All right. Thank you very much. Great. Appreciate it. Thank you for joining us on this afternoon session. Up here on the stage, we have members of the Immunocore team. It's been super busy for you guys. We were just talking about that, both coming away from ASCO over the last few days where you presented new phase II data in melanoma for PRAME. You've been at investor events and then, obviously, very busy here at the conference. So that means, David, you're all warmed up and prepared with great answers. So we have David Berman, CSO of Immunocore. We also have Brian Di Donato, CFO of Immunocore, to step in as well. So they're going to tag team. First question, I guess, maybe just for the team. Obviously, at ASCO, you presented the data.
Their market has obviously reacted in trying to interpret the data and is a little bit confused as to how compelling the PRAME data is. David believes that the PRAME data is very compelling and is going to work in phase III. So maybe just walk through the two or three things you see in that ASCO PRAME data that suggests that this PRAME is going to work in melanoma and that the world is more than just overall response.
Yeah. Michael, thanks for inviting us today. I think there's two key points we learned from KIMMTRAK, which was our first approved ImmTAC bispecific. And that was, number one, RECIST response rate underestimated the benefit because we saw an 11% RECIST response rate better survival hazard ratio of 0.51. And the second was that the disease control rate was the best way to predict the significant PFS. So with that in mind, when we think about the brenetafusp, which is the generic name for our PRAME program, we saw some very compelling monotherapy activity in heavily pretreated patients. While we did see some RECIST responses, we saw most of the benefit in the form of disease control. And we saw a very promising disease control rate of about 56%. That compares very well to PD-1 plus LAG-3 in a similar population.
Of course, we're a monotherapy, and that was a combination. Number two is we made a very important discovery about T cell fitness being an important predictor of brenetafusp activity. And we found that T cell fitness improves as we move into earlier lines. And that gave us confidence that although we see great activity in second-line, we believe that will even be higher in first-line. And finally, in our first-line phase III trial, we are combining two agents, each with independent mechanism, each with independent monotherapy activity, brenetafusp plus nivolumab, and we believe to have at least additive activity. So that was essentially the key reasons to believe from the data. I'm happy to go into more.
That the control arm is mostly going to be single agent Nivo.
Correct.
So that your drug is PRAME plus Nivo, that should be Nivo given the totality of the data.
Correct.
Can you maybe let's walk through some of those. So the first one, which I think is most important, is just understanding that you believe the efficacy is really a snapshot of the disease control rate, which is people who are having tumor shrinkage as well as the RECIST criteria. Remember, 30% is just this arbitrary number that defines RECIST criteria, but that there's a 56% rate of people having, I guess, disease control or 36% with confirmed tumor shrinkage plus PR, which is still a high number. And that overall, that is the numbers of patients who are seeing benefit. Can you just elaborate on that? And that's really based on the fact that, hey, go look at KIMMTRAK, that is basically the same thing: low response rate, also high disease control rate, and led to a whopping huge survival benefit.
So one of the interesting observations we see in the spider plot over time is this closed umbrella, if you can imagine a closed umbrella, where the patient's tumors essentially hover between +20% and -30%. And if you look at, and we primarily see that in the PRAME positive patients. In patients whose tumors are PRAME negative, so you wouldn't expect the treatment effect, their tumors continue to grow. So this is essentially the phenotype that we see, which is this disease control. Now, we also have additional reasons to believe that this is not just good performance patients with slow growing tumors, which is one thing you would be thinking about. Number one is half of these patients with disease control actually have tumor reduction. Now, it doesn't reach the -30%, but tumor reduction does not happen spontaneously. It is only a treatment effect.
Number two is we see 42% of patients who are PRAME positive have a ctDNA molecular response. We saw with KIMMTRAK and our other platform programs that molecular response and ctDNA is a better surrogate because it is independent of radiographic assessments. And then this concentration of benefit in PRAME positive and PRAME negative serves as the negative biological control increases the reason to believe.
So on the first part, again, you believe that people who are having disease control or some tumor shrinkage is clearly having biologic effect. And is that because if you look at other melanoma studies, people either have tumor shrinkage or it's quickly going to increase tumor size? And you see all that within the first three to five months if you go look at it. I haven't looked at every single spider plot for all these other recent studies, but if you have melanoma in the first few months, the tumors are either going to be growing or if you get an active drug, it's going to be shrinking. Therefore, you shouldn't be having people in the middle, which is, and you have a lot of those people.
Exactly. So Michael, exactly what you mentioned, which is in a traditional drug, you see an open umbrella, which is you see some patients going down and many patients progressing. Our phenotype is just different. We saw this with KIMMTRAK, this closed umbrella, and we see it now with brenetafusp as well.
By the way, pretty sick patients. These are second and third line, if not greater.
Correct.
Pretty beat-up population.
They progressed on all available therapies.
Yeah. So they're definitely either going up on the tumors, the tumors are getting bigger. In this case, you're not really seeing that.
Exactly.
Okay. And then the second was ctDNA. So biological rationale, molecular evidence that you believe. Now, if I go look back at other studies, do other drugs have ctDNA in their response? I don't know. I don't know. Let's look at this.
Yeah. Yeah. So the ctDNA data hasn't been as widely studied in melanoma as comprehensively for other mechanisms. So it's hard to compare. We just have our own internal data. But interestingly, ctDNA molecular responses have been reported for in lung cancer with the Friends of Cancer Research using different companies' checkpoints. So I think there is growing evidence that ctDNA molecular response is a measure of treatment effect.
And by the way, for those who are not into ctDNA, when you say a response, what is the definition of a ctDNA response? You said what percent of people had a ctDNA response?
So 42% of PRAME positive had a response. Now, because this is a novel endpoint, Michael, it was not clear what the depth of reduction is to call something a response or not. So in the KIMMTRAK program, what we did is we looked at multiple different levels of reduction all the way to clearance from 10%, 50%, 68% reduction from baseline. And what we found with KIMMTRAK and uveal is actually there was almost a direct correlation. The greater the reduction, the longer the survival. We found molecular response, which is defined as a half log reduction to be a good threshold. Doesn't mean it's the best, but it's a good threshold because patients who have a molecular response are better, have definitely longer PFS and survival. So it is.
If you're going to have a half log reduction, you have the math and the analysis that says that those people are going to have longer PFS.
We see that with brenetafusp as well, that there was an association with longer PFS and survival.
For those who don't know their log scales, how much, what percent reduction?
68%.
68%. So this is an assay, blood assay. You're looking at ctDNA, you're measuring it. You take it at end of treatment?
We take it, no, it's early. So it's by week nine. And we use either the Guardant platform or the Natera. So we use validated platforms.
It's a 68% reduction of the quality of the patient.
From baseline.
Of the ctDNA of that cancer.
Correct.
Okay. And then therefore, again, you believe that the disease control, which is the closed umbrella, people are getting tumor shrinkage, may not be necessary 30%, but a lot of disease control, that leads to better PFS. Number two, ctDNA, significant numbers of responders, 68% drops. That should probably lead to something's going on that's not spontaneous. And that all should lead to longer PFS. Now, one of the things is that people would say, okay, well, David, maybe PRAME plus Nivo will be Nivo alone. Seems reasonable. However, in the United States, the dominant standard of care, or at least a very common standard of care, is Opdualag. So here, at least in the United States, to be commercially relevant, you'd want to be at least as good as Opdualag. Do you believe you're better than Opdualag?
I do. Let me walk you through the rationale for that.
Okay.
Number one is it's important to compare in an apples-to-apples population. So there was a study, Relativity-047, which studied Opdualag, which is PD-1 plus LAG-3 in a similar population. In that study, the disease control rate was 40% and the median PFS was 3.2 months. For brenetafusp, in a similar population, the disease control rate was 56%. So that's a monotherapy brenetafusp compared to.
A combination of LAG-3.
A combination. Right. So now in first-line, which is our PRISM-MEL phase III study, it's Nivo plus PRAME versus Nivo plus LAG-3. The Nivos essentially cancel out because that's going to be the same activity. So it's now, can PRAME compete against LAG-3 because the Nivo is going to be the same? And as I mentioned, I think PRAME monotherapy is superior to Nivo plus LAG-3. So I'm willing to bet that PRAME plus Nivo will be superior to PRAME plus, sorry, to Nivo plus LAG-3.
And so that, again, based on that, there's a chart on this for people who want to refer to slides. You can pull it down on the website. Correct?
Yeah. Slide 32.
Slide 37. In a second line population, Opdualag, which is a combination that has got 40% disease control and three months PFS, you're a monotherapy.
Correct.
You have 56% and basically three months PFS. Adding Nivo to your drug is going to be better. That's going to push it higher.
Correct.
Okay. And then tell us the design of the phase III and how long it's going to take to get this. So we go, okay, maybe I believe you, David and Brian, maybe I believe it will work. Is the result coming soon? How long does it take? What do you have to do to show this? Because I think investors are like, how long do I have to wait for the data?
So fortunately, we have started, officially started randomizing into the trial. We expect it to be a three-year trial. So starting in 2024, we expect to complete the trial in 2027.
Data would be in 2027.
Yeah, we expect that, yes.
That's PFS.
PFS and survival.
I mean, PFS and, oh, it's first line. So the difference between PFS and survival is material.
Yes.
PFS is the primary endpoint that reads out in 2027.
Correct. And then survival would be the secondary endpoint.
Right. And I guess one of the questions too is the design of the study, going back to it, yes, it's going to take a few years. Is this mostly OUS? And what visibility, first of all, why is it mostly OUS? And second of all, is that challenging to enroll these studies?
So it will be mostly ex-U.S., probably 90% ex-U.S. based on the historical. The Relativity-047 was 90% ex-U.S.. It's my sense that these first-line melanoma trials are heavily run outside the U.S. I don't think it will be a challenge because we're providing nivolumab study drug as part of the BMS collaboration, but we're providing it to the sites. And so, you know, and then there's the incentive because it's the option of being part of a potentially novel combination.
Well, I mean, if you're outside the United States, you're guaranteed to get some Nivo in either arm.
Yes. So there will be 56% chance of getting Nivo plus an additional benefit. Correct.
Yeah. So you're getting a drug and then you have the opportunity to get a better drug. So why would it be OUS? And the reason I ask is because there's not a lot of options. In first line, same thing. You could get Opdualag. I mean, I guess in this case, the United States, it's a 50% chance of being randomized to Nivo plus an investigational agent where here you'd get Opdualag. So it's more of a different standard here. You have a chance, investigational versus that if you didn't enroll in the study, you're guaranteed to get Opdualag.
Correct.
That kind of a thing.
Yes.
There's a well-established.
Well, no. Just to be clear, the control arm is Opdualag in the United States.
Right. So in the United States here, if you enroll in the study, it's a 50% chance to get Opdualag, 50% chance of getting Nivo plus investigational agent. If you didn't enroll in the study and I went down to my local cancer center, I'm going to get Opdualag. It's going to be reimbursed.
Understood.
100% certainty you get that. Outside the U.S., you're guaranteed to get a drug. That's not that easy to get that. And you're guaranteed to get that drug in both arms.
Correct.
Yeah. Then in terms of site enrollment in this, I think you were making some comments at ASCO in previous or last few days that it's just mostly blocking and tackling and it takes time to enroll these sites, et cetera. Opdualag took 36 months, so it's not.
Correct. Yes. Yes.
Okay. Now, while we wait for that, there's also next quarter ovarian cancer data. So PRAME, ovarian cancer, and then lung cancer. After coming away from melanoma, David, what should we expect and how will we just, is this press release? Is this medical conference? How will we disclose new ovarian cancer data for PRAME?
Yeah. So ovarian is third quarter and lung is fourth quarter. We're on track for both of those. It will be at a Congress presentation.
Okay. So like ESMO, and it's just our own Jefferies estimate. So at ESMO, there could be data. So it would be, okay, there's an abstract or it's a late breaker or it's an abstract. They'll wait for that. And then there'll be a press release with the data and updated data cutoff because the abstract is older. Is that?
Correct. Yeah. Just like for ASCO.
Okay. And then what should we expect? How many patients? And given what I just saw, melanoma is ovarian easier, harder, same thing. What do we expect in that result?
Yeah. So in terms of number of patients, it's going to be just a little bit south of the number of patients with melanoma. It's going to be both monotherapy and chemo combo. There'll probably be more, relatively more chemo combo than we saw PD-1 combo because there's lots of different chemotherapy combinations. It is a very heavily pretreated population, so platinum resistant. And so we're studying chemotherapies used in the platinum resistant setting.
Yep.
That's the initial data we're going to share for ovarian third quarter. We also are studying ovarian earlier stage and earlier combinations because as we've talked about, we believe this platform will work even better in early stage. That data won't be available for third quarter. That's for next year.
Okay. There is second plus line plus 40-ish patients. Those are monotherapy and combo with chemo.
Correct.
Those data sets may be somewhat similar to the melanoma. It's a closed umbrella.
Correct.
Disease control may have some responses, but a lot of disease control and that that mechanism supports why you'd be excited. But in combination with chemo, is that a reason to be more excited because that's what everyone gets? You know, talk about that.
Yeah. So the additional data will be ctDNA and T cell fitness or carrying through the same themes from the ASCO melanoma. I think the path forward for ovarian is likely to be one of two. I think, you know, when we speak to ovarian investigators and we share the melanoma data and the early ovarian data, their impressions are this type of drug that has a stable disease would be great in the maintenance setting where, for example, Avastin is a similar type with low response rates, but it's a disease stabilization. Or the alternative is a chemotherapy combination in the platinum resistant setting. So those are the two opportunities that I think play to the platform strength.
Yep. And then you obviously would like to move it to first line. So if you were to run a phase III, if you were to run a phase III, it's safe to say that it probably would be a first line study, which means you need to wait for more first line data. And then you need to figure out what that data looks like to support it because like no one would enroll it unless they had data to look at.
I think that's correct. So earlier line setting. Correct.
So there would be no decision on a phase III. Like we're not going to go to ESMO and find out, oh, they're starting phase III because you do, like is that first line? Because by the way, the melanoma thing, you guys started phase III and announced it like a year ago or so, or second half of 2023. We were like dumbfounded and like, whoa, you guys, how did you do that?
Yeah.
That's not going to happen with this.
Yeah.
We need to see more data.
We need to see more data. That's correct. It's when we feel we have enough confidence that the drug has sufficient activity and that we understand the development path that we can make.
That sounds like combination with chemo, earlier line settings, still working on more of that data.
Correct.
More data next year on that.
Correct.
Okay. And then lung cancer. Everyone's always excited about a new lung cancer drug, but it's also an even more challenging tumor than melanoma and ovarian. And so what would we expect in the fourth quarter? Same thing, medical congress?
Yes. Medical congress. I think it's, we're still enrolling that cohort. So ovarian obviously is a little more advanced. But I would characterize it as the signal detection phase, signal seeking phase in lung cancer.
Signal seeking, meaning what does that mean?
Well, with ovarian, we saw a signal already, right? We saw activity and responses. With lung, we saw ctDNA reductions, but we hadn't seen actual responses yet. So I think we're still in the signal seeking phase for lung cancer.
I see a closed umbrella with lung cancer. I mean, how many patients? Is it like 10, 20, fourth quarter? Not 40.
Yeah. It's still ongoing. I don't want to commit to a number yet on lung. More comfortable in ovarian right now.
Okay. We just want more data before we're investing in a bigger lung cancer.
Exactly. Yes. No, we want to be careful how we make these decisions.
Yeah. Yeah. All right. So that's all coming over the next six months in PRAME. KIMMTRAK. Now with KIMMTRAK, a drug that's doing very well across the globe. I don't want to spend too much time about that because the results are doing well and continues modest growth. Fantastic. Great. I don't want to underestimate the fact that you have an approved drug doing almost a $300 million run rate generating cash for you guys. That's great. But there's a possibility that we could significantly expand the market opportunity by more than four, five, six, 8x by going to cutaneous melanoma from uveal melanoma. And you announced last week that you have, well, why don't you go through the announcement of what's going on with KIMMTRAK in cutaneous melanoma? There was an update last week.
Yeah. So we had high confidence in studying KIMMTRAK in cutaneous for a couple of reasons. One is we had the survival benefit in a different type of melanoma, uveal. They both expressed gp100. Number two is we had two phase I studies with very promising survival. So that's why we did the phase II, III.
A lot of patients, by the way. I was dumbfounded by that too.
Yes.
Like 40, 50.
Yes.
It's a lot of patients.
Yes. We have a lot of cutaneous. We have both PD-1 naive cutaneous melanoma, KIMMTRAK as a monotherapy, and we have PD-1 refractory KIMMTRAK plus checkpoints. Now we didn't know, this study is enrolling essentially last line patients who progressed on everything. We didn't know at the time whether KIMMTRAK monotherapy or KIMMTRAK plus checkpoints would be better. So we decided to do a phase II, seamless phase III, where we would use the phase II data to inform whether to drop one of the phase III arms. It was never meant to be a futility because we had a lot of confidence in the trial.
Because of the recent accrual patterns and the accrual being very recent rapidly, the survival will not mature now until next year, at which point the phase III would have, we would have rolled into the seamless phase III and it would have been mostly enrolled. We decided to approach the FDA with a very novel idea, which is to convert the phase II that's already randomized 120 patients into the phase III. Essentially expand it. The FDA agreed to that. The outcome is we're not going to drop an arm. We're actually going to test both arms versus the control. We're going to actually achieve our primary endpoint 12 months earlier than what we had already planned. It seems like a win for us to do that.
The consequence is we're not going to look at that original 120 patients because they're now part of the phase III. And the idea, Michael, is that we will finish the base cases, finish randomization no later than first quarter of 2026, although we're trying to see if we can bring that in. And then the final endpoint would be in the second half of 2026.
So basically, Wall Street was thinking we could get later this year an initial look at, you guys would, and potentially modifying or dropping the arm in the KIMMTRAK phase II and then getting some survival data in 2025. You just went straight into the phase III part and said, "Look, we got 120 patients. Let's not look at the data. Let's just enroll another 130 patients." I think 130. And go right to phase III. So we're kind of just glossing over the phase II and going right to the phase III.
Yeah. And it's very important to understand the reason. The reason is because the survival now, because the accrual happened recently, the survival would only mature now in the first half of 2025, at which point the phase III would have been mostly enrolled. And so deciding to drop an arm at that point wouldn't make sense. It's mostly enrolled.
Let me repeat that again. In Q2 of 2025, there could have been a look at some OS on the phase II. So now you're going straight into phase III. That's going to take completion of enrollment by end of 2025. So the data for survival may be 2026.
Second half 2026.
Yeah. So basically, if you just wait another 18 months, you would get a phase III result.
Exactly. So it's, if you wait a year, I would say, or 18 months, you'll get phase III OS as of phase II.
Rather than Q1, rather than spring or summer of 2025.
Exactly.
If you wait till summer of 2026, you will get phase III data. Why not wait a year? More patients, phase III.
Yeah. Yeah, that's exactly right. Also, it turns out, if we do it this way, Michael, that final endpoint is up to 12 months faster than had we done it the original way. So it just seemed for us to be the prudent way to do it.
Now, how do I think about powering? Because now there's three arms. There's KIMMTRAK PD-1, there's KIMMTRAK, and then there's chemo. Nothing.
Yeah.
Oh, it's clinical trials. So by the way, it's pretty rare that you have clinical trials where the control arm is someone else's clinical trial. But there's not a lot of options for cutaneous melanoma. So appreciating that KIMMTRAK is probably active in cutaneous melanoma, refer to the data. And I'm fairly confident it would work. My point is you have two investigational arms. So usually it's one arm versus a control. You've got two investigational arms versus control. How do you think about that?
Yeah, it's a good question. So there'll be about 170 patients per arm. And we are just finalizing the statistical analysis plan. We need to run it by the FDA to get their agreement. But likely it'll be probably one of the arms versus the control arm. And then there'll be a secondary. If that's positive, the alpha will roll over into comparing the second arm to the control arm. So we will sequentially test both arms.
I mean, this has to do with the laws of combinability, the rules of combination. You're going to have two active drugs against a control. The FDA would say, well, what's the contribution of those two drugs? Well, the other arm is one of the other drugs. So you can see if two is better than one, then there you go.
Yeah. PD-1 by itself shouldn't have activity because the patients by definition are refractory to PD-1.
Why do we need the PD-1 on there?
This comes back to the original trial we did, which is KIMMTRAK plus AstraZeneca's checkpoints post PD-1, where we saw the great one-year survival of 75%. It was because we didn't do the contribution of components then. We don't know whether we would have had the same result with KIMMTRAK alone or whether PD-1 is adding on or whether PD-1 is adding on.
Do you believe it's adding anything in a post PD-1 world?
Probably not, but I don't know that for sure.
Okay. Right. So it may or may not work, but the second arm is KIMMTRAK alone.
Exactly. And so we will now rigorously answer the question, does PD-1 unlock more activity from ImmTAC? We'll be able to answer that in a randomized fashion.
Very good. Well, Brian, David, thank you guys very much. There's also some HIV data coming in. I will refer to, since we're out of time, to take a look at that data coming up, which is very interesting.
Looking forward.
A lot more data coming up. Thank you guys for spending time with us. It's been a long week. It's obviously very helpful to run through all this. Thank you guys very much.
Thank you for hosting us, Michael.
Thank you.
Thank you.