EMA outlines approach to single pivotal trials for marketing authorization

Experts from the European Medicines Agency and its committees have outlined the agency’s approach to determining the number and design of pivotal clinical trials needed to support marketing authorization, amid discussion over whether a single trial should become the default in drug development.

The Baseline

• European Medicines Agency experts say the number of pivotal trials should be determined by the scientific questions, evidence gaps and patient needs rather than a predetermined standard.
• A single pivotal trial can be sufficient when supported by compelling results and the overall evidence package.
• The authors recommend early scientific advice to determine appropriate evidence-generation strategies.

A new article published in Nature Reviews Drug Discovery examines how the number of pivotal clinical trials should be determined when generating evidence for marketing authorization. Written by experts from the European Medicines Agency (EMA) and its committees, the article sets out an approach based on the disease and patient population, existing evidence, remaining uncertainties and the scientific questions to be addressed, with patients’ needs informing evidence generation.

In February 2026, then-US Food and Drug Administration (FDA) officials Vinay Prasad and Martin Makary proposed in the New England Journal of Medicine that one pivotal trial should become the standard or default approach. Their proposal cited advances in trial methodology, biomarkers and statistical methods, alongside the potential to reduce development costs and timelines. The FDA subsequently issued revised draft guidance clarifying circumstances in which one adequate and well-controlled clinical investigation, together with confirmatory evidence, may satisfy the substantial evidence of effectiveness standard.

The discussion comes as EMA scientific advice approaches its 30th anniversary. Over this period, scientific advice has evolved alongside regulatory science and guidance, including consideration of the design and number of pivotal trials used to support marketing authorization applications.

The authors argue that evidence planning should begin by identifying the scientific questions that need to be answered, based on the disease and patient population, available evidence and remaining knowledge gaps. Patients’ views and needs should also inform this process.

“The number of trials to be conducted is therefore not a starting point but the outcome of careful consideration of the evidence needs,” the authors wrote.

The question has previously been considered by European regulators. Around 25 years ago, the EMA’s human medicines committee, now the Committee for Medicinal Products for Human Use (CHMP), considered whether a single pivotal Phase 3 trial could be sufficient. Following internal deliberation and public consultation, guidance established that there is no formal requirement to conduct more than one pivotal trial.

The authors noted that for applications relying on a single pivotal trial the results would need to be particularly compelling in terms of internal and external validity, consistency, clinical relevance, statistical significance and data quality. The totality of evidence, including early clinical, preclinical and pharmacological findings, is also considered when assessing whether a medicine’s benefits outweigh its risks.

Replication may provide a rationale for conducting two or more trials, particularly when the benefit–risk balance remains uncertain. However, the authors distinguished replication from simply repeating an identical study. Its scientific purpose is to examine the robustness of effects under different circumstances, which may involve different populations, clinical settings or comparators.

They cited examples in which separate trials addressed different scientific questions. For nivolumab in melanoma, pivotal trials assessed efficacy in previously untreated and previously treated patients. Pembrolizumab was evaluated against chemotherapy in one pivotal trial and an alternative immunotherapy in another as the treatment landscape evolved. Separate trials are also generally conducted in treatment-naive and previously treated populations in HIV.

Other sources can contribute to the overall evidence package. Real-world data (RWD), for example, may provide supporting evidence, although the authors noted that it may not always be sufficient to provide independent confirmation.

The article also considers patient access and the argument that a single-trial default could reduce development costs and timelines. The authors pointed to existing EU mechanisms intended to facilitate medicine development and access, including PRIME, accelerated assessment, conditional marketing authorisation, authorisation under exceptional circumstances and incentives for orphan medicines.

Scientific advice is central to these initiatives and has been shown to increase the likelihood of successful medicine development. The authors also highlighted the involvement of patients and downstream decision-makers, including health technology assessment bodies, in informing evidence generation.

“There is no debate about whether single pivotal trials can be sufficient; they can and increasingly are,” the authors wrote.

Rather than working towards a predetermined number of pivotal trials, the authors recommend that developers begin with patients’ needs and the scientific questions to be addressed, consider factors including evolving standards of care and developments in data collection and regulatory science, and seek scientific advice early.