Seamless clinical trials could aid development of new drugs for rare cancers, according to a Friends white paper and experts who spoke at the group’s annual meeting. While there are a wide range of practical challenges, regulatory support may make such flexible trial designs more widespread.

Regulatory background: Cancer drug development

• The past decade has brought about a revolution in cancer treatment. Immunotherapy’s emergence helped to shift the standards of care across a wide range of tumors, while developers found more and more success tailoring their treatments to the genetics of cancer. Now, newer technologies like antibody-drug conjugates, bispecific antibodies and radiopharmaceuticals are giving physicians powerful new tools to curtail certain cancers.
• The FDA has met this research activity with regulatory innovations designed to stimulate and expedite oncology drug development. Among the agency’s initiatives are programs designed to collaborate with other health authorities, develop new trial endpoints, modernize how drug doses are selected and encourage the testing of promising drugs earlier in the course of disease. At the center of these efforts is the Oncology Center of Excellence, which also oversees the agency’s Real-Time Oncology Review program.
• Yet cancer clinical trials, while streamlined in many cases, remain time-consuming and expensive. They are shaped by patterns of research formed over decades of testing blunt interventions like chemotherapy. For instance, developers often initially evaluate prospective drug candidates in patients who have run out of other treatment options, or whose cancers have progressed despite multiple lines of therapy. And early studies typically seek to establish a “maximum tolerated dose,” even when such an approach might not be as suitable for more targeted medicines. [Read AgencyIQ’s analysis of a recent FDA workshop on the topic here.]
• The standard, step-wise approach to trials is another longstanding pattern that regulators and developers alike aim to modernize. Generally, developers run studies in phases, first assessing a drug candidate’s safety and dosing, then exploring efficacy before running the kind of larger study regulators require for approval. This framework has advantages, allowing researchers to gauge a drug’s potential and risks before launching another trial, but it also slows development and requires more regulatory interactions.
• While combining certain phases is now a common practice, the FDA has in recent years issued limited guidance to support wider adoption of so-called seamless, or continuous trials. This type of study allows testing to continue across trial stages without stoppage. A 2022 final guidance addresses the use of expansion cohorts in first-in human trials for oncology products, while a 2025 FDA draft guidance adopts the ICH E20 guidance on adaptive trial designs [Read AgencyIQ’s analysis of the expansion cohort guidance document here, and the ICH E20 draft here].
• FDA Commissioner MARTIN MAKARY appears interested in promoting seamless trial designs, and has spoken often in recent months about how such studies could help make research more efficient. He’s also hinted a major new proposal from the agency could be coming. [Read what AgencyIQ knows about Makary’s plans here.]

A new white paper plots a roadmap for seamless trials in oncology

• The Friends of Cancer Research held their annual meeting in Washington on Nov. 4, 2025, spotlighting three white papers the group put together on seamless trial designs, combination drug development and multi-regional clinical trials. Each white paper was produced by a working group that included patient advocates, academic and industry experts, and staff from the FDA and National Cancer Institute.
• This analysis will focus on seamless trials; AgencyIQ will soon also have a detailed breakdown of the day’s discussion on multi-regional clinical trials here, and previously took a deep dive into the FDA’s July 2025 guidance on combination drug development in oncology here.
• In its seamless trial white paper, Friends describes these studies as trials that integrate “multiple, sequential stages of drug development” within a “single framework.” These stages, typically enumerated as Phase 1, 2 and 3, comprise assessments of safety, dose escalation and optimization, and efficacy. Seamless designs can either be “operationally” seamless, consolidating the conduct of testing under a single protocol, “inferentially” seamless, when investigators analyze data combined from multiple stages, or both. Seamless trials may also be adaptive, permitting researchers to change dosing levels, refine eligibility criteria or even modify endpoints.
• The white paper focused on drug development in rare oncology, which it said presents distinct challenges for developers. By definition, rare cancers affect fewer people, meaning recruitment for these trials can be challenging. And a standard, phased trial plan ups the difficulty by requiring drugmakers to repeat much of the recruitment and enrollment process for each distinct phase. Researchers may also understand less about the pathophysiology of a rare tumor than they do a common cancer, making it more important that clinical trials for rare oncology products expand disease knowledge as well as test prospective drugs.
• Seamless trials can be a good solution for these problems. They can help maximize the information collected from any given patient enrolled into a trial, for instance, or allow a sponsor to adapt the overall study’s design upon gathering preliminary evidence from an early cohort. A seamless trial organized under a single protocol also avoids the delays and inefficiencies that result from winding down and starting up studies phase by phase, which may include obtaining new clearances from institutional review boards as well as reinstituting robust recruitment efforts and keeping investigators engaged across diverse institutions and geographies, for example.
• As the white paper makes clear, developers can deploy seamless trials to accomplish different research objectives. Early-phase trials, for instance, might combine standard dose finding with efficacy signal-seeking so sponsors can better plan their registrational studies. As an example, the white paper points to the FIGHT-101 study run by Incyte to test the tyrosine kinase inhibitor pemigatinib in solid tumors.
• Seamless trials may also support registration in their own right, as did studies of the RET inhibitors pralsetinib and selpercatinib, and the ROS1 inhibitor repotrectinib. The pralsetinib study, dubbed ARROW, was recently referenced in a paper published by the FDA and the American Association for Cancer Research on use of adaptive approaches to optimize drug dosing. [Read AgencyIQ’s analysis of that paper and two others also written by FDA and AACR here.]
• More broadly, the group’s white paper outlines a general model for seamless trials meant to support registration. Successive dose escalation cohorts can identify a target dose that’s used in expansion arms evaluating efficacy across variables like biomarker status and prior lines of treatment. An integrated review of safety and efficacy across those cohorts could, with regulatory support, back a regulatory submission, or provide evidence for designing further tests. The outlined model could also incorporate a randomized assessment comparing a drug to a placebo.

Workshopping challenges to seamless trials

• Panelists who spoke at the Nov. 4 annual meeting described seamless trials as helpful tools, but highlighted challenges to implementing them. Among the challenges discussed were the greater upfront burden in planning a multi-phase study, the risks adaptations pose for a trial’s statistical rigor, and practical hurdles with managing trial sites and drug supply.
• On planning: “You don’t have the luxury of waiting until your Phase 1 or Phase 2 trial concludes,” said HARINDRA ABEYSINGE, vice president and head of oncology in Johnson & Johnson’s global regulatory affairs team. “You have to design it at the outset and things like your dose, patient population, sample size and endpoints have to be well-defined in advance.” Abeysinghe also noted how drugmakers planning a seamless trial need to take care ahead of time to protect the study’s statistical rigor in the event changes are later made. CHRIS TURNER, chief medical officer at Nuvalent, made a similar point: “In a seamless design, you need to focus on how to use other aspects of a patient’s data.” That might mean building assessments of drug response in earlier, he suggested, so a patient enrolled into the first phase of a seamless trial can provide data toward efficacy as well as safety.
• On statistical design: Seamless trials offer developers greater flexibility, especially so if the design builds in adaptive features to change elements of the testing plan. But that comes at a cost, noted YING YUAN, chair of the University of Texas MD Anderson Cancer Center’s biostatistics department. “Whenever we do a new adaptation, it’s a chance to make a false positive,” he said. Here, Yuan was describing the statistical phenomenon known as Type 1 error: The repeated “looks” at data that can happen in certain study designs may falsely make random results support superiority of one trial arm over the other.  Computer simulations can help model how changes may affect the study’s statistical rigor, Yuan added.
• Rapid adaptations can risk turning a trial into a case series, warned LOLA FASHOYIN-AJE, a senior vice president at Parexel and until earlier this year head of the clinical evaluation office within the FDA’s Office of Therapeutic Products. “The risk I see with the implementation of seamless design comes in reacting in almost real-time to information being collected,” she said. “When you’re doing that without a plan, you obviously risk introducing a lot of multiplicity issues in the overall investigational plan,” she added, referring to the problem that can result when multiple statistical tests are run on the same dataset.
• On practical hurdles: Clinical trials depend on sites to run them, and on manufacturers being capable of providing sufficient drug product for patients to take. Both practicalities can become more difficult in a seamless trial, according to Nuvalent’s Turner. The biotechnology executive noted that study researchers often specialize in a specific type of investigation, such as conducting the kind of intensive pharmacokinetic analysis common in early-stage trials, or disease-specific parameters in later tests. “If you’re doing a Phase 1/2/3 trial, the sites you need in the beginning of the trial may not be optimal later,” he said, suggesting sponsors could stagger when sites open.
• Acceleration through phases in a seamless trial can also stress a sponsor’s drug supply. “If the trial is going well and then you need to rapidly expand, there’s a big time lag,” said Turner. “How do you design the right amount of drug … so that you’re not wasting too much money in advance, but you’re not running out of drug and inhibiting the seamlessness and having to pause because you’re waiting for drug to be made?”
• Taken together, the discussion emphasized how sponsors need to invest in a development strategy for seamless trials to work well. “There’s obviously always an urgency in the rare cancer space to develop a product and to do it very quickly, efficiently,” said Fashoyin-Aje. “I do think it’s really important to take a step back and put in the necessary time to develop that comprehensive strategy from beginning to approval. … Even when you’re not in the clinic, there are opportunities to engage with regulators and put forward a plan.”

Analysis: Broader adoption may be around the corner

• The white paper and corresponding panel discussion serve as something of a 101 course on seamless clinical trials. In their white paper, the working group authors write how they hope it will be used as a “practical foundation” to inform trial designs that can benefit rare cancer drug development. The conceptual models they outline in the paper could be of particular interest to sponsors interested in designing their own seamless investigations.
• The gains from adoption could be large. Sponsors are increasingly taking longer to start a new cancer drug trial after finishing the prior phase, according to recent data from IQVIA’s Institute for Human Data Science. Last year, the total of between-trial intervals for an average oncology program was about 10 months, significantly higher than the two to three months IQVIA found was the case at the end of last decade. IQVIA also noted that trial enrollment timelines in oncology are double those of other therapeutic fields.
• Regulators could provide a further tailwind. While Makary’s been somewhat vague in sharing details about his continuous trial proposal, his consistent mentions of the topic suggest it’s fairly high on his agenda, even if not quite ready for release. Cloud-based clinical trial data might provide further opportunity for regulators to monitor and stay ahead of a drug’s development, potentially expediting later review.
• Speaking at the Milken Institute’s Future of Health conference on Nov. 6, Makary again referenced this theme. “’Why do we have phases done in separate giant applications? You don’t go to college … and then apply again for your sophomore year with a 100,000-page application,” he said, using a now well-worn metaphor. “We can use new technology to run continuous trials and have the regulators look at endpoints in the cloud.”
• Regulatory guidance on Bayesian statistics may help. In short, Bayesian statistics allow researchers to compute probability using prior knowledge – such as previously obtained experimental data – and, importantly, to update their calculations as new information is acquired. Applied to clinical testing, this approach can be particularly useful for conducting seamless trials. The ARROW trial, for example, used what’s known as a Bayesian Optimal Interval to identify a drug dose that might not have been selected if a more traditional dose-finding design had been implemented. (See slides 17-21 in this FDA-AACR presentation here for more details, and this previous AgencyIQ analysis for further discussion of Bayesian techniques in dose optimization.)
• The White House’s Office of Information and Regulatory Affairs recently cleared a new FDA draft guidance document on using Bayesian statistics in clinical trials, suggesting its publication is imminent.
• Improving the trial infrastructure around medical innovation may help the oncology field take another step forward. “We have the drugs, but we’re not done yet,” said ROY HERBST, deputy director of the Yale Cancer Center, during a panel later in the Nov. 4 meeting. “We have to find the right patients, the right biomarkers, the right trial design.” Anna Abram, a senior adviser at Akin who spoke on the same panel, foreshadowed Makary’s Milken Institute remarks by calling cloud-based data sharing between industry and regulators “the next revolution in clinical trial development.”

To contact the author of this article, please email Ned Pagliarulo ( npagliarulo@agencyiq.com).
To contact the editor of this item, please email Kari Oakes ( koakes@agencyiq.com).  

Key Documents and Dates

• Seamless Clinical Trial Designs in Rare Cancers: Leveraging Operational and Adaptive Strategies to Accelerate Drug Development
• Friends of Cancer Research Annual Meeting, Nov. 4, 2025
• YouTube recording

https://home.agencyiq.com/article/0000019a-5b33-dcb0-addf-5fff5d810000