AstraZeneca PLC (AZN) Earnings Call Transcript & Summary

Welcome, ladies and gentlemen, to AstraZeneca's investor conference call, ASCO GI 2022. Before I hand over to AstraZeneca, I'd like to read the safe harbor statement. The company intends to utilize the safe harbor provisions of the United States Private Securities Litigation Reform Act of 1995. Participants on this call may make forward-looking statements with respect to the operations and financial performance of AstraZeneca. Although we believe our expectations are based on reasonable assumptions, by their very nature, forward-looking statements involve risks and uncertainties and may be influenced by factors that could cause actual results to differ materially from those expressed or implied by these forward-looking statements. Any forward-looking statements made on this call reflect the knowledge and information available at the time of this call. The company undertakes no obligation to update forward-looking statements. Please also carefully review the forward-looking statements disclaimer in the slide deck that accompanies this conference call and webcast. There will be an opportunity to ask questions after today's presentations. [Operator Instructions] And with that, I will now hand the call over to AstraZeneca.

Hello, and welcome, everyone. My name is Josie Afolabi in the Investor Relations team at AstraZeneca. We're delighted to be hosting this investor and analyst science conference call, where we'll discuss key data presented at this year's ASCO GI Symposium. Please move to Slide 3, next slide, please. As part of our panel speakers today, we have Dr. Bruno Sangro, investigator on the HIMALAYA trial, Professor and Head of the Liver Unit and Hepatobiliary Oncology Area at the Clinica Universidad de Navarra; as well as Dr. Arndt Vogel, Steering Committee Member for the TOPAZ-1 Trial and Professor for the Department of Gastroenterology, Hepatology and Endocrinology at the Hannover Medical School. I'm also joined by AstraZeneca colleagues who will be presenting this evening: Susan Galbraith, EVP of Oncology R&D; and Dave Fredrickson, EVP of Oncology Business Unit. For Q&A, we will be joined by Cristian Massacesi, the Chief Medical Officer and Oncology Chief Development Officer; as well as Niko Andre, Global Franchise Head of the Immuno-oncology and Hematology. If you could move to Slide 4, please. Here is the agenda for today's call. Susan will start with an overview of ASCO GI 2022; followed by Dr. Vogel, who will take us through TOPAZ-1. Dr. Sangro will go through the HIMALAYA data, which will be followed by Dave giving commercial context. And finally, before we get to Q&A, which will be chaired by Dave, Susan will take us through for what is next for AstraZeneca as we enter into the gastrointestinal space. And now moving to the next slide, I will now pass you over to Susan.

Thank you, Josie, and welcome. Appreciate your time. Please go to the next slide, Slide 6. So this slide gives you an overview of the strategic investments in our oncology portfolio: on the left-hand side, obviously, our approved drugs; on the right-hand side, the pipeline of mechanisms that we're investing in. The left-hand side of this wheel focuses on those that are focused on killing cancer cells by activating the immune system, of which the immune checkpoint inhibitors are a key part; on the right-hand side, the different mechanisms that we're investing in that kill cancer cells directly. What I would say is the immuno-oncology portfolio that we have is a critical component of building really effective combination therapies that we aim to use as early as possible in the treatments of disease to improve long-term survival benefit. So with that as context, if you can turn to Slide 7, I just want to talk a little bit of an overview of the presentations that we had at ASCO GI 2022, and it was a pleasure to be able to be at the Congress in person this year. We had 21 abstracts with 4 oral presentations: 2 proffered oral papers that we'll go into more detail. Obviously, the highlights being the HIMALAYA and the TOPAZ-1 data that we're going to share with you shortly, and we also had data from Study 22. And then Enhertu, we had updates on DESTINY-Gastric studies and the Destiny colorectal study in addition. We go to Slide 8, please. So before we go into the presentation of the data, I just wanted to point out that the key Phase III trials that we're presenting today were presaged by careful evaluation of these assets within Phase II context. So on the left-hand side, we have the data from Study 22, which was a randomized study in hepatocellular cancer, which looked at different schedules of the combination of tremelimumab and durvalumab. And it was on the basis of both the biomarker and the safety and efficacy data from this study that led us to go forward with what is now called the STRIDE regimen of a high single dose of tremelimumab at 300, plus durvalumab. What you can see on the left-hand side of the figure is that you can look at both CD4-positive cells, which is what's called Canon 1; and CD8-positive cells, Canon 2. You see high levels of both that have also a high Ki67 proliferation index, which is shown on the right-hand side and substantially higher in the T300+D that in the T75 plus durva combination that was looked at. This was also supported by the improved response rate and durability of response with a median OS from this study at 18.7 months on the STRIDE regimen versus Imfinzi at 13.6 months. On the right-hand side, you can see that there was also a Phase II work that was done, an externally sponsored study run by Dr. Oh from South Korea who presented the data from TOPAZ at ASCO GI. And this was a reasonable-size Phase II study enrolling patients with first-line biliary tract cancer, looking at the safety and efficacy of the combination of Imfinzi, plus or minus tremelimumab, in combination with gemcitabine and cisplatin. The combination was well tolerated, generated a good response rate, good durability of response. And in this single-center study with Asian patients, obviously, from South Korea, the median OS estimate was 18.1 months for Imfinzi and gemcitabine/cisplatin. So with that, I'd now like to hand over to Dr. Arndt Vogel, who will take you through the data from Imfinzi TOPAZ-1. Thank you.

Yes. Thank you very much for inviting me, and it's really my great pleasure to discuss with you the TOPAZ-1 study. And before I start and to put the study a little bit in context, I would like to emphasize that this is only the second global Phase III study in BTC with positive results. The first one with the ABC-02 study, comparing gemcitabine/cisplatin to our old standard-of-care gemcitabine, and we observed a significant improvement in overall survival. And this is now the standard of care for more than 12 years. So now the TOPAZ study has been reported at ASCO GI last week. And can I get the next slide, please? So this was a global Phase III study, as mentioned before. When we talk about biliary tract cancer, it's always important to understand that we are actually treating different types of cancer: intrahepatic cholangiocarcinoma, extrahepatic cholangiocarcinoma and gallbladder cancer, and I will come back to this in a second. Here, all 3 types were included into the study as you usually see it in trials in BTC. This is a first-line study. Accordingly, patients should not have received any prior therapy and -- but adjuvant treatment was allowed. But here, there was a time in between of more than 6 months required before inclusion into the clinical trial. The study was a 1:1 randomization. Our old standard-of-care gemcitabine/cisplatin was the control arm, and then we evaluate the additive efficacy of durvalumab to this chemotherapy backbone. As you would expect in a tumor type with a very poor overall survival of just around 1 year overall survival [ VOS ], the primary endpoint of the study and the study aimed to see a significant improvement. Secondary endpoint, is another GI -- oncology trials, PFS; response rate; duration of response; efficacy by PD-L1 status, which we could also discuss later; and of course, safety. The next slide, please. These are the demographics of the patients. And I think it's typical for global Phase III study, median age around 64 years. In respect to region and race, you can see there was -- this was well balanced, and around 50% of the patients were recruited outside of Asia. Then in respect to the primary tumor location, that is something what I emphasized in my -- in the first slide, we have 3 types. And usually, we see in clinical trials with systemic therapy, the majority of patients with in intrahepatic cholangiocarcinoma and the same is here, more than 50% of patients with an intrahepatic cholangiocarcinoma included, followed by extrahepatic cholangiocarcinoma and gallbladder cancer. When you look at adjuvant studies, you will see more extrahepatic and gallbladder cancer and fewer intrahepatic cholangiocarcinoma. The majority of patients presented with initially unresected disease, and only 20% were included after surgery. In respect to PD-L1 expression, there are different scores to evaluate PD-L1 expression here. The TAP score was evaluated, which looked at the positive cells per tumor area. And you can see here, a cutoff of 1% was chosen, and the majority of cells -- a majority of tumors had a TAP score of more than 1%. Next slide, please. So I think this is the most important slide of my presentation, overall survival. And I think it's very clear that this is a palliative study with a significant improvement in overall survival. You can see the median, 11.5 compared to 12.8 months, which -- with a hazard ratio of 0.8 and a p-value of 0.021. When we look at IO trials, you must look at 2 points. One is other curves -- crossing is there an early detrimental effect of IO. And specifically, when we use only IO, this is not the case. And the second point, if we look at this, when and how are the curves separating, and we are always going to see the state of the curve with an increase of patients within long-term survival, and this is exactly what we see in the study. The median is always difficult in IO trials, curves are separating after 6 months, and they stay separated for the rest of the follow-up period. And you can see the survival rate at 12, 18 and 24 months, which were all higher in the durvalumab when compared to the chemo only. Next slide, please. Not only overall was improved, also progression-free survival was improved. You can see here, 5.7 to 7.2 months, hazard ratio of 0.5 (sic) [ 0.75 ] also and a significant improvement. And same point here, curves are separating. We do not see an early detrimental effect. And when we look at the PFS rate at 6, 9 and 12 months, we do see long-term responders in the IO arm. Next slide, please. Another point which is always interesting and important also in the palliative setting is whether the tumor is responding. I mean, this is important for the -- for us, as physicians, to -- but also for the patients because it can relieve symptoms. And in the past, you have seen similar to the people in the ABC-02 study that response rate with gemcitabine/cisplatin is rather low, depending on the study, around 20 months. Similar here, 18.7 months. And this response rate was significantly improved by the addition of durvalumab to 26.7 months. And as you would expect with IO combination, duration of response was also increased compared to the chemo-only arm. Next slide, please. Another point is when we talk about additive efficacy if there's additive toxicity, and there always needs to be a right balance. And I apologize a little bit for this busy slide. There are a lot of numbers. But when you just pause a second and when you just look at the event rate and percent numbers, you can really see that with the addition of durvalumab, we do not really have an increase in the event rates, treatment-related events, grade 3 and 4 events, serious TRIEs leading to treatment discrimination, which all was very, very stable. I mean, if you look at these numbers, it would be hard to guess which arm would be the durva arm and which one would be the placebo arm, which is good. Next slide, please. So with this, I would already like to conclude the data. I have mentioned in my introduction only the second positive trial overall Phase III study in the first-line setting in BTC and is the first study reporting positive data with the addition of immunotherapy in biliary tract cancer. The TOPAZ trial met all its primary -- or met its primary endpoint at the interim analysis. And as I've shown you, we have seen a statistically significant and also, I think, clinically meaningful improvement in overall survival. We did observe an increase in response rate, and also progression-free survival was improved by the addition of the checkpoint inhibitor. In contrast to the efficacy data, we did not observe additive toxicity when add checkpoint -- the checkpoint inhibitor to gemcitabine/cisplatin. There were no new safety signals which is really reassuring that we can use this combination in clinical practice. So my final conclusion would be durva plus GemCis is a clearly effective first-line treatment and could certainly become the standard of care for patients with advanced BTC. Thank you very much.

Thank you, Dr. Vogel. I'm now going to hand over to Dr. Bruno Sangro, who's going to talk us through the HIMALAYA data. Dr. Sangro?

Thank you very much. Now it's my turn to talk about HIMALAYA. Next slide, please. Let me first start by saying that this is the study that has recruited the largest population, more than 1,000 patients of unresectable advanced and intermediate-stage HCC. The study population included those patients that today we consider candidates for systemic therapy, patients with metastasis or macrovascular invasion or those with disease limited to the liver in the intermediate BCLC B stage for forever reason is they are not eligible for locoregional therapies, patients who are in ECOG status 0-1, and importantly, have no cirrhosis or a compensated cirrhosis in Child-Pugh class A. Additionally, here, an upper endoscopy was not required to enter the trial. This large cohort was randomized at the end of the day into 3 arms because one cohort arm was closed during the trial because the Study 22 showed clearly that it was not expected to be superior to durvalumab or sorafenib. So the primary objective was to test the combination of a single priming dose of tremelimumab plus the usual regime of durvalumab, STRIDE or T300+D versus sorafenib. And the -- a key secondary objective was to test monotherapy of durvalumab in comparison with sorafenib. And of course, there were a number of additional secondary objectives. Next slide, please. These are the baseline characteristics of the patients. I would like you to focus on 2 important points. One is the distribution by region, and the second one is the distribution by viral etiology. As they come together, importantly, the majority of the patients came from rest of the world, including Japan, versus Asia, excluding Japan. And this is in opposition to other trials published in the last 5 years. And because of that, the -- this large cohort represents very homogeneously the heterogeneity of the viral etiologies with around 30% of patients being infected by hepatitis C, another 30% by hepatitis C and 40% of the patients not having any kind of background viral infection. On the right-hand part of the slide, you can see that patients have indeed a quite advanced stage of disease with most of the patients having extrahepatic metastasis and a significant proportion having either macrovascularization or high levels of Apolipoprotein, both factors that drive poor prognosis in HCC disease. Next slide, please. This is the main result of the trial. The survival curves for T300+D versus sorafenib. You can see that they start separating after 4 months in favor of the immuno combination, and the separation evolves all the time and results in a statistically significant but also technically meaningful improvement in survival. Median was -- being 16.4 before the combo as opposed to 13.8 for the sorafenib monotherapy. But what is most important is that this increase in the separation over time leads to a benefit in terms of the rate of survival at different landmark points, that it goes from 15 at 18 months to 50 at 36 months. And it's remarkable, this 30% survival rate at 3 years in a population of so advanced unresectable HCC. Next slide, please. When we compare durvalumab monotherapy versus sorafenib, the non-inferiority was achieved. And you can see that the -- indeed, the curves start separating at around 9 months. And as a matter of fact, if we look at long-term survival, again, there was a benefit for durvalumab in comparison with sorafenib. Although overall, the survival was not statistically different with a median of 16.6 for durvalumab versus a 13.8 for sorafenib, hazard ratio of 0.86. Next slide, please. In terms of progression-free survival, as expected, there were no significant differences. But again, starting from month 6, there was a benefit in terms of progression-free survival for the T300+D versus sorafenib, resulting in fewer patients having progressed at 2 and 3 years after treatment initiation. Next slide, please. And these came together with, I would say, also expected but nevertheless never proven benefit in tolerability. If you focus on treatment-related adverse events of any grade or grade 3 and 4, they were fewer in the T300+D versus sorafenib arms, indicating the good tolerability of adding this single priming dose of CTLA-4 inhibition on top of durvalumab. Next slide, please. If we now focus on specific hepatic and bleeding events, which are very important in the HCC setting where the cirrhotic background impacts a lot in the tolerability of any treatment, the hepatic events were -- the rates of hepatic events, all rates and grade 3 or higher were comparable in T300+D versus sorafenib. And there were no relevant numbers on bleeding events in the -- both in the durvalumab monotherapy but also in the T300+D combination. Next slide. So the conclusion was clear in these global heterogenous population that is truly representative of patients with unresectable HCC. Single priming dose of tremelimumab plus the regular interval durvalumab with the STRIDE regime provided a statistically significant improvement in overall survival versus sorafenib, median OS being 16.4 versus 13.8. And note also that the STRIDE appeared to provide a long-term survival benefit with, as I mentioned, this landmark, 36-month overall survival of 30%. The overall survival for durvalumab monotherapies was non-inferior with a favorable benefit-risk profile, which is a meaningful endpoint as well. And both STRIDE and durvalumab monotherapy had certainly manageable safety profiles with lower rates of grade 3/4 treatment-related adverse events and treatment-derivative adverse events leading to discontinuation than sorafenib with no increase in liver toxicity or bleeding risk. But altogether, this means that the STRIDE regime and durvalumab monotherapy may represent certainly new treatment options for patients with unresectable HCC. And with this, I will now hand over to Dave Fredrickson.

Dr. Sangro and Dr. Vogel, thank you both very, very much for joining us. And obviously, both will be with us as we move into the Q&A. If you go to the next slide, I just wanted to take an opportunity, as Josie had suggested I would do on the front end of this, to put both of these studies into the context of the existing real-world practice and competitive context in order to paint a picture for where we see the commercial opportunities to be. So if we take a look at what Dr. Vogel walked us through, talked about the fact that this is the second very important Phase III study that's been positive. The last one was over 10 years ago, and that was with the GemCis combination, which is now standard of care within this patient population. Obviously, while that was an important improvement a dozen years ago, despite that advance, only 5% to 15% of patients are alive at 5 years. And that important landmark that we lay out is also serving as the context within which why the long-term survival data that, again, Dr. Vogel spoke to, is so important. So again, to put this into some very patient-friendly terms, when we take a look at the results from this study at 2 years, 1 in 4 patients are alive on the Imfinzi chemotherapy arm versus 1 in 10 alive at 2 years on the GemCis-only arm. And so those data, together with the safety data, which, again, I think that, as Dr. Vogel highlighted, really, I thought quite nicely that if shown in a blinded fashion, you wouldn't be able to determine which one of the data sets was the control arm and which one of them was the experimental arm, speaks to the really important fact that there's no AE-related increases as you move from one to the other. So this is, first, on Topaz. Secondly, as you move then to the next slide into HIMALAYA and what Dr. Sangro walked us through here on the next slide, Again, on HIMALAYA, innovative IO regimen. First time that we've seen a single priming dose of CTLA-4 in the STRIDE regimen, also the first IO-IO combination in first-line advanced unresectable liver cancer. Liver cancer is the third leading cause of death worldwide. You can see some of the numbers that we've got on the bottom left of the slide in terms of epidemiology. So certainly, it is certainly sizable. And again here, 5-year survival in the advanced HCC setting, below 10% and fairly dismal in the mid-single digits. We've heard estimates of as many as 40%, 50% of patients that are deemed to be ineligible for VEGF-based therapies due to either bleeding risks or severe liver toxicity concerns. I think, again, that's part of the reason that the HIMALAYA and STRIDE regimen specifically is so important. Once again, you can see at 3 years in patient terms, 1 in 3 patients alive at 3 years versus 1 in 5 on sorafenib. I'd also point out that we've certainly seen from the IMbrave150 study 18-month landmark survival, but that study does not a any long-term follow-up data in it that allows us to be able to also look at the same 3-year time point. I'm not sure that we'll get a chance to see those, and so I think that it is important to note that, again, as Dr. Sangro said, that this separation, most importantly, is that we see benefit in the rate of survival at these landmarks. And it is certainly our expectation to continue to follow this study, as well as TOPAZ, at the subsequent landmarks that come behind it. And then finally here, as we think about the side effect profile. It's a clear efficacy, safety and simplicity for patients. The single priming dose of CTLA-4 brings with it a relatively simple approach, and there's no increased bleeding risk or severe liver tox seen within the trials. Again here, we think that the fact that, as Dr. Sangro quite nicely said, fewer grade 3, 4 treatment-related adverse effects in STRIDE versus sorafenib with that single priming dose. And so this is how we put those into context and really does represent the first 2 really important Phase III readouts in the GI cancer space where durvalumab, Imfinzi, we hope, now really has a strong beachhead for the future, and it builds a portfolio beachhead also as we look at the other modalities. And with that, Susan, maybe I'll turn to you for you to talk about what's coming next and what we're building upon this foundation and start.

Thank you, Dave. So if you could just turn to the next slide, to the next one. So I just want to put into context the substantial patient number opportunity that was represented by gastrointestinal cancer, as you can see on this chart. We obviously, at AstraZeneca, already have a substantial presence in breast cancer and in lung cancer and growing for prostate cancer with the opportunity to shout-out for ASCO GU, which is coming next month, as well as our presence in hematologic malignancies, particularly CLL with Calquence, and of course, Lynparza in ovarian cancer. But GI cancers really do represent a substantial burden of disease, and I think one of the things that has been challenging in GI cancers is the ability for the immune checkpoint inhibitors to show activity here. This is a start with TOPAZ and HIMALAYA but by no means the end of our ambition here. And I also think there are also opportunities within segments of this patient population for other mechanisms that we have within the portfolio, including, of course, the antibody-drug conjugates. But as shown by the activity of Lynparza in maintenance treatment of pancreatic cancer, there is a significant fraction of patients, including in biliary tract cancer, for example, of patients that have got deficiencies in DNA damage response also. So I think when you look at the spectrum of the modalities that we've got and the spectrum of the opportunities within GI cancer, I'm very confident that in the next few years, we can make a difference within this disease also. If you could go to the next slide, please. So this is just to summarize the ongoing trials that we have for Imfinzi and Enhertu in a variety of different gastrointestinal cancers with the associated data potential readout. And again, we're not stopping with the agents that we've got in late phase. We have several early-stage programs that are also progressing, including with our TROP2 ADC, dato deruxtecan, And I'm very pleased to say that we had a first patient in-human dosing recently for AZD8205 or B7H4 ADC, which is our own proprietary topoisomerase warhead, and this is a target that is also highly expressed in cholangiocarcinoma. We also have ongoing programs for oleclumab and the anti-LIF antibody that we licensed from Northern Biologics in pancreatic cancer. Next slide, please. So I think at this point, we're now going to close the presentation and open up for Q&A. Dave is going to moderate. Thank you.

Yes. Thank you. I appreciate that. So we've got 25 minutes or so for questions, and I can see that we've already got a couple that are in the queue. And so with that, let me start, first, Simon Baker, please, at Redburn.

I'll try and be disciplined because I do have quite a few questions, but I'll limit it to 2. Firstly, a question for Dr. Sangro on the role of viral etiology on efficacy in HIMALAYA. There was quite a difference between B and hep C. So some thoughts on that would be much appreciated. And then a slightly more general question, I suspect, for Dave and Susan. Clearly, you've learned a lot about the optimum dosing of tremelimumab over the years. I just wonder what implications this has got for -- beyond the indications we've seen here for other settings where CTLA-4 combination with INFINITI could be useful.

Thank you, Simon. I appreciate it. So let's go exactly in the order that Simon suggested. Dr. Sangro, in terms of the role of viral etiology as it relates to how you take a look at the HIMALAYA data, and then Susan will follow with you on sort of implications for STRIDE and HIMALAYA on the rest of the CTLA and IO thinking.

Yes. So this is an interesting question because you're probably referring to the fact that post hoc analysis on viral etiology showed lower effect or no effect in -- potentially no effect in the -- in patients with hepatitis C. But let me say that this was largely based on the fact that patients with hepatitis C were -- had a different magnitude of liver functional reserve. And when you adjust it by ILD, then the effect was no longer there or not as strong as and again shows the fact that as statisticians says, if you want to have in a post hoc analysis a negative result, you only have to do a number of them, and the trials are powered, to answer the question, irrespective of viral etiology. I think the important thing is -- for HIMALAYA is that all the etiologies were significantly represented in the overall cohort.

Thank you, Dr. Sangro.

Which is not the fact for other trials.

Exactly. Thank you. Susan?

Okay. So in terms of the -- how I see the use of the STRIDE regimen elsewhere, antibody indicated on the Slide 31 beyond what we've currently presented, we do have ongoing EMERALD-1 and EMERALD-2 studies in locoregional hepatocellular cancer and in the adjuvant setting. I think given the data on the activity of other mechanisms, it is helpful to have a regimen where the safety and tolerability is as good as was seen with the STRIDE regimen in HIMALAYA. And this gives us the opportunity to think about additional combinations that we might want to do, both within GI cancers, but potentially beyond that. So I think I think it's a very helpful data set that we will undoubtedly be taking a good look at and looking at the other places that we can utilize across the portfolio of trials that we want to start.

Great. Thank you, Susan. Thanks, Simon. [Operator Instructions] So now next question is over to Deutsche Bank. I don't know if it's Emmanuel or a member of the team, but over to Deutsche Bank.

Yes. This is Manos Mastorakis on behalf of Deutsche Bank. Yes, first of all, a couple of questions. If you could help remind us the approval and reimbursement status for Imfinzi in China after the SCLC approval last summer? And what has uptake been like in SCLC? And then a follow-up would be is there still any realistic route to a commercial opportunity for PD-L1s opportunity in China, for example, in liver cancer?

Perfect. So in terms of where we are with China for Imfinzi, just by way of reminder, we've got approvals for PACIFIC as well as for CASPIAN. PACIFIC is the Stage III post-CRT data set. CASPIAN is the small cell lung cancer. We are not included, just again for important context, as part of the National Reimbursement Drug Listing for Imfinzi within China, so the utilization that we get and then the reimbursement is through private. We have seen, I think, in the past quarters that we've reported to good utilization of Imfinzi within the private pay market. And I think that we are seeing that we still are enjoying first-mover advantage as it relates to Stage III setting and not yet seeing competitors catch up, though I think that that's probably going to start to come under pressure as some Chinese PD-L1s are beginning to get close to having the potential for approvals within that setting. The same is true within small cell. We have seen some good uptake within the private sector for small cell. Again, I think this is a place where we see a reasonable amount of competition that's coming. To your last question, and then, Niko, I'll invite you if there's anything that you want to add to that, we do see liver cancer as an important area of differentiation within China. And so we are absolutely going to be discussing the HIMALAYA and the TOPAZ results with Chinese authorities. This is an important part of why we moved into the GI space is because of our strength within China. And we think that we've got something that is innovative and an important part of an opportunity in front of us within China. So certainly, we look at the China opportunity is an important commercial one as part of both of these studies. Niko, anything you would like to add to that?

Yes. Thanks so much. Maybe 2 quick points. One is important. There's no regime on the NRDL listed for extensive-stage small cell lung cancer. And since we presented the 3-year overall survival update for CASPIAN, which is a very important data point for the clinical validity, we see actually good research data, particularly in China, with regards to the Imfinzi IO plus chemo use for the CASPIAN regime.

Super. Thanks, Manos. We've got a question from the webcast from James Gordon. What portion of liver cancer patients do you estimate are ineligible to receive treatment with Avastin? And do you see all of these patients being eligible for the HIMALAYA regimen? Perhaps let me ask first Dr. Sangro, if you could speak to this in terms of what your experience and view is on this. And then I invite Susan and Cristian if there's anything you wanted to offer on the same topic. So first, though, Dr. Sangro, to you.

Yes. The amount of patients that are ineligible for Avastin today based on our -- on the experience that we've gained over the last couple of years in HCC is around 20% to 30%, I would say, but that means totally ineligible. I think the question here and -- will those be eligible for HIMALAYA? Yes. I mean, the only reason for not being eligible to an IO therapy today are active, relevant autoimmune diseases, which is a small proportion of the population. And I think what is important is not how many patients would not be eligible for Avastin but how many patients will be eligible for these combinations that seems to be, in my opinion, better options because those patients that might be eligible but have to go to a difficult pathway of performing an endoscopy, maybe finding varices, go to ligation, et cetera, et cetera, and delaying treatment by weeks or months, those patients that are formally eligible would probably be not suitable for anything that is not as easy as this trial.

Thanks, Dr. Sangro. Susan, do you want to add to that?

Yes. So I have the opportunity to have discussions with some investigators at ASCO GI. We the discussed at [ Dr. El-Jawahri ], I asked him exactly this question following the presentation. His comment was very similar to what Dr. Sangro has just said, which is that there's a significant issue with performing the endoscopy necessary to determine some of the potential bleeding risks on varices. And just the logistics of organizing, that's not a trivial thing, particularly in the COVID era. And his estimate was around consider the 40% that I think Dave had on his slide. So it clearly varies depending on the area of practice and is a matter of judgment but also convenience. I think the one thing I would say is that in terms of the view about the HIMALAYA regimen wasn't as much or solely focused on the safety concerns that might choice upfront. But actually that potential for the long-term benefit in the long tail, it's been seen with the addition of CTLA-4 in many other settings, including in melanoma, et cetera, is something that is meaningful for patients. And I think it's backed by the biomarker data. So I think that's an important consideration as well as the safety concerns in terms of the choice between the regimen. Cristian, do you want to make any other comment?

No. I think you and Dr. Sangro called it. Just one reminder related to IMbrave study that ultimately is the basis for the use of bevacizumab, together with atezolizumab, in that indication, the study was excluding patients with untreated or completely treated esophageal or gastric varices, all patients with prior bleeding within a 6 month of initiation of the treatment. And despite that, there were quite a good number of bleedings. So I think in clinical practice, also this is an important aspect that ultimately will impact the treatment of these patients.

Yes. Thank you, Cristian. I mean, I spoke to this on the slide with respect to the simplicity, in addition to the safety, and I think Dr. Sangro's point absolutely underscores this as well. When we have opportunities to talk to treaters within the first-line advanced HCC space, really, the severe bleeding risk and liver toxicity and the need or lack there for scope is a critical consideration. And so I think that this is something that we hear quite a bit about and I think factors into the decision-making. I'd also say that, at least within the U.S., we see somewhere in around 40% of patient share within this space going to atezo-bev. So I think also, that's an indication that there's a significant portion of patients that are being treated, either with another targeted therapy or chemotherapy and that there are a number of patients for whom the existing VEGF combo is not providing the option that they're looking for. Next question is Mattias Häggblom from Handelsbank.

Two questions, please. Firstly, with regards to HIMALAYA. Given the numerically similar median OS for the combo durva/treme arm and the mono durva arm, how will physicians select for which patients to use the combo and for which could the mono be just as good but with a better tolerability profile? And maybe a point of clarification. Does that plan also test for [ non-inferiority ] between the 2 arms as well as one of the secondary endpoints? I may have missed this. And then secondly, with regards to an article published in Clinical Cancer Research this past week that received some attention, where the offers looked at the number of clinical trials involving combinations with checkpoint inhibitors and other therapies, the offers concluded based on the analysis that additive benefits were proven but not to negate as such. So in light of your ongoing combination trials within oncology, I'm curious to hear what hurdles are there for you before exploring combination in the cancer population? And when is a combination not the right approach?

Great. I appreciate that. So maybe perhaps can we start first, either -- so Susan, I think it's probably worth just addressing quickly the statistical aspect of monotherapy and the inferiority and superior design, so we can talk through that, or Cristian speak to it. And then perhaps you can offer perspective, but then we can ask Dr. Sangro as well around -- something that we didn't get much time to talk about in the prepared remarks, which is any view on those patients for whom the monotherapy might be an interesting option. Start first, Susan and/or Cristian.

So in terms of the trial design, the trial is designed to be powered for each arm to be compared with the control, sorafenib, rather than be compared with one another. But nonetheless, contribution of components is always something that's important to demonstrate in discussion with regulatory authorities. So there are subsequent analyses beyond those that have been initially presented that are a normal part of the analysis that we do and the discussions that we have with regulatory authorities. I think when you look at the totality of the data that we've got across Study 22 and the HIMALAYA study and the data that's been seen previously, particularly in terms of the sustainable tail with the landmark analysis at 3 years, there is demonstration of contribution of components, in our view. And that will be the basis of the discussions that we have with regulatory authorities. I think in terms of the comment that you made about the publication on checkpoint inhibitors, plus or minus chemotherapy, whether you're killing different cells within the cancer patient population and whether there's true synergy. I think it's very difficult in those trials to demonstrate synergy which is measured in a very particular way. What has been shown repeatedly is that you actually got added benefit from the 2 components of the combination together versus either separately, and I think the important thing to recognize as well is that the alternative is to give things sequentially rather than as combinations. And a lot of patients, particularly in late-stage cancer trials, don't make it through to second-line treatment, and that's often forgotten when these analyses are done. So I think there's actually a robust data set supporting the combinations of immune checkpoint inhibitors with chemotherapy across a number of different settings. And I think in terms of your question about the choices we make about combinations, we're looking for that -- both safety -- the feasibility of giving a combination effectively as well as the efficacy. Often in preclinical studies, you're looking at synergy for efficacy but forget about the potential of the synergy for toxicity as well. And I think that therapeutic index widening is something that's really important. And again, there were some good examples today with the TOPAZ data showing enhanced activity but no added toxicity as a great example of that.

Thanks, Susan. So Dr. Sangro, just quickly on this. I mean, so obviously, the numerical improvement along the overall survival for durva versus sorafenib didn't translate into statistical superiority. But nevertheless, I think this is the basis of Mattias' question. How do you look at the durvalumab monotherapy arm in terms of any considerations for patients were it to be included as part of the label down the road?

Yes. I think there would be very few patients whom one would consider durvalumab monotherapy versus the STRIDE combination. And this is because, with STRIDE, you have very relevant long-term efficacy, 30% 3 years survival rate with a very tolerable safety profile. And so the difference between the combo, just adding one priming dose of tremelimumab and durvalumab monotherapy is a bit higher incidence of the same profile of adverse events. So you're not broadening the range of adverse events. They are just slightly more frequent. And so for the general community, I would not consider because of this good tolerability of the combination monotherapy. Maybe in some cases where you have some concerns about potential immuno tox. For instance, a patient with past ulcerative colitis or any of the autoimmune disease, well, maybe those, but those are very few patients, I think the combination is set to become the standard of care for most patients.

Thank you, Dr. Sangro. I think it's a really, really important point. It gets back to the benefit-risk and the innovation and simplicity of the STRIDE regimen and the lack of toll, if you will, the CTLA-4 addition but the benefit and the detail that goes along with it, and that's something that we've heard a lot of. So thank you. The next question on the webcast comes from Richard Parkes. And Niko, I'm going to ask you to speak to it. Can you discuss the challenges to driving uptake in the BTC market? It's seen as a modest opportunity, but the epidemiology numbers suggest it might be more meaningful. What percent of advanced BTC patients currently receive chemo? Are there other challenges to integrating durva into the treatment of these patients? So maybe, Niko, your thoughts. And then certainly, Dr. Vogel, any other comments, after Niko responds, in terms of how you see the utilization of chemo within this group. Niko, please.

Yes. Thanks a lot. So I think we're looking at around 90,000 to 100,000 patients in the U.S., France, Germany, Spain, U.K. Japan and China that are considered treatment-eligible. And probably around 70% of them actually are chemo-eligible. One important point I want to make, we're not talking about combination of durva with one standard-of-care chemo. We consider CisGem the standard-of-care chemo. So we also base on several failed attempts to improve on that standard of chemotherapy with other combinations of chemotherapy. Noteworthy to mention, FOLFIRINOX trial from the ECOG collaborative group that just showed that even FOLFIRINOX didn't really improve that standard. And having said that, we certainly will look at this as adding durvalumab directly to that chemotherapy standard with more or less no exclusion situation, meaning that every patient that is eligible for CisGem usually should be eligible for durvalumab CisGem.

Great. I mean, maybe just quickly, Dr. Vogel, are there many of your BTC patients that you are treating with something other than chemotherapy?

No.

I mean, are there patients that are getting something other than GemCis or perhaps one of the other chemos that are out there? Or are the majority of those patients ones that are indeed drug-treated?

So I mean, first of all, I would like to make a point that BTC was really a resected tumor. We had only few opportunities now really clinical trials from pharmaceutical companies and not too many options. And this has really dramatically changed in the last years specifically because we do find a lot of druggable alterations, so there's clearly a new excitement about biliary tract cancer. And with this, I think we do see more biliary tract cancers, not only because they are going up in numbers due to the risk factors like fatty liver disease, for example, similar to HCC, but also because many tumors of unknown origin are now reclassified and more correctly classified as biliary tract cancer, specifically for those adenocarcinomas in the liver. So I think the numbers are going up and also the willingness to treat these patients. And actually, I would think that in our clinic, more than 90% of patients are treated. We have a bias because we are academic centers. But in general, gemcitabine/cisplatin is a very well-tolerated chemotherapy backbone. And with that, I think there are hardly any patients that we would use only gemcitabine, which is usually suggested for more frail patients. In some countries like France, GEMOX is recommended as another platinum-based chemotherapy, but I think the globally accepted standard of care is gemcitabine and cisplatin. And overall, I'm quite convinced that more than 80% of patients are candidates to get this kind of treatment.

Super. Thank you very, very much. To Berenberg and Luisa Hector. Please, Luisa.

I had a question on the kind of the tail of the responders. Are they still on Imfinzi? So I'm just wondering the longest duration. Are you seeing use in these patients of 1 year or more? And I'm sorry if I missed this, but what is the performance like when you break out the patient pool by PD-L1 expression? And then maybe, finally, Dave, very much a commercial question, but -- and I know you've touched on it a little bit. But what do these studies really do for the Imfinzi growth profile? So I know you're not going to give numbers. But when I look at these 2 data sets, it looks like they will really drive the ex U.S. sales to be significantly ahead of the U.S. Obviously, we need to allow for some potential locals in China. Is that a fair assumption? And then perhaps if we look at the current label, again, is it fair to assume a little bit more growth on the current label? Perhaps that's sort of a bit more muted now. And then the TOPAZ and the HIMALAYA study can really make meaningful contribution on top of that.

Thanks, Luisa. Appreciate the question. So maybe we'll start with Cristian. Cristian, would you be able to speak to maybe even Luisa's first 2 questions where you can talk about kind of patients in HIMALAYA that -- and those that are still on therapy at the landmarks also speak to the PD-L1 data that Dr. Sangro spoke to. And then I can come back to the commercial piece.

We still have around 15% of patients that are ongoing on the STRIDE regimen, so we will have another data cut to confirm this tail that, of course, is the strength like they presented of the outcome. So still, these patients with -- we have a very long follow-up in our trial, still there is a 15% of patients that are ongoing. So for a cause for OS. We will provide another data cut, and then we will decide what will be the right timing. The second question on PD-L1, Dave, was more...

Just to review, did we see any differences in the forest plots by PD-L1 status, which is something that Dr. Sangro had a chance to just show us quickly, but it's just a question on the forest plots. And I think Luisa maybe missed it as we went through it.

Yes. Not really, but I would like also Dr. Sangro to comment on this. And we didn't see a real -- this is not a real driver for this population. So Dr. Sangro, do you want to comment on this or this is also your take on this?

Yes, absolutely. These -- the results from HIMALAYA echo the results from any other trial with checkpoint inhibitors in HCC, where there was a slight increase in response rate and overall survival that was absolutely nonsignificant and not clinically meaningful. So PD-L1 is not a way to select patients for checkpoint inhibition in this particular case for STRIDE regime.

Thank you.

Thank you. That's quite clear, I think, in answer to Luisa's question. So Luisa, I think that what I'd like to offer on the commercial opportunity question is maybe just answer this in 2 parts. The first is that, together, HIMALAYA and TOPAZ, we think, represent, in and of themselves, significant opportunities for us to be able to build good growth on Imfinzi. I think it's also important to note -- excuse me, for Imfinzi and treme, I think it's also important to note that as we take a look at treme, treme has protection for 12 years on -- in the U.S. on the biologics side and 10 years data exclusivity within Europe. That's an important aspect of how we think about the way in which we'll, of course, approach pricing and the way in which we'll think about the value that can be maintained through this combination study. Also, Susan spoke to the fact that we've got a number of important GI studies that are also going to be coming in on the heels, the ones that we're reading out here. We made some pretty sizable, I think, and important investments in our IO portfolio, in general, in the 2017-2018 time frame. And we're seeing an opportunity with these 2 studies but also the EMERALD studies to hopefully see that we're able to have a first-mover advantage into the liver cancer space, which is something that is obviously a big opportunity if we're able to indeed get these data as quickly to patients as we hope that we will. Next question online is from Jefferies and Peter Welford. Please, Peter, go ahead.

Yes. I'll be quick. Just 2 quick ones really. Firstly, on the TOPAZ-1 study. I wonder if Dr. Vogel, in a similar vein actually, too, HIMALAYA. Can you just talk about the PD-1 expression. So I appreciate, is it post op into the subgroup analysis, but it does look as though the majority of benefit there was potentially driven by those PD-L1 expressers. Curious whether or not is that something you think that is worth pursuing further. And then secondly, just coming back to China. And I don't know -- curious, is there a CTLA-4 yet approved within China? And how do you think about potentially using treme, I guess, as your path onto the NRDL? Or is that really not viable?

So Dr. Vogel, would you like to speak to your view on the PD-L1 patient subtypes within TOPAZ-1? And then we'll come back to the other questions that are more commercial in nature about China, and we'll handle that within AC.

I think I can make it really short, and it's very much in the direction what [ Professor Langlord ] just said for PD-L1 expression in hepatocellular carcinoma. In cholangiocarcinoma, it's even more difficult. We do not really have an established antibody. It's a very heterogenous tumor, not only between the tumors, but also with a very high intratumor heterogeneity, which makes the assessment of PD-L1 expression extremely difficult. You have different cells that express PD-1 expression, which might have a different impact on outcome. And with that, it's not even a weak biomarker. And I think there are some subgroup analysis have been performed and have also been shown at ASCO GI, but there was no clear trend to an enrichment of patients that would benefit from checkpoint inhibition in BTC. So I think it's an interesting biomarker. It's a weak biomarker, and it's too early to draw any conclusions. And I think it will probably make -- that will not make it into the clinic.

Thank you very much. On the question, Peter, with respect to CTLA-4, this is the only -- and I think that this is the presentations that we spoke. This is the only combination data that we've seen that has been IO-IO positive combination data within the liver cancer space. And so I think that, that remains something that's certainly an important part of an advantage that we have with this study relative to the landscape that exists within China. We are still working through very much how we want to think specifically about our local reimbursement strategies. But I think that if you just take a look at where the differentiation lies within China, as you know, and we've talked about extensively, there are many, many, many checkpoint inhibitors, either approved or on the NRDL, including a long list of domestic. There's far less competition in the CTLA-4 space, although that will probably pick up, but there's also no one with the data in the combination space in liver. So that's really the place that we'll look to, to see if we can create meaningful opportunity there in China, and we're full speed ahead on that given the importance of this disease within Asia itself. Moving next on to a webcast question, and I know we're a bit over time. I think this may be our last question. Matt Weston, Credit Suisse. How should we think about pricing given the short course of treatment in STRIDE? Secondly, what's patent protection for treme? Thirdly, how do you see the opportunity given the low price of local PD-L1 competition? And Josie tells me there is a couple more. We'll take maybe 1 or 2 more after this, if possible. I think, Matt, I addressed the majority of these in the comments. We're still working on how we think about pricing. I do think that, even though it is a short course, we will think about treme in the context of how we best get the value from STRIDE given also that this will likely represent the first global approval for tremelimumab. And obviously, POSEIDON is the other that we're working with. There's an opportunity to set the price with HIMALAYA, and so we'll have to work on how we want to make that work out. But we've got some ability to be able to think about the way in which we introduced the sequence of that and whether we set price with HIMALAYA or set price with POSEIDON, particularly ex U.S. And so I think we can think about this as a course of therapy question and not necessarily on a per dose. Patent protecting -- protection 2031, '32, depending on the jurisdiction. And I think that I already spoke to the third point. So a question from Seamus Fernandez. Any thoughts on the pancreatic cancer data with Seagen's CD40 and thoughts on this mechanism? Susan, I think that's a good question for you. Secondly, can you discuss the commercial prospects for treme? Is there something unique to IgG2s versus ipi's IgG1? And then thirdly, can you update on your CTLA-4 bispecific as well. Companies highlighted this program 2 years ago has been deprioritized. So Susan, any -- actually, I'd ask if you could, if you might be able to take all 3 of those and comment on those maybe in the order in which they appear. Even though the second question is a commercial question, I think it's really an MOA question.

Yes, I understand. So I think in terms of pancreatic cancer data with Seagen CD40, I think it's still early. We need more data really to really understand the dose and schedule aspects and the mechanistic in that context. As we all know, pancreatic cancer is a particularly difficult setting. I think beyond just stimulating the T cells in that setting, there need to be other components of the regimen, which is what we're trying to address with some of the other combinations that we've got. So I think it's just early, to be honest. In terms of is there something unique to IgG2 versus ipilimumab's IgG1 approach? Honestly, I think there's been a lot made of the Is type differences, but it really isn't data clinically that it actually makes a difference. If you actually look at all of the tremelimumab data versus ipilimumab, it's actually very similar for similar patient populations in terms of mechanism of action. So I think what's important is that the CTLA-4 mechanism of action, there's some fundamental differences about that versus PD-1. I think you're getting a broader stimulation of T cells, which is one of the reasons why you get a broader array of the immune-related adverse events and this really important component of the priming of the immune response, which is again something that the STRIDE regimen was designed to optimize for that priming response, which leads to durable tissue resident memory T cells that have very long durability. I think that's the thing to take away from the HIMALAYA data and the biomarker data that supports that. In terms of the CTLA-4 bispecific, just as a reminder, the design of that molecule was designed to have the ability to bind CTLA-4 only in the presence of PD-1, so to help tune it more to activating T cells within the tumor and less activity in terms of activating T cells within some of the normal tissues and to provide a bit more of a therapeutic window. I think this is an important program for us. It remains important. Getting the right dose and schedule selection, as we've already shown with the tremelimumab data, I think, is really important. We continue to evaluate in Phase II this program in a number of different tumor types, and we'll be happy to show the data later once those are available.

Super. Thanks, Susan. So I think we've got one other question from the webcast, which I think is really a nice one actually for us to close on. And let me just thank Doctors Vogel and Sangro for your time with us. I know we've gone a few minutes over, so thank you for doing this. Recognizing that we're asking a question of both of you as it relates to HCC patients, could you describe in the -- your experience what percent of HCC patients are surviving at 3 years in your practice? So I think really the question is asking do you find performance of the control arm in HIMALAYA to be representative of what you're seeing in real practice? Or indeed, have you seen perhaps any improvement as a result VEGF combos? I think it's seeking to try to understand the HIMALAYA/STRIDE results in the context of what you are seeing today in your real-world setting. So maybe first, I'll ask Dr. Vogel, maybe you can speak to it. Then Dr. Sangro, as you presented on the study, I'll let you have the last word on that.

I think it's a very interesting and relevant question, and I think we really have to acknowledge here the rapidly changing field in hepatocellular carcinoma for many years. We just had one drug available for sorafenib, and this has really tremendously changed in recent years. And I think many of us are now starting to reevaluate our data in real-world cohorts to really better understand the impact, not only of the IO-based combinations in the first-line setting, but also the impact of sequential therapy on overall survival. Of course, you always have patients with 2 diseases, the underlying liver disease and liver cirrhosis, which has an impact on overall survival, but having more opportunities definitely offers prospect for more long-term survival. So if you ask me in the past, we -- this is rough, we have seen hardly any long-term survival. There were some. Usually, I wrote a case report on them when we observed one in the clinic. And today, we see it certainly more often, yes. And with the longer survival, we have, of course, also more opportunities of sequential therapies. And then it's difficult to distinguish is it just the first-line effect? Or is it the effect of the sequential therapy, probably a combination of both? And so to make it short, I think we do see in the clinic an increase of long-term survivals, which we have seen more than we have seen in the past, definitely more.

Thank you, Dr. Vogel. So Dr. Sangro, to you. And I didn't mention, this question came from [ Richard Parkes at Robeco ]. Dr. Sangro, please, on the same question from Richard, if you could, please.

Yes. To echo Arndt's comments, I would say, I mean, overall numbers I think -- overall, I think 10% or fewer would survive 3 years in clinical practice. But if you focus on those patients that are fit for clinical trials, which is a proportion of the whole cohort, then probably 15% to 20% would be alive. Has the other combinations increase that to some extent? Yes, absolutely, yes. But I think what is important is not only these landmark 3-year survival data. And as Arndt mentioned, we are seeing now patients that are alive, not only 3 years, 4 years, 5 years, 6 years from the start of treatment. And these long-term benefit is something that is absolutely unique to immune checkpoint inhibitors combinations.

Thank you, Dr. Sangro. I really -- I think that's a good and important place to stop on. We've long been very much intrigued by the biologic hypothesis of immune memory with CTLA-4 and the promise of long-term durable responses for a certain subset of patients. Obviously, the toxicity and side effect profile has been one of the challenges in getting there. And I think that, that speaks to why the STRIDE regimen is something that was as important as it was for us to be able to find a way to bring that to patients in the way that we have. And so let me just take a moment to once again thank Doctors Vogel and Sangro for your time with us today. We really deeply appreciate it, as always, to thank my colleagues from AstraZeneca, who also have joined us today. And then lastly, thanks to all those in the -- those who've dialed into the call. We appreciate very, very much your interest in AstraZeneca. Thank you for joining us in this special ASCO GI edition of a touch point that we've had, and we do very much look forward to having many more conversations on the advancements that we're making into the GI cancer space given how important it is as an area of high unmet need across the globe. Thank you all very much.