Amgen Inc. (AMGN) Earnings Call Transcript & Summary

My name is Christie, and I will be your conference facilitator today for Amgen's Conference Call from the European Society for Medical Oncology 2021 Virtual Congress. [Operator Instructions] I would now like to introduce Arvind Sood, Vice President of Investor Relations. Mr. Sood, you may now begin.

Thank you, Christie. Good morning, everyone. Thanks for joining us this morning. So we presented important data at the ESMO meeting yesterday, looking at a combination of our KRAS G12C inhibitor LUMAKRAS with our EGFR inhibitor Vectibix, or panitumumab, in advanced colorectal cancer. We continue to lead with the largest and most comprehensive development program in patients with the KRAS G12C mutation. We look forward to going through this data in additional detail this morning, and we'll also use this opportunity to discuss our broader emerging oncology portfolio. Joining me this morning are Dr. David Reese, our Executive Vice President of R&D; and Dr. PK Morrow, our Vice President for Global Development. Both Dave and PK will make some prepared comments, after which we should have ample time for Q&A. So with that, over to you, Dave.

Thanks, Arvind, and good morning, everyone. Thanks for joining us today. If, Lauren, we can advance to Slide 3. I'll provide some brief remarks. Today's focus will be on the LUMAKRAS development program, but we will make some remarks on other important programs, where we've had progress since we last got together, including the bemarituzumab program as well as AMG 757, our DLL3-targeting BiTE in small cell lung cancer. Can you advance to the next slide, Lauren? And one more? So I think today's data represent the approach that we put in place several years ago, building a portfolio of first-in-class, and we hope always best-in-class, molecules against very high-value targets, focusing on combination therapy and then rapidly advancing key programs as we see promising data. As I mentioned, today's focus will be LUMAKRAS in combination with Vectibix, but I will make some remarks on some of our other important programs, where we do have updates to provide. Next slide, Lauren. This is simply a reference slide that gives you an overview of the broad and deep portfolio. And of course, we'll have much more to say on many of these molecules in the coming months. Next slide. Now we're quite pleased with the progress that we've made with LUMAKRAS. I would say, in fact, extraordinarily pleased. We've had an extremely positive reception to the launch in the U.S. Awareness is up dramatically. The testing rates for the KRAS G12C mutation are now in the 70% to 80% range. So very rapid uptake, with a large number of the -- large percentage of the reference laboratories now, including this, in the standard panel and readout. Within the last week or so, we also received approvals in the U.K. and Canada. Those were countries that participate in Project Orbis. And there are multiple ongoing regulatory reviews in jurisdictions around the world, including Japan and the EU, at the EMA. There are now about 3,000 patients treated across the program. And we are pressing forward with the largest combination therapy development program in the field, including now exploring some triplet combinations. Next slide. Now this is another reference slide that gives you a view of the combination program, in particular, as well as key monotherapy studies. If you focus on the middle of the slide, where it's a CRC, colorectal cancer, today's focus will be on the LUMAKRAS-Vectibix combination, about which we are quite enthusiastic and will be pressing forward. Next slide. I did -- given the interest in the program, did want to give a preview of milestones that you can expect through the remainder of this year and then on into the first half of 2022. We will be initiating, based on the data that you'll see in a moment, a Phase III study in combination with Vectibix in patients with third-line or beyond colorectal cancer. We're working hard to get that study up and running in the coming months. We're also launching a Phase II trial, as we have mentioned before, in lung cancer in the first-line in patients that have biomarker-selected disease with either STK11 mutations and/or PD-L1 negative tumors. A little later in the year at one of the major medical meetings, we'll be presenting additional combination data with both an EGFR small molecule inhibitor, afatinib, as well as a MEK inhibitor. As we look to next year, through the first half, we expect to have the top line data from the confirmatory Phase III trial. This is a head-to-head trial against docetaxel, Taxotere, in patients with second-line and beyond lung cancer. And in addition, data from a Phase II monotherapy study in patients with solid tumors other than lung cancer and colorectal cancer, so the -- that large number of tumors where there's a low frequency of G12C mutations. And this will inform whether any sort of basket indication is potentially feasible with the molecule. And then finally, additional combination data in the first half of the year will include checkpoint inhibitor data, data with at least one SHIP2 inhibitor and potentially other cohorts, all of which are actually enrolling quite well. And additional ones may have data available in the first half for presentation as well. So with that, I'm going to transition things to PK, who will actually review and put into context the data. Before -- next slide, Lauren. Before I do that, I do want to provide updates on a few of the other programs. As we have mentioned, we've had very good discussions with regulatory authorities in the bemarituzumab program. This, again, targets first-out-of-the-gate patients with FGFR2b receptor-positive gastric cancer, about 30% of gastric cancer. We are launching that Phase III program by the end of the year, which will include at least 2 and possibly 3 studies, combining with the backbone regimens that are used commonly around the world, both chemotherapy alone as well as a triplet combination with a checkpoint inhibitor. In addition, we are planning to launch, by the end of the year or the very first few weeks of next year, a Phase II signal-seeking trial in squamous non-small cell lung cancer with bemarituzumab. So that program is moving along on track and quite quickly. Acapatamab, the PSMA-targeting BiTE, also known to you as AMG 160, has completed enrollment in a dose-expansion cohort. We remain encouraged by the clinical data we're seeing. We are looking for longer follow-up, really, through the course of this year, and then we'll provide guidance on that program. But we are preparing for a potential movement into the registration phase with this BiTE as well. And then finally, tarlatamab, or AMG 757, again, a BiTE-targeting DLL3 in small cell lung cancer. We have had good discussions with the regulatory priorities and are now planning to initiate a potentially pivotal Phase II trial. That trial will [indiscernible] doses of the BiTE going forward. More details on that as we get to launch. But we are working very hard to launch that trial in the coming months and are quite pleased with the program, and that it is moving into a potential registration phase. So with that, over to you, PK, to focus on the LUMAKRAS and Vectibix data.

Perfect. Thank you so much, Dave. And if you can advance to the next slide. So I'm now going to walk you through the following 3 conference abstracts, beginning with our recent colorectal cancer data with sotorasib in combination with Vectibix, or pmab, which is then going to be followed by brain metastases data, and then I'll end up with our genomic profiling data in KRAS G12C in non-small cell lung cancer. So next slide. So first, I'd like to review very briefly the recent data from CodeBreaK 101, focusing on the efficacy and the safety data from the combination of sotorasib with Vectibix. As you recall, the CodeBreaK 101 study enrolled patients with locally advanced or metastatic KRAS G12C in colorectal cancer, who received at least one prior line of therapy. And the primary endpoints focused on safety with some key secondary endpoints, including overall response rate, as you can see here, duration of response, PFS and OS. The study is for dosing sotorasib at a dose of 960 milligrams, in combination with the approved dose of Vectibix, which was 6 milligrams per kilogram, intravenously every 2 weeks. And we will review both Part 1 and Part 2 of the Cohort A data, which included patients who had received KRAS G12C therapy as well as those who were naïve to KRAS G12C inhibitor therapy. Next slide. So this slide reviews the fact that these patients had received a median number of 2 lines of therapy, and 5 patients had received prior sotorasib therapy. You can see here the median duration of therapy was 10 months. Next slide. We did not see any dose-limiting toxicities during the DLT period, and most of the treatment-related adverse events were Grade 1 to 2. So you can see here, of the Grade 4 -- excuse me, of the 4 Grade 3 treatment-related adverse events, 3 were Vectibix-related, which were expected, including hypokalemia, rash and acneiform dermatitis, very consistent with the known safety profile of Vectibix; and 1 Grade 3 event was diarrhea and believed to be related to sotorasib. Overall, the combination was extremely well tolerated. Next slide. So this is a very important slide. Let me walk you through it from right to left. So you can see in the right-most column that overall -- the overall response rate, which included both confirmed and unconfirmed responses, was 27%. And we have data that -- at this point, all but one of these responses are now confirmed, and we're waiting the last scan. In particular, in the pseudo-naïve column, which is in the middle column, patients demonstrated an overall response rate of 33%, including the confirmed patients as well as those who had an unconfirmed partial response. In addition, an encouraging 67% also demonstrated stable disease, leading to disease control rate of 83% in this cohort. In the left-most column, this demonstrates the fact that even in a cohort that included sotorasib-pretreated patients, we saw a disease control rate of approximately 75%, which was primarily due to the approximately 63% of patients who had stable disease. I'm now going to double-click into the cohort responses in the next slide. Next slide. So this slide reviews Part 1 Cohort A, which includes 5 patients who had received prior sotorasib. Let me walk you through this slide a little bit. So the top graph represents the tumor shrinkage in each patient in the cohort, and the bottom graph demonstrates the time on therapy for each patient. And so the bars that had the diagonal or hashmarks on them represent the patients who have previously received sotorasib. So for example, if you look at the green bar, you can see that this patient experienced a partial response. And the bottom graph demonstrates that this response correlates with the fact that the patient was on therapy for almost 12 months and remains on therapy today. So you can see that while 4 out of 5 patients with prior sotorasib treatment demonstrated tumor shrinkage, the degree of tumor shrinkage, which ranged from 15% to just under 30%, didn't quite meet the bar for partial response and, thus, was classified as stable disease. Next slide. This is Part 2, which included only sotorasib-naïve patients. So the cohort initially contained 18 patients, but one patient was not able to be included in the disease assessment as their scans are not currently in the database. And the remaining 17 patients, you can see that 15 out of 17, or 88%, demonstrated a decline in the target lesion, with 3 reaching the threshold required for partial response, but the majority being classified as having stable disease. And as we talked about on the previous slide, if you follow each response column down to the corresponding treatment duration, you can see that of those patients who experienced a decline in the target lesions, 14 remain on therapy. So overall, looking at these slides, in both cohorts, you can see the combination of sotorasib and panitumumab led to tumor shrinkage in the majority of patients, accompanied by a tolerable safety profile. Next slide. So now we will move to the first of 2 abstracts presented at the World Conference on Lung Cancer. Next slide. The data that I'll review now is from a post-hoc analysis of the CodeBreaK 100 study. And this study disallowed patients with brain metastases that had either been resected or irradiated at least 4 weeks prior to enrolling on the study. They were allowed to enter CodeBreaK 100 as long as they've had Grade 2 or lower residual neurologic symptoms, they were on a stable dose of steroids, and they've had an MRI within 30 days that showed no new lesions. We wanted to examine the response to sotorasib in both the target as well as the nontarget lesions. And specifically, target lesions were considered lesions large enough for repeated measurements. And some of the nontarget lesions were visible, but too small for very accurate repeat measurements. Next slide. So this slide demonstrates that the lung cancer patients with brain metastases exhibited a disease control rate of approximately 78%, and those without brain mets demonstrated an 84% disease control rate. The rates of Grade 3 and Grade 4 treatment-related adverse events, as you can see here, were very similar between the 2 groups. And there were no fatal treatment-related AEs. Next slide. So we performed central analysis of the brain metastases using RANO, or Response Assessment in Neuro-Oncology, which demonstrated that 2 patients with a valuable brain metastases had a complete response while in sotorasib therapy. And 12 demonstrated stable disease, with an intracranial disease control rate of approximately 88%. Next slide. This is a nice side-by-side MRI comparison slide of a 63-year-old man with stage IV KRAS G12C-mutant non-small cell lung cancer, which show progression of disease on 2 prior lines of therapy, which included chemo as well as anti-PD-1 therapy. And you can see a complete response in the brain mets. Next slide. So now I'm going to review the genomic profiling data from the CodeBreaK 100 study also presented at WCLC. Next slide. We can see here that baseline tissues from CodeBreaK 100 were analyzed for potential genomic profiles that could correlate with early progression, late progression or ongoing disease control. Next slide. While some of these trends we're seeing and analysis, you can see here from left to right, there were no clear genomic signature that were seen to predict response to sotorasib nor resistance. Next slide. And the same is seen on this slide. Next slide. So in conclusion, the data presented today really reinforced the fact that combination of LUMAKRAS plus Vectibix is safe and tolerable in chemorefractory patients with KRAS G12C-mutant colorectal cancer. The adverse events seen in the study were very consistent with the known adverse events for LUMAKRAS as well as Vectibix. And in terms of the efficacy, as mentioned previously, the study demonstrated an overall response rate of 27%, with 33.3% in the LUMAKRAS-naïve arm or cohort. So as a result, we are actively planning and will execute a Phase III study in patients with third-line plus CRC, which is planned for end of this year. The last 2 abstracts demonstrated that LUMAKRAS was associated with the systemic durable anticancer activity in patients with non-small cell lung cancer with stable brain mets. And to further explore this, we're currently enrolling a cohort of second-line plus non-small cell lung cancer patients with active untreated brain mets. And finally, our analysis at CodeBreaK 100 really demonstrated that, at this time, there is no single genetic signature that predicts for LUMAKRAS responses in patients with non-small cell lung cancer. Next slide. So in summary, LUMAKRAS development program demonstrates, first, the fact that we have really moved forward with unparalleled development speed, accelerating and thinking of development in terms of hours and days and rapidly advancing the largest and most comprehensive clinical program with nearly 3,000 global patients, thus far. We have also demonstrated commercial success in oncology, as Dave alluded to previously, and we have a strong commitment in supporting patients with non-small cell lung cancer with our program. We think that we are truly uniquely positioned for success with colorectal cancer with this established portfolio, pursuing triplet combinations with chemo plus Vectibix or MVASI, and our CodeBreaK 101 master protocol as well as our Phase III study that is planned to initiate later this year. And with that, I'll move it and turn it over to Dave. Thank you.

All right. Thanks, PK, for that excellent review of the data. Again, very encouraged by what we've seen here and are moving forward with the Phase III study. So at this point in time, we will open it up for Q&A. Christie, can you go ahead and remind everyone how to submit questions here? And then we can get started.

[Operator Instructions] Your first question comes from the line of Michael Yee with Jefferies.

David, can you hear me?

Yes. Go ahead, Mike.

I had a question around maybe putting this in the context of single-agent Vectibix and the data in the label and trying to compare and contrast the addition of the combination regimen. So maybe you could talk about, A, what you think -- how you think this compares to monotherapy? And then, B, if you were to run a head-to-head study, which you're planning, what you think needs to be shown to be clinically meaningfully different and better and how you would think about that?

Great. Thanks, Mike. I'll start with that, PK then can add comments as she wishes. First of all, remember that Vectibix and EGFR inhibitors are not used in KRAS-mutated colorectal cancer now because they are generally ineffective. So the Vectibix monotherapy contribution here would be considered quite minimal. So I think this is good. And then when we look at the LUMAKRAS monotherapy data, I think there's no question in our minds that we're seeing a synergistic effect here of the 2 agents. And biologically, we have some preclinical data indicating that, that could, in fact, be the case. In terms of the bar for moving forward, I think the data we're seeing are quite promising. As you know, for standardly available therapy, response rates are typically quite low, in the low single digits, with a relatively short duration of response. I think if we're able to replicate these sorts of data going forward, we would be very comfortable in the third-line and beyond setting. We'll say a little more about that study design as we get ready to roll it out. But I'm quite pleased, very pleased with the data we're seeing and how it's likely to stack up against standard of care. PK, would you like to add anything?

Yes, absolutely. So I totally agree, Dave. I think the preclinical data that we had previously looked at with the combination has really been validated in this clinical study. And we've had encouraging meetings with regulatory authorities also in terms of looking at potential beyond additive potentially synergistic effect of common -- combining sotorasib plus Vectibix. So we've always known that we are going to actually advance -- I can hear a little bit of an echo for some reason -- that we were going to advance sotorasib in combinatorial regimen with CRC. So this really reinforces our key development strategy here.

Okay, thanks.

Your next question comes from the line of Jay Olson with Oppenheimer.

Congrats on these results. Could you please remind us about the differences between Vectibix and ERBITUX? And especially any important differences when combining Vectibix with LUMAKRAS in this chemorefractory CRC setting?

Yes. I mean if you look at the data generated now over 1.5 decades with the EGFR inhibitors, they behave in a very similar fashion in the clinic. And so I wouldn't expect dramatically meaningful differences in combination with the G12C inhibitor between either Vectibix or ERBITUX. Thanks, Jay.

Your next question comes from the line of Umer Raffat with Evercore ISI.

I had 2 quick ones, both somewhat confusing, so I figured I'll ask. First is, when I look at Part 1, that green bar, the 1 responder, it says they had some nontarget lesions. But clearly, the target -- the main lesion met the criteria response. But when I flip to the following slide, Part 2, there are those 3 additional unconfirmed responses, but they also seem to have nontarget lesions. So I guess, I'm just trying to understand, were they new lesions or not? I'm trying to figure out if they could potentially become responses, or is that not going to be possible? That was first. And secondly, and this goes back to something, Dave, we've discussed in the past as well. As I look at AUC playing out between the Cohort A versus the Part 2 of CodeBreaK 100, it does move around quite a bit, at least, initially. And I'm curious where you guys stand on -- how comfortable you guys feel on whether sort of a once-daily approach truly does deliver the target AUC you guys have in mind?

Yes. Thanks, Umer. I think I can handle those questions. So as PK mentioned, in fact, all of the responses but one have now been confirmed since the data cutoff and the submission of the poster for presentation. The final patient is awaiting confirmatory scan. So you can mark most of those as confirmed, and there are additional patients who could potentially convert. Just to remind everyone, the response criteria count a specific number of target lesions. There are other nontarget lesions that are followed but don't count in the numerical assessment of tumor shrinkage. And this is a standard approach, according to the most recent RECIST criteria. In terms of target coverage, we're very confident based on the accumulated data that we've got, that getting over a particular threshold exposure for a period of time is sufficient to ensure continuous inhibition of the signaling molecule through the 24-hour dosing period. We're very confident that we're achieving that. In my mind, I think we've answered that question, and we're moving on. And we really feel quite confident with the once-daily dose.

Your next question comes from the line of Geoff Meacham with Bank of America.

Congrats on the data. Just a couple of quick ones. When you look to the Phase III in colon, just want to get your perspective about what the potential points of differentiation could be for this combo? Is it speed of response? Is it depth of response? Just trying to maybe extrapolate some of the data that you've seen, thus far, to what could play out in a larger registration. And then, more broadly, for the program, the data that you guys have on 20 -- Slide 25 and 26 with a number of different sub mutations, is that going to be what drives the combination strategies for other targeted therapies going forward? Or I'm just trying to get a sense for how you guys prioritize LUMAKRAS in different combination settings, whether it's chemo combo, IO combo or another targeted therapy combo. There's a lot to do, and so I wasn't sure what -- how you guys were able to sort of rank these in terms of strategic priority.

Yes. Thanks, Geoff. Two great questions. I'll start and again ask PK to comment. In terms of what we find promising, what we think will be important clinically in the third-line colorectal cancer setting, you asked about speed and depth of response, both of those are important. Again, if -- I think if we're able to replicate these data, we will win convincingly on response rates. But critically, for these patients, duration of response and duration of disease control is, really, probably the most important outcome measure and then, ultimately, overall survival. With standard therapy, the median progression-free survival is often measured on the order of a couple of months. And so our goal here is to really try to do much, much better than that to make it clinically meaningful impact for these patients. So with that, I'll turn things over to PK. She can comment additionally on the colorectal, but also then the mutational patterns that we're seeing and how that guides our thinking in terms of further development.

Yes. Thanks, David. I completely agree. So our colorectal strategy will be driven on those clinically meaningful endpoints, particularly overall sort of response rate and duration response. And we feel confident, based upon the data that I've just presented to you, that, that treatment duration and that the ability or the demonstration of continued confirmation of responses is something that can be encouraging towards our Phase III trial. In terms of looking at sub mutations, that's quite interesting, and I think that we will continue to review this data on an ongoing basis. But our strategy will really be driven upon the potential for potential additive or synergistic effects of combinatorial strategies in colorectal cancer as well as in other cancers. And that's why, as was presented recently this morning, we will be also looking at important combinations, for example, with SHIP2, SOS1 and others.

Great. Thanks, PK. Yes, I think it's fair to say we have not -- there are various resistance mechanisms described, but there are none that are dominant. And there is not the equivalent of the gatekeeper mutation that you often see in receptor tyrosine kinases. So the biology is clearly more complex here. And as you say, Geoff, we're sorting through that.

Your next question comes from the line of Alethia Young with Cantor Fitzgerald.

I was following on the genetic mutation. I guess, I was just curious if you guys were surprised that there seemed to be a very strong relationship here. And I guess, what does that bode for other combinations? I mean do you think there are doublets or does it mean only triplets?

Yes. I think it's probably all of the above. There are some settings where doublets may be sufficient, and there are some settings where triplets will be required. And we're looking at triplets in colorectal cancer, for example. As PK mentioned, we've got triplet combinations ongoing right now with backbone chemotherapy, LUMAKRAS and then either Vectibix or bevacizumab, which are really the standard regimens used in colorectal cancer, and will allow us to then, should those data look promising, really move into earlier lines of therapy. So thank you.

Your next question comes from the line of Kennen MacKay with RBC Capital Markets.

Congrats on the data. I was just wondering if you could talk a little bit about the de-escalation of the sotorasib dose. How many patients were de-escalated from 960 to either 720 or 480? And was that distinct from those 3 dose modifications that we saw due to TRAEs? And then just a quick follow-up question. On your comprehensive global clinical development slide, are you running distinct trials combining LUMAKRAS with pembrolizumab and your biosimilar pembrolizumab, AMG 404, or only the biosimilar?

If I can...

Go ahead, PK. Why don't you handle these questions on both on the dose reductions and then the checkpoint inhibitor combinations?

Yes. Maybe I can just get started, then you can add as much as you [ wish for ], Dave. So first of all, on the question about de-escalation. So we did start a dose level 1, which is the full dose of sotorasib and full dose of Vectibix. And since, as you saw, there were no DLTs per protocol, we deemed that they're ready for Phase II dose, and we've used that in the expansion cohort. So there was really no need to explore lower sotorasib doses. Now when I talk -- let me double-click a little bit to talk about some subjects who require dose reductions of sotorasib or panitumumab due to adverse events. So just briefly, there were a few patients, for example, who, due to, as I mentioned in the slides, skin rash, particularly, had to have some drug interruption and particularly of Vectibix. And in some cases, the drug was resumed or was resumed at a slightly lower dose. For sotorasib, we also had to have a slight discontinuation due to the diarrhea, but then we were able to resume a full dose without recurrence of that event. We also had some patients with fatigue and hypokalemia for which we had a dose reduction, but it was really small numbers of patients in these examples. So I'll just stop there, first of all, and just see if there are any additional questions.

Yes. And this is also very consistent with the single-agent experience with both molecules. So I think probably the big message here, again, Kennen, is that no real additive toxicity that we saw with the combination. In fact, we were quite pleased with the overall safety and tolerability profile.

Your next question comes from the line of Matthew Harrison with Morgan Stanley.

I was just wondering, in the combination approaches, especially as you look at other doublets and things like that, are you looking at any additional biomarkers or genomic analysis to think about how to potentially, I guess, enrich the population further? And -- or do you think it's just best to sort of look at an all-comers G12C population?

Yes, Matt, great question. Yes. No, we're doing extensive tumor profiling, molecular profiling, sequencing of tumors, really to try to further tease out predictors of response and resistance. As we've mentioned, there is no clear signature as of to date. And everyone should take away that this -- that the biology here does not perfectly mirror the biology of receptor tyrosine kinases, like EGFR, where resistance patterns can be predicted, and there are key mutations that are clear -- that have a clear ability to confer resistance. Here, it is, so far, we have seen a grab bag, but we are doing a very extensive biomarker program to further dissect these data. Thank you.

Your next question comes from the line of Geoffrey Porges with SVB Leerink.

First question, could you just give us a sense of the median duration of response that you've observed with the combination of panitumumab and sotorasib? It looks as though there are quite a few patients who are still on therapy. And then go back to the question -- I want to go back to the question on panitumumab alone. Dave, have you looked through your database to specifically look at patients in the panitumumab clinical experience who had G12C mutations? Because G12C is just a fraction of the KRAS mutations. And as you point out, pmab isn't very effective, unless monotherapy or with chemo in patients with mutated KRAS. But that's a sort of grab basket of a lot of different mutations. So how certain are you that pmab alone wouldn't have conferred better responses in the G12C subset?

Yes. A couple of things there, Geoff. In terms of median duration of response, it's too early. Because the vast majority of patients remain on therapy, and so we haven't reached median duration of response. So we'll continue to follow these patients. And speaking to our investigators, we're quite encouraged by what we're seeing in terms of disease control. We are going back and trying to tease out the specific G12C component. As you mentioned, there are about -- there are 6 or 8 different RAS mutations, broadly, that occur in colorectal cancer, including G12C, which is a small fraction. If you look at the overall response -- single-agent response rates in all-comers in colorectal cancer to Vectibix monotherapy, it's on the order of 15%, 18%. So it's -- I think it's exceedingly unlikely that there is a dramatic Vectibix monotherapy contribution here to what we're seeing. And in fact, our guess is that Vectibix monotherapy against this population would add very little. But we will take a deeper look at that and try to pull-through some of the data we've accumulated over the years.

Your next question comes from the line of Chris Raymond with Piper Sandler.

I just wanted to probe a little bit on a couple of things that you have already kind of -- you've talked about a little bit. But David, I know you've answered the durability question a little bit, but maybe if I could ask you to answer a little bit more directly. I know you've said standard of care's a couple of months. What are your KOLs saying sort of the bogey here in terms of clinically meaningful in terms of durability response for success? And then maybe a second question on dosing. I know you've talked about what you think the Phase III dose is. But you're also evaluating a 240-milligram dose in non-small cell lung cancer. Is there any plan for a lower dose sort of exploratory study in this setting as well?

Thanks. Again, on the duration, as we mentioned, the median progression-free survival is typically a couple of months in third-line and beyond therapy. So our hope would be that we would be substantially -- that we'll be able to do substantially better than that. The -- I don't want to put a stake in the ground in terms of the bar prior to any sort of regulatory discussions. But again, our -- we're not trying to add a few weeks here of an increment, thinking that's not the sort of outcome that we're shooting for. So our hope, based on the data we've seen, would be that we would be doing substantially better than that. And in terms of the second half of the question, let me just turn that to PK and see if she wants to comment.

Yes, absolutely. So thanks, Dave. I think, obviously, we are very open to ensuring that the best -- the optimal effective dose of sotorasib is being used. And we don't assume that particular doses is the best dose for colorectal versus lung cancer. So we have had discussions with the agency about dose -- identifying the appropriate dose in sort of -- for sotorasib for colorectal cancer in our Phase III trial.

Your next question comes from the line of Colin Bristow with UBS.

Congrats on the data. A couple of quick ones from me. I was just curious if you have any resistance mechanism data for the CRC patients and how this compares to the non-small cell. Is there, again, are you seeing no dominant mechanism? And then secondly, Mirati is going to have a top line update on adagrasib in second-line monotherapy lung on Monday. If they replicate the prior day to ensure a mid-40s response rate, I'm curious how you see this playing out competitively, and what you'd point to in terms of LUMAKRAS' potentially -- potential points of differentiation?

Yes. And -- so I'll take those. In terms of the resistance mechanisms in colorectal cancer, we're actually accumulating those data right now. And we're doing molecular profiling on the patients that you saw presented in today's data as well as patients that will be enrolled in the Phase III trial going forward. It is very clear that the co-mutational pattern in colorectal cancer in the G12C setting is different than lung cancer. And so one of the key questions that we will look to answer is to whether that provides insights in colon cancer into patterns of response and resistance. In terms of comparative data, I don't want to speculate, other than to say we're -- with each week that passes, I've become increasingly confident in our molecule. I would look to any data set to ensure that it's a center of view, an intention-to-treat analysis and that you get the most appropriate sort of apples-to-apples look as you think about data sets. I'm very, very confident in where we stand with a once-daily molecule and the broadest combination program in the field, and the drug that will soon be available in multiple jurisdictions around the world. So a very large global program.

Your next question comes from the line of Robyn Karnauskas with Truist.

Just first, how quickly do you see these responses? I know people have asked like how many more you think will go into a response. I'm just curious about like how deep do you think they are. And just maybe a quick comment on your announcement of the BI collaboration for the SOS1 pan-KRAS. What have you seen in preclinical models? What kind of responses -- what do you expect to see when you put that combination into the clinic?

Yes. I'll handle these questions, Robyn. Thank you. As I mentioned, all but one of the responses were confirmed. And over time, we'll look to potential deepening of those responses. That's a pattern that we've seen in our monotherapy -- in the monotherapy setting, both in lung cancer and colorectal cancer, that these responses can deepen over time. So obviously, that's a key question to ask as we go forward, but we're quite encouraged by what we're seeing at this point. In terms of the SOS1 combination, given its position in the signaling cascade, there's good biologic rationale here. And so we're quite interested in this combination and really getting it going in the clinic and seeing what sort of efficacy data that we can produce and so on. So thank you.

Your next question will come from the line of Dane Leone with Raymond James.

Congrats on the data. It looks like you guys cleared what the clinical community was looking for with this combination on both efficacy and toxicity. So congrats on that. That's a great step forward. Just a few details here that I think people are trying to get clarified. Firstly, what was the data cut actually for the poster? And then when you're referencing the point when the unconfirmed responses have now been confirmed at that data point, can you just clarify how many responses are ongoing? I know when we look at the poster, it looks like one response wasn't ongoing as of that data cut. So I'm just curious if you're referencing the post data cut for that, confirming more responses. How many are currently, today, since the last time you've looked, ongoing?

So I can take that. So one is the data cutoff was August 6, and apologies for not having that on the poster. In terms of the confirmation, that's pretty much late-breaking information. I'll have to go back and confirm with the team how many are ongoing.

Yes. But based on our discussions with investigators, the large majority are ongoing, and we'll provide updated data in the coming months as we got longer follow-up there. But we've been quite encouraged by the duration, overall, Dane.

Your final question comes from the line of Michael Schmidt with Guggenheim.

I just had one on your development strategy for LUMAKRAS in colorectal, and perhaps if you could speak to your decision to pursue a Phase III study with this combination in a third-line setting. Your competitor, Mirati, is obviously doing a similar study in the second-line indication. And perhaps comment on what your plans might be in the [indiscernible] of colorectal cancer for this combination?

Yes, sure. As I mentioned, we've got triplet combinations ongoing now with standard regimens, backbone chemotherapy regimens, including FOLFOX and FOLFIRI, which are, by far and away, the most commonly used combination regimens, either in combination with Vectibix or MVASI, bevacizumab or bevacizumab biosimilars. So we really are covering, I think, all of the major regimens. And those would be used to advance into earlier lines of therapy. So thank you for the question, Mike. All right, Christie, if that's the end of the queue, we can wrap up. Any final questions?

There are no further questions at this time. Do you have any closing remarks?

No. I'd like to thank everyone for getting together. In summary, we're quite pleased with the data today, not only the overall response rate, but what we're seeing preliminarily on duration, and certainly, tolerability of the combination has been quite good, moving us to move quickly into a Phase III pivotal trial. And then I think we've got some nice progress in the other programs that I mentioned: bemarituzumab moving forward into its pivotal phase, Phase III trials as well as signal-seeking trials; and then the AMG 757 program, tarlatamab, moving into a potentially pivotal Phase II trial across a couple of doses; as well as a solid tumor BiTE moving into a registration path. So thanks very much. The Investor Relations team will be available for additional questions, as always, if you've got them. And we look forward to talking to you again. Thank you.

Ladies and gentlemen, thank you for participating in Amgen's Conference Call from the European Society for Medical Oncology 2021 Virtual Congress. You may disconnect at this time.