Okay, great. Welcome everyone to the first day of the Virtual Oncology Summit hosted by Citigroup. I'm Yigal Nochomovitz, one of the biotech analysts. The next session is with Rain Oncology, and it's my pleasure to have with me the Co-Founder, Chairman, and CEO, Avanish Vellanki. Welcome, Avanish. Thanks so much for taking the time.
Thanks for having us, Yigal.
I think a lot of people are pretty familiar with the story. Maybe it'd be helpful if you could just do, though, a quick, you know, two to three minute elevator pitch on the company, and then we can jump into the key questions on your upcoming big phase III readout.
Absolutely. Happy to. Rain Oncology is a precision oncology biotech, of course, based in the Bay Area, and our lead program is a small molecule oral inhibitor of the MDM2 P53 interaction. The goal here is to focus on reactivation of the guardian of the genome, something we think is gonna be tremendously important across a multitude of tumor types.
We believe that this molecule has the first well-tolerated treatment algorithm of any program in the landscape. We're excited to have jumped into a phase III study in patients with dedifferentiated liposarcoma. Unfortunately, a tumor type that has gone really underappreciated by investors, really because it's not well understood.
Unfortunately, patients have a very poor prognosis, and the standards of care do not adequately provide a meaningful avenue for a lengthy treatment cycle. We started a clinical trial, a global registrational phase III study in middle of 2021. We completed that in July of 2022, called the MANTRA Study, and we expect the phase III readout sometime next quarter.
Because of the larger vision for our program, milademetan, we believe the opportunity goes far beyond liposarcoma in a multitude of indications that could become MDM2 dependent. We also have the MANTRA 2 study ongoing, enrolling patients today in the U.S., in patients that have a certain degree of MDM2 gene amplification. We also plan to start the MANTRA 4 study around middle of the year.
Our first combination study to look at milademetan in combination with Roche's atezolizumab in advanced solid tumor patients, that have lost function of the gene CDKN2A and therefore, post the liposarcoma strategy, we have two basket trials, each with a unique biomarker strategy.
All right. Awesome. Let's start with the tip of the spear, so to speak, the MANTRA trial, which is reading out any, I guess, Q2. Obviously the key questions are tell us what are the reasons why we should be confident in the outcome, a positive outcome, and of course, you know, to the extent that you can comment on what you see as the residual risks or potential unknowns that are obviously hard to handicap, be great to kinda cover that also. Thanks.
Sure. The reasons for confidence, I think are twofold. One, going back to dedifferentiated liposarcoma as a tumor type that has an unmet need. The standards of care are unfortunately for patients rather poor, with a median progression-free survival of around two months. We know from other trials in this space that a two-month median PFS is approximately the PFS that we see with placebo.
The standards of care effectively are giving nothing to these patients. The first reason for confidence in the outcome of the phase III, Yigal, is the hurdle for success is extraordinarily low. We effectively have to beat a level of efficacy that's equivalent to placebo. That's reason one. Reason two for our confidence is the prior clinical data.
There's no data, nothing to give investors confidence like prior patient experience on the drug. We licensed this molecule from Daiichi Sankyo. Daiichi Sankyo ran a very large phase I study when they were initially identifying the dose and the appropriate schedule.
107 patients, 53 of those patients were dedifferentiated liposarcoma. Across all 53 patients, before they optimized the dose, before they optimized the schedule, they saw a median progression-free survival of 7.2 months. 7.2 months, tripling to quadrupling of progression-free survival versus that historical data for the standard of control, standard of care. I think that gives us a lot of confidence.
Daiichi went further, they did a lot of the heavy lifting here to quantitatively model and put into practice a specific schedule that they felt that solved the historical challenges of this landscape of MDM2 P53 interaction inhibitors. Through this optimization work, they saw that they were able to go to a much higher dose, bring AEs meaningfully down to the point where grade 3/4 cytopenias were relatively modest.
The bigger vision here for milademetan in this program is far beyond liposarcoma. The bigger vision is if you have a first well-tolerated MDM2 P53 reactivation inhibitor, is the ability to combine with a wide variety of agents to, for the first time, add P53 reactivation to the arsenal of mechanisms that could help counter cancer. That's the bigger vision.
Liposarcoma, again, we think is lower-hanging fruit because of the low hurdle for success with the standard of care and the prior clinical data that shows a tripling to quadrupling, versus trabectedin, our control arm.
Awesome. Tell us a little bit more about that. Just how you anchoring that, I think it's two or 2.2 months PFS for the Yondelis and the HALAVEN. Just reference, you know, where does that come from and do the KOLs generally agree that that's kinda what you would achieve with those drugs in DDLPS?
Sure. It comes from the pivotal study for trabectedin and Yondelis. The registrational study for trabectedin included patients of mixed tumor type. It was leiomyosarcoma and all subtypes of liposarcoma. It was a mixed bag. Liposarcoma is not a uniform mix of tumor types or biology.
But when they teased out specifically the dedifferentiated liposarcoma patient population from their pivotal study, they saw a median PFS of 2.2 months. We don't see really any publication that is a less accurate predictor than that study. We think that is the most accurate predictor of that study, the 2.2 months. Now, having said that, this was approved back in 2015.
Is it possible that there's some optimization that has taken place from docetaxel to extend the treatment benefit of trabectedin? Of course, it's possible. We hypothetically modeled a 50% improvement on PFS from a 2.2-month PFS to three months. 50% improvement in PFS, is that possible? It could be possible. Does that provide us a little margin of error? Absolutely.
We assume a three-month median PFS in the trial design for the pivotal MANTRA study to account for some flexibility. Is it possible that trabectedin has a 100% improvement from a rigorous phase III trial? We think that's less likely. That's where we get confidence that the bookends for what trabectedin will show, you know, the upper end should be around three months.
To your second question, again, you asked about, do docs think that trabectedin could do worse? This is actually a really important point, and this came out with several other investor meetings that we recently had. There's various anecdotal comments from clinicians where they seem to suggest that trabectedin might do better than 2.2 months.
What's really important in that question from investors to clinicians is to highlight and interrogate the specific subtype of dedifferentiated liposarcoma. Our co-founder, Dr. Robert Doebele, who couldn't join us today, brings up on a regular basis that once upon a time in lung cancer, they thought treatments for squamous and adenocarcinoma were the same. It wasn't until Lilly's pemetrexed that you could actually get differential levels of activity.
We believe that clinicians don't necessarily see the subtypes. They don't necessarily are aware of the subtype differential level of response until you actually have a treatment for those subtypes. Today, there is no differential treatment for those subtypes. When we hear anecdotal comments that trabectedin might do better than 2.2 months, it's because of experience with trabectedin in the myxoid subtypes of liposarcoma where they do better.
Myxoid subtypes of liposarcoma do respond far better to trabectedin than dedifferentiated liposarcoma. When investors ask specifically about dedifferentiated liposarcoma, I think you will find, I think investors will find, that docs will have to scratch their head a little bit and recall because there is a dramatic difference in sensitivity to trabectedin based upon the subtype of liposarcoma. I wanted to clarify that because that is a question that comes up repeatedly in our discussions with investors.
Yeah. The other thing that comes up sometimes when I talk to investors, and maybe you can clear this up, is, you know, you have the well-differentiated, dedifferentiated phenotype. In your trial, you're enrolling patients that have some element of DD histology. Is that correct? I mean, or does it have to be exclusively DD, or as long as some of it is DD, then they qualify. How does that work?
That's exactly right. They need to have some dedifferentiated liposarcoma. Whether it's exclusive or not exclusive, doesn't matter for patient inclusion to the study. The official inclusion criteria is a diagnosis of dedifferentiated liposarcoma with or without a well-differentiated component. That's the official inclusion criteria, you have to have DD liposarcoma.
To take a step back and talk about the biology, I think it's important to mention that well diff and de-diff are on a spectrum. De-diff is thought to come from well diff. Where you do see some pure well diff tumors, they're not eligible for our study. When you see DD, you most often also see well diff. It's important to understand the biology that one progresses to another. They are both MDM2 gene amplified.
Okay. You don't see pure DD without WD, right? Is that your point?
Most often you see them together, correct?
Okay. when you reference the 2.2 PFS, those are patients that obviously had DD, but they probably or most likely had FWD also.
They most likely had some degree of well diff as well. Correct.
Okay. Okay. Very, very helpful. Okay. Now in terms of some of the background, if you could just kind of provide just a little bit of a maybe very, very quick history lesson. You know, how did you come to acquire this asset from Daiichi? You know, what did you see in the data that was interesting? They encountered some roadblocks, I assume, in terms of how they wanted to develop the drug, which you've pushed through. It would be helpful if you could just kind of give us just a quick background on that aspect.
Yeah, happy to. We saw the Daiichi Sankyo data set from their phase I study, the U-101 study in 2019 or so. We approached Daiichi Sankyo for interest in this molecule, at which point they very politely said, "Thanks, but no thanks." They weren't interested in licensing the molecule.
They had full intentions of developing the molecule on their own. We said, "Good luck." That was the end of the conversation, the first time around. What happened was, Daiichi had extraordinary success with their ADC franchise with trastuzumab deruxtecan and Enhertu, they saw fantastic data in the metastatic breast cancer trial, to the point where they were beginning to realign the entirety of their focus to being an ADC company. Appropriately so. It's fantastic data from Enhertu.
Because of that, they began to reevaluate the homes for various small molecules that they had in their pipeline, including this one. They came back to us the second time around, late 2019. They asked us if we were still interested. Of course, we were. That's how we were able to license the program. It's really because of the opportunity presented itself that Daiichi had success on the other side of their business. From our side, why we were so excited, it wasn't necessarily because of liposarcoma.
They had this clinical data in liposarcoma, which looked interesting, but what was more compelling was their tolerability profile. Now, we've talked extensively at lengths, Yigal, historically, about the problems that other companies have had with MDM2 inhibitors. MDM2 inhibition is nothing new.
Big Pharma spent a lot of money and spent a lot of time on this mechanism because restoring P53, the guardian of the genome, is a good thing. Theoretically, half of all cancers could benefit from it. That's why Novartis stepped in, Roche, Amgen all stepped into the space. What we think was not appreciated or anticipated at that time, those early-stage molecules, was the fact that restoring P53 or altering that MDM2 P53 balance in the bone marrow is a very delicate matter. The P53 MDM2 interaction is very tightly controlled in the marrow, and bumping it, offsetting it slightly, is gonna lead to ablation of blood cells, your platelets, neutrophils, red blood cells. That's why you see cytopenias as an on-target toxicity from this class of drugs. Completely on target.
The more potent of a MDM2 inhibitor you have, the more cytopenias you would expect to see. Completely on target. They didn't go back to the drawing board, but other companies did, Daiichi Sankyo included. The reason we got so excited is that if milademetan is the first molecule to truly address the cytopenia rate, then it's not liposarcoma what we should be excited about.
Liposarcoma is the low-hanging fruit at the tip of the iceberg. It's a small patient opportunity, about 1,400 patients in the U.S. It should, we hope, prove the thesis that now improving the tolerability profile allows patients to stay on therapy, stay on drug, moderate those dose reductions, dose interruptions, and now ultimately become combinable with a wide variety of other agents.
I think you were very quick to point out on several notes, Yigal, that after liposarcoma, everything that we're doing is bigger. MANTRA-2 is a bigger opportunity. MANTRA-4 is a much bigger opportunity. We fundamentally believe today, where there is no P53 reactivation commercial strategy, and as most tumor types truly do engage in a combination regimen, we can fill a huge void for clinicians by serving as a backbone of a combination regimen.
That's why we're excited. We got excited by the safety data first. The liposarcoma efficacy data was secondary. We thought that data, the liposarcoma strategy, would get us to market quicker, but the tolerability data would allow us to expand far more widely.
Okay. Makes sense. I think you sort of answered my question as to why you chose to pursue liposarcoma, even though there are other opportunities as an anchor, for the franchise. On, in terms of the thrombocytopenia, maybe you can get into a little bit more specifics as to, you know, what kind of levels of tox have been seen historically with the MDM2 class. You mentioned some of the pharma players that have tried. What have you achieved with the very, very specific three on, 11 off schedule, that has dramatically ramped that down?
Yeah, happy to. Historical data for grade 3/4 thrombocytopenia really being the core one, actually, you see differential levels of thrombocytopenia, neutropenia, anemia, the cytopenias of all these different molecules historically. One may have more thrombocytopenia, and one may have more neutropenia, but the class has been rather profound.
For this specific molecule, thrombocytopenia was always more the focus. We've seen other agents with thrombocytopenia grade 3/4 rates of 40%-50% or more, historical levels of eight grade 3/4 toxicities. That was substantial, 40%+. We've seen neutropenia in 30%, 35%+ levels for other competitor molecules.
What we see for milademetan was that when it was optimized for this three days on, 11 days off-Despite going to a much higher dose, 260 milligram, which is the highest or one of the highest doses of all the schedules evaluated, we saw thrombocytopenia coming all the way down to 15%. We see anemia effectively going to zero, neutropenia at a roughly 5% rate.
In our view, that profile, the 260 milligram dose, three days on, 11 days off, was the best AE profile that we've ever seen for any drug in this class. Now we'll see what some of the new generation MDM2 inhibitors show as new molecules are getting explored in the clinic and they devise dose schedules that are appropriate for them.
So far, I think we remain confident that we have the most inherently combinable agent out there. Today, the prior phase I study showed a 15% grade three or four thrombocytopenia rate, which is exciting to us.
It's interesting. I mean, what is it more specifically about the mechanism that allows you to, you know... It's an unusual situation, right? Where you can improve on the tox, but you're not sacrificing, apparently you're not sacrificing on efficacy. How does that happen? I mean, that's an unusual situation, no?
It's the recovery time, right? It's the short period of significant dosing, maximizing Cmax to induce apoptosis, backing off and allowing some sort of recovery time that is a function of the molecule's half-life, maybe some tissue accumulation properties, optimizing for this dance.
I think every player in this MDM2 landscape is trying to get a certain dance right between a short period of high dosing to induce cell killing, but then allowing the bone marrow to come back. Those platelets come back. In the specific workup that Daiichi did, they found that platelet recovery rate over 11 days was sufficient. That's what was optimized here. Now, different competitors that are out there are some molecules that are far more potent than milademetan.
Greater potency carries greater heme tox, and maybe that there's a much longer period of an off cycle to allow a longer recovery period. I think everyone is trying to get this dance right. It's hard to predict before seeing the data in the clinic. I think the Daiichi data did a great deal to give us confidence that there's now a significant number of patients who have experience with this specific dosing schedule, where we have confidence in what the safety profile is gonna look like.
Okay. Just kind of tracking back to MANTRA. I think you said three months assumption for control. Just, you had a choice, I assume, between Yondelis and HALAVEN. Just explain the reasons for your choice of trabectedin. Also just if you could just go through what was the powering on Mila for what you're assuming for the active, and then remind everyone, given, you know, given the change to have the data in the Q2, so everyone's clear on that.
Sure. Sure. Let me start with the statistical assumptions first. It's powered to show a doubling of PFS benefit, so a hazard ratio of 0.5 at 94%. The assumptions are three months for trabectedin, six months for milademetan. A very strong powering. The reason, going back to your first very question, why do we pick trabectedin? We believe trabectedin is preferred versus HALAVEN across the U.S. and Europe. From our conversations suggest that it is the preferred agent. Although both agents have roughly comparable median progression-free survival, there's more physician experience with trabectedin, right? It was the earlier drug approved. Ultimately it came down to what we thought physicians use most commonly today, despite roughly equivalent data.
Truth be told, there was a point that because of the very modest PFS, there was a discussion in the early days about actually having a placebo-controlled study, which we ultimately decided not to do because of challenges with enrollment. The two months, again, for either trabectedin or Halaven is roughly comparable to placebo. We went with trabectedin because it was a more commonly used agent with greater physician experience.
Yeah.
I'm sorry, Yigal, your third question was?
Just to clarify the timing. I mean, you pushed it slightly.
Yeah. As per the trial design of the MANTRA study, we are waiting for 105 progression events. We enrolled our last patient in July of 2022. Our internal modeling suggested that we would get to around 105 patients by end of 2022. By end of 2022, December, we did not hit the 105 events. When we looked again on January 30th, three weeks ago now, we still did not achieve the number of events. When we account for the timing to clean and validate the data, we didn't feel that we were gonna achieve a 1-2 readout timeline. That was the reason for pushing it off. We have, you know, we believe 2Q is far more likely.
Okay. Just going back to my earlier question, when you in-licensed from Daiichi, I mean, was there a big debate internally at Rain about whether you would even, you know, do the liposarcoma phase III or that was kind of a given, and then, you know, you would expand later into... We'll talk about it later in the hour about MDM2 amp and CDKN2A.
Yeah. Well, I think, you know, liposarcoma presented a very rapid path to market because they had already demonstrated the data. They already demonstrated the safety profile in liposarcoma patients. We already had a good amount of clinical data. We thought getting to market quickly is gonna help this molecule in terms of establishing a foothold in MDM2 inhibition, to be a thought leader for this mechanism, broadly speaking, get physician experience. As you know, we didn't really waste any time in starting trial number two, which was the MDM2 amp MANTRA-2 study. When we looked at the MANTRA-4 CDKN2A loss study, what we were waiting for was actually confirmation that from other companies that we were not gonna see any clinically meaningful overlapping tops between our drug and checkpoint inhibitors.
As a reminder, MANTRA-4 is our first combination study, milademetan, and Roche's atezolizumab. We think checkpoint inhibition should be synergistic with MDM2 inhibition. We know that a competitor presented data at ASCO 2020 that looked at an MDM2 inhibitor plus pembrolizumab with no overlapping toxicity, and both drugs could be dosed effectively at full dose. When we saw that data, it gave us confidence to pursue this strategy. Of course, it's for a sufficiently large patient population, 45,000 patients in the U.S. Go back to your point, Yigal, we didn't wanna terminate liposarcoma because it was the lowest hanging fruit out there in our view. Again, we think first to market is a very powerful advantage to have.
Okay. No, that totally makes sense. That kinda segues into some of the next questions regarding the competitive landscape and the competitive positioning. I mean, one of the key questions we get obviously is with respect to, you know, how you stand relative to the BI molecule, which is obviously not nearly as advanced as you guys. I believe they're doing a first line trial or that's their initially stated plan. You know, you've made the point in the past that you would be able to take share in the first line following a positive MANTRA trial.
Just that would be very helpful to understand the thinking there and also, you know, to the extent you can comment around the competitive positioning versus the BI MDM2.
Yeah. Let me start with the competitive positioning first, and we'll get back to the frontline of the conversation.
Okay.
Here's the competitive positioning with BI. We don't necessarily know yet because they haven't presented an immense amount of data. We started the MANTRA study in liposarcoma. They started a study called Brightline-1 in liposarcoma. We started MANTRA-2 in MDM2amp, and then they started Brightline-2 in MDM2amp. There is a little bit of a, you know, a quick to follow strategy from BI here. They are a very potent molecule. They're, you know, molecularly they are a more potent molecule than milademetan. As a result of that, their dosing schedule focuses on a much longer period of recovery time versus milademetan. We're out three days on, 11 days off. They're on a dose one day, 20 days off. Dosing once out of a 21-day period, once every three weeks.
Again, all molecules in this space are trying to find the right way to do the dance. Is their schedule going to ultimately lead to meaningful efficacy with a superior tolerability? We don't know yet. Early data from their study suggests greater levels of thrombocytopenia, neutropenia, and anemia versus the prior phase I data for milademetan. Early data. We'll see how that evolves over time. In liposarcoma, early data suggested across all doses that there was comparable efficacy in liposarcoma, which is good to see that there's multiple levels of validation of MDM2 inhibition in this tumor type.
They presented some early data on MDM2 amplified patients, which suggests that to us that they're seeing some activity specifically in biliary and pancreatic patients, which is where they have publicly announced an expansion, which is all good to see. I think more than anything else, I think the strategy from Boehringer legitimizes MDM2 inhibition, broadly speaking, across a multitude of cancers. Competition's a good thing. Competition's a good thing to increase awareness across the board. Now, in terms of the frontline market, the inclusion criteria for the MANTRA study states that patients need to be on at least one prior therapy, which includes a prior anthracycline. If a patient's on a prior anthracycline, and they progress, they could receive our drug as a second therapy.
That first therapy could be in the, in the adjuvant or the neoadjuvant setting. We are aware that in the liposarcoma treatment paradigm, giving frontline anthracyclines, docetaxel or doxorubicin around the time of surgery is common practice. If a patient were to receive this as adjuvant or neoadjuvant care, they could receive milademetan as the first-line metastatic therapy. Upon progression after surgery, they could receive milademetan as the first metastatic treatment that they experience. Take that into perspective, along with the fact that this is a small market and a meaningful outcome in the MANTRA study, we think will encourage adoption across the treatment landscape. We don't think that with a 1,400 patient population in the U.S., there's gonna be a lot of debate.
We think that the first market drug is likely to take it all, as long as there's a reasonable PFS from the pivotal study. That's our view.
Okay. It's not. Just so everyone's clear, It's not a second-line metastatic trial. It's a second therapy trial. How do you say it? I mean.
Correct. That's the confusion that we saw out there.
Yes.
It is a, the prior therapy-
Right.
Amp for cyclins does not need to be given in a metastatic setting.
Right. Okay. Very helpful. Now the other question which I think would be great to answer is, you know, obviously, you're doing the phase II liposarcoma, but you have bigger plans, which you've articulated already, in larger populations. You know, if for any reason MANTRA doesn't work, I mean, I think it has a very good shot. We've done the modeling. You've seen our modeling. Nonetheless, you know, things happen. If for whatever reason it didn't work, you know, how would you approach the question as to why that really shouldn't be a negative for looking at these other markets like MDM2amp or CDKN2A? Why would you still be optimistic about those?
For the CDKN2A, I mean, again, that is our first combination study, we are expecting synergy with the other agents. That could provide a different level of outcome. But in all reality, and I'll be very transparent here that if MANTRA was not successful, if we did not show a meaningful improvement in PFS versus trabectedin in an MDM2amp population, at the same time, we have MANTRA-2 ongoing. We have visibility in terms of whether or not we're gonna see activity there at the same time. I think the question you're asking is, you know, with the data on hand, at the time that MANTRA reads out, you know, is there a...
If MANTRA doesn't work and MANTRA-2 doesn't show activity, we'd be very concerned, right? We would be very concerned if that was in fact the actual outcome because those are MDM2 amplified, MDM2 de-dependent tumor types. If that was the case as a monotherapy, that would call into question whether or not MDM2 ambition is legitimate as a monotherapy.
Mm-hmm.
CDKN2A-LOS, again, our third trial is a combination strategy which still holds a different opportunity to see a different outcome. I think that's how to answer the question, Yigal, in terms of whether or not there's gonna be a read-through for another. We do expect a read-through that if a tumor type is MDM2 dependent, whether it's because it's MDM2 gene amplified or MDM2 protein overexpressed, we do expect that that does signal that MDM2 inhibition should work to different degrees, right? One tumor type may be more sensitive than another tumor type. We do expect a read-through. We think the biology focusing on where MDM2 is playing a central role should signal where an MDM2 p53 reactivation inhibitor can be effective.
Okay. Just one more specific one on MANTRA, I think people are asking me is just, can you just quickly run through in terms of baseline demographics for MANTRA versus the prior phase I that Daiichi did, how similar are we talking here in terms of the enrollment?
We don't have visibility in terms of the patient baseline demographics yet. We have no visibility into the MANTRA study other than the quarterly data safety monitoring board that have always recommended to pursue the study. We have no insight into how these patients look. We can say that we designed the study to mimic the prior phase I study as best we can. Every attribute of the MANTRA study design was meant to reproduce exactly the phase I study. There's been no change with the exception of the prior therapy. Daiichi did allow treatment-naive patients on. We are not allowing treatment-naive patients on. That's the only change. In terms of patient baseline demographics, I can't comment on that yet as I have no visibility.
Okay. I think also this is a good opportunity for me to give you a chance to address one of the criticisms of the phase I is that there were about a third that were treatment naive, and I guess some people were saying that that may have influenced things. You've done that analysis and you have a good counterpoint to that. Maybe can you just walk through that?
Yes. Because the Daiichi trial included some treatment-naive patients, the question we received was maybe that was carrying some of the activity. Now that was the Daiichi dataset as presented by Daiichi Sankyo. We did not have the ability to go public with different cuts until this recent publication. Because of where investor questions were focused on teasing out specific sub-subsets of patients, we published the subset analysis in the Journal of Clinical Oncology several weeks ago, which now teases out how treatment-naive patients and treatment-relapsed patients did at various levels across the study. What the subset analysis showed is that, first of all, across all 53 patients, regardless of dose, regardless of schedule, there was a median progression-free survival of 7.2 months across all patients.
That's part of it. When you tease out the treatment-naive patients, the treatment-naive patients did do better. Treatment-naive patients did 14.6 months median PFS, and the relapsed patients just under six months. Yes. That's across all doses, all schedules. When we look specifically at the dose and schedule that we're most concerned about 260 mg, three days on, 11 days off. The treatment experienced and the treatment-naive patients were about the same. The relapsed patients had an eight-month median PFS. Treatment-naive patients, about 7.4 months, about the same. In that, in that 7.4-month median PFS that we always talk about, treatment-experienced and treatment-relapsed patients did about the same. There's only five patients that were treatment naive, pretty small number.
We would expect treatment-naive patients to do better, as evidenced from the 14.6 months across the broader study of all treatment-naive patients across all doses and schedules. I think now we have definitive data to suggest that treatment-naive patients did not carry the Schedule D optimized dose and schedule cohort.
Okay. Got it. That makes sense. All right, let's move on, just in the last remaining few minutes here. Let's talk about some of the other larger opportunities. First of all, for those less familiar, can you just kind of give us the nomenclature, you know, how you're defining the MDM2 amplified basket, how much amplification? Is there a routine genetic testing for this? Is there a tumor diagnostic? Just walk us through all the basics. It's not a market that maybe people are super familiar with.
Yeah. We're looking at MDM2 copy number as a selection criteria to define MDM2 gene amplification for patients across all solid tumors, advanced solid tumors. This is our first basket study, the MANTRA-2 study. We're using next-generation sequencing to identify copy number. MDM2 is on all of the standard commercial panels, so this is not a commercially difficult thing to identify. We are using a copy number threshold of eight. We started initially at a copy number 12, and we dialed it down to eight after seeing some patients with activity that were below 12 but above eight. There was reason to believe that we didn't need as stringent of a copy number as 12.
The rationale for starting high in the beginning was to make sure that we see something, because if we had not seen any activity at a copy number 12, then this may not have been a legitimate strategy to pursue. The way that we were encouraged by some early signs of activity, based on the interim data that we released last November. Copy number eight copies or greater, is the criteria based upon local testing at the sites for enrollment. All of those samples are tested independently at a central third party at Tempus. Tempus is our partner for the MANTRA-2 study in anticipation, in the hope that if we see good activity across tumor type, we're gonna need a third-party central confirmation for potential expansion and registration.
That's why Tempus has been there from the very start of the trial in hopes that we see good activity across tumor type. That's the MANTRA-2 study. We have not provided guidance on subsequent disclosure of data. What we have said, and I'm looking at time here, Yigal, tell me if we need to wrap up. What we have said is that we will disclose the data when it tells us what to do next, whether it is a front door tumor-agnostic strategy, whether it's door B tumor-specific expansion strategy, or whether it's door C that we need to look at a combination therapy because in the event that we're not seeing good activity as a monotherapy. There's a couple different ways to win with the MANTRA-2 strategy.
We think it's an exciting new strategy to select patients.
Okay. At this point, you're not ready to provide a timeline for data for that one for 2023?
Correct. Not yet.
Can you comment at least broadly in terms of, you know, the breadth of not necessarily activity since you can't comment on that yet, but just in terms of MDM2 amplification across solid tumors. Are there certain tumors where it's more common or is it really across everything?
Lung, breast and bladder comprise the vast majority of where we see this amplification. When we look at patients with eight copies of MDM2 or greater, it comprises about 8,000 patients per year in the U.S. annually. Lung, breast, and bladder comprise 6,000 of that 8,000.
Okay. How does it work with the biology, if you know? I mean, apologies, this might be a very technical question, but, you know, once they're diagnosed, is it just sort of a fixed copy number or as their disease gets worse they start generating more copies of the MDM2? Is that...
That's a great question. Whether or not we start seeing a modulation or change in copy number with treatment line. That's a great question. We don't know yet. We don't know. We are looking at patients who are advanced patients who have exhausted available commercial therapies for their specific tumor type and broad variability there. So by the time that we see that MDM2 amp is quantified, we're seeing, you know, eight or above, presumably, at that specific stage. We don't know what that tumor looked like when it was newly diagnosed.
Okay. I know you can't talk about data timelines, but just in terms of the structure of the study, the basket study, what can you say about just the basic blocking and tackling of that?
Yeah. Yeah. 11 sites in the U.S., major academic centers today. It is all at the identified dose and schedule, so all of our studies are at the 260 mg dose. There is no more dose finding going on, so this is not a dose-finding study. It's targeted for a 65 patient target accrual rate. Response rate is the primary endpoint.
Do you have some sort of, you know, sort of stratification, so you wanna get a little bit from each of the tumor types you mentioned? Is it just you'll take the first people in the door, and then how does it...
Yeah, there is a cap on tumor types so that we don't over-enroll any one specific tumor type, so absolutely. We haven't identified what that cap is. But it is aimed to be broadly supportive of a tumor-agnostic label, which was the initial strategy. I think based upon recent FDA guidance, as you know all too well, the FDA wants to see that you have broad representation across multiple tumor types. I think you'll recall, Yigal, when we presented data from the first 10 patients, every patient was from a different tumor type. I think that's very much in line with how we've been enrolling the study.
I guess just one last question on the science, not necessarily from the clinical data, but from your preclinical work or earlier studies. Is there any reason to believe that the higher the copy number, the more likely Mila would be effective, or is it not like that? Or is it more like a, you know, like a threshold effect, where it's as long as it's above X, then you're in a good position for showing efficacy?
Yeah, we definitely think it's a threshold effect. We've said publicly at the time of the interim data release that at that time, we did not see any meaningful correlation between copy number above a certain level. I think as once you cross that copy number eight, maybe to copy number 10, whatever that level is, around that mark, that tumors, we think, will become more amenable to therapy. We don't believe necessarily that a copy number 20 patient does worse than a copy number 30 patient, for example.
When you drop the cutoff from 12 to eight, how big, how much bigger did the opportunity get, just out of curiosity?
Yeah. The opportunity grew by about 30%.
Okay. Okay, very good. All right, in the last few minutes, let's hit MANTRA-4. This is obviously the largest opportunity by, I guess, maybe an order of magnitude or close to that. Talk about the strategy for the combo, and I think you're doing a dose de-escalation versus a usual dose escalation. Just can you walk us through the logic of that?
This is a phase I/2, initially set to identify the safety profile of the combination regimen, 260 mg of milademetan on the Schedule D that we've identified, at the labeled dose of atezolizumab, about 1,680 milligrams IV every four weeks. Starting at full dose of both drugs, if there is a dose-limiting toxicity, we will step down the dose of milademetan from 260 down to 200. If we see another reason, the dose reduced down to 160. I think we've seen from other, as we, I think, alluded to, other companies that have presented combination studies of MDM2 inhibitors and checkpoint inhibitors, that there have been no reasons to suspect overlapping toxicities.
If in the first three patients we don't see any dose or any toxicity at the full dose, we will begin the expansion with the additional 27 patients. This study effectively becomes a 30-patient signal finding study at one dose of milademetan and the combination agent. Response rate well is certainly the focus here. If we can check the box on safety, we hope to see responses to 45,000 patient opportunity in the U.S. That is meaningfully larger than everything else we're looking for. This is a patient population that had previously been treated with and relapsed from a checkpoint inhibitor, so that we get some degree of comfort that any activity that's observed will be from milademetan or the synergy from the two and not derived purely from the combination agent.
It's again set to enroll 30 patients across the U.S. We said we're gonna start this trial in the middle of the year.
Okay, just remind everyone, what's the agreement with Roche? I mean, is this one of these, you know, basically clinical drug supply agreements, or is there more in terms of potential economic partnership here?
This specific trial is for a supply agreement only. It's to confirm the safety. We like to think, and we remain hopeful that if there's good activity here, that the collaboration opportunities can take multiple shapes and forms.
Okay. Are you able to comment yet on timelines for this one for data, or is it too early?
It's too early. We'll start the study first and get a read on enrollment.
Okay. beyond those three programs, MANTRA-2, MANTRA-4, what can you say about other BD efforts, or do you have any internal discovery ways to move beyond mila?
Yeah. Beyond these three indications, I think if MANTRA is supportive, there is a slew of additional indications that we're evaluating internally preclinically as to validate the opportunity. Much of that analysis is being done by competitors, we think 2023 is gonna be a busy year for this class of drugs, not just in the solid tumor space, but also the heme malignancy space. We'll see how some new agents look there. That will certainly affect our thinking about where appropriate invested capital should be deployed. There's always BD opportunities. Of course, we can't talk about those today. We want to make sure that we get past a de-risking event with MANTRA data first. Of course, just continue to work on RAD52. We're excited about RAD52 as a preclinical program.
We slow-rolled the RAD52 research program during 2022 to focus all resources on the clinical strategy for milademetan. There's nothing new to share with the RAD52 program just yet, we would expect to build a pipeline through both internal and external BD efforts to supplement milademetan to diversify the portfolio.
Okay. Just in the last minute or two here, just really quick, just tell us the what's the financial status in terms of your runway and ability to fund the company?
Pro forma for the Q3 of 2022, we haven't reported our Q4r results yet. Post our November financing at $147 million. We believe that gives us a runway well into 2025 to run and complete all these three studies. It is likely not sufficient, however, for a commercial effort in the U.S. around a liposarcoma detailing effort. For all the clinical development work for milademetan, we have a lot of comfort that that takes us through completion of all three trials.
Just one last one just on a, I guess, strategy. You just decided to make a slight change to the name of the company. Just, it makes sense, but just maybe just remind everyone why you did that.
We think it'll be, it better talks to the strategy for the future, not just for milademetan, but the overall priority for the company. We are a cancer company. We are a precision cancer company. We wanted to be very specific as we start expanding the pipeline, about what the focus for the company is gonna be and what it remains.
Perfect. All right. Well, that was great. Thank you so much, and we will be chatting, I'm sure, very soon. Thanks, Avanish.
Thank you, y'all. Appreciate the time.
Sure. Bye.