Immunocore Holdings plc (IMCR) Earnings Call Transcript & Summary

Peter Lawson. I'm one of the biotech analysts at Barclays. Welcome to Barclays Global Healthcare Conference in Miami. Really pleased to have up on stage with me management from Immunocore. And I've got Brian Di Donato, CFO; and also David Berman, Head of R&D. I'd love to talk through kind of KIMMTRAK, PRAME, you guys pivot back a little bit to melanoma.

And with that, the first question would really be around kind of the commercial launch for KIMMTRAK and kind of the growth drivers that we should be thinking about. And also kind of any headwinds that we should be contemplating in uveal melanoma.

Yes. First, Peter, thanks for having us. Great to be here. Great to be in South Beach. We're really pleased with the launch of KIMMTRAK in the metastatic uveal melanoma, the only approved therapy for metastatic uveal melanoma. Just to remind you, we have an off-the-shelf bispecific that targets a protein called GP100. And our bispecific, it is designed for a specific HLA type, it's HLA-A02, which is about 50% of the Western population. So we launched about 2 years ago in the United States, and we're launching in Europe on a rolling basis as we get reimbursement agreement with countries. Really pleased with the launch in the United States. We are the first line standard of care now in the United States. The duration of therapy is like 10 months, which is exceeding the Phase III trial. The real world is better than the Phase III. And we have more to grow. So initially, as a biologic, the academic centers were the first to roll out KIMMTRAK, and now we're -- now a majority of patients are treated in the community. So we expect incremental growth in the United States from here. So just to remind you, our sales last year were $239 million net globally, with over 70% of those coming from the United States. And then in France and Germany, we launched shortly thereafter. And we feel like we have great penetration in those markets and also first-line standard of care there. We now have reimbursement agreement with Canada, Australia, Spain, Italy, and those countries are rolling out now. But Peter, the European reimbursement environment, talking about headwinds, is really challenging. So, very challenging in Europe right now for reimbursement. So those prices are a fraction of the United States price. So I think you should expect, from a net sales perspective, most of our growth to still come from the United States.

Are there other countries we should be thinking about whether launches, whether it's outside the EU, to kind of help bolster ex U.S.?

Yes. So as I said, Canada and Australia, we reached reimbursement agreement and -- but those will come online throughout this year. But those will be just incremental to what the United States is.

And how penetrated is it? Are you in the kind of the community setting as we think about whether it's the U.S. or ex U.S., both? And then, where can that hedge in '24?

Yes, since this is the first approved treatment for metastatic uveal melanoma. The patient population in the United States, we estimate, is roughly about 500 that are HLA-A02 positive, and about the same, maybe slightly more in Europe. What we know is what -- how many patients we have on drug, and we know we have still more room to grow in the United States. We don't know exact numbers. We know that, in the community, physicians traditionally have treated with IPI NIVO so it's great to see KIMMTRAK now being the first-line standard of care.

Where's the real world treatment duration? And what does it look like in like first line versus second line, if you've got that kind of level of granularity?

Yes. On the Phase III, the duration of treatment was initially 7 months, and then with the 3-year follow-up, it expanded to around 9 months. We can see that with the -- in the commercial market, in the real world market, we're now over 10 months duration of therapy. The reason for this is KIMMTRAK has a very -- in our 3-year survival follow-up, it has a really consistent tail of patients that are living quite a long time. Somewhere around 25% of patients continue to be on drug and do extremely well. So you would expect, as that tail -- if that tail continues at 5 years, et cetera, that duration of therapy, the mean duration of therapy, you would expect to extend.

Got you. And then just with emerging competitors in that, whether it's first-line indication, kind of how we think about that taking share or disrupting your opportunity there?

It was great that KIMMTRAK in the Phase III had a hazard ratio of 0.51 versus investigator choice. An investigator choice is around 82% KEYTRUDA so PD1. So this is pretty remarkable. We also know that in small tumors, tumors less than 3 centimeters, the hazard ratio is 0.36. So treating patients earlier is even better. So this is an amazing survival benefit for these patients. And I think going head-to-head with KIMMTRAK in the HLA-A02 positive population would be very challenging. I know there are other companies that are looking at the HLA-A02 negative population, which is about 50% of the population, and that's great for patients.

Got you.

Yes. Sorry, just to add that we also are moving into the adjuvant setting with our ATOM trial. So not only will we have the indication in metastatic uveal melanoma, but our hope is to also have a global indication in adjuvant high-risk uveal melanoma, to set the standard of care there as well.

And kind of as we think about, I guess -- to pivot but just over to the PRAME side of things. Kind of interesting data set potentially in 2Q, kind of how much -- how many patients should we expect to see? And what should we be focused on?

Yes. So Peter, we haven't commented on the number, but what we have said is it will be a sufficient number of patients in order to inform both clinical activity and safety. And that's -- will be monotherapy data. We will also have combination data with an anti-PD1. But there, it's a handful of melanoma patients, and then some other patients as well. So it's primarily a safety data set because all the patients have been previously PD1-treated. So it was really just to show that you can safely combine with an anti-PD1. So that's how I would talk about the numbers of patients.

Got you. And so that combination with PD1, it's not the idea of making cold tumors hot again or...

So there's 2 ways to combine with an anti-PD1. One is where you have anti-PD1 as a standard of care, and that's where we're combining with anti-PD1 in first line in our Phase III trial, just you're combining with a standard of care. The other way is the scientific, which is will T cells upregulate PD1 in the heavily pretreated setting and will an anti-PD1 further unleash the impact activity? So it's a scientific question. And so that's one of the questions that we need to ask. And we're asking that in the randomized KIMMTRAK TEBE-AM trial.

Perfect. And then the -- when we think about the cutaneous melanoma PRAME data, what's the success look like around that data?

So there are 3 questions that -- or 4 questions I typically ask when I'm developing a drug in a given tumor. Number one, is it active? Number two, what percentage of the patients are having clinical benefit? Number three, what is the form of that clinical benefit? And number four, what is the durability of that benefit? Those 4 questions are critical in order to determine the next steps. So for F106C in cutaneous melanoma, which is the second Q data set, we certainly know the answer to number one is, yes, we know the F106C is clinically active. Number two is we believe and will show that a sufficient number of patients are benefiting. So we'll be able to show that. Now the form of benefit, I think, is really important. And it's actually quite interesting. For checkpoints for anti-PD1s in melanoma, the form of clinical benefit is in partial responses and durable PRs. What's emerging for F106C is that the form of clinical benefit is not only partial responses, although we do see some, but it's also stable disease and disease control. And the combination of stable disease and partial response is, in a sense, progression-free survival, which is why we have confidence in PFS in the first-line Phase III trial. One of the points just to add in terms of stable disease, because sometimes people think a stable disease just means a very slowly growing tumor. It could just be a patient who was enrolled who just had a slowly growing tumor, they would have had that anyway. What we're seeing with F106C, and we saw this to a certain degree with KIMMTRAK, although I think more with F106C, is we are seeing patients who are having tumor reduction that doesn't hit the minus 30% threshold. And so it's called a stable disease, but they're behaving like partial responders. So they are formerly called stable disease, but they are, in essence, having a treatment effect. And I think part of the challenge is that the resist criteria of a minus 30% threshold, which was designed for chemotherapy, probably does not adequately categorize responders for F106C in cutaneous melanoma. But that's essentially what -- so we are guiding to looking at partial response plus stable disease, which is in essence, PFS, progression-free survival.

And just thinking about competitive data within that space of the differentiation, I guess, beyond the mechanism, beyond the mode of activity for your molecule versus Immatics' is T cell approach and their bispecific approach.

So their T cell engager approach, we've only seen preclinical data, and there's no clinical data so I think it's hard to compare. And even when clinical data comes out then, it's great to see more TCR bispecifics. It will probably take some time to sort it out. If you think of anti-PD1s into the clinic and it takes years to differentiate them, and there may not be a difference, just because there's differences in clinical trials and patients enrolled, et cetera. In terms of TCRT or TILs, I think those are 2 completely different platforms. On one hand, those autologous approaches have some patient enrichment because you're excluding patients who are rapidly progressing, there's the chemotherapy that you have to give, and then there's the IL-2. And they take the T cells out and they make them very immune-fit exogenously and then put them back in. So there are differences in the platform that I think are hard to compare. I think our approach, which for F106C in cutaneous melanoma is this disease control approach lends itself very nicely to moving into early alliance. And we saw that, by the way, with KIMMTRAK, we saw that as we move from second line to first line with KIMMTRAK, we saw a tripling in the circulating tumor DNA clearance. And as we move to smaller tumors, as Brian mentioned, we saw an even better survival benefit. So our platform lends itself to moving into earlier stages and then eventually into the adjuvant space.

Interesting. Okay. And yes, you feel there's a barrier, essentially, for these TCRTs that may not be able to get into these earlier.

One can never say never, but I just think practically, we've had some examples where our trial has been at the same centers as TCRTs and the investigators have said to us, I intended to put this patient on a TCRT, but they're progressing too rapidly so I'm going to put them on your trial. Just to give you an example where our approach when we do it is essentially an ITT approach, everyone who comes on is included, where sometimes autologous T cell therapies, it's patients who can't make it on are just not even put onto the trial. So I think it's going to be challenging in that type of approach when you have to rely on ITT and you have to move into a randomized setting and you have to do an ITT analysis, it can, in theory, be challenging for an autologous T cell approach.

Interesting. Do you ever think you could end up using a bispecific as like a bridging therapy to TCRT?

I think they use chemotherapy as a bridging. I think, for me, the vision is that for F106C and for all our platforms, it's combine them with standards of care, and we can talk about the combinations and why that makes sense. We haven't yet seen anything that we cannot combine. So it can be combined with chemotherapy with checkpoints. Our approach is to move into earlier lines, and then with the ATOM trial, eventually moving to adjuvant. And I think that's where this platform really will work well because there's no long-term toxicities. Whereas with checkpoints, there is the risk of long-term chronic toxicities.

Interesting. Do you think you could -- yes, maybe we should talk about the combination therapies and like if you can rotate in and out other standard of cares on the backbone of a TCRT?

I think so. I think that -- and we saw this with anti-PD1s, by the way, in lung cancer, which is probably the best example. There was -- if you remember the randomized Phase III trials in second-line lung cancer versus docetaxel, there was a very modest resist response rate. Some of the trials didn't even have a PFS benefit, although there was a survival benefit. Then when they moved into first line with chemotherapy, all of a sudden in first-line chemo plus anti-PD-1s is a very successful opportunity. And so I think there's something about chemotherapies plus T cell therapies that really unleashes opportunity. And that's why I'm particularly interested with F106C, where we're studying with chemotherapies, and we'll share that data with the ovarian cancer readout in third quarter.

Perfect. Just thinking about the ovarian cancer readout, what do we want to see in that data set? What's the real bar in a disease where there's multiple drugs emerging? So just curious how it's going to fit there.

Ovarian cancer is certainly more crowded and complicated from a development standpoint, platinum-sensitive, platinum refractory. It's, of course, a tumor where immune therapy has not had much success before. So when we think about those, if you think about those 4 parameters, those 4 questions, is F106C active in ovarian? Yes, we've shown a partial responses in the original data we showed a few years ago. The question we're asking now is what percentage of patients benefit, how do we characterize that benefit and the durability as a monotherapy. But we're also asking about the combination with chemotherapy. And I think that for us, the opportunity will be to show that the drug is active in ovarian cancer, F106C, and to show that it can be moved into earlier lines with chemotherapy or as a maintenance because I think that plays to the strength of this platform of disease control.

Got you. And the Immatics data that we're going to see this year, is there some things for their bispecific, for the PRAME, is there something we should be looking for in that data set that you'll be honing in on?

Certainly interested to see it. And once again, I think it builds the excitement for T cell -- TCR engagers. We're just at the beginning of this journey of TCR bispecifics. If you think of ADCs, that's taken 10 years to get to where we are today. So I think there's lots of opportunities for improvement. We, ourselves, are very excited about what we have now. There's going to be other approaches that come in over the next couple of years. And I think we'll just have to see where things go. In particular, for their platform, I'm going to be -- I expect there to be clinical activity. I expect it to be well tolerated. I want to look at the types of patients they enroll and the PK, the dosing frequency.

Okay. Perfect. And then the -- as we think about this kind of extended half-life variance that you're also developing, how do those kind of fit into the developmental path?

This is an exact example of what I mentioned, which is there's going to be iterations of development. We pioneered this with KIMMTRAK, which is a very great drug for patients with uveal melanoma with the survival benefit. We're equally excited about F106C, that as part of our continuing optimization that we will continue to do in our research labs, we have this half-life extended version. The question has always been what does the half-life extended bring you? Clearly, or at a base case, it can bring you less frequent dosing as a monotherapy. And that will be per patient convenience. And that's at a base case what we expect. The other question is, will it bring increased efficacy? That's an unknown. With our KIMMTRAK, which had an even shorter half-life than F106C, we saw the hazard to survival benefit with the hazard ratio, it's hard to imagine how you would get an even better survival benefit, but anything is possible. So we have our eyes open. But at the base case, I would say we're focused on taking the insights we've made from F106C and then applying them to a half-life extended version just to develop it faster.

And then in 4Q, we get the update or initial data for KIMMTRAK in cutaneous melanoma. How much data are we going to see? What should we be focused on for that data?

The Phase II, just as a reminder, it's in previously treated cutaneous melanoma. It's got 3 arms, KIMMTRAK mono, KIMMTRAK plus anti-PD1, and then a very innovative control arm. So it's a good randomized study. The Phase II part has a dual primary endpoints of circulating tumor DNA and overall survival. The ctDNA will be mature in fourth quarter, and that's what's going to trigger the analysis. That's an exploratory type endpoint because, although we validated ctDNA with uveal melanoma, it's still exploratory in cutaneous. We have to run large trials and show that it correlates with survival. So it will be an indicative adjunct decision-making for us. The survival is clearly going to be the best way to capture the benefit of KIMMTRAK. We acknowledge it will be immature at this fourth quarter analysis. But what we'll be looking at is not necessarily comparing it to the control arm, because it will be early. We could do that, but it will be early. What we're really looking for is, are the Kaplan-Meier curves between the KIMMTRAK mono and the KIMMTRAK anti-PD1 the same? Or are they clearly separated? Because what we want to really do is make a decision whether to drop one of the arms, ideally drop the KIMMTRAK plus anti-PD1, just to keep it simple as a KIMMTRAK monotherapy. So if the 2 KIMMTRAK arms are essentially overlapping, we'll drop the anti-PD1 arm. If there's differences, then we'll choose the one that has the greater benefit. If -- and we'll use the ctDNA as an adjunct. If we can't make a decision, we seamlessly roll into a Phase III so we can always drop one of the arms later. We'll also look at that data as the survival matures, and say, do we need to adjust the sizing of the Phase III? Maybe we don't need 170 patients per arm. So those are the kinds of questions we're asking in terms of what data we'll release in the fourth quarter, which is perhaps your question, we haven't decided. Probably a press release -- and we haven't yet decided exactly what the type of data because we need to preserve that for Congress, but there will be some information put into our press release. That will be clear.

Perfect. I'd love to pivot to the kind of autoimmunity side of the story, it's fascinating. How are you differentiated in the, again, a relatively crowded space?

We're very excited about what our autoimmune platform can bring. Essentially, what we're doing -- what we're interested in, what our unique angle is bringing immune suppression to the localized tissue or cell that's under inflammatory attack, rather than the sledgehammer of global immunosuppression. And the way we do that is by having our bispecific that's targeting, and the TCR, the T cell receptor, is designed to target a specific tissue, for example, the pancreas for type 1 diabetes. And the effector in here does not recruit and activate T cells like in our cancer program. It actually turns T cells off. And it turns T cells off by PD1 agonism. So the effector end is the mirror image opposite of an anti-PD1, which releases the PD1 break. We're going to push on the PD1 break. So in essence, we're going to coat the target tissue in a PD1 agonist invisibility cloak. Any T cell that approaches that inflamed tissue will become exhausted. That's essentially the unique approach. And it differs. There are some other PD1 agonists that are being developed. There was one published in New England Journal for rheumatoid arthritis. The difference there is that those are systemically active anti-PD1 agonists. Ours are local tissue PD1 agonists, tissue tethered.

When can we see, I guess, updates around whether it's strategy or preclinical data around the autoimmune program?

We have published on the type 1 diabetes program, and that one is in GMP manufacturing now. So it's -- we have a line of sight to the clinic. And the nice thing there is that it is a relatively small Phase I/Phase II to see whether you can actually have some type of effect on delaying the progression or onset of Type 1 diabetes. In terms of our other program, which we're quite excited about because this is our skin immune suppression, so skin-tethered bispecific. This is the first time we have developed a bispecific that is not HLA-restricted. It's a nonclassical target. And so it is a TCR bispecific that could be developed for anyone in the world. It's not HLA-restricted. And this could be for atopic dermatitis or psoriasis or any inflammatory skin disease. This one we'll probably publish on this year or early next year to share more preclinical data. And that's heading towards GMP manufacturing.

And kind of the evidence that's kind of driven you to using antibody-based or T cell engaging approaches that have kind of solidified a role for you in autoimmunity, is that kind of the CD3 antibodies? Just if you could elaborate upon what those kind of functional bits of evidence that drove there.

Well, the -- we know that the targeting end works. We know that you can develop these bispecifics to target a cell or a tissue. And so we've showed that with KIMMTRAK, with F106C, and we know it can also target normal cells because with GP100, with KIMMTRAK, we see rash and we see vitiligo. So we know we can target normal melanocytes. In that case, we're recruiting T cells because it's an anti-cancer drug. For the effector side, on the other end, the reason to believe, I think, comes in 2 forms. One is the data from the anti-PD1 world, where if you treat patients with anti-PD1, they can develop a series of autoimmune diseases, because they unleash T cells that are autoreactive and one can see colitis, hepatitis and endocrinopathy. So there's a whole range of autoimmune diseases that result from PD1 blockade. So we said, if you do the reverse, instead of PD1 blockade, if you actually PD1 agonism, could you actually reverse some autoimmune diseases? And then the anti-PD-1 data that was published in New England Journal last year from Eli Lilly, which showed that there can be a treatment effect for rheumatoid arthritis by systemic PD1 agonism. Of course, the challenge there is that there is a hypothetical risk long term. There was no acute toxicities, but long term, there is a risk of turning off anti-cancer cells, which for our approach we are less worried about because that's limited to the tissue and nonsystemic.

Fantastic approach. So thank you so much.

Thanks a lot, Peter.

Pleasure speaking.

Thanks for having us.

Thank you.