Precision medicine has fundamentally changed how diseases are defined and how therapies are developed — shifting eligibility from clinical phenotype to molecular profile, and fragmenting patient populations in the process. However, with that promise comes complexity: new therapies are increasingly targeted, diseases are redefined by molecular profiles, and patient populations are more fragmented than ever. Traditional trial designs — built around single-disease, single-therapy assumptions — are structurally misaligned with the realities of genetically stratified drug development. master protocols have emerged as a structural response — trial architectures designed to accommodate the complexity that precision eligibility, targeted therapies, and fragmented patient populations introduce.
Master protocols are trial designs that test multiple drugs and/or multiple subpopulations in parallel under a single protocol, without the need to develop new protocols for every trial. These designs break away from the one-drug, one-disease model. They have been most widely adopted in oncology and rare disease research, where matching targeted therapies to small patient cohorts is key. Regulatory bodies have recognized this shift. In December 2023, the FDA released draft guidance on master protocols for drug and biological product development, providing recommendations on trial design, analysis, and submission documentation. This signals growing institutional support for these approaches.
Master protocols reduce redundancy, reduce the operational and regulatory complexity of running parallel trials, evaluate multiple therapies within a single overarching framework, and reduce unnecessary burden on patients by limiting their exposure to interventions they are not suitable for. The three main master protocol designs — basket trials, umbrella trials, and platform trials — each address a different dimension of this structural challenge.
Basket trials group patients with different diseases who share a common molecular alteration, allowing researchers to test a single therapeutic intervention across various conditions.
This design has been widely used in oncology, where mutations in the same oncogene may be present across various types of cancers.
For instance, the MyPathway trials featured multiple “baskets” to evaluate the efficacy of pertuzumab and trastuzumab in patients with solid tumors with amplification, overexpression, or mutation in HER2. The trial's primary endpoint was objective response rate (ORR), with survival and safety outcomes as secondary endpoints. By grouping patients based on shared genetic alterations rather than disease, basket trials create a more efficient route for testing targeted treatments. This is especially useful when the alteration of interest is rare in any one condition.
The biotech company Alltrna is extending the utility of the basket trial approach beyond oncology. Alltrna is developing therapeutic tRNA molecules designed to target stop codon disease by restoring protein production. For the next phase of drug development, the company is using basket trials to evaluate the safety and efficacy of a single tRNA molecule across hundreds of diseases that share the same underlying mutation which creates a premature stop codon. Instead of designing separate drugs and trials for each condition, this disease- and protein-agnostic strategy could reduce development costs and eliminate the need for separate drug candidates and trials for each individual condition.
Umbrella trials, by contrast, enroll patients with the same disease and stratify them into subgroups based on specific molecular or genetic markers. Each subgroup receives a treatment tailored to its unique biomarker profile. This approach enables a comparative evaluation of several targeted therapies within one disease area.
Because eligibility for each arm depends on a patient's biomarker status, umbrella trials require robust molecular screening workflows. Genetic testing, result interpretation, and counseling must be coordinated at scale to ensure patients are assigned to the correct treatment arm without unnecessary delay.
Lung cancer illustrates this well. Previously approached as a single disease in clinical trials, it is now understood as a collection of molecularly distinct subtypes. In precision medicine trials for lung cancer, patients are stratified by subtypes such as adenocarcinoma and small cell lung cancer. Each subgroup receives targeted therapies based on specific mutations, often involving EGFR and ALK. Evaluating multiple interventions within a single trial structure improves efficacy and accelerates drug development.
Platform trials represent the most flexible of the three designs. They incorporate multiple therapies and arms within a single, ongoing framework, allowing for seamless adaptation based on interim data. As data is collected, treatment arms may be dropped due to lack of efficacy or safety concerns, while new interventions can be introduced.
A notable example is the STAMPEDE trial, which began in 2005 to evaluate novel treatment strategies for men with high-risk, locally advanced or metastatic prostate cancer. As a multi-arm, multi-stage platform trial, STAMPEDE allows for the continual assessment of new therapies added to standard androgen deprivation therapy. More than 10,000 patients have been enrolled across several research arms, with ineffective treatments removed based on interim failure-free survival analyses.
STAMPEDE illustrates how a platform design anchored in a consistent standard-of-care control group can generate reliable conclusions about new therapies across a large enrolled population — without restarting the trial each time a new arm is added. Evaluating multiple treatment arms within a single, shared control-group framework would require substantially more patients, administrative infrastructure, and elapsed time if pursued through separate sequential trials.
Platform trials have since been applied beyond oncology. The REMAP-CAP trial, designed to evaluate treatments for community-acquired pneumonia, became one of the most prominent examples of adaptive platform design when it was rapidly expanded during the COVID-19 pandemic. Its multi-arm, multi-stage structure allowed new treatment arms to be added and evaluated in near real-time, demonstrating how platform trials can respond to urgent clinical questions while maintaining statistical rigor."
| Trial Type | Focus | Key Characteristic |
|---|---|---|
| Basket | One drug, many diseases | Groups patients by shared molecular alterations. |
| Umbrella | One disease, many drugs | Stratifies patients by specific biomarkers. |
| Platform | One disease, many drugs | Flexible, ongoing design with adaptive arms. |
The growing adoption of these trial designs reflects a broader trend toward more efficient, personalized, and data-responsive research. A 2019 review found that the number of basket, umbrella, and platform trials had increased nearly threefold in just five years. The FDA's 2023 draft guidance on master protocols further signals that regulators are prepared to support these frameworks with clearer pathways for design, analysis, and submission.
However, the promise of master protocols depends on the infrastructure behind them. Each design places significant demands on patient identification, biomarker screening, and cohort management. Basket trials require finding patients with shared molecular alterations across different disease populations. Umbrella trials require robust genetic testing workflows to stratify patients into the correct treatment arms. Platform trials require sustained engagement as arms are added or removed over time.
Without coordinated systems for recruitment, genetic testing, and long-term participant engagement, the operational complexity of these designs can offset their scientific advantages. As master protocols become more central to precision medicine, the ability to identify, qualify, and retain the right patients across multiple arms and geographies will increasingly determine whether these trials deliver on their potential.