Welcome to Partnology’s Biotech Leader Spotlight Series, where we highlight the remarkable accomplishments and visionary leadership of biotech industry pioneers. This series is about showcasing the groundbreaking strides made by exceptional leaders who have transformed scientific possibilities into tangible realities. Through insightful interviews, we invite you to join us in following the inspiring journeys of these executives who continue to shape the landscape of the biotech industry. This week we are recognizing:
Ben Buelow is CEO and Co-Founder of 92Bio. Ben and the 92Bio team are bringing their industry leading expertise in therapeutic antibody discovery and development to create the next generation of multi-specific antibodies for the treatment of cancer. He previously served as CEO of Ancora Biotech (2022-2024) and Chief Medical Officer of Teneobio from 2016-2021, when it was acquired by Amgen. He has guided five antibodies from concept to clinic since 2016, with two now in phase 3 registrational trials. He has published ~25 peer-reviewed papers and has presented at numerous national and international scientific conferences. Ben completed his residency and surgical pathology fellowship as chief resident at UCSF in 2016. He received an MD with honors and a PhD in Immunology from the University of Washington in 2011.
You helped build Teneobio into a $900M acquisition and then led Ancora. After those experiences, what problem did you feel still wasn’t being solved that made you start 92Bio?
What led us to focus on developing multispecifics for solid tumors is the belief that T-cell engagers still haven’t reached their full potential in that space. We saw an opportunity to take on one of the field’s biggest challenges.
For our size, I believe we have one of the most experienced teams anywhere working on concept-to-clinic T-cell engager development. I think our track record speaks for itself. When you look at the number of molecules we’ve advanced into the clinic, the large partnership deals we’ve completed, and the fact that at least one—and likely two—of those programs appear to be on track to become approved drugs, it’s a unique accomplishment for a team of our size.
What’s especially exciting is tackling difficult scientific problems with a small, highly experienced group. Being lean allows us to stay nimble, move quickly, and make decisions efficiently while ensuring everyone has a voice in the process. That combination of speed, collaboration, and deep expertise is a real advantage.
Our experience isn’t just about designing innovative T-cell engagers from a scientific standpoint—it’s also about efficiently advancing those molecules into the clinic and maximizing their probability of technical success. Translating a promising concept into something that actually works in patients is where much of the challenge lies. As they say in the patent world, it’s about reducing the idea to practice. That’s an area where our team has developed a particularly strong and differentiated skill set.
That expertise is especially relevant in solid tumor T-cell engager development. It’s easy to look at the success T-cell engagers have had in hematologic malignancies, where combining a reasonably effective CD3 binder with a well-functioning tumor-associated antigen binder often produced a functional molecule. I’d say we’re seeing a similar trend with the shift into autoimmune diseases using many of those same approaches.
Solid tumors are different. The margin for success is much narrower. The geometry of the molecule, its architecture, how it’s assembled, and the specific building blocks you choose all become critically important. You have to be far more precise in how you design these molecules.
That’s why we believe our experience is so valuable. It allows us to apply a level of precision and development expertise that’s particularly well suited to solving the unique challenges of T-cell engagers in solid tumors. For us, that’s both an exciting scientific opportunity and an area where we believe we can make a meaningful impact.
You recently dosed the first patient in the Phase 1 study of NTB-928. What were the biggest challenges getting from idea to first patient?
The first thing that came to mind when you asked that question was that it’s a bit like asking someone to choose their favorite child—which part was the biggest challenge? Without being too flippant, I’d say that every stage of a drug’s lifecycle, including for NTB-928, presents its own unique challenges. They’re so different from one another that it’s hard to compare them directly or say which is the most difficult. But using that framework, I can break it down by stage.
From a business perspective, the biggest challenge was simply getting the drug back to a point where it could continue on its path toward the clinic. NTB-928 was originally developed as TNB-928 at Teneobio. When Teneobio was acquired by Amgen, that program went with most of our other assets. However, for strategic reasons, Amgen decided in 2025 not to continue developing the molecule, while we remained convinced of its potential.
That meant the Amgen team and our team had to develop a framework for bringing the asset back under our umbrella. Amgen was an excellent partner throughout that process. They were flexible and genuinely wanted to see the program continue under a new organization. Even so, it required a great deal of creativity and collaboration to structure an agreement that met everyone’s needs while allowing the program to move forward.
From a scientific and preclinical development perspective, the challenge centered on the target itself. FRα is a validated antibody target—as evidenced by the success of multiple antibody-drug conjugates—but there are currently no other FRα-targeting T-cell engagers in development of which I’m aware. That’s because the target presents theoretical risks of on-target, off-tumor toxicity.
Our challenge was designing a molecule specifically to overcome those risks. We describe that work extensively in the Avanzino et al. paper outlining the preclinical development of NTB-928. I think the final molecule is really a testament to the Teneobio team’s engineering capabilities. More broadly, it reinforces the idea that how you build a T-cell engager can be just as important as which target you choose.
On the IND-enabling side, we faced a different challenge. Like all of the Teneobio molecules, the T-cell engaging arm of NTB-928 is not cynomolgus monkey cross-reactive. That’s a somewhat unconventional approach among T-cell engager developers, although I believe it’s well supported by the data.
A well-known paper by Saber et al. from the FDA essentially concluded that, regardless of whether you have cynomolgus monkey data, first-in-human dosing for T-cell engagers should be based on the MABEL (Minimum Anticipated Biological Effect Level) approach. We incorporated that guidance into our development strategy. The key challenge was ensuring that regulators understood we had assembled a robust safety package and that our proposed starting dose was both scientifically justified and appropriately conservative. At the same time, we wanted to position the program to reach therapeutically relevant dose levels as efficiently as possible.
That naturally leads into the clinical challenge. Because T-cell engagers typically use a MABEL-based starting dose, you begin treatment at a much lower dose than you would for many traditional therapeutics, which often rely on the NOAEL (No Observed Adverse Effect Level) to determine the starting dose. As a result, you need many more dose-escalation steps before reaching potentially active dose levels.
Designing that escalation strategy becomes critically important. You need to protect patient safety while also moving efficiently enough through the dose levels that you don’t enroll an excessive number of patients at clearly subtherapeutic doses. Striking that balance between safety and efficiency is one of the defining challenges of early clinical development for T-cell engagers.
Looking beyond ovarian cancer, what excites you most about where this platform could ultimately go?
If you look at our track record and the evolution of 92Bio, it should come as no surprise that the goal has never been to build just one product. The vision has always been to create a platform that enables the rational assembly of different molecular modules to develop therapies for solid tumors with a high probability of technical success.
I’ve touched on this in some of my earlier answers, but we truly believe that how you assemble a molecule—which binding moieties you choose and how those components interact with one another—is where the real “secret sauce” lies. That’s especially true for difficult indications like solid tumors and AML.
While we’re very product-focused and advance one molecule at a time into the clinic, each clinical program generates data that helps us better understand what our preclinical work was actually telling us. Those insights then feed directly back into the platform, creating a continuous cycle of learning and innovation.
Ultimately, what we’re building is a feedback loop where we can take what we learn about T-cell engagement, immune cell activation, co-activation and co-stimulation, and solid tumor immunobiology, and apply those insights to design even better molecules in the future.
To me, that’s where the most exciting opportunity lies. Alongside others working in this field, we’re beginning to unlock what it really takes to design T-cell engagers capable of delivering the depth and durability of response we’ve already seen in hematologic malignancies—and ultimately bring that same level of success to solid tumors.
If someone looked inside 92Bio today, what would they notice that’s intentionally different from every other early-stage biotech?
I want to preface my answer by sharing something one of my mentors told me early in my career. He said that if someone else has had—or is pursuing—the same idea you have, that’s actually a good thing. If you come up with what you think is a novel idea and then discover someone else has had it too, you should feel encouraged because it validates that it was probably a worthwhile idea. On the other hand, if you’re the only person who’s ever thought of something, you should ask yourself why. The most likely explanation isn’t that you’re the first person to think of it—it’s that other people have already tried it, and it didn’t work.
Because of that mindset, we start with the assumption that other people have good ideas. We’re not shy about looking at what others are doing—whether scientifically or operationally—and saying, “That’s interesting. Maybe we should try something similar.” We pay attention to what’s happening in the field, read the literature, and learn from what others have done.
That’s especially true when it comes to how we run the business. Scientifically, there’s a stronger incentive to push outside your comfort zone and pursue genuinely novel approaches. But when it comes to building and managing a company, a lot of smart people—particularly in larger biotech and technology companies—have already experimented with different ways of operating. We regularly learn from that work, whether through the literature, consultants, or conversations with experienced operators. Some ideas work well for us, others don’t, and we’re perfectly comfortable discarding what doesn’t fit.
That said, I think there are a couple of things that really distinguish how we operate.
The first is an unusual level of transparency. We believe our team is our greatest asset. If we hire someone, we trust that they have both the intelligence and the judgment to be part of the inner circle. We’re not a company that believes in keeping secrets from one another. In fact, I think we share more information across functions than many companies our size.
We don’t believe in silos. If you want a continuous feedback loop of innovation, people need to understand what’s happening across the organization. That extends well beyond the scientific teams to functions like finance, operations, and IT. There’s also a long-term benefit: it’s how you leave people better than you found them. If they never gain exposure to parts of the business outside their own function, they never have the opportunity to grow into broader leaders.
The second principle is something we’ve borrowed from the Silicon Valley technology ecosystem: becoming comfortable taking calculated risks when the expected value is high. Our philosophy is to innovate in areas that are promising but still underexplored or technically unoptimized.
I’ll give you a few examples. When Teneobio first entered the T-cell engager space, we certainly weren’t the first company to pursue T-cell engagers. But during our discovery work, we identified a naturally affinity down-tuned CD3-binding domain. At the time, others had explored affinity tuning, but no one had been willing to take that approach into the clinic. We did, and ultimately benefited from making that decision.
We’re following the same philosophy at 92Bio. NTB-928 is a good example. Many groups have looked at FRα and concluded that the risks outweigh the opportunity for T-cell engagement. We see those same risks as engineering challenges. If you can solve them, we believe FRα becomes an exceptional target.
Our second program, NTB-921, which engages gamma-delta T cells, follows a similar pattern. Other groups have explored this biology, but when we look at the field, we see technical limitations that we believe can be overcome. Our goal isn’t simply to work on established ideas—it’s to solve the problems that have prevented those ideas from reaching their full potential. More broadly, that’s how we think about innovation. We look for opportunities where the key risks have already been identified, because solving those risks has the potential to unlock enormous value.
There are plenty of organizations willing to explore high-risk ideas in the preclinical setting and generate exciting molecules. Far fewer are willing to make the much larger commitment required to test those ideas in patients. Historically, that’s been one of our differentiators, and it’s a philosophy we’ve carried into 92Bio. If we believe an idea is compelling enough, then we believe it’s worth putting to the ultimate test—evaluating it in patients.
Multi-specific antibodies have become one of the hottest areas in oncology. Where do you think the field is still underestimating the biggest opportunities—or biggest engineering challenges?
I think the biggest opportunity being underestimated for T-cell engaging multispecifics is in solid tumors. A narrative has emerged that T-cell engagement simply isn’t well suited to solid tumor biology, physiology, or fluid dynamics. But from my perspective, the hard data supporting that conclusion are still fairly limited.
We’re still in the very early stages of antibody-based therapeutics for solid tumors, whether you’re talking about ADCs, T-cell engagers, or even CAR T cells. CAR T cells are antibody-based in the sense that the chimeric antigen receptor itself is typically derived from an antibody. Across all of these modalities, the number of molecules that have actually made it into the clinic remains relatively small.
Because of that, I think the field risks abandoning certain approaches too early based on narratives like, “ADCs work here, but T-cell engagers don’t.” I’m not convinced the data actually support conclusions that broad. In fact, that disconnect is exactly the opportunity we’re pursuing.
The biggest engineering challenge in solid tumors is that there is no one-size-fits-all solution. You have to be very careful not to overfit your interpretation of success or failure from a small number of programs.
Ultimately, the clinical data are what matter. Preclinical systems are inherently artificial, and depending on how you design an experiment, you can often generate the answer you’re looking for. The real question is whether a molecule benefits patients. But because clinical development is expensive and time-consuming, those datasets are inevitably small. That means we should be very cautious about extrapolating the performance of one molecule to an entire target or therapeutic modality.
PSMA is a good example. A number of T-cell engagers targeting PSMA failed, and for a period of time the conclusion was that PSMA was simply a poor target for T-cell engagement—and perhaps for ADCs as well—and that radioligand therapy was the only viable approach.
Then Amunix developed what is now VIR-5500, which appears to have challenged that assumption. The lesson is that the target itself may not have been the problem. We may simply not have found the right molecular architecture or engineering solution yet.
That’s why I think the next generation of T-cell engagers requires a much more sophisticated approach. It’s no longer enough to take your favorite CD3 binder, combine it with your favorite tumor-associated antigen binder, and perhaps add a co-stimulatory element.
You have to think very precisely about the target, the indication, the architecture of the molecule, and how each component interacts with the others. The goal is to design the molecule specifically for the biology you’re trying to address and give it the best possible chance of succeeding in patients.
You’ve now helped move four antibody programs from concept into the clinic, with two reaching Phase 3. What decisions during early development have you found determine whether a program eventually succeeds?
I really believe that this idea of reducing innovation to practice—moving drugs from concept to clinic and designing molecules with a high probability of technical success—is where our team has built its reputation. I’ve spent a lot of time thinking about how to turn that philosophy into a repeatable system that we can continue to leverage as we grow.
I’ll just rattle off a few ideas that I hope are useful. They’re probably not new to everyone, but they’re the principles that have stood out most to me.
The first is to be relentlessly data-driven. You have to strive to be dispassionate and let the data guide your decisions. That’s something everyone says, but it’s much harder to do in practice. Define your criteria for success and failure before you run the experiment—not after you have the results. And don’t be afraid to terminate programs if the data aren’t compelling.
At the same time, be willing to bet on differentiated ideas, even when they challenge conventional wisdom. F2B is probably the best example of that. It’s the CD3-binding domain we developed at Teneobio. At the time, many people questioned why we’d prioritize a molecule with a better safety profile rather than simply maximizing activity. The prevailing view in oncology was that the tumor is what ultimately kills patients, so toxicity can be managed.
We believed something different. Our hypothesis was that better safety would create a wider therapeutic window, and that wider therapeutic window would ultimately allow us to achieve better efficacy. That was a fairly unconventional idea at the time, but it proved to be the right one. The same philosophy informed our work on CD19 and continues to drive programs like NTB-928 and NTB-921 today.
From a discovery perspective, we’ve found that one of the most effective ways to create value is to identify a specific technical challenge within a field and then design the simplest possible solution to overcome it. If you can successfully solve that problem, you’ve often created something that can extend beyond a single molecule and become the foundation of a platform.
Again, F2B is the classic case study, but the same principle applies to our gamma-delta T-cell platform. NTB-921 is being developed initially for one indication, but if the underlying engineering strategy proves successful, there are many additional indications where those same advantages could apply. That’s how you create a virtuous cycle—using the success of one program to accelerate the next and continually improve your platform.
Another lesson is to think about clinical development from the very beginning. Even during early discovery, you should constantly ask yourself how today’s design decisions will influence how the molecule ultimately behaves in patients.
Practically speaking, that means engaging regulators early. Use FDA Type B and Type D meetings to get feedback on areas where you’re uncertain. Don’t wait until you’ve locked yourself into a development strategy before seeking input.
Similarly, design your clinical trials with the end goal in mind. For me, there are really three questions every development program has to answer:
- Does it work?
- Is it safe?
- Can it be dosed in a practical way?
It’s easy to become consumed by biomarkers and mechanistic questions. As scientists, we’re naturally curious about exactly why a drug behaves the way it does, and those questions absolutely matter. But particularly in a small biotech, the first priority is demonstrating clinical benefit. Does the drug produce meaningful responses in patients? Once you’ve answered that question, then you can spend time understanding the underlying biology. Until then, much of the mechanistic detail is secondary.
Another lesson is that relationships and communication are some of the most valuable assets in drug development. Whether you’re working with CROs, investigators, clinical sites, consultants, or vendors, everything ultimately comes down to people.
It sounds almost simplistic, but it goes back to something we all learned in kindergarten: treat people well. Be respectful. Remember that there are human beings on the other side of every interaction—including, most importantly, the patients receiving your therapies.
That naturally leads to another principle: make the patient and provider experience central to your Phase 1 trial design. If patients want to participate in your study and investigators want to enroll patients, everyone benefits—including your company. A well-designed study isn’t just scientifically rigorous; it’s also practical and patient-centered.
The final point is about people.
Retaining great talent sounds like one of those phrases you’d find on a motivational poster, but I think it deserves much more attention. Let people participate. Give them real ownership. Listen to ideas from every level of the organization, including your most junior employees. Be transparent. If you trust people enough to hire them, you should trust them with meaningful context about the business.
I also believe strongly in the “campsite rule”—leave people better than you found them. Give them opportunities to learn, grow, and take on responsibilities they haven’t had before. Keep your organization as flat as possible, empower people to make technical decisions, and when those decisions lead to success, make sure they receive the credit.
There’s obviously a lot more that could be said about building successful biotech companies, but those are the principles that have shaped how we’ve tried to build ours.
