Daiichi Sankyo Company, Limited (4568) Earnings Call Transcript & Summary

Thank you for joining Daiichi Sankyo's conference call to present ASCO 2020 highlights. Please note that today's call will be recorded. Dr. Manabe, please go ahead.

Thank you for joining, ladies and gentlemen. This is Sunao Manabe, President and CEO of Daiichi Sankyo. Welcome to our conference call to present ASCO 2020 highlights. We wanted to hold the live event in Chicago as we did last year, but hosting the conference call this time. First and foremost, we would like to express our deepest sympathy to all around the world that have been adversely affected by COVID-19. We sincerely wish for speedy recovery of those affected. Also we pay our heartfelt respect to the medical staff and the others around the world that are working tirelessly for the affected patients and their families. Today, I will first present our progress in R&D over the last year and the highlights from ASCO and then provide some views around our mid- to long-term strategy. Following that, Antoine, our Global Head of Oncology R&D, will provide overview of our data presented at ASCO and then the current status and future plan for oncology R&D. Please see Page 3. 1 year had passed since we held our first ever on-site event at ASCO for investors. During this year, DS-8201 was launched in the U.S. in January under the brand name, ENHERTU, for breast cancer and ENHERTU was also launched in Japan last week for breast cancer. We have also seen significant progress outside of breast cancer. For gastric cancer, we submitted NDA in Japan at the end of April and received breakthrough therapy designation as well as orphan drug designation in the U.S. in May. For non-small cell lung cancer with HER2 mutation, we received breakthrough therapy designation in the U.S. in May. This is the third breakthrough therapy designation for DS-8201, following breast cancer and gastric cancer. The alliance with our partner, AstraZeneca, is going very well, and new critical trials are being prepared. Page 4 show the progress of DS-8201 for breast cancer and gastric cancer. For breast cancer, as I mentioned earlier, ENHERTU was launched in U.S. and Japan for third-line treatment. The 3 Phase III trials are steadily progressing. And we are discussing development plans with AstraZeneca for earlier line treatments. For gastric cancer, we presented favorable result from the pivotal Phase II study at ASCO. Again, NDA was submitted in Japan in April. And based on the SAKIGAKE designation, we obtained for this indication, the review period is expected to be 6 months or less. As for the U.S. breakthrough therapy and orphan drug designations, it's apparently extremely rare to obtain these designations with Asian population data only and we think this reflects the high expectation for this product from the regulatory agency. Page 5 show the progress for lung and the colorectal cancer study. Positive results were presented at ASCO for both studies. The response rate for non-small cell lung cancer with HER2 mutation was 61.9%. And for HER2-positive colorectal cancer, the response rate was 45.3%. Both are very strong response rates. For non-small cell lung cancer with HER2 mutation, we obtained breakthrough therapy designation in the U.S. We will further accelerate development and will pursue indications for additional cancer types, such as lung and colorectal, which are cancer type with no approval anti-HER2 therapies. Please see Page 6. In addition to DS-8201, we have made significant progress over the past year for DS-1062 and U3-1402. For DS-1062, interim data were presented at ASCO. Compared with the last year, the number of cases increased significantly, and the development is moving forward nicely. The lung cancer trial for U3-1402 is progressing nicely, too, and internal data will be presented at the future conference. We are currently reviewing our development plan following the U3-1402 breast cancer program. Please turn to Page 7. As I presented, the potential of the 3 ADCs are steadily increasing, and we have changed our R&D strategy to 3 and Alpha. Based on this strategy, we are moving forward R&D activities to maximize the value of the 3 ADCs with priority. In Alpha, we are focusing on R&D activities with the potential to change the current SOCs for sustainable growth of our business. Please see Page 8. Our business growth scenario has become clearer as 3 ADCs progressed steadily. The challenge of -- the challenge for our U.S. business outside of American regions, formerly reported, was a lack of products to drive further growth. With the progress of 3 ADCs and earlier-than-planned launch of ENHERTU, in particular, we now have core pillars for regrowth of our U.S. business, which was not going at the LOE of olmesartan, the anti-hypertension franchise. In Europe and in ASCA, Asia, South and Central America, our previous plan was to grow around the anti-coagulant, LIXIANA. With the progress of 3 ADCs, we now have expanded our portfolio for accelerated growth. Please turn to Page 9. We have started to put together our 5-year business plan that starts from April 2021. And with a strong R&D strategy, 3 and Alpha, and the powerful driver of business growth, I am confident that we can put together a high-quality plan. Our plan -- no, I'm sorry, our view is that the potential in ENHERTU/DS-8201 is now higher than originally estimated. And we will nurture this product as the growth engine for the entire 5-year business plan. For DS-1062 and U3-1402, we will focus on accelerating the development and will also consider utilization of external resources versus utilization of internal resources only to take the best strategy for long-term growth. We will enhance our capabilities for development operations through the new arrangement with our CRO, Syneos Health, and our capabilities around manufacturing through additional capital investment. And on R&D activities for sustainable growth, I will directly get involved where important and steer the activities. We will announce the 5-year business plan in March or April of next year. With that, I will now hand over to Antoine. Antoine, please?

Thank you, Manabe-san. My name is Antoine Yver. I'm the Global Head of Oncology R&D and the Chair of Daiichi Sankyo Cancer Enterprise. It is a pleasure to be again with you, although, only virtually. Today, this ASCO -- sorry, in this particular time, are very special for Daiichi Sankyo as we are now well recognized as a critical global oncology player and a source of what I believe is an excellent, innovative and rigorous science which is meaningful to patients. So next slide, 11. Here is today's agenda. We will cover DS-8201, the value, the data, the plan. We'll move on to our next DXd-ADC flagship, 1062, our investigational TROP2 ADC using the same DXd platform. We will carry on with the Alpha portfolio, the R&D transformation, a little bit about the impact of the SARS-CoV-2 pandemic and finish with forward-looking news view before taking questions. Next slide and next slide, Slide 13. Now in 2020, beginning the fifth year after launching Cancer Enterprise, the onward and upward journey continues. We have made and are making a steady and strong progress. We saw the delivery of strong value. We are progressing fascinating new science, which is a direct effect of our innovative pharmacology and we continue our transformation to meet the opportunities ahead of us. Slide 14. Since last year ASCO 2019, we created tangible value by delivering 3 new drug approval: ENHERTU in the U.S. and Japan; TURALIO in the U.S.; and VANFLYTA in Japan. We've uncovered fascinating new science on the role of receptors in the ADC pharmacology, an aspect which, so far, has delivered many researchers in pharma, probably because the ADC themselves didn't have the broad spectrum and precision in payload delivery that the DXd technology allows for. This is true for HER3, as I've explained at our R&D Day in December last year. It is also clearly true for HER2 in lung and colorectal, as well now as for TROP2 in non-small cell lung cancer. These past years has also seen Daiichi Sankyo Cancer Enterprise received 2 additional breakthrough therapy designation by the U.S. FDA, both within HER2 in lung and gastric cancer, respectively, and most recently, as Manabe-san said, in the very late April, we completed the submission of the first supplemental marketing application for ENHERTU this time for gastric cancer in Japan, just 2 months after the first approval. So very proud of progressing the portfolio rapidly. Next slide, Slide 15. 2020 is also a year of continued transformation to support the delivery of our portfolio. This is true in particular for our global development operations. The fiscal 2020 forward-looking R&D budget increases by 16%, directly funding the ADC demand. We've just announced an exceptional collaboration with Syneos Health. I'll explain further later. We've also defined last fall the Alpha strategy to secure our longer-term future beyond 2030. And our colleagues in manufacturing demonstrate and build scale and agility to meet the biologic demand. Next slide, Slide 16. Just excuse me a second. Thank you. Now moving to 8201, the value. Slide 17. In the U.S., DS-8201 was born for a first-ever approval after a short 113-day review by the FDA, which was right after the last R&D Day in December. We just -- our timing was just a little too short. We have so far delivered an outstanding market launch sequence. The indication in the U.S. is clear and does not specify the need for prior treatment with T-DM1. It is clear that, with the recent approval of tucatinib, the second line and beyond treatment sequence in HER2-positive metastatic breast cancer has significantly evolved. DESTINY-Breast03 study or the Phase III versus T-DM1 will further establish the pivotal role of ENHERTU and gating combination to support first line and second line are underway or in preparation and we're observing the clear positive effect of ILD Safe Use measures. Next slide, 18. In Japan, the first approval for breast cancer was received in March 2020 after a rare 6-month review for our first approval. And as Manabe-san said, the second submission was made in April under SAKIGAKE designation, and we expect a swift 6 months' or less review. We've also launched an early access program for gastric cancer in Japan. In the European Union, we expect the EMA CHMP to validate our breast cancer submission very soon through the very first part -- of the first half of fiscal 2020. Next slide, 19. Now the data in first gastric cancer. Next slide, 20. The DESTINY-Gastric01 study, the randomized study of DS-8201 versus an active treatment control in patients with advanced gastric or gastroesophageal junction cancer and it was just presented at ASCO and simultaneously published in the New England Journal of Medicine. This is the second time that DS-8201 had simultaneous New England Journal of Medicine publication in 6 months after the DESTINY-Breast01 study was published in NEJM at San Antonio Breast Cancer Symposium last December. Next slide, 21. We presented the primary analysis results for Cohort 1 in the study. 187 subjects randomized 2:1 to DS-8201 at 6 milligram per kilogram every 3 weeks versus the physician choice of irinotecan or paclitaxel. 100% of subject had received trastuzumab, 86% had received taxane, 72% ramucirumab and 33% also on anti-PD1 or PD-L1. Next slide, 22. Here are the results. The primary end point was overall response rate. The most relevant measure is the RECIST-confirmed response rate, at 42.9% in the DS-8201 arm versus 12.5% in the active control arm. In this heavily pretreated advanced and relapsed or refractory cancer population, of particular interest is to observe 10 cases of complete response by independent review. Of 10 of the 51 responses were complete, 20% of all responses were complete responses. The median duration of response was 11.3 months for the DS-8201 arm versus 3.9 months for the active control arm. On the right-hand side of this slide, you can see the corresponding waterfall plots. And active control did have some tumor reduction effect, but this was quite shortly as shown by the short 3.9 months' duration. Next slide, 23. On these lines, I've shown the overall and progression-free survival analysis. Both Kaplan-Meier plots show an early and sustained compression of the curve. The median overall survival in DS-8201 was 12.5 months compared to 8.4 months in the active control arm. The corresponding highly relevant and statistically significant hazard ratio was 0.59 plus 41% risk reduction of death and a p-value of 0.0091. The 1-year survival rate estimate was 52.1% in the DS-8201 arm versus 28.9% in the active control arm. What about safety? Next slide, Slide 24. Here is a safety summary. And now typical profile for DS-8201 with neutropenia, anemia, thrombocytopenia as well as decreased appetite and nausea, common to both arms but in excess in the DS-8201 arm. 51 -- I'm sorry, 15.2% subject discontinued treatment in the context of the treatment-emergent adverse event and 32% and 62.4% dose-reduced or dose-interrupted, respectively. [ ITTs ] ILDs were observed in 9.6% of subjects with no death. Next slide, 25. Now we move to non-small cell lung cancer in subject with HER2-mutated tumors. Next slide, 26. We have presented at ASCO the result for the Cohort #2 in the red box. The Cohort #1 is still recruiting and too early to report. DS-8201 was administered single agent at a dose of 6.4 milligram per kilogram. This is the main data, but not the only data, which supported the recent breakthrough therapy designation by the FDA. Next slide, 27. 42 subject in this Cohort 2, 1/3 from Asia with 45.2% of subject with central nervous system metastases at entry -- brain metastases at entry. Next slide, 28. 90.5% had received NSCLC platinum-based treatment, 54.8% anti-PD-L1/PD-1 and 19% docetaxel. Next slide, 29. The confirmed overall response rate by independent central review was 61.9%, including one complete response. The median duration of response was not reached at data cutoff, with a lower limit of a 95% confidence interval at 5.3 months. The median progression-free survival was so far estimated at 14 months. Clearly, in addition to response rate, durability is key in that setting. Next slide, 30. Here is the corresponding waterfall plot. Of interest is that not one subject had the best response -- or best change of a target lesion as a measure above the baseline value. The dotted line of 30% decrease, which corresponds to a former response, is barely seen towards the left of the plot. Next slide, 31. Here, you have the overall survival and the progression-free survival Kaplan-Meier plot as seen here, and they are consistent with the anti-tumor effect just described on response rate and durability. Next slide, 32. Here is the overall safety summary. Nothing new. The median duration of treatment was almost 8 months. And next slide, 33. Here is the bar graph of the most frequent treatment-emergent adverse events. Even known exception of anemia and neutropenia, most events were a Grade 1 to a Grade 2. Next slide, 24. And here is ILD, 11.9%, all cases being Grade 2, low Grade 3, low Grade 4 or low Grade 5. And now finally for DS-8201, next slide, 35. It's a result of the open-label Phase II study in HER2-expressing metastatic gastric -- I'm sorry, metastatic colorectal cancer. Slide 36. Here is the study design. We have reported on Cohort A, 53 subject with HER2-positive by IHC 3+ or IHC 2+ and ISH+. At the data cutoff, Cohort B and C with lower expression level were only very partially enrolled and not reported here in full. Next slide, 37. Here are the subject characteristics. 98% were confirmed RAS wild type. Slide 38. Here are the prior treatment summarized. This subject has received a median of 4 prior lines of therapy for very advanced disease. 100% had received irinotecan, 5 fluorouracil/capecitabine, oxaliplatin and an EGFR-targeting monoclonal antibody. 75% received also bevacizumab and 30% on new dedicational use of an anti-HER2 agent. Next slide, 39. The confirmed overall response rate by blinded independent central review was 45.3% with, again, 1 complete response in the heavily pretreated advanced colorectal cancer cases. The median duration of response was not reached at the data cutoff. Next slide, Slide 14. Here is the waterfall plot. The light blue indicates the lower IHC expression level, and we can clearly see it segregating towards the lower end of the activity. The small orange triangle indicate prior treatment with HER2. They are distributed all over. And clearly, prior HER2-directed treatment did not make these tumors less sensitive to 8201. Next slide, 41. And now the spider plot documenting durability. Next slide, 42, progression-free survival, overall survival, too early and immature with noticeable sensoring still early in both curves. Next slide, 43. Here is a safety summary. Nothing atypical, except, obviously, the 2 cases of treatment-associated death with 1 Grade 5 ILD at the cutoff. Next slide, 44. Here is a bar graph with the most frequent and severe TEAE, again, a typical profile. And next slide, 45, which is more important, are the details on ILD. At the cutoff, a total of 6.4% with 2 cases of Grade 5. We did not -- and an additional one I mentioned in the footnote, we did not identify any exposure related or over risk factor. But obviously, the prior heavy exposure to topo I isomerase is of interest. Also of interest is that the study was conducted and the data cutoff was in August of last year, meaning that the ILD Safe Use campaign were mostly not yet implemented during the majority of study compounds. Next slide, 46. Now we are moving to the forward plan for 8201. Next slide, 47. And as has been presented before, an immediate effect of our collaboration with AstraZeneca has been the ability to mobilize more means, meaning more people and more dollars, and the ability to further hedge the risk. In practice, this has allowed us jointly to enrich the forward plan from 17 planned studies under Daiichi Sankyo sole sponsorship to 43-plus studies, covering 4 of the 5 major cancer killers, namely breast, lung, gastric and colorectal cancer. We also progressed the immuno-oncology combination and tumor-agnostic approaches. Next slide, 48. This is 1 of 2 slides covering in broad term our development plan with more detail than what had been presented before. Here, on this Slide 48 is for both breast and gastric cancer. I will not describe the full detail, but just draw your attention on the typical HER2 breast at the top, the Phase III post neoadjuvant as well as a first-line metastatic breast cancer, both monotherapy and in combination. For HER2 low critical are the first-line Phase III monotherapy in chemotherapy-naive patient and the first-line monotherapy in high-risk patient, more details will be provided later. For gastric cancer, here is a second-line HER2-positive Phase III study. Next slide, 49. Here, you have for lung cancer a critical first-line Phase III in HER2 mutation and a more comprehensive and more demanding program in HER2-expressing non-small cell lung cancer, which represents approximately 20% of all non-small cell lung cancer. It is more demanding because the impact is probably less dramatic than the one we've observed with HER2 mutation, again, attracting to the fascinating science of HER2 and the receptor biology in ADC. On this slide, colorectal also calls for subsequent monotherapy Phase II as well as Phase I combination trials, leading to further plan. Next slide, 50. We're now describing the key data delivery for DS-8201. First, the DESTINY-Breast02 Phase III study in HER2 metastatic breast cancer versus standard of care as an event-driven final analysis, which is now projected for the second half of fiscal 2021. The actual date will depend on observing the number of events. The more active drug, the later we observe the required number of demand. Next, DESTINY-Breast03, the Phase III in HER2 breast cancer versus T-DM1 is projected to read out in first half of fiscal 2021 for planned interim analysis. Next, in DESTINY-Breast04, a Phase III study in HER2 low versus standard of care is projected for second half of fiscal 2021. Lastly, on this slide, we're also very pleased to progress our PD-1/PD-L1 combination trials, with to-date a total of 72 subjects treated with at least 2 distinct immune checkpoint inhibitors, of whom 64 subject are in expansion cohorts. Tolerability is excellent, and we are very pleased here. Next slide, 51. We now move to the second DXd-ADC flagship, our TROP2 DXd-ADC, with a specific drug antibody ratio of DAR4-enriched part. We will spend some time here, in particular, to go beyond the data presented at ASCO and provide additional data and future plans. Next slide, 52. First, the data, as we presented. Next slide, 53. Here is the study design, exactly similar to that applied to the early development of 8201 and U3-1402, our HER3 ADC. A couple of key things in red box. First, on the left, you will note that subject enrolled in this study provided a fresh tumor biopsy at study entry that were not selected for TROP2 expression. Second, top right, this study has now enrolled nearly all the subject required to perform a robust dose justification over a wide range of active dose from 4 to 6 to 8 milligrams per kilogram for a planned total of 180 subjects across these 3 doses. As was done before for 8201, in particular, justifying the dose is the one place in drug development we do not rush and we do not need to rush. A strong dose expectation is an essential hallmark of successful alternative drug development in oncology. It was key for DS-8201. And in the previous life of mine, it was key for another major lung cancer drug, which one of my teams and my former company developed ultrafast and provided sweeping results exceeding expectations. I will not name the drug. Next slide, 54. Here are the key demographics of the 138 subjects for whom we have at least 1 dose and safety data on. So most important are on the bottom left in bold. 63% of subjects failed 3 or more prior line of treatments for the recurrent refractory non-small cell lung cancer. 88% have received and failed anti-PD-1 or PD-L1. 91% have received and failed platinum-based treatment. On the right, you will note that the median duration of treatment at data cutoff has been very short, with only 2 cycle or 42 days since the third dose. It means that many of the 50% or more of subject enrolled have entered this study very recently. Next slide, 55. Here are the safety result. The MTD, maximum tolerated dose, has been defined at 8 milligram per kilogram with mucositis or stomatitis as a dose-limiting toxicity as predicted by monkey study. The adjudicated ILD has been observed so far in 5.8% of subject with 2 Grade 5, both of early onset and Cycle 2 and Cycle 3, respectively. Let's -- next slide, 56. Now a few slides on efficacy. First, the overall response rate by blinded independent central review. Of the 138 subjects enrolled, only 85 were evaluable for efficacy, meaning, that they had at least one scan at 6 weeks after the first dose or had discontinued treatment or study before 6 weeks for any reason. It shows that a lot of subject had very recent enrollment. In this 85 evaluable subject, responses were observed at 4, at 6, and at 8 milligrams per kilogram, with a response rate so far at 27%. As you will see, it is in all likelihood underestimated. Next slide, 58. Here is the waterfall plot, color coded -- I'm sorry, the spider plot. Again, we see activity across the 3 doses. And interestingly, complete responses are now observed at each of the 3 doses, of 4, 6 and 8 milligram. You will note also that this spider plot is now on the basis of blinded independent central review. Next slide, 58. Here is the waterfall plot, which is color coded by dose. This, in particular, shows why it's so critical to continue to intensely study and justify the best dose because it's clear that all 3 dose contributes to the shape of the waterfall plot. Next slide, Slide 59. Here is before-and-after image after 6 weeks in the subject treated at 8 milligrams per kilogram. This subject had multiple massive abdominal metastases. So first scan at 6 months -- I'm sorry, at 6 weeks, showed a massive 41% shrinkage by RECIST. And this particular subject has been treated for more than 10 months. Next slide, Slide 60. This is probably a very important slide, which we did not present also. This is a swimmer plot for the 23 subject who had response and we mentioned earlier. On the X-axis is the duration in months from the start of treatment. Treatment is color coded with blue for 4 milligrams per kilogram, red for 6 mg per kg and green for the 8 mg per kg. The first time the response is observed is marked with a star. The most important observation on this slide is to note that 7 of the 23 responses, or 30% of all observed responses are delayed. They are not observed at the first scan at 6 weeks. The 2 of these responses are observed first at 3 months; 4 additional responses are first observed at approximately 4 and 4 to 5 months after the first dose; and even 1 response was first observed only at 6 months. Next slide, Slide 61. Here are the responses by TROP expression. You will recall that subject provided a fresh tumor biopsy at entry but were not selected for TROP2 expression by IHC. By looking at membrane IHC score on the left or by boxplot on the right, it is clear that TROP2 expression does not appear to predict response for -- to 1062. Next slide, Slide 62. We now move on to the clinical development plan for DS-1062. We are confident that 1062 can follow a fast-to-market development in non-small cell lung cancer in patients who have failed immune checkpoint inhibitors and platinum-based chemotherapy. On this chart, the green study on top is a study from which we just presented ongoing results. This study continues to accrue, as I said, and then we hope by year-end to complete accrual in the 180 across the 3 different doses. In this study, we will add soon a cohort in triple-negative breast cancer as TROP2 is, obviously, a validated target in this setting. Next is the first pivotal study post platinum and post I/O without actionable mutation, which will test 2 doses of 1062. That study, in our view, will be the pivotal study to support accelerated approval. Next is the intended post-marketing Phase III commitment to convert an accelerated approval into a full approval. Next is a Phase II signal-seeking study with 2 doses as well in non-small cell lung cancer, but this time with actionable mutation after failing TKI, chemotherapy plus/minus I/O. And of course, with very high priority, we're pursuing the immuno-oncology combination, starting with pembro combo, for which we just announced in collaboration with Merck. This is an ambitious initial program for the first and fast wave in non-small cell lung cancer, with a hint of test in triple-negative breast cancer. There's, obviously, a lot more to come once we've further progressed. So next slide, Slide 63. So what does this mean for 1062? First, we've observed a 27-plus percent overall response rate in unselected last-line post platinum in 91% and post I/O in 88% of subject. We've observed complete responses by independent review. Durability of response, continued tumor control and late-occurring responses are unique features in this particular setting. We are working with a drug which has an active dose range, which covers a widespread from 4 to 8 milligram per kilogram, which leaves plenty of room to define the best justifiable dose. There's a clear, fast-to-market path and focus on non-small cell lung cancer all-comer activating mutation post TKI and earlier line in combination with I/O as the obvious way to go. And I've said before, we also will now expand into the triple-negative breast cancer by year-end. Next slide, 64. Now moving to the Alpha strategy more briefly. Next slide, 65. The Cancer Enterprise CE-Alpha Strategy embraces the concept of a string of pearl and maintain a healthy development flow of high-value drug candidate. We want to identify and rapidly scale up and accelerate any high-value drug candidate from our outstanding research capabilities. First and foremost, we actually want to maximize the intrinsic quality and value of the pearl and progress our multiple ADC platforms. We will maximize the value of the portfolio beyond the 3 big ADC by performing a rigorous assessment using model drug development principle and the 5R criteria, by rigorous prioritization and by rigorous value generation. Next slide, 66. The strategy will use a hybrid model deploying internal and external collaborative development options, in particular, through academic research organization, ARO collaboration. We have already made significant progress by bringing now our fourth in the full, and now most recently, our fifth DXd-ADC in clinical stage: B7H3 last fall; and more recently, GPR20 ADC, our first-in-class in GIST. Next slide, Slide 67. Now let's move on the R&D transformation, slide 68. We were proud to have announced an exceptional agreement with Syneos Health, a world-class, world leader, premier CRO. And together, we will realize the promises of our ADC pipeline through exceptional study delivery operation. The key characteristic of these collaborations are listed here. So key for me is about combining an early and personal engagement of the CRO with our own Global Project Team, including for Daiichi Sankyo governance interaction, so that we break the service provider mentality and culture and we create and foster a culture of trust, transparency, rigor and collaboration. We'll also capitalize on Syneos Health's very unique focus and success at the critical investigational-type level. Next slide, 69. In summary, we expect to collaborate and leverage each of these strengths to deliver the 3 leading ADC assets faster. Next slide, Slide 70. Now moving to commenting on the context of the SARS-CoV-2 pandemic and how it affects our current R&D operation. Next slide, 71. On this slide, you see bar graph for the key oncology studies, each bar representing monthly accrual from January 2020 until May, May being an estimate when this slide was drawn. We definitely saw an impact in accrual with site redirecting efforts to handle the COVID-19 impact on health care system locally. We are projecting now -- or recovering to level close to prior pandemic as the care of patient with advanced cancer cannot substantially be more deferred. Next slide, 72. Here is a more qualitative and quantitative assessment. Approximately 70% of sites are currently closed for on-site monitoring. 1/4 of the sites has suspended enrollment. And on that, only 1.4% of subjects so far are at risk in terms of compliance with clinical requirement and treatment dosing -- with respect to treatment dosing. We currently estimate a minimum delay of 1 to 2 months for completing study. We, obviously, track COVID-19-related protocol deviation. None of this deviation to date have substantially affected any study integrity. We also have longitudinal serum banking to assess retrospectively the occurrence of the COVID SARS-CoV-2 infection as such an infection might affect the subjects' sensitivity to certain AE. And so far, the rollout of our next wave study is not substantially affected. Next slide, 73. We're now moving on to forward-looking step in news. Next slide, 74. Here, the top part of that slide refers to previously mentioned. What I've not mentioned so far is for U3-1402, our HER3 DXd-ADC. We're planning update at ESMO in San Antonio for lung, EGFR and breast cancer, respectively. We also have news coming for Valemetostat or DS-3201, our dual EZH 1/2 inhibitor. With the start of a pivotal study in the second half of fiscal 2020, we expect an approval in Japan for Axi-Cel, the CAR-T relicense for Japan from Gilead Kite in B-cell lymphoma. We also expect to announce the submission in Japan for DS-1647, our viral vector for malignant glioma in first half of fiscal 2020. This completes our formal presentation, and I believe we will now take questions.

Thank you, Antoine. We will stop here and go to Q&A.

[Operator Instructions] The first question is from Mr. [ Richard Wagner ] from Wolfe Research.

Can you hear me?

Yes. We can hear you.

[ Richard Wagner ] with Wolfe Research. My question, looking at your clinical development plan as outlined, I see a preponderance of settings where in HER2 could be employed as a single agent. But in some cancers, first-line non-small cell lung cancer would be one or gastric cancer, another multi-agent chemotherapy is the standard of care. So my question relates to in HER2 as a potential new standard in those settings, would that require a combination with chemotherapy and how feasible do you believe that would be? And second, if not, does that materially raise the threshold for success in those settings if in HER2 would need to displace multiple chemotherapeutic agents?

Antoine, please.

Sure. Thank you. Thank you for the question. I think, yes, this is an important question. We do not believe that we necessarily have to combine with chemo, especially in HER2 mutation, given the level of activity that we've observed. But we also do believe that, especially in the HER2-expressing, where the response rates which we saw in Phase I was less, we probably will need to combine with chemo. No. There's nothing wrong with combining with chemo. We are actually doing this right now as we speak, so combining with chemo, combining with the immunotherapy and other agents in breast and other tumor types. And the reason for this is because the DXd technology is, in essence, a smart chemo therapy. And the principle of combining a smart chemotherapy with another chemotherapy agent is actually the same as what we do when we combine typical chemotherapy agents. So there's really nothing wrong in principle in trying to combine the 2. There were more uncertainty in combining the DXd technology with immuno-oncology checkpoint inhibitor, especially because of the ILD. And this is why we started this extremely early, and we are pleased to have observed what we observed far. We didn't give detail, except to say that we're very -- we were very much into the expansion phase. So we, obviously, didn't see a signal preventing us from moving into large expansion phase. So really, the biggest challenge probably was in combining with I/O. But obviously, not every chemo combination is feasible. That is the principle of chemo combination. We know how to do this. We've done this for 30 or 40 years, combining chemo with chemo. So combining chemo with a smart chemo is probably very feasible. But you're right. I mean some setting will require chemo combination, but not necessarily always. And for instance, in breast cancer, first line, we -- one of the options we're pursuing is a bold monotherapy approach versus the triplet of trastuzumab, pertuzumab and docetaxel, which is standard of care. And the reason is because this triplet of HER2suppression plus chemotherapy is essentially what we do, but more effective -- potentially more effectively with our own drug.

The next question is from Ms. Melissa Gallagher from T. Rowe Price.

It's Melissa from T. Rowe Price here. I just wanted to ask a little bit more detail about DS-1062. And obviously, you've outlined to us that around 30% of the responses that have eventually been seen so far were delayed responses. Are you able to give us an indication of the response rate in those patients that have been followed for longer? So for those that have had at least 2 scans. So maybe we might get some reassurance that the actual response rate, the final response rate we see could be higher than that currently seen for the data set. I don't know if you have any data you can share on that.

No. We did not get the data like this data. The data we have presented -- the data cutoff as of March of this year, we did not subset the analysis for patients who have at least 2 tumor.

Okay. And then on the ILD cases that have been seen for 1062, have you disclosed that by dose level since you updated the number of patients?

No. We have not. And the top of my head is that I honestly don't have that information in my head. What I do know is dose is still definitely a factor in frequency of ILD that is true across the different ADCs. So it's more actually more than dosage exposure, which is because you have variability, you have PK variability, inter-patient variability at the given dose. So exposure is definitely a factor in the frequency for the severity, the severe event. Grade 4, Grade 5 are too rare to really know what are the predictive factor other than the prior intensity of the product treatment. So -- but no, we have not -- I honestly don't know. But on the flip side is that is the ability to justify those and pick a dose across a very wide range. If you recall, for 8201, we were playing between 5, 6 and -- so we only had 1 milligram per kilogram between the 2 doses that we were testing, and then we picked one. We just picked one. Here it is -- we have a much, much broader range. And even for 8201, we -- the choice of the dose, which -- I mean, the reasoning and the evidence behind the choice of the dose, which we published at San Antonio a year before, not last year, but the year before, was actually quite complex because it looked at exposure. It looks at the death response, frequency response, time to response, durability response, PFS as well as, generally, AEs, general AEs, AEs at ILD frequency severity. All of that was taken into account, and this is why we settled on 5 [ mg. ] Here is -- for 1062, we have an active very much wider active dose range. And we just need to accrue more data to then consider which dose we will bring forward. So I don't know if I'm sure. But an obvious question is which 2 doses will you bring or will we test in the 2-dose studies that drive this one. I don't know, it may be 4, it may be 6, it may be 6, it may be 8. It may be actually -- we can do -- we will do in clinical testing. So we may have 5 and 7, for instance, as 2 doses being tested. So we have a range of different doses that we will be able to test in these studies.

So just on that point on, Antoine, I mean, obviously, you've detailed that you're expanding the 4-milligram and 6-milligram dose cohorts. Presumably, that expansion has already started. Obviously, the data cutoff we've seen of March, presumably, you've recruited more patients into the cohort already. I'm just sort of conscious of timing and with COVID-19 slowing the situation down. Presumably you will have enough data to still make the decision on which doses you want to take forward into that Phase II before the end of the financial year.

Yes. That is correct. So we will have enough data to at least bring 2 doses. Will it be -- yes, but we also continue -- we will continue to accrue more data because you're right. I mean, the COVID-19 is just slowing down at least by 1 or 2 months and potentially more as it carries on. We just don't know how the reopening will affect. So at the release, we have 1 or 2 months now. So -- but we will have enough data to carry -- to launch the Phase II, but we also continue to commit more data on the 4, 6 and 8, and then it will be in clinical research.

The next question is from Mr. [ Mu Chen ] from [ Genco Partners ].

First, congratulations on this very exciting piece of data both on HER2 and DS-8201 and 1062. Second question about TROP2. So 8 milligram, compared to the abstract, we actually added 24 patients. Just want to get a sense sort of like how long on average those 24 new patients can follow up.

Sure. Yes. So you're right, we've added a lot of subjects very recently. And I think the exact number is the number of cycle initiated in -- at 8 milligram, which is only 1 or 2 weeks, I think, 22 or 23 subjects. I only had that. So we're very, very front-loaded in subject to have very minimal follow-up.

Okay. Got it. And also I noticed, compared to the abstract, you added like in the result -- basically, you added like 24 patients in the 8 milligram, probably 1 -- just 1 or 2 subjects in 6 milligram and 0 in 4 milligram. But you mentioned all the 3 doses, you may consider sort of pushing into the to Phase II. But based on the subject added, it seems like you didn't prioritize 4 milligrams. Is that the right thinking?

Initially, we did not -- so the initial thinking was to expand at 8 and 6. The very first thing was to expand at 8 because the 8 was defined as the entity. This is why the cohort at 8 is the leading cohort, because that was a natural thing to do, but we did have activity at 6 and 4, and that's why 6 was also pushed beyond. And until we realized that we had visible and substantial activity at 4 and it was probably a wise thing to further expand our experience at 4 mg per kg. So this is why all of 3 cohort expansions are actually not simultaneous because they were triggered by -- as we were observing the higher dose effect. So this is why we have much more subject enrolled at 8. We are enrolling now at 6. It's almost completely enrolled at 8 now, the target number, and then we're enrolling now 4 and 6. And back to Melissa's question, which is, will we have enough here? We will have enough, at least for 2 of these doses, to make a decision and will continue to accrue for 4, 6 and 8.

Got it. Got it. So probably in the next readout, we will see more patients in the 4 milligram. Is that right?

Yes. Probably, yes, mostly in 6. So the next readout is -- as we're planning, but again, with all uncertainty attached to the SARS-CoV-2 pandemic, we're planning the next -- at World Cancer -- World Conference on Lung Cancer in January of next year. That's our next readout for this particular project. And by then, we definitely will have more at 8 and 6 [indiscernible] some at 4. Maybe -- still, it may be still a little too early for 4 to have a full 50 patient that fall within a follow-up.

Next question is from Mr. Hidemaru Yamaguchi from Citigroup.

Can I ask a question?

Please.

Okay. The first one is the 1062. It's too early to say, but can you give me some background, the patients, who is in this setting? If you get anything, if you don't get anything, what is our kind of expected duration or OS for those people? Then you're getting only a few -- 27 of the patients -- or 23, I think the patient, which is showing some spider plot, but can you give the duration of response from those charts, if you can? That's the first one.

Sure. So you're right, it's too early. And Slide 60 shows you how many hours you have. And so you have a median duration of these responses, which is in the order of 6 months plus, because of actually below the 6 months, you have most subjects continue to carry on, except 2, if you remember well. So obviously, it is too early to give a directional response from the first dose. So it's just -- it's, unfortunately, too early to really speculate what it is. At the lower doses where we had responses, you can see at the top of that chart is that it carries into the 10, 12 months after starting the dose [ over ]. Again, you have to remember that the duration of response is typically computed from the first observed response. So you actually have to take it from the start onward, not from the start of treatment onwards. The start of treatment onwards is what [indiscernible] I mean. In patients, we have, I mean, remember that the first line for lung cancer, unselected. The combination of pembro chemo gives you 48% response rate. And the chemo alone gives you something like 28 -- something like 19 response rate, platinum, platinum pem gives you 18, 19 response rate. So the difference in the 2 in first line is 28% response rate. 28% response is what we observed after pembro or [ mutated part ]. Yes, it's a pretty solid trend. I mean, there's a lot. I mean, if you -- obviously, the top 2 IMU 132 was approved on the 33% response rate in triple negative. Olaparib was approved, and I got it approved a few years back in fourth line lung -- ovarian cancer in 30% response rate. When you have that, I mean, not everything can be 60% response rate.

Okay. That's right. Okay. Can I have another question? You're talking about tucatinib, which oral HER2, which is now approved from Seattle Genetics and getting into the second line, the combination, I think the Herceptin and other oral agents. When you talk about the tucatinib, we may change the kind of a treatment landscape in the second line. From your perspective, what do you need to do in the near future or with the HER2?

So we need to do a few things in the first line, second line. Obviously, we -- the very first thing to do is to give a formal comparison versus T-DM1 because that's an anchor in that setting, and we will have that readout. The second thing is to, obviously, is to combine with tucatinib. So we are planning to do that combination, definitely planning to do the combination. So that is, obviously, in the plan. We did not describe this in the clinical development plan slide because that's an obvious combination that we are going to do. So we'll give more details once we do it, but it's imminent. The first thing to do is to actually look at the first line, which I've described. And in the first line, we obviously are approaching this business where the collaboration with AstraZeneca is allowing us to take -- to have more efforts. Because in the first line, you can approach it in 2 ways. One is to say in HER2, in and by itself, is a combination of in HER2 suppression for the trastuzumab component, and a smart chemotherapy, which is obviously for more powerful than tucatinib. And I know tucatinib. I put it on the market. But it's far more. So one line of thinking, which AstraZeneca and other leaders have subscribed to is just to go head-to-head monotherapy, so in HER2 versus a triple combo. And -- but it's -- the bar is extremely high because the triple combo is extremely effective. The other approach is to do combination. And we, obviously, will do combination. And then one of the key question is, do you need to double to suppression pertuzumab -- trastuzumab; or pertuzumab in HER2. Obviously we don't need the docetaxel component. So this is also how we are approaching this first line. And then finally, as I said to begin with, is that the tucatinib combination is a very obvious thing to do because the tucatinib is, obviously, a great drug, a fabulous drug, to be honest. Overall, our experience with patients who have CNS metastasis at entry is actually quite impressive as well. But combining the 2 mechanism is the obvious thing to do. So we are doing this as well. So it's really a multipronged approach in how do you redefine the -- or how we could redefine the HER2 metastatic breast cancer first and second line. I'm sorry. It's a long answer, but that's how we [indiscernible].

The next question is from Mr. Tony Ren from CLSA.

Could you hear me?

Yes. Thank you.

So Antoine, so yes. Congratulations on the very strong data. So my question are in HER2. So at ASCO, I think we see that from multiple studies, basically, the safety profile shows that maybe somewhere between 6% to 14% of the patients would develop interstitial lung disease. So my 2 questions are, first, so perhaps with more clinical experience with the agent, are you guys seeing ways to manage this AE in interstitial lung disease? And also my second question is related to this. At theory, is there a way to identify who is going to develop ILD?

So thank you. Thank you for your feedback, and thank you for this question. I mean, yes, obviously, we are working very, very hard on both. So to your first question, is there a way to improve management? Yes, there is. Because we continue to see cases where the early sign and symptoms are missed and treatment is continued or and the steroids actually are initiated and the steroid treatment, high-dose steroid, is obviously a standard of care on ILD, but the steroids treatment are initiated and then tapered down too quickly just because -- and so clearly indicating more about the need to be serious about ILD is key. And it's very difficult because these patients -- and I met patients, with the 8201 and had the life transformed. I still remember back in January, the last trip I did, I was in France. There was an 82 years old lady who had metastatic breast cancer and stage-failed everything and was dying. And she was receiving and chatting with me, 14 -- I mean, 14 months, so 16 to 17 infusion of 8201 would transform a life. So patients and doctors push to continue treatment despite minor sign and symptoms continues to be an issue. So we really have to continue to improve the awareness that you don't play with an ILD. You just don't -- because ILD can go down very rapidly if you continue to expose patients and continuing to expose patient with the drug is probably really not good. And it's also very important to be -- to deliver treatment intensely. Now to your second question is, can we predict? Yes. We have more experience now. What seems to be emerging more than what we had seen before is the prior treatment. So the degree of prior treatment seems to be more and more a factor-predicting ILD, which we did not see initially because probably we didn't have enough patients, which is actually reassuring, because then that means that when we go into the earlier line and the early breast cancer, early disease, it's -- it reduces the risk and be better -- better measures, better management measures also reduces the severity. And thus, we will -- it's reassuring. And I can tell you, I mean, the -- the different curability groups in Europe, in the U.S, they are desperate and very keen to bring this drug into the early breast cancer, adjuvant, post adjuvant setting. So we are actually not very extremely keen to progress it. And we are keen to progressing. So we are very, very keen to test the drug in the early breast cancer stage even though a lot of these patients are potentially -- there's still a lot of patients who are at high risk of relapse and are willing to take highly active drugs to try to improve their outcome.

Yes. In breast cancer, I think that's a prime example because we -- CLSA hosted Roche or a virtual conference a couple of weeks ago. And so Roche's feedback in HER2 is largely supportive, complementary, but they also said that breast cancer is an indication with very good prognosis, right? I mean, you cannot have a 14% pneumonitis, right? So -- and I also think at San Antonio, we heard that physicians was saying, KOLs were saying that people do not respond to steroids when given -- right, in the face of ILD. Is this what you guys are seeing as well?

So number one, I'm not commenting on Roche's comments generally speaking. What I can comment is, number one, it's -- when you have an IDSF of 83%, 84%, 85%, it means that 1 in every 7 or 8 women are actually going to relapse, and that -- so this is, obviously, when you're a 95-plus percent, this is -- the majority of patients are cured and then none are -- so you're trading with a risk -- really your risk, which is much, much lower. But when the risk is in the 15% range, it's actually very close to the risk of adjuvant breast cancer therapy in the early 2000s when women were given, in particular, Taxotere combined with tricyclines, for instance. I still remember filing this and you had treatment-induced deaths through AML and MDS, which were exactly in the same order of 1%, 1.5%. And still, women are at that time -- and now we see that because the adjuvant surgical alliance, the NSABP, the GBG in Germany and all the big breast cancer group are desperate to progress the testing of ENHERTU in the context of the adjuvant breast cancer high risk. And we -- obviously, high risk, high risk means that you have an IDFS which is greater than 10%. So we agree that when your IDFS is very few -- small low single-digit and the risk of 1%-ish risk of death because of ILD is a nonacceptable rate. So we agree with that. But having 10%, 15% ILD leading to treatment discontinuation, but a risk of severe ILD, which is obviously only 1% or less, is perfectly acceptable seemingly to these big large collaborative group as well as patients themselves as we hear. So we don't necessarily agree that early breast cancer is not an option for ENHERTU. Early breast cancer is not cured by far, but there's still many women dying or relapsing from that early breast cancer deserve to be treated. And then your question about what is the -- the corticosteroid not being active. I don't know who says that -- I mean, the common practice and good medical practice says that ILD is sensitive -- and it's not always sensitive, but it's definitely sensitive to intensive steroid treatment. But clear is that, in many cases, doctors tend to shy away from intense and/or make it too short of a duration, and that's where it doesn't work or it works much less. But when it is intense and of the right duration, it actually does work. It does not work always, but it does reduce the worsening of the conditions. So I don't agree that it's not -- but what all of that what shows is that there is an education needed about the risk as well as education needed about the potential and the potential benefit and risk and the need to do the right thing. Is that helpful, Tony?

Yes, yes, yes. Could you just elaborate a little bit in very quickly on the dose and the duration of corticosteroids?

No. Not now. I'm sorry. I'm just not prepared. I know we have a detailed [indiscernible] on how to do it. I'm sorry. I want to [indiscernible].

[Operator Instructions] The next question is from Ms. Caroline Stewart from Bloomberg.

Just a quick couple of questions on 1062. The fact that the responses are kind of irrespective of TROP2 expression, why is that? That's my first question.

If I had the answer, there's absolutely -- that's what I said is that there is a fascinating new biology of the role of receptors in ADC, which is actually not limited to 1062. We don't know why in HER2 or HER2. We don't know why an HER2 mutation in lung cancer leads to a response rate in excess of 60% as we have observed and reported. Whereas in colorectal, it's the IHC 3+ which carries the effect. Whereas in lung cancer, the HER2 IHC has actually less response rate compared to mutation. And then you go to HER3, as I described in -- at R&D Day, we know now that the ADC interacts with the HER3 expression. You also have good dimerization, which modifies the receptor expression and the trafficking, which leads to variability in internalization and tolerability. For TROP2, there is absolutely no clue. We just don't know. We don't know what TROP2 is. We don't know the natural ligand of TROP2. We -- when we look at TROP2 total protein expression, total RNA [ sway ] effect, we looked at cytoplasmic. We look at all the different things, it's just nothing predicts responses. And more importantly, I think -- for TROP2 is -- and that's not new. I mean, if you look, the IMU 132 was actually approved by FDA without a companion diagnosis. Even though it's a TROP2-targeted agent, it doesn't require a selection of patients on the base of an expression of a marker. So we're looking at the market. We are actually working very hard to try to understand something, which is also very unique, which is, you may have noted is that in the response rate for 1062, we also have -- you can see that on the spider plot. It's actually, we have a number of patients whose best response is progression. And that is or you also see that in the waterfall plot. And that is not typical for the DXd platform. So we're trying to understand if we can also predict that as opposed to trying to predict the responses on the basis of our TROP2 expression. Bottom line, it's just we don't know. We're working very hard. We have collaboration, active collaboration with a major transitional medicine cancer center in the world to investigate this. But it's a fascinating big question that, because of that pharmacology, now we can have a product question, but I don't have the answer.

So with the -- okay. So with the -- again, with 1062, for the 1 -- it looks like 30% or so and 1/3 of patients who are late responders, are they -- and I'm sorry, I didn't -- I'm not -- toggling back and forth between the swimmer plot and the spider plot. But the patients who respond late, are they just like patients who are border line and don't qualify as responders until late? Like are they improving, but not responders until the later month? Or is it -- or how does that look?

Yes. It is that. It's a continued regression of the tumor size up to the threshold of reaching the 30% threshold from baseline value.

Okay. But they're actually -- they're close for a while before they actually hit that point then, right?

Yes. Yes. So they're not early progressors and then suddenly responder. Tumor control and tumor control, tumor control and then they [indiscernible] very different response.

Okay. And then lastly, just in HER2. Since the U.S. label is not restricted to post T-DM1, are you -- I don't know if you can answer this. Can you -- is there a use being -- is there use basically of second line? Do you see a proportion of the use already of the drug in second line rather than post KADCYLA?

Don't quote me on that as a positive answer. I do think we see, but I'm not positive about that. I also do know that insurance company across the country now have accepted off-label in lung as well as in gastric. I know of 2 -- at least 2 cases where we -- where there were requests for early access or permission to use access to the drug. And then both requests were actually suspended because the insurance company had agreed to provide the drug even though it was off-label. So we suspended the process.

Okay. And then just -- actually, just last quickly. Do you expect any -- I mean, the results were so strong in gastric, but that was an Asian population, you wouldn't expect a very different response in kind of non-Asian patients would you in the gastric setting? I know you had other gastric study that's ongoing right now.

So we do not expect that like in science, you have to test your expectations, but you're probably right. We do not expect. But there are 2 reasons for this. Number one is that this gastric study we just published was in Japan and Korea. We have [indiscernible] and I know [indiscernible] pretty well. And I actually did the elaborate study within, which was a [indiscernible] almost positive studies. And the second reason is because the majority, I don't have the exact number in my head, but probably 2/3 of the subject in the Asian gastric study did not have the primary cancer resected, unlike pure Japan gastric cancer with a lot of patients are screened for gastric cancer and have the primary cancer resected, and they have a very different tumor if and when they relapse. There was a primary tumor in place with the majority of patients and which makes them far more consistent with what -- the experience is in the West where -- in the West, the very vast majority of patients do not have primary resection. We also know that the West gastric study is an open-label study. So we do know that patients are responding. That -- but I don't have enough -- and I will not give you a number until we have a read. So do I expect the result to be portable from one population to the other? Yes. And that's probably why the FDA was actually very keen to see the data and gave us pretty quickly the breakthrough designation because they also came to the same conclusion, which is the likelihood of the result to be reproduced in the West is high. What we're doing now is how to bridge that, the development between East and West, and we are actively discussing with the FDA and EMA on how to bridge the eastern experience with the West population and make it generalizable to the West. But that's a long answer. The short answer to your question is, do we expect to produce? Yes we do, but we have to verify that.

The next question is from Ms. Melissa Gallagher from T. Rowe Price.

I had a couple of follow-up questions on 8201. And I think you've sort of touched on this question in your last answer. Obviously, you have breakthrough designation for colorectal and small cell -- non-small cell lung cancer in the U.S.

No, no. Gastric, not colorectal.

Sorry, sorry. Gastric, my mistake. But I'm not entirely clear which data set you plan to file on. Because obviously, in Japan, you've been able to file. But I think when you were sort of mentioning bridging studies, I'm assuming we may not have to wait until the Phase II studies report? Or I mean, how should we think about what date you can actually file in the U.S. for those 2 indications?

So we are, obviously, actively discussing for both indications. We're actively discussing with the FDA on what they would like to see. So for lung cancer, we have an ongoing study, which we started, and we -- which will supplement what we've seen now and when -- so there are different scenarios. We have the planning analysis. We have the plant analysis. They -- ASCO would be willing to suggest that they want to see an intermediate payer cut and so on and so forth. So we really are actively discussing with the FDA on lung cancer, what it is that would constitute an appropriate data package. So it's too early to say what is the basis because, bottom line, you need enough subjects, enough durability, enough to demonstrate durability and enough -- for gastric cancer, it's a different question. The gastric cancer is how do you predict -- I mean, obviously, we have a firm overall survival. So clinical benefit is proven. The question is how do you generalize these results to the gastric patient population in the West or in Western Europe and in the U.S.? I mean, U.S., as the FDA accepts foreign data, license drug purely foreign data. So it's not that it is a foreign data. It is how do you bridge. And you have a few things. One is you have a question of actually PK. You may have noticed that the dose of gastric is 6.4 versus a dosing -- versus 5.4. The reason for this is that we knew from the trastuzumab experience is that both in the West and in Japan, actually, trastuzumab PK is modified in gastric cancer. And you achieved the same systemic exposure at 6.4 compared to 5.4 in lung cancer. But you still have to verify that in gastric cancer in the West or higher construct for ENHERTU as opposed to purely trastuzumab. So that's one bridge that we definitely have to do, but we already have the data for that. The second is the bridge on -- are the patients with respect to resection baseline similar enough to Asian experience to make the results predictable. Again, this is something that, how much do you have to submit so the FDA files and then accept and then says that this is post-approval commitment versus pre-approval requirement. This is all the subject of negotiations. And they just -- we want to do the right thing. We don't want to push a treatment which is not properly documented in the West. We're just want to do the right thing. And that's part -- that's the benefit of [indiscernible] and a good relationship with the agent.

And can I squeeze in one question?

Sure.

So also on 8201, if I look at the colorectal, the lung and the gastric data, in the colorectal population, it looks as though the sort of frequency of dose reductions and dose interruptions appears to be less despite being a fourth line population. Is there anything we can conclude from that? Is it to do with the status of the patients and their health and how they respond to it there? Or is that -- just don't overinterpret that?

I would be careful. That's overinterpreted. It still is regular small data set. I would we be careful in that. No. No, sure. I mean, I'm giving you -- I mean, I'm careful about not talking and interpreting this. It's still too small. I mean, 150 [indiscernible] worth of data.

[Operator Instructions] As there are no more questions, we will close the Q&A session.

Okay. Then we would like to close today's conference call. Ladies and gentlemen, thank you again for your interest and participation. Stay safe, everyone. Goodbye.

This is the end of the call. Thank you for your participation.