Okay. Good afternoon, and welcome to Barclays Global Healthcare Conference. Do email me or my team if you have questions. My name is Peter Lawson, and I'm one of the mid-cap biotech analysts at Barclays, cover and focus around oncology companies. Really delighted to have with us on stage, Troy Wilson, CEO of Kura Oncology. I guess with that, I'd love to kind of touch upon definitely the menin inhibitor.
Sure.
Kind of how we should expect kind of data to unfold. You know, how much more data should we expect to see? I guess that's maybe earmarked for EHA.
Yep.
Yeah, EHA.
Yeah. First of all, Peter, thank you to you and Barclays for the invitation. In terms of the update, we've guided toward midyear, obviously you have ASCO and EHA. It'll be one of the two. EHA would be a logical venue, just so you get exposure to European HemOnc physicians, right, who may not have made it over for ASH. In terms of the data that you'll see, the focus is gonna be on the phase 1b cohort in NPM1- mutated AML, primarily with a focus on durability. If you remember, at ASH, we had 20 patients, 30% CR rate with full count recovery, the duration of response at that time was still immature. It was about four to four and a half months.
That was as of an October 24th data cut, October 24th, 2022. If we fast-forward to midyear, you know, you're probably looking at an additional six months. That'll give both a perspective on the potential durability, which I think was a question mark. We'll give an update of all the NPM1- mutated patients on study, it'll also be more of an apples to apples comparison with the competition.
Gotcha. Thank you. I mean, so where do you think that bar is for durability to be kind of best in class and build a sufficient barrier of entry?
Yeah. I don't know that there's an absolute number. Our competitor at ASH showed a duration, a median duration of response of nine point one months, that was a composite number. Don't think they broke it down, I can't remember, by genotype. You know, that certainly at that time at ASH, when our duration of response was still immature, we said we thought that was a number that we could probably meet or exceed. I, you know, I think anything you can do, if you think about the overall survival for NPM1 mutant patients in the relapsed/refractory setting's about six months. If you have a DOR that's, you know, approximately, you know, six months, nine months beyond, that's clinically meaningful.
The longer the better. I don't think it's necessarily a winner take all, but, you know, we'll see. We'll see when we present the data in mid-year.
Good.
What it looks like.
I guess questions that we've been getting recently inbound are just like the idea of resistance mutations developing.
Yep.
Kind of how much does that hinder the use of the drug?
Yep.
Can you know, repeat dosing with a different drug?
Yep.
Et cetera?
We've not seen a lot of resistance mutations. We're still analyzing the data. You know, I'm aware of a kind of a handful of cases, and of that handful, a couple were patients who presented with a resistance mutation after having been on another drug, another menin inhibitor. I do think, you know, if AML is like every other cancer indication, I think the resistance arises from, you know, from basically an inability to clear extramedullary disease. It makes sense, right? If you're keeping constant drug pressure and you still have tumor cells around, those cells are going to find a way to become resistant.
Our thesis for ziftomenib is you get, because of the tissue penetrance, because of the very good tolerability, we've said consistently, we see clearance of extramedullary disease in both the KMT2A and the NPM1. The penalty in KMT2A is differentiation syndrome, which I think we'll come to. If you clear that disease, then there's no, there's no tumor to develop resistance mutations. Maybe it's early days, but it's not something, you know, it is possible we do see it. I wouldn't yet describe it as a major problem for ziftomenib. I think that might be a point of some differentiation from the competition, but it's early days. We're talking about, you know, phase I datasets.
Gotcha. Do you think a covalent menin inhibitors can kind of help alleviate that?
Um.
Does that become a problem?
I don't because one of the things that ziftomenib does effectively, and this is in contrast to one of the non-covalent competitors, is it degrades menin. It inhibits menin and it degrades it. I don't think we know enough yet, Peter, as to what the covalent menin inhibitor does in patients. You know, it's less potent in vitro, but who knows? We're, you know, I think we're expecting to see some data a little bit later this year. I'm interested to see.
Yeah. No, exactly.
As you know, we investigated both non-covalent and covalent. That wasn't a point of differentiation.
Gotcha. Okay. Do you get responses on patients that have a mutation that's been acquired from a different inhibitor?
We didn't see responses. You see blast count reduction, you see symptomatic improvement. It hasn't been a large number. It's been, you know. I can think of two specifically. We weren't able to drive them to a full response. No. Ziftomenib is, if a patient has a resistance mutation that does impair the binding of ziftomenib. It actually impairs the binding of all. We've, you know, we've tested what we believe are all the menin inhibitors across all the various sponsors, and they all seem to be blunted by the mutations that are arising at the interface that have been described.
Gotcha. Thank you. As we think about these MLL rearrangements.
Mm-hmm.
Kind of how many are there? Does that change the dynamic of if you can inhibit all those rearrangements or if it's a subset of those rearrangements?
I don't think we know enough yet. I don't think we have enough clinical data to know. It's, you know, it's always, you're always reluctant to talk about all or none, right? In, in oncology, it's all is always a CR. I think the best course of action for the MLL rearrangements will be combination, right? Bringing a couple of different drug pressures, maybe it's chemotherapy plus a menin inhibitor, maybe it's, you know, maybe it's something like venetoclax plus a menin inhibitor. That's how you're gonna drive your deepest and best response. It's what we know. What we know now is menin inhibitors are showing promising clinical activity. There's, there's a path forward for one of our competitors as in registration. Ultimately, you need to be in combination.
That's everybody sort of acknowledges that if you want to provide the best benefit for patients.
Gotcha. As we think about differentiation syndrome.
Mm-hmm.
Do you think that in any way correlates with durability or response, or is there a, is that a good secondary read to do anything with that?
I think it correlates with clinical activity. What we've shown, what we showed at ASH is, you can see that you're engaging the tumor, you can see blast count reductions. The reason you're not taking patients on is because they're not staying on therapy due to differentiation syndrome. As we've said on our last earnings call, our plan is to take those patients in combination with standard of care. I think, hopefully by the end of this year, we'll show we can combine safely and tolerably, and we can mitigate differentiation syndrome in the KMT2A population. In the NPM1 population, it's not an issue. Our goal is by the end of this year to have data out that says combination effectively solves that problem.
Gotcha. Having a covalent version, do you think that kind of amplifies DS, or do you think it's, you know, there's a different mechanism or action going on there that perhaps reduce DS?
Um.
Just curious on your read on the read.
I don't know. Without human data.
Yeah.
you can't, you can't. It's very difficult to model DLTs preclinically. It's really something that, you know, mice don't complain of DLTs. It's something you really have to see in the clinic and we've not seen any clinical data from an irreversible inhibitor. I would be guessing at this point what they see. You know, other covalent inhibitors, you do see. I mean, you see diarrhea, right? You see GI toxicity. We've not seen any clinical data at all. Let's wait for that particular sponsor to release clinical data, and then we can all take a look at it.
Gotcha. I remember you saying in the past, you kind of passed on a covalent.
Yeah.
Inhibitor.
Yeah.
Why was that?
We passed on it because at the end of the day, menin turnover is about every six to eight hours. In our experience, you needed to dose repeatedly in order to drive activity. There was no advantage relative to a reversible inhibitor. As we've seen in the case of some of the KRAS inhibitors, having a potent electrophile on it can not only create GI toxicities, you know, such as diarrhea, but it can also hit off targets. For chronic therapy, it wasn't obvious to us that the irreversible or the covalent inhibitors had much more to offer. Certainly something you could pursue. The chemistry is very clear, but the reversible inhibitors are driving. I mean, we have a 30% full CR rate, right?
Syndax has a 27% CR/CRH rate in the KMT2A. I'll be interested to see if the covalent inhibitor can even match that. Let's see. Let's see what we see.
Gotcha. phase II.
Mm-hmm.
How's that enrolling, and kind of how should we think about data folding out from that?
It's enrolling well, not surprisingly. People will recall the phase 1b enrollment was very strong in NPM1. We enrolled, I believe it was 14 patients in three months in the 1b with about 20 clinical sites. you know, we're gonna pretty much double that site footprint. The sites are still in the process of coming online, but we've seen very strong interest, strong enrollment. When we announced that the study had started, we took pains to say we had dosed multiple patients, plural, and that has continued. Given that one of our competitors, probably our nearest competitor, has pushed its timelines out on NPM1 enrollments, I think, and given the strong pace of enrollment we've seen, maybe we're three months behind, three to six months.
I, you know, I think, we plan on taking advantage of every development and regulatory strategy to accelerate development, and I think we, you know, we have a strong chance of catching, and potentially, getting ahead of them in the NPM1 cohort.
Got you.
We'll see.
I guess that difference is coming up a couple of times about the 85 patients versus 60 odd.
Yep.
Kind of a what's the right number? Why was yours larger?
Ours was larger. Yeah. I mean, the right number is whatever the you know, however much risk the sponsor wants to take. Our 85 was driven by safety, not driven by activity. In our opinion, and this was not something the FDA imposed upon us, it was something we volunteered, you wanna have a sufficient safety database that you can support registration. Typically in the FDA's mind, it's about 100 patients overall at the, you know, at the recommended dose. We thought 85 was kinda the right number, and it was a number that we thought was eminently achievable given the enrollment we'd seen in the 1b.
We haven't been specific about the parameters around the, you know, the efficacy parameters or the statistical plan, but it's, we are going to take advantage of everything we can to go quickly.
The kind of the driven by safety, is that because there's something fundamentally in the data that we should be thinking about?
No, no, no. It's, it's not that we've seen something. It's more that, you know, you wanna have enough patients that you can say conclusively you've ruled out the incidence of any kind of safety event. In our experience, you know, 85 is the right number. You can always, if you set it higher, you can always go faster, right? The risk that you run is if you set it too low and you have to expand it, then that sends a very strong signal. You know, we've tried to be a bit more conservative as a company, not take, you know, unnecessary risk. At that point in enrollment, Peter, there's not a huge difference between 64 and 85 patients.
If you're enrolling, again, go back to the 1b experience, 14 patients in three months, you're talking about a month or two, in enrollment once all the sites are up and running. There's no, there's no real material difference in time, but it's just a more substantive package, and it's one that, you know, a we think is the right one to submit to the agency.
Gotcha. The potential, I know you previously spoke about the potential enroll, including those phase I patients into the phase II.
Yep.
Kind of what was the disconnect or how did that conversation evolve?
We didn't take that to the agency. We basically left it as a review issue. We didn't ask the agency to opine on that, in part because it was in the context of Project Optimus, and we had enough going on with Project Optimus. Clearly, you know, if you take a 30% CR rate with a 20-patient denominator, you can do, you know, just, the statistical analysis and figure out, you know, our trial has a 95% probability of success, you know, assuming those parameters keep up. We didn't feel at that time in the, in the type C meeting with the FDA that we needed to introduce the variable of asking them to include phase 1b patients.
Because in all likelihood, it would have slowed everything down. We wanted to get that answer so we could get on to enrolling the phase II. You know, dealing with the FDA is as much art as it is science. We just made a judgment call. We didn't, however, foreclose the opportunity of including them. I just don't think we're gonna need them because enrollment is so robust that, it just adds more to the body of data. The other thing to keep in mind is our activity continue, you know, through the phase 1b, you know, continued to get better and better, and I think we're getting better in both the NPM1 and KMT2A of knowing how to manage DS. If you have a pristine phase II data set, that's gonna be reflected in your label.
folks won't be able to come back and say, "Well, what about differentiation syndrome? What about this? What about that?" It is art, a little bit of art, a little bit of science.
Okay. There's no You wouldn't at the end, towards the end think, "Okay, we can now include these patients to phase I." Or you wouldn't know?
You will include them as part of the filing. The question is, can you include them in the primary efficacy analysis? You don't have to make that decision today.
Gotcha.
There are statistical tools you can use if it allows you to go faster. If, for example, you're able to take advantage of something like breakthrough therapy designation, there are ways you can go more quickly with the agency. We've been, you know, pretty guarded about exactly we're doing. Our goal is to go as fast as possible. We think we've got the best-in-class agent in NPM1, and we are pulling out all the stops to get there as quickly as we can.
Just thinking about that, how are sites going for enrollment and up and running, et cetera?
Yep.
Where are you for Europe as well?
Sorry.
For enrollment of new clinical sites.
The U.S. sites come up more quickly. The European sites typically take, you know, an additional 3 to 6 months because they have to go through country review and then ethics review. The U.S. sites will have them all, I think, up and running certainly in the first quarter. That's gone extremely well. As I said, the intent is to kinda double the footprint from what we had in the phase 1b. Approximately 40 sites globally. Interestingly, we are also in study startup for the first combination study, and we have between the KOMET-007 protocol, which is seven plus three in venetoclax, and the KOMET-008, which is primarily gilteritinib, we have a waiting list of sites.
I mean, we have more sites that want to participate than we probably can take on at this point. Very, very strong interest. That's where you're gonna see us solve the problem with KMT2A as in combination. We have sites lined up ready to go.
Gotcha. Thank you. Has the thinking evolved around the rearrangement and the differentiation syndrome? Has, you know, physician feedback helped or kind of how you think that happens where you get a delta in responses?
Well, yeah, I mean, we know why it happens. It happens because the patients aren't able to stay on therapy because they are seeing symptoms of differentiation syndrome. Everything we've seen from, you know, experience with prior drugs, for example, IDH inhibitors, when you combine an IDH inhibitor with azacitidine, you cut the rate of differentiation syndrome in half. That's just, you know, that's fairly simplistic. We think we'll see the same, if not better here. All the docs tell us as well, it's a logical thing to do. It's what they want to do to be able to drive deeper and more durable responses.
We've just sort of decoupled the KMT2A piece for now. We'll come back, I think, in the second half of the year and give people an update on can we combine safely, we expect we can, and does it mitigate DS. If it does, I think it'll support the thesis that we have, you know, the most potent menin inhibitor out there.
Gotcha. Has that idea resonated well with physicians?
Yes.
Has there been any worry that it's working well in the NPM1, but, you know, we've got differentiation syndrome, so does that cause any holdback in any way?
No. I think they want to know how to. You need to make management of differentiation syndrome routine. While we could spend more time on monotherapy, we've said, "You know what? This is a world of combinations. Just combinations will mitigate the problem, both by tumor debulking and by clearing white cells and blasts. This is the way to go." The KOLs, the physicians have been, you know, unanimously supportive. I'll just remind you, I mean, people forget this. KMT2A is 5%, NPM1's 30%. You know, our focus is on the 30% of the population that's, in the near term, gonna drive the greatest value. We'll come back and by the end of this year, I suspect, and have more to say on the KMT2A opportunity.
As soon as anybody's in combination, everybody's gonna be in combination because AML is a disease of polypharmacy, right? It wasn't venetoclax, it was venetoclax plus azacitidine. It wasn't seven, it was seven plus three. Combinations are the name of the game. You have to do them to drive durable, deep responses, whether you're considering transplant or not.
Okay. Before we transition over to tipifarnib.
Sure.
I have to ask just about Silicon Valley Bank kind of exposure, not only to Silicon Valley, but also I guess regional banks and non-bank lenders, kind of how you kind of think about that.
We put out an 8-K. We don't have any material exposure to SVB. Now with the Fed's action, you know, all depositors are gonna get money back. Even our immaterial money, you know, we know is not at risk. I think all companies will consider their, you know, can they either through diversification or insurance protect against something like this happening? I think that's just gonna be part of what we do. We don't have any exposure that I can think of to, you know, regional banks, non-bank lenders or anything like that. Other than all the noise, you know, it isn't affecting us in any way.
Gotcha. Gotcha. Then just tipifarnib.
Yep.
kind of what we should expect to see, I guess the combination strategy. We hopefully see something mid-year and what we should look at.
Yeah. Well, yeah. What you're referring to is the combination of tipifarnib plus alpelisib.
Yes.
In PIK3CA mutant head and neck. We are working toward an optimal biologically active dose. To my knowledge, we haven't yet seen any DLTs, so we continue to dose escalate, which is interesting. The we have to reach that OBAD, optimal biologically active dose, and then probably, Peter, we have to do a sufficiently sized expansion cohort to be able to say, "Okay, what's the effect size and the path forward?" I will just, you know, use this as an opportunity to introduce, we now have the IND open for KO-2806. We expect to dose the first patients in Q3. You'll see a press release, a curtain raiser on the next two opportunities that will actually be with KO-2806. I think I can tell you they are KRAS mutants and tyrosine kinase inhibitors in renal cell carcinoma.
Both of those we are going after a uniquely farnesylated protein that's driving resistance either to KRAS inhibitors or to TKIs. You'll see more about that science at AACR here in I think April.
Thank you. That's kind of post the G12C kind of development.
It's not post. Taking just a quick step back, when we first started with farnesyltransferase, we thought, "We need to go after a farnesylated oncoprotein." Well, there's only one, HRAS. What we've learned subsequently is when you hit a cell with a highly potent drug like a KRAS inhibitor, like a TKI, like a PI3K inhibitor, EGFR inhibitor, you induce certain proteins that are farnesylated. Those proteins drive resistance. In the case of KRAS, when you hit, and this is well documented, you treat with a KRAS inhibitor, the PI3 K pathway, PI3K mTOR AKT gets activated. mTOR is under the control of a protein called RHEB. RHEB is farnesylated. By blocking RHEB, you can blunt the resistance to KRAS inhibitors, drive deeper and more durable responses. Same thing in RCC.
When you hit with a TKI, you induce VHL. VHL is under the control of, surprise, RHEB. RHEB is farnesylated. It's our understanding now of how to use farnesyltransferase inhibitors in cancer is don't use them as the primary hammer. Use them to address the resistance that arises when you hit them with these well-understood commercial stage drugs. With both KRAS and RCC, we're optimistic 'cause there's a quicker path to data, potentially a quicker path to market relative to some of the other opportunities we've looked at.
Okay.
More to come.
Yeah. You'll be combining it with a MEK inhibitor.
You'll be combining. Exactly. That's the key.
also, sorry, a MEK inhibitor as well.
Everything's a combination.
Anything that's got resistance.
Yes, everything's a combination.
Perfect. With that, thank you so much, Troy.
My pleasure. Thank you, Peter. Thanks.