What I Heard When NAMs Became a Conversation, Not a Webinar
By Ray Dogum, Chief Editor, Drug Discovery Online
I attended a unique kind of event yesterday: not a webinar in the usual sense, not a panel, and not another slide-heavy session where the audience waits quietly for the Q&A box to open. Advancing NAMs Together: An Interactive Exchange of Insights & Expertise, hosted by the American Association of Pharmaceutical Scientists (AAPS) and BioIVT, was designed around participation.
From Presentation to Participation
After a brief setup on the current landscape for new approach methodologies (NAMs), attendees were asked to talk 1:1 with each other (randomly assigned) for 10 minutes at a time. We were asked to compare experiences, test assumptions, and surface the practical questions that still shape whether these tools make it into real decision-making.
That structure made it feel safe to be blunt and honest about the current state of NAMs. Most people in this space already believe that human-relevant models, organoids, microphysiological systems, computational tools, and predictive toxicology have enormous potential. The harder and more useful conversation is about where they are (or aren’t) already having impact, what kind of evidence makes them credible, and why adoption remains uneven across organizations, use cases, and regulatory environments.
The Real Barrier Is Behavior Change
What I heard repeatedly was that the barrier is not primarily scientific. It is cultural, operational, and sometimes psychological. One participant from a major CRO framed the biggest challenge as the limited appetite to do something different after decades of relying on familiar animal-based approaches. He told me his clients, “don’t really have the appetite to do something different from what they’ve been doing for the last 50 years. How do we get from here to there? Do we need to do both [animal studies and NAMs]?”
That point captures the real tension in the field. NAMs are often discussed as replacements, but in practice, many groups are still asking a more cautious question: can these methods add confidence where uncertainty is high? Let’s face it; animal testing is not predictive for humans.
For toxicologists and preclinical safety teams, context of use came up again and again. A model may be scientifically impressive, but that does not automatically make it useful for a specific decision. Variability, technical characterization, and budget realities all influence whether a method can be implemented responsibly. In CRO-driven work especially, NAM adoption can be difficult when the need for deeper characterization was not built into the original plan.
Another theme was visibility. Sponsors and researchers are not always trying to solve the same problem, and technology developers may not have a clear window into the specific uncertainties that matter most to decision-makers. At the same time, the number of available platforms keeps growing. Without better ways to benchmark performance, match technologies to use cases, and define what “fit for purpose” really means, the field risks creating more choice without creating more clarity.
Trust Has to Be Built Case by Case
The regulatory conversation was just as candid. Some participants saw European agencies as more open and adventuresome to NAMs whereas the FDA is more conservative in applying them to consequential decisions. Whether or not everyone would describe the landscape that way, the perception itself is important. Developers need to know what kind of data will be trusted, when animal studies can be reduced or avoided, and how evidence from human-relevant systems can support safety evaluation, excipient qualification, or concerns such as nitrosamine impurities in drugs.
My biggest takeaway is that advancing NAMs will require more than enthusiasm for innovation. It will require a shared inventory of the uncertainties: where animal models are weakest, where human-relevant systems are strongest, where orthogonal assays can build confidence, and where regulators, sponsors, CROs, and technology developers are still speaking slightly different languages. The promise of digital twins, predictive toxicology, and human-based models is real, but promise alone does not change practice. Trust is built through use case studies, standards, transparent limitations, and repeated proof that a method can answer the right question at the right time.
The organizers informed us that the conversations we had were transcribed and analyzed using AI to generate a summary was shared with participants. A recording of the event will not be available; however the perspectives were captured anonymously to identify recurring themes, differing viewpoints, and questions that remain unsolved. To make NAMs translationally relevant at scale (and since I’m a fan of speed networking), we need more of these kinds of interactive exchanges.