Hi, good afternoon, everyone. I'm Farzin Haque, one of the biotech analysts at Jefferies. It's my pleasure to introduce Yuval Cohen, CEO of Corbus Pharmaceuticals. This is a fireside chat format. Thank you for joining us today.
Thank you for having us.
For those who may be new to the story, can you provide an intro to Corbus?
Sure. Corbus is based in Norwood, Massachusetts, just outside of Boston. We're a very small company. There are about 40 of us. We really like being small. We have a very unusual pipeline. The two assets in development that we get most questions about could not be more different. One is a Nectin-4 ADC targeting solid tumors, specifically head and neck and cervical. The other one is a small molecule oral CB1 inverse agonist targeting obesity. They have absolutely nothing to do with each other. I think that's part of the logic. We wanted to have a very diverse pipeline, and I guess we got one.
Cool. Starting with the oncology asset, you presented some encouraging data in the second line head and neck and cervical cancer at ISMO. For those less familiar with the story, can you summarize the data before we dive deeper?
Sure. We are one of so many companies that licensed an ADC out of China. Our partners are CSPC. CSPC are, I believe, the third largest pharma company in China. They have what we think is a really exciting ADC platform. We licensed the Nectin-4 MMAE ADC. The story starts in early 2024 with the first Chinese data coming out of China, which is CSPC's responsibility. Starting at the beginning of this year, our own Western data started coming out. Importantly, we do not mix the two data sets. We have two completely independent development programs. Ours is primarily U.S. and European sites. Theirs is our Chinese sites.
What we showed at ISMO was really the first time we had what starts to feel like a critical mass of patients from our dose escalation and then dose optimization study. What we had were data on three tumor types of relevance. I'll start with the one that, in a sense, is not actionable for us, but is important certainly in terms of our positive control, and that was bladder. Bladder is the tumor type that was important because Padcev, the poster child of a Nectin-4 MMAE ADC, is approved in bladder. It is within the bladder tumor context that we can compare, especially safety. To pursue bladder as a standalone company is going to be incredibly ambitious, given how dominant Padcev plus Keytruda are in that field.
The other two tumor types are ones that are very actionable for us and data that we showed. The one was in head and neck, looking at monotherapy second line plus. The other tumor type, which is a younger data set, is in cervical cancer, looking again at second line monotherapy. That is probably as good as an introduction as we can.
On the efficacy side, you reported 48% ORR that was confirmed and unconfirmed at 3.6 mg per kg, and then 33% at 2.7 mg per kg. But they were all confirmed.
Yep.
Do you see meaningful differences in efficacy in terms of response and durability between these dose cohorts?
We think so, but it's probably a little bit too early. The number one question that we've now been asked, which is a very sensible question, is, when will we have durability? I'm encouraged by that question because that's really the last question you kind of ask. I think we've managed to convince that we're seeing meaningful clinical responses in both doses. Now the question is, right, are they short-lived? Are they longer-lived? That data set, we think, Farzin, will mature in the middle of next year at a medical conference. By that stage, we should have enough data to look at durability. That's really the last big question to be asked for head and neck. Later on next year, we'll also have the same question to be answered for cervical.
ASCO most likely for head and neck?
I didn't say it, but it's not a crazy idea. It tends to happen in the middle of the year. We're trying to be very disciplined about it and build a routine and a culture where we communicate through conferences. We think that's a very wise way of companies to disclose data and for the data to be meaningful. We have one big last question left. We want to answer it as conclusively as we can.
Got it. And then on the safety side, like you had some rash and peripheral neuropathy, but they were low. And then keratitis was it remained high despite Prophy eye drops. So what have you learned about the Prophy effectiveness and the patient compliance from the site investigators?
The first thing, Farzin, is I'm going to a little bit emphasize the first part, which is we have markedly low peripheral neuropathy in skin. In terms of peripheral neuropathy, remember that that is Padcev's biggest obstacle. In fact, not just for Padcev, if we look at the landscape of these Nectin-4 MMAE constructs and other MMAE constructs that do not even target Nectin-4, it's that peripheral neuropathy that is so devastating. It tends to be irreversible. It sadly tends to be oftentimes untreatable, and it will translate to the patient having to discontinue a drug. Having such a low number is really wonderful. It also fits with emerging data out of China, which is very reassuring. It is also a moat around us for head and neck.
The ability of the other Nectin-4 assets to move into our swim lane in head and neck is probably quite limited because of their peripheral neuropathy levels. As they say, there's no such thing as a free lunch. What we're getting as the price, so why is that low peripheral neuropathy happening? It's because our ADC is so stable. It holds on to that MMAE. What is the cost of that stability? Ocular tox. We see that in many other stable ADCs. It's very similar. It's the surface of the eye, so it's dry eye and keratitis. It's not the inside or the back of the eye. There's sort of a wonderful spectrum. On the lower end of the spectrum is Padcev. We forget, but Padcev, if you look at the label, has about a 44% ocular tox. It's just that nobody cares.
I do not mean that flippantly. It is really not a problem for the patients. It is mild. It does not lead to interruptions or to discontinuations. On the extreme opposite end would be Blenrep, that just got approved, for example, with its whopping 95% ocular tox, of which 77% is grade 3. Yet just got full approval. There is the middle ground. We look somewhere between Tivdak and Elahere, for example. We all use the same strategies. They are wonderfully unoriginal. They all come from recommendations from FDA. What are the strategies? If you have active eye disease, do not enroll that patient into the study. Sensible stuff. Use prophylaxis, as you mentioned. The prophylaxis sounds very fancy. It is really eye drops. That is about as fancy as it gets. It is given prophylactically around the time of the administration of the drug.
If the patient is developing an episode, it is given throughout the episode. Last but not least, what we have seen our physicians do, but what others do, for example, with Tivdak, et cetera, is dose interruptions or dosing holidays, right, to keep the patient on the study while they are experiencing an episode. The episodes tend to be on the earlier side, not always, but they tend to be. It is sort of a wave function. On the other side of that, they tend to resolve. What was interesting is when we looked at the interruption and started to sort of put numbers to them, the actual magnitude of the interruption is really extraordinary. For the 2.7, it is just an extra day of pause on average. For the 3.6, it is just an extra six days of pause. That seems to be enough. It is real.
It's manageable. It tends to be reversible. In the sort of hierarchy of things that would bother oncologists, it's certainly not as high as things like peripheral neuropathy or the type of skin side effects we see with some of the other Nectin-4 MMAEs.
Got it. Just wanted to ask on the discontinuation aspect. There was, you mentioned the dose interruption was like we have like 43% at 3.6. What were the main AEs driving the interruptions? You mentioned that. What about the five patients that discontinued at this dose?
It's usually ocular. I mean, occasionally you'll have a patient that discontinued because something happened to the patient related to their disease. But the drug ones are usually the ocular.
Got it.
What's interesting, Farzin, is we've seen another sort of phenomena evolve. As sites and physicians get used to this, we're seeing a reduction in those adverse events and in those dropouts, et cetera. Initially, when we started, we had to educate. To be fair, we had no data. We had to educate physicians or the sites that this is not just a Padcev me too. It's actually a very different drug. It has certain things that you don't have to worry about that you do with Padcev. On the other hand, you really do have to take the ocular thing prophylaxis seriously. As they get more familiar with it, those rates improve. That, of course, is going to be counterbalanced by patients staying on the drug for longer. We'll see how it evens out.
Makes sense. Enrollment in the head and neck was strong. What does that indicate about the demand in the second line plus head and neck space?
Yeah, there's not a lot there. Second line is pretty grim, sadly. There are currently no good options. It's really sort of try chemo again type of thing or try pembro again. They're grim. They're about 10% ORR and lower. That's what is so attractive to us about the second line. What's interesting is the EGFR therapies might actually make it even worse. Here's what I mean by that. EGFRs are going to dominate that front line, those EGFR combos. That's really where they're going to shine. They're not going to play a role, certainly not a meaningful role in the second line therapy because the assumption is they're not going to work one after the other. If you're a patient in the future, you'll have the choice of potentially three different EGFR therapies. Whichever you want to embark down, that's probably your EGFR therapy.
That's actually going to mean that we might have more patients reaching second line just because more of them survive the front line. At the moment, the second line therapy space in the US is around 24,000 patients. We've seen some modeling from third parties indicating that that might actually increase to as many as 36,000 patients. Again, not because there are more patients with head and neck. It's because more of them survive that front line thanks to EGFR. Certainly, our really above average, to say the least, our incredibly encouraging enrollment is a testament to the fact that physicians are telling us once a patient has failed front line, there's really not much to offer them.
In terms of competitive space, any other ADCs they're looking after in the second line or any other mechanism that could fit?
Certainly on the ADC, I'll start by who's not there and then talk about who's there. Pfizer's two attempts, one with Padcev and the other one with their PD-1 ADC, are not progressing forward. That's our understanding from Pfizer. What remains are Pyxis with a very different modality. We'll see what that data looks like. It is completely orthogonal to what we do. Like us, it's orthogonal to what the emerging front line is going to look like. That should be very interesting to look at. As I mentioned before, it is unlikely that the other Nectin-4 MMAE constructs can move into head and neck because of their peripheral neuropathy. That doesn't mean they won't, but it just means it's unlikely. If they do, it's probably not quite as attractive.
The Topo 1s are probably not attractive just because of the track record of Topo 1 in general in head and neck.
Makes sense.
We try.
Are you seeing any notable differences in response between HPV positive or negative patients or PD-1 high or low? Because that was an issue with the EGFR targeted therapy.
Very much. The answer is no. Now, again, small numbers, emerging data set. As we get more patients, we're very keen on stratifying that. We're not seeing differences. We're seeing responses in all. Theoretically, this is so wonderfully orthogonal, right? It is a Nectin-4 targeting agent. Those should be insensitive. If anything, there's a pretty robust literature that HPV positivity correlates with an overexpression of Nectin-4 as one of the oncogenes that gets overexpressed in the HPV positive patients. We've certainly seen responses in HPV negative as well.
Nice.
Farzin, as you think about it, certainly opens some interesting doors beyond the second line therapy, right? There are emerging potential blind spots even in that front line therapy. The two that you mentioned are certainly two very sensible ones. That is before we talk about what would our data set look like in our combo study that started this summer with Keytruda.
When will you see that data?
Next year, we have three data sets, as I mentioned. The first one will be monotherapy, second line plus middle of the year, cervical, and then the combo data later on. They are all coming in the second half of next year.
You also talked about the registration study start potentially in mid-2026. What could the trial design look like versus Mersana's second line registration study?
Hopefully exactly the same. That is, it is one of those cases where we do not want to be particularly creative. Mersana and then later BioAtla with their subpopulation are both doing the same very sensible design. We would be quite keen to have the same design. We're delighted that FDA has already granted us fast track. That makes that conversation, that dialogue even more productive. We look forward to updating everyone on that, hopefully in the first quarter.
Got it. We have just a quick question. A big sales estimate for you in the second line?
It's super. I mean, you can do some interesting back of the envelopes. But if you think about 24,000 patients in the US at the moment and figure the same thing in the EU, and if you benchmark that to the price of standard ADCs and/or Padcev, without those words coming out of my mouth directly, it's a meaningful market. Unlike cervical, where we acknowledge is a much smaller population, right? Cervical in the US is about the metastatic refractory is about 3,000 women a year. Having said that, astonishingly, wonderfully, Tivdak, with all its limitations and challenges, both in terms of efficacy and tolerability, Tivdak made $300 million last year worldwide, which I would it's not a number you would normally have thought they would make. Yeah, head and neck is certainly much larger than that.
Got it. Switching on to the 913 asset. That's in development, starting with the SAD and MAD data. What can you share about the latest progress so far?
Yeah, we'll have SAD, MAD data coming out before the end of the year. That's almost here. A couple of important reminders. A, it's a SAD, MAD study. Remember, everyone, this is not oncology now. SAD, MAD and the rest of BioPharma are usually healthy volunteers, right? They're not patients. There's nothing wrong with them. Why do you do SAD, MAD? You want PK, you want safety. It involves very, very high doses. Just to give you a bit of an idea of the scale, Rimonabant was efficacious at 20 milligrams a day. Malonabant is shockingly efficacious at 10. SAD, MAD can go up to hundreds of milligrams a day. You'll never give that to a patient. You're doing it in the context of a SAD, MAD because you want to see the PK. You want to push the envelope.
You want to see what makes this thing tick from a safety and tolerability. It is in the phase 1 unit in the U.S. We only do studies, and we will always only do studies in obesity in the U.S. It's a little obsession that we have, but it has served us well. Non-US obesity data tends to be very odd. We don't want that. The MAD is just seven days of dosing. It's pretty much the fastest MAD you can get. The idea behind it is to open the door to a 90-day study that will start before the end of the year, which is in obese non-diabetics. That study is all the things you would want to see. A, what does the safety look like over 90 days? Ninety days is plenty to catch all the safety that worries us. That's the first one.
The second one is in 90 days on a CB1 inverse agonist, obese patients will lose weight. The reason I can say it with such conviction, every single CB1 inverse agonist ever tested in humans has always led to weight loss in 90 days. I think the interesting question, though, is how much weight loss, right? Rimonabant is really surprising. And remember, we are cousins of Rimonabant. In DIO mouse studies, which we ran with Rimonabant, we made Rimonabant, we ran it in DIO. In DIO, it looks like Rimonabant. And we look like Rimonabant. And pretty much, I'll be honest, everything looks like Rimonabant. What was so shocking with Rimonabant is how spectacularly effective it is in humans. It's a lot more potent than Rimonabant.
To give you an idea, Rimonabant at a year, 20 mg a day, was maybe, maybe 6.5% on a good day, maybe 6.5% weight loss. Back in the day, that was exciting. Today, that's less exciting. Rimonabant at half that dose got there in four months. Rimonabant looks about as efficacious as Orforglipron, maybe more efficacious than Orforglipron. That's very surprising because nobody knows why. We have no idea how efficacious we'll be. We will see weight loss. There's no question about it. We just can't figure out how much weight we'll see. There's only one way to do it. We got to run the experiment. It's that simple.
What is the minimum threshold to be clinically meaningful?
Oh, that's easy. The FDA tells you that. That is 5% in a year. Is 5% commercially viable in a year? No.
Just checking back on the safety, the initial SAD data is for seven days only, you mentioned, right?
SAD is a single day.
Single day. MAD is seven days.
Seven days of dosing, correct.
On the neuropsych AEs, what should we expect? Because the time frame is just too short.
It's a very short time frame. Now, the Rimonabant data had paper very helpfully provided Kaplan-Meier curves for AEs. We will never show that. Nevertheless, that was very nice of them. They had plenty in the first week. What was really surprising about Rimonabant, one of the most surprising things about their data, and it was not there in the press release, was how much GI they had. They had a 50% dropout rate in their top dose, most of it driven by GI. The GI was 95% happened in the first week. Again, we have no idea why they had so much GI. It is very atypical of a CB1 inverse agonist. If our GI is quiet in a week, I feel much better about that. We'll see.
Have you mentioned how many dose levels you have planned for the phase 1 B study?
We have not. Everyone's assumption is we're basically copying the design of Novo. It's a very sensible design. So there's placebo. There's a number of cohorts. They did 120 days. We're going to do 90 days. Honestly, 90 is more than enough. With the Kaplan-Meier's for the adverse events, after a month, pretty much nothing happens.
Got it. And then just quickly on that, we have two minutes left, on 601 program.
Yep.
When do you plan to share from the phase 1 study results?
We had a study in progress at SIDC. We had a poster. It's halfway through. We'll wrap it up. We're setting a very high bar for it. The bar we're setting is we'd like to see some monotherapy. We just don't have those higher doses yet. It is IO, and I think all of us in this industry have become very skeptical of IO assets without monotherapy. We'll see what we get. The other thing is we're really quite keen to see what Pfizer got. Pfizer finished their study last December. The Pfizer monoclonal antibody is very similar to ours, right? We're both targeting alpha v beta 8. They finished a largish study, several hundred patients, and it's been awfully quiet. Yet the program is still live. We're quite keen to see what they got so we can, as it were, compare notes.
Yeah, it's not as a burning priority for us as are the two other assets. It's also the one where we actually don't know if it's going to work or not. The other two, I think we're pretty comfortable they're going to work. The question is, have we solved a certain problem? The latent TGF beta hypothesis is still out there for the jury to decide on.
Why hasn't Pfizer disclosed the data yet?
I don't know. Ask Pfizer. I mean, quite seriously, we're quite, in other words, why have they not disclosed the data but not killed the program? Don't know. It's an interesting question.
Almost out of time to wrap up, maybe talk about the cash position and then the key catalyst that investors should focus on for the next year.
We have just under $180 million in the bank now. As I have mentioned, three key data catalysts next year for 701, and then a very big catalyst for 913 in the middle of the year, which is the 90-day dose range finding study. That should keep us busy.
Definitely. Always a good conversation. Thank you all.
Farzin, thank you so much. Thank you, everyone.