This week, BioNTech presented the industry’s most in-depth look so far at
how PD-(L)1xVEGF bispecifics
are being developed in cancer at its R&D Day in New York City.
Following the event,
Endpoints News
sat down with two of the company’s co-founders, CEO Uğur Şahin and CMO Özlem Türeci. The discussion covered BioNTech’s partnership with Chinese biotech Biotheus on the experimental cancer drug pumitamig, how Şahin and Türeci measure success, and why BioNTech ultimately decided to buy out its partner.
Şahin and Türeci also talked about turning that buyout around into a $11 billion-plus licensing deal with Bristol Myers Squibb, an agreement that has the companies set to significantly compete in this new immuno-oncology space. The following interview has been edited substantially for length and clarity.
Max Gelman:
I wanted to start with the Biotheus buyout: Was there something that made you want to acquire it instead of just partnering? And how did that come about?
Uğur Şahin:
It was a combination of things. We had a great experience in the collaboration with Biotheus, and we learned relatively quickly that this is a very competent team. We also understood during the process that this asset (pumitamig) has really a huge opportunity, huge potential. And we wanted to ensure that we can control everything.
Özlem Türeci:
It was clear that we would proceed with the asset, and milestone payments would be an ongoing topic, and it was easier to buy Biotheus instead of pay that amount.
Gelman:
After the acquisition, you made a pretty quick licensing deal with Bristol Myers. Why was it important to do this quickly, and were there any other discussions with other potential partners?
Şahin:
It was because it was two sequential steps. Beginning in 2025, we realized this is really bigger than what we can do alone. We had already started preparation for three registrational trials, but going through the list of potential indications, we came up with 15 potential indications, and we were interested in combinations. We said, if we partner, one key reason is accelerating. Therefore, waiting until the end of 2025, or the beginning of 2026, was just too late.
Gelman:
At the beginning of 2025, was this realization something that happened because of the acquisition, or were you already starting to consider it earlier?
Şahin:
You can say this was a bold move to acquire the company. But at the end of the day, it was completely logical. What was also interesting is that in the first six months of 2025, more and more datasets matured, making it very clear that this is something which is a class of molecules.
Gelman:
Were there other interested parties who wanted to take a more slow, deliberate approach? Or was Bristol Myers’ ability to accelerate the deciding factor in finding a partner?
Şahin:
We were really interested in a foundational partnership, and during the conversation with BMS, it turned out that this partner really checked all the boxes. That does not mean that the other partners would not have qualified in the same way. But at the end of the day, you sign as you proceed, and BMS was just very quick.
Gelman:
How many companies did you end up talking to for a partnership?
Türeci:
No comment. (laughs)
Şahin:
Yeah, no comment. It was significant.
Gelman:
Is there a ballpark estimate you can say?
Şahin:
No, you’d have to define “talking to.” We had significant conversations, but not too many, really. If you take our criteria for a partnership, there are not that many that can fulfill that.
Gelman:
Okay, switching gears now. Can you contextualize these bispecifics with regard to immuno-oncology as a whole? Is this something that’s going to be the next frontier of I-O, or is it more seeing some exciting early data but maybe it doesn’t fully pan out, like
TIGIT
or
CD47
?
Türeci:
It is the next-gen I-O in a very broad sense, that’s how we view it. After the PD-1s and PD-L1s brought all this progress, the next question was, “What is the next-gen I-O?” The TIGIT molecules and so on were not the right thing because of the lack of the pan-tumor aspect. With this class now — and this is not only our opinion — it’s ripe for broad pan-tumor development, and could replace the first-generation PD-1s and PD-L1s.
Gelman:
Why do you feel that combinations are the most potent approach here?
Türeci:
In many indications, there’s a paradigm of combining the I-O with chemotherapy. When we say combine, that means with novel-novel combinations, we want to address these two components of a validated duo. It’s also not sufficient to have just an incremental benefit; it’s really about going beyond this incremental effect. This will be only feasible by combinations which are rationally designed and synergistic.
Gelman:
Is there a number that you would consider to be more-than-incremental, in terms of extending survival times?
Şahin:
It’s the hazard ratio, every hazard ratio which goes below 0.7 is significant. If you can get something which goes to 0.6 or 0.5, it’s transformative. But can we come into a cancer area where we see a stabilization of deaths? In most indications, we are ending up with 20, 30, 40% stability. And the question is, can you go higher? Can you create a stable stabilization at a higher level?
(Editor’s note: In oncology trials, hazard ratios measure the impact on survival times of an experimental drug, when compared to a placebo or control. For example, when looking at progression-free survival, a hazard ratio of 0.7 means the experimental drug improved the risk of disease progression or death by 30%, relative to the comparator. A hazard ratio of 0.6 equals a 40% improvement, 0.5 equals a 50% improvement, and so on.)
Gelman:
Would you consider, then, the 0.7 hazard ratio to be the benchmark for success for pumitamig’s studies?
Şahin:
Success is if you have a hazard ratio which is below 0.8 because, formally, you usually have two endpoints: progression-free survival and overall survival. For PFS, the bispecific class in many indications is in the range of 0.6 or 0.5 hazard ratio, and for OS, if you have a better disease control, this is already a success.
If you can combine that — a better OS reached in a fully powered clinical trial, and a hazard ratio below something like 0.8 — you would say, I have something new. So this is what we want to accomplish. But is it the end of what we want to accomplish? I don’t think so. In our view, it’s a beginning, bringing in a new backbone into immuno-oncology.
Gelman:
You’re testing pumitamig in triple-negative breast cancer, an area where other developers have not disclosed plans and where anti-VEGF therapy has not traditionally been used. How do you see pumitamig’s potential here?
Türeci:
This was already in the package when we started our cooperation with Biotheus. We have seen clear, compelling data in the Biotheus dataset. So this evidence, which we also published, was the reason why we moved ahead with TNBC.
Gelman:
Is there any irony you see in that you’re now competing with Pfizer after having developed the Covid-19 vaccine with them? This was one of the most fruitful pharma collaborations ever, and now in this next frontier of I-O, you’re competing against them.
Şahin:
No, it’s just likely that this happens. If you have such a compound that ultimately could become the next-generation compound, you will compete against the largest pharmaceutical companies. So therefore the likelihood is high that you compete against Pfizer as well.
(Editor’s note: Endpoints spoke with BioNTech on Tuesday. And on Thursday, Bloomberg reported that Pfizer is looking to sell its remaining stake in the biotech.)
