basket trials investigate a single therapy across multiple tumor entities or patient subgroups defined by the same molecular biomarker—regardless of the primary tumor. Umbrella trials evaluate multiple therapies within a single disease, with patients stratified into different treatment arms based on molecular profiles. Platform trials are the most flexible form: they allow new treatment arms to be added and ineffective arms to be dropped during the ongoing study without re-approving the entire platform. All three types share a common control group, shared biomarker testing procedures, and a central data infrastructure.
Regulatory framework
The FDA and EMA have published specific guidance for master protocols. The FDA guidance “Master Protocols: Efficient Clinical Trial Design Strategies to Expedite Development of Oncology Drugs and Biologics” (2022) describes requirements for study design, statistics, and regulatory communication. In the EU, Regulation 536/2014 governs the approval of complex study designs via CTIS, whereby each new sub-study or new treatment arm can be submitted as a Substantial Amendment without having to re-apply for the entire platform. This significantly reduces regulatory burden and accelerates activation of new arms. Close alignment with authorities through Scientific Advice or pre-submission meetings is particularly recommended for master protocols.
Bayesian methods are increasingly used in platform trials because they can leverage ongoing evidence accumulation across all arms more efficiently than classical frequentist approaches. Response-adaptive randomization enables dynamic adjustment of randomization probabilities in favor of more effective arms, which is ethically attractive but places additional requirements on blinding and the DSMB. Pre-defining and publishing the statistical analysis plan is particularly important for master protocols to avoid allegations of data dredging.
Statistical considerations
Master protocols entail specific statistical requirements. Joint control of multiple hypothesis tests increases the risk of false-positive results (alpha inflation), so appropriate multiplicity adjustments must be defined in advance—for example, Bonferroni correction or familywise error rate control. Adaptive designs within platform trials enable interim analyses after which arms can be stopped or expanded according to pre-defined rules. An independent Data Safety Monitoring Board (DSMB) monitors all arms simultaneously and must consider both individual arm safety and platform-wide patterns. The statistical and methodological plan must be fully pre-specified before study start and be transparent to all involved authorities.
Operational advantages and challenges
The greatest operational advantage of a master protocol lies in the shared infrastructure: a single eTMF, a shared CTMS, a central randomization and stratification platform, and shared laboratory resources for biomarker testing reduce overall costs and significantly accelerate the start-up of new arms. Sites only need to be qualified and activated once and can then participate in multiple arms. At the same time, master protocols are more complex to manage: coordination among multiple sponsors (in collaborative studies), management of sublicenses, and clear allocation of responsibilities require robust legal and operational governance structures, ideally defined in a platform charter.
Master protocols proved particularly valuable during the COVID-19 pandemic: platform trials such as RECOVERY (UK), SOLIDARITY (WHO), and ACTT (NIH) were able to evaluate multiple therapies simultaneously within weeks and deliver results in months rather than years. These experiences have significantly strengthened global regulatory acceptance of master protocols. For CROs with expertise in complex study designs, master protocols offer attractive business opportunities, as sponsors increasingly seek experienced partners for platform governance, biostatistical planning, and multinational regulatory management.
Frequently Asked Questions
What is the difference between a basket trial and an umbrella trial?
A basket trial evaluates a single therapy (e.g., a BRAF inhibitor) across different tumor types that all share the same molecular feature. An umbrella trial, by contrast, evaluates multiple different therapies within a single tumor entity—patients are assigned to the different treatment arms based on their molecular profiles. Both use biomarker-based stratification, but the direction of the research question is opposite.
Does every new treatment arm require a new regulatory approval?
In the EU, a new treatment arm can be submitted as a Substantial Amendment to the existing master protocol—a full re-approval is not required. The EMA and national competent authorities review the new arm under the accelerated amendment procedure. In the US, sponsors align new arms with the FDA via IND amendments. The exact approach depends on the study design and the jurisdictions involved.
Is a master protocol only suitable for oncology?
No—although master protocols are most widely used in oncology, they are increasingly being applied in other areas as well: infectious diseases (e.g., COVID-19 platform trials such as RECOVERY or SOLIDARITY), rare diseases, and more recently neurological disorders. The concept is applicable wherever multiple therapies are to be compared against a shared platform control or a therapy is to be investigated in biomarker-defined subgroups.