Everyone, I am Imogen Mansfield. I'm a Biotech Analyst at Cantor, and I am delighted to be joined today by two legends, the CEO of Eikon, Roger Perlmutter, and Dr. Tim Yap, who is a legendary clinical trialist and physician at MD Anderson . Welcome. Thank you for joining to discuss PARP1 with us. We will be going through PARP1 biology, discussing the unmet needs left by the first generation of PARP inhibitors, and then going through Eikon's programs and spending some time at the end talking about Werner helicase. To kick us off, Dr. Yap, would you like to tell us a little bit about your background and the trials that you're involved with for the PARP1 field, and I guess synthetic lethality more broadly.
Just to confirm, we've just discussed this offline, but Dr. Yap is involved in several programs with Eikon and will not be sharing any non-public information today.
Thanks so much, Imogen. Good morning, everyone, and really thank Imogen for the kind invitation. Great honor to be here with Roger. My name's Tim Yap. I'm at MD Anderson. Been here for 10 years. Before that, I was at The Institute of Cancer Research and The Royal Marsden Hospital for about 12 years, where I've had a big interest in synthetic lethality, been involved with the development of many first-generation PARP inhibitors, and then, of course, now the new generation PARP1 selective agents and also many other DDR agents. Also, I guess, Werner helicase that we'll be covering as well. Wonderful to be here, and thanks so much for the invitation.
Great. Thank you, Dr. Yap. Roger, I'm sure that many of our listeners have met you in the past, but it would be great for you to share your background as well and give us a quick overview, a two-minute overview of Eikon.
Thank you, Imogen. Thanks very much for making this possible. It's a great privilege to be here with Tim Yap. He's modest, but he is a towering figure in clinical research, and it's a privilege to be here with him. I'm a physician scientist. I worked for many years studying signal transduction in hematopoietic populations when I was a professor at the University of Washington in Seattle. Moved to Merck to try to apply some of those lessons to drug discovery. Had a brief, well, 12 years at Amgen, taking a small, little company called Amgen and turning it into a great big giant company called Amgen, which I then retired from. Had no intention of doing that kind of job again, but was recruited back to Merck to lead research and development there again, and did that for a number of years.
My focus has been quite broad in the treatment of infectious disease and malignancy and also metabolic disease. But I have focused quite a lot on methods to improve cancer treatment, and in particular, the introduction of checkpoint inhibitors. We did that at Merck. When I say we, I mean everybody at Merck. I was just there to applaud. But we did that in a quite robust way, and I think really it transformed cancer care. That led me to think about how to proceed still further in this area, and it was at that point that I began working with the company that became Eikon. The premise of Eikon is that we can use super-resolution microscopy to visualize individual proteins in living cells. In doing so, we can develop novel methods for screening for active pharmaceutical ingredients that would alter the behavior of proteins.
It turns out that all drugs change protein motion. In fact, that's largely how they work. That wasn't something that we knew about before we got started with this. It took quite a long time for us to build the instruments necessary to apply super-resolution microscopy fruitfully to drug discovery. But we have now, for some years, reached that point and have been able to use our platform to introduce what we think are going to be important new medicines into the therapeutic armamentarium. We'll talk about one of those, EIK1005, which is a Werner helicase inhibitor, perhaps towards the end of the period today. But in the meantime, we also were able to use our tools fruitfully to look at some areas in DNA damage repair, particularly, because that's an especially good area for us to apply our tools to.
We were able to identify some very, very selective PARP1 inhibitors. One of the things that my colleagues and I had done when we were at Merck is we had worked together with the group at AstraZeneca to develop olaparib, which is the leading dual PARP inhibitor. Tim knows that drug very well and has used that very effectively. We felt that it was possible to find drugs that could have similar therapeutic efficacy but a superior safety profile, particularly with respect to hematologic toxicity, which frankly is limiting for the dual PARP inhibitors. We were able to identify two very good series of molecules from a small company in China called Impact.
We partnered with them and have developed 1003 and 1004, a brain-penetrant one, 1004, which Tim has been working with, I think at MD Anderson, and then a non-brain penetrant one, 1003, which is somewhat further along, but both of them are making good progress in the clinic. We're optimistic that they will turn out to be meaningful improvements for patient care, and that's, I think, much of what we're going to talk about today.
It is. For those who haven't seen, we'll also be getting updates for both of those programs at ESMO, and we're going to talk about that as well. I guess to just set the scene, Tim, would you be able to remind us what synthetic lethality is, and kind of the basics of what PARP enzymes do?
Yeah. All right. Let's start off with a tough question, huh? It's never easy to explain synthetic lethality without some slides, but I'll have a go. Essentially, if you think about two genes, X and Y, essentially if both genes are working and viable, you essentially get a viable, functioning cell with no issues. If you have an issue with one of them, loss of function of one of them, the cell still lives and there are no issues. But if you actually have an impact or a non-functioning X and Y gene, that is when the cell will undergo apoptosis and cell death, essentially. An easier way to think about things is to think about synthetic lethality as a table with four different legs, each leg representing a DNA repair pathway, and that could be base excision repair pathway, which is controlled by PARP.
PARP stands for poly ADP-ribose polymerase. There are several different members of the PARP family, but really it's PARP1 and 2 that are relevant to what we're all discussing from a pharmacological perspective. If we go back to that table, one leg represents base excision repair that's controlled by PARP. Another leg could be homologous recombination, which is the error-free way of repairing DNA double-strand breaks, and that's something controlled by, for example, BRCA, the BRCA genes, BRCA1, BRCA2. If you think about how this applies to the clinic, how does this table apply to patients? We know that there are patients with BRCA1, BRCA2 alterations. This could be inherited. For example, Angelina Jolie has a BRCA1 mutation. These patients, we know are at great risk of developing different cancers. That includes breast, ovarian, prostate, pancreatic cancers, and also other malignancies as well.
If we're dealing with an individual who has a high-risk, penetrant, germline BRCA1 or 2 mutation, they have a non-functioning homologous recombination pathway. In other words, their table is standing on three legs rather than the four legs. If you think about that three-legged table and you come in with, say, a pharmacological inhibitor of, say, PARP, you can actually block that other leg that controls base excision repair pathway. If you administer a PARP inhibitor to a patient who has a BRCA mutation plus a cancer, that's going to lead to the taking out, if you like, of that third leg. That table's not going to stand on two legs, and that table's going to collapse. In other words, that tumor cell is going to undergo apoptosis and cell death because you're taking out that third leg.
If you think about the normal cells in these germline carriers, they have a functioning homologous recombination. Their BRCA gene is functioning, and so that table is still standing in the normal cells with three legs without any issues, versus the tumor cells, which will collapse and die and undergo apoptosis because you're taking out the third leg. It is that difference between the cancer cells and the impact that you can have with a PARP inhibitor in those individuals that will lead to that whole synthetic lethality and the real personalized medicine, if you like, to kill the cancer cells without necessarily impacting the normal cells. I hope that
Yeah
summarizes things like
That's great. Within PARP, what is the difference between the function of PARP1 and PARP2? We'll get to talking about the first generation agents, but they all inhibit both of those enzymes.
They do. They impact the base excision repair pathway.
Importantly, from years of research done by many different groups, we know that PARP2 as an enzyme is critical for erythropoiesis. By inhibiting PARP2, that will lead to potential increase in anemia, neutropenia, thrombocytopenia, so myelosuppression because of the impact on erythropoiesis. Really when we think about how we can optimize cell kill, it's PARP1 inhibition that we want to achieve without necessarily hitting PARP2 as an enzyme. That provided the rationale for developing these new generation PARP1 selective agents over just the first generation ones, which target both PARP1 and PARP2, and even some of them target PARP3 as well.
That more selective pharmacological inhibition of PARP1 versus PARP2 or PARP3 or the other PARP enzymes really, I guess, provides us with a novel way of developing a new class of agents, which is cleaner, without off-target effects, and which really focuses in on the PARP1 inhibition and cell kill. The expectation here is that this will lead to greater PARP1 inhibition in patients because as you think about it, if you're taking off the PARP2 inhibition, you can actually drive much higher doses of a PARP1 inhibitor into patients, and therefore achieve much greater pharmacodynamic inhibition in patients. The hope here is that you can actually keep patients on a PARP1 selective inhibitor for much longer period of time without dose interruptions, without dose reductions, because of the PARP2 liability.
Yeah. Okay, great. We will dig into PARP1 selectivity more in a second. I had one other question for you on the role of mutations. Beyond BRCA, there are other mutations that are on label for several of the approved PARP inhibitors. How do those mutations impact the sensitivity of those cells to PARP inhibition?
Yeah, great question. If we think about it, actually, BRCA1 and BRCA2, which is the often talked about alterations, these are really just surrogates for what we are really targeting. And what we are really targeting is homologous recombination deficiency. That is what we are targeting in cancer. And we need to ask ourselves, what are the other surrogates that also lead to homologous recombination deficiency? And it is not just BRCA1 and BRCA2, it is also PALB2 mutations that we often see in breast cancer. It is also RAD51C, RAD51D. There are also other alterations such as FANCA, a whole sleuth of different alterations from CHEK1, CHEK2, ATM, so on and so forth. And we have seen that in prostate cancer, for example, the approval of olaparib and other PARP inhibitors, it is not just in BRCA1 and BRCA2, but in about 14, 15 different alterations, all of which lead to homologous recombination deficiency.
The ultimate goal, of course, and this is probably beyond the scope of this discussion, but it would be wonderful if we could actually identify and develop functional homologous recombination deficiency assays, which is what we are working on here at MD Anderson, and also many other companies are working on this as well, to really develop a functional assay that will identify tumors in patients that have homologous recombination deficiency directly. And that could help us go beyond just the BRCA1, BRCA2, and all of the other alterations that I have already mentioned. So instead of identifying patients who will respond through these different surrogates, we can actually directly identify patients with HRD.
Yeah. It has been a bit of a bumpy ride for the first-gen PARPs and the respective HRD-positive patients versus negative with the label changes for olaparib. I guess there, when patients who are HRD negative, so homologous recombination deficiency mutation negative, what would be the rationale for including those for a PARP inhibitor? I guess, is it because we just are not able to tell if there is a defect in the homologous recombination pathway?
Yeah. I really see the changes in the label, for example, in ovarian cancer, in a resistant setting, that is really more fine-tuning of the patient selection. It is not that these drugs were failures in all of those other indications. We need to remember that the first indication was in ovarian cancer in a maintenance setting. Not in the frontline, but really in patients who have already had other therapies, including platinum-based chemotherapy agents. And so, with the SOLO-1 trial from AstraZeneca leading to frontline approval of olaparib, obviously, if someone is already had a PARP inhibitor, that does change how we respond in terms of the resistance that has already set in. Giving patients another first-generation PARP inhibitor after someone has already been exposed and resistant to another first-generation PARP inhibitor, or even the same one, that, to me, does not quite make sense.
The indication changes really, to me, reflect that change with the indications and the labels, with the SOLO-1 trial showing great benefit in patients with frontline BRCA1 and 2 ovarian cancer.
Yeah.
That, to me, was more of a fine-tuning of the different PARP inhibitors and the patients that would stand the greatest chance of responding.
Yeah. If we think about, we'll get to the future landscape and the opportunities for PARP1 inhibitors in more detail, but in terms of the mechanism by which this would work, there's some literature about chemotherapy agents sensitizing patients to PARP inhibition. How is that working?
Yeah. There are similarities, of course, between platinum-based chemotherapy agents like carboplatin, cisplatin, in terms of the impact on DNA damage and cross-linking of DNA strands, and leading to cellular apoptosis. We need to address upfront that there are similarities between platinum-based drugs and PARP inhibitors. Of course, the big differences are that PARP inhibitors are orally based drugs that are given as pills. Number two, they are better tolerated versus chemotherapy agents. Because they are better tolerated, these are drugs, pills that we can give to patients for a very protracted period of time without the side effect that we see with platinum-based agents. As we think about that, it makes sense to actually try and combine chemotherapy agents plus PARP inhibitors.
And even as we think about the current landscape, combining ADCs, antibody drug conjugates, which are obviously, I would call it, and Roger's very humble about his achievements with Pembro. I would say, when he was at Merck, we were in the era of PD-1, PD-L1 inhibitors, and we are certainly now in the era of ADCs. There are so many different ADCs out there. A big challenge now is how do we enhance and build on the first generation ADCs that have already been approved, for example, ENHERTU, Dato-DXd, TRODELVY, all of those drugs. A good combination to think about, and combinations need to be front and center of what we do. Combining PARP inhibitors and ADCs, particularly Top1 payload ADCs, is a very rational approach, and we can talk about that if you want, the mechanisms.
But certainly combining a Top1 payload ADC plus, a PARP1 selective agent certainly makes a lot of sense from a scientific perspective and hopefully from a clinical perspective as well.
Yeah. I guess quickly, we will get you to explain that to everyone. So what is the mechanism by which a Topoisomerase I payload would sensitize a cell to PARP inhibition?
Yeah. So as we think about how Top1 payload ADCs work, essentially, there is a formation of the Top1 complex that binds to the Top1 molecule, and that basically screws up the whole repair mechanism and stops transcription, translation in cancer cells and leads to cell kill. However, there is a molecule that comes in to actually remove that Top1 payload from the actual damaged DNA. Guess what that molecule is? That molecule is PARP1. So wouldn't it make sense to actually come in with a PARP1 selective inhibitor to stop that molecule from coming in and removing that Top1 complex? That will help actually maintain the damage that one is already achieving with a Top1 payload ADC. So to me, as we think about rational combinations, because right now everyone is combining everything, with an ADC, I don't think that is the right thing to do.
I think we need to go back to science and first principles and to really match up lab research with translational research and clinical research. Certainly a great partner and a rational partner to combine with these Top1 payload ADCs would be a PARP1 selective inhibitor for all of the reasons I mentioned.
Yep. But the overlapping hematoxylin being the primary reason?
Well, from a mechanistic perspective-
The mechanistic, yeah.
think about the efficacy.
Yeah.
The molecule that removes the Top1-
Yeah
complex from the DNA damage in these cells-
Yeah
really is PARP1. It's really-
Yeah
PARP1 that we need to address.
Mm-hmm. Yep.
That is the reason and the rationale for combining a PARP1 selective inhibitor with-
Yeah
an ADC.
Okay, cool.
But if I could jump in, I-
Yeah
would say that we wouldn't mind having a dual PARP inhibitor, PARP2 inhibition as well, if it weren't for the fact that that limits, to a considerable extent, our ability to dose patients with the drug because of the underlying immunologic toxicity. And Tim, you can speak to that very clearly, but patients have a lot of difficulty, even in the maintenance setting after platinum treatment, as we did with olaparib initially. They have a lot of difficulty staying on drug. The anemia is significant. And for many of the dual PARPs, there's very substantial neutropenia and also platelet effects, thrombocytopenia. Olaparib is, I think, the best tolerated. But even so, it's hard to maintain people on this.
If it is really true, we believe it is, but if it is really true that by selectively inhibiting PARP1 and not PARP2, we can improve tolerability, then we can approach all of these important areas that, Tim, you've outlined.
Yeah.
Yeah, to build on what Roger's just said, which I completely agree with, it's also about maintaining patients at meaningful doses at the RP2D, at the recommended doses that have already been shown to be effective.
Yeah.
If there is a way of developing more selective agents without the myelosuppression, we can avoid dose reductions, dose interruptions, because all of that will impact ultimately on the efficacy of the PARP inhibitor. That's the whole rationale, really, for the development of these PARP1 selective agents. We're thinking long term here. We're thinking about survival. We're thinking about PFS and OS. If we can maintain patients at their starting dose of the PARP inhibitor, wouldn't that be great? We've already seen from other PARP1 selective agents, from other companies that you can achieve much greater pharmacokinetic drug exposures as well, with a PARP1 selective agent versus the first-generation PARP inhibitors. The truth is, when we were developing the first-generation PARP inhibitors, we did see DLTs, a lot of myelosuppression at higher doses.
Even at the recommended doses, we still do see myelosuppression, as Roger was alluding to. Therefore, if we could actually come up with a way of dialing out the PARP2 and really focusing on the PARP1 inhibition, that will help create an effective drug, but importantly, a better-tolerated drug with higher drug exposures, better PK, better PD, and to keep those patients on that particular dose and schedule without dose interruptions and dose reductions.
Yes, also we can go into first-line therapy, in principle, much more readily with novel chemotherapeutic agents because we won't be adding the myelosuppression to them, which is often a consequence of new therapies, including antibody-
Yep
drug conjugates.
Yep.
The idea of combining these together is very appealing. This all has to be demonstrated, of course. We don't know-
Yep
this is going to work.
A few more questions on the mechanism. Would we expect any efficacy loss from not inhibiting PARP2? I guess in the context of the first generation of PARP inhibitors, they all do hit both PARP1 and PARP2. Any thoughts there, Tim?
Yeah. Great question. I think we could use early clinical data that have already been generated by AstraZeneca. These are data that have already been presented with saruparib. Even with their PARP1 selective agent, by dialing out the PARP2, they are still seeing impressive efficacy in those clinical trials. I think that in itself answers and addresses that question.
Okay. Yeah.
I think the criticism is that they see myelosuppression that is not very different from what they saw with olaparib, and you published some of that data. The exact reason why that is true is unclear, but it may be that the selectivity of saruparib as the first of the PARP1 nominally selective inhibitors is not quite selective enough, and that is still enigmatic, I think it is fair to say.
Yeah.
I guess the other thing for PARP inhibition that we have not touched on yet is the concepts of trapping versus inhibition. For the four approved PARP inhibitors, they also differ quite substantially on that scale. Help us understand the differences there. Which one is more important, or do we need both?
I think there are different uses for a trapping PARP inhibitor versus a non-trapping PARP inhibitor, but you quite rightly state that the four approved first-generation PARP inhibitors are all good trappers. The one first-generation PARP inhibitor that does not trap PARP is veliparib, and that is a drug that is not approved as a single agent. We do not tend to see monotherapy activity even in BRCA-mutated cancers, or at least a lower rate of responses. Certainly from a trapping perspective, you do need some PARP trapping as you think about the development of these drugs as monotherapy agents. To really drive cells into apoptosis, you do need some trapping there. As you think about combinations, that is where probably trapping becomes less important.
It is still important, but you probably do not need as high of a trapping capability or capacity versus, say, a monotherapy approach because you are basically building and enhancing the efficacy of a cytotoxic drug, whether it is platinum, whether it is an ADC. It is a different function that you are using the PARP inhibitor with.
Roger, I guess as you were evaluating EIK1003 and EIK1004, what was your kind of goal in terms of trapping and the profile that you saw there?
Well, I think as Tim says, trapping seems to be the easy part.
Yeah
in a way. Most of the agents that you find do trap the poly ADP-ribose polymerase enzyme on DNA. It's sort of a feature of many of the DNA damage repair enzymes that if you inhibit their activity, one of the things that you do is you drive them onto DNA in a non-productive context, and then they just sit there, and that's what we call trapping. I think that the enzymatic activity, which involves the transfer of the ADP to the proteins in the area, broadly speaking, I think the enzymatic activity is important. I think that that's a critical factor, ultimately, in driving efficacy. For us, what we wanted to see was selectivity at the enzymatic level. Trapping selectivity clearly has been achieved.
But enzymatic selectivity was harder, and that was what was so appealing for the molecules EIK1003 and EIK1004, which are between 650 and 800-fold selective at the enzymatic level. It turns out, and this goes clearly beyond our conversation today, that PARylation, this business of transferring the ADP-ribose, is a broad regulatory mechanism that's used in a whole variety of different settings. As Tim mentioned, there are 17 of these enzymes encoded in the genome. Not all of them are poly ADP-ribose polymerases. Some of them are mono ADP-ribose polymerases with a similar structure. But all of them are involved in regulating various aspects of gene expression, DNA manipulation, DNA damage repair. So there's a lot of running room there scientifically to understand how we can modulate a PARylation as a way of changing responses to other therapies, I think.
Yeah. So before we go onto the indications that you are pursuing with EIK1003 and EIK1004, in terms of the four approved PARP inhibitors, could you give us a quick snapshot of the strategy there for how they've gone into different tumor types? Because it is quite complicated. They're in lots of different tumors in various combinations or monotherapy, maintenance versus treatment settings.
Yeah, I mean, it's a lot.
Yeah. How do you kind of map it out for yourself?
Yeah, I think of it in terms of the four canonical cancers where we do have PARP inhibitor labels. So ovarian, breast, prostate, and pancreatic cancer. Olaparib is approved in all four of those indications. As we think about which setting in ovarian, it is in the frontline maintenance setting based on the SOLO-1 Data. In breast cancer, it is in two different settings, in the adjuvant but also in a relapse setting. In prostate cancer, it is approved in many different indications now, and that is where it is getting very busy because of approvals, not just in the monotherapy relapse setting, but also in combination with the new hormonal agents that are out there, like abiraterone and enzalutamide, and it is a very busy space with many different combinations that are ongoing.
In pancreatic cancer, it is approved in a maintenance setting, in a frontline germline BRCA-mutated tumors in patients who are responding to their initial platinum chemotherapy agents. So olaparib is approved in all four of those indications. Obviously, you also have niraparib that is approved in ovarian cancer. You also have talazoparib. It is approved in breast cancer. What is the fourth one?
Oh-
Tala-
Olaparib, talazoparib-
Rucaparib-
Rucaparib.
Rucaparib is approved in ovarian-
Yeah
cancer as well. Yes. There you go. So four different indications there, and four different PARP inhibitors approved in different tumor types and different settings.
Yeah. With this sort of complex backdrop, Roger, how did you think about developing the compounds at Eikon across this messy-
Yeah
map?
In a way, the complexity that Tim has referred to is made much simpler by the hypothesis that we are addressing, which is it the case that a rigorously PARP1 selective molecule so removes the myelosuppression hematologic toxicity associated with PARP inhibition, that we can both improve persistence on therapy in individuals, for example, with ovarian cancer who have responded to platinum, but that we can also advance upstream in treatment paradigms and combine these drugs with chemotherapeutic agents that have their own myelosuppression associated with them. It's very straightforward. The first question is it true that selectivity for PARP1 really does improve tolerability? For us, the issue was first, obviously you do a first-in-human study. You look at the sensitive populations, ovarian, breast, prostate, pancreatic, although pancreatic is much less responsive to PARP inhibitors generally.
You look at all of those and you ask, can we continue to advance dose beyond the initial target occupancy and not see hematologic toxicity? That is the approach we took with EIK1003 and EIK1004. EIK1003 is substantially ahead, and once we demonstrated that, yes, it appeared we could with very modest effects on erythropoiesis, then we said, "If that is true, we ought to be able to use these drugs in combination." Some of those data have been presented, for example, at the recent American Society of Clinical Oncology meetings, combinations with taxane. Other combinations like abiraterone, which Tim mentioned, we hope we will be able to present at ESMO. We will, in fact, be able to present at ESMO coming up in October.
You also have another combination study ongoing with carboplatin and paclitaxel?
Well, that is correct. Yes. I mean, for us, the most rigorous test would be the ability to go into a first-line setting
for example, in ovarian cancer and say, in ovarian cancer, being treated with very active chemotherapeutic agents and there are meaningful responses in those circumstances. Can we improve those responses and improve durability by adding a highly selective PARP1 inhibitor, in this case EIK1003?
Yeah.
Those studies are also underway.
Tim has brought up the combination with ADCs as being particularly compelling with Topoisomerase I payloads. You don't have any ADCs maybe yet. How are you thinking about the potential for ADC combos for your compounds.
Yeah, I think that there's great hope, just as Tim says, for being able to combine a highly selective PARP1 inhibitor with an improved safety profile with an ADC, and to take advantage of the delivery of the chemotherapeutic payload more directly. It's important to remember that an antibody drug conjugate is just what it describes. It's a drug that we know, a drug that was used before without the antibody component, but it is now, we believe, or we hope, in most cases, has an improved therapeutic index by virtue of the antibody targeting. So our first responsibility is to show that in the setting of chemotherapy, in chemotherapeutic combinations, we can use these in combination. If that's true, expansion into the ADC populations, and particularly Top1 inhibitors, as Tim Yap mentions,
Yep
that seems like it would be a straightforward and very desirable thing to do.
Yep. One more sort of strategic rationale question for you, Roger, and I guess also Tim. The brain-penetrant PARP1 inhibitor that you have seems like it could have quite a lot of potential in breast cancer, where patients with BRCA mutations are more likely to have brain mets.
At least from my research, niraparib seems to be the only brain-penetrant PARP1/2 inhibitor, and it is not approved in breast cancer. Is that the most logical place to take that program?
Well, yes, of course. I think that, and Tim, for sure, you can comment on this. As therapeutic options evolve, we are seeing more and more breast cancer patients on therapy who, as they begin to fail therapy, present with CNS metastases. One would like to be able to address that. Having a brain-penetrant molecule was therefore something that we and others thought was very important. EIK1004, we think, has the potential to address that. We are enrolling such patients in our initial studies, and we will have the opportunity to see whether or not there is a demonstration of treatment effect. Of course, all of these things will be used in combination, just as Tim said. It has to be the case. Combination therapy is essential for control of tumors.
Yeah, to build on that, as Roger said, brain mets are a real area of unmet need. It is not just good enough to dig them out with surgery or to zap them with radiation, because when we do that, we are only treating the brain mets, but we are not actually treating the systemic problem-
Yeah
that are underlying the growth of and the development of the brain metastases. Here at MD Anderson, we're really focused on developing systemic ways of treating not just the brain mets, but also the extracranial disease at the same time, right? To avoid the whole game of Whac-A-Mole, where you're only-
Yes
treating the brain metastases, but you're not treating the extracranial disease. The optimal therapy or management for brain metastases, to me at least, would be a systemic approach as opposed to just focusing on radiation or surgery, which may, again, not always be possible with brain metastases, right? I think it's a really important development, and really raising the bar, I guess, for all of our patients with brain metastases.
Yeah. Tim, are there any other combinations that you find particularly compelling, either that Eikon is pursuing or could pursue beyond ADCs?
Good question. Yeah, we've been testing many different combinations. Maybe I can just talk about the general principles of what you would want from a combination. I think number one, it's always important to think about the whole rationale behind why you would want to combine it. Is it because the combination may actually prevent resistance from coming in, or can it reverse the resistance from coming in? Number two, will it build on the actual efficacy of a drug by introducing a PARP inhibitor there, and can you enhance that baseline efficacy with just the monotherapy drug? Number three, a really important point to bear in mind during the development of any combinations are contributional components. How do you prove that two drugs is better than one drug?
I think that's really important even during the early development of these combinations, because you don't want to be moving into a very large, costly, randomized, large phase III trial. You really want to have sufficient confidence that you're actually doing the right thing with a very active combination as well. Then finally, therapeutic index is, of course, very, very important. That will impact on how you actually combine these drugs. Do you combine them concurrently? Do you combine them sequentially with a gap scheduling? I think as we think about modern oncology, we need to think about better ways of doing combinations. It's not just a concurrent A+ B equals success, because you may get an overlap of the underlying mechanisms that you're actually inhibiting, and perhaps a sequential approach might be a better strategy for combinations.
I think we need to be more creative, think outside the box to really generate effective and well-tolerated combinations.
Yeah. Let's talk about your programs at Eikon, the PARP1 inhibitors, 1003 and 1004. At a high level, Roger, could you tell us about how you approached signal finding for efficacy and safety in these studies, in terms of, we've talked about this many times, but the HRD status. Then signal finding for safety when you have patients who have had many prior lines of therapy and may have residual anemia from chemo.
Yeah. All great questions. First of all, it's important to recognize that just as we've been saying all along, dual PARP inhibitors work. People are receiving them, and they benefit from them. For gosh sakes, we don't want to introduce something that is not going to be active. Initially, with EIK1003, our first entrant, we said, well, we're fundamentally interested in the issue of hematologic toxicity. We will look in individuals who have the canonical genetic defects, either inherited or tumor-specific, that are associated with favorable treatment outcomes with PARP inhibitors. We'll look at those patients. But we recognize that anyone who has that and is being treated at a world-class center likely has been treated with a PARP inhibitor, so they are going to be PARP experienced. Nevertheless, for reasons of safety ascertainment, that's perfectly okay.
We enrolled patients in the study who had the canonical genetic abnormalities, should be PARP sensitive, most of whom had been previously treated with PARP. As we were able to find evidence of activity and a more favorable, what we believe to be more favorable, toxicity profile, these are all single-arm studies, which are multiple ascending dose studies. We are simply looking to see how far can we push the dose. We began to see patient responses, and that enabled us, just from the standpoint of ethics, to say, we can treat PARP naive patients because we are having the legitimate PARP kind of effect, where actually we are seeing tumor responses in this setting. Of course, from what I have said, in these early studies where we are simply trying to see, does the drug have appropriate pharmaceutical properties? Is it absorbed in patients?
We had confidence it would be but had to demonstrate that. Can we achieve drug levels that are active? Does it persist? Those kinds of things. In the process, we were able to demonstrate, yes, there is activity, and we were able to increase the patient population to include patients who had not previously seen a PARP inhibitor and hence were PARP naive, as they are referred to. In all, we analyzed quite a large number of patients, and that gave us considerable confidence then to go on and do additional studies in combination, some of which we have just mentioned in passing. I should say for EIK1003, pharmaceutically, it is extremely well-behaved. We have explored doses from 10 mg to 160 mg in EIK1003.
There is an adverse effect, an off-target adverse effect associated with that molecule because it has, in general, it is true of all drugs that while they are specific for one particular target, they may hit others. We knew that EIK1003 would potentially have an effect on an enzyme called a phosphodiesterase that is involved in many things. But its inhibition is first seen as an acceleration of heart rate. The heart rate is normal. The conduction system is normal, but the rate increases. We monitored for that very carefully. We are able to find out when that happened and to adjust our doses so that we do not have that problem. The half-life of the drug permits easily once daily dosing. It is about a 24-hour half-life. That makes it looks very good. In terms of the routine safety and tolerability issues, it looked just fine.
That enabled us to begin, as we said, these combination studies. Our first combination study was, in a way, a swing for the fences combination, and that was to combine with paclitaxel at full dose. As Tim knows well, we had tried for a long time to combine olaparib with taxanes in the treatment of second-line ovarian cancer, for example. In patients who fail platinum therapy, those patients, at least initially in the first approvals, patients who failed platinum therapy were platinum insensitive, would go immediately to a taxane as second-line therapy. Those patients are not eligible for PARP inhibition. Only the platinum-sensitive patients are eligible. The question was, could we combine a PARP inhibitor with a taxane? The answer was, we tried a lot. We did a lot of studies. We adjusted dose and schedule. But the overlapping hematologic toxicities were just intolerable.
It just couldn't be used. We thought, well, here, if we have a much less hemotoxic molecule, EIK1003, can we combine it with full-dose paclitaxel in the second-line treatment of ovarian cancer and breast cancer? We did that study. We reported it at ASCO recently, and the answer was absolutely, we could do that. The combination is tolerated. With full dose paclitaxel, along with full dose intensity of EIK1003, yes, you see hematologic toxicity, but the hematologic toxicity is what you would expect from paclitaxel. Now, we have to do a great deal more to demonstrate that that is an effective regimen, but the response rates in heavily pretreated populations looked really quite good.
In fact, in our study, more than 90% of the patients had already received a prior taxane before they enrolled in our study, and yet the vast majority of patients either responded or had stable disease. This is a decent-sized study on the order of 70 patients or so. I would just say from our experience years ago, doing the same thing with olaparib back when Merck and AstraZeneca were working together on the development of that drug, we were never able to achieve something like that. Even with complex sequential regimens of olaparib plus taxane.
Yeah
That opens up the possibility for moving a PARP inhibitor into the second-line therapy setting in combination with taxanes, and of course, stimulated us to do, as you've already mentioned, Imogen, the combination study with platinum and paclitaxel in a first-line setting.
Yeah
Very encouraging results, but early, very early.
Yeah.
Early on in the drug development process. Tim, I don't know if you have any thoughts about that.
Yeah, I guess-
No, I think. Yeah, go ahead, Imogen.
I was going to ask you a specific question. Tim, do you think we have evidence that these drugs are not causing anemia?
Which drugs?
PARP-
Both drugs
selective PARP1 inhibitors.
Again, based on data that have already been presented with other PARP1 selective agents-
Yeah
PARP1 inhibition in itself, just thinking about the mechanism of action, it should cause a little bit of myelosuppression. It's not unexpected. I think it would be naive of us to think that there'll be completely zero myelosuppression just because one is inhibiting PARP1. But what is important is that it should be different, if we like, versus the PARP1/2 inhibitors. By dialing out the PARP2, you do save on a significant amount of myelosuppression there. That will allow one to actually combine a PARP1 selective agent with different agents.
Yeah. How do you both think about interpreting that information given they're small studies, patients have had different prior therapies, different PARP exposures, different mutation status?
Well, as an early phase guy, it's always tough. Because everyone's thinking about a phase III-level type of efficacy in a very homogenous
Yeah
population and probably a very early population, and so expectation's always high.
And that is okay. However, I think one needs to be very careful when interpreting early phase clinical trial data. Again, I am just talking about things in general. As you said, these are heavily pretreated patients. They are usually different cancer types, so a heterogeneous group of patients. They are not your phase III population ultimately, and that is the whole point of, I guess, later phase trials, is to really home in on your optimal population. To identify what we hope to be a more homogenous population of patients so that one can really see what that optimal signal of efficacy is. I think we know enough about the biomarkers to identify certain groups of patients, and now it is really about how can we bring these drugs into earlier lines, where these are patients who are less heavily pretreated. By definition, they should respond even better.
The efficacy rate should be higher. Also focus in on certain tumor types and certain mutations.
Yep. Okay, great. To round us up on PARP1, Roger, could you remind us what should we be expecting to see at ESMO for both of these programs in terms of the number of patients in the different cohorts?
Yeah. I think the most important thing that you will see from Eikon at ESMO, with respect to the PARP inhibitors, is the EIK1004 data, broadly speaking. This is, again, an enormously selective PARP1 inhibitor.
Yeah.
Getting a feel for how that behaved in its multiple ascending dose escalation studies. We have completed a fairly large patient population. I think one always worries when you have a brain-penetrant molecule, that there can be central nervous system specific adverse effects.
Which can range from simply mood changes or changes in sensorium to actual seizure disorders or worse. Those are important safety issues to look for and you like to have a large patient population that you have exposed, so you have some confidence that you can move forward, then to the more selected areas as Tim mentioned.
Yeah.
I think that the other data which people are really going to pay quite a lot of attention to, is the combination study of EIK1003 with abiraterone. Again, as we have discussed, there are now good data with respect to talazoparib and niraparib in combination with novel hormonal agents in the prostate cancer setting. We were eager to see early on what would it look like to combine EIK1003 with abiraterone. It was one of our early study designs. We have completed a fairly large study population. I think people will be interested in seeing what those data look like and comparing them to essentially analogous kind of data that are, of course, much larger data sets, phase III data sets, for example, from the TALAPRO-2 Study.
Yeah
et cetera, that give you a chance to see how those things work together, and in particular, give you a chance to assess, "Gee, what is that like in terms of the tolerability of the combinations and the extent to which there are dose interruptions as-
Okay
mentioned or the need for transfusion therapy because of suppression of red cell production.
Yeah. So in the last 3 minutes, let's touch on the Werner helicase inhibitor, EIK1005, which was the first program to come kind of fully off of the Eikon platform. So Roger, in one minute, could you tell us why it's been so difficult to develop Werner helicase inhibitors and how you were able to make maybe a better one?
Well, first of all, credit to the group at Novartis for having generated the first one by stitching together a set of molecules that bound poorly but could be linked together to create a higher affinity molecule. A set of, let's call them first generation Werner helicase inhibitors have been tested. What we know about those is that they're active, and they are active particularly in the setting of microsatellite instability, which again relates to the ability to resolve higher order structural problems that occur as you're replicating DNA. When there is mispairing of repetitive sequences, those repetitive sequences can expand or contract in the genome. That instability is readily detectable. There are easy clinical assays we can use to detect those.
Patients who have microsatellite instability are actually, their tumors have different characteristics irrespective of histology, frankly, and they're very responsive, as it turns out, to checkpoint inhibitors, and we showed that many years ago.
Yeah
with pembrolizumab. That said, there are patients who don't respond, and there are patients who relapse. There's good reason to believe that a Werner helicase inhibitor would improve therapy in that setting. There's a desire to have a good, well-tolerated Werner helicase inhibitor that's potent. EIK1005, we were able to obtain using a totally different approach, which was to screen for things that changed the motion characteristics of the Werner helicase using
Yeah
our tools.
We obtained a kind of remarkable series of molecules from which EIK1005 was derived orally bioavailable. We believe these molecules will be shown to be well-tolerated. We know that the molecules have a very long half-life. In fact, the profile of the molecules preclinically was so good that we went initially into healthy volunteers, just to see
Yeah
what we had. EIK1005 has a more than nine-day half-life and can be given on a once-weekly format. It could be given more frequently than that at lower dose, but once-weekly seems a well-tolerated regimen. We're now in the process of looking at activity for the Werner helicase in patients who have microsatellite instability. We're also looking at patients who don't have that but might still respond. Those data will be presented for the first time at
Yep
ESMO in October. A pretty exciting development of a well-tolerated, quite potent Werner helicase inhibitor that could have an important influence on treatment.
Yeah. Before we let you go, Tim, I know that you told me something interesting about this. What do you think that investors are missing about the early data from Roche, with the response rate being around 14%, but quite a high disease control rate?
Yeah. I think number one, everyone gets excited by overall response rates, and there's the usual bar of 30% to get excited by that. These are really based on traditional landmarks and cytotoxic agents, ADCs. A Werner helicase inhibitor is a very different agent, different mechanism, and certainly, as you said, there is a disconnect between the overall response rate observed with both the Roche and, well, Vividion plus the Novartis compounds in the phase I trials that have already been presented. But there is a very long tail on the curve in terms of the durable DCR, the disease control rate. You're basically putting patients on trials, and they're actually remaining on study for a very, very protracted, durable period of time.
And that, to me, is a very meaningful parameter from a clinical standpoint because essentially, as you think about the ultimate goals, we're trying to prolong survival, prolong PFS. And to me, that's what we're seeing at an early stage in these clinical trials. Most importantly, the drugs have been well-tolerated, mainly GI toxicities, but no myelosuppression, no other surprising toxicities as well. I think-
Yeah
Overall, these are very, very promising agents.
Great.
Yeah.
Well, thank you all for joining, and thank you both for sharing lots of very valuable nuggets ahead of ESMO. Have a good day, everyone. Bye.
Thanks, everyone.
Thanks so much, Imogen. Bye-bye.
See you, Roger