Welcome to day two of Guggenheim's Healthcare Innovation conference. Really pleased to be joined by Eikon Therapeutics. I'm Seamus Fernandez, one of the analysts here at Guggenheim. To my immediate left is Roger Perlmutter. Many of you know Roger from his days at Merck and at Amgen, prior to that, developing breakthrough therapies like Repatha at Amgen and a number of other therapies there. Uniquely, KEYTRUDA, which I believe is now the largest ever oncology product developed and a transformational immuno-oncology therapy. To his immediate left is Freddie Bowie, the CFO. Roger, I think this is just a great opportunity to dive in and talk a little bit more about what came out of ESMO.
I think I just kind of want to jump right into the clinical story of Eikon, because that's really what I think is uniquely exciting, where you've taken this platform and brought it to a number of unique therapeutics.
Mm-hmm. Yeah, ESMO was an important meeting for us. We had seven abstracts at ESMO, which is a lot even for a large company, and certainly for a small company, an enormous number, spanning four major programs, each one of which had different aspects. For us, the meeting was an extraordinary opportunity to interact with key opinion leaders. There were 40,000 attendees at ESMO this year in Madrid. Key opinion leaders, of course, and also our clinical trialists. It's worth mentioning that Eikon, of course, a small biotech company, is kind of a big story. We have over 1,000 patients enrolled in various clinical trials, and we do all of this work ourselves. We do all of the clinical work ourselves. It's not done with a contract research organization. We have our own regulatory, safety, quality, biostats, clinical operations, clinical data management.
The entire thing is done by us. We're busy, let's say it that way. The range of different advanced malignancies, because we focus on advanced malignancies, and our desire is to take people who are really quite ill and permit them to lead their lives for a much longer period of time. We are active in breast and prostate, in ovarian, in non-small cell lung cancer, in melanoma, in gastric, in colorectal, in endometrial, a huge range of different programs addressing important unmet medical needs for patients with advanced malignancy.
Yeah. Great. So, at ESMO, if there were the key datasets that you would highlight, we can kind of go through each of them individually.
Sure. Happy to.
So, let's go through, first and foremost, your 001 asset, or EIK1001 asset, the TLR7/8 dual agonist. Just in terms of the mechanism and the systemic dosing profile, what have we learned so far about this, and what's unique about this TLR7/8 opportunity?
Right. Okay, so EIK1001 is a systemically administered dual agonist of Toll-like receptors seven and eight. To really put it in context, we have to step back for a minute and talk about checkpoint inhibitors.
Yeah.
Really, almost from the beginning, the observation, first of all, that checkpoint inhibition works was a big surprise. Because although immune surveillance as an idea, that is that one thing that your immune system does for you aside from the principal role of protection against infectious agents, of course some of those do cause malignancy, but infectious agents, is that they, as Macfarlane Burnet hypothesized many, many years ago, they are responsible for engaging in a surveillance activity and eliminating malignant cells. I have to be honest with you, that never made any sense to me. People who have immunodeficiency do not get solid tumors. That is not what happens.
Yeah.
There is not a lot of evidence, or there was not a lot of evidence that that mechanism was in play. The availability of checkpoint inhibitors, and from the very first time I saw what became pembrolizumab, now KEYTRUDA, the surprise was, of course, it is not a vaccine. It is not stimulating an immune response. It is revealing a preexisting immune response.
Yeah.
That is what it does. And the question almost immediately was, why does it not work in everyone? And the answer is, it works in a minority, unfortunately.
Yeah.
A minority of patients are responsive to checkpoint inhibitors. And it seemed straightforward to work through that mechanism, because it ought to be the case, although we used to argue about it a lot early on, it was almost certainly a T-cell-mediated recognition event, T lymphocytes recognizing something about tumors, maybe neoepitopes, maybe something else. And for that reason, it should be pretty straightforward to look in a patient and say, "Right, here is what your immune system looked like when you were diagnosed with your malignancy. Now we gave you KEYTRUDA, and your immune system changed.
Yeah.
Because what happened was, T- cells reactive with the tumor expanded, and we can identify them. By the way, every T- cell is marked with its own unique receptor structure. We can enumerate them one after another. This should be easy. It should be duck soup. The answer is, not so easy. Been doing this now for more than a dozen years, can't figure out how it works. We have no idea how KEYTRUDA works.
Yeah.
I mean, we know what the receptor is, and we observe what happens, but we can't really correlate that with the change in the repertoire of lymphocytes directed against the tumor.
Yeah.
Never been able to do that. We therefore can't figure out why is it that people don't respond, the majority don't respond. What's wrong with them, and what can we do to help them? We do have a hint, though, that could be helpful, and that is, as we tried to profile what the differences were, and there are many differences between responders and non-responders. Responders, in general, they have tumor-infiltrating lymphocytes. Non-responders, often not so much.
Yeah.
There's an inflammatory profile there. Initially, we focused on PD-L1, which is one of the two ligands for PD-1, the target of KEYTRUDA, and noted that it always made sense to me that if the tumor is under immune selection, and that's, as I say, we're not priming an immune response. It had to be there already. So if there's immune selection going on in the tumor, then you ought to select for tumor cells that express more PD-L1, wouldn't you? I mean, it seems like it because.
Yeah
they would defend. The tumor simply wants to survive and make more progeny. It seemed like that ought to be true. It was natural, it seemed to me, to say, "Let's look in the PD-L1 high population, because those should be the most responsive." Famously, that's what we did, and famously, that proved to be successful, particularly in lung cancer. In the PD-1 high population, monotherapy with KEYTRUDA gave remarkable response rates, and we presented that actually, those data actually at ESMO back in 2015. We've known this for a long time. Actually, it turned out, though, that it wasn't so simple, because the fact that they were PD-L1 high, it might have reflected immune selection. I can't tell about that. PD-L1 is very much entrained, expression is entrained by gamma interferon.
It was really, there's a gamma interferon response element upstream of PD-L1, the PD-L1 gene, and it was really gamma interferon more than anything else. After profiling every cytokine and chemokine that we could, gamma interferon was really what it came down to. Then the question was, "Well, okay, so what's stimulating that?
Yeah.
Ultimately, it led me to say, "Well, I'm not sure I'm going to figure this out in my lifetime, but what I want to do is try, if I'm going to try and make non-responders responders, I need to make a pro-inflammatory environment.
Yeah.
That was really what drove the combination therapy with chemotherapy.
Yeah.
Right? Because people said, "Well, chemotherapy, it's an immunosuppressant." Well, not really, because chemotherapy, we've sort of, through a lot of hard work, identified chemotherapeutic agents that kill tumor cells more successfully than normal cells. You get more cell death in and around the tumor. Cell death is very pro-inflammatory. Very pro-inflammatory.
Yeah.
Nothing like blood bleeding into subcutaneous tissue to give you an inflammatory response.
Yeah.
Cell death is very pro-inflammatory. That led me to think, "Well, what kinds of other pro-inflammatory signals could we use?" And we tried a bunch of things, and so did everybody else in the world, particularly STING agonists, RIG-I inhibitors, et cetera, et cetera.
Yeah.
None of them terribly effective. All of them had to be given intratumorally, as opposed to chemotherapy, which is given systemically. And it made sense to me to think, "Well, if we could give a pro-inflammatory stimulus systemically, we could expand the T- cell repertoire and get better immune responses." All of this is getting to the fundamental question of why TLRs and why the way we use them. Toll-like receptor agonists have been used intratumorally by a lot of folks. A lot of attempts have been made to stimulate inflammation in tumors by direct administration. We know Toll-like receptors are active drugs. Imiquimod is an approved drug for topical administration.
Yeah.
in basal cell carcinoma, squamous cell carcinoma of the skin. Similarly, resiquimod, which is really EIK1001, resiquimod is a dual TLR7/8 agonist, and it is active in that same range of tumors, and also active in cutaneous T-cell lymphoma. It is quite a potent stimulus of immune reactivity that leads to tumor cell destruction. The thing is that what you want to do is use it systemically, because, of course, the T- cells that you are trying to recruit, the additional specificities that you want to expand, they are not in the tumors.
Right.
They are in the central lymphoid tissue, in the secondary lymphoid tissue, lymph node, spleen. You need to get systemically, but how are you going to give a TLR agonist systemically? They are very potent, and they cause cytokine release syndrome, which can be fatal. Way back when at Merck, now many years ago, a little company called Seven and Eight Biopharmaceuticals approached us with the idea that they had identified a therapeutic window for a TLR7/8 agonist given systemically, and they had seen that window in preclinical species, and they wanted to do it in humans. That frightened me, because of the cytokine release syndrome could, of course, be fatal.
But I eventually said, "This is worth exploring, but we are going to do it extremely carefully, and it will only be done in NCCN hospitals in the U.S., where we have a crash cart in the room, and et cetera." They began that dosing under, Roy Baynes and I were supervising. It is really Roy, let us be clear, not me. Roy. Roy was supervising the broader clinical program. He is an extraordinary clinical development specialist and an extraordinary oncologist. We did see cytokine release syndromes in those early studies, but we noticed that they were only in the heavier patients. That suggested a simple pharmacologic trick, which was, why don't we dose by body surface area instead of by weight? Because if you dose by weight, of course, it expands by the cube because we are three-dimensional objects
Yeah.
Where surface area is two-dimensional. That meant that you could put a cap on the top dose. When we did that, since that time, we have not seen any Grade 3 or above cytokine release syndrome, and we have now administered what is EIK1001 in more than 600 patients. So we have a lot of experience. We are very comfortable with the safety profile. We do see Grade 1, 2, or what we call Grade 1, 2 cytokine release syndromes. Those are fever and chills, but only in low single digits. They are rare, and it is not necessary to hospitalize patients. It is not necessary to give them vascular pressor support or volume support. So the patients tolerate it very well. We had evidence already from a variety of different settings within our own studies that there was single-agent activity.
The late Emmett Schmidt was the one who pointed out many years ago that the thing that best predicted what could combine with the checkpoint inhibitor was single-agent activity. TLR7/8 agonists, EIK1001 being the example, have single-agent activity, can be administered systemically. We had a non-small cell lung cancer, a renal cell carcinoma that had failed all other treatments that responded to monotherapy. We showed it could be used in combination with pembrolizumab. All of that work was done while we were at Merck. Then Seven and Eight Biopharmaceuticals kind of ran out of funding, and they came to us after we had started Eikon and said, "Would you be willing to take over the program?" We said, "Yes, we would." We have expanded it enormously.
All of which leads us to an important piece of data, which is that we had never used what is EIK1001 in combination with chemotherapy, and we were not sure that was safe. So we had to do a safety study, which is called TeLuRide-005, which we presented complete results for at ESMO. This is a very simple single-arm study entirely in the U.S. where you combine standard of care therapy for squamous or non-squamous, which is KEYTRUDA plus chemotherapy, as we showed many years ago, still the standard of care, and then we added EIK1001 to see if it was tolerable. The answer was yes, it was tolerable, and that is what we are looking for. The study enrolled 72 patients. We have safety data. People have been on the study now for a long time.
We have patients who have completed two years of therapy with the triplet. So extremely well tolerated. What we learned from that was that response rates improve both in non-squamous and squamous, as does durability, particularly in the squamous population. We had never presented mature squamous data before, but we did at ESMO, where the overall response rate in our response evaluable population was 72%, which is really quite high.
Yeah.
When you consider what you do with the standard of care gives you roughly, typically, in most studies, something in the mid-50s. More importantly, the disease control rate, which includes those people who have stable disease, was 100%.
Yeah.
This is investigator determined. I mention that and emphasize it because investigator determined, investigators are really tough, right? Whereas blinded independent central review, particularly in non-small cell lung cancer, you just measure the sum of linear dimensions and you call their responders or non-responders. The investigator actually comes in and says, "How are you feeling?" If the person says, "I'm not feeling so great, I think it's coming back," that's a non-responder.
Yeah.
Immediately, that person has progressive disease.
Yeah.
To have 100% means that in the view of the investigator, every single patient is benefiting in some way.
Yeah.
Because they were progressing at the time they began getting therapy. That's kind of a remarkable thing. The duration of response, again, cross-study comparison, so be careful. This is not a sensible thing to do. But in most studies, the duration of response, the combination checkpoint inhibitor plus chemotherapy is on the order of seven months in squamous cell carcinoma. For us, it was 12.9 months. Small study. Could it be play of chance? Could it be different patients? Who knows? But it's pretty remarkable. Those data got a lot of attention at ESMO. It's a very long-winded way of answering your question.
Yeah. In terms of the phase III, the decisions heading to the phase III, what initiates a phase III program fully with EIK1001?
Well, we really have a single study in melanoma and in lung. It is a seamless phase II, III study.
Yeah.
The reason we structure it that way is it is much faster to enroll all of the patients in a single study. Of course, we do not have phase II data other than the TeLuRide-005 study. We do not really have phase II data to support, for example, the melanoma study. We have a handful of patients, but we do not have a controlled, randomized controlled study. What we do is we basically run a phase II study, and we have an independent data monitoring committee that supervises it. The data monitoring committee recommends various things at different interim evaluation points. We have two such studies, one in melanoma, which is the most advanced, and one in non-small cell lung cancer, which is spooling up. The melanoma study has enrolled extremely rapidly. It is enrolling very, very well. It went through its first interim analysis, which was a dose collapse.
We had two different doses of EIK1001. We collapsed to one. That was the first charge to the data monitoring committee. The next interim analysis, which is coming up relatively soon, is should you proceed to phase III? That is having enrolled, at that time, roughly half of the patients total, rough numbers. The question is really not how many you have enrolled, it is how many events do you have? Because it is all an event-driven trial.
Yeah.
You have a certain number of events that triggers an analysis. That puts you in this interesting situation of if the drug's really working and you don't have very many events, you can wait a long time for the next interim analysis.
Right.
We should be so lucky. I'd be happy with that.
Yeah.
At any rate, it's an event-driven trial. When we get enough events, the independent data monitoring committee will opine on whether proceeding to phase III is appropriate. The main thing they're concerned about, and this can happen, by the way, in the best of families, you can accumulate a certain number of adverse events for other reasons. If there was an imbalance in how sick the patients were that were enrolled and you got five immediate deaths very early on in your active arm, it's going to be really hard to catch up. You might have difficulty showing an improvement in overall survival. We're powered for overall survival if you have a lot of early deaths. So the independent data monitoring committee will be evaluating that kind of question. They have stopping rules, in essence.
Assuming that we don't flip one of those stopping rules, then we'll proceed and enroll the rest of the study. For the non-small cell lung cancer, the story is essentially the same, but there's a subtle difference, and that is we did a phase II study. It's EIK1005.
Yeah.
It is a sufficiently large study that you can begin to look at it carefully, and FDA decided to look at it carefully. FDA came back to us and said, "You know, if you get data like this in your randomized controlled study, we agree you should do a randomized controlled study, and you should do a registration-enabling study. That is what we propose, should we?" They said, "Yes." They said, "But if you get data like this, then it is incumbent upon you as sponsor to ask us to review this for accelerated approval.
Yeah.
Right? Based on response rate and duration. They actually put that in the minutes of the record of contact of our interaction with them, which is very unusual in my experience.
Yeah, that is interesting.
What that means is that FDA, you have to always remember when you speak with regulatory agencies, they see everybody's data, right? They see your data, but they see everybody's data. Always listen very carefully to regulators, and when regulators say, "It is incumbent upon you as sponsor," I take that seriously.
Yeah.
There is the possibility that the non-small cell lung cancer study, which will enroll much more quickly because unfortunately there are a lot more patients with lung cancer than with melanoma, that will flip a switch there, and because we get sufficient events and because those events are different in the two arms of the study, that we will have an accelerated approval, which would advance that program by about a year compared to the melanoma study.
Yeah, that could be a game changer.
Yeah.
Let's go to the PARPs. I mean, my look at your PARP inhibitors in terms of what was revealed at ESMO is almost unquestionably the best tolerated PARPs that I've ever seen. Can you just help us understand what is uniquely exciting about having an almost pristinely well-tolerated asset that also can penetrate the CNS.
Yeah.
The way that EIK1004 does?
Yeah. I think our whole goal in the highly selective PARP1 inhibitors is to move from maintenance to treatment. That's the whole goal.
Yeah.
Right now, dual PARP inhibitors like olaparib, and we at Merck helped to develop olaparib along with our colleagues at AstraZeneca. Olaparib is an excellent drug. It is indicated for maintenance of a chemotherapy treatment effect. Ovarian cancer, for example, first-line treatment for ovarian cancer is platinum or platinum plus taxane. If you're a responder, olaparib is a sensible treatment. If you're a non-responder, olaparib isn't indicated. It's not indicated in a non-responder.
Yeah.
If we could use a PARP inhibitor in combination with chemotherapy, we could use it in treatment, but we can't do that because the hematologic toxicity overlaps the hematologic toxicity, the bone marrow suppression that you see in chemotherapy. You can't use it in that context. What we showed at ESMO this year is we presented full data for the combination of EIK1003, the first non-brain penetrant, but very highly selective, 650-fold selective PARP1 inhibitor. We could combine it with paclitaxel in second-line treatment of ovarian and breast cancer, and it was well tolerated. In fact, just the fact that you can do that is quite remarkable. There is no PARP inhibitor, to my knowledge, other than ours, that's ever been combined with full dose, 80 milligrams once weekly paclitaxel. It just doesn't work because.
Right.
The heme tox is too great.
Certainly not dosed more than once.
No, that is right. I mean, it is just not tolerated. We tried with olaparib. I mean, it is not tolerated. Here, we were able to tolerate it, and we had high dose intensities for not only for EIK1003, but also for paclitaxel with a very impressive waterfall plot. The response rates are really quite good. Sure, you get plenty of heme tox from paclitaxel at that dose, but we are not seeing any additivity. As you say, EIK1004, which had its coming out party, if you will, at ESMO, at EIK1004, the heme tox is almost negligible. We had only a single patient with Grade 3 anemia, which is really very remarkable. For olaparib, which is the best tolerated of the dual PARPs, the hematologic toxicity, Grade 3 anemia is about 30% of patients in the typical first line, first in human population.
That looks really good. I think that is the result of PARP1 selectivity. We chose these compounds, EIK1003 and EIK1004, because they are PARP1 selective. I admit the possibility that there is just something special about their chemistry that I do not understand that makes them less heme toxic. It is important to remember that virtually all of the PARP inhibitors out there look quite a bit like olaparib. They are chemically structurally quite similar, sometimes differing only by a single atom. It is possible there is something else that is not PARP-related that has to do with that toxicity. But just, you said it, we will take you out on the road with us. EIK1003 and EIK1004 are the best tolerated PARP inhibitors ever, and they have very good efficacy profiles, so they look quite good.
Yeah. Maybe just to wrap up, in terms of funding all of these programs, I know that there is a lot of different opportunities. Freddie and Roger, how are you looking at the different potential non-dilutive opportunities to fund the various programs as we go forward?
Yeah. Thanks, Seamus, for the question. We have a strong balance sheet, so as of 6/30, we had $531.2 million of cash, and are roughly investing around $70 million a quarter in those programs, largely focused on the clinical spend to support, again, as Roger mentioned, advancing. In addition to a strong balance sheet, we obviously have a number of strategic opportunities to partner, program financing, royalty financing opportunities. I think what is great about this company is that we do have so many shots on goal, which gives us the opportunity to take these assets and partner them non-dilutively, and also still have, again, a wealth of other programs and an engine that is generating new molecular entities on a kind of annual basis.
Right.
It is an exciting time in the company.
Great. We did not have an opportunity to cover the internally developed programs
Yeah.
That are advancing into the clinic now, but what is uniquely exciting is that the platform and the sort of chemists at Eikon are now pumping out internally developed programs to sort of expand the opportunity.
Yeah.
A lot of excitement. For anybody who wants to hear more about the Eikon story from Roger, he will be keynote speaker at noon today with Mathai Mammen, so we are going to have a great discussion.