Writing a master protocol
A master protocol study has several parallel substudies that share a common overarching framework, and its protocol has to divide its content between what every substudy shares and what each one adds.
Why master protocol designs arose
In immuno-oncology, new agents with diverse mechanisms of action raised the possibility that combination therapies could lead to greater depth of response and prolonged survival. The large number of potential combinations created an issue of practicality for industry, health authorities and clinical investigators, who share the goal of understanding which agents bring the greatest value to patients, so a clinical trial framework was needed that assesses novel combinations robustly across a broad range of patient populations within a tumour type and lets combinations be evaluated relative to one another. One strategy for evaluating combination therapies has been a complex clinical trial design whose separate parts could in effect be perceived as individual clinical trials but are elements of a single protocol, characterised by extensive adaptations such as planned additions of new investigational medicinal products or new target populations. The master/subprotocol concept is one such design.
In response to COVID-19, the research community started thousands of clinical studies to understand the disease's pathology and identify safe and efficacious treatments, and the absence of a coordinated response and the fragmented approaches that followed impeded health policy decision-making and resource allocation. Estimates suggested that only 6% of COVID-19 clinical studies in the US were expected to yield actionable data, the primary barriers being poor enrolment, because overlapping studies competed for similar patient populations, and studies conducted without the robustness that regulatory approval needs. During the pandemic, master protocol study designs were shown to be a more structured and sustainable approach to clinical study evaluation. A master protocol design standardises study design and operating procedures, and the efficiency and uniformity this brings facilitate the parallel development and parallel evaluation of several interventions.
Early master protocol studies in oncology targeted prevalent biomarkers and genetic subtypes to address several clinical questions within one overall study structure. Their use then increased steadily and branched out into other therapeutic areas, such as Alzheimer's disease (DIAN-TU), Ebola (PREVAIL II) and community-acquired pneumonia (REMAP-CAP). In early 2020, COVID-19 accelerated their adoption, as governments and researchers established far-reaching master protocol studies to address the public health crisis, among them the World Health Organization's Solidarity Trial of repurposed antiviral drugs, the US National Institutes of Health's ACTIV network of four fast-track focus areas for treatment, ANTICOV, the largest COVID-19 study conducted in Africa, and RECOVERY, the UK platform study that demonstrated that dexamethasone and tocilizumab improved the survival of hospitalised COVID-19 patients.
Recent years have seen clinical study protocols introduce more complex design features into the gold-standard randomised controlled trial, to make investigational drug development more efficient. Such complex protocols often involve a master protocol alongside disease-specific subprotocols. Cancer has been the clinical setting that has most often used novel study protocols, but a survey in 2021 found master protocols in infectious disease, neuroscience, immunology and rare disease settings, the most common design being the basket trial.
Two medical writers predicted in 2021 that after the pandemic master protocol studies would be an integral part of clinical research and play an important role in providing actionable data to support health policy and resource allocation, and that medical writers and study teams would be expected to be fluent in developing clear and coherent protocols for them.
Defining a master protocol study
Regulators and key opinion leaders agree that a master protocol study has several parallel substudies that share a common overarching framework, but in 2021 how these studies were defined and categorised had not yet reached maturity, and the lack of a clear definition and classification had obstructed wider adoption of the designs. The definitions of the US Food and Drug Administration (FDA), the Clinical Trials Facilitation and Coordination Group (CTFG) of the Heads of Medicines Agencies and EU Patient-cEntric clinicAl tRial pLatforms (EU-PEARL) showed a continuing evolution in understanding, and EU-PEARL's description was the most comprehensive.
The terminology of the protocol document was evolving too, and only the CTFG and EU-PEARL described master protocol content and platform study content, respectively. To separate the study design from the document, the terms core protocol and subprotocol have been proposed for the common and the substudy-specific content.
| Term | Description | Reference |
|---|---|---|
| Master protocol study | A single overarching design developed to evaluate several hypotheses, whose general goals are to improve efficiency and establish uniformity through standardisation of procedures in the development and evaluation of different interventions. Under a common infrastructure, the master protocol may be divided into several parallel substudies, with standardised study operational structures, patient recruitment and selection, data collection, analysis and management. | EU-PEARL 2020; Park et al. 2019 |
| Protocol scaffold | A visual aid for planning how the protocol content will be distributed between the core protocol and the subprotocols. It is most easily presented by extracting the protocol template's table of contents and indicating whether content is in the core protocol or the subprotocols, whether it is repeated, and whether it is complementary. | — |
| Core protocol (document) | The protocol document describing the content of the overarching study design that applies to all substudies, such as a general introduction to the master protocol study, common objectives and endpoints or estimands, the rationale for conducting the master protocol study, and common administrative, regulatory and operational elements. Also called the master protocol. | — |
| Subprotocol (document) | The protocol document or content specific to an individual substudy. Synonyms include intervention-specific appendices, domain-specific appendices, study modules and comparison protocols. | — |
Kinds of master protocol study
Woodcock and LaVange of the FDA gave the initial classification of master protocol studies as basket, umbrella or platform designs. Its application to real studies suggests that the classification was incomplete, with only 57% of the studies in a systematic review correctly classified, and EU-PEARL expanded the definitions to include a matrix design and a multi-arm multi-stage (MAMS) analysis framework.
| Name | Description | Reference |
|---|---|---|
| Basket | A study designed to test a single intervention in different populations defined by disease stage, histology, number of prior therapies, genetic or other biomarkers, or demographic characteristics. | Woodcock & LaVange 2017; FDA 2018 |
| Umbrella | A study designed to evaluate multiple interventions, administered as single drugs or as drug combinations, in a single disease population. | Woodcock & LaVange 2017; FDA 2018 |
| Platform | A study designed to evaluate multiple interventions in the context of a single disease in a perpetual manner, with therapies allowed to enter or leave the platform on the basis of a decision algorithm. | Woodcock & LaVange 2017 |
| Matrix | A study that is both an umbrella study and a basket study, including analyses in multiple disease subtypes. Many platform studies are matrix studies with the additional feature that, as the study progresses and interventions leave it, new interventions may enter, and the study has no initially fixed duration or sample size. | EU-PEARL 2020 |
| Multi-arm multi-stage (MAMS) | An analysis framework that can be used with umbrella or platform designs. It analyses study results in a group-sequential framework, controls the overall type I error and is attractive for studies intended for regulatory submission. It avoids features more problematic for regulatory submission, such as response-adaptive randomisation, subgroup analysis and longitudinal modelling. | EU-PEARL 2020 |
A master protocol study designed with a fixed number of populations or interventions is a basket study (a single intervention in several populations), an umbrella study (several interventions in a single population) or a matrix study (several interventions in several populations). A study designed with the ability to add or stop substudies prospectively is a platform study.
Master protocol designs have led to great advances in cancer therapy. A basket trial investigated imatinib in 186 patients with 40 different malignancies, ranging from solid tumours to haematological cancers, a common treatment combination being investigated across several disease cohorts and outcomes assessed in the context of the relevant genetic mutations. The umbrella trial NCI-MATCH investigated whether treating cancers according to their molecular abnormalities was effective, and assigned each of its 1,593 participants to one of 38 subprotocols. Matrix studies, like platform studies, can remove interventions and include new ones as the study progresses, and they need not have a fixed duration or sample size. Platform studies have a fluid structure that allows several targeted therapies to be studied in populations with similarities such as a common disease. The platform trial STAMPEDE, in men with newly diagnosed advanced prostate cancer, included almost 12,000 participants from its start in 2005 and in 2025 was ongoing with recruitment closed, and its first results demonstrated improved disease control and life expectancy from adding docetaxel or abiraterone to treatment regimens; since then its fluid structure has allowed many other strategies to be tested.
Dividing the content between the master protocol and its subprotocols
In a master/subprotocol design for evaluating combination therapies, the master protocol applies to every combination treatment selected for evaluation under a tumour-specific study, and defines:
- the overall study plan
- the background and rationale
- the study design and duration
- the inclusion and exclusion criteria
- the time and events, including all procedures, laboratory tests, pharmacokinetics and pharmacodynamics that are not treatment-specific
- the statistical plan
Further treatment combinations can then be introduced into the study through subprotocols appended to its master protocol, each with the information appropriate to the combinations or contemporaneous controls being added. A subprotocol details the specific study treatments and contains:
- the background: the scientific rationale, from preclinical and clinical data, for evaluating the additional combinations
- preclinical toxicology on the single agents
- a clinical safety package for new agents, including monotherapy safety information
- combination safety data on at least six participants to support the protocol-specified dose, which may come from a different patient population or tumour type
- the drug dose and administration
- adverse events and dosing modifications
- treatments and evaluations, including treatment-specific procedures such as pharmacokinetics, which the master protocol does not hold
Guidance documents generally describe a master protocol that describes the study design in detail, with subprotocols that describe the disease-specific aspects to highlight the differences between them.
Choosing the protocol structure
Basket, umbrella and platform protocols have different focuses and several possible structures, and the structure is worth defining with the study team before protocol preparation starts. As of 2021, guidance for conducting master protocol studies had focused on the operational implementation of the protocol and given little attention to its structure, although the structure makes decisive contributions to how the substudies are submitted, updated and reported. The need for flexibility in protocol structure to accommodate master protocol designs was recognised in 2015, and since then the adoption of the SPIRIT recommendations and the TransCelerate Biopharma Common Protocol Template has consolidated industry protocols around a common framework more amenable to standardised document structures for master protocol designs.
Templates for developing master protocols are available, such as EU-PEARL's, and more detailed guidance comes from the US FDA and TransCelerate Biopharma, but how useful the templates are depends on factors including the disease setting and the experience and expectations of the study development team. By 2023, guidance on master protocol trial designs had begun to emerge, such as the European Medicines Agency's concept paper on platform trials, and EU-PEARL released a provisional version of a template tailored to platform protocols in 2021 and in 2023 announced a final version for later that year.
The complexity and variability of master protocol designs preclude a one-size-fits-all approach, so the chosen structure has to balance the study's needs against the resulting trade-offs, a decision that can affect study conduct and data integrity if it is made poorly. Where subprotocol content is minimal, a standard protocol structure would be most appropriate, but it can soon become complex and difficult to understand as more content is added. Appendix or annex subprotocols and independent subprotocols are comparatively clear for larger studies with more substudies and for studies with few substudies of substantial content, and independent subprotocols give additional flexibility when recurrent or parallel amendments are anticipated over the life of the study.
| Protocol structure | Description and structure example | Benefits | Risks |
|---|---|---|---|
| Integrated subprotocols | Substudy protocol content is integrated within the core protocol structure. Example: Section 5, Study Population; 5.1, Inclusion Criteria (IC); 5.1.1, IC for Substudy 1; 5.1.2, IC for Substudy 2; 5.2, Exclusion Criteria (EC); 5.2.1, EC for Substudy 1; 5.2.2, EC for Substudy 2. | Easy to implement. Maintains the standard protocol structure. Reduced structural complexity compared with the other structures. Minimal impact on protocol development processes and timelines. A single document for future amendments. | Comprehension reduced as substudy content increases. Can substantially increase the length of the standard protocol structure if there are numerous substudies. No single location for substudy information, so all study staff need a detailed review and comprehension of the protocol. Difficult to amend if new substudies are required. |
| Appendix or annex subprotocols | Substudies are provided as separate appendices or annexes to the core protocol. Appendix example: core protocol; Appendices 1–9; Appendix 10, Subprotocol 1; Appendix 11, Subprotocol 2. Annex example: core protocol with Appendices 1–9; Annex Document 1, Subprotocol 1; Annex Document 2, Subprotocol 2; the annex documents are submitted under the same submission number. | Clearer comprehension when there are several subprotocols with substantial content. Maintains the standard protocol structure. Clear division of subprotocol information, useful when not all sites are enrolling across all subprotocols. Easy to amend if new substudies are required. A single document for future amendments. | Redundant if substudy information is brief or few substudies are planned. Reduced flexibility for amendments as the number of substudies increases, since substudies cannot be updated independently and amendments are queued: subprotocol 2 could not be amended while subprotocol 1 was undergoing an amendment. Risk of repetition, redundancy or conflicting statements in the subprotocols compared with the core protocol if not managed correctly. Moderate impact on protocol development processes and timelines. Increased reporting complexity, as a single study report is required. Increased study disclosure complexity, as all substudies need to be summarised and submitted simultaneously. |
| Independent subprotocols | Substudies are provided as independent subprotocol documents and registered separately. Example: core protocol, submitted alongside subprotocol 1 and subprotocol 2; Subprotocol 1, EudraCT number 2021-xxxxxx-01; Subprotocol 2, EudraCT number 2021-xxxxxx-02. | Clearer comprehension for several subprotocols with substantial content where the core protocol information is limited to summary operational details, which is most common for platform and matrix studies. Maintains the standard protocol structure. Clear division of subprotocol information, useful when not all sites are enrolling across all subprotocols. Easy to amend if new substudies are required. Subprotocols can be amended independently and submitted in parallel if desired. Independent reporting of each substudy. Study disclosure less complex than summarising all substudies together. | Redundant if the core protocol contains most of the content, where appendix or annex substudies are preferable. Core protocol updates affect several submissions that need to be updated in parallel. Risk of repetition, redundancy or conflicting statements in the subprotocols compared with the core protocol if not managed correctly. High impact on protocol development processes and timelines, with more upfront planning and time required from team members. Increased administrative burden if several amendments are conducted in parallel. |
A review of the literature in 2021 found no preferred protocol structure by study design, but field experience from the STAMPEDE and FOCUS4 platform studies supported appendix or annex subprotocols or independent subprotocols for platform studies.
Registering, submitting and reporting a master protocol study
In a master/subprotocol design, each study, master and subprotocols alike, can be identified by single EudraCT and IND numbers, with all its elements linked by a single research hypothesis, an important regulatory feature that a sponsor's global regulatory function may consider carefully. Each subprotocol is then submitted as a substantial amendment for separate regulatory and ethics committee review before it is implemented.
The regulatory requirements of the study are a further consideration for the protocol structure, and in 2021 submission guidelines were limited, since only the FDA and the CTFG had released guidance on master protocol studies. Both recommend one of two submission structures: a single submission with several substudy protocols under a single EudraCT or NCT number, which suits integrated or appendix or annex subprotocols, or independent subprotocols, each accompanied by the common master protocol and submitted under its own EudraCT or NCT number. The submission strategy depends on the operational needs and the long-term considerations of the overall master protocol study.
Complex clinical trials are most often early exploratory trials in relatively few participants, so the limited safety data make transparency even more of an obligation. One potential obstacle to transparency in a master/subprotocol design is that, when every subprotocol is registered under the same EudraCT and IND numbers as the master protocol, the information from each completed subprotocol becomes available only after the end of the entire trial. This limits the technical obligation to report the treatment arms to one year after the last patient's last visit in the entire trial, which reduces the documentation burden but increases the need for robust and ethical reporting strategies.
For a complex design registered as one trial, sponsors are strongly advised, for timely and transparent reporting of key information, to engage the health authorities and propose periodic safety and status reports that summarise the current study status, including:
- how many participants have been enrolled, randomised and treated
- which arms have been closed or newly opened
- the proposed plans for the next periodic interval, including known amendments and upcoming subprotocol initiations
- the overall safety parameters, such as adverse events, serious adverse events, discontinuations and deaths
- an assessment of the overall benefit-risk of the trial for each amendment that adds a new subprotocol, addressing how all the risks will be mitigated
Sponsors are also strongly advised to include data from closed subprotocols in the appropriate investigator's brochure. Communication within the sponsor's regulatory and clinical organisations, and robust interactions between those organisations and the relevant health authorities, are critical to realising the promise that combinatorial clinical trial designs hold against cancer.
Running the development of a master protocol
Master protocol studies are highly complex, and the complexity and variability of their protocol development can test even the most experienced medical writer and study team. Standard protocol templates and approaches are often inadequate for the complexity and the many configurations of a master protocol study.
Confirming the team
Confirming the study team members before protocol development begins is an important task, which starts gathering individual expertise around the study's collective objectives, although it is often not the medical writer's responsibility. It can be challenging, in particular for a team managing its first master protocol study, and unlike for a traditional protocol, identifying every member in advance may not be straightforward, because the team structure depends on the ambition of the study design, the number of interventions and patient populations, and the countries involved:
- Study design: the design may include adaptive elements, decentralised components or digital health technologies, and engaging the relevant expertise early minimises the risk of substantial changes late in the protocol's development.
- Several interventions or participant populations: depending on the organisations involved, there may be several representatives for one function, so a study of several interventions may need representation from each intervention group, such as medical professionals or study or programme leaders, and a study of several populations needs adequate representation for each population to ensure the design suits it.
- Geographic footprint: as in every multiregional study, the regulatory requirements of the countries in which the study is conducted may influence the protocol, and a master protocol study may need additional discussion and engagement with regulatory agencies or regulatory professionals during the protocol's development.
Agreeing the structure
Making sure every team member knows and agrees the protocol structure before development begins reduces the risk that conflicting opinions about it arise later, from unfamiliarity with master protocol designs or with the study's requirements, and extend review cycles or require additional drafts. Both can damage the team's decision-making and efficiency, which may extend development time and reduce overall quality. The medical writer can initiate early discussions to identify the most suitable structure, on these questions:
- Does the structure give readers optimal clarity and coherence? The readers of a master protocol may engage with it not as one study but as separate substudies: two readers of the same document may approach it with different participant populations, interventions or study schedules in mind, and the question is whether the structure lets each find the relevant substudies easily.
- Will information be repeated, or will a single source be cross-referenced throughout? There is a strong argument for cross-referencing a single source rather than repeating information within or across the core protocol and the subprotocols, since duplication breeds inconsistency, although not everyone shares this view. Where information is repeated, the team should clarify what essential content is repeated, such as overarching objectives and endpoints or eligibility criteria, how team members will comment on its repetitions, and how its consistency will be controlled.
- What information is specific to the core protocol and what to each subprotocol? For example, whether every substudy follows the same schedule of assessments, and whether there is a core set of eligibility criteria with additional criteria for each substudy.
Once the structure is provisionally decided, the medical writer may develop a protocol scaffold to show the team what the structure will look like, which can minimise the risk that the team rejects the structure during its revision and requires substantial changes midway through development.
Agreeing the timelines
Agreeing timelines up front, together with the structure and the scaffold, is an important step in aligning expectations while allowing enough time for protocol development. Two approaches to protocol development have been proposed:
- A parallel approach, which develops all the content together as standard protocol development does, with additional time for content development and review.
- A staggered approach, which submits leading content, such as the core protocol, for review first and trailing content, such as the subprotocols, once the leading content has been reviewed.
A parallel approach is preferable in some circumstances, for example where several indications are involved and it helps to engage every team member at the same time, while a smaller team covering every substudy would likely benefit from a staggered review, which keeps reviewers from being overburdened. The team can also consider whether any planned absences or work requirements would interfere, and whether there will be periods when all the content has to be reviewed together, such as when the protocol is close to being finalised.
A multi-indication master protocol
One team developed a master clinical study protocol to evaluate the clinical activity of a new drug across several indications. The master protocol held the common trial elements, and the disease-specific aspects were presented as separate substudy protocols. To enhance clarity and avoid confusion, the team outlined the overall protocol structure at the beginning of the master protocol, and because of the protocol's complexity it briefed internal reviewers and the quality control team before their evaluations.
The writers prioritised clear, unambiguous language and a consistent structure, aiming for simplification to make trial implementation and execution efficient. The protocol was submitted and approved in several countries, and no regulatory authority or ethics committee raised issues about its structure, complexity or readability. The flexibility of the master protocol enables compliance with country-specific requirements while keeping a harmonised global protocol, and allows adaptations as the study progresses. Careful documentation of amendments and version relationships will be essential for quality assurance. The team concluded that the experience underscores the importance of strategic planning, regulatory alignment and cross-functional collaboration in implementing complex clinical trials, and that the protocol's development demonstrates the potential of innovative trial designs to accelerate drug development.
A pivotal rare-disease protocol with detailed subprotocols
A pivotal phase 3 study, double-blind, randomised, multicentre and placebo-controlled, was to investigate one therapeutic agent in two rare diseases, which led to a master protocol with subprotocols. In a somewhat unusual approach, the team prepared detailed disease-specific subprotocols rather than a detailed master protocol, each subprotocol holding all the information usually expected in a protocol for a randomised controlled trial. The master protocol then served as a summary document that presented an overview of the study, highlighted the differences between the two subprotocols, and gave the rationale for a master and subprotocol approach rather than separate studies. In this arrangement, the subprotocol becomes the document mainly used by the investigator, and the master protocol supports regulatory oversight. The team considered that detailed subprotocols promote clarity for investigators, because all the information required for conducting the study was in a single document rather than split between a master protocol for common aspects and a subprotocol for disease-specific ones. The approach aimed to make the clinical investigation process efficient and to offer optimal clarity in process and design to both study investigators and regulatory reviewers.
The comparison of the two subprotocols highlighted the differences that would be expected, chiefly in the disease inclusion and exclusion criteria, the sample size calculation and the stratification variables.
| Similarities | Differences |
|---|---|
| Similar disease characteristics, facilitating the use of many common endpoints | Each subprotocol has its own control group, because of potential differences in disease progression profiles or rates between the conditions |
| Common database set-up | Sample size calculation |
| Comparable visit and assessment schedules between the subprotocols | Stratification variables |
| Use of a single independent safety monitoring committee | Disease inclusion and exclusion criteria |
The complex approach was intended to rationalise operational aspects, allowing the common database set-up, the comparable visit and assessment schedules and the single independent safety monitoring committee.
Giving each subprotocol its own control group was a major difference from many other complex-design studies, which use a common control group across substudies to accrue a larger body of data on the investigational agent and to ensure that as many participants as possible receive the potentially beneficial treatment.
The sponsor submitted the protocol as part of a clinical trial application through the Clinical Trials Information System, and it was approved. The regulators' comments were as expected for a phase 3 protocol, with minor changes required, and the master and subprotocol structure was approved without resistance to the concept. The master and subprotocols were all to be registered as one clinical study, due to start in 2025, and the protocol was to be submitted to further countries and regions.