All right. It is my pleasure to introduce Claire, CEO of Bicara, and AJ, the Head of Corporate Development and Chief of Staff for Bicara. So welcome. I believe this is your first Cowen-Sachs conference?
It is. Thank you for having us.
All right. So welcome to the first conference, and it's a pleasure to be hosting you. Before I kick off with the Q&A, I'll let you guys introduce the company and also kick off with the opening remarks.
Happy to. As Richard highlighted, my name is Claire Mazumdar. I'm the CEO of Bicara. Bicara is a clinical-stage biotech company focused at this intersection of what we call targeted tumor modulation. We have developed bifunctionals or multispecifics that are really meant to drive this concept of tumor penetration. This is really exemplified by our lead program, which is known as ficerafusp alfa, but I'll refer to it as ficera, which is an EGFR-TGF beta trap. The real idea behind this molecule is to use the EGFR arm to target to the tumor itself and drive the TGF beta inhibition within the tumor microenvironment and allow that tumor microenvironment remodeling that allows the immune cells in synergy with immunotherapy to really go after and penetrate the tumor.
Our focus has predominantly been in solid tumors that have a very fibrotic tumor microenvironment or immunosuppressive tumor microenvironment, with a perfect example of that being HPV negative head and neck recurrent and metastatic head and neck. This is a tumor type where we know that even in the frontline setting, immunotherapy has pretty poor response rates and overall survival. Pembro mono therapy in the recurrent and metastatic setting only has about a 19% response rate and a 12-month median overall survival. It actually looks far worse in the HPV negative subset, where median overall survival is actually closer to nine months with pembro mono therapy. Our focus has been to demonstrate, even in the frontline setting in combination with pembro, improved response rates, but most importantly, improved durability that would lead to outsized overall survival.
About two weeks ago at ASCO, we shared some really exciting data in frontline, recurrent, and metastatic HPV negative head and neck, where we saw an immunotherapy-like duration of response, at more than 21 months, which is in some ways unheard of in such a large patient population. 50% of our responses had these median DOR beyond 21 months, which led to an outsized overall survival of 21 months. Maybe just the last big headline from that data set that we presented was really we're seeing consistent depth and durability. 80% of our responders showed greater than 80% depth of response or greater with a 21% complete response rate. That is what really led to an outsized overall survival.
Fantastic. Yeah.
Maybe just I think you maybe asked the question about 80% tumor shrinkage in one of our calls. I think one of the interesting things about the depth is how it's translating to that durability. I know we had provided a subsequent analysis that evaluated what does the durability progression-free survival in OS look like amongst the different types of responders, those with those characteristic deep responses versus those without. One of the interesting, I think, pieces to tie it all together is you're seeing that outsized durability that Claire mentioned in those deep responses, which is translating to that improvement in PFS and OS.
Got it. We'll be diving deeper into that discussion later. Just to cut back to Bicara's overall development, you guys have a phase II/III study ongoing. Maybe just a quick status update on how that's been going. You guys also have an umbrella of other studies looking at different types of head and neck, I think CPS, CMR.
Yeah, happy to speak to that.
Maybe just go over that too.
Yeah. As you alluded to, Richard, given the exciting data we saw in our phase I-B, we started to build a confidence that there was a path to an accelerated approval in the frontline setting of head and neck. Earlier this year in February, we dosed our first patients in our FORTIFI-HN01 study. This is a seamless phase 2-3 study comparing, in a blinded fashion, so a double-blind placebo-controlled study where we're looking at ficera plus pembro versus pembro plus placebo. This study is designed in such a way to look at an interim analysis for accelerated approval with the ultimate confirmatory endpoint being overall survival. Beyond that pivotal study, we are looking at a number of other subsets of head and neck where we've seen some exciting early signs of activity.
These include HPV positive smokers, which is one that we had seen some, in fact, two complete responses early on in our phase I-B and are looking at a larger cohort. The other one we're quite excited about is actually the CPS zero recurrent and metastatic setting. This accounts for about 15% of the frontline market where pembro monotherapy has no responses. Typically, these patients seek chemotherapy regimens. Even on chemo, the response rates are quite low. What builds confidence that we're excited about this data set is at ASCO this past two weeks ago, we shared actually very strong responses, durability, and overall survival in the CPS low, which are the 1-19 that were consistent with the CPS high.
I think that's, again, unique to our underlying mechanism of action where the TGF beta arm is driving responses even in the more immunosuppressive CPS low. We think that the CPS zero category, which will be about a 20-patient data set that we plan to disclose next year, will speak to the potential in these more immunosuppressive phenotypes.
Got it. Okay. Before we jump into the very exciting data, can you talk about the background of ficera , how it was developed, the history of it, what was the why these EGFR and the TGF beta, and then what was the original application it was intended for?
Yeah. The focus was always to go after solid tumors, the squamous-like carcinomas that have a strong rationale for the inhibition of both EGFR and TGF beta. What we know today is typically squamous-like carcinomas have greater than 90% EGFR overexpression. These include, of course, head and neck, but colorectal cancer where we've seen EGFR use, squamous cell carcinoma of the anal canal, and cutaneous squamous cell carcinoma. Pancreatic cancer is another area as well that has high EGFR overexpression. This molecule was really designed to go after tumor types where we see very immunosuppressive tumor microenvironment. This comes from this concept. We saw the first wave of IO therapies, but what people started to realize is that not all patients responded to IO therapies, and those tended to be in very immunosuppressive or fibrotic tumor microenvironments.
I would say there became this class of companies that were focused on really remodeling the tumor microenvironment where ficera fits in. The concept here was to use two well-validated oncology targets, EGFR and TGF beta. A key issue with TGF beta inhibitors in the past has just been the inability to show high efficacy where it mattered at the tumor. There has never been a tumor-targeted TGF beta inhibitor before. That was really where ficera came in, used the EGFR arm to maintain the functionality of an EGFR inhibitor, but direct that TGF beta inhibition where it mattered at the tumor.
What we showed about two months ago at AACR is that we're actually able to see that tumor microenvironment remodeling where typically in these immunosuppressive tumor microenvironments, you see almost a barrier of immune cells that are unable to penetrate the tumor. When you treat with ficera plus a PD-1, you almost get that one-two punch to dig into the tumor. That is what we always hypothesized would lead to deep and durable responses. This was really designed for tumor types like head and neck where that durability was going to be the ultimate predictor of prolonged overall survival.
I see. Got it. Based on the ASCO data, again, very impressive data with 21 months of OS never seen before, and then very deep responses across the patients that responded. What do you think is, when you look at the EGFR and also the TGF beta, and then you guys also have ADCC functionality across these different mechanisms, how much are the different arms contributing to it? How much do you think the TGF specifically, TGF beta arm is adding to the EGFR and pembro? How is that adding to the, and also in what ways, to the data that we saw?
Our hypothesis was always in many ways that maintained ADCC from the EGFR arm of the molecule, so cetuximab, was really going to drive responses. Where we thought the TGF beta arm would contribute was really on prolonging durability that would lead to prolonged overall survival. We would see that also with the depth of response. I think we see that today when we compare to other EGFR inhibitors, whether it be cetuximab and those investigator-initiated studies that combine PD-1s with cetuximab, we see a much greater depth of response and a far improved durability. That is why, to your point, we are seeing that best-in-disease overall survival at 21 months, which in HPV negative head and neck, that was more than doubling what you would expect with nine months only on pembro.
I see. Okay. One area of debate regarding the TGF beta is that I think historically there has not been a lot of success. Then you guys showed that success at ASCO. Your hypothesis for the TGF beta is that it improves immune response through synergy with anti-PD-1s and reduces the immune suppression. I noticed that in the ASCO deck, you guys changed the language slightly and now say that it enables tumor penetration.
Penetrate.
Has your understanding of how TGF beta changed in terms of how it?
No, I mean, I think it's exactly consistent with how we always believed we would be able to remodel the tumor microenvironment. The data we shared at AACR really with spatial transcriptomics really focuses on this exact concept that we know in tumor types like head and neck, you have this very fibrotic immune-excluded tumor microenvironment. If you can combine a PD-1 with a TGF beta inhibitor, you can really go after the tumor microenvironment. To your earlier point, one of the key differences with our approach to TGF beta inhibitors is they have not typically worked well in the context of oncology. We always believed that, sorry, we always believed that the key difference was the fact that ours was tumor-targeted.
Using the EGFR arm to drive it directly to the tumor, which is why the phospho-SMAD2 data that we shared at AACR is the key to all of these pieces. In fact, in many ways, when we first began the company for Bicara, it was unfortunate that Merck KGaA and GSK had discontinued their PD-L1 /TGF beta. What we do know about that molecule is they were never able to definitively demonstrate tumor-target engagement by looking at phospho-SMAD2, which is why even in our dose escalation and the dose expansion you saw at AACR and ASCO, it was very important for us to demonstrate that at the doses that we see, we are seeing very high inhibition of phospho-SMAD2. That to us is really connecting the dots.
We're seeing this inherent tumor penetration and synergy with the PD-1 that allows for this one-two punch and the immune cells to infiltrate the tumor. That's leading to prolonged durability that's then leading to overall survival. The translational biomarker data we shared at AACR to me is really the focus on connecting the dots and following the science.
Yeah, I completely agree. I think it's encouraging not just to see, as you are seeing the TGF beta inhibition in actual paired tumor biopsies from patients treated, that T cell infiltration as well. For those specific case studies to be some of our deepest and best responders, we've seen a few complete responders as well as deep partial responders. It's really encouraging to us to kind of address this issue of tumor penetration that's challenged a lot of solid tumors. I think it's perhaps one of the reasons why you don't see great response rates to immunotherapy in head and neck the way you do in other cohorts and other solid tumor types.
I see. Okay. Based on ficera 's mechanism in terms of synergizing and enhancing pembro, and we know pembro works in both HPV negative and positive patients. Why does not ficera add that efficacy to the HPV positive patients beyond what pembro shows?
Yeah. I mean, I think what we will say is we do see high efficacy in the HPV positive population. In fact, we see two complete responses in our small N, which means that we are really driving deeper responses even in the HPV positive. Our focus on the HPV negative is really to increase and go after the patient population with the highest unmet medical need, but also importantly where we think we are going to see the highest difference in efficacy. I think we have seen this case in so many oncology trials. I always point people back to Merck versus BMS. By going after the CPS population and doing patient selection, Merck got every approval with pembrolizumab. While by going into an all-comer population, there was a big detriment and there were a lot of failed studies with NEVO.
For us, while we do see activity in HPV positive, which I think is consistent with the underlying mechanism, we know that HPV negative have higher levels of EGFR and higher levels of TGF beta. In many ways, we consider that our patient selection strategy to improve our probability of success for our pivotal study.
Yeah, I agree. The one thing I would add is I think we've had the benefit of seeing a wealth of literature and data around whether or not EGFR targeted therapies are effective in HPV positive disease and head and neck. We view these as two very distinct diseases with separate biology. I think no matter which study you look at, numerically the response rates are much lower for EGFR targeted therapies in HPV positive disease. Actually, when you look with monotherapy, there's almost no activity at all. I think for our purposes, we had the benefit not just of assessing some activity, as Claire mentioned in our own cohort, but just the totality of the data in head and neck really points to this refined patient selection approach in HPV negative only to really enhance that probability of success in a phase three trial.
Yeah. I think just echoing one of the key pieces that AJ brought up, it's not only that we just believe that these are two distinct diseases. You talk to any KOL, and they'll tell you that today it's well understood that patients respond differently, both on response rates, overall survival, and should be treated as two distinct diseases. It was really both following the data sets that came before us from panitumumab, Amgen's large studies, or other EGFR inhibitors as well as our own that really convinced us that the population that would most derive benefit was in the HPV negative population. By following the science and the field in thinking of these as two distinct diseases.
Got it. Okay. For head and neck, and also throw in CRC as well, do you believe TGF beta is a better target compared to, say, LGR5 when you add that on top of EGFR?
Yeah. I mean, I think there's a very clear biology of TGF beta in head and neck. It's well understood that these pathways are, especially in HPV negative disease, there is a very overexpressed TGF beta itself as well as TGF beta signatures. We see these more epithelial to mesenchymal or EMT phenotype signatures in patients who typically do not respond to immunotherapy. These are really the hardest patients. Again, I think where TGF beta plays a role in colorectal cancer is very much as a resistance mechanism to EGFR targeted therapies. It's known to be the number one driver of resistance to cetuximab and other EGFR targeted therapies. I think today, while I do believe LGR5 was identified in a screen in colorectal cells, it's not well understood what its role plays beyond stem cells.
While I do think TGF beta, while it hasn't showed strong efficacy because of this lack of tumor-targeted nature, is a very well-validated oncology target and applies to far more tumor types beyond just colorectal and head and neck. We know it plays a very important role in pancreatic tumors, in gastric tumors, and the like.
I see. Okay. Back to the ASCO data. Based on the data set that you guys generated, how do you see ficera compared to the SOCs and then among the other assets in development? Do you believe that it has to be best in class in order to win in head and neck?
Yeah. I mean, I think we have shown a best in disease profile in HPV negative when it comes to overall survival, which is the ultimate important endpoint for head and neck oncologists. Where I would say we always hypothesized we would be the most differentiated, and where we have demonstrated that today is really on a focus on durability. Durability, we are more than doubling the durability of EGFRs, and I would put both cetuximab and petosemtamab, which is in development, in that category. Ultimately, seeing these more deep, more durable responses, we think is really a best in disease profile. What we've heard from investigators is those deep responses fundamentally are leading to significant symptom relief, sometimes within a few days. I think what people forget is that these are patients who have a very hard time eating, have a very hard time speaking.
They are typically on significant pain medication. We are hearing in one case of a patient who deepened to a complete response, but within 24 hours went off pain medication and could start eating again. It is almost completely unheard of. We think being able to provide patients an option where you can combine ficera plus pembro and tell them you have a one in four chance at a complete response, that is end symptom relief, quality of life is a huge aspect. Not only do we think we will be best in disease in terms of overall survival, but in terms of symptom relief and providing patients a chance at complete tumor regression, I think is an important aspect of our profile.
Right. Another area that I was pleasantly surprised looking at the data was the patient population. I think in the ITT population, in the total population, that included HPV positive, I think the ECOG score was 45%. Here you show 37%, which means that these patients are more severe, so harder to treat patients, and you were able to show the type of efficacy you did for these patients. Does it mean that the HPV positive patient had an ECOG score of greater than almost like 60% to kind of balance that out too?
No. No, what I will say again is it's important to look at the totality of the baseline characteristics. ECOG status is one aspect, but we look at CPS score. We've seen a great balance. We see patients with a low percentage of our patients have distant meth only. In fact, the biggest predictor of response and overall survival to pembro, if you look at the KEYNOTE-0 48 data, is the presence of distant meth only. In fact, if I look at keynote, it's at the forest plots in terms of hazard ratio. It leads patients with only distant metastases have outsized overall survival compared to patients with local regional disease. In KEYNOTE-0 48, we saw about 30% of patients with distant meth only. In our study, we see about 23% of patients.
In LEAP-010, where we saw pembro control arm outperform KEYNOTE-048 to about an 18-month median overall survival, they had a 60% rate of distant meth only. That actually is, in fact, one of the biggest predictors rather than ECOG status for predictors to overall survival. The fact that we are actually worse than KEYNOTE-048 in terms of distant meth to us really speaks to location of tumor, some of the target lesions for bulky disease, the predictor of bulky disease, CPS status, and ECOG status. All of those together really speak to this was a very, in some ways, a completely unpreselected patient population with a high, typically a worse overall survival.
Yeah. I think maybe, Richard, I think you, Jane, and the team have done always a really good job thinking about the impact that baseline characteristics can have on not just responses, but also survival. I think, as Claire mentioned, there are number ones outside of ECOG, such as the CPS score, patients who have local regional disease versus distant mets. I think another one is also what the baseline tumor burdens look like in these patients. We had put out some baseline characteristics that showed that about half of our patients actually have bulky tumors in excess of 5 centimeters. I think outside of the baseline characteristics themselves is also being able to demonstrate the activity across those different patient subsets.
In our own data, I think we were really encouraged to see that across CPS status, across patients with bulky disease, we're seeing still really high response rates as well as really strong duration of response and OS across those CPS subtypes, for instance. I think that's been encouraging for us as we've evaluated the data across the different patient subsets.
I see. Got it. I am going to ask you a theoretical question. If you were to include pembro in the study, how do you think the pembro would have performed?
In this patient population, I think it's not necessarily just a theoretical question. If we look at KEYNOTE-048 and we look at real-world data sets that have very similar patient baseline characteristics, we see in particular in the HPV negative subset, a median overall survival of close to nine months. That to us really speaks to how we think about this patient population based off of, to AJ's point, not only CPS status and not only ECOG status, but some of the target lesions, as well as importantly distant meth versus local regional. In that totality of the data set, we do typically see very large data sets of 300 patients in the U.K. and I think 600 patients here in the U.S. from the Flatiron data set that really show consistency in these difficult-to-treat patients at about a median OS of nine months.
Got it. When you look at the totality of the data, how do you think this data set compared to, I know, the FACO study, the cetuximab plus pembro study? I think they show about 18 months OS in the overall and also in the HPV negative patients with 57% ORR. However, only 88% of those patients were in the first line. Sorry, in the first line. Then there were some patients in the second line setting. Again, very small numbers, probably not good to compare across trial, but just overall, your overall thoughts about how that compared to just cetuximab plus pembro.
Yeah. So we've had a chance to speak quite a bit to Dr. Sacco at UCSD on her data set as well as Dr. Christine Chung with her Nevo plus cetuximab data set. Again, to your point, these are both small data sets. In Dr. Chung's data set, she saw in the HPV negative a median overall survival of about 10.6 months. In the total population, she saw a 20-month median overall survival. That was, in that case, very driven by the HPV positive. The key piece that differentiates ficera plus pembro versus both of those data sets is really the durability and the depth. We continue to see more than a doubling of the depth of response compared to both of those data sets, and importantly, about a doubling of duration of response.
That is where we think we continue to feel confident that the durability really is consistent with our prolonged overall survival, which in many ways to us de-risk our ultimate endpoint of OS in a randomized study where you'll have the pembro control arm. The key differentiator is really durability.
Great. At ASCO, or right before ASCO, your competitor showed the one-year landmark OS rate of 79% for the ITT population compared to your 61% for the HPV negative patients. The market obviously reacted quite aggressively between the two. How do you compare the result for the one-year rate? Is that anything meaningful that could be predictive of the overall or the end outcome?
No. I think what we've seen today is the importance of mature data as a much more statistically stable endpoint, especially in these small data sets, single-arm data sets. These landmark numbers are typically point estimates, and you see significant confidence intervals. What we do know is there have been many examples where these short-term landmark analyses have not led to outsized median. We've seen that, unfortunately, with molecules like the CD47 for ALX, where they saw an 88% 12-month landmark OS that led to, again, an inferior OS compared to pembro control. What we would highlight is, while we were disappointed, of course, in the stock reaction, there are so many confounding variables to short-term landmark analyses. It can be very influenced by baseline characteristics, but also important subsequent therapies.
The feedback we've received from KOLs is really the importance of longer-term durability data and overall survival as predictive of what to expect in our pivotal studies. While we were disappointed, I do think people came to realize that our mature data set, the fact that we were seeing, while we did see patients, these are typically very sick patients in the HPV negative who do not survive past 6 or 12 months, and yet about half of those were surviving if they responded to ficera really speaks to the long-term durability. Again, I would point to all the prior studies of PD-1 combinations with lenvatinib, with IDO inhibitors, with TIGIT, with LAG-3, where we saw improved response rates and improved short-term landmark analyses, none of which translated to meaningful OS improvements.
Many of these KOLs have been burnt by those studies that it was important for, I think, ASCO to us really speaks to the positive reaction we heard from KOLs of just the strength of our data set really speaks to them wanting long-term durability data.
Now with that said, based on the phase I-B, based on only 28 patients, how reliable or how robust is that result? Because it is very impressive to show greater than 21 months. When I look at the OS curve between, I think, month 12 to 22, there are four, kind of depending on where the patient could land, and the OS rate could change quite a bit. Just thinking on the small number, I mean, how robust do you think is this data set when you try to translate to, say, phase II or III with over 600 patients?
Yeah. I mean, I think the importance is you have to look at that in the context of the totality of the data that we put out at ASCO. The fact that the durability is consistent, so that patients who are responding to ficerafusp are having these durable and deep responses, that is the ultimate predictor of overall survival. In fact, the fact that the majority of our patients today are still alive because of their time on ficera, not on the subsequent therapy. In fact, we have very few patients who've gone to subsequent therapy because they still are in response with ficera plus pembro. I do think all of the caveats taken around small ends for both ficera and pembro on single-arm studies.
What continues to build confidence for us is we will be sharing in the next 6-9 months 60 additional patients of HPV negative, recurrent, and metastatic that continue to speak to a consistency and larger end. I think showing 90 patients in the front line builds a very strong conviction that those data sets will replicate in a pivotal study. I think we know any phase I, phase II data sets will, of course, deteriorate in a pivotal study. That is why we also think it is important that we size our study accordingly. We have listened to the investigators around the proper sizing of the study. I think that is also an important risk between ourselves and our competitor where their study is quite small to really see that difference.
Okay. Let's spend the final couple more minutes on your phase II/III study. You guys are always thinking of the study. You guys have a couple interim analysis built into the study. I think the first one was a 60-patient study that you mentioned, 60 patients. The second interim is ORR for 415 patients. The final one is the primary analysis of the OS for 650 patients.
Maybe just to clarify, the first interim analysis for dose optimization does not use statistical analysis. The real statistically powered analysis is the interim analysis for accelerator approval.
I see. Okay. Maybe just how do you land on that design? What is that path for accelerator approval? What does that require? What does that look like?
Yeah. I think the first part is the dose selection. As we think towards an accelerated approval, once one of the non-optimal dose arm gets dropped, there's really three aspects to that. It's obviously the response rate endpoint that we want to show, but also really, I think a key piece just understanding, especially in head and neck prior trials where you've seen response rates that aren't necessarily translated to OS, for instance, Sleep 10. At that accelerated approval, being able to show durability data, six months' worth of follow-up on all patients, and also a descriptive analysis around the OS. I think those are key pieces of the accelerated approval, obviously with the confirmatory endpoint for the ultimate phase III being on an overall survival endpoint.
Maybe to one part of your question as well, this was designed as a seamless phase II/III so that the dose optimization portion or the phase II portion actually contributes to the interim analysis for the accelerated approval. We chose to do that in that way to be able to really accelerate those timelines and get to an approval earlier than had we done those as separate phase II and a phase III.
Okay. Got it. We saw a very different patient population between KEYNOTE-048 and LEAP-010.
Exactly.
That led to very different results from the.
Driven predominantly by that distant metastases.
Right. Now looking between these two studies, how are you thinking about your phase II/III in terms of the population that you want to recruit? Is there a specific type that you are, and how are you going to control that? Is it using caps? How do you try to achieve that patient population?
To your exact point and what I was alluding to earlier, even in the baseline characteristics we shared in our phase 1B, our population is actually quite similar to that of KEYNOTE-048. If anything, actually, the phase 1B population was sicker in terms of who would respond to pembro monotherapy given we were seeing less distant metastases. In LEAP-010, as I alluded to earlier, we saw about a 60% number of patients who had distant mets only. That was, again, given the toxicity profile of LEAP-010, they excluded patients typically with vascularized locoregional disease and predominantly enrolled patients with distant mets only. Given we do not have the same exclusion criteria as LEAP-010, we anticipate our data set or our patient baseline characteristics to look far more similar to those of the HPV negative KEYNOTE-048.
Fantastic. I actually have a lot more questions to do another fireside chat with you, but we'll save it for another time.
Perfect.
Claire, AJ, thank you so much for coming, and thank you so much for participating in your first Cowen-Sachs conference.
Thank you, Richard.
Thanks for having us.
Thank you.