Amgen Inc. (AMGN) Earnings Call Transcript & Summary

Outstanding. Listen, guys, pleasure to have Amgen management join us to -- and a great first day at our conference. We have -- we were scheduled to have Arvind and Rob from the global development side, and I just learned that Peter is joining us as well. So thank you to all of you for being here. Maybe I'll turn it over to you, Peter, just to kick things off, and we'll jump right into it.

How are you doing, Umer? It's great to be here. It's nice to hear you and Eric and Mike. We appreciate the invitation as always. So thank you. And I'd just like to maybe share a few remarks on behalf of all of us at Amgen. I mean we feel like the company is well positioned for attractive long-term growth. Our 3 main drivers are volume-driven growth of our innovative brands and our biosimilar portfolio. We feel great about that. Secondly, some successful near-term launches: LUMAKRAS, as we all know, has brought -- been successful and most importantly for patients with the KRAS G12C mutation, non-small cell lung cancer; Otezla, mild to moderate, we're just a couple of weeks away from the PDUFA date; teze, early PDUFA date in 2021. So near-term launches, we're excited about, and then rapidly progressing our R&D pipeline. And we're joined today by Dr. Rob Lenz, who heads up all of our global development efforts with Dave Reese. So we're fortunate there. We can hear more about that. Third quarter, just to refresh everybody's memory, revenue went up 4% year-over-year, 8% volume growth, record quarterly sales of a number of key products. We were excited about that. Prolia grew 15%. EVENITY was up over 100% year-over-year. And importantly, Repatha was up 33% year-over-year. So we're pleased with that. Otezla was up 13%. LUMAKRAS and the biosimilars, MVASI, KANJINTI, the hem/onc portfolio grew 12% year-over-year. So we're pleased with that also. Maybe turning over then to 3 -- the 3 near-term launches. I won't interrogate those too much right now because that's probably something worth going through with Rob. LUMAKRAS, the launch is going well, and we're excited about that, generated $36 million of sales in the quarter, cumulative sales of $45 million, to refresh everyone's memory through the end of the quarter. Otezla, mild to moderate. We talked about coming teze. And so we see a number of sequential launch opportunities coming through 2025, 2026 with biosimilars, with seeing data on bema in gastric cancer and the atopic derm, AMG 451 through the collaboration with Kyowa Kirin. So that's exciting. Turning over to the pipeline. We'll talk about oncology, I'm sure, and the focus we have on lung cancer. We've got LUMAKRAS, AMG 757, solid tumor BiTE. So targeting DLL3 in the small cell lung cancer, where there probably hasn't been a lot of standard of care in a lot of years. So we're excited what that may mean for patients with squamous cell non-small cell lung cancer. Rob can talk to you about that and our Phase Ib study of bema directed against that FGFR2b receptor. So then GI cancer. We talked about bema prostate cancer. They acquired Teneobio. We've got AMG 160 also, and we've got AMG 509 targeting the STEAP1. So Rob's certainly available to talk about what we're trying to accomplish for patients with prostate cancer. Then maybe turning over to -- we leveraged our strength this year in the BD transactions. We have 5 transactions, different types, all the way from the acquisitions of Rodeo and Teneo and Five Prime to the collaboration with Kyowa Kirin and then the opportunity to remain engaged in that strategic collaboration with Neumora in neuroscience. And then finally, our continued execution and strong balance sheet are going to provide us, and they do provide us, with the firepower to deliver attractive long-term growth. Always allocating first to innovation, great internal, great external innovation; and then two, returning capital to our shareholders. This year, we grew our dividend 10%. We started our dividend in 2011. It's grown meaningfully since then. And then we also have indicated that based on our original range of 3 billion to 5 billion for share repurchases, we expect to end up at the upper end of that range. So we're -- continued strong quality of execution, integrating in the opportunities that we are involved in, in terms of collaborations and acquisitions, strong quality of -- strong innovation and then finally, of course, disciplined capital allocation. So with that, Umer, let me turn it back to you. I'm sure you've got a lot of great questions, especially for Rob.

Okay. Fantastic. So maybe let me kick things off with you, Peter, on some of the points you laid out. I feel like it's -- and I was mentioning this to Arvind the other day as well. It's the first time in a long time where there's starting to be a lot of questions coming in on Amgen. I wouldn't necessarily say there's any sort of a consensus or anything, but there's some questions coming in around sort of the stacking of Otezla, the ENBREL and perhaps denosumab franchise, LOEs happening later in the decade. Whether or not some of those are certainly happening later in the decade is a whole different conversation. But I guess how would you characterize -- I know you mentioned long-term growth. So from your seat -- again, I'm not necessarily getting towards any sort of a guidance or anything, but do you see the growth trends? And you mentioned long term. So do you see that going well past 2025?

That's a great question, Umer. And we've said -- Bob and I and the team have consistently said that we think the company is well positioned for long-term growth. We really like the cards we have right now. We like the launches we've got with LUMAKRAS, with teze. We like this series of biosimilar launches over the next 4 or 5 years. We like bema, data coming and being hopeful for patients, good data in the middle of the decade. Same with AMG 451 in atopic derm. So we like those cards a lot. And we also like our multispecific or bispecific portfolio. And we talked about AMG 757. We've got 160 in prostate. So we like what we've got there. And I would -- 2 of them are kind of turned towards -- we feel really good about Repatha and what it addresses, and we're working really hard on that and executing. We think that's a strong opportunity for us for -- over the long haul. We think Otezla is our -- we defended that through 2028. We feel really good about Otezla mild to moderate indication. We think it's not just a great indication and maybe opens up a significant large patient population there but also gives our field reps an opportunity to go back out and really work -- continue to work Otezla. We're very excited about it. We'll continue to defend all of our products as they come off. As you know, with biologics, you defend them and you keep a couple of extra points every year. It's really, really important, and we love that kind of execution. We're going to continue to do that. So we like the cards we've got. This year, we did 5 transactions, a couple of acquisitions, collaborations. We continue to execute on the collaboration with BeiGene. So that's important. So we like what we have. We'll continue to -- as I said, we've got the firepower and the ability to bring it in and still meet our objectives from an earnings prospect. To think back that we -- this year, I think we shared that we've absorbed about almost $200 million of operating expense in these deals, and we'll continue to operate our businesses efficiently and effectively as we can for our shareholders.

Got it. And Peter, maybe just before we leave this point, just 2 quick follow-ups on that. One, would there ever be a consideration? I know Amgen has done this in the past. Maybe it's for Arvind as well because I know Arvind has defended this over the years. And then you guys did hit it very early. Would there ever be a consideration towards the long-term guidance?

Umer, that's a fair question. So we will continue to gather feedback from analysts and our investors. I think Arvind and the others here for many, many years, and I think the last time long-term guidance we shared around here, Arvind, they've been around 2014 or so. And if we believe there's some kind of disconnect between our long-term outlook and that of the investment community, we'll certainly consider that. But I would just say we like where we're at right now. We listen carefully to our investors. It's really important for us to do that, and we'll continue to work at it that way.

Okay. Fantastic. And just we -- before we pivot into the R&D more specifically, the only other quick one I wanted to touch up on with you also, Peter, is have you guys laid out in dollar terms the type of opportunity we're talking with the TSLP? I know with dupi in asthma, it's multiple billions. Is that a reasonable expectation at TSLP in the $3 billion to $5 billion range? Or is that too high?

I don't think we're prepared to talk about the ranges on that. We -- Arvind, you might, or maybe Omari, have what we've talked about on this. We just see it as a great opportunity and up and down the eosinophiliac stack. Arvind, do you want to jump in?

Yes. I mean we have not provided product-specific guidance, Umer. But suffice it to say, when we think about severe asthma, it's a large population. And the fact that we are going to be able to address the population regardless of the patient's eosinophil phenotype, this is a product that's going to be applicable to all comers. We view this to be a meaningful opportunity for tezepelumab.

Makes sense. Okay. Fantastic. So it's a meaningful opportunity. All right. So I want to jump into the pipeline, and I also want to bring in Mike and Eric as well. But maybe just to open it up, Rob, the question that a lot of people are asking but also quasi confused about is what should we expect on the PD-1 combination data. And let me also caveat that by saying I think the fact that EGFR combination data got presented early on and the PD-1 combo data has been sort of -- I don't want to say held back, but it's more slower in terms of the disclosure. I think expectations in general are not super crazy high on that. I guess how would you level set expectations as we head into next year? Presumably, we might see this cards turn over at ASCO as well.

Yes. So I think we've shared the plan is those data would be at a point that we're ready to share in the first half of 2022, preferably at some medical conference. So that's still the plan. Certainly, there are a number of examples in the targeted therapy space where combination with checkpoint inhibitors has been challenging. We have what we feel is a pretty robust ongoing Phase I combination cohort inclusive and not exclusive to PD-1s and taking a pretty comprehensive approach there to determine what will lead to a safer, more efficacious regimen. And that includes combination. It also is interrogating things like sequential dosing, say, with the PD-1. And there are some analogues of that in the targeted therapy space that have been effective. There's been -- there's some interesting preclinical data showing that LUMAKRAS can actually increase infiltrating tumor infiltrated T cells. So that might support a paradigm where one would try LUMAKRAS first, followed by checkpoint inhibitor as one example. So these are the types of data that we're continuing to generate in the ongoing Phase I combination cohorts.

So maybe not necessarily day 1, PD-1 plus KRAS, maybe more sequenced of sorts?

So we're evaluating both and a combination of those. It will be an empirical-based, data-driven decision on what we feel the optimal regimen is once those data are more sufficiently mature.

Got it. The other one is, Rob, as we look at some of the data that Mirati shared on 5 out of 7 -- although it was not all confirmed, 5 out of 7 ORR in first line. Is that representative of a larger data set or harder to read too much into it? And I guess what would be the bogey for a KRAS mutant, for a KEYTRUDA mono in that type of indication?

Yes. I won't comment specifically on the Mirati data other than just comment on it's a small data set. As you alluded to, it's part of a larger, I think, 40-patient cohort in combination. So -- and the data was shared via a press release, so it's lacking the sort of typical detail that one would want to have before commenting on it or trying to extrapolate at all. So I'll refrain from doing that and just anticipate hopefully seeing those data in a scientific forum that will allow us to make a more informed kind of assessment of it.

All right. And then maybe just your overall take, Rob, on whether a KRAS plus PD-1 plus chemo is doable. Or is that too toxic and the doses have to be taken down more? So is it just KRAS and PD-1?

Yes. So I'd say too early to say. It's probably not surprising that one might expect the more combinations to be associated with greater toxicity. That's sort of what happens in oncology drug development. But the question is probably best addressed through generating clinical data of KRAS with chemo, KRAS with checkpoint inhibitor, establish the safety and hopefully antitumor efficacy in that population before then assessing whether or not sort of a triple combination could be a viable path.

Okay. Okay. Maybe switching to TSLP then. The SOURCE trial, that was the oral steroid-dependent trial in asthma. I guess how does -- what was sort of the thought process on -- the goal behind that specific trial? And how does that change or not change anything from a TSLP for a tezepelumab perspective?

Yes. No. So I'd say we remain very confident around the efficacy and safety profile that tezepelumab has been able to demonstrate in severe asthmatics, as Peter mentioned, including efficacy across the broad population of severe asthma patients regardless of the baseline eosinophil status. In terms of specifically the evidence of efficacy of asthma patients taking oral corticosteroids, we recently presented data from a pooled analysis of patients who participated in the NAVIGATOR. That was the Phase III in the pathway Phase IIb trials in order to have sufficient sample size and look specifically at what was the treatment effect in those type patients who are taking concomitant oral corticosteroids, right, getting at that question of what is the efficacy in those patient populations. And what we showed was there was a 41%, which is highly clinically meaningful, reduction in exacerbation in those patients taking concomitant oral corticosteroids. So again, this definitely gives us the confidence that tezepelumab does, in fact, work in patients with severe asthma taking concomitant corticosteroids. And it also gives us further conviction around the reason for the failed oral corticosteroid taper study that you referred to, SOURCE, was really due to design aspects. And certainly, when we shared those data with experts in the field and clinicians, they understand that in a trial where about 50% of the patients on placebo were able to reduce their oral corticosteroids by 90% to 100%, there were some fundamental problems with that trial.

So it was not the active arm.

Right. Yes.

Makes sense. Okay. That's super helpful. All right. Fantastic. Mike, did you want to touch upon something before we move on from the TSLP topic?

No. I'm good. I'm good.

Okay. All right. The other one is, Rob, which one among the other indications would you peg as high probability indications for the TSLP program? Like candidly, I would have thought atopic derm would have worked, and I was very puzzled when it didn't. But I realize the indication focus now is beyond atopic derm. So I'm curious, which ones would you characterize as high-value, high-priority indications from your perspective beyond asthma?

Right. Well, given TSLP is really at the top of that inflammatory signaling cascade, we believe there could be multiple potential diseases to target there. We already have 3 trials that are underway. The ones -- the Phase III trial enrolling patients with chronic rhinosinusitis in nasal polyps, and that's one that I -- if I could say, that we have a high degree of confidence, I think it was your question around. And that's based on some sub-analysis Phase III trial that we recently presented at the European Respiratory Society. And that showed that teze reduced the exacerbation, improved lung function and also improved nasal polyp symptoms in the subset of those asthma patients who also had nasal polyps, right? So -- and it resulted in a very robust, mid-80s, sort of 86%, I think it was, exacerbation rate reduction in that population. So that gives us a lot of conviction that in that ongoing Phase III stand-alone trial, we are likely to sort of replicate those findings and demonstrate meaningful benefit in that patient population. We also have 2 ongoing Phase II studies: one in COPD, obviously, challenging disease but massive upside potential there; and then chronic spontaneous urticaria, which is ongoing. With atopic dermatitis, we -- the scientific rationale behind that we thought was compelling. That's what compelled us to pursue that Phase II trial in atopic dermatitis. And there's evidence of biologic activity for sure. It just wasn't [ through ] that we thought was sufficient to advance that product into Phase III, especially when we are seeing data from programs like the OX40 program in atopic dermatitis where we believe very compelling Phase II data. So the resources better applied towards partnering and advancing aggressively program with that rather than advancing teze for atopic dermatitis.

Fantastic. Okay. Got it. All right. And then maybe the last one, I've noticed there's at least one program at -- there's at least one inhaled program at AstraZeneca right now. I know we were hosting there. [ We were part ] of Astra this morning, and we talked about their inhaled dupi and inhaled JAK. And I know Novartis also has an inhaled program. That one happens to be an inhaled TSLP. I guess how do you see any possible competitive risks from those programs? And again, those are inhaled, not monthly injectables.

Yes. Well, all of those programs are extremely, extremely early. I'd say it's probably a bit premature to comment on that. I look forward to seeing data, in particular, in exacerbation data from the reasonably sized either Phase IIb or Phase III trial. But I'd say everything we've seen with tezepelumab, the data that we've seen, the efficacy and the safety profile and the fact that teze will be pretty established in the marketplace before the launch of any of the inhaled program should they move forward gives us a lot of confidence in that. And TSLP is -- an anti-TSLP program is something that any moving forward would be potentially partnered with AstraZeneca.

Got it. Okay. Got it. But Astra does have a freedom to operate on their own inhaled dupi. You guys don't have any economics on that, correct?

Correct.

Okay. Fantastic. All right. Fantastic. Perhaps moving on then to FGFR. I guess how would you interpret the PFS in the U.S./EU relative to the overall trial? And I only ask because like there's a hazard ratio tracking like closer to 0.4 in the sort of rest of world and China, and there's a hazard ratio closer to 0.9 in the U.S./Europe. And I couldn't tell if that's because the scans are being read by different people or there's some other reason contributing to the FGFR efficacy data.

Right. Yes. Well, it's quite common to see regional differences in efficacy, particularly in the smaller trials. So I would just say I'd be very cautious in over-interpreting differences in subgroups. And I would add that our Phase III program will be of sufficient size and sample size to understand if there are, in fact, any real regional differences. I would just add. You mentioned that was the hazard ratio difference in PFS. The difference in overall survival between the regions was really much less in terms of hazard ratio for overall survival in the U.S. and Europe of 0.66.

So that sounds like there's a read -- like central read versus noncentral read issue going on. Was it central reading or no on that trial?

That was -- those were not centrally read.

I see. That's what it is then.

Well, I don't know. Again, I've seen enough of these trials to know that dissecting relatively small Phase II subgroups is problematic and rarely doesn't do that in Phase III. But obviously, in the Phase III bema trial, the scans will be read centrally.

Yes. What will be the timing of Phase III?

So we've already started 2 of the Phase III trials, both the chemo combo as well as the chemo/nivo combo trial that's recently just started.

Got it. Okay. I'm sorry, I meant a readout timing.

We'll give more updates on that as, like I said, we've just kicked off the start of those. So I think we'll give more insight on that as those programs advance.

All right. Very cool. OX40 program, I feel like among a lot of the data, much of the data shared by Amgen in the last many months, this one really got people's attention from, okay, so this is starting to look differentiated over a lot of other stuff and in a bit of a niche of its own, not sort of gunning after 10 other competing programs of similar nature. How would you characterize, first of all, the tolerability profile of this program?

Yes. Well, one, it is what we believe is going to be a first-in-class monoclonal antibody. Specifically -- and hopefully, we can share some of the -- what we feel could be interesting, compelling data that we saw from Phase II that might be differentiated, specifically on the tolerability. The most common side effect is the fevers, less so the chills. It's clearly what I call a first dose effect, upwards of mid-80s to high 80% of patients across the dose experienced it on that first day of dosing or the day after. And then the vast, vast majority of them resolved within 1 day. It didn't result in discontinuations. These are generally mild, something that we've informed patients about, just so they're aware that it's not something that we consider a safety effect. It's more of a sort of a tolerability first dose effect.

Got it. I guess, Rob, would you -- but I have to admit, when I first read that table on tolerability, I was looking at it and wondering if it has, at least in some patients, very COVID shock like symptoms going on in the first day. But I also wondered at the same time, if you take sort of NSAIDs, can't you mitigate much of the pyrexia, et cetera, and the chills?

Again, some patients with COVID, obviously, have very extreme reactions. That's clearly not something that we've seen in this trial. Again, patients sort of lead with their [indiscernible] and discontinue the trial due to the [indiscernible]. It's probably the first thing that I look at when I look at safety data. And in terms of potential prophylaxis or mitigating measures, definitely something we're interested in looking at. That wasn't formally assessed in the Phase II study.

Got it.

Rob, I'm sorry, I just wanted to interject a question. I noticed with regards to the safety profile of the OX40 drug, when I first looked at the safety table, I noticed a couple of canker sores, which would kind of suggest kind of a weakened immune response, which would kind of be commensurate with that decrease in T cell level. Is there anything to be worried about in that regard? Or are these canker sores just fleeting and really kind of a fleeting sort of phenomenon that we're seeing?

Yes. That's described as aphthous ulcers, but yes, the lay term sort of canker sores that they saw in a small number of patients. Again, patients describing them as mild. They're self-limited, didn't result in discontinuation. Mechanistically, what's driving that, not exactly clear. Could be through an NK. You mentioned NK cell [indiscernible] mechanism. That's certainly probably what's mediating the ADCC mediated cellular depletion of the mechanism, which that cellular depletion component, we believe, is likely what's underlying the ability to potentially dose at a very infrequent basis. But in terms of what's underlying the aphthous ulcers or canker sores, not sure exactly mechanistically. But we're self-limited, not leading to discontinuations for those patients.

Got it. Got it. And sorry, if we already touched upon this, but a big question we're receiving from investors is how it compares to Sanofi. Obviously, Sanofi hits the -- hits the ligand, I believe. Now Sanofi hits -- is different -- in a different regard with that. In terms of the EASI-75, the data look kind of comparable. But I know you guys enrolled a very recalcitrant population. Maybe you could kind of speak to where will you see that vis-a-vis Sanofi's program efficacy-wise.

Yes. Well, not surprising. I won't compare the efficacy of our program to the Sanofi program. Obviously, there are often differences in terms of the trial design, the patient population that are enrolled. I find the helpful thing, to be honest, to look at is placebo effects within a trial for diseases like that. It gives you a sense of what the degree of sort of how recalcitrant or refractory these patients are. You'll see from the OX40, AMG 451 Phase II study, there was a very, very small placebo effect. The EASI-75, almost no placebo effect, and the EASI-90 certainly suggested that this was relatively recalcitrant patient population. About 14% of the patients in that trial had previously been exposed to biologics, most of which were dupilumab. So the consistency there suggests pretty refractory patient population. Perhaps it's being driven in part by the underlying previous biologic exposures. And in spite of that, we saw what we feel is very compelling and very, very consistent treatment effects across all the doses that were evaluated. So that gives one a lot of confidence when we study that broad of a dose range, and we see highly consistent statistically significant effects on the primary and secondary outcome measures. So we...

Rob, can I ask a really crazy question on this? Why can't this be an atopic derm vaccine?

Haven't thought about that.

Just take it twice a year. Don't take it. My point is you could...

This program? I'm sorry I thought about why couldn't someone...

No, no, no. I meant this one. Like position it like kind of like the PCSK9 and the RNAi side. Like why can't this be just twice a year and it's all you do and do nothing else?

Yes. Yes. So I mean definitely, mechanistically, we feel that based -- mechanistically and on the Phase II data that we saw, it certainly suggests that infrequent dosing, particularly in maintenance, say, could be a path for this. And that's definitely something that we'll be interrogating formally in the Phase III program. In terms of what that duration is, that'll -- those discussions and plans are ongoing for the Phase III program. But I think the general concept of potentially relatively infrequent maintenance dosing and that being a meaningful point of differentiation and benefit vis-a-vis patients for sure is one of the few areas that we thought were compelling with this data set.

Got it. So -- okay. So like for example, would it -- quarterly or biannual, is that on the table?

We haven't solidified the Phase III designs yet, but it's a -- we're definitely contemplating less frequent than currently available or, say, soon to be approved -- potentially soon-to-be-approved therapies there. So of course, ultimately, the data will determine what is, of course, the true optionality in the end. But I'd just say it's great to be in a position of a molecule that has that kind of optionality.

Yes. Would you run a dupi refractory trial specifically?

I think we will include patients who have previously been exposed to and fail biologic therapies for sure. I mean that's clearly one of the areas of unmet need in this disease area. How we go about doing that, there's different ways to kind of skin that cat, if you will. But the intent is to study and include that type of patient population in the Phase III.

Because, Rob, one thing you mentioned is it was a very refractory population. But when I look at the baseline, it was about 10% to 15% prior biologics. I guess how should we square those 2?

Well, that's where I tend to look at how the placebo patients do, right, and show there's regression to the mean but there's also a true placebo effect in these trials or with any -- almost any trial where you have relatively subjective outcome measures, right? And we see this a lot in inflammation, neuropsychiatry, et cetera. And when you see a trial that has that little placebo [ effect ], to me, it's suggestive that you have a relatively refractory patient population that's not sort of spontaneously getting better over the course of the trial but really sort of just saying this is a relapsing/remitting type of illness. [ What you've been waiting ] when you have a patient population is that consistent in the placebo arm, it suggests that you've got a tough population.

Right. Maybe in the last 5 minutes or so, let us hit on a couple of BiTEs. Sorry, Eric, do you want to start with the BCMA BiTE, for example?

Yes. So on the BCMA, we were curious about where we are on the trial resumption. And what dose showed the signal that prompted the FDA to start this discussion on safety?

Yes. So we've had discussions with FDA, and that trial is back up and running again with the modified dosing schedule. There's 2 cohorts that are up and running now. And I'd say we're very encouraged by what we're seeing there in terms of both efficacy as well as lack of meaningful CRS. So encouraged by that. We've also started the subcutaneous cohort recently there as well and interested to see how those data look as well. So I think some key ongoing data from those cohorts, and we'd expect those to be more mature and be able to share some of those results next year.

Rob, what was it that happened in this trial? Like nobody really knew what happened.

Well, it's -- the BCMAs have been associated with cytokine release syndrome that they all have. And the key with the BiTEs in general are optimization of both the dose, be it through stepped dosing or other ways to sort of mitigate Cmax-related cytokine release syndrome, particularly true and somewhat unique with the hematologic malignancies. So...

Right. Rob, do you remember how your overall response rates were tracking? Like in the mid-30s out of the 55 patients, but that was across doses. But then at the very best dose, I think it was 6 patients and 5 out of 6 had a response. And I guess that was a question. As we compare versus Pfizer and the J&J bispecifics, should we be using that 5 out of 6 as the comp? Or should we be using your mid-30s as the comp? Like what's the right number? And again, this will partially depend on the dose you have to move forward with.

Right. Yes. You remember correctly that it was 83% ORR in -- when I think at the time was the latest dose group there. And the doses that we have ongoing now are very, very similar in terms of ultimate exposure. So sort of how you get there is a little bit different. And so I wouldn't over-index to 83%...

But you're seeing that higher ballpark though.

Yes. I would say if it's exposure-related, which it seems to be, and we've achieved that with that dose, then one might expect something in that ballpark.

Got it. Okay. And then, Rob, am I correct in interpreting that your DLL3 BiTE is one of the more promising ones among the emerging BiTEs?

Yes. No, absolutely. We're excited about the data that we've seen so far in the ongoing Phase I study. And we recently just kicked off the Phase II third line small cell lung cancer trial there.

I see. Anything we should be concerned about from a CRS perspective on that one? Because I feel like the response rates on this one are definitely starting to look very intriguing at 30-ish percent.

Yes. No, we've seen promising response rates. And I think equally, if not more, important is we've seen quite impressive durability. And there's patients who now have response that's lasting over a year, which in small cell lung cancer, that's extremely rare. And so that we find very encouraging. DLL3 is in AMG 757. This is a program where we don't see significant amounts of CRS as the dose-limiting toxicity. So it's become very evident to us that CRS as the dose limiting toxicity is highly dependent on the tumor type, on the tumor burden as well as on the actual target. So one can't generalize across the bispecific programs. And here's a perfect example where we're really not seeing CRS as a dose-limiting toxicity.

Got it. Okay. Got it. So less concern on CRS on this one is what it sounds like. And I guess, Rob, theoretically, you could be in a position to go to Peter and say, okay, we're going to set aside x amount of dollars for an expansion cohort, which could potentially be a registration on DLL3. Is that a possibility as we head into 2022?

Yes. No, it's a definitive reality. So we -- sorry, that's what I alluded to. We have -- we started just that, call it, an expansion cohort for Phase II in third line small cell lung cancer. And that's moving forward, and it's potentially registration-enabling, obviously, depending on what results we see in that trial.

And Umer, that's a definite yes, by the way.

All right. So -- got it. So this is just starting up. Is that right?

Yes. That's correct.

Got it. I don't think most people are tracking this. So this could be a potential readout as early as -- I mean, gosh, you could have data certainly by ASCO 2023.

I won't comment on specifically when it would read out, but it's literally just recently started. And so things [ are up in the ] clinical trial.

How many patients, Rob?

It's -- I can't remember if that's been shared yet, but it will be consistent with a sort of expansion cohort supporting a potential accelerated approval.

And these are refractory to everything?

I'm sorry. This would be...

Refractory to everything?

Well, third line patients. And there's no sort of consistently see what's in second line sort of...

Yes. FDA bar in this indication has been so low, and we saw that with the ZEPZELCA approval. So this looks like this is -- okay. Interesting. And maybe one last one. PSMA bispecific, I don't know, for whatever reason, I feel like it's not -- ORR is not quite at the level of being a stand-alone or maybe the data is too early or the dose is too low. I guess how would you characterize that?

Yes. So we have a dose that we're evaluating currently in an expansion cohort that's ongoing. Clearly, I think it's too -- the data are not sufficiently mature yet to comment on it. Those data will be continuing to mature. I think we'll be in a position to share the results of that in the first half of next year. And then I think as Peter nicely laid out, that's our AMG 160 program. But we also now have with the Teneobio acquisition AMG -- what we call the AMG 340, that's the low-affinity CD3 PSMA engager. Anti-PSMA programs do have cytokine release syndrome associated with it. So as the AMG 340 dose escalation data accumulates in parallel with 160 expansion cohort, I think that'll be very complementary data set to give us some optionality around what the most appropriate path forward is. And then simultaneously, AMG 509, our STEAP1 inhibitor, which is in dose escalation as well, we find to be quite encouraging, the data that we're seeing so far. So there's sort of this confluence of data that will be coming out in the first half of next year across our prostate portfolio. And I think it's going to be a pretty exciting time to understand again what that optionality is in terms of which molecules do we see addressing which patient populations or lines of therapy moving forward.

Got it. Peter, Rob, Arvind, any drug I should have asked you about that we didn't bring up?

I think you've covered the gamut, Umer.

Yes?

Yes.

Outstanding. Well, listen, on that note then, thank you again for joining us. This was very, very helpful. Thank you again both for high level as well as for more drug-specific comments. And I guess I'm going to have to do more work on DLL3.

Thank you and -- for inviting us. And you and Mike and Eric, we wish you and our investors and everyone very safe holidays. It's a lot going on out there. So...

Sounds great.

Remain safe. We thank you for the opportunity to be with you as always, Umer.

You, too.

Absolutely. Thank you, guys. Great seeing you as always.

Thanks a lot. Great seeing you.

Likewise.