Amgen Inc. (AMGN) Earnings Call Transcript & Summary

Good afternoon, everybody, and thank you once again for joining us for the Fourth Annual IO Summary that coincides with the SITC conference. I'm Yaron Werber, Biotech Analyst here at Cowen. And it's a great pleasure to moderate the last session of the day with Amgen. We have 2 gentlemen that really need no introduction. David Reese is Chief Medical Officer; and Arvind Sood is VP of Investor Relations. Gentlemen, thanks so much for joining us. We appreciate it.

Thank you, Yaron. And just for clarity, I'm Head of R&D, not, in fact, our Chief Medical Officer.

That is correct. I've been so used to saying Chief Medical Officer all day today, I apologize.

No problem.

So we have a packed agenda, and we'll be pretty efficient over the next 40 minutes or so. What I wanted to drill down first is your BiTE platform, and you've made a lot of improvement over time from the ones that early on that sort of had a shorter half-life to the extended half-life. Can we -- let's spend a minute on that, and specifically, literally structuralize, how is the HLE different in the Gen 1 bispecifics? And what does that mean for the PK of the molecules, and ultimately, their activity as well?

Yes. No, that's a great question, Yaron. Thank you. Let me just start by saying, we are increasingly optimistic about the BiTE platform. Based on what we're seeing, we can delve into some of the recent clinical data here in a moment. But as you mentioned, we've transitioned from a first-generation format, which is BLINCYTO is required to be given by continuous infusion to a second-generation format. We looked at a large number of potential structures, a couple of dozen structures at least and settled on a structure that retains the compact core of the first-generation molecules adds in Fc domain to give it an extended half-life with the goal being really weekly, every couple weekly dosing, ultimately. That will vary by target, which is, I think, something to understand, and that applies to, I think, all such molecules. But we're very happy with what we've seen so far in terms of things like pharmacokinetics and basic behavior, and we're learning a lot, and I'll be happy to go into that in a little more detail in a moment.

And how does the -- maybe in terms of structure, how does the Fc domain differ ultimately from other self-complementary FEVs or other platforms?

Yes. I mean I can't comment on their structure, specifically, some of which are published, some of which aren't. I'll say that, again, we evaluated preclinically a wide array of formats, not only to optimize biologic activity but also for manufacturability. And one of the things that I very much like about our platform, it's off the shelf, of course, but we've industrialized the process from target identification and validation through molecule generation with an eye to being able to produce in a highly efficient fashion, using some of our advanced manufacturing techniques, these molecules. So that industrialization, I think, is also a really important component here.

And what does that mean to manufacturing yields and the ability to have consistent sort of separation of the homodimers and the heterodimers during manufacturing?

Yes. I mean, the manufacturing process, I think, is quite elegant now to give us high-purity, high-yield molecules. And remember, we're giving these in small doses. So you don't -- we don't need to make metric -- a metric ton of BiTES would be enough for 1,000 years. And so that part, I think, is very, very efficient and should be quite competitive from the cost of goods perspective.

Okay. Got it. And in terms of -- is there a difference from the first generation and potency? Or is it really pharmacokinetics and scale?

The answer is unsatisfying because it depends. And it depends more on the target than the format. And so that's the part that remains somewhat empirical in the clinic. Nevertheless, even in cases where we need larger doses, you're talking milligrams, not grams or hundreds of milligrams, but a small number of milligrams for a max dose. So the ultimate dose is going to be small in almost any instance. You muted yourself here. There you go.

Let's shift over to the data that's presented at SITC -- this -- obviously, at this meeting. Specifically in small cell lung cancer, relapsed/refractory patients, with your DLL3 compound, which is delta-like ligand 3. The first question I had is DLL3, in addition to small cell lung cancer, what are the tumors is it expressed on?

Yes. So it's almost ubiquitously expressed in small cell lung cancer tumors. It's expressed in a high percentage of neuroendocrine tumors as well. And at a molecular level or actually an embryologic level, small cell lung cancer and neuroendocrine tumors share a common lineage. So that is not perhaps surprising, but that's another area of potential interest.

Okay. And so this is early data from the first one, the Phase I dose escalation component. Obviously, you're looking at safety and you're looking at efficacy. I think 38 patients had a valuable disease. You had a 16% PR, another 3% unconfirmed PR and the -- in 5 out of 6 responses, there the median follow-up was 8.8 months, and I think they have been fairly durable. You have not reached the MTD yet. So what's your initial conclusions from this initial dataset?

Yes. I'm quite enthusiastic. It's probably worth putting the disease into perspective, small cell lung cancer is a disease for which treatment hasn't really changed in decades. There's standard chemotherapy. The drugs used are the same ones I used as a fellow on the wards many years ago with minor modifications. Checkpoint inhibitors have had a measurable effect, but it's quite modest. So there hasn't been, I would say, a large impact of immunotherapy. And to see these sorts of responses in patients with highly refractive -- refractory disease is quite encouraging. And in some of the later cohorts, there's a suggestion that, that response rate is going up. So -- we think there's -- given the huge unmet medical need here, there's a real opportunity we're continuing dose escalation and dose optimization. But I -- this is a molecule I'm really interested in right now.

And at this point, the -- in terms of safety, really, all CRS were essentially Grade 1 and 2, right? You haven't seen Grade 3 or higher. You haven't seen any neurotox from what I can gather?

Yes. Neurotoxicity is much more common with the CD19-directed molecules. And I think it's worth again pointing out that the toxicity profiles, even in CRS, Cytokine Release Syndrome, are often very target specific or target related. That's a point we can just never forget.

Okay. And so the dose escalation is still ongoing. And once you choose the right dose than you do a dose expansion?

Yes. The trial is designed to permit the robust dose expansion, what's more or less a Phase II study. And we'll provide guidance when we think we're getting to that stage. But again, quite encouraged with where we are right now.

Okay. And is the thought that this is ultimately going to go into a, let's say, potentially a Phase II sort of pivotal, sort of like sotorasib. Initially, monotherapy and then potentially move up and get combined with chemo? Or...

So we're thinking about that. One possibility is checkpoint inhibitor combination, so that's obviously something we're thinking about. And then another approach is to advance it into earlier lines of therapy. Standard therapy here is combination chemotherapy, which actually induces remissions in a large percentage of patients. The problem is relapses occur typically quite quickly. The large majority of patients relapse and most of that -- those relapses occur in less than a year, often months, after initial therapy. And so one thing that we're thinking about is can you then introduce some BiTE sequentially. And think about it as a minimal residual disease setting. We know that disease is there, can you eradicate it? So once we have a fixed-dose going forward, these are the sorts of things we're going to be thinking about in terms of the development program.

And as checkpoint inhibitors shown single-agent activity in small cell?

Checkpoint inhibitors have shown some activity with chemotherapy in small cell. But again, I would say the incremental effect has been quite modest so far.

And in terms of other good preclinical animal models that you can rely on to look at depth of response? Or is it just too hard?

That's a little bit of a challenge because you need to reconstitute a human immune system in many ways. We do have, I would say, probably the best preclinical models, at least, that we're aware of, including animal models where we've reengineered parts of the immune system to carry human CD3, for example, to give us a sense that's quite beneficial in toxicology as well as efficacy studies.

Okay. Great. So let's move to 701. There's going to be updated data at ASH. That's the CD3 BCMA bispecific. The response rate across all dose levels -- and this is really going to be the first time we really say clinical data. We've obviously seen data with 420 in the past, which is the first generation. This is the HLA. The ORR across all patients dose level is, what, 36%. But at the high dose, the response rate was 5 out of 6 or 83%, which now gets definitely at the high end of bispecifics. And I think, possibly the highest, obviously, in a small cohort so far. What other data, we're going to get at ASH? I imagine probably a more robust sample, probably a bigger end.

Yes, exactly. I mean, you'll see more patients. It's just a more fulsome dataset. Everyone knows the abstracts have to go in well in advance of the meeting, of course, in a dose escalation trial. Of course, in the interim, we're accumulating additional data, and so you'll see some of that. But again, here, we're quite happy with what we're seeing. And as we 0 in on a dose, we'll determine the best pathway forward here. It's obviously a competitive area and that colors our thinking as well. So the molecule is behaving how we wanted it to behave.

Have you selected a dose? Or are you still doing the dose escalation?