Good morning, everyone. I'm Steve Brady. I'm the CEO of Tempest Therapeutics, and this morning I am joined by my colleague, Tom Dubensky, our President, Sam Whiting, our Chief Medical Officer, Anne Moon, who heads up Product and Project Development, and Nick Maestas, head up. Thanks, Nick, who is the Vice President of Finance and Strategy. We are really fortunate this morning to be joined by a group of distinguished thought leaders. They're the real reason you're here, not to listen to any of the Tempest folks. First, Mark Yarchoan from Johns Hopkins is going to be speaking on Tuesday about our lead clinical program, TPST-1120, PPAR α antagonist. Just note there, this is our first clinical data that Tempest has presented, and then he's gonna hand the phone off to Susanna Ulahannan from the University of Oklahoma. She's gonna be speaking about our second clinical program, TPST-1495.
Then finally, Jason Luke from University of Pittsburgh School of Medicine is going to be talking about our TREX-1 program and innate immunity. Then originally, Toni Choueiri was going to be joining us for the Q&A session at the end. For those of you who saw Toni yesterday at ASCO, he lost his voice in the middle of a panel, so he is unable to join us this morning. Well, we didn't talk. We communicated via text this morning 'cause he doesn't have a voice. For those of you who wanna speak with him after the event, he's absolutely willing to discuss the programs.
As you might imagine, he does have pretty strong opinions about what Tempest is developing. Just give him a couple of days to get his voice back. That was the only request before you give him a call. We are gonna have, as I said, a Q&A session at the end, so if we could please hold questions until then. We do have folks online as well using a chat function to submit questions, so it'd be great if we could keep each of our speakers on schedule. Tempest is a public company, so please refer to our filings in the event that we make any forward-looking statements today. Again, by way of brief introduction because you're here to hear our speakers. Tempest is a clinical-stage oncology company.
Currently, we're focused in all small molecule therapeutic development. That is not a bias. It's just the right way to hit each one of the targets that we are hitting. Each one of these programs, and I would say we're fortunate to have such a robust pipeline, are not bound by a single hypothesis or pathway or discovery perspective. Each one of these stand on their own two feet. The one thing linking them all together is that they were selected because we think that they are the right way to hit areas of unmet medical need, and they all have the potential to hit a broad patient population. The main topic for this ASCO is TPST-1120. It is a PPAR α antagonist. I stress the antagonist 'cause we're all familiar with this pathway.
To our knowledge, this is the only PPAR α antagonist in clinical development. We think that's relevant from both a competitive perspective, but also it's one of the things that factored into our collaboration with Roche, and we're gonna talk about a pretty, you know, an exciting study that we have ongoing with them. There's a strong mechanistic rationale to put this molecule together with atezolizumab and bevacizumab. Secondly, this is the first molecule, as I mentioned, that for which we are doing a clinical presentation. We're excited about that, our first program is producing clinical benefit for patients, and Dr. Yarchoan's gonna talk about that today. Our second program in the clinic is behind TPST-1120. It started actually in the middle of the pandemic, TPST-1495. It's in the prostaglandin pathway, an arising pathway of interest.
As I'm sure you know, there have been data at SITC. There were data at AACR, implicating signaling in this pathway, multiple potential mechanistic points where you can intervene. Very excited about this program. Like TPST-1120, to our knowledge, we're the only dual EP2 / EP4 antagonist in the clinic. There are multiple EP4-only antagonists. For our view, that's necessary but insufficient. Dr. Ulahannan is gonna talk about this program, which is ongoing right now in monotherapy and combination studies, across the U.S. Finally, our third publicly disclosed program targets TREX-1 in the STING pathway. To give Tom the credit that's due .
Dr. Dubensky over here was the first person to lead a program that resulted in a STING agonist being put into clinical study, and we think that the TREX-1 program is really the evolution of what we've learned in the past. This program has fallen out of favor somewhat with investors given some of the clinical data, but STING remains of significant interest in clinical research, and especially on the corporate side, there are multiple programs. We just put out some of our first preclinical data at AACR, and there were other posters there as well. Then finally, we have an undisclosed cancer target we believe completely novel. We licensed in IP from University of California, Berkeley, and we announced that in the fall.
Well, this is moving forward as well, and we'll probably announce the target at an upcoming R&D meeting, down the road. We really wanna make progress on that obviously before unveiling it to the rest of the world. With that, I'll turn it over to Dr. Yarchoan. Thank you.
Morning. So I'm Mark Yarchoan. I'm an Associate Professor of Medical Oncology at Hopkins, and today I'll be presenting for the first time ever the phase I trial of TPST-1120 on behalf of all the co-investigators here. This will be an oral abstract here at ASCO on Tuesday for those of you sticking around. Free fatty acid oxidation is a key metabolic adaptation that occurs in many tumors. It induces tumor cell proliferation, induces angiogenesis, and then critically induces immunosuppression in the tumor immune microenvironment. PPAR α is a gene that many people may be aware of. It's a master regulator of fatty acid oxidation.
Inhibiting PPARα is a novel strategy to reduce fatty acid oxidation, inhibit tumor growth, and critically to enhance antitumor immunity by relieving immunosuppression in the tumor microenvironment. TPST-1120, it's a small molecule oral inhibitor of PPARα that has greater than 250-fold selectivity for PPARα over other PPAR isoforms. This is some preclinical data just showing the dual mechanism of PPARα antagonism. On the left, you can see that both ablation of PPARα in tumor and PPARα ablation in the host mouse, both of them independently reduce tumor growth. When you combine PPARα inhibition in tumor and in the host, the tumor volume is reduced nearly 100%.
On the right, this is further validation that PPAR α is really an immunomodulatory target. If you take the bone marrow from a PPAR α knockout mouse and actually transplant that bone marrow into a tumor-bearing wild-type mouse, you can carry over that or recapitulate that tumor control just through bone marrow transplantation, which again, I think is strong evidence that PPAR α signaling in immune cells is a critical immune target. This is data from TPST-1120 that shows that it has both direct antitumor effects through direct tumor killing as well as immunomodulatory effects. On the left, you can see these are three different HCC cell lines exposed to TPST-1120, and then there's a direct tumor killing effect that's dose dependent.
Below that, this is an RCC model, [xenograft] model again showing very nice tumor control through direct tumor killing of TPST-1120. I think importantly, this is also an immune target. On the right, these are TIL isolated from MC38 tumors, and what you can see is that consistent with the mechanism of action of PPAR α antagonism, there's a markedly reduced lipid uptake from dendritic cells, which is a tolerogenic dendritic cell phenotype. On the right, T cells from these tumors have markedly reduced evidence of exhaustion, so lower PD-1, LAG-3, and TIM-3 expression. This is both a direct tumor killing effect and an immunomodulatory target.
Given that TPST-1120 appeared to be immunomodulatory, there was a hypothesis that this agent might combine well with anti-PD-1 therapy, even synergize with anti-PD-1 therapy. Sure enough, shown here are the growth curves for a combination of TPST-1120 and anti-PD-1 therapy. In green, you can see that there's much better tumor control than monotherapy or vehicle, suggestive that these agents combine very well. Actually, in this experiment, a few of the mice actually cleared their tumors. These mice appeared to develop T cell memory against tumor because they resisted tumor rechallenge. Actually, transfer of splenocytes from these mice into other mice continued to control MC38 tumors. Again, very strong evidence that this is a immune-mediated phenomenon that this induced excellent T cell memory.
Now the data everyone's been waiting for. This is the TPST-1120 phase I study design for TPST-1120, both in monotherapy and in combination with nivolumab. The primary endpoints were safety, PK/PD, and then a preliminary assessment of efficacy. You can see that all tumor types were eligible to participate in the monotherapy portion. In the combination portion, eligibility was restricted to three tumor types, hepatocellular carcinoma, renal cell carcinoma, and cholangiocarcinoma. These three tumor types were selected based on data from the TCGA showing that these three tumor types likely had an enrichment of fatty acid oxidation versus other tumor types. The other thing I just wanna call your attention to is that in both parts prior treatment with anti-PD-1 therapy was allowed.
These are not patients who are naive to anti-PD-1 therapy, okay? This used a standard 3 + 3 design. In monotherapy, there were five dose levels ranging from 100 mg twice a day up to 600 mg twice a day. In the combo portion, there were four dose levels from 200 mg twice a day to 600 mg twice a day. DLT period was 21 days for monotherapy, 28 days for combination, and the PD-1 agent in this case was nivolumab at the standard 480 mg every four weeks. These are the baseline patient characteristics. I think the things I would call your attention to is in the monotherapy group, enrollment was really dominated by late-line pancreatic cancer, okay? That was the dominant tumor type. Followed by cholangiocarcinoma.
It's hard to imagine a more challenging group of tumors to enroll in a phase I trial. In the combination arm, again, it was restricted to the three tumor types I've mentioned previously. I think the thing I would call your attention to is that these were heavily pretreated patients, a median of three prior lines of systemic therapy. The majority of patients in the combination arm had prior exposure to anti-PD-1 or anti-PD-L1 therapy, including all the patients with HCC and all the patients with RCC. Not too much to say about the PK here. I think what's critical is that there's a linear dose exposure relationship shown on the left for both monotherapy and combination.
On the right, you can see that the steady-state profile showed that all the doses studied in combination were above the IC50 adjusted for protein binding, with the exception of the lowest 200 mg twice-a-day dose. You know, I put a couple of patients on this therapy in my clinic and my experience with it reflects what's in the charts here, which is that this is a very well-tolerated agent overall. Patients had some low-grade fatigue, diarrhea, nausea, abdominal pain. Again, these were low-grade grade 1, 2 events. In the monotherapy arm, there was a single grade 3 event, which was hypertension. In the combination arm, there was a single patient with arthralgia, a single patient with LFT abnormalities, and one patient with muscle spasms.
Again, grade 3 toxicities. There were no grade 4 and no grade 5 toxicities in either arm in the study. There were no DLTs during the dose escalation portion, and the recommended phase II dose for both arms was 600 mg twice a day, the highest dose studied. This is the monotherapy data. On the left is a swimmers plot for all patients, and then on the right is a waterfall plot for the 19 patients who were evaluable for response. The disease control rate in the monotherapy part was 53%, so over half the patients with stable disease as a best response. You can see there were four patients who had some reduction in tumor volume not meeting criteria for a partial response.
Again, I think the thing that I would call your attention to is that the majority of patients who enrolled in the monotherapy portion were late line, third line, beyond pancreatic cancer or cholangiocarcinoma, where, you know, candidly, as someone who treats these patients, stable disease is a big win here. You can see on the left, a number of patients really had meaningful, prolonged, stable disease with this agent, clearly showing that this agent has activity as monotherapy. Shown here are two patients, both with IDH1 mutated cholangiocarcinoma. That's a mutation found in about 25% of intrahepatic cholangiocarcinoma. You can see that both patients had stable disease as a best response. One patient, TPST-1104, was able to stay on drug for approximately 10 months with some reduction in tumor volume.
You know, the reason I call this out is that this is a subgroup of patients where we have pretty good data from recent modern clinical trials about how they do. As many of you know, ivosidenib is approved for IDH1 mutated cholangiocarcinoma. Ivosidenib was approved on the basis of a 1.6-month prolongation in PFS, and a response rate in the single digits. Again, small numbers of patients, but just comparing what we're seeing here as monotherapy versus other approved agents, in a similar sort of group of patients, I think that it's clear that this agent seems to have activity that looks quite favorable. The other reason that this is called out is that IDH1 mutations, interestingly, at least pre-clinically, may enrich for fatty acid oxidation.
Actually there are some pre-clinical experiments that are ongoing with our group and others trying to understand if this could be a biomarker to select for TPST-1120 sensitivity. Then I think this is the data that everyone's been waiting for the most, and this is the combination therapy data for TPST-1120 plus nivolumab. Now as a reminder, these are patients who had prior exposure to anti-PD-1 or anti-PD-L1 for the most part. There were three PRs on this study for an overall response rate of 20%. All the responses occurred in patients who were at the higher dose levels. Actually if you look at patients with higher dose exposure to TPST-1120, the response rate was 30%.
Among the RCC cohort, the response rate was actually two out of four, so 50%. Again, critically, both of the RCC patients who responded had prior exposure to anti-PD-1 therapy and had progressed on anti-PD-1 therapy. I will move on here. I think pictures speak a lot more than words. Shown here at the top is one of the responders. This was a patient with metastatic RCC who received ipilimumab plus nivolumab in the first line setting, had some stable disease and then eventual progression. Then went on a TKI cabozantinib, then went on everolimus in third line and then enrolled on our study in fourth line. Again, re-challenging with nivolumab, something that they had already progressed on.
You can see that the patient had a really quite remarkable response. There's a reduction in innumerable lung lesions that are shown here for an overall response assessment of partial response. This response is still ongoing at the latest disease assessment time point, which is about a year out from when the patient started. I'll just note this patient blew through, again, three lines of prior therapy, none of them with any sort of prolonged stable disease control. This is much more durable and fourth line than anything this patient had before. Subject 2208 is another renal cell patient who responded. This patient was treated at our institution, and I happen to have met this patient. This patient received a PD-1 inhibitor, pembrolizumab, plus axitinib in first line.
Had progression, went on a TKI in second line, cabozantinib, with eventual progression, and then enrolled on this study, and had a partial response. The patient actually came off of therapy because the patient had a pericardial effusion at baseline. A short time onto therapy, reevaluation of the tamponade showed that, probably the patient should have a pericardial window, even though there wasn't true progression there, just 'cause there was some concern. Unfortunately, it was a very complicated hospitalization. The patient had to come off therapy, but clearly this patient had a nice response to therapy, after multiple lines of prior therapy.
I think the overall thing that I would call out here is that these are two patients who had prior exposure to anti-PD-1 therapy, progressed on anti-PD-1 therapy and then were rechallenged with PD-1 therapy in combination with TPST-1120. Both patients responded, and for me, this provides strong evidence that TPST-1120 overcomes primary or secondary resistance to anti-PD-1 therapy. This is another patient who responded here. This is a patient with metastatic extracranial cholangiocarcinoma, who had multiple lines of systemic therapy prior to enrolling on study. This patient had a partial response. You can see the scan showing an interval reduction in a liver lesion. Critically, this patient's tumor was PD-L1 negative, was mismatch repair proficient, and had a tumor mutation burden of less than 10 mutations per megabase.
For those of you who don't treat cholangiocarcinoma, this is an extremely immune-resistant type of cancer. The response rate to PD-1 as monotherapy is in the single digits, especially for patients with low TMBs and who are mismatch repair proficient. I think this is strong evidence that TPST-1120 can reprogram the tumor immune microenvironment. As someone who treats hepatobiliary tumors, I'm really excited that clinical development of TPST-1120 continues. This agent has been incorporated into Roche's Morpheus phase II study, which is investigating bev-atezo in combination with other agents, in a randomized phase II study design. Bev-atezo is currently the preferred first-line standard of care for HCC.
It's really exciting for me that TPST-1120 is now being studied in the first line with a contemporary, modern, preferred first-line regimen in a major tumor type. This should be a real readout about what TPST does in an important tumor type. Up to 60 patients will be enrolled in the combination arm of this study. The primary endpoint of the study is response rate. In conclusion, I'm very excited about this compound. TPST-1120 is a first-in-class oral inhibitor of PPAR α . In this study, TPST-1120 demonstrated safety and excellent tolerability. For me, this study showed that TPST-1120 has some activity as monotherapy and likely more activity in combination with anti-PD-1 therapy.
The responses observed in this study in patients with prior progression on anti-PD-1 therapy, for me, provide strong evidence that TPST-1120 works as hypothesized from the preclinical models, and that it can reprogram the tumor immune microenvironment and overcome resistance to anti-PD-1 therapy. I think critically, development of this agent continues with a planned phase II trial in combination with atezolizumab and bevacizumab and HCC, and hopefully potentially in other tumor types as well. Those discussions, I believe, are ongoing, and I really look forward to seeing the future development plans of this agent. I will be happy to take questions at the end of the session, but I'm going to turn it back over. Oh, perfect. Thank you very much.
Hi, I'm Susanna Ulahannan, and I'm a Medical Phase I Trialist.
I'm also a GI Oncologist at the University of Oklahoma. I'm gonna talk to you about TPST-1495, which is a dual, first-in-class, dual EP2, EP4 prostaglandin E2 receptor antagonist.
We know for many years we've been looking at the prostaglandin E2 pathway, and we know that it is a very important pathway. You know, the question is really how do we target this pathway in the best way, just looking at toxicity and also looking at efficacy. The prostaglandin E2 has both promoting the tumor, and it also suppresses the activity through its four receptors, EP1, EP2, EP3, and EP4. The EP1 and EP3 are tumor-suppressing and immune-activating. On the contrary, the EP2 and EP4 are tumor-promoting and immune-suppressive. When we look at NSAIDs, they block upstream. They block all four of these receptors. They also block other bioactive lipids, which causes mainly cardiotoxicity, which has been limiting in COX-2 inhibition.
Despite all this research, we still don't know how to optimally target the PGE2 signaling pathway. The TPST-1495 is a first-in-class, highly specific antagonist that inhibits only the promoting EP2 and EP4 receptors. It's oral therapy, and it targets both the tumor cells and the immune-suppressive cells. Next slide. This is a TCGA chart just looking at the expression of EP2 and EP4 in multiple different tumor types. We see here that both of these receptors are highly expressed in multiple tumors. We see here, for example, pancreatic cancer and colorectal cancer both have very high expression of both these receptors, and these are cancers with very high unmet need.
This kind of supports the rationale that both these receptors needs to be blocked. Consistent with this that we talked about before is that TPST-1495, it directly inhibits the tumor growth without being dependent on the immune function. It is shown here in an animal model where there was a patient-derived CRC adenocarcinoma that was orthotopically introduced into the colon of an immune deficient NCG mouse model. These mice were treated with TPST-1495 or vehicle. You see here in the next panel here that after nine weeks, when we looked at the tumor burden, that mice that were treated with TPST-1495 had a significant tumor reduction both in lung and liver metastasis.
In addition to directly stimulating tumor growth, prostaglandin E2 suppresses immune function through the EP2 and EP4 receptors. Here it is illustrated how EP2/ EP4 increases the immune-suppressive cells. The MDSCs, the M2 macrophages, and T regulatory cells. It also decreases the dendritic cells, both its maturation and differentiation. It also decreases NK cells and cytotoxic T cells. Here, on the top panel, you can see how dual EP2/ EP4 inhibition with TPST1495 reverses the immune-suppressive activity of PGE2. In the top panel, there is, it's an in vitro experiment looking at human monocytes inhibiting EP2 and EP4 with the TPST1495. That's the blue line. Then there is inhibition with either EP2 or EP4 at a single blocking agent in the black line.
You can see that there's a significant reduction in M2 macrophages and an increase in dendritic cell differentiation and maturation, and it's significantly better with the TPST-1495 with the dual block compared to the single-agent blocking. On the bottom panel, it uses an immunocompetent mouse model, the CT26 tumor model, and what it shows is that TPST-1495 has single-agent antitumor activity that leads to this CD8 tumor influx. You see here the CD8 influx of T cells compared to the control. There's a strong mechanistic rationale for combining the TPST-1495 with an anti-PD-1. Everyone wants to combine with a PD-1 inhibitor, but it actually makes sense in this case. The association we see here that the PGE2 is a potent suppressor of immune function in the tumor microenvironment.
It's also associated with both primary and secondary resistance to immune checkpoint inhibition. With the TPST-1495, it blocks the immune suppressive signaling of the PGE2, and it stimulates the anticancer immune function. There's been some recent data in the Topalian Lab at Hopkins, where they have shown that the COX-2 and PD-L1 are induced by non-redundant signals and represent independent and potentially complementary therapeutic targets. The experiments that you see here on the left is, and again, a immunocompetent tumor mouse model, the CT26, where you see the combination of TPST-1495 with a PD-1 checkpoint inhibitor compared to either of them single agent, that there is a synergistic effect with the combination.
Here is a mouse model, the APC-min mouse model, and this mouse model stimulates spontaneous human colorectal cancer through heterozygous mutation of the APC tumor suppressor gene. We know in colorectal cancer that the APC mutation is one of the most frequent mutations. In this mouse model, you see here how they have blocked. The black one is the vehicle, and then we have the TPST-1495 in the blue, and then single receptor blocking either EP4, EP2, or with celecoxib. There is a significant decrease with TPST-1495 compared to just blocking EP2 or EP4. Here is looking at the survival in these mice, and we see again that the TPST-1495 has the best results. Interestingly, this red line is combining a EP2 receptor with the EP4 receptor.
We see that these mice did much better than blocking either of those two independently. There's a growing evidence of positive feedback loops between oncogenic signaling transduction pathways and the PGE2 expression. The PIK3CA mutation is something that we have looked at for a long time both in CRC and squamous cell head and neck. We know that these patients who have a PIK3CA mutation seems to benefit more from aspirin and NSAIDs and NSAID therapy in preventing cancer. We see here is that that might be a biomarker to predict which patients will respond to TPST-1495. Here is the phase I clinical trial. It's a first-in-human. It's TPST-1495-001.
This trial is currently ongoing, and we are looking at different dose schedules and also looking at the optimal dose. In the monotherapy, we have multiple dose levels. We initially looked at BID dosing and then looked at QD dosing, looked at intermittent dosing, and so currently ongoing. When we reach that recommended phase II dose, the plan is to go to dose expansion cohorts in endometrial, squamous cell, head and neck, and also this PIK3CA basket, which might be a biomarker for efficacy. There's also the combination cohort here with pembrolizumab that's ongoing in multiple tumor types. Again, the plan is once we reach a recommended phase II dose, is to go to MSS colorectal cancer, endometrial, squamous cell, head and neck, and again, a PIK3CA basket.
In summary, the prostaglandin E2 stimulates tumor cell growth and suppresses anticancer immunity through the EP2 and EP4 receptors. TPST-1495 is a first-in-class, potent, and selective dual antagonist of EP2 and EP4 that does not inhibit the immune-stimulating EP1 and EP3 receptors. TPST-1495 has immune independent and immune-dependent antitumor activity in preclinical models and overcomes PGE2-mediated immune suppression more actively than with the single antagonist of either EP2 or EP4, or for that matter, COX-2 inhibitors. The enrollment is ongoing in the first-in-human phase I trial to look at the optimal dose and schedule. The safety profile so far currently is looking good. I've treated quite a bit of patients on this trial. We're looking at finding the dose and then going to expansion.
The potential cohorts that we have looked at in the expansion makes sense from a biomarker standpoint and also looking at the literature and the biology of the PGE2 pathway.
All right. Good morning. Thanks everybody for coming out this morning. I'm Jason Luke from UPMC Hillman Cancer Center, also at the University of Pittsburgh. I'm really excited to give a presentation today. I really, personally, I've had a long-standing collaboration with Tom and now Steve. I really think it's one of the best biotech companies that I work with, and people in the room know that I work with a lot of them. As witnessed on this slide, which I show actually, only for the purposes of emphasizing that, you know, the field goes up and down, but the fundamental science that underpins what we're trying to accomplish doesn't, right? The trends go back and forth, and I don't want people to really lose sight of the fact.
What I highlighted here, you know, obviously I have a lot of interactions, but all the companies in bold are companies that are still pursuing STING agonism as a high priority therapeutic intervention. I think, again, a lot of interest goes up and down in the field, but it's. I think it's really important to emphasize that there's fundamental biology underpinning that pathway, really makes it a high priority, and it's not at all the case that people have given up on trying to target this pathway. In the context here, I'm actually gonna talk about TREX-1, but I wanna also differentiate the idea that TREX-1 is not just a STING pathway therapeutic agent. It actually has other potential therapeutic aspects to it, and that's gonna be very important as we kinda get along.
I'm gonna give a little bit of background that I see actually a lot of friends from the investment community who have probably seen these slides many times, but for people who have not and who are on the web, I'm gonna actually go over it a little bit just to make sure to level set. I'm gonna remember to say next slide. I'm currently on slide four, which is T cell inflamed gene expression tumor mutational burden. Just for everybody who's, you know, maybe not, you know, thinking about immuno-oncology all the time, you know, where is the field now? Well, it's really kind of not as far as we had wished, but it is somewhat far in that we do have a pretty good handle now which patients are most likely to benefit from anti-PD-1 as a monotherapy.
On the left-hand side, you can see a schematic around sort of simplistically what is going on. The immune cells come in contact with the tumor, PD-L1/PD-1 interaction, interrupt that get response. From many datasets, we now know that the presence of an interferon response, and on the right-hand side, you can see this as measured by a T cell inflamed gene expression score, which is sort of you can surrogate for PD-L1 expression or high levels of tumor mutational burden, both strongly correlate with outcomes to anti-PD-1, such that those patients that have high levels of both are those that are most likely to respond. That's been shown by many groups now.
I think it's underappreciated how important that is, because even at a minimum, that would be a good way to stratify patients, and we actually don't use that all the time. Now, that being said, those are clinical observations that can inform who gets treated, but we'd actually use that to dive into thinking about the biology that underpins that then to think about therapeutic targets and further drug development strategies. Now on the next slide, immunobiology of the T cell inflamed and non-inflamed tumor microenvironment. What I show is trying to break this down a little bit and turn this from an observation in clinic into thinking about how do we use this then to move forward. This is my favorite slide that I've ever seen in my entire life. I left the Harvard Medical School because of this.
Because simplistically speaking, this actually explains what we're trying to accomplish in immuno-oncology, right? On the left-hand side, we see a tumor that's highly infiltrated by CD8 T cells. Because the CD8 T cells come in, they elaborate interferon gamma and other cytokines and chemokines, which drive the accumulation of all the things we see in hot tumors. As an extreme opposite, we see non-T cell inflamed tumors then in the middle there, and those are tumors that lack a spontaneous T cell infiltrate. Because there are no T cells that go in the tumor, none of the rest of this stuff happens. This is something that's really obvious, and yet almost 15 years into immuno-oncology, people still don't understand this, and I don't know why.
You have to get T cells into the tumor first in order to have all the things happen downstream. It's not a one-off thing, right? The T cells come in and all these things come downstream. There's been a lot of work done. I already showed you data from gene expression profiling, but this was data from my mentor, Tom Gajewski's work, even dating back into the, you know, 2,000s, looking at chemokine gradients in order to predict which patients had a highly T cell inflamed tumor microenvironment. The reason I show you this was even back in the day when we were using gene arrays. You can see that you could predict out which patients were most likely to have an immune infiltrate.
Observing that, then you could dive into the rest of the gene expression data and look for other associated pathways that could potentially predict response. As I mentioned now, it's been shown with anti-PD-1. On the bottom right is data for pembrolizumab. Patients who have a high interferon signature are gonna do much better in terms of overall survival and progression-free survival compared to those that don't. On to the next slide. We use that observation. Again, now this is old data now, but everybody knows this, but it's still important to profile the TCGA, which at the time was a novel thing to do when thinking about gene expression profiling for immune genes, to show that there was this spectrum of hot and cold tumors across all the different tumor types.
I emphasize it here because it really shows the opportunity space that remains in immuno-oncology. On the top, you see a heat map of all tumors from TCGA, but on the bottom you see all the different tumor types split out individually. The red dots are the hot tumors, the blue dots are the cold tumors. I think just to emphasize that where we see the most activity for anti-PD-1 is in the red dots, but only like maybe half of the red dots and none of the blue dots. There's a lot of patients and a lot of tumor types that are still out there that could benefit from anti-PD-1. You know, you might say, well, but if they're blue dots, then PD-1 is not going to work.
Well, that might be true, but the question then is the immunobiology possibly there that we could still actually get at those patients? We would argue the answer is yes. In that same paper from back in 2016, we looked at the presence of antigen in cold tumors. On the top, you see the non-synonymous mutational burden between T cell signature high and low tumors. Sorry, I went on a slide, people on the line. We even went a step further and synthesized candidate HLA-A201 binding peptides and looked at the presence in hot and cold tumors and showed that there was no difference. In other words, the number of antigens is. The antigens are there, but for some reason, the immune system has not mounted a response effectively.
Rather, when we looked at who is gonna what actually associates with the presence of an immune response is actually the presence of antigen presentation machinery, type one interferon, dendritic cells, and that's indicated by the expression of BATF3 and IRF8, which are genes that are associated with dendritic cell function. Moving on then to the next slide, this sort of sets up our paradigm then. For tumors that lack a spontaneous immune response, what can we do to get it started? There are a lot of different approaches we can think about here.
In this cartoon that we published some years ago, you see a tumor on the upper left where you can do things, and so activate the STING pathway or generate DNA damage, which will allow the immune system to observe the cancer and then drive type one interferon priming of immune cells in the lymph node and the trafficking of T cells back to the tumor to inflame tumors. How would we do that? Well, the canonical pathway that's been shown in the context of human tumors with T cell responses is the STING pathway. There are many pattern recognition pathways that are associated with immune responses to bacteria and viruses. In human beings, the STING pathway is strongly associated with upregulation of chemokine gradients that attract T cells in patients. Moving on, what about STING?
STING is a little complicated. This is actually a slide that I ripped actually from Tom, 'cause Tom is like one of the leading experts in this field from a few years ago when he gave a presentation at SITC. I won't go over all that except to say that preclinically and translationally, STING checks every box that you want in drug development. There's a human phenotype. The mouse models look unbelievable, right? You get this abscopal-like effect in many different mouse models, right? I'm not gonna go over the data because I think everybody probably knows this already, except to say it's complicated because in clinic, we haven't quite seen the efficacy that we want.
There's another side to the story, which is that a number of people have published on the idea that overactivation of the STING pathway actually can drive cancer in and of itself. How do we thread this needle? How do we think about this, that STING is associated with an effective immune response, yet too much STING causes cancer? I think we don't fully understand that yet, but I think it, again, raises sort of this high level complexity. Given that, in a previous life, Tom was still quite intelligent but running other ideas in another company, we led the first in human development of the first thing, I guess, to go into human beings, which was ADU-S100 or MIW815, and I've moved the slides now. I'm now on the one with the waterfall plot.
What I emphasize here is these are all the patients in the phase I trial, and as has already been presented, quite heterogeneous group of advanced solid tumors, many different tumor types, all refractory. What you can see, though, is that when you look at the sum of all target lesions in all these patients, that's the blue line. You know, the majority of them get worse when you aggregate. When you look actually at the injected lesion where we put the STING agonist, almost all of them actually shrink. As a sort of a case report, this was one of my patients who had a very large neck lesion, a head and neck cancer. You can see we injected this tumor, and over the time, it involuted.
This patient who had progressed on anti-PD-1 as their immediate prior therapy is on this therapy about 9 months. Once they progress, we put them back on PD-1, and they started re-responding again. It's a case report, again, emphasizing that we can do this. There are some patients out there where this works, and that was associated with translational findings that you would expect, like the infiltration of CD8 T cells as well as other interferon-associated effects, genes expression, et cetera. One more case, moving on to the clinical case vignette. This was a patient that we treated on the sister study to the phase I. This was the combo with anti-PD-1.
This was a lady with a large melanoma that was refractory on her right hand that we were injecting and then treated concomitantly with anti-PD-1, and you can see this over time. Eventually, it was completely gone and biopsy proven to be completely gone, and we took this woman off, and to the best of my knowledge, she's never had a recurrence. One final going on, STING agonism can drive systemic responses. You know, the issue here, what we've seen is that when we inject, it works where you inject, but what about everywhere else? Well, this is data from the H3 [Bioscientist] E7766 STING agonist that we presented at SITC last year.
Patient of mine who had esophageal cancer with a very large neck mass, where we injected the lymph node up underneath the ear, and we did see treatment effect in the neck. That was associated with the elaboration of all the cytokines, et cetera, that you would expect. A big induction of interferon alpha, but the concomitant downstream activation of interferon gamma, IP-10, and then induction in the peripheral blood of all the different immunomodulatory genes that you would expect being associated with STING agonism. The reason I raised this case was this patient actually had a systemic response to the injection of just the one lesion in the neck.
You can see that in the red arrows now on the CT scans slide in the deck, where you can see the patient's disease in the nodal basins as well as the pulmonary metastases actually shrank. This patient was on the treatment for about eight months, and unfortunately, after progression, he passed away, but this was really quite emphasizing the idea that in some patients this can work even as a systemic therapeutic agent, but really, you know, suggesting we gotta get this pathway engaged more broadly in cancer. Moving on to STING agonism development. There's a lot of development in this space, and this is from a review that we put out last year looking at the various different molecules. Again, just to highlight that people have not given up on STING agonism by any means.
Essentially, every version of STING agonism you can think of between intratumoral, IV, subcu, antibody drug conjugates, anything you can think of is still getting developed in this space. But you know, that it remains kind of a tough nut to crack, and we're not sure how to do it. There's a fair amount of interest now in thinking about regulatory pathways that control STING that you might be able to drug systemically as opposed to having to do intratumoral injection. Two of them that are a high profile I'll point here. One of them is ENPP1, which is a phosphodiesterase that regulates the degradation of cGAMP. A second one, coming back to the point here, is the exonuclease TREX-1 that breaks down double-stranded DNA in the nucleus.
Why are we really interested in that? Well, as I just mentioned, TREX-1 is chewing up DNA, which can release cGAS and that lead to downstream activation of this pathway. It's been quite potent. We don't have to go into all the literature. There's a lot of literature showing that this will activate this pathway. Playing off of what I just said about the potency of the STING pathway, this is a potential systemic way to drug the STING pathway to sort of bring down what you might call the activation potential around the pathway to make it more likely to trigger in the context of other interventions. In addition, however, I mentioned the beginning, TREX-1 does more than just regulate the STING pathway.
TREX-1 is also involved in the tumor cell viability in the context of homologous recombination, where TREX-1 has a role in allowing that process to go forward, and therefore drugging this pathway actually has the potential to induce synthetic lethality in the same way that we think about PARPs and other DNA damage response approaches. TREX-1 kind of sits at a really interesting intersection between immunity and DNA repair. I'll just highlight one other thing, which is in the context of STING agonists, we've done a lot of work looking at the intersection of irradiation and anti-PD-1, because preclinically, there's a gigantic literature that suggests that radiation and immunotherapy should marry well together.
This is another complicated story, but I'll just talk about it very briefly, which is we pursued a study thinking that we don't really believe there's a lot of abscopal stuff going on, but rather than, and I guess the oligometastatic data from earlier at ASCO actually probably proves that point. Really what we were interested is irradiate as many tumors as possible actually with the idea that anti-PD-1 might actually enhance the irradiation. Where you irradiate, you're actually gonna allow the immune response to be bigger if you give anti-PD-1. In this clinical trial, we made the observation of, I think, actually for the first time, which is surprising as a medical oncologist in radiation oncology, that, kind of obvious, but in the patients we irradiated, if the irradiated lesion shrank, that was associated with the overall survival of the patients.
That seems like kind of obvious, but that's we never really thought about it that way before. You can see the difference here. Responders had a median. This is advanced solid tumors with six lines of prior therapy, a very progressed bunch. You see they live much longer relative to patients with a mixed response and patients with non-responder. When we went in the study, we actually did pre and on treatment biopsies after the irradiation. Again, apparently that had not been done in the field before, but we observed kind of probably what you'd expect, but really mechanistically pulls this thing together, which is that in the patients who benefited, what we saw was the induction of type one interferon and dendritic cell pathways, so like, as you would expect, driving de novo immune responses.
You can see the canonical pathways and upstream regulators, which include antigen presentation and dendritic cell maturation, and then upstream regulators around interferon gamma. The irradiation would inflame the tumor microenvironment that can be then enhanced by anti-PD-1. TREX-1 is really well-positioned then as a therapeutic target that should be a very high priority to the field. Tempest has done quite a bit of work over the last couple of years to develop their TREX-1 program. You can see here data that they have developed, and on the left-hand side, you see exposure in this CT26 model to either control a STING agonist S100 or doxorubicin chemotherapy.
After exposure, you get induction of TREX-1, and you can see that on the protein level as well as expression level on the bottom left-hand side. In an experiment in an implantable CT26 tumor model, you can see in the middle, the right panel, they implant the tumor and start the TREX-1 inhibition early on, but then deliver chemotherapy at day seven and then follow the mice out over time. You can see the growth curve on the bottom left, where the TREX-1 inhibitor by itself doesn't have substantial efficacy, but in combination, you can see that it improves the doxorubicin treatment effect quite dramatically.
You can see that in terms of the tumor volume, both over time as well as on the right-hand side by day 25 in terms of overall tumor reduction. Just to summarize then, TREX-1 is a very interesting and it's a high priority therapeutic target. Tempest is actually, to my knowledge, the first in the space to go forward with a TREX-1 inhibitor, and hopefully very soon coming into clinical trials. STING agonists may provide a means to initiate immune responses and promote endogenous T cell responses, but the first generation of STING agonists actually have not really moved the needle, and the space remains open for drug development. Intratumoral deliveries remains kind of a big problem.
Here we're talking about an oral or a potentially systemic approach that could activate this pathway without having to deal with that. TREX-1 targeting represents a novel systemic approach to optimize STING agonism, to exploit synthetic lethality, and potentially to combine with many other therapies. Thank you very much, and I'm gonna hand it back to Stephen.
Thank you to all three of our speakers. We're gonna open it up to questions, and we have some off the chat function, but let's start within the room first. Algenay is gonna hand you a mic so our folks listening in can hear your question as well.
Thanks for doing this presentation. This is Maury from Jefferies, and thanks for taking my questions too. I guess starting off, for the monotherapy data for TPST-1120, I'm just wondering if you're seeing an increase in PD-L1 expression over time in those patients. I guess any other PD biomarkers that you're seeing there that could suggest that those patients are gonna be responders for a checkpoint inhibitor.
Yeah. Should I?
Yeah.
Can I come to the mic? Is that better?
Actually, yes, if you guys don't mind. I mean, whatever's easier. If you wanna.
You know, for passion.
The question was about biomarkers from the monotherapy portion of eleven twenty, and that work's ongoing. We don't have anything to show yet. I think, unofficially, what I'll say is that we've started to look at PBMCs from the study and just done some RNA expression data, and there does appear to be, at least in the subset that we've looked at, some changes to PPAR-related genes in peripheral lymphocytes that match the preclinical expectations of what we'd see, but a lot more to come, hopefully. We don't have serial biopsies, by the way. So we can't tell you whether PD-L1 expression of the tumor increases.
Should I just do a follow-up question as well, I guess?
While you have the mic.
For the two RCC patients, were they PD-L1 low at baseline?
Well, first one was PD-L1. Second one, I don't know if we can get that.
Repeating Sam's response, the first patient was, and we're not sure about.
Was PD-L1 negative.
Yeah.
The second patient, I'm not sure that PD-L1 staining was done, actually, on the second patient. Yeah.
We can't say that neither of them respond to prior checkpoint any company. With that.
Right. For the IDH1 observation, is that what are the next steps with that? Yeah, just wanted to.
Yeah. Some of that work is ongoing in actually my lab, so we're trying to understand whether there's some other data out there that IDH1 mutations are associated with increased fatty acid oxidation. Potentially this could be a subgroup of patients that are more sensitive to TPST-1120. We've made a mouse model of IDH1 mutated cholangiocarcinoma and are doing some comparisons of efficacy versus the parental cell line that doesn't have the mutation, and then trying to really understand why IDH1 mutations would induce fatty acid oxidation. We know that these tumors induce 2-HG production and totally cause metabolic rewiring of the tumors, but the mechanism is not totally teased out about why it would induce fatty acid oxidation.
We think if we understand the mechanism better, it would give some greater confidence that this is a tumor to target.
For proportion of patients in those different tumor types, can you talk a little bit about that? Just a number of patients.
The question was about the proportion of tumors in the monotherapy or combination or what's-
With the, I guess, tumors that you're gonna pursue that have IDH.
I should say I'm an expert in hepatobiliary tumors, so forgive me, I don't know a whole lot about IDH1 outside of my world. You know, IDH1 mutations are common in cholangiocarcinoma. They're found in about 20%, so about one in five intrahepatic cholangiocarcinomas, which is the dominant subtype in the U.S. These tumors tend to be not responsive to systemic therapies. There is an IDH1 inhibitor that's approved, it's ivosidenib, but the efficacy in cholangiocarcinoma is, I think, modest. You know, there's a randomized study, it showed a very small PFS benefit, about 1.6 months, no overall survival benefit, and it's approved and used.
You know, I think this is an opportunity potentially to develop something that could be more effective. Although, again, I have to defer to the Tempest team because they're the ones footing the bills, but probably a little more preclinical work to validate this biomarker would make sense.
Thanks, Maury.
I'm gonna just be handing the microphone back and forth, Mark. Matt Phipps, William Blair. Thanks, guys. Were any of the CCA patients in the combination, did they have IDH1 mutations, in the TPST-1120 plus pembro combo?
I can take that. One that we know of, who also did well on with the combination therapy.
As in durable stable disease?
Durable stable disease. Correct. It wasn't the responder, but it was another patient who did quite well with frontline study.
Mark, would you want to try combining this eleven twenty with one of those IDH1 inhibitor? Are you looking at that preclinical? Does that make any sense? Or would you rather try to say, "Okay, we could push immunotherapy, try to get these durable, more durable type responses?
That's actually the exact question we're trying to answer pre-clinically, so more to come. I think there's reason to think that dual targeting would make sense.
Yeah. I can just add, from a development standpoint, we'd be happy with either. There's clearly a path forward, either in combination with an IDH inhibitor or with IO. We are agnostic at this point and Mark is helping us to address the best mechanism of synergy in that pathway.
Different size patient populations, though.
True, yeah. The IO combo could be an all-comers study. We're very interested in both approaches, and hope to get data to support where to go.
Similarly, as far as just the background of some of these patients in the combination, any kind of beta-catenin mutations in those HCC patients? I know they're all lower doses, but just wondering if you've had a chance to look at that. Then do you think there's like what do you think is going on in RCC? I mean, you've talked about it is maybe there's some fatty acid, but are there particular pathways that these, you know, their VHL mutations or something, is that present or what do you think?
Okay. I'll say what I can about that. We did have a beta-catenin mutant hepatocellular carcinoma who had multiple stable disease scans. In general, as I think everybody knows and you can see on the data plots, we treated very few patients with hepatocellular carcinoma on this study. Also, for what it's worth, they were all treated at the low doses. We think the data suggests that there may be a dose response where it would've been ideal if they'd been treated at higher doses. We did have a few patients with HCC who had a decent response and a beta-catenin mutation, which Mark was not able to go into all the mechanism behind why that's important to us.
This is another potential predictive biomarker for the drug, and that, for what it's worth, that individual patient, did well. RCC is interesting because, I personally don't understand all of the intricacies of how IO works in RCC. It does seem different. For example, PD-L1 expression is not clearly as dominant of a predictor in RCC. RCC, however, is very metabolically driven. That's known. It's a great space for metabolic targeting drugs. And there are actually impressive data, preclinical data in RCC models showing, the really potent activity of inhibition of fatty acid oxidation, and PPAR α inhibition in RCC. Without knowing the exact mechanism, it probably relates to the metabolic profile of RCC.
The nice thing, I guess, about TPST-1120 is that we've got both the fatty acid oxidation pathway in the tumor, but also the ability to hopefully activate the immune system and improve upon that checkpoint therapy.
Thanks. Dr. Luke, last question for you. So with this TREX-1 inhibitor, would you like to combine with radiation versus chemotherapy? I guess maybe throw PD-1 on top. And then, you know, I think there was a lot of worry initially when the STING agonists were coming out, any systemic exposure could cause massive CRS and things like that, but I don't know if it's really played out as bad. You know, there are companies that are looking at systemic STING agonists now. So maybe thoughts on a systemic STING agonist versus systemic TREX-1. And then last part of this long question, there is. I guess is there a TREX-1 human phenotype of like a severe autoimmune formulation? So do you worry about if you hit it too hard, you might push it all the way to that?
There is a human phenotype of TREX-1 loss. It's not an extreme autoimmune phenotype, but it is there. In terms of TREX-1 versus STING, I think the idea here is that TREX-1 inhibition systemically is something you might be able to titrate and then bring sort of on and off in a way that could be advantageous relative to the combinations you might think about. To your question, it's kinda hard to know about safety considerations when there's no data in clinic, right? When we started the first in-human studies of STING, yeah, everybody thought the patients would like melt when we like gave STING agonist.
That's why FDA made us start at like, you know, doses that were like, you know, even the mouse wouldn't touch, it wouldn't touch the mouse kinda thing. I think the answer there is we probably don't know the answer. But I think TREX-1 is very attractive because of all the potential combinations. Now, which one would you want to pursue? I think that's a development question actually, 'cause you make a pretty good hypothesis as to all of those being pretty useful. I think the initial approach would be to, you know, position, I would guess, but I'll defer to the Tempest folks. You know, kind of the same standard drug development path that we've kind of been seeing here.
Then if you can actually think you are showing the pharmacodynamic effects you would expect, then you have like a whole world kind of open up in front of you, because if it's really just bringing down the potential to activate the pathway, you're not actually really putting the patient at systemic risk immediately. You have to do something else to kinda turn it on. And that is potentially quite advantageous relative to a potentially toxic direct agonism of the pathway in a powerful way. But I think some of these questions, it's hard to know 'cause they're comparing some future where we have no data to, you know, what we've got. There are also. Sorry, one more question online. Somebody asked about comparing the datasets from the Novartis data to Merck and all this kind of stuff.
What I'll say there is that the data reporting in the field of STING agonist is a gigantic mess, and I would not recommend that people take their time to try to do that. Just for reference, when we went to put the paper out, we had to basically go back and scrub all the CRFs because it was a gigantic mess. Merck has, functionally speaking, not disclosed any real data other than one poster five years ago or something. I don't think you can really compare across. That's one problem with this field, is that we're still kind of neophyte trying to move out of preclinical mouse models into humans because literally, that CCR paper I mentioned is the only report in the field of humans being exposed to STING agonism.
We just don't have enough evidence or data yet to really start to really understand what would be the best way to do this.
I have to still maybe because there.
Tom, yeah, they need to hear you online.
The Tempest team thinks about translation for this molecule, and Jason was just touching on that, is that we really think one of the exciting features of the biology of TREX-1 is that it does participate in DNA repair. You can imagine a first-in-human study and actually TREX-1 helps tumor cells avoid immune recognition by preventing activation of the STING pathway. We would like either with particular combinations, Jason talked about radiation, but also particular mutation profiles in tumors like [BRCA1], for example, where there's known to be an upregulation of TREX-1, and obviously deficient DNA repair that we can target TREX-1 in that scenario, both to at least ostensibly inhibit tumor proliferation, but at the same time, obviously activate the immune response too.
Hey, it's Matt Biegler, Oppenheimer & Co. Could you just talk maybe about how you for eleven twenty arrived at the 600 mg dose for the recommended phase II dose? Are you looking at saturating certain targets or is there some kind of biomarker that we know that we're really hitting FAO hard? Thanks.
Yeah. Thanks, Matt. This, the starting dose for the dose escalation was based upon preclinical predictions and was, you know, sitting in that between the efficacy and the toxicity profile for the drug. We went up to good exposures based upon preclinical modeling in patients. From that perspective, we're definitely comfortable with the 600 mg dose. I also can say that that except at the very lowest dose, you know, our trough exposures are above the IC50, protein binding IC50 of the target. That said, it's a super tolerable dose of the drug in patients at all the doses that we've tested.
It's been a challenging drug to determine PD from the lipid perspective, which is the mechanism of action of the drug. We think about the possibility of dosing higher, which would clearly be feasible. The good news is we have activity, we have responses at the current dose. I think in our pocket as we reformulate, as we make tablets and we go from capsules to tablets, we also could be exploring higher doses and potentially even more efficacious doses. I'd say it's not a complete answer, but we are comfortable that we're based upon all the data, we're in a good range for the dose. We'll simply say that super tolerable.
At this point, it looks like we could push the dose if we wanted to.
I think it was Matt. Did you ask about why the RCC patients might have responded or was it more? In any case, getting around Toni's lack of voice, this is his response. "RCC is the epitome of a metabolic disease. If you look at all the TCGA, it is one of the cancers with significant alterations of the metabolic pathways. PPAR α is a TF for genes that includes enzymes involved in the energy metabolism in RCC, even independent of VHL." So he asked me to read that to the room.
Technology getting around biology.
That actually. To just add to that, one of the interesting things from the TCGA analysis, VHL is a clear cell gene pathway, you know, loss, and that's 85% of clear cell RCC. The overexpression of PPAR α and associated FAO genes very high in chromophobe and papillary renal cell carcinoma as well. Again, see, just showing, I think, the metabolic perspective of this particular indication independent of VHL.
Any other questions in the room? I'm sorry, any other questions in the room? Algenay right behind you. No, in front of you now.
I'll keep asking questions. [Cyn], you haven't had a chance to get asked one yet. Is there any? You know, you clearly showed why EP2 and EP4 are better than either one by itself. Clearly, you know, people have tried COX inhibitors for a lot of clinical trials, very big things, and it hasn't really worked. Then there are some data presentations we've seen from EP4 solos that, you know, maybe have had some signals. I'm wondering if there's anything that you take away from those datasets as you read through the dual EP2/EP4. Then lastly, there's now some [HPK1] inhibitors moving into clinics, which are in the prostaglandin pathway.
Wondering if, you know, I guess how you think about going for EP2 there, that part of the pathway versus EP4, if you've thought about that?
I think, you know, with all the data that we showed with the combination, it's pretty preclinical. You know, we don't know for sure, but preclinically, it really looks like blocking both of them makes more sense than one alone. Doing the EP4, which is in the clinic, right? EP4 alone, but doing the dual with the EP2 and EP4 makes more sense. Yes, I mean, I use a lot of aspirin and I use a lot of, you know, in colorectal cancer for prevention. You know, I think, in colorectal cancer and in head and neck, this is something that we've been using for a long time. How can we do it optimally, right? Without toxicity and be more effective.
I think just based on the preclinical data, this is very promising. In the clinic so far it's been very well tolerated. If you look at the inclusion criteria, it's quite strict, like everyone is when you're first in human, right? Like, they cannot be on any anticoagulation. They cannot have bleeding risk. Just like with Avastin, you know, if everybody was very careful, but then you release more and more. It's very restrictive inclusion criteria. So far what we've seen with the daily dosing, it's been very well tolerated. Patients have very good quality of life. We have seen, you know, that patients are doing well on it.
Just sensitive to everyone's time. Let me take care of some of these questions from the internet. Mark, this first one's for you. Of the RCC responders, were they refractory or not tolerating the checkpoint inhibitors?
The two RCC responders had both received anti-PD-1 therapy previously. One received ipilimumab-nivolumab, the other received pembrolizumab. Both had progression on those agents. They had stable disease, then progression, and then got TPST plus nivolumab, and both had true deep PRs.
Okay. This is related to another question you had received, and this is also for us. What are the plans for translational data to show the immune component, especially in such late stage patients?
Do you wanna take that?
I'll take that. Mark said that one of the key points about our phase I program, we have maybe 1 on-treatment biopsy from our phase I study. You know, the target that we wanna look at is the tumor. Addressing PD-L1 expression, for example, is one of the questions not gonna be possible from the study. I will say that 'cause I always talk about the Morpheus study and the importance of that study. That phase I/II study, which is awesome from the perspective of a development even to approval in HCC and how to inform that, also has biopsies pre- and on-treatment.
It also has multiple biomarker analytics planned and characterization of beta-catenin in all the patients. For the question of what's the best PD we can do, I would say that's coming and it's not too far away because we're gonna have, you know, more than 40 patients treated in a response rate by the end of this year. The biomarker work that's been done from the Tempest phase I has been primarily Nanostring gene expression, and it's primarily looked at lipids. I will just say it's been challenging in the lipid profile of patients to tease out, you know, kinda reproducible signals.
Even though we've had some patients who've done well, who've had a, like a classic PD response, it's just been hard to translate that across multiple patients.
Jason, historical question for you. Could you please remind us, weren't the Aduro data patients also worse off than the Merck data, which gave the STING killing perception to investors and physicians alike?
That's why I tried to answer that question before.
Yeah.
This is broadly misunderstood in the field, right? There are just two posters that have been presented. People made kind of broad, sweeping generalizations around the field based on almost zero data. Well, I won't say more. It is very little data, and you surely shouldn't be trying to make really broad generalizations based on what's out there.
Thank you. Any other questions in the room? We were good online? Okay. Well, seriously, thank you everyone. I know this was early, and you have very busy ASCOs. Thank you to all three of our speakers and to Toni via text. We really appreciate it, and obviously, we're happy to follow up any questions after this, so please have a good rest of your ASCO.