Good morning and thank you all for joining Tempest Therapeutics' conference call. At this time, all participants are in a listen-only mode. As a reminder, today's conference call is being recorded. Following management's remarks, we'll hold a question-and-answer session, and at that time, the lines will be open. I would now like to turn the call over to Nicholas Maestas. Please go ahead.
Thank you, Operator, and good morning, everyone. Thank you for joining our call to discuss the updated positive data from the global randomized trial of amezalpat or TPST-1120 in first-line liver cancer. For your reference, a press release covering the news is available on the company's website at TempestTX.com. Joining me on the call today are Steve Brady, President and Chief Executive Officer, and Sam Whiting, Chief Medical Officer and Head of Research and Development. Turning to slide two, during this call, we will be making forward-looking statements that are subject to risks and uncertainties. Our actual results may differ materially from those described. We encourage you to review our risk factors in our most recent SEC reports on Forms 10-K and 10-Q, which can be found on our website at TempestTX.com.
While we may elect to update these forward-looking statements in the future, we specifically disclaim any obligation to do so even if our views change. With that, let me pass the call over to Steve.
Thank you, Nick. Good morning, and thanks to everyone for joining our call to discuss the exciting new data from Tempest's amezalpat program, previously known as TPST-1120, in first-line liver cancer, also called HCC. As you may remember, the first clinical data from the amezalpat program was a phase I oral presentation at ASCO in 2022, while a randomized global phase I-B2 study was ongoing that combined amezalpat with atezolizumab and bevacizumab, the standard of care in first-line liver cancer, and compared it to that standard of care alone. In the phase I presented at ASCO, we saw interesting clinical benefit with amezalpat monotherapy in difficult-to-treat late-line tumors and observed robust responses when amezalpat was combined with nivolumab, notably in patients who had been treated with checkpoint inhibitors before but had never responded.
We're highlighting this because those data support one of the hypotheses behind this program, that is that amezalpat could potentially enable a long-sought-after grail in cancer drug development, that is, to turn cold tumors hot. We'll go deeper into this when Sam discusses the updated randomized data. Moving to slide number three. At its inception, amezalpat was developed to target PPAR-alpha. PPAR-alpha is expressed on both cancer cells and immune suppressor cells, so our hope was that amezalpat would have a dual mechanism of action, that is, work both on the cancer cells directly and target the suppressor cells that keep the immune system from doing its job to fight tumors.
Given this potential dual mechanism of action, also known as MOA, we reasoned that it could bring benefit to a broad scope of patients who are normally difficult to treat with traditional IO, including those with the so-called cold tumors. We're excited about the data we've observed on this front, both from those earlier slides in the phase I combination study and even more so in the phase II data. On slide four, you can see the different tumor types that express the PPAR-alpha target. All the way on the far right is liver cancer as the highest expressor, so we've been interested in liver cancer since the beginning as a potential place to help patients. In addition to liver, you'll also notice kidney cancer and cholangiocarcinoma as high expressors, two types of cancer where we observed benefit to patients, including durable responses, in the phase I study.
We believe amezalpat has potential in these and other cancers, although the focus today is on first-line liver cancer. Another area that we've been thinking about since the start of the program is the role of beta-catenin in PPAR-alpha regulation generally and in liver cancer specifically. Although amezalpat doesn't target beta-catenin, the literature shows that mutations in beta-catenin upregulate PPAR-alpha, which made us interested in the potential for increased benefit to patients. Said differently, although amezalpat could help liver cancer patients generally because the target is broadly expressed in the tumors, we wondered if it could also help those with the mutation in the beta-catenin gene even more. Based on the data to date, the answer to both questions is yes, and we believe this provides built-in advantages in the overall liver cancer population when amezalpat is combined with standard of care.
We've had two prior announcements of data from this liver cancer study, the first in April 2023 and the second in October 2023, which, as you know, was received very well. The April discussion was an early look at some top-line values such as objective response rates and number of patients on therapy and on study, whereas in October, we discussed a more comprehensive per protocol analysis. We've been very happy to watch the data improve over time, with the October discussion embodying a broad set of both clinical and biomarker data that not only show an improvement for patients in the amezalpat arm in every category when compared to the standard of care arm, but also exciting biomarker data that we believe illuminate the way amezalpat is helping patients.
However, in the fall, we didn't yet have a key piece of information, which is the median overall survival, or also known as median OS. OS is the primary approval endpoint for first-line liver cancer worldwide, and although it had been reached at 15 months in the control arm as of the earlier data cut, the patients on the amezalpat arm were living longer, and it had not yet been reached. At the time, we did have a very favorable hazard ratio of 0.59 for the amezalpat arm over the standard of care arm, so we have been eagerly looking forward to the next data readout. We're excited to report to you today that as of the recent data cut in February, we have a median overall survival in the amezalpat arm of over 21 months, which is a six-month improvement over the standard of care control arm.
As you likely know, this is double what is largely viewed as clinically meaningful, so we are thrilled about the potential for patients here. As Sam will discuss, in addition to the absolute values, the survival signal embodied by the hazard ratio was also largely maintained in the 10 months since the prior data analysis, which is really exciting as we prepare for the phase III. So with that, I will turn it over to Sam, our CMO and Head of R&D. Sam?
Thank you, Steve. Moving to slide five, as a reminder on the study design, this is an ongoing global randomized standard of care control phase I-B2 study operationalized by Roche in clinical collaboration with Tempest. It is conducted in patients with unresectable or metastatic HCC being treated in the first-line setting. Eligible patients are randomized to receive either amezalpat plus standard of care atezolizumab and bevacizumab, which we'll refer to as atezo plus bev from here on, or atezo plus bev without amezalpat in the control arm. The study enrolled patients at 26 sites in seven countries in the U.S., Asia, and Europe. Study eligibility criteria matched with Roche's IMbrave150 approval study for atezo plus bev in first-line HCC. Per protocol, 40 patients were randomized to the amezalpat arm and 30 patients randomized to the atezo plus bev control arm.
An important aspect of this randomized design is that it models, in a phase II study, what would be reproduced in a confirmatory phase III study. For example, the patient population is regionally diverse, spanning U.S., Asia, and EU, as well as will be needed for the confirmatory study, and the key design attributes and selection criteria match the prior atezo /bev registrational study. Looking forward, we believe the straightforward path is to enlarge the study size for appropriate power to achieve regulatory approval statistics. As Steve said at the October 2023 presentation, the top-line data were very strong across multiple efficacy and safety endpoints, but they were early for the key regulatory endpoint of overall survival in the amezalpat arm, which was doing quite well and had only reached 25% of the survival events.
The latest data cut reflects an additional 10 months of maturity over the top-line analysis and a median follow-up time of over 17 months. This cut is also over 14.5 months after the last patient was randomized and allows us to report an overall survival with the comparisons of medians having been met. Shown on slide six are some of the key takeaways from the updated data. The patients in the amezalpat arm experienced a hazard ratio, or HR, of 0.65 for overall survival, reflecting a 35% reduction in the risk of death compared to the standard of care atezo plus bev arm. Importantly, this hazard ratio barely moved in 10 months since the primary analysis when it was effectively 0.6 and now is 0.65.
Also important is that an HR of 0.8 is generally considered clinically meaningful by treating physicians and regulators for the overall survival endpoint, so you can see that the result here of 0.65 is much better than is needed for the regulatory target. The patients in the amezalpat arm experienced a six-month improvement in median overall survival at over 21 months compared to 15 months in the standard of care control arm. Similar to the HR, this six-month improvement in the median survival is larger than needed for a clinically meaningful outcome, which is typically over three months. Also notable is that 50% of the patients on the amezalpat arm, 20 out of 40, remain in survival follow-up with a subset of those patients still on therapy compared to 30%, or 9 out of 30, on the control arm.
We were also pleased to see in the recent data cut that one patient on the amezalpat arm who was in a partial response by RECIST criteria at the top-line analysis had continued to respond and has reached a complete response, that is, a complete resolution of all measurable cancer. This was an immunocompromised patient by tumor biomarker analysis, and we believe would be a challenging patient for the standard of care without amezalpat. Also important to doctors, patients, and global regulators, the safety profile is manageable and consistent with what we observed in the phase I study. Let's dive into these points further.
On slide seven, important in any randomized study, the patient demographic and baseline characteristic data highlight a patient population that is generally balanced between the two arms, including some factors that are numerically potentially advantageous to the control arm outcome and fewer that are potentially advantageous to the amezalpat arm. This balance between arms supports the conclusion that amezalpat is the critical factor driving the difference in outcome between the two study arms. When we dig further into these subgroups at the level of survival outcomes, we see consistent benefit to amezalpat in all categories. More to come on that shortly. Slide eight is a nice summary and reminder that the amezalpat arm is winning across all efficacy endpoints, most importantly in overall survival, but also in difficult-to-treat patients who are PD-L1 expression negative and who have immunologically cold tumors.
Now, onto the new survival data shown on slide nine is a Kaplan-Meier curve that we would love to see replicated in our proposed phase III study. There was an early separation of the curves showing survival benefit as soon as three months after starting treatment. There was also consistent separation of the curves reaching to the far right-hand edge of the Kaplan-Meier. For those of you who don't regularly stare at Kaplan-Meier curves, it's important to note that there were 20 patients alive in study follow-up at the time of this data cut. None of those patients were enrolled early enough in the study to have reached 27 months at the time of that data cut, which is the reason why the curve drops off at the right-hand edge despite 50% of the amezalpat patients still being alive.
The key data points for comparing the two arms are the hazard ratio, reflecting every point on the curve, that is 0.65, and the landmark at the median survival, which is a six-month improvement. These results show meaningful improvement in survival coming from the addition of amezalpat to standard of care and the maintenance of this benefit over the last 10 months of data maturity. We believe this bodes well for a successful pivotal study to follow. In addition to the survival benefit in the overall population, as Steve noted, we have been very interested in how patients do with treatment based upon their individual immune and beta-catenin gene status, as those two populations give insight into the proposed mechanism of action of amezalpat.
Our pre-clinical data and the biology of the PPAR-alpha pathway targeted by amezalpat suggest that the amezalpat arm could have activity in immune cold tumors, which are normally challenging for immune therapy, and could work well in tumors mutated in the beta-catenin pathway because that mutation increases expression of the PPAR-alpha target. Slides 10 and 11 show tumor responses for each evaluable patient on the atezo /bev control and the amezalpat arm, respectively, overlaid over the immune status of that patient's tumor, and on slide 11, the beta-catenin status as well. On slide 10, the atezo /bev responses clearly cluster in patients whose tumors are PD-L1 positive and/or immunologically inflamed. This result is consistent with data from other immune checkpoint inhibitors and not just atezolizumab. On slide 11, the amezalpat arm responses occur irrespective of PD-L1 status or the immune phenotype.
This is a unique profile not seen with the atezo /bev standard of care or with other immune checkpoint inhibitors, and we believe reflects the unique mechanism of amezalpat added on to atezo /bev. Finally, although we observed that the confirmed response in the amezalpat arm rose in beta-catenin mutated tumors from 30%-43% with a 100% disease control rate, patients responded well to the amezalpat therapy even without the beta-catenin mutation because the drug target is generally overexpressed in liver cancer. I'll show similar data for overall survival a few slides from now. Slide 12 shows an analysis recently completed for the new data cut that we think is an important and positive view of the amezalpat activity. Shown here is the hazard ratio for overall survival in key populations that are predictive or prognostic for survival in first-line liver cancer.
This graph, called a forest plot, shows the hazard ratio for OS centered on one, which would be equal survival in the two arms, and deviating to the left of one if the amezalpat arm is better, and to the right of one if the atezo/bev control arm is better. While the numbers in each group are not large, the trend consistently favors amezalpat across the categories in both higher risk and lower risk populations. This is the pattern that we want to see in a forest plot, that the investigational treatment helps patients broadly across multiple categories and characteristics. These include the immune-positive and immune-negative tumors, the better and worse ECOG performance statuses, the high and low alpha-fetoprotein tumor markers, and the high-risk tumors invading the perihepatic vasculature or that have spread outside of the liver, and the low-risk tumors that don't.
And finally, liver cancers are rising in the setting of chronic viral infection, predominantly hepatitis B virus, which is very common in Asian patients, as well as the cancers arising in the absence of viral infection, which is predominantly alcohol use or metabolic diseases, which is more common in North American and European patients. On slide 13, we divided the amezalpat arm into two curves representing beta-catenin mutated and beta-catenin non-mutated, also called wild type, and contrasted those two curves with the control arm patients. The dotted blue curve are patients with a beta-catenin mutation, and the solid blue curve are wild type patients. As you can see, the patients with the mutation do better than the overall population with a median OS of 23.5 months. Just like the improved response rate, this result supports the proposed mechanism of action of amezalpat.
But clearly, the wild type patients are also doing very well. They maintain the overall population benefit of 21 months, and clearly, they exceed the control arm patients. So amezalpat appears to benefit both populations, and both populations are the patients that we will be targeting in the planned confirmatory phase III study, aiming for a result that leads to FDA approval and global approval. We've spent a lot of time receiving the patient-level data since receiving the patient-level data, preparing for our upcoming FDA meeting, and studying patient safety. Slide 14 summarizes the safety data from the updated cut and is very consistent with the previously reported top-line analysis. As was seen in our own phase I study, amezalpat has a manageable profile of adverse events, mostly in the lower severity, Grade 1 and Grade 2 range.
Adverse events of high grade, or meeting the regulatory definition of serious, are very similar in the two arms. Fatal adverse events are also similar in the two arms, as are adverse events leading to interruption or discontinuation of study treatment. Only one patient of the 40 treated with amezalpat, 2.5%, reduced their dose of amezalpat treatment, which we are very pleased with and believe demonstrates a tolerable safety profile even in the setting of triplet therapy with atezo /bev. We view these data as having a positive read-through for the planned phase III global study. All of this leads to designing the right trial to win global regulatory approval for amezalpat in combination with atezo plus bev in patients with first-line HCC, shown on slide 15.
The rules for that approval start with running a quality randomized phase II study that reflects the approval trial population and shows a meaningful benefit in an approval endpoint that, if replicated with a larger sample size, would be statistically significant. That is a checkmark for us with this randomized phase II study design and the overall survival benefit shown in the amezalpat arm. Then, we need to expand that trial to a sample size that is powered to demonstrate a clinically meaningful benefit in the overall survival endpoint with statistical significance. We believe that the hazard ratio for overall survival that is clinically meaningful for doctors, patients, and regulators is around 0.8, or a 20% reduction in the risk of death. And the phase II results we were discussing today is 1.5x better than that target.
We believe that the improvement in the median OS that is clinically meaningful for doctors, patients, and regulators is around three months, and the phase II result is twice that. Those numbers give us a buffer in the study design to boost the probability of a successful phase III, that is, demonstrating a statistically significant and clinically meaningful benefit. The key aspects of the patient selection strategy and patient management are taken directly from the phase II randomized study to replicate that result. We believe we also have the opportunity in the pivotal study to incorporate an early look at the overall survival endpoint that could shorten the time to readout by a number of months if reached, which we plan to discuss with the FDA and global regulators.
We have been speaking with HCC doctors from around the world to discuss these data and the pivotal study design and are gratified to be batting 1,000 so far on their interest in these amezalpat randomized data and their desire to participate in the pivotal study. We're excited to be on course to open that study in the first quarter of 2025. And with that, I will turn the call back over to Steve for his concluding remarks.
Thank you, Sam. As you can see on slide 16, liver cancer affects a large number of people, and it's an aggressive cancer with high mortality rates. There is a significant need and opportunity to improve patients' lives with new therapies. The cumulative data package for amezalpat, consisting of clinical benefit for patients and biomarker data that flips back to the foundational hypotheses of the program, filled the team here with confidence about the path forward.
We look forward to discussing the proposed phase III study with FDA and global regulators. And based on current estimates, even without an early stop for survival benefit, we believe we could have top-line data in less than three years from the first patient enrolled, a relatively short time given the potential magnitude of the benefit we could see from the amezalpat combination therapy. This concludes our team's prepared remarks, and we're happy to open the call for Q&A.
Thank you. Ladies and gentlemen, if you have a question or a comment at this time, please press star one one on your telephone. If your question has been answered and you wish to move yourself from the queue, please press star one one again. We'll pause for a moment while we compile our Q&A roster. Our first question comes from Edward Tenthoff with Piper Sandler. Your line is open.
Great. Thank you, guys. Great data. Just exciting to see this continue to build on what you showed last year. I wanted to kind of get a sense, obviously it makes sense to move forward into the phase III and I'll be curious to see what the FDA has to say. Is this a trial that you guys think you can do on your own? Has Roche expressed interest? Would you potentially design the data with other anti-PD-1, PD-L1s, and TKIs to maybe make it so that this could be used broadly in first line? Thanks so much for answering the questions.
Yep. Hi, Ted. And thanks so much. We appreciate the congratulations. So Roche is obviously absolutely aware of the data and knows what we're planning to do. It is important to note, though, that we have 100% both the right and the ability to do this study alone. And in fact, I think this team's and our board's view has evolved over time as the data have just continually improved. If we had spoken last year, I think we felt more, you could say, backed into a corner to have to do a partnership. But given the experience on the team and the strength of this data package, our base case is to do this alone. And not only do we have the legal right to do that, but we have the ability to do it.
With respect to subbing in different molecules, one of the philosophies inside this company is, with respect to getting a drug approved, is to design a trial with the highest probability of technical success. So as you know, a lot of people will design smaller studies to try to save money. In this case, we're doing what we would call or proposing a proper phase III that will be powered to win. And we believe this is more likely than not to win. But the second part of that is to not change the variables that made your phase II successful. So while there is a mechanistic argument to say sub in a PD-1 instead of atezo and to put in a TKI because of anti-VEGF components or anti-angiogenic components, we believe the highest likelihood of success is to just expand what we just won in the phase II.
That makes a ton of sense, and I'd love to see you guys run this yourselves. It seems like a really good return on investment for $1 billion opportunity for a three-year trial. So thanks for the additional color.
Yep. Thank you, Ted. Appreciate it.
Our next question comes from Maury Raycroft with Jefferies. Your line is open.
Hi, good morning. I'll add my congrats and thanks for taking my questions. For the forest plot data that you showed, it looks really good even with low numbers of patients there. Assuming FDA agrees to your phase III design, can you talk about strategy to help ensure that the phase III patient baseline profile with consistent proportions across the subgroups can look similar to the phase II? Maybe talk in particular about the immune cold tumors or the approximate 60% of patients who are PD-L1 negative, because some of the comp studies that we're looking at in HCC are only enrolling about 20%-40% PD-L1 negative. So just maybe talk about being able to get that higher proportion of patients in your phase III.
Thanks, Maury. One of the important points of running a large-sized randomized study is randomization is allowed to, by its nature, equalize patient numbers between the two arms. We will have four different stratification factors in the study to balance what are considered the dominant prognostic outcome factors. Those will actually be very similar to other approval studies. The PD-L1 negative patient population, we believe, and really the literature supports, is almost always in the 60%-70% range across the liver cancer first-line population.
That's in a number of real-world studies and in retrospective studies. There have been a few studies, I believe, where the PD-L1 negative population was in the 40%. For example, actually, the atezo pivotal study was 43% PD-L1 negative. But even this study was 60% PD-L1 negative in our arm, sorry, in the control arm, and 67% PD-L1 negative in the amezalpat arm. We are expecting that. The other thing just to note, what we think is encouraging from the forest plot is that the amezalpat arm has good benefit straddling both sides of the immune equation, for example. So the hazard ratio for benefit is actually even better in the PD-L1 positive patients than it is in the PD-L1 negative patients. We see this as a win, really, for the amezalpat arm, irrespective of the immune status of the patients.
We just focus on the fact that the immune cold tumors and the PD-L negative tumors are a very challenging population for most IO therapy historically. And that gives us kind of a unique profile with the amezalpat drug. I don't know if I missed part of your anything else in that.
No, it sounds like you're confident you're going to be able to enroll consistent subgroup populations into the phase III as the phase II.
We do, yeah. So PD-L definitely. It's just a super common status. The beta-catenin mutation also is in the literature between 20% and 40% of liver cancer. What we're most comfortable about is this is another category where we look like we beat the control arm on both sides of the mutation status. And so we absolutely will pre-specify and look at the outcomes in mutation patients versus wild-type patients. We believe that those analyses will be published, excuse me, will be published. If they're advantageous for the amezalpat arm, that ultimately, as doctors have next-generation sequencing in hand when they treat patients and they look at the mutation profile of patients, if they see a beta-catenin mutation, they will, and they know that that's actually a marker that's important for the amezalpat drug that will be getting those patients diverted to us.
Got it. That's helpful. Maybe one other quick question. For the three-year timeline for the phase III, that's helpful that you guys are providing that. Can you talk more about when the futility analysis in the phase III could occur and when the early efficacy analysis could occur and what the drivers of those analyses will be?
Yeah. So what I can tell you is that when they normally occur is kind of available when you look. Futility analyses are generally early in the study, or as early as is reasonable, but not so early that they are at risk for a false outcome leading, in the worst case, to early inappropriate termination of the study. Those typically come at about 30% of the planned events, in this case, planned overall survival events. On the flip side, the early look at efficacy is generally delayed until a signal there is considered mature enough to be reliable. Those typically come, and this is again with regulatory feedback, those typically come at about 70%-80% of the projected overall events.
We're anticipating that, A, this early readout could be positive for us because the stats tell us that the outcome would actually not have to be as good as our phase II results for an early stop, and also that it would shave a number of months off of the time of enrollment or the time to readout doing the analysis that way.
Got it. That's helpful. I'll hop back in the queue. Thanks again.
Thank you.
Thanks, Maury.
Our next question comes from Aydin Huseynov with Ladenburg . Your line is open.
Good morning, everyone. Thank you for providing the details and congrats with the data. So appreciate the three-year guidance on the trial. But could you provide any more details in terms of how much it is going to cost for Tempest to run this trial?
Yeah. So we're not revising the estimate that we've given previously. Aydin, so it's over $120 million-ish. I think we said. I can say that we have been continuing to optimize costs since we began this process and realized that we had a phase III ready asset, and we will continue to refine that. The nice thing is that with the strength of this data package now, we believe it's eminently financeable. And because we think this is more likely to win, it actually flips over to why we said, in response to an earlier question, that our base case is to retain all the commercial value and do this with all haste.
Okay. Understood. Any plans for other indications for 1120?
Right now, our base case is to do this, is to run the study independently. That's not to say that something couldn't come after this announcement. It would just have to be because we think this is going to win and because the potential benefit, value creation on the other side is so significant to shareholders, to give up commercial rights now, we'd have to be a very exciting deal. Now, of course, on the other side, although we view liver cancer as an indication into which a growth-stage biotech could launch into North America, globally, you could see us doing something ex-U.S. or obviously strategic transactions could always happen. But yes, we do see an eventual opportunity for partnership with respect to global distribution because remember, this is largely an ex-U.S. indication. So yeah, it's absolutely a viable event in the lifecycle of the amezalpat franchise. Yeah.
Yeah. Understood. And the last question for me. So yeah. Sorry, go ahead.
Oh, I was just going to say it sounded like the other part of your question was other potential indications for the drug. We actually are very interested in additional indications based going back really to our phase I data. The primary kind of the lead second candidate is renal cell carcinoma. As you probably know, there's a lot of similarities biologically in the drugs that are used to treat liver cancer and renal cell carcinoma. So the standard of care, for example, in first-line renal cell carcinoma is an anti-VEGF therapy and immune therapy, just like in liver cancer. That gives us, we think, the opportunity to use the same kind of strategy and same successful combination in renal cell carcinoma.
It would be predicated on funding that study, but honestly, we've mapped out something similar in RCC that's the renal cell carcinoma in, for example, a randomized phase II first-line study where we could look at a standard of care IO VEGF doublet plus or minus amezalpat and ideally show a signal that's strong enough to warrant moving this drug into first-line RCC as well.
Okay. Thank you. This is very helpful. Yeah. And the last question that I want to ask is, I think you mentioned previously you were on business development conversations with large biopharma companies. So what were the most common asked questions or comments from your potential partners at the time? I mean, other companies that were on PD-L1, PD-L1, PD-L1 agents that they would potentially want to test in HCC?
That's a great question. And I want to be clear. We're not stopping all business development conversations. We just feel like because of the data, we have the strength to do this alone and preserve the commercial rights to the future where we could command an even higher value for shareholders. What is the most common question? I would say it was before we had the survival data, right? I mean, that is the primary global regulatory endpoint. And yeah, and now that we have it, I mean, the conversations take on a different tone. And like I said, even though we think it's in the best interest of everyone to run this alone independently, again, there could absolutely be a deal on the other side.
Or if something really interesting that we think is in the best interest of the program and shareholders happens sooner, we will absolutely take it seriously. But yeah, definitely, Aydin, we are not stiff-arming the large multinationals. We just believe the right course to generate the most value is to go quickly as a nimble biotech.
Okay. This is very helpful. Thank you and congrats again on the data.
Thanks, Aydin.
Appreciate it.
Our next question comes from Joe Pantginis with H.C. Wainwright. Your line is open.
Gentlemen, good morning. Thank you for taking the question. Congratulations on the data. We think the numbers speak for themselves. So the predominant questions that I have are forward-looking, if you don't mind, and maybe one data question. So first, do we expect anything further out of Roche going forward?
So we think that the next cut from this study, and obviously, thanks to Roche for transferring the data to us. We appreciate it to enable our FDA conversations, which was the primary reason we received this data cut. We think the next data cut will be when the study is done. Remember, Sam said that even though we have this beautiful hazard ratio and a curve that he would like to see replicated in the phase III, 50% of our patients are still online with a group still on drug. So the good news is continuing. We expect that when these patients are all no longer in survival follow-up, we'll have a final CSR-type data analysis. That's when we would think would be the next logical time to discuss the data. We don't expect that until sometime next year.
Got it. Then, look, the phase III looks pretty straightforward. So I guess I would ask the following. What do you feel, if any, are the key outstanding questions that you still need to address with the FDA? And then also, as you prep for the initiation of the phase III, how would you define your drug supply and availability?
Yeah. So on the first part, when Sam talked about designing the right phase III from a sizing and ability to hit statistical significance, the other selections we've made in designing the phase III are all, to quote someone who said this when they looked at design right down the fairway, it's the FDA, so you never know. But we don't expect any major bumps because all of the assumptions in that phase III we think are rational and actually, for lack of a better word, traditional. And then knock on wood, whatever this table is made of, our CMC process has been great. We do not see any problems with drug supply. And we have a great head of chemistry in CMC who is on top of it, and we expect that to be all on time.
Got it. Got it. And I guess my one data question, if you don't mind, is, well, I don't know if you have access to this data or if you've seen this, but if you were to look at a swimmer's plot of these patients' responses, how would you describe the kinetics of the response curves? Are there any particular patterns, or do patients respond at somewhat variable times? And also because you also have a patient that now has also converted into a CR as well. So anything about the kinetics of response?
Yeah. It's an interesting question. I guess when we've approached it is that we're seeing a pretty early separation of the survival curves between the two arms at around three months.
That includes early responses as well. I do think that the kinetics of the response are consistent with the mechanisms of the three drugs together. And even if he's above, can have early responses in patients. One of the classic patterns of tumor response for immunotherapy, as you probably know, is a delayed response. Excuse me. As the immune system takes hold and starts to control tumor, that can take a while sometimes to start. Some of the classic IO-only regimens, anti-CTLA-4 drugs like Yervoy, the benefit comes way out on the kind of right-hand edge. So it is interesting to see our complete responder is somebody who at over nine months of follow-up was still in a partial response, and then subsequent to that has continued to have tumor shrink.
His tumor reduction was about - 50% something at the top line, and now has hit, obviously, a 100% reduction. So that's a late evolution that's pretty classic for the immune mechanism. But we think that this is on par with what we would expect.
Fantastic. Appreciate all the comments, guys.
Thank you, Joe. Appreciate it.
Our next question comes from Matt Phipps with William Blair. Your line is open.
Morning, guys. Congrats on this nice consistent data update here with more follow-up. Great to see. I'm curious, as you think about the phase III trial, you mentioned a little bit earlier, but do you think you need to have any sorts of those stratification factors such as age enrollment or some of the viral etiology?
Okay. Yeah. So we anticipate that enrollment will be, I would say, brisk in Asia, for example, in China because of the size of the global population and incidence in Asia. We have absolutely intended to make sure we meet the minimum enrollment in both China and Japan.
Environmental etiology and other factors. Are we getting past any of them?
Oh, I'm sorry. I was thinking, I thought you were asking about regional enrollment. Okay. So for the patient characteristics, we don't believe we need to cap any of them. I think we're in a lucky position where we can basically take advantage of the frequency of these characteristics in patients and let that run. For example, the viral versus non-viral. There's a benefit to amezalpat shown in the phase II in both populations, and we don't need to fuss with that.
Okay. Thank you. And just curious to think your thoughts on the recent CheckMate 9DW study and if that has impacted your thoughts around kind of market share for different therapies in the front-line.
Yeah. Thank you. So the CheckMate study, this is the BMS's study of nivolumab and ipilimumab combined together in first-line liver cancer. That was reported out top-line data at ASCO in early June. It was a nice result. They have a positive result from the perspective of overall survival. It was actually, honestly, like we were just talking about before, very much like an IO regimen. The benefit ultimately was on survival and not earlier endpoints like progression-free survival. How that's going to play out is not certain. As a physician, I'm an advocate of funds that help patients getting into the clinic, and I think that this regimen will.
The combination of ipi and nivo has been approved in a number of other indications over the years. I will say it never picked up a dominant position because of the toxicity profile of this kind of all IO and somewhat non-specific mechanism of the anti-CTLA4. Many of the doctors we talked to think that this regimen, A, will be approved and B, will compete well with the STRIDE regimen with AstraZeneca's durvalumab and tremelimumab combination. It's really the same exactly as the drugs. But it's not clear that it's going to move any significant way into the atezo/bev market, for example, because doctors are more used to dealing with costs of atezo/bev and being able to manage that.
Great. Thanks, Sam. And congrats again, guys.
Sorry, Matt. We couldn't hear that last part.
Sorry. I said congrats again. Sorry for breaking up.
No problem. Any additional questions? Okay.
Our next question comes from George Farmer with Scotiabank. Your line is open.
Hi. Good morning. Congratulations from me on your results. I wonder if you could comment a bit on the maturity of the data that you presented. It looks like only 50% of events have occurred in the triplet arm. And given that, how do you see these results evolving over time? And does this have any impact on your powering assumptions? And then secondly, could you comment a little bit on the competitive landscape now that there's a number of studies out there that have read out? And what do you see going forward as being some other things that we maybe should watch out for? Okay.
So I'll take the first part, and Steve will take the second part. One of the things that's notable about the survival curve for this study is that there's 20 patients still alive and ongoing on the amezalpat arm. Ultimately, that's a good sign, right? That means that more than 14.5 months since the last patient started dosing, we still have 20 patients that are ongoing and alive. Included in that group, obviously, is patients who are even still on drug therapy. The vast majority of the patients who are alive are already beyond the median survival of the control arm. So no matter what happens to them, we have a regimen that looks superior to the control arm. It's possible to look into the patient-level data at the patients who are alive and censored and try to get a feel for what's coming.
I hesitate to say too much about that, but I will say that included in the patients who are alive and ongoing, and this may be as obvious, are some patients who are doing really well. The complete response patient with no measurable cancer left. Some of the deep partial responses in the -50% to -60% range are ongoing. We're optimistic that they're going to continue to march out alive and continue to extend that survival curve to the right. Yeah. I would just say that in terms of the design, we think that all of this bodes well for their design. And as I said before, I think one of the important points is we're being conservative in our assumptions in order to maximize the PTS for a positive phase III.
Yeah. And I would add on the competitive landscape, George. So Sam just touched on the recent ASCO data. I would add that I have the same experience with respect to the effect of the BMS data on shifting the competitive landscape. What I heard from talking to PIs and KOLs when I was there is that they didn't think it was going to shift into the atezo/bev market share piece. And whether that's because of the tox that Sam noted or different patient population between the two studies, the two therapies making them not comparable, that's open to debate. I think what's important from an evolving competitive landscape and what we're going into and what we'd be launching into on the other side of a successful phase III, at least in my experience, this is rare.
To our knowledge, we're going to be the only other ongoing randomized phase III going against the dominant standard of care.
Usually, that's not the case, right? You've seen in the past few years, there have been attempts to take market share from Roche, but they haven't exactly, well, they haven't worked. It's not that they haven't exactly worked. They haven't worked. So we'd be launching into a largely unchanged commercial landscape than what we're seeing effectively now if the points that we've made hold true. So we think it's a really unique opportunity.
Okay. Great. Thanks very much. And congratulations again.
Thanks, George. We appreciate it. Yeah. We're really excited about the data.
And I'm not showing any further questions at this time. I'd like to turn the call back over to management for any closing remarks.
Thank you. And thanks, everyone, for joining the call today. As we've said multiple times, we're very excited about these data and the amezalpat program, and we look forward to keeping you informed of our progress. With that, operator, please feel free to close the call.
Ladies and gentlemen, this does conclude today's presentation. You may now disconnect and have a wonderful day.