BeOne Medicines AG (ONC) Earnings Call Transcript & Summary

Hello, everyone, and welcome to BeiGene's Early Development Pipeline and Research Virtual Event. My name is Craig West, and I'm the Head of Investor Relations at BeiGene. We have the great pleasure of having with us today both of BeiGene's Co-founders, Dr. Xiaodong Wang and John Oyler. In addition to being founders of BeiGene, Dr. Wang is the Chairman of our scientific advisory board, and John Oyler is BeiGene's Chairman and CEO. In addition, we have Dr. Lai Wang, Senior Vice President, Head of Global Research and APAC Clinical Development; Dr. Eric Hedrick, BeiGene's Chief Adviser. And in addition, during the Q&A, we'll have Dr. Jane Huang, CMO of Hematology and Dr. Yong Ben, CMO of Immuno-Oncology; Dr. Howard Liang, CFO and Chief Strategy Officer, will be our host for today. After the presentations, we'll take questions from the audience. [Operator Instructions] As a reminder, today's call will be recorded. I would now like to introduce your host for today's conference, Howard Liang. Howard, let me turn the call over to you.

Thank you, Craig. I'd like to welcome everyone to our R&D event. As many of you know, in the last few years, the investor focus on our story has been the progress for 3 late-stage candidates. We recently provided an update on BRUKINSA, our BTK inhibitor, and tislelizumab, our PD-1 antibody, around ASCO following Phase III data presentations. We've been getting increasing interest in our pipeline behind the 3 commercial and registration stage assets. We're very excited to host this event to showcase both our research organization and our pipeline. Before we get into the presentation, I need to remind everyone that we'll be making forward-looking statements, and our business carries certain risks. Some of these are discussed in our filings with the SEC and the Hong Kong Stock Exchange. So please see those filings for discussion of the business and the risks. Today's presentation will start with opening remarks by Xiaodong, which will be followed by video presentation, introducing our research organization, its people, capabilities and facilities. We'll then provide an overview of our research programs and internally developed assets. Focusing on TIGIT, BCL-2, OX40 and a potentially first-in-class molecule in late preclinical development, targeting HPK1. Eric will then cover in-license programs. He will just highlight a few in a broad portfolio. Sotorasib or AMG 510 partnered with Amgen, sitravatinib in collaboration with Mirati, and zanubrutinib or ZW25 from Zymeworks. John will conclude today's presentation with an overview of our strategy before we open up the call to questions from analysts. I would like now to turn over the call over to Dr. Xiaodong, our Co-Founder and Chair of Scientific Board of BeiGene. Xiaodong?

Thank you, Howard. BeiGene was founded 10 years ago with a goal of best medicine for most people. As for the most people part, it's obvious why we are in China because China has world's 1/5 of population and cancer patients of U.S., Europe and Japan combined x2. What may not be so obvious is China also have now talent for scientific research. Over the last 10 years, many scientists that are working in both academics and pharmaceutical industry has returned to China and to form a variety of biotech companies. And in BeiGene, we believe the best path to best medicine is to follow science. And we believe is core -- our core research principle is to follow science, which means we believe in evidence, and we believe in logic. We don't believe hearsay or simply portfolio declarations. And we pursue every research program with best-in-class differentiations, and we are not afraid to terminate the subpar programs during any time of the development. And so far, BeiGene's research team has produced and delivered 2 drugs approved and the third one at the filing stage and all with clinical differentiated properties, and we also have 11 internally developed molecules advanced into clinic, all with clinical differentiated prop -- preclinical differentiated properties. And we believe our research team in BeiGene will continue to expand and deliver in the future. Now I'll give back to Howard.

Thanks, Xiaodong. So at this point, I would like to introduce our research organization with a video presentation. [Presentation ]

I'm Lai Wang, Senior Vice President, Head of Global Research, and Asia PAC Clinical Development. BeiGene's first decade has been extremely productive. Under this relatively short period of time, the company has built an exceptional research organization, has developed a proven track record of success. That, in turn, has resulted in a robust early pipeline. We're going to be discussing each of these points in more details today. BeiGene has built an excellent research organization that is now equipped with a team and other capabilities to allow us to target more types of cancer with cutting-edge technologies and multiple therapeutic modalities. We have one of the largest research team in China with a full suite of capabilities in-house, and much of the team have had long tenures at BeiGene. The research organization has advanced 11 in-house discovered molecules into the clinic. Of these programs, 2 have already been proven -- or being proved for multiple indications, the third is close to being filed. The engine that we are here today to describe in more details than ever before has produced a robust pipeline of differentiated candidates. This included TIGIT and OX40 antibodies and our B-cell inhibitor as well as a potentially first-in-class HPK1 inhibitor. There are a number of internal combination opportunities that we are investigating under the TIGIT program is progressing well, and we plan to accelerate it into Phase III as quickly as we can. Let me give you a brief overview of the broad and deep set of research capabilities that the company has built over its first 10 years. BeiGene's research organization includes 4 major departments: biology, chemistry, biologics and the translational science. Within this 4 major departments, there are many functional groups. These groups work closely together to form an integrated research engine. We also implanted a portfolio management system to help drive all the programs and prioritize within our portfolio. At BeiGene, we aim to build seamless transition from research to manufacturing and to clinical development. Safety, formulation, [ early ] [indiscernible] and the translational teams are all involved in the programs very early on. We believe this ensures better transition and a faster time line. In addition to immuno-oncology therapies, target therapies for solid tumor and hematology oncology, BeiGene has now expanded its therapeutic area into immunology and inflammation. In tumor therapeutic modality, in addition to traditional small molecule and monoclonal antibodies, BeiGene has invested heavily in several new areas in the last couple of years, including PROTAC, bispecific antibody and antibody drug conjugates. Our collaborations with Ambrx on their site-specific drug conjugation and Zymeworks on their bispecific antibody platform are part of this big initiatives. PROTAC is like bispecifics small molecule. On the one hand, it binds to the target protein and on other hand, it binds to E3 ligase that will lead to ubiquitination and degradation of the target protein. Not shown here, yes, we are also investigating the use of the newest technologies such as cell therapy and mRNA. The research team is currently pursuing 10-plus potential best-in-class and the first-in-class projects with a plan to double that in 1 year. BeiGene's first research center was and still is the building you are seeing here on the left in Beijing. Because it resides in the Changping District, we often call ourselves, we are the champions. We began with a small team that grew towards 200 in the first 8 years. We have been in growing mode in the last 18 months. Today, our team is 350-plus people, and we're running about 12 full preclinical programs. We're planning to expand the team and are building a new research center in Shanghai that is intended to effectively double the size of the team under the number of programs in about 1 year. BeiGene's research team has a proven record of picking the right targets and producing high-quality molecules. In a short span of 10 years, actually more like 9 years as we didn't really start the first experiment until December 2010, 2 homegrown molecules have received market approval. BRUKINSA, our potentially best-in-class BTK inhibitor, is now approved in the United States for patients with relapsed/refractory mantle cell lymphoma and in China for patients with relapsed/refractory mantle cell lymphoma or CLL. It is the first China-discovered molecule that was approved by FDA and granted breakthrough therapy designation. BRUKINSA was designed to overcome the off-target and low absorption issues associated with ibrutinib. It is more selective and achieved a much higher exposure in body than ibrutinib. The safety profile of this drug was shown to be favorable in the head-to-head study in the form of non-Hodgkin lymphoma called Waldenstrm's macroglobulinemia. Our PD-1 inhibitor, tislelizumab was designed to minimize the bonding to Fc gamma receptor or macrophages and avoid T-cell elimination. It is the only approved PD-1 antibody that is designed in such manner. On the development front, tislelizumab has been approved in Hodgkin lymphoma and urothelial bladder cancer in China and the filing for frontline use in both non-squamous and the squamous non-small cell lung cancer and the previous treated HCC have been accepted by the NMPA. Total 16 pivotal trials have been initiated for tislelizumab, 7 of them are global studies. Our third molecule, pamiparib is at the filing stage. This molecule is differentiated by, a, its ability to get into the brain and b, not a drug pump substrate, preventing a potential resistant mechanism that has been reported for other PARP inhibitors in the clinic. Overall, BeiGene's first decade has been extremely productive and in its relatively short period of time, the research organization has advanced 11 in-house discovered molecules into the clinic. We have established a track record of picking highly relevant targets such as PD-1, BTK and the PARP. Today, I can assure you, this is just the beginning, and that there are a whole lot coming. Seen here is BeiGene's robust early pipeline that now encompasses 28 assets, 8 of which we have global rights to. And we plan to continue to add to this list from our internal efforts as well as from collaborations such as the deal with Amgen. As not all the programs from our work with them have found their way onto this slide yet, I'll be speaking about 4 selected internal discovered assets highlighted here, namely our programs in TIGIT, OX40, BCL-2 and HPK1. You can see here that our programs present a wide variety of ways to intervene therapeutically in the tumor microenvironment. Shown on this slide, we have agents targeting MDSCs, T Regulatory cells, dendritic cells, NK cells, cytotoxic T cells and angiogenesis. These agents indirectly kill the tumor cells by activating the immune microenvironment. Many of them have potential to be combined with PD-1 or tumor cell targeting therapies. In addition, we also have several molecules that directly target tumor cells. As shown on this slide, our tumor-targeted therapy pipeline is broad and it targets a variety of tumors, including glioblastoma, non-Hodgkin lymphoma, non-small cell lung cancer, small cell lung cancer, gastric cancer, colorectal cancer, biliary tract cancer, pancreatic cancer, prostate cancer, melanoma, multiple myeloma, ovarian cancer, ALL, AML, CLL and breast cancer. With that, let me turn to our first molecule and tell you about our TIGIT program. There has been very encouraging proof-of-concept data reported by Roche for their TIGIT antibody in combination with their PD-L1 antibody at this year's ASCO. BeiGene's TIGIT antibody BGB-A1217 is one of the most advanced TIGIT antibodies with the right Fc format and the PD-1 combination RP2D identified. It is by far the more potent than Roche's TIGIT in the preclinical studies. Competent Fc is believed to be essential for efficacy base based on preclinical data. A1217 combination with tislelizumab has been very well tolerated in the clinic, combo RP2D has been identified. We believe we are in position to compete globally, and this could be a transformative opportunity for BeiGene. A broader registration program is being planned. TIGIT has 2 ligands, PVR and PVRL2. Once bond, it induces a inhibitory signal in T-cells and NK cells. BGB-1217 -- A-1217 blocks this interaction, freeing up PVR and the PVRL2 to bind to CD226, which activates co-stimulatory pathway in T-cells and NK cells. In this way, TIGIT antibody provides both a blocking immune inhibitory signals and immune cell stimulatory signals show CD226. One of the highlights of this year's ASCO is Roche's tiragolumab and atezolizumab combination results in first-line PD-L1 positive non-small cell lung cancer. As shown in the red box, by adding tiragolumab, overall response rate was increased from 24% for atezo alone to 66% for the combo in patients with PD-L1 TPS score equal or higher than 50%. On the right bottom, there is a clear separation of the 2 PFS curves of the mono and the combo treatment. Hazard ratio was 0.3, which is very impressive. These results clearly show the benefit of adding the TIGIT inhibitor in this setting. Here, in the left panel, we show that an anti-TIGIT antibody with a competent Fc region on the left graph is effective at the tumor growth inhibition in this model. In contrast, an anti-TIGIT antibody without this Fc function has no apparent effect on tumor growth in the right graph. The mechanism of action to explain this data is still being elucidated, but data that leads to this conclusion has been independently produced by others. There are multiple explanations for this observation and shown on the right are 2. One possibility is that Fc and Fc gamma receptor co-engagement enhances the quality of immune synapsis, leading to increased T-cell responsiveness. The second MOA is that Fc and Fc gamma receptor engagement on myeloid cells creates a pro-inflammatory tumor microenvironment. I mentioned that our program is one of the more advanced program of the many that have been started. The slide shows there are 11 TIGITs in the clinic, but only 3 have both wild-type IgG-1 backbone that is shown in preclinical study to confirm maximum efficacy. And the recommended Phase II doses for the combination used with PD-1 or PD-L1. The 3 that are most advanced are indicated in Red, BeiGene, Merck and Roche. BeiGene's BGB-A1217 is fourfold more potent than tiragolumab in binding assay, the first-in-human study initiated last year. It's a Phase I study examining A1217 in combination with tislelizumab in the advanced solid tumors using a dose escalation and expansion design. The combination was well tolerated, and we have identified our Phase II combo dose. Full target occupancy was observed in peripheral at the lowest dose level. We are planning to move forward aggressively with a broad registrational program. We're going to turn now to one of our newest molecule, our BCL-2 inhibitor. BGB-11417 is a very potent BCL-2 inhibitor with potential to overcome resistance to venetoclax, the first approved BCL-2 inhibitor. Venetoclax is approved for at 400-milligram dose for CLL and 600-milligram dose for AML. It is being explored at 800-milligram in breast cancer. These are very high doses and could present significant issues with pure burden. For a more potent molecule like 11417, there is opportunity to be dosed at higher if needed. 11417 is more selective than venetoclax for BCL-2 relative to BCL-xL. It demonstrated a large therapeutic window in animal studies. Its exposure in NOAEL dose is close to 30-fold higher than predicted human therapeutic exposure. It is well-positioned to be combined with BeiGene's own PTK inhibitor, zanubrutinib. Dose escalation studies ongoing: Currently, it is at the 80-milligram dose level, which is predicted to be equivalent to 400-milligram of the venetoclax. Combination with zanubrutinib is planned to be initiated in the second half of this year. This table shows a variety of biochemical and cellular assays in which 11417 showed much stronger potency versus venetoclax, in both wild-type background as well as in the G101V mutant cells, which is mutant associated with venetoclax resistance in the clinic. It is worth to point out, 11417 has better IC50 against the wild-type BCL-2 than G101V. It suggests, we probably need a high dose in the clinic to inhibit the mutant form. The bottom row shows the selectivity profile of the 2 compounds. This graph show that 11417 works in both wild-type under the G101V mutant xenograft models. This mutant model on the right was a knock-in mutant our research team created. As noted on the previous slide, G101V is a mutation that has shown venetoclax's resistance in the clinical setting. In both models, the higher dose of 11417 produced the best efficacy. And in the mutant model, the curve of high-dose of 11417 separates early [ gastric cancer ]. BCL-2 inhibitors are not only useful for hematological malignancies, BCL-2 inhibitor is also a promising therapeutic strategy for breast cancer and small cell lung cancer. BCL-2 is often overexpressed in ER-positive breast cancer and combining venetoclax with endocrine therapy has shown tolerable safety and notable activity with a response rate of 54% at the RP2D. However, at the chosen 800-milligram dose in combination with tamoxifen, there was a significant pure burden with venetoclax that 11417 might not have. Interestingly, during inhibition of the CDK4/6 and BCL-2 pathway has also shown augmented response in ER-positive breast cancer in preclinical primary tumor models. In small cell lung cancer, there's early data showing BCL-2 inhibition provides benefit, and the preclinical data is correlative in showing that high BCL-2 expression can further sensitivity of small cell lung cancer to venetoclax. Let me now turn to another antibody therapeutics, our anti-OX40 antibody, BGB-A144 -- A445. This is one of my favorite programs. BGB-A445 is the one and only anti-OX40 antibody in the clinic that does not disrupt OX40 and OX40 ligand engagement. By doing so, it not only retains OX40 ligand's reverse signal in the antigen-presenting cells but also achieves maximum OX40 activation in the T-cells by keeping the natural ligand stimulation. A non-OX40 ligand blocker is widely efficacious in various preclinical models, including PD-1 resistant models. In addition, this molecule can be combined well with many other immune modulators under chemotherapies in preclinical studies. The Phase I dose escalation is ongoing, and the combo with our PD-1 is planned to start in the second half of this year. This slide shows the putative difference in activation characteristics of the antibodies that do and do not block OX40 ligand binding to OX40. The top right panel shows, as you move from left to right, with increasing concentration of ligand-blocking antibody, it will disrupt the interaction between OX40 and OX40 ligand, leading to loss of the reverse signaling in APC antigen presenting cells. In addition, due to lack of natural ligand stimulation, you won't be able to achieve the maximum OX40 activation in T-cells. In contrast, in the bottom right panel, you see that a non-ligand blocking antibody can achieve stronger T-cell activation by retaining the ligand stimulation of OX40 as well as the OX40 ligand reverse signaling APC cells. Here in the top panel, we compare this 2 type of -- these 2 types of OX40 antibodies on tumor cell of host in OX40 knock-in-mice. On the left is BeiGene's anti-OX40, BGB-A445. And you see a nice dose response, where increasing dose of the antibody led to deeper tumor growth inhibition. The TGI bar graph shown right, in contrast, a reference ligand-blocking OX40 antibody in the clinic showed the hook effect. Both the lowest and highest dose levels provided no tumor growth inhibition. The tumor growth inhibition was only observed with the middle dose. As we all know, drug exposure varies from patient to patient. With this type of U-shape curve, it is very difficult to identify an optimal dose for a patient population. In addition, I want to point out that the maximum tumor growth inhibition was lower for the reference antibody compared to our OX40. The bottom panel shows objective responses that have been presented from clinical trials of competitors' OX40 antibodies. And interestingly, most responders occurred at the low dose levels, either as single agent or in combination with a checkpoint inhibitor. There's one main point on this slide, which is that BeiGene's A445 is the only... [Technical Difficulty] Sorry, I just accidently on mute myself. There's really one major point on this slide, which is that BeiGene's A445 is the only OX40 antibody in the clinic that is known not to block OX40 ligand interaction with OX40. Inhibrx molecule is of unknown format at this moment. We believe this distinction has the potential to be important based on the data that I just showed you. Here, we show 13 mouse syngeneic tumor models. In each model, we test the effectiveness of a non-ligand blocking OX40 antibody as well as a PD-1 antibody. Black line is the vehicle group, the red line is a PD-1 antibody, and the blue line is the OX40 antibody. The top row includes 7 models that responded to both OX40 and PD-1. In most cases, OX40 antibody was more efficacious than PD-1 antibody. In the middle row of the graph, we show the data from 4 models that only responded to OX40, but not to PD-1. So these are PD-1 resistant models. And in the bottom, there are 2 models where neither mode of intervention worked. In sum, OX40 antibody was active in 11 out of 13 models, whereas PD-1 antibody was only active in 7. This is a busy slide, but we don't need to spend a lot of time on it. The purpose of the slide is to show that OX40 can be beneficially combined with several other therapies. Shown here, we tested our OX40 antibody in combination with PD-1, PI3K delta, sitravatinib, TLR9 agonist and PD-1 chemo combinations, all have better results than the individual therapies by themselves. BGB-A445 is in a Phase I dose escalation, dose expansion study that we have the screen shown here. After the early dose levels, each group will have tislelizumab added to their therapy. From this data, we will select the recommended Phase II dose, and that dose is planned to be advanced into the expansion portion of the study with either monotherapy or in combination with PD-1. Turning to one of our newest molecule. Let's look at BGB-15025, which is a potential first-in-class HPK1 inhibitor. BeiGene started work on this program a few years ago. HPK1 is a key negative feedback regulator of T-cell receptor pathway. 15025 is a highly specific and potent HPK1 inhibitor. We have observed interesting combination activity with PD-1 and the preliminary tox studies suggests there is a wide therapeutic window. We plan to file an IND later this year and expect it to be in the clinic early next year. This graph shows the HPK1 signaling pathway that occurs downstream with the T-cell receptor and the effect that occurred when the pathway is stimulated. As part of this pathway, HPK1 normally phosphorylated the SLP-76, which is critical adaptive protein downstream of the TCR machinery. This phosphorylation caused SLP-76 to undergo proteasome-based degradation, which leads to T-cell inactivation as shown on the right. There has been increasing evidence in the literature showing that HPK1 plays a critical role in T-cell activation and antitumor activity. The panel on the left shows CD4 and CD8 cells stimulated with anti-CD3 treatment. Cell with wild-type HPK1 produced a modest interferon gamma production upon activation. But there's 3 other cell types, which have none or decreased HPK1 catalyst activity or have increased interferon gamma production relative to the wild-type. Experiment with mice that have no HPK1 activity is shown on the right. In this panel, wild-type mice showed a quick tumor growth throughout the study period, whereas the HPK1 kinase dead mutant mice showed the tumor regression. This slide shows that our lead compound has demonstrated significance in vitro and in vivo activity. Such as, in the panel on the left, which shows that our compound caused an increase in IL-2 production in peripheral blood mononuclear cells. The panel to the right shows 15025 resulted in inhibition of phosphorylation of SLP-76 with a nice dose response, as one would expect based on the mechanism action for this molecule. In this slide, we are showing you the combination of HPK1 with PD-1 in the preclinical model. 15025 has some antitumor activity by itself as does the PD-1 antibody. But when combined, the tumor growth inhibition is maximized. On the right is shown the count of animals treated with indicated experimental therapies and the 28% of the group received the combination treatment were tumor free, as opposed to 4% with PD-1 alone and 13% in the 15025 monotherapy group. I hope I convinced you today that BeiGene has a robust discovery engine. BeiGene's first decade has been a productive one. The research organization is now 350-plus strong and boasts a full suite of research tools that do not limit us to any 1 type of treatment modality. We can bring the latest cutting-edge technologies such as PROTAC, ADC and bispecific antibodies to build on the biological problems we seek to solve. Further, our capabilities are full and internal, such that we are not relying on outsourcing. The sum of what we have been building has resulted in 11 clinical programs today from internal discovery engine, that have been quite productive and produced that molecules that we have just described to you. We have a TIGIT program that is one of the most advanced with a compound with advantageous design, and we are moving forward with our selected dose. We also have a BCL-2 inhibitor, we believe, has a potency advantage and potentially can overcome venetoclax resistance, and it can function as a great partner for our zanubrutinib. We have the only anti-OX40 antibody that is non-ligand competing, which we believe could be an important distinguishing feature. And finally, a new program we have just talked about for the first time today is our efforts in HPK1 inhibition, which could be a first-in-class molecule that prevents T-cell exhaustion. With that, I would like to turn the presentation over to my colleague, Dr. Eric Hedrick, our Chief Adviser, to take you through some of our in-license programs. Eric?

Great. Thanks, Lai, and thanks to everyone for joining us today. I'm now going to move on to our partner programs, which contributed a significant amount to our overall development pipeline. Go to the next slide. Over the past 5 years, BeiGene has really been quite active in pursuing a variety of clinical and commercial partnerships. And we've utilized our unique China development and commercialization capabilities to enter some development partnerships for a number of interesting molecules. As a result, we now have over 25 molecules across 9 collaborations and these very well round out our internally developed assets. In doing so and as we approach this partnering, we focused on a few things, not only complementing our internally discovery programs but we're also focused on diversifying our therapeutic modalities and perhaps the Amgen partnership with its broad set of the BiTE assets is the best example of that, but the -- our pipeline -- our partner pipeline is obviously not limited to those assets. Another focus was expanding the combination potential with our marketed drugs, zanubrutinib and tislelizumab. And finally, we really wanted to take advantage of our specific development expertise in China with a focus on cancers that are of importance in Asian population. So the interesting thing is that many of these collaborations are approaching a state of maturity, such that we expect several of these molecules to be in late-stage development within the next 6 to 18 months. So here graphically, you see the partnered programs in the context of our overall pipeline. Obviously, our partnership with Amgen contributes a significant proportion to this portfolio but the number of interesting candidates is by no means limited to the Amgen pipeline. For the purpose of today's presentation, I'll be highlighting 3 of these programs that have either entered or approached late-stage development: AMG 510, sitravatinib and ZW25. And these are examples of the approach that we've taken overall to our development partnering. You should also be aware that there are several other collaborations listed here that are earlier stages of development that we'll be describing over time. So on to the first highlighted program, which is sotorasib or AMG 510. Obviously, this is a program that's generated a significant amount of interest and excitement in our community. Most of this audience will appreciate that mutated KRAS as a drug target has been historically very difficult. And it was only until a discovery or breakthrough by scientists of Amgen in the identification of a surface group that allowed for a potent covalently binding inhibitor to be identified. The KRAS G12C mutation is the most frequent of the KRAS mutations in non-small cell lung cancer. I'd also note that although the proportional frequency in Chinese patients is expected to be less than in western patients, we estimate that the number of patients diagnosed annually as an absolute number is roughly similar in the U.S. and China. Most will also be aware that this drug is established proof-of-concept in both non-small cell lung cancer and colorectal cancers harboring the KRAS G12C mutation. And the program is advancing very quickly. At this point, both Phase I and Phase II studies in these diseases have completed enrollment. BeiGene's particular role in the partnership is the conduct of clinical activities in China. And this would either be in the form of China-specific trials or more commonly, global Amgen-sponsored trials in which China will participate. Given the nature of the diseases being investigated, we'd expect that our contribution to the overall program will be significant. China will be participating in trials starting in the fourth quarter of this year. And as a reminder, potentially registrational Phase II as well as the initial randomized Phase III study in KRAS G12C mutated non-small cell lung cancer are currently ongoing. Next slide. So as a reminder, seen here is the data presented by Amgen at ESMO last year in non-small cell lung cancer showing a high frequency of durable responses with this agent as a monotherapy in the setting of G12C-mutated disease. And next slide. And again, proof-of-concept data with AMG 510 has also been established in KRAS G12C-bearing colorectal cancer and this data is from the most recent ASCO. We'll now move on to sitravatinib, which is a program that we're pursuing in collaboration with Mirati. So sitravatinib is a multi-kinase inhibitor. It was of particular interest to us, not only because it potently inhibits VEGF receptor, but it's also a potent inhibitor of tumor-associated macrophage kinases such as AXL, Tyro3 and MerTK. Those familiar with our development of tislelizumab will realize that we have a particular interest in the role of tumor-associated macrophages in cancer immunotherapy and so the binding profile or the specificity profile of this molecule is of particular interest to us. With sitravatinib, we believe that inhibition of these tumor-associated macrophage targets, which appear to be critical in the production of an immuno-tolerant state in a tumor, may be an important adjunct to checkpoint inhibition with tislelizumab. At this point, proof-of-concept data has been produced with -- for combination of this agent and checkpoint inhibitors in non-small cell lung cancer, urothelial cancer. And we've recently shown a study conducted by BeiGene that we've established, what we believe, proof-of-concept in platinum-resistant ovarian cancer, which is interesting, obviously, because this is a relatively checkpoint inhibitor resistant tumor type. BeiGene responsibilities again in this partnership are similar to what is profiled with Amgen. We're responsible for China development. And I think in this case, our ability to really expand the scope of the early program by working in China has already been realized. We're conducting broad Phase Ib studies in a variety of tumor types. Globally, the program is entering potentially registrational trials in non-small cell lung cancer. Go to the next slide. So as I mentioned, the original proof-of-concept data for sitravatinib as an adjunct to the checkpoint inhibitor nivolumab in patients who had previously failed checkpoint inhibitor therapy is shown here. In the setting, the response rate was 16%, and a vast majority of patients achieved sustained disease control. And with the caveats of this being a single-arm study, longer-term outcomes such as PFS and overall survival seem encouraging. Next slide. And interestingly, in what I mentioned a moment ago, when you look at data that we've just reported from our Phase Ib study in platinum-resistant ovarian cancer was interesting in that the unconfirmed PR rate for the combination was 41%. We presented this data at ESMO last year. And again, it sort of highlights that this combination may be applicable, not only in checkpoint inhibitor sensitive tumor types, but also potentially checkpoint inhibitor resistant tumor types. And then finally, we'll move on to the bispecific HER2-targeted antibody, ZW25. This is the development collaboration with Zymeworks. Next slide. So ZW25 or zanidatamab is a bispecific antibody that was developed by Zymeworks, targeting both the trastuzumab and pertuzumab domains of HER2. So potentially, this agent as monotherapy has the potential to provide additive activity of both of these antibodies. This type of activity has been well demonstrated already in preclinical models and the Phase I study of ZW25 suggested that this type of activity may potentially translate clinically, at least at the proof-of-concept level. Again, BeiGene is responsible for China-based development activities. And I think as an example of our overall capabilities, we've also established a manufacturing collaboration with Zymeworks that has been quite productive. This molecule, again, has already entered late-stage development. In fact, a potentially pivotal trial in HER2-expressing biliary tract cancer is now active. Additionally, Phase II combination studies, which will ultimately support Phase III trials in both HER2-expressing breast cancer and gastroesophageal cancer are underway. Of course, in -- especially in terms of the gastrointestinal cancers, given their prevalence in China, we would expect that BeiGene will have a significant role in the registrational programs globally. And so lastly, hopefully, it's providing an illustration of how BeiGene's partnering strategy has supplemented our internal discovery efforts to put us in a very favorable position in terms of our overall clinical development pipeline, with candidates of interest in their own right as well as their combination potential with our existing marketing agents. So with that, I will conclude and turn it over to John for some concluding remarks. John?

Thank you, Eric. Appreciate it. In summary, our internal R&D platform is generating a robust, sustainable pipeline. It's proven, it's working in cutting-edge areas, it's impactful, and it's broad and growing. BeiGene is building unique, sustainable competitive advantages. Most importantly, our clinical acceleration and lower cost with a team of over 1,350 across the globe for China-inclusive global trials. The ability to combine with our internal platform assets, such as PD-1, BTK and PARP. Our science and medicine-based commercial team of over 1,300 people with 6 commercial products, an excellent pipeline and leadership of Dr. Wu Xiaobin, and our internal manufacturing and preclinical capabilities, including biologics manufacturing. Our past collaborations are successful. They leveraged these competitive advantages, and they expand our portfolio. And capacity exists for this to continue to be a major source of growth for BeiGene in the future. BeiGene is well-positioned to act quickly and capture future internal and external breakthroughs. I want to talk about recent accomplishments and upcoming milestones. This is really hard to keep track of. If you look at the past 10 months, even for me, it's almost impossible to keep track of. But if you look at this, for preclinical assets that we've advanced into the clinic in the last 10 months, it's 4; for trials enrolled, there were 5, fully enrolled; 3 Phase III data readouts; 4 NDA filings; and in terms of approvals or launches, there were 6. We added 25 assets through collaboration. And in addition to that, we helped build out our organization, for example, in biologics manufacturing or in business development. As we look at the next 18 months, it's a similar story. There have been -- there is a plan to have 4 early data readouts. For potential Phase III readouts and potential filings, 9 are expected. And for potential NDA filings or regulatory discussions, there are 7. And our commercial portfolio is expected to grow from 6 to 11. It's a really, really exciting time to be part of BeiGene, and it's just a pleasure to have the opportunity to work with the team that's accomplishing so much. In terms of the future vision of BeiGene, we do believe that right now is a once-in-a-lifetime period of transformation in our industry. It creates opportunities for small players to become the industry leaders. We believe that BeiGene is one of the best positioned companies for this time and for this period of transition in the world. On this journey, BeiGene is striving to do the following: first of all, become an oncology and scientific leader. Secondly, to expand beyond oncology into other areas of need. Third, to continue to build our sustainable competitive advantages. We've talked about them on the previous slide, but I think that this is truly something that if you want to understand who BeiGene is, it's not just the assets, it's the capabilities and the competitive advantages that we've grown as an organization. Fourth, we're striving to become the best global clinical organization in the world. This is the biggest issue of our industry, clinical trial cost and time. It's the most important thing to be addressed, yet most organizations in the world are asking third-parties to build this capability for them. At BeiGene, we're very focused on this. It's the biggest problem to solve, and you'll continue to see us try to build sustainable, competitive advantage here. And lastly, as Xiaodong said from the beginning, our goal and our vision was to transform the industry to bring better medicine to more patients across the world more affordably. And that's what BeiGene is set out to do. Thank you all, and we'll move to questions and answers.

[Operator Instructions] Our first question comes from the line of Tony Ren.

Just a couple of very quick ones. So first on TIGIT. So we know that Roche is testing its tiragolumab with atezolizumab in the SKYSCRAPER trial, but only in PD-L1 expression over 50% cohort within the small cell lung cancer. Is this a similar strategy that you guys are going to pursue as well for your TIGIT asset? The second one is on BCL-2. We know that venetoclax has a lot of TLC, which -- tumor lysis syndrome, which makes dosing a little bit challenging. So have you guys seen TLC in your studies? The next one is on OX40. So we know that AstraZeneca, Pfizer has discontinued their OX40 -- excuse me, discontinued their OX40 program. And I think the asset from Eli Lilly is not very active. It was about 4% response rate. So I just want to hear how you guys are thinking differently? And lastly, it's not exactly related to your early pipeline, but I just want to get a sense of ABRAXANE supply situation in China. So this is Tony Ren from CLSA.

Thanks, Tony. It's Howard. So I think the first question is TIGIT? No, sorry, which one was...?

Yes, first one is on TIGIT. Yes. So SKYSCRAPER is testing PD-L1 expression in over 50% set. We already have Keno 24 (sic) [ KEYNOTE-240 ] there, right? So I just want to see if you are doing the same?

Sure. Lai, could you start answering the questions?

Sure. In terms for the TIGIT, the question is about whether we're going to PD-L1 try it on 50% patient population. At this moment, as you can probably guess, with all the competitive landscape, we are not in a position probably to discuss about the detail of our origin trials. But what I can tell you is, we're moving very aggressively towards a broader registrational program. And also, I just want to point it out, BeiGene has initiated 16 pivotal studies for our tislelizumab, so those will function as a foundation for our future add-ons.

Okay. Great. BCL-2?

Jane, could you -- do you mind asking -- addressing that question, whether we're seeing tumor lysis syndrome?

Yes. We are in our early days of the Phase I trial, and we have not yet seen tumor lysis syndrome. However, it is a hallmark of the drug being potent. And as we approach the higher risk tumors, such as CLL or mantle cell, we have a plan in place in order to address that.

And Lai, on the third question on OX40?

Yes. I think as we're talking about today, our OX40 is unique in a way that it does not disrupt the OX40 and OX40 ligand interaction for preclinical model. It shows very different pattern of dose response compared to OX40 antibody that will block ligand interaction. As you pointed out, many other clinical OX40 antibodies hasn't really shown much clinical activity. Their response has been low. But we do believe those might be due to its wrong format. And this OX40 is quite different. We are really eager to see its activity in the clinical, either as a single agent or in combination with PD-1.

John, could you address the ABRAXANE question?

Sure. I think that ABRAXANE currently is not being commercialized in China. That is due to a manufacturing issue that is not related to BeiGene. It's associated with a third-party manufacturer for BMS Celgene. And from that perspective, that's an issue that they are handling and dealing with to try to get it back active in the market as soon as they can. And so that's probably all I'm able to say about that at this point in time.

Yes. Got it. I appreciate that. So if I may just tag on to one more. So how is BRUKINSA selling in the U.S., particularly with the pandemic situation. It right now probably makes it quite difficult to promote it.

Tony, if I can take that. So we, as you noted, BRUKINSA is in early-stage of launch, and there is a complication due to the pandemic. But I think the launch has been on track, going well from our perspective. So we'll just update you when we report additional financial data for the second quarter.

Okay. We'll move to the line of Ziyi Chen.

This is Ziyi from Goldman Sachs. Yes, I was really excited to see those new details on those early-stage discovery stage assets. So just a couple of questions. First, just on HPK1 inhibitors. Could you explain the differentiated points for TIGIT, BCL-2, OX40 or -- for HPK1 inhibitors because we know that there are other players who are working on that, like Celgene working with Nimbus on that and a few other players. So we try to have a better understanding of how differentiated it is to other different HPK1 inhibitors. That's my first question. And second, I think BeiGene is actually working on both in-house discovery and also licensing models. So will any of the licensing assets actually block your own research effort on developing like second-generation or better version, for example, because now we're already working on the KRAS, the AMG 510, other assets. So is that going to be any of the restrictions for BeiGene to develop the next-generation KRAS? So will that also apply to other like same-target innovation? And my third question is more on R&D budget because probably, if you look at all the major pivotal trials ongoing, they are probably in 2020 and 2021, we're going to see a lot of large pivotal trials on tislelizumab is going to be completed. So that means, going forward, probably we're going to have more resources to be allocated to some of the early stage projects. So if we look at across 3 major late-stage assets and versus some early-stage clinical development and also versus our efforts on the research and the licensing, what could be the break -- roughly break down, like 60%, 70% on 3 core assets, 20% on early stage or 10% for the research side? Just trying to get a sense.

Lai, would you take the HPK1 question? And John, can take the PD question. I'll address the R&D question.

Yes. For the HPK1, as you pointed out, there are quite a few other companies who are working on HPK1, but at this moment, nothing is in the clinic. And this is not an easy target to make a good molecule against. So BeiGene really started on work on this target a few years ago and we started with HER2 screening. We were able to identify a lead and then further optimized based on that. We believe we have a very potent and also selective HPK1 inhibitor in our hand. And also, this has been demonstrated by in vitro, in vivo activities as well as through the human PK pathogen, it has really nice human PK profile. And in addition to that, from the early tox study demonstrated, this molecule has a good therapeutic window. So it's difficult at this moment to really comment on how all the different HPK1 inhibitor will behave in the clinic because nothing is in the clinic yet.

Xiaodong, do you have anything to add either on this program or our programs in general? Xiaodong, you are on mute.

I think Lai said it very well. We certainly are aware there are quite a few other players on the table. And as BeiGene always does, we profile their molecule as well if we know their structure, and we strive for the high bar. And we already don't just with a single blow we have a combination of blows just try to achieve what we want to achieve.

John, on the PD question, whether our collaborations block our internal research.

Sure. I think that, look, when you build collaborations and you're working with people, you try to, in each of those collaborations, align interest and create scenarios that are going to be win-wins and where neither of you need to be worried about the behavior of others. I think we strive to do that in all of our collaborations. The exact restrictions, those are always deal terms that are typical to the industry and happen and exist. And so those are in probably any licensing arrangement that you see. So from that perspective, naturally, that's the way you want to structure partnerships. All that said, there is plenty to do in the world. And I'm unbelievably proud of the research team that we have, you can look at the accomplishments and all that it's moving through the pipeline. But we're still only scratching the surface of things that we could possibly be doing or working on from that perspective. And there are areas where there's disease expertise that other people are better than us at. And so certainly, we prefer to work with them and have the A team in that area of expertise working on it than trying to build that from scratch. So we're very comfortable with that sort of relationship, and I think you'll continue to see that. Thanks so much.

Thanks, John. Ziyi, on your R&D question, the evolution of the programs, I mean, you're right. As John showed in his slides, some of these trials that we have ongoing for BRUKINSA and tislelizumab, we'll be reporting the data, some, by the end -- at the end of next year. So this is -- and clearly, we're -- we have -- at the same time, we have increasing number of earlier-stage assets moving to pivotal development. I think it's a nice transition with continued pipeline movement. I think in terms of the specific breakdown, research is -- we're increasing investment in research, but the overall R&D budget is still relatively small. Clearly, as John noted, industries spend mostly on clinical trials, actually late-stage clinical trials. So we do have -- I think, the focus will increasingly shift towards this next early-stage targets. It's a little bit difficult to say specifically the breakdown because some of these products, for example, we do TIGIT in combination with tislelizumab, it's hard to say whether that's a tisle program or a TIGIT program, same with the BCL 2 with BRUKINSA.

Our next question comes from the line of Yaron Werber.

Can you hear me now?

We can.

Okay. Terrific. Thanks, Craig. So I have a question -- a couple of questions. First one on OX40, can you give us a sense of how do you design the antibody in terms of binding to the ligand, so you bind to the epitope. So you won't actually prevent the binding -- of the ligand binding? And then I have a couple of commercial questions.

Lai?

Yes. Maybe I address the question. This is really through the screening assays. We perpetuate to have assays to -- of course, during screen also have ligand-blocking antibodies. And with -- then we compare the activity of the ligand-blocking antibody versus the ligand-nonblocking antibody. We're actually seeing an increased T-cell stimulation with ligand-nonblocking antibody. And with that kind of the assay and then we're further digging into the mechanism of action and we believe this is really due to which I illustrated today is ligand -- natural ligand binding will help to achieve the maximum activation on the T-cell side of it. And on the other hand, the ligand also will have a role in the APC cells. So if you disrupt the ligand and OX40 interaction, you will lose the APC activation. So this is really through the screening effort to identify antibody and specifically only bind to OX40, but does not disrupt the ligand interaction.

Okay. Great. And then a question on ZW25 or zanidatamab. In terms of the Phase III or the registrational-enabling first line, the HER2 gastroesophageal cancer, I assume that's going to be a head-to-head. Is it a head-to-head against Herceptin and chemo? Or do you envision a different kind of trial design?

Eric, could you answer that, please?

Yes. Sorry, Howard. Maybe I'll give -- if it's okay give Ben the first shot at this and then I'll add in.

Yes, Ben?