Amgen Inc. (AMGN) Earnings Call Transcript & Summary

My name is Jason, and I will be your conference facilitator today for Amgen's conference call from the 2022 World Conference on Lung Cancer. [Operator Instructions]. I would now like to introduce Arvind Sood, Vice President of Investor Relations. Mr. Sood, you may now begin.

Actually, Jason, this is David Reese, Head of Research and Development at Amgen. I'll be helping to host today's conference. We'll provide an oncology clinical update on some of our lung cancer programs, based on data that were presented here in Vienna at the World Conference on Lung Cancer over the last couple of days. We know there's a lot of interest in these data, and so we'll move into data summaries very quickly here. Can we advance the slide, please? Here is our safe harbor statement and today's agenda. After very short remarks from me, Dr. Bob Li will present an update on the LUMAKRAS combination studies presented here, including both checkpoint inhibitor combinations as well as the SHP2 combination trial. Following that, Dr. Luis Paz Ares will present data, very promising data, we believe, from Tarlatamab, our DLL3 targeting BiTE from the first-in-human study; that followed by Dr. Jean-Charles Soria, our Head of Oncology Clinical Development, who will provide some concluding remarks, and then we'll move into the Q&A session. As I mentioned, we have quite a bit of data to cover today, and we know there'll be many questions. So I'll turn things over now to Dr. Li to summarize the LUMAKRAS combination studies. Dr. Li?

Thank you, Dave, for the introduction and the opportunity to share with you the 2 abstracts on the sotorasib combinations. First one is in combination with PD-1 or anti-PD-L1 immunotherapies. Next slide, please. So combining sotorasib with immunotherapy has been very encouraging in preclinical models showing the potential synergy in increasing CD8-positive T-cell infiltration in the tumor cells and in mouse models. It's led to dramatic tumor shrinkage and durable responses. And this is the clinical trial that we brought it into the clinic to see if this approach can help improve the care of patients. So in this CodeBreaK 100/101 study, we've had a flexible dose finding design. We know that historically, that combining targeted therapy and immunotherapy has been challenging. So we allowed escalation also deescalation of sotorasib doses from 960 down to 120 milligrams as clinically appropriate. We also allowed the combination with the anti-PD-1 pembrolizumab or anti-PD-L1, atezolizumab, a both in the concurrent regimen when you just combine the 2 drugs together or using a leading regimen where you can give sotorasib for 3 weeks or for 6 weeks as a single agent and then followed by combination therapy with atezolizumab or pembrolizumab. We primarily looked at safety in this Phase I study, but also secondary endpoints, including overall response rate, duration of response, disease control rate and so forth. And this is an analysis of data after a long period of follow-up with a median of 12.8 months. Next slide, please. So we treated a total of 58 patients. And in this study, more than 2/3 of patients had already undergone anti-PD-1 or PD-L1 immunotherapy prior to study entry, so a lot of pretreated patients. But there were some immunotherapy naive patients as well. As you can see, median -- there was a median prior line of therapy of 1, but up to 7 prior lines of therapy. And the PD-L1 expression range is -- are also detailed on this table. Next slide. So first, we combined with the high doses of sotorasib at 720 and 960 milligrams. As you can see, there certainly was a high incidence of treatment-related adverse events, primarily asymptomatic elevations of liver enzymes, AST and ALT, and this is -- these were reversible side effects with dose modification and corticosteroids. But certainly, from a drug development point of view, this is not a tenable combination for further development. So that we -- those reduced it down to 360 and down to 120. And you can already see numerically a reduction of these elevated liver enzymes with lower doses. So that was the first experience. Next slide, please. We then looked at the lead-in versus the concurrent regimens. So whether it's atezolizumab or pembrolizumab, you can see that the leading regimen had a lower risk of treatment-related adverse events in Grade 3 and 4, especially the elevated liver enzymes. In this experience, we saw about these hepatotoxicity events occurred primarily outside of the dose-limiting toxicity window of the first 3 weeks of the combination. It actually happened on cycle 2 and cycle 3. It took time for that to happen. But the 97% of those events were resolved with corticosteroids, treatment discontinuation modification. And those were largely asymptomatic elevations of liver enzymes. This occurred in patients with immunotherapy naive as well as immunotherapy pretreated settings. Next slide, please. I want to also point out there were no treatment-related deaths, and there were no Grade 5 events throughout the whole study. So given this experience, we then looked at the leading of sotorasib in lower doses in combination with pembrolizumab. As you can see here, this is a better toxicity profile compared to when we started. If you look at sotorasib 240-milligram dose, there was still about 20% rate of Grade 3 or more elevated liver enzymes, but certainly much better overall safety data and tolerability for patients. Next slide. This is the overall waterfall plot, but this is a mixed bag of different cohorts and sub cohorts but overall showing a response rate of 29%. But bear in mind, many of those patients were treated with earlier cohorts and could not have to discontinue the study drug. So didn't have enough time to get drug exposure. But overall, we saw a median depth of response of 51% and a disease control rate of 83%. If you look at the swimmer plot on the right-hand side. Of the 17 responders, the median duration of response was 17.9 months. So this is encouraging as it's longer than what's reported with a single agent. And 8 out of the 17 patients are still responding at the data cutoff. These responses were similar in immunotherapy naive or pretreated patients. Next slide, please. This is just a spider plot showing the cohort of patients with sotorasib leading and then a combination with pembrolizumab. You can see these durable responses were observed in this plot. And 3 patients out of the 7 who confirmed response are still going on at the time of the data cutoff and still having able to receive the combination of sotorasib and pembrolizumab. So those were tolerable and durable for these patients. Next slide. So in conclusion, in this mostly immunotherapy pretreated patient population but also 1/3 patients with immunotherapy naive, sotorasib in combination with atezolizumab or pembrolizumab led to a higher incidence of Grade 3 to 4 treatment-related adverse events compared to what we know from monotherapy experience. Now those were primarily asymptomatic elevation of liver enzymes, and they occurred outside of dose-limiting toxicity window, so usually cycle 2, cycle 3. And nearly all of them were resolved with corticosteroid, treatment modification and drug discontinuation. There were no Grade 5 toxicity, no treatment-related deaths. And we also learned that there are mitigating strategies that were effective, such as leading cohorts, leading sotorasib, especially at lower doses that were then combined with the immunotherapy that led to durable clinical activity and deep responses and some of them are still ongoing. And the median duration of response of 17.9 months gives us a glimpse of encouraging future. And due to this experience, a low-dose sotorasib as leading, followed by combination with pembrolizumab is now being further studied as a potential first-line patient -- first-line treatment for these patients. Next slide. I will then now turn your attention to the sotorasib and SHP2 combination experience. Next slide. So the genomic operations in RTK pathway is actually being identified as a putative mechanism of resistance to sotorasib. In the resistance study that I presented at ASCO just a couple of months ago, more than half of the acquired genomic alterations in sotorasib resistance actually belong to the RTK pathway. And also in mouse xenograph models we showed -- that we saw that combination with sotorasib with a SHP2 inhibitor, impaired RTK signaling to RAS, and basically, it also prevents the activation of the GTP loading of the KRAS protein and therefore shut down the signaling and enhanced antitumor efficacy. So we, based on this science, we set out to do a Phase I trial combining sotorasib with RMC-4630, which is a small molecule SHP2 inhibitor and primarily focusing on non-small cell lung cancer, but we did do a cast of wide net looking at a variety of solid tumors. Next slide. So this is a study design. It's pretty simple. Those are patients who underwent approved treatment and generally anti-PD-1, PD-L1 or platinum-based chemotherapy already. And we even allowed prior KRAS G12C inhibitor. So sotorasib as a single agent, for example, and those have progressed after that, we're able to enter this study. So this is a dose finding study. We started with 100 milligrams of RMC-4630. And due to the longer half-life, the dosing is 2 days every week. So day 1, day 2, every 7 days, or you could do day 1, day 4 every 7 days. And we escalated to 140-milligram and 200 milligram as well for the RMC-4630. For sotorasib, we gave the FDA approved dose of 960 milligram daily. And in this study, we looked at safety as a primary endpoint, but also secondary endpoints, of course, very keen to look at our overall response rate, duration of response and so forth as an early look into potential efficacy. Next slide, please. So in this study, we treated a total of 27 patients across a range of solid tumors and 11 of those had non-small cell lung cancers that were KRAS G12C inhibitor. As you saw in the 11 patients, they were pretty heavily pretreated, 3 prior lines of therapy as a median, up to 6 prior lines. And 5 of the 11 have previously undergone KRAS G12C inhibitor therapy. So a lot of pretreated patients. Next slide. This is the safety profile. In terms of the treatment-related adverse events, there were no surprises. We know that edema and diarrhea were cause effects of SHP2 inhibition, and that's primarily what we saw. And overall, it's been well tolerated. None of them led to drive discontinuation and they were managed, some of them with those reductions. But overall, there's no treatment-related deaths or anything unusual. Next slide. So looking at the efficacy of patients -- in patients with non-small cell lung cancer. This is the 11 patients. In all the 11 patients, there were 3 partial responses. So that's a 27% response rate in this early cutoff. But looking at the KRAS inhibitor naive patients, all 3 actually belong to that group. So that's of 6 patients, 3 patients responded, 3 out of 6, so 15% in a very small subgroup. But disease control was achieved in 7 of the 11 patients, and there were certainly, this suggests that in the KRAS G12C inhibitor naive settings, the effects were more pronounced and looking promising. And 2 of the 3 patients who had partial response had on -- is still having a response at the time of the data cutoff. Next slide. So in conclusion, there's preclinical rationale and putative mechanism of resistance supporting the combination of sotorasib with the SHP2 inhibitor. And this is the first-in-human study of sotorasib plus SHP2 inhibitor, RMC-4630 appearing safe, tolerable in these patients, heavily pretreated with chemotherapy, immunotherapy and many of them with prior G12C inhibitors. But in this experience, no Grade 4 and no fatal treatment-related adverse events we've seen. And there were very few treatment-related dose drug discontinuations. So we saw promising and durable clinical activity in the p.KRAS G12C-inhibitor naive setting in non-small cell lung cancer. And based on this finding, we -- the Phase II study of sotorasib in combination with RMC-4630 is now ongoing. And we hope to see more data as time goes on. And this is looking specifically at the G12C inhibitor naive patient population. I'll now turn the floor to Dr. Luis Paz Ares, who will tell you more about the Tarlatamab naive experience. Thank you.

Okay. So I'm Luis Paz Ares and I'm always happy to give you an update on the Tarlatamab development. So we can go to the next slide, please. So small cell lung cancer is accounting for some 15% of the lung cancers, there's one of the most aggressive solid tumors. That typically happens in patients with heavy smoking history, and has a very aggressive biology. Indeed, some 70% of the cases are metastatic and these are not diagnosis. This disease is typically responsive to chemotherapy. That typically happens with tumors that are highly preemptive as response to chemotherapy it only happens in cells that are replicating all the time. However, responses are short-lived, are very transient, and therefore, the prognosis on those patients is quite good. Let's say, for those patients with extensive disease, which accounts on 70% of the cases, as we said before, median survival is in the rate of a year and survivorship long term is happening very, very well. Importantly, as very few advances in the treatment of the small cell lung cancer over the last 2 or 3 decades. Now first line, we use chemotherapy, platinum plus etoposide and a PD-L1 inhibitor. The PD-L1 inhibitor really improved survival, but truly very modestly. Median survival is improved by 12.5 months. And that impacts in about 10% to 15% of the patients in the long term. And in second line, we have very few advances. We are still using mainly drugs that were approved more than 20 years ago. Recently, we have the approval -- the conditionated approval by FDA of Tarlatamab, but there is no data that show a clear improvement in the outcome is mainly a safety benefit, I would say. So if we go to the next slide, you will see here the basis for the use of BiTE and in that case, Tarlatamab in the treatment of small cell lung cancer. BiTE molecules actually are more tumors have engaged the old T cells of the host of the patient to attack and eradicate the tumor cells. In that case, the Tarlatamab has, I mean is an anti-DLL3 that binds to DLL3, which is typically ligand spreads in the cancer cells, in that case, most of the small cell lung cancer cells expressed in the surface of the cell, this molecule, DLL3, which is typically very -- has very little expression in normal cells. And when it happens, it doesn't happen typically in the surface. So that makes that is a very nice target for this carrying of our drugs. On the other hand, the Tarlatamab molecule has binding space for CD3 that is engaging the T-cell by putting together the 2 cells. The T-cell gets activated and therefore, inducing cell killing by getting the tumor cells to get into apoptosis. Importantly, the Tarlatamab has Fc domain to improve the pharmacokinetic of the drug to extend half life. Go to the next please. After the drug has been shown activity in preclinical models, we have started the first-in-human sometime a goal actually. And we are here presenting, we just presented at the World Lung Cancer Conference, the updated, the most updated data on some more than 100 patients with a meaningful life of 8.5 months of this Phase I trial. Phase I trial where we started with a very low dose of 0.003 milligrams, and we have gone into the highest dose of 100 milligrams. And actually, we have an expansion cohort with 100 milligrams. As you may imagine, this is a Phase I trial first-in-human. Primary objective was to evaluate the safety and tolerability in that particular tumors in small cell lung cancers induced to drive. And of course, to get an entity and a recommended dose for further study in Phase II trials. As a secondary objective, we have to characterize the pharmacokinetics and evaluate preliminary the efficacy, the antitumor activity of the compound. And of course, we also like to evaluate immunogenicity of Tarlatamab and assessed biomarkers in that particular case. If we can go to the next slide, please. Some 106 patients I told you have been included. Inclusion criteria were patients with small cell lung cancer that had confirmed diagnosis and progressed to at least 1 line of therapy, including platinum etoposide. And in the case that when the patient is treated, the anti-PD-L1 was approved the patients who have received as well the PD-L1 therapy. Indeed, 60% of the standards had not availed with that particular therapy at the time those patients were reported. Patients who have good PFS and had measurable disease. And patients were allowed to have brain metastases if they were asymptomatic or they had been treated. As you see here, no prior therapy have been permitted over the last 4 weeks and the patients should have HIV or interstitial disease. Basic characteristics are typical for those type of trials in the relapsing setting. As you see here, median age is 64 years. Good PFS for most of the cases some half of the basin has had 2 lines of therapy. Some 1/3 of the patients have had 3 or more prior lines of therapy and some 50% of patients -- actually 49% have had not only chemotherapy but also PD-L1 treatment this is co-related here. And as you see here some 27% of the patients had brain metastases and 51% of the patients hepatic disease, hepatic metastases at the time of being included into the diseases type. We'll go to the next, please. In terms of safety, I can tell you the profile is quite valuable as compared particularly to what is expected from BiTEs. Now we have some experience with BiTEs in hematological malignancies, but as some other type of cancer. And I would say the profile of this specific BiTE Tarlatamab is available. We have seen, of course, some side effects, but in terms of cytokine release symptom, which is, let's say, one of the more, I would say, toxicities we were somehow nervous about. I would say, half of the patients have had some side effects of that, but only in 1 case was Grade 3 and no cases of Grade 4 or Grade 5. For most of the cases, cytokine release syndrome was either sometimes of mild hybridation, very responsive to antibiotics. Some fluid sometimes, in some cases, dexamethasone, let's say, steroids, only 8 patients were given tocilizumab, which is a specific therapy, we give for the patients that are a Grade 2 or Grade 3 of the cytokine release symptom. So I would say basically to be so. And as I said, in most of the cases were restricted to Phase I. In terms of neurological events, we have seen typically mild, Grade 1 headache, nausea, some 50% of the cases, very few cases are Grade 3 and those typically happen in patients that have a prior radiation to the brain. We have had some cases of neutropenia, including 9% Grade 3. We are not very clear on the mechanism. The only thing I can tell you, no patients have had fever. So we didn't see patients of the important, the clinical relevant publication, which is neutropenic fever. We go to the next, please. This is the activity we have seen, which I would say is pretty remarkable. We have responses confirmed in some 23% of the patients, 2 CRs, 22 PRs, 37% of the patients have had tumor shrinkage of at least 30% of the volume, I would say. And disease control rate, which means responses plus disease stabilization we're seeing in some 52% of the patients. So I would say a reasonable and encouraging response rate in this setting that I said is in patients with heavily treatment before entering here. We'll go to the next, please. Duration of the response is pretty remarkable, remember that some 24 patients had confirmed responses. That response was typically identified very shortly shopping after treatment was started less than 2 months; median time for response was 1.8 months, and median duration was 13 months. And still, of course, we are following up those patients as some half of the patients are still responding, 11 patients are still responding at the time of the last data cutoff. Go to the next, please. More importantly, maybe also the impact on survival. You see here survival in the semi-treatment population is about 13.2 months median survival. And of course, this cohort should mature still on time. But I would say the relevant duration of the response seems to impact on the survival of this patient growth. Go to the next, please. So just let me to finalize, to conclude, I think this first DLL3-targeted immune therapy, the first DLL3 BiTE that show a clinical evaluation in Tarlatamab has shown promising activity with remarkable responses duration, 23% of the patients responding with a median duration of 13 months, and importantly, impacting on survival 13.2 months. We have seen over the last decade a number of the drugs with responses in the rate plan to even 35%, 40% in some cases. But median survival always in this relapsing setting has had been in the range of 8 to 9 months. And that is the more remarkable data for me, the impact of survival. Of course, that is now personal opinion. The safety profile is pretty acceptable, particularly with each account. The nature of this immunological type of drag and potential registration of Phase II studies already undergoing in the second, on prior line of study that patients with response time. So I think that's my last slide. I will be happy to take any questions that you may have later. So I'm just passing the microphone to Dr. Soria.

Thank you very much, Luis. If I can have the next slide. So we remain confident on LUMAKRAS future. This is the KRAS G12C inhibitor, with the largest and broadest clinical program. We are exploring LUMAKRAS combination with over 10 different partners as well as in various tumor types, including non-small cell lung cancer, colorectal and pancreatic cancer, and we are currently discussing different potential paths to first-line non-small cell lung cancer with LUMAKRAS. Next slide. I would like here to highlight our excitement and confidence on Tarlatamab as a potentially transformative therapy for small cell lung cancer, as well as other tumor types, including neuroendocrine prostate cancer. If we respect on small cell lung cancer, this is truly a disease that while chemo-sensitive as highlighted by Dr. Paz Ares remains a disease with very poor survival, very few patients survive beyond 1 year. And it's a significant number of patients up to 70,000 patients in first line and relapsed who could be addressable and treated and benefit from this T-cell engager. Put into context the data we have shared to you is truly transformative. I mean, patients who are in third line small cell lung cancer usually have a median survival of 5 months and response rate is below 2 digits. Here, we have a confirmed response rate of 23%, sustained with a median duration of 13 months and a median overall survival of 13.2 months. This is why we are now exploring the activity of this compound even outside of small cell lung cancer and notably in neuroendocrine prostate cancer, which is the phenotype that usually appears on patients who develop resistance to all monotherapies and chemotherapies. So these 2 molecules, sotorasib and Tarlatamab, give us confidence on our oncology thoracic portfolio. With that, I pass the mic to Dave Reese.

Thanks, Jean-Charles, and thanks to all of our presenters. I know we have a lot of interest. So Jason, could you remind everyone how to ask questions and then we'll go ahead and open up the line.

[Operator Instructions] Our first question is from Matthew Harrison with Morgan Stanley.

I guess it would be helpful to understand in sotorasib and combination, I guess 2 parts. So first, why you're expecting or what kind of efficacy benefit you're expecting to see when you move into a true frontline population who hasn't been exposed to prior G12C inhibitors. And then secondly, why you're confident that the lower dose with the lead-in wouldn't degrade maybe that efficacy advantage.

Thanks, Matt, I'll ask Bob to weigh in here in a minute. I think as Dr. Li noted in his presentation, one of the keys is determining whether there is a clear therapeutic window with the lead-in low-dose approach, if there is, and then you're able to get the combination in, we would expect enhanced efficacy based on what we've seen. But that I think is the hypothesis we need to test going forward. Dr. Li, it would be great to get your perspective here as well.

Thanks, Dave. I agree. We hope to do better for patients. So in the first-line setting, we hope to really maximize the duration of benefit with these targeted drugs. And in the first-line setting with standard treatment with chemotherapy or chemo immunotherapy, it's just not good enough. And the median progression of free survival being 9 months, approximately that. And with KRAS mutant patients still being an unmet need, and this was still highlighted at the recent FDA presentation at ASCO, we hope to really improve the duration of benefit. That's in progression-free survival, duration of response endpoints. And we are seeing a glimmer of hope in this Phase I study of this combination, seeing a longer duration of clinical benefit in some patients. And we have -- we highlighted in the World Lung presentation, one of my patients who had a CR going on more than 2 years, and she still continues. She had a lot of initial liver toxicity with the infiltration of immune cells into the liver. But in the long run, she's doing well. So we really hope to translate that benefit in the first-line setting. But you raised a very good question. How confident are we that the toxicity is not going to impede on the response, on the durability of response that we did see in the initial cohort experience. That certainly was not a no-go with the high doses. We did see numerically a dramatic reduction of Grade 3 or 4 treatment-related toxicity with liver transaminitis in the leading cohorts. I suppose I don't have that crystal ball on a larger population of patients on the trial with cohort expansion. What are we going to do is that going to translate to a much better first line treatment for a patient. We have certainly hope but we have to do the trial, and that's what we're doing with the current expansion.

Our next question comes from Michael Yee with Jefferies.

We had a follow-up question, of course, the former question was on efficacy and then also on safety, how you feel about the onset of the liver tox. If you look at that case study, there's liver tox early on, but then later on at 5 and 6 months, there's another sort of spike. I was just wondering if you can comment on either the mechanism or the time of onset I know does it matter if it's sequential or not? And I know you had some data on that at, I think, 50 to 70 days. So maybe just talk about the onset and maybe the mechanism upon when it comes on.

Yes. Dr. Li to comment on that again in just a moment. Yes, clearly, one of the things that's quite clear is that the onset typically with outside of the dose-limiting toxicity window in the large majority of cases, and in some case, could be a late onset perhaps due to some sort of immune timing, the mechanism remains unclear to us right now. But this is why I think adequate numbers of patients followed for an adequate duration is really essential to understand the profile of these regimens. Bob, it would be great to get your perspective here as well.

Yes, thanks, Dave, I agree. This actually came as a surprise because when you do Phase I trials, we set a dose-limiting toxicity window for purpose because that's when you're going to see the adverse events happen. And if they pass that window, you're fine. The next cohort of patients can enroll -- but in this case, we didn't see any dose-limiting toxicities. The DLT window is completely fine. The patient went through it, maybe with Wave 1 liver enzyme elevation that we saw with the single agent. But by the time they get to dose number 2 -- cycle 2 and cycle 3 definite things went off in the case report that you alluded to in the presentation, it happened after the second dose and then the third dose, and then it just went sky high with the transaminitis. So it's important not to just call it a day at the end of the DLT window and say, look, it's a safe combination. We really have to have long follow-up period in this study, and that's why we only presented the data after a median of 12 months a follow-up. It's not good enough to just say, okay, DLT, we passed, we said that's not good enough. And it's the immune infiltration and the immune toxicity are somewhat mysterious. And we need to do the trial to approve it. And we saw Grade 1 in the DLT in that case. And then by the second dose, it went to Grade 3. We have to immediately stop the drug. Steroids, it reverts back to normal, completely normal. Okay, so okay, maybe we can just rechallenge. We challenged with a slightly lower dosing, the 960 going down to 720. And then it went up again the liver enzyme went straight up, as you saw in the graph that we alluded to, and then we have to back off again. And we went down, went down and eventually 240 and the patient is still taking 240. This is more than 2 years on with the CR. So this is a learning experience for all of us, and we have to really do the due diligence and do the trial expanded and to observe carefully and take care of these patients. So then we can confidently say what is the right strategy for first line.

Right. Thank you. And it's worth pointing out that this is not the first time that delayed toxicity has been reported for checkpoint inhibitor combinations, particularly with targeted therapies. So a lot to learn here, and we're going to really explore these things in the ongoing dose expansion.

Our next question comes from Jay Olson with Oppenheimer.

Maybe for David Reese, if you could please talk about the feedback that you've received from KOLs on these data presentations at World Lung? And then specifically for Tarlatamab, given the results you presented today in the context of the poor prognosis and limited treatment options available for patients with small cell lung cancer, can you talk about the regulatory pathway and where the bar is set for an accelerated approval?

Yes. Sure, yes. Thanks, Jay. Yes, I mean in terms of the feedback, I think in the LUMAKRAS combination the checkpoint, the sense is that we've seen this before with multiple targeted therapies, combining with checkpoint inhibitors. The mechanisms remain poorly understood. I think there's a general sense it's worth attempting the lead in the lower dose followed by then the sequential use of a checkpoint inhibitor. But I think the general sense is that it's an open question. In terms of Tarlatamab, I've had a chance to talk to quite a number of investigators, and of course, we have Dr. Paz Ares here as well, very, very enthusiastic. In fact, our sites -- they're looking for more thoughts on the trial. Many of the investigators are proposing investigator-sponsored studies. There was just, I think, some real excitement here, and the general notion that this drug has a real shot at changing the natural history of this disease. So this is a molecule that we're very keen on. And as Luis presented, the safety profile has been quite nice so far. And so we're moving forward. In terms of the regulatory pathway, we have a potentially registrational Phase II ongoing right now. And then we've got trials open in both the second line and in the first line with standard of care therapy. So our goal is to attempt to develop this drug across lines of therapy, again, to really change the natural history. So Luis, I don't know if you want to add anything to that.

I think I was on mute, sorry about that. So I cannot hear that, I mean, the enthusiasm is pretty clear. I mean, I see everyone in the context of small cell lung cancer really like to do trials with drug in the relapse setting, first-line setting. And they are, of course, like competitive groups, academic groups. We are thinking about different proposals to maybe try to convene something to get there or to help us to get them. Secondly, it's not only doctors, it's also patients. I'm telling you my experience. I was the only surgeon in the country, I was receiving patients with small cell lung cancer for from everywhere in Spain. Every week in our teleconference doctors were fighting for the slots to include their patients because there was really assignment because of the early data that has been released over the next -- the last 2 years in this. So I guess, also the patients on their circuits, on their patients, advocacy groups they are really enthusiastic about having a drug in a setting where it's so desperately needed such as in small cell lung cancer. This is with terrible prognosis where the improvements over the last 2 or 3 decades has not been as good as in non-small cell lung cancer. So I don't -- I think here, doctors and patients are in the same page in terms of really having high hopes with that particular drug.

Our next question comes from Salveen Richter with Goldman Sachs.

You're planning to initiate a Phase III study of LUMAKRAS in chemo and frontline PD-L1 negative non-small cell lung cancer and then working on approaches for PD-L1 high and low expressers as well. Can you just discuss what you've seen here and give us a sketch of your strategy?

Yes. Thanks, Salveen. That's a very good question. And one thing that we like to do is think of patients according to their molecular tumor profile with non-small cell lung cancer for these purposes in kind of 3 big buckets, those who have PD-L1 negative tumors, those that are PD-L1 low to intermediate and those that are PD-L1 high. We've generated what we think is very interesting preliminary data in using LUMAKRAS in combination with standard chemotherapy. And as you know, we'll be moving forward with a Phase III study in patients with PD-L1 negative tumors where the addition of checkpoint inhibitors provide the benefit, but it's fairly modest. And we will provide guidance as the year goes on in terms of our specifics of that study design and the launch timing and expected data timing. We've had very good discussions with regulators regarding that trial. For the remaining patients who have some PD-L1 expression, of course, we're going to pursue the LUMAKRAS checkpoint inhibitor combination, as described today. And then the question is whether some of the other combinations as we generate experience such as the SHP2 combination, generates the sort of efficacy data that would indicate that, that is worth exploring in the first line as well. So I think many avenues, but it's very likely the therapies are going to be customized according to the tumor profile.

Our next question comes from Evan Seigerman with BMO Capital Markets.

Congrats on the data update. So while we have data on IO naive and IO screening patients in the trial that you presented, do you have any data kind of looking at the difference in TPS? I'm specifically talking about doses greater than 50% and doses 1% to 49%. And then was there any noticeable differences between PD-1 inhibition and PD-L1 inhibition?

Yes. I'll ask Bob again to comment here in a minute. I would say, Evan, that once we start upsetting the data, the numbers become very, very small. So I think I would be hesitant to draw conclusions about distinct differences between the immunotherapy naive versus exposed population. Bob, why don't we get your perspective here. But to me, it's just a little hard to say right now.

I agree with Dave. It's subgroup of subgroup analysis. We're coming to sort of comparing 1 of 2 versus 2 of 3. So really no way of making any statistical comparison. So we need to just expand the trial and do a proper analysis with beta sample size. Thanks.

Our next question comes from Umer Raffat with Evercore ISI.

Candidly, I'm a little confused today, so I'll ask 2, if I may. First, at the very start of the presentation, it says median duration of follow-up is 12.8 months. But then when I look at the slide on how long was the median duration on the combination, it's only 0.7 to 2.5 months. And even the median duration of sotorasib is like, I don't know, 3 to 5 months, depending on which 1 you look at. So I was really confused sort of the 12.8 months of follow-up versus the duration of combination? And then secondly, as I think about sort of the reason to do this IO plus KRAS combination is ideally for perhaps moving upstream, but also for incremental efficacy reasons, and I know the ORR at your lower doses from Phase I data was tracking right around 25% to 30% based on the early disclosures that were done as of 2020. And it's kind of in the same ballpark as well as reported for this data set here. And I'm just trying to figure out how to interpret those 2.

Yes, thanks, Umer. I'll ask Bob to comment here on the first part of your question where he's taken quite a bit of time to look at these data. In addition to response rate, and many of the patients came off study drug quite early. So that's 1 thing to keep in mind. Of those who are able to continue the duration of response is 1 thing that we're looking at, as Dr. Li highlighted in his presentation. And I think that's one of the key things that we will look at in the expansion. Bob, do you want to go ahead and provide some perspective here?