Thank you for waiting, everybody. We'll now start Daiichi Sankyo's ASCO 2025 highlights. This is Ken Asakura from Corporate Communications, and I will be the moderator today. First, about the language for this event: presentation will be in English, and Q&A will be in both English and Japanese. Simultaneous interpretation is provided, so please click the interpretation icon at the bottom of the screen and select either English, Japanese, or original audio. When original audio is selected, you will hear the original sound. English presentation slides will be shown on the Zoom screen as well as live streaming screens. The slides have been uploaded to the IR presentation material page of the Investor section on our corporate website, so please view or download as needed. Our presenter for today is Dr. Mark Rutstein, our Head of Therapeutic Area Oncology Development. We will take questions after the presentation.
Mr. Hiroyuki Okuzawa, our President and CEO, will open the meeting and will join the Q&A session. Please note that this event will be recorded. With that, let's start the event. Okuzawa-san, please.
Hi everyone. I'm Hiroyuki Okuzawa, President and CEO of Daiichi Sankyo. Thank you for joining our scientific discussion highlighting our data presented at ASCO 2025. At Daiichi Sankyo, we remain firmly committed to our 2025 goal of becoming a global pharma innovator with a competitive advantage in oncology. This year, ENHERTU is entering a new phase of growth with three key breast cancer trials, DESTINY-Breast09, DESTINY-Breast11, and DESTINY-Breast05, aiming to expand treatment into earlier lines. Today, we are pleased to discuss with you DESTINY-Breast09 data as the first among these three trials. For that way, TROPION-Lung02 presentation, including QCS data, is offering you an early insight into its product value ahead of AVANZAR readout later this year. Our deliverables extend beyond these two products. We continue to develop innovative therapies for patients in need. Now, I would like to hand it over to Mark Rutstein.
Thank you, Okuzawa-san. I'm Mark Rutstein, Head of Therapeutic Area Oncology Development at Daiichi Sankyo. Very pleased to be here. In fact, excited to have the opportunity to share with you data from ASCO. Indeed, the theme of this evening is about the quest to change standards of care for patients with unmet medical needs. This evening, I will walk through key data from ASCO, but before I do that, I'll talk a little bit about our R&D strategy. Could we have the next slide, please? Great. Thank you so much. Just a reminder of the overview of our R&D strategy, and that included expanding, which is to pivot from the focus on lung and breast cancer into additional tumors, but also earlier lines of therapy.
At this ASCO, we have the pleasure to present data from DESTINY-Breast09 in an earlier line of HER2-positive metastatic breast cancer for ENHERTU. We also, approximately one month ago, announced a positive press release for DESTINY-Breast11, a phase III trial for ENHERTU in HER2-positive neoadjuvant breast cancer. This represents association with our expand strategy into earlier lines of therapy. Later this year, particularly in the second half of fiscal year 2025, we should have the top line results of the high-risk adjuvant HER2-positive study, DESTINY-Breast05. Next slide, please. This is a map of our breast cancer program, and you can see it is very extensive.
Just to orient, at the top for HER2-positive disease, we have the frontline DESTINY-Breast09 trial we have read out at ASCO that we will discuss together, and the other early phase trials I just referred to, Breast11 and Breast05. We also have a broad program for Dato-DXd in triple negative breast cancer. In the first half of fiscal year 2025, we expect to read out TROPION-Breast02, which is listed in patients who are not eligible for immunotherapy. Of course, in dark blue down the bottom represents our approval, our approved indication of Dato-DXd in Breast01 study in hormone receptor positive HER2 negative disease. We continue to study on the far right of the slide in green, we continue to study HER3-DXd in this disease state as well. Next slide, please. This is a map of our lung cancer program.
As you can see, we have a number of trials for Dato-DXd, our TROP2 ADC in non-small cell lung cancer. We have, of course, ongoing and accelerated approval submission based on TROPION-Lung05 in the upper right of the slide, and we will have PDUFA date in July of this year. In addition, we have several ongoing front line trials, including TROPION-Lung08, Lung07, and AVANZAR. Notably, AVANZAR is expected to read out in the second half of the calendar year of 2025, and I will show data today from phase I-B study T-Lung02, which is supportive to these front line trials. In addition, we have our B7-H3 DXd ADC I-DXd represented in yellow to the bottom right and a small cell lung cancer program that's ongoing.
In addition to that, of course, ENHERTU is approved in segments of lung cancer, including mutation positive disease and late stage HER2 positive 3 plus expressing lung cancer. We have a front line program as well, not only with DESTINY-Lung04 in the mutation positive population, but also a newly announced trial, DESTINY-Lung06 in the HER2 expressing population. We continue to study in green also HER3-DXd. Next slide, please. Now I will describe the highlights from ASCO. Next. This is what I will cover, and indeed, it is quite ambitious to cover this content. In some cases, I may not cover slides completely, but please feel free to ask any questions that you may have. If we could go to the next slide. I want to first start with ENHERTU and DESTINY-Breast09. Next.
DESTINY-Breast09 is a frontline randomized global trial in HER2-expressing metastatic breast cancer. This is data that we present here that was presented at ASCO also today for the T-DXd and pertuzumab arm. Next slide, please. The design is here. Patients who had received no other prior systemic therapy for metastatic breast cancer, HER2-expressing, they were randomized to T-DXd and placebo, fundamentally T-DXd monotherapy, T-DXd and pertuzumab, and THP, keeping in mind that THP has been an entrenched standard of care for approximately a decade, and that includes taxane, Herceptin, and pertuzumab. You can see the primary endpoint, progression-free survival, and key secondary endpoints, including overall survival. Next, please. Today, we will present data from the planned interim analysis with the data cutoff present on the slide for the T-DXd plus pertuzumab and THP arms. Next.
At this interim analysis, there was a maturity of approximately 38%, and we showed data from the interim analysis for, as mentioned, T-DXd plus pertuzumab. The T-DXd monotherapy arm remains blinded until the final PFS analysis at a later date. We do show overall survival here, and this is with 16% maturity. Next, please. The patient demographics are here. I would highlight the majority of patients IHC 3 plus, half the patients with de novo disease. In addition to that, you can see demographically a global trial with approximately half the patients enrolled from Asia. Next slide, please. Here are the primary endpoint results. T-DXd plus P showed 40.7 months of median PFS. That's more than three years.
That regimen also improved median PFS more than a year compared to the decade-long standard of care of THP, a very clinically significant result with clear evidence of separation of the curves, hazard ratio 0.56, representing a 44% reduction in the risk of progression or death. Next, please. Similar pattern for investigator assessment. The previous slide in the endpoint was based on independent radiological assessment, but investigator assessment corroborates that result as well. Clinically significant PFS with significant separation in the curves. Next slide, please. Here are the subgroup analyses, and basically the statement would be across various key subgroups, there is consistency of benefit in terms of progression-free survival. A few key subgroups highlighted there, including hormone receptor status, and you can see evidence of PFS benefit regardless of hormone receptor status, as well as various other subgroups on the slide. Next, please.
Here is the response rate, and you can see an 85% response rate for the combination regimen of T-DXd and pertuzumab. Notably, 15% of patients had a complete response, which means no evidence of radiologic disease on scan. I would also note to the right that for those that responded, there was a median duration of response of more than 39 months. Quite significant observation. Next slide, please. Here is the overall survival with 16% maturity. You can see a very early positive trend, but there is a lot of censoring in the curves, as expected with 16% maturity. Though I do not show you, what was presented today was PFS2, which is progression or death on the next line of treatment with a hazard ratio of 0.6, which is supportive and gives a window into what overall survival might look like when more mature. Next slide, please.
Safety is here. Overall, the safety profile of TDXD is consistent with what we know for this drug in metastatic breast cancer. You can see in a comparison compared to THP, there was a higher rate of dose interruption, a higher rate of dose reduction with TDXD plus P, and there were a few more episodes of related fatality. As I'll show later in the slides, there were two fatal ILD events. Overall, this is a safety profile that shows nothing particularly new and nothing surprising and is consistent with what we know about this drug in metastatic breast cancer. Thank you. Next slide, please.
Here are the adverse events most common, and I would simply say that they are consistent with what we would expect based on what we know about TDXD in the past and based on what we know about consistency with Pertuzumab and TDXD as a regimen. Next slide, please. Here are the adverse events of special interest, and I noted earlier there were two fatal ILD events. The total rate of ILD was about 12%, and overall, again, consistent with what we may expect in a metastatic HER2 positive breast cancer population with ENHERTU. Left ventricular dysfunction down below, there was one grade four event, and actually there was 11% versus 7% for TDXD and P versus THP. Next slide, please.
In summary, TDXD plus P demonstrated a statistically significant and clinically meaningful PFS benefit, superiority to THP, which has been an entrenched standard of care for approximately a decade. The other endpoints that were looked at, including response rate, durability as well, were very supportive, and the early overall survival data, though immature, suggested a trend, an early trend, safety being consistent with the known profiles of TDXD and pertuzumab. TDXD plus pertuzumab may represent a new first line standard of care for patients with HER2 positive metastatic breast cancer. Next slide, please. Next is DESTINY-Gastric04. This is our phase III trial of second line in HER2 treatment and HER2 expressing metastatic gastric cancer. Briefly, an estimated 5%-17% of gastric cancers are HER2 positive, and the second line standard of care is ramucirumab and paclitaxel.
DESTINY-Gastric04 was conducted to evaluate T-DXd against ramucirumab and paclitaxel in patients with HER2-positive metastatic gastric cancer. Next slide, please. Here is the design. Patients had received prior therapy here. They were randomized into T-DXd at 6.4 milligrams per kilogram q3 weeks, remembering that this is a higher dose than that in breast cancer, against ramucirumab and paclitaxel. This time, primary endpoint and overall survival, secondary endpoints included PFS. Next, please. In the hierarchical testing, OS was tested first as the primary endpoint, followed by PFS, and then confirmed response rate. The analysis we present today, and that was presented at ASCO, is an interim overall survival analysis. At the time of the data cutoff, there were 266 overall survival events, and at the overall survival interim, superiority was achieved. Next slide, please. Here are the demographic and baseline characteristics.
We can see that this study was conducted in Asia, Western Europe, and mainland China. This study was not conducted in the U.S. in that HER2 was approved in the second line setting in the U.S., but does not have full approval in some other regions or approval in the second line. In addition to this, we can see the majority of patients were IHC 3 plus, and the prior treatments included immunotherapy only in about 16% of the population. Next slide, please. Importantly, we show subsequent anti-cancer therapy because this is a primary overall survival endpoint, which can be confounded by subsequent anti-cancer therapy. You can see in the ramucirumab plus paclitaxel arm, about 25% of patients received prior HER2 ADC, including about 20% with ENHERTU itself.
This is very meaningful because we can see that ENHERTU is able to deliver an overall survival clinically meaningful benefit despite such post-progression therapy within HER2. Next slide, please. Here is the primary endpoint. We can see a hazard ratio of 0.7, statistically significant and clinically meaningful, with a 3.3-month difference in median overall survival. Certainly confirming ENHERTU as a second line standard of care in HER2 expressing metastatic gastric cancer. Next, please. Here are the overall survival subgroups, and we can see largely consistent results across the subgroups, including IHC status, geography, and other aspects. Next slide, please. Here is the secondary endpoint, PFS by investigator. Hazard ratio 0.74, 6.7 months versus 5.6 months, also clinically significant, supportive of the overall survival primary endpoint. Next, please. Here is overall response rate and durability. We can see 44% for T-DXd, 29% for ramucirumab and paclitaxel.
You can see on the right the median duration of response separated, higher in T-DXd versus ramucirumab and paclitaxel. Next slide, please. Here is the safety summary. Overall, T-DXd had a manageable safety profile. If we would compare the various AE categories on this slide, we can see a similar incidence of drug-related grade 3 or higher AEs, serious AEs, treatment discontinuations, as well as the fact that there were a similar rate of related deaths in the two arms, a bit higher on the T-DXd arm. Next slide, please. Here are the drug-related adverse events that are most common. We can see that the most common events in the T-DXd arm included fatigue and AEs related to GI or hematologic nature. This is consistent with what we already know about T-DXd. Next slide, please. Adverse events of special interest included, of course, interstitial lung disease.
Very importantly, and considering the higher dose of 6.4 mg per kilogram of T-DXd, there was one grade three event and no grade four or five events with a total rate of ILD of 13.9%. And there were 2.5% of patients who had a left ventricular dysfunction event similar to the control. Next slide, please. In conclusion, this data confirms T-DXd as a second line standard of care in HER2 expressing metastatic gastric cancer. This is the first randomized trial with a HER2 directed agent showing a survival benefit. And we can see the 0.7 hazard ratio for OS, supportive secondary endpoints, manageable safety profile with no new signals identified. So confirmation of this global standard of care. Next slide, please. Now we're going to shift to Datopotamab deruxtecan or Dato-DXd, our TROP2 ADC. Next, please.
This is a slide which provides an overview of our combination studies in non-small cell lung cancer. As you can see, there are trials with various immuno-oncology agents, including pembrolizumab, where we study TROPION-Lung02, the data presented at ASCO, which I will share with you momentarily. The data that I will show supports TROPION-Lung07 and Lung08 frontline trials that are ongoing. We also combine Dato-DXd with durvalumab, the PD-L1 inhibitor from AstraZeneca. I will share some results from NeoCOAST-2 in the neoadjuvant setting shortly. We also know that AVANZAR, the frontline trial that I said would read out in the second half of the calendar year 2025, also contains durvalumab, as well as the Lung04 trial. There are also combinations with rilvegostomig, AstraZeneca's TIGIT PD-1 bispecific, in two ongoing phase III trials, Lung12 and Lung10.
Lastly, to round out the combination strategy, we include combinations of Dato-DXd with osimertinib, the EGFR TKI, and we have ongoing pivotal trials Lung 14 and 15. Next slide, please. I am now going to walk through data from Lung 02, and we have quite a bit of data to present. Next, please. Lung 02 is a phase I-B study of Dato-DXd with pembrolizumab, with or without platinum chemotherapy in non-small cell lung cancer. Patients enrolled had both second and frontline non-small cell lung cancer, and the focus of today's presentation is the frontline population. You will see in the study that more than one dose of Dato-DXd was used, 4 mg per kg or 6 mg. You will also see that patients may have received carboplatin or cisplatin. The primary endpoint was safety and tolerability, but we also, of course, look at efficacy. Next, please.
Here are the baseline characteristics for the frontline patient populations that received the doublet of Dato-DXd plus pembrolizumab, 42 patients, and the triplet of Dato-DXd and pembrolizumab plus platinum. I want to point out some imbalances here that I think are important to understand in terms of reviewing the efficacy results. First of all, there is a higher proportion of patients with brain mets in the triplet regimen. In addition to that, you will see a higher proportion of patients in the triplet arm received the lower dose of 4 mg per kilogram, as opposed to the higher dose being more common in the doublet regimen. The majority of patients had PD-L1 less than 50%, and the majority of patients had non-squamous histology. Next slide, please. Here is the safety summary. Overall, with the doublet and triplet regimen, there was a manageable safety profile with Dato-DXd in combination.
You will see that about a third of the patients in the doublet and the triplet arms did discontinue Dato-DXd. There were grade greater than or equal to 3 related adverse events in about 10% versus 17% in the doublet and triplet respectively. When we consider ILD, I think it's important to note there were no grade 4 or 5 ILD events with this Dato-pembro combination, either doublet or triplet. There were two patients with grade 3 ILD with the doublet, one patient with grade 3 ILD with the triplet, and again, nothing higher than the grade 3 and mostly lower grade. Next slide, please. Here are the treatment adverse events that were most commonly occurring in the frontline patients. In green is the triplet. In blue is the doublet.
We can see that the most frequent adverse events of any grade across the regimens were nausea, stomatitis, anemia, decreased appetite, and pneumonitis. Hematologic toxicities, not surprising, with platinum being contained in the triplet, were more common in the triplet cohort. Next slide, please. Here is the efficacy in the doublet cohort, reminding you of a 42-patient subset, response rate nearly 55%, median duration of response 20.1 months. You can see the PFS curve on the right with the doublet at 11.2 months. Next slide, please. Here is the efficacy of the 54 patients treated in the frontline treated with the triplet. You can see a 55.6% response rate, median duration of response 13.7 months. You can see the progression-free survival curve on the right, with the triplet showing 6.8 months median PFS. The next slide. Thank you. Now we turn to response by PD-L1 status.
Importantly, if you look at the doublet or the triplet, there were responses in both PD-L1 less than 50% population and the PD-L1 greater than 50% population. However, and as mentioned when I discussed the baseline characteristics, only a minority of patients that received either the doublet or the triplet had a PD-L1 50% and higher expression, therefore a limited population at that high expressing side. Next slide, please. Now we'll turn to the TROP2 QCS assay because in TROPION-Lung02, we did conduct a retrospective exploratory analysis of the TROP2 QCS assay looking at normalized membrane ratio in the TROPION-Lung02 population. We had previously stated that immunohistochemistry TROP2 expression using the conventional assay did not enrich for response.
We had also showed in a retrospective analysis of the phase III second line non-small cell lung cancer trial, TROPION-Lung01, we saw the potential for the QCS TROP2 assay to predict response to Dato-DXd. As a reminder, and as previously presented at the World Lung Cancer Congress, we can see the steps in the QCS assay. First, there is the immunohistochemistry scanning with the conventional assay. Then there is the whole slide imaging. Then there is the quantitative continuous scoring assisted with artificial intelligence to identify analytes within the cell and within the slide that may help predict response to Dato-DXd. After two training sets, including the phase I TROPION-Lung01 and then the TROPION-Lung01, it was determined that the biomarker cut point is represented on the right.
Positivity being that greater than or equal to 75% of tumor cells with TROP2 normalized membrane ratio less than 0.56 and negativity being less than 75% of tumor cells at that level of TROP2 NMR. That ratio basically is a ratio of the overall membrane expression divided by the membrane expression and the cytoplasmic expression. The importance of this assay is that the patients who are biomarker positive tend to have a higher component of cytoplasmic expression. Next slide, please. This is a reminder of what was presented at the World Lung, the biomarker evaluable population of TROPION-Lung01, 352 patients that were retrospectively and in exploratory fashion analyzed for TROP2 QCS NMR. Indeed, TROP2 QCS NMR did predict outcomes in this exploratory analysis in TROPION-Lung01.
Importantly, the same cutoff for the TROP2 QCS assay that was used in TROPION-Lung01 is used in this retrospective analysis using the assay in TROPION-Lung02 that we're now going to look at together. Next slide, please. Here is PFS by TROP2 QCS NMR in the frontline biomarker evaluable population. This would include patients that receive both the doublet and the triplet. As you can see, TROP2 NMR positive patients had a higher median PFS compared to TROP2 NMR negative patients in this exploratory analysis with a hazard ratio for PFS of 0.62. You can also see a higher response rate in the biomarker positive population in this exploratory analysis. If we go to the next slide, please. Now we're going to break out the doublet and the triplet looking at this TROP2 QCS NMR assay.
In the frontline population receiving the doublet, we see a similar trend toward higher PFS in the NMR positive population versus the QCS negative NMR population, 21 months versus 11 months. I would note that the sample sizes are small. However, the trend is there with a PFS hazard ratio of 0.5 and a response rate higher in the NMR positive compared to the NMR negative population. Next slide, please. Here is the biomarker evaluable population with the triplet. Again, limited sample sizes, but the trend remains 8.2 months versus 5.5 months and a PFS hazard ratio of 0.67 with a higher response rate of nearly 62% versus 52% in the NMR negative population. Next slide, please. In conclusion, TROPION-Lung02 phase I-B study of Dato-DXd plus pembrolizumab with or without chemotherapy demonstrates durable anti-tumor activity in frontline population, including patients with higher and lower PD-L1 expression.
Tolerability of these regimens, doublet and triplet, was as expected based on known profiles. In the retrospective exploratory analysis of QCS NMR TROP2, we see a trend toward prolonged PFS and also higher response rate, as I showed, in patients that have NMR positivity. These data support two ongoing pivotal trials, TROPION-Lung07 and TROPION-Lung08. Next slide, please. As mentioned, I'm going to show data from NeoCOAST-2. Now we turn to the use of Dato-DXd, our TROP2 ADC, in combination with durvalumab and chemotherapy, this time in resectable neoadjuvant setting in non-small cell lung cancer. Next slide, please. Here is the schema, which is somewhat complex.
As you can see in this phase I/II study in patients that have stage 2A or stage 3B resectable non-small cell lung cancer prior to surgery, are randomized to a number of regimens with neoadjuvant courses of four cycles every three weeks. I will not really speak to the data of arm 1, which has oleclumab, which is a CD73 inhibitor from AstraZeneca. I will also not speak to the arm 2 data using monalizumab, which is an antibody against NKG2A from AstraZeneca. Our focus here today will be on arm 4, which is Dato-DXd plus durvalumab, the PD-L1 inhibitor plus platinum chemotherapy. You can see the patients then undergo surgery, and then they continue on durvalumab monotherapy after surgery for up to a year. Next slide, please. Here are the baseline characteristics. As mentioned, this is an early non-small cell lung cancer population.
You can see the majority of patients were at stage three as opposed to stage two, but we did enroll patients both stage two and stage three. The majority of patients also had adenocarcinoma, and the majority of patients also had a TPS score, PD-L1 expression score greater than or equal to 1%. Again, I'm going to focus on the safety and efficacy only from arm 4. Next slide, please. In this study, and I'm sorry it's a bit small, we're going to start on efficacy in the upper left. In the blue bar curve, we can see arm 4 with a pathologic complete response rate, a key efficacy endpoint, of about 35%, and a major pathologic complete response rate of about 63%. As a reminder, pathologic complete response rate is no evidence of tumor in the resected tumor or surrounding lymph nodes during surgery.
Major pathologic response rate is less than 10% of viable tumor at the time of surgery in the nodes and in the resected tumor. This is a common endpoint in the neoadjuvant setting. Those two endpoints are common in the neoadjuvant setting. I'll now turn to the bottom left where we look at pathologic complete response and major pathologic response in arm 4 in the blue. You can see that in arm 4, the MPR and the pCR rates were quite similar in those two PD-L1 expression subgroups, TPS greater than 1% or less than 1%. In the bottom right in the blue, there were higher pathologic complete response rates and major pathologic response rates seen in patients with squamous carcinoma as opposed to adeno. I do recall, though, squamous population was smaller than the adeno population. As you can see, there are overlapping intervals in the bars.
Next slide, please. Here are the safety results. You can see the safety results are broken down by phase. In the neoadjuvant phase, you can see the grade greater than or equal to 3 adverse event rate was about 24%, with about 19% of patients having a related event. Discontinuations were about 11%. Overall, a manageable safety profile prior to surgery. This is very important because when we give a neoadjuvant regimen, we look to see if we disrupt the ability to then undergo surgery. Indeed, in this trial, we saw the safety did support neoadjuvant treatment and was consistent with what we would expect in historical control or ability to go then on to surgery. The most common adverse events on the right are stated there, including hematologic events, some constitutional events, as well as events that we know with Dato-DXd.
Overall, consistent with what we might expect. As you can see in the dark blue, a minority of events of grade three or higher. Next slide, please. In conclusion, focusing on arm 4, there were numerically higher pathologic complete response rates with Dato-DXd plus platinum plus durvalumab compared to standard of care and promising major PASCR rate as well. To give a sense of standard of care, it is approximately 25% or low 20% in this setting. Although we do not have a cross-trial comparison, this is the standard of care referred to here. As I mentioned, the PASCR rate was 35% in this study. Data from arm 4 showed a manageable safety profile. Encouraging pathologic response rates were observed regardless of PD-L1 status.
The pCR and the MPR rates in arm 4 are the first reported for any antibody-drug conjugate in the neoadjuvant setting and confirm an encouraging efficacy signal and a manageable safety profile for Dato-DXd combination in this very early stage of non-small cell lung cancer. I did not share with you the ctDNA clearance data that was presented here. This is a biomarker that was assessed and there was evidence in reduction in presurgical ctDNA associated with pathologic response. That is an exploratory analysis, and I did not have time to present today. Overall, we find this data encouraging for Dato-DXd-based regimen in neoadjuvant early lung cancer. Next slide, please. Now I'm going to turn to HER3-DXd. Next, please. You may recall we issued a press release recently notifying that we would withdraw the HLA-1 submission.
HLA-1 was a phase II study for accelerated approval of HER3-DXd in late line EGFR mutation positive non-small cell lung cancer. That withdrawal announcement was based on the fact that in this phase III study that we talk about now, TROPION-Lung02, we did not achieve statistically significant overall survival. Now we're going to look at that data together. Next slide, please. Here is the study design. As a reminder, this is advanced EGFR mutation positive non-small cell lung cancer. Patients had to receive one to two prior lines of an approved EGFR TKI. Most of the patients were required to have received osimertinib, which is a clear standard of care third generation EGFR TKI. Patients were randomized to monotherapy HER3-DXd or platinum-based chemotherapy. The primary endpoint was PFS. We will show you the primary analysis of PFS.
We will also show you a more recent overall survival third interim analysis. Overall survival is a secondary endpoint. Next slide, please. As a reminder, the primary endpoint being PFS, you can see the primary PFS analysis results, which we will show now, had a cutoff of May 31, 2024. The key secondary endpoint overall survival analysis had a more recent cutoff of February 28, 2025. We are going to show you different cutoffs in time for the primary PFS analysis and the overall survival analysis. Next slide, please. Here are the baseline population characteristics. Overall, without going into much detail, they were fairly balanced between the two arms and were pretty typical of what we see in an EGFR mutation positive non-small cell lung cancer population. Consistent with the eligibility criteria, you can see that about 90% of people did receive prior osimertinib.
In addition to that, about a third of patients had brain metastasis. If we move to the next slide, please. Here is the primary PFS results. As you can see, the hazard ratio for PFS is 0.77, with a p-value of 0.11, not achieving statistical significance. However, the median PFS in the HER3-DXd arm was 5.8 versus 5.4 in the platinum control arm, demonstrating only a very modest improvement in median PFS. There is some separation in the curves. Next slide, please. Here is the efficacy according to subgroup. Overall, fairly consistent across key baseline factors with a higher hazard ratio in the L858R activating mutation for EGFR activating mutation compared to the exon 19 deletion. However, you can see that these do overlap in terms of the hazard ratio ranges. Next slide, please. Here is the response rate data.
We can see there is a higher response rate of HER3-DXd at 35% compared to 25% for platinum chemotherapy. There is also, if we see down below, a pretty consistent and comparable median duration of response for HER3-DXd at 5.7 months compared to platinum chemotherapy at 5.4 months. Next slide, please. Here is the safety summary. Overall, HER3-DXd had a manageable safety profile. However, in some of the AE categories, including the treatment-related SAEs, including dose reductions and also interruptions, we do see a somewhat higher rate of adverse event compared to platinum chemotherapy. Next slide, please. Here are the most common adverse events with HER3-DXd and platinum-based chemotherapy, with HER3-DXd in blue and platinum-based chemotherapy in orange. There were some higher rates of hematologic toxicity in the HER3-DXd arm compared to the platinum-based chemotherapy arm. Notably, these generally occurred early and were temporary, in most cases not associated with clinical sequelae.
For those that did have a grade three or higher platelet count reduction, each arm had one bleeding event. Overall, the safety profile here and the types of adverse events are consistent with what we know about HER3-DXd. Next slide, please. Here are the adverse events of special interest, particularly ILD. You can see that the overall rate of adjudicated ILD was low. The treatment-related rate was 4.8%. There were two fatal adjudicated related ILD events at 0.7% and one grade three event. Next slide, please. Here is the overall survival data. Just as a reminder, the data cut for this was later than the PFS. We included this data in here as an important secondary endpoint result. It was also referenced in our press release. Here you can see there is no improvement in overall survival with HER3-DXd versus platinum chemotherapy with a hazard ratio of 0.98.
Next slide, please. In conclusion, at primary analysis, HLO2 demonstrated a statistically significant improvement in PFS, but only a modest improvement in the median. The recently available third interim analysis data of overall survival did not show an improvement compared to the control arm. What I did not show you for the sake of time was that HER3-DXd did have activity in intracranial disease. You can see the intracranial ORR was fairly high in the HER3-DXd arm compared to PBC, the platinum-based combination. The safety profile was generally manageable and consistent with previous reports. There were two fatal ILD events, with the overall rate of ILD being relatively low. Importantly, we will look at biomarkers in this data set, including conventional immunohistochemistry, as this is a phase III data set that is rich to further understand which patients might benefit most. Next slide, please.
I'm almost done with the presentation, and I'm going to show first the map of our HER3-DXd studies because what I'd like to do is convey the message that even though we consider the HER3 and the Lung02 trial to be a setback, very importantly, we think this asset has significant potential. Indeed, setbacks are inherent to the clinical research process, and we will persevere. We continue on in non-EGFR non-small cell lung cancer with studies in combination with pembrolizumab. We also continue on in breast cancer, both in the HER2 positive and the triple negative breast cancer settings. Some of you may recall at the last ESMO that we presented a French study, ICARUS-Breast01, with about 100 patients in hormone receptor positive second line breast cancer with more than a 50% response rate.
Indeed, we will explore additional tumors, including about approximately 15 tumors in total. More to come with this asset. Next slide, please. I'm going to round out the discussion with mentioning a phase I asset program that was presented here at ASCO. Next slide, please. We have a phase I study of DS-2243 in HLA-A2 ESO-1-directed bispecific T-cell engager. Next slide. Here you can see a cartoon showing this bispecific T-cell engager, the first at Daiichi Sankyo, which again targets HLA-A2, HLA-A02, and ESO-1, and also engages T cells. With this molecule, we believe there's broad applicability as there's high to moderate expression of NY-ESO, the target in synovial sarcoma, liposarcoma, non-small cell lung cancer, and urothelial carcinoma. Indeed, in preclinical models, we see evidence of anti-tumor activity that is robust.
At this ASCO, we presented the phase I study design, which I won't show you here for the sake of time, but I thought I would mention this molecule to express that we are indeed looking at the diversity of the pipeline and increasing its breadth and depth. Next slide, please. I would like to conclude by thanking you for your attention. I would like to conclude by reiterating the data that was presented at ASCO is consistent with our mission of transforming standard of care. We will continue in that mission. Indeed, we look forward to presenting further data in the future that demonstrates the progress of our pipeline to address unmet needs for patients. Now I believe we'll move to a Q&A. Thank you kindly for your attention.
Thank you, Mark. We will now take questions. You could ask your questions in English or in Japanese. If you have multiple questions, please ask one at a time for optimal translation. Please limit the number of your questions to two at most. Now, please raise hand if you have questions. The first question is from Yamaguchi-san from Citigroup. Please go ahead.
Hello. Can you hear me?
Yes.
Okay. Great. The first question regarding TLO2. You cover the topic and a little bit of sub-questions, but a covered benefit after QCS is much higher, much longer than the total breadth benefit, even though it's not really comparable. There is also the imbalance of the patients. Can you give me why there is a case? Also, OS benefit looks the same. You did not touch about the OS benefit, but do you think OS benefit is also a positive sign? Thank you. That is the first question regarding TLO2.
Okay. Sure. So indeed, we did see relatively comparable overall response rates between the doublet and the triplet. We also did see a longer median PFS in the doublet compared to the triplet. We did not show overall survival particularly. We showed the response rate and the median PFS. I believe it's important to recognize that this is a phase I-B study. The interpretation of the efficacy results is directional, such that with the sample sizes we have, what we're looking for is a directional signal. Indeed, with both the doublet and the triplet, we see evidence of durability and also regardless of PD-L1 expression.
One reason, actually two reasons, and I think I might have alluded to those when I presented the baseline characteristics, that we may see a longer PFS, median PFS in the triplet arm, relates to the fact that the lower dose of 4 mg per kilogram was more prevalent in the triplet arm. The brain mets as well. Overall, I would say both arms are directionally encouraging towards supporting our pivotal trials, considering small sample sizes and considering some baseline imbalance in factors.
Thank you. The second question is quickly on the DDO9. The data was pretty impressive. The conference was really excited. One thing to note is that the comment data who commented on that trial results was a little bit cautious about how to sequencing the treatment given the current THR and other CD46 are also available, which is more of the maintenance wise. Can you give me a comment? How do you think about those comments from the comment data? Thank you.
Yes, thank you. The commentator addressed a couple of issues. One was sequencing, and I think one was what should be the optimal course of T-DXd. Indeed, we study T-DXd plus pertuzumab until progression. We do not have data that demonstrates what is the best sequence of use of T-DXd, whether one would start with T-DXd or start with something else and then T-DXd. We also did not study a limited number of cycles of T-DXd. Therefore, it is difficult to comment what data may inform an alternative approach. That being said, I do think the questions that were raised are important. These are questions that we will consider going forward in our development program.
In addition to this, I would note when we think about the sequencing, because I think you specifically asked about that, if you would look across the conduct of all the metastatic HER2 breast cancer trials that have been conducted for HER2, what we see is a trend of increased median PFS when we go from later line to earlier line. It does appear that the most robust and strongest clinical benefit is when T-DXd is used upfront. That is some consideration for how we might think about the sequence. Again, without having specific data to inform sequence, it is challenging for me to comment on the merits of an alternative strategy. I would say that I would need to reflect back on the robust results and benefit risk profile we did provide using T-DXd and pertuzumab treatment to progression.
Okay. Thank you. Thank you very much.
The next question is from Hashiguchi-s an from Daiwa Securities. Hashiguchi-s an, please go ahead.
[Foreign language]
Thank you very much.
Hashiguchi-san, can you please repeat your question?
[Foreign language]
Thank you very much. The first question is about TL02 demonstrated QCS data, which is very encouraging. How I'm going to link this to the success of the AVANZAR study, I'd like to have your comments. AVANZAR study, primary endpoint in its assessment, QCS biomarker is going to be used or not? You have not made any comment on this. This is going to be useful for the studies beyond AVANZAR or also in AVANZAR. By leveraging the data, you can increase the probability of success. I appreciate your comment.
Yes, thank you. There are several comments. The first is to look at the retrospective exploratory analysis that we did in TROPION-Lung01 with the QCS assay. Indeed, it suggests a trend toward improvement in PFS and response rate with both doublet and triplet with the NMR or biomarker positive population. We believe that those results are supportive of the conduct of AVANZAR study, AVANZAR being a frontline trial in non-small cell lung cancer regardless of PD-L1 expression with a triplet regimen. It is not the same as the one in TROPION-Lung02, but similar because Dato-DXd plus durvalumab, PD-L1 inhibitor instead of PD-1 inhibitor, and platinum. Indeed, the pre-planned analysis in AVANZAR includes the QCS biomarker positivity.
To answer your question, the primary endpoint structure of AVANZAR allows us to be able to look at efficacy in both the biomarker positive population for TROP2 and separately the intent-to-treat population that is not biomarker selected. The analysis in AVANZAR is pre-specified, though it will be a retrospective analysis. Overall, we do believe that the data from TROPION-Lung02, because it does contain a triplet of Dato-DXd, immunotherapy, and platinum, as does AVANZAR, we believe that data is supportive of AVANZAR. We believe that the QCS assay use has the potential to increase the probability of success for AVANZAR. Of course, we will need to await the data in the second half of the calendar year of 2025. I believe you also asked about the use of the TROP2 QCS assay in other indications, if I'm correct.
We are very interested in patient selection in general, and certainly are interested in learning from the use of this TROP2 QCS assay, not only in lung cancer, but also potentially in other indications. This is an exploration underway, but we have no particular data to describe in other tumors at this time. Thank you.
[Foreign language]
The second question is about HER3-DXd. This time, based on the study results, based on the patient population and their backgrounds, the efficacy could be different according to what was suggested. Predictive biomarker importance is now clearer, according to my interpretation. EGFR mutated NSCLC is not the only disease, but also HER3-DXd, other indications under development, predictive biomarker would be needed as well. Do you think you can say the same thing for other diseases, predictive biomarker candidates, anything which can be a promising candidate already in sight?
Thank you. Indeed, HLO2 for the intent-to-treat population did not produce the overall survival benefit that we had hoped. We are looking carefully at the HLO2 data set, which is a large randomized global data set, and exploring potential biomarkers. One of the biomarkers of interest to us is immunohistochemistry, standard immunohistochemistry for HER3. Indeed, we are looking at that and other potential biomarkers, including immunohistochemistry, not only in EGFR mutation positive non-small cell lung cancer, where HLO2 was performed, but in the other tumors that I mentioned. I did not call out specific tumors, but I said we are testing, I think, up to 15 tumors. We have announced at previous times that those tumors span GU, gastrointestinal, and other tumors.
In the context of those early phase signal finding disease-directed cohorts, we will indeed look for a biomarker patient selection using immunohistochemistry, but also other approaches.
[Foreign language]
Thank you very much.
Next question is from Tony Ren-san from Macquarie. Please go ahead.
Hello. Can you guys hear me?
Yes.
Okay. Perfect. Yeah. Thank you for the opportunity. Going back to TROPION-Lung02, I want to ask about the toxicity if you go to slide number 46 and number 47. Let's go to slide 47 here. Forty-six here, sorry. On slide 46, you really see very high rates of interstitial lung disease and pneumonitis, right? Twenty-six%-ish, roughly. This is much, much higher compared to TROPION-Lung01. I just want to see why we have almost a tripling of the ILD rates. Do you think that is attributable to the PD-1 involved here? On slide 47, I may have missed it, but I did not see any ocular toxicity, which is often talked about. That is my first question.
Yeah. To the ILD rate, you're right. On slide 46, there is the higher incidence rate of ILD than we've seen in TROPION-Lung01. I believe that this is possibly related to the fact that Dato-DXd itself, when given, can be associated with ILD, and pembrolizumab and other immunotherapy agents in that class also have the potential to cause pneumonitis. However, I would point out, which I think is quite important, and during my presentation, that despite the rate that we see, we can see that the rate of grade 3 and above ILD was actually relatively low, and there were no grade 4 or grade 5 events. I think this is important. The majority of events, though, if at higher rate than seen in TROPION-Lung01, were low grade.
The other thing I would say is that this phase I-B study is an important way for us to learn about the management of ILD and other adverse events of special interest in an early phase setting and apply those learnings to later phase setting. I think that's also something to consider. You asked about ocular events as well. The ocular events are actually represented there on that same slide 46. They're not on a, I guess they don't make it onto slide 47, but they do make it, they're highlighted on slide 46. Here again, we see a low rate of grade 3 ocular surface events, which is important.
Okay. Got it. Yeah. Now I see them. Now I see them on slide 46. Yeah. Talking about managing ILD, yeah, I want to go back to what Yamaguchi-san said earlier today about the presentation of DESTINY-Breast09. I would say that the discussant, Dr. Claudine Isaacs from Georgetown University, was not particularly keen on this regimen, let's just put it that way. She mentioned the escalation of treatment. I want to ask you about the 12.1% ILD rates there with two fatalities involved, right? I noticed the trial started in 2021. I believe you guys were already rolling out the ILD management mitigation strategy at the time already, right? How did we still end up with such high rates of ILD, including grade 5s?
Yeah. I would say, first of all, we take ILD very seriously, and we've been learning on a consistent basis. I think it's important that in the clinical study conduct over time, we have certainly had a heightened surveillance. We have addressed eligibility criteria for people who could be at risk. We have developed dose management criteria, including stopping and withholding at ILD and also administering steroids. In the post-marketing setting, we have also had education and labeling to address ILD. What we see in ILD rates, the 12.2%, is really consistent with what we've seen for ILD with ENHERTU in metastatic breast cancer. This is actually not a new finding or not a disproportionate finding with respect to overall incidence or severity in this population. When we think of the benefit-risk profile, I mean, we have to consider both benefit and risk.
This is a regimen which had more than three years of median PFS. This is a regimen which had more than one year of median PFS improvement compared to the standard of care. This is a regimen where if you responded and 15% of patients did have a complete response with no evidence of radiographic disease, you responded for a median of 39 months. This is evidence of very significant clinical benefit. On the risk side, yes, there's ILD, which is a known adverse event of interest in HER2. The level of risk that we see is not different or disproportionate compared to what we've already known about the drug. Overall, we believe there's a favorable benefit-risk profile. When I turned to the discussant, I don't want to speak for the discussant, but my impression was that the discussant was not unsupportive of the regimen.
I think what the discussant highlighted were some questions that she had about the regimen as we would kind of move forward and think about its best use. In the way that we studied in HER2 plus pertuzumab in frontline HER2 metastatic positive breast cancer, the data suggests a very favorable benefit-risk profile. In accordance with the presentation by the initial presenter, this could potentially represent a new standard of care in this setting based on robust efficacy. The discussant did raise some interesting questions.
Okay. Yeah. I certainly agree that I think Dr. Tolaney from Dana-Farber was far more supportive of the regimen. Yeah. Okay. That's it for me. Yeah. Thank you.
The next question is from Sogi-s an from Sanford Bernstein. Sogi-s an, please go ahead.
Thank you very much. I have also a question regarding TROPION-Lung02. It was really striking to see the triplet data for TROP2-positive PFS. Also, you disclosed the duration of response within non-squamous for triplet. The data that was included in the abstracts, so duration of response in non-squamous, it was 18 months. I assumed that that was probably mainly TROP2-positive patients, non-squamous. Yesterday's presentation that ASCO, you have shared with us, the triplet TROP2-positive PFS, 8.2 months, but not the duration of response. Given that the overall response rate was above 60%, the majority of patients are responding.
My question is, do you think that the difference between this 8.2 months and 18 months, obviously, we are not really comparing the apple to apple here, but does it suggest that the TROP2 positive in basically the drug's efficacy between non-squamous TROP2 positive and squamous positive difference, that is leading this difference? What else do you have as a hypothesis?
I wanted to make sure, thank you for the question. I want to make sure I understood the question. You're saying a higher durability with the QCS assay in the retrospective analysis compared to without the assay? I want to make sure I understand the comparative groups that you were referring to when you talked about greater durability versus lesser durability. Can you repeat what those two comparisons are? I want to make sure I understand. Thank you.
Yeah. Sure, sure. Yes. I'm comparing the data you included in the abstract, which is without QCS, but non-squamous duration of response, and the data included in the presentation.
I do not have the abstract in front of me. When we show in this presentation, we show duration of response according to doublet and triplet regimen, so without QCS. In the doublet, we have duration of response of 20.1 months. In the triplet, so this is non-QCS, we have duration of response of nearly 14 months. In both cases, with the doublet and the triplet, in the non-squamous population, we see significant evidence of durability in the response. What we present here are the doublet and triplet broken out and not together. I apologize, I do not have the abstract in front of me to see the 18 months that you are referring to.
What I would say is that in the dataset for TROPION-Lung02, whether it be the doublet or the triplet, we do see evidence of durable response in patients who have both low PD-L1 expression and high PD-L1 expression. I think that's important to support the TL07 and TL08 populations. We did not show duration of response in the QCS analyses, but we did show PFS and response rate, and we did show PFS curves. I think even though we don't have duration of response in the median expressed in the same way, according to the PFS that we see, we believe there's certainly durable response according to QCS selection as well.
Because you seem to want to refer to something in the abstract I don't have in front of me, then maybe we can follow up about that unless you want to continue with another question. Thank you.
Yes. It'd be great. We can have something to follow up on this. The second question is that this TL02 is really striking in terms of really strengths of the doublet data. Are you now with AstraZeneca considering running a doublet in the first-line non-small lung cancer setting?
Yeah. We already are. Thank you for that question. Indeed, the data suggests that the doublet is active and the triplet is active. Again, there were some baseline sort of imbalance in brain mets, and there was a higher use of the lower dose in the triplet. Because of that, it's a bit hard to interpret. The study itself wasn't built to compare the two arms, actually. However, we do certainly conclude from the data directionally in TL02 that the doublet as well as the triplet can be viable regimens and supportive of our ongoing pivotal trial program. To your question, we actually do have doublet deployed in the phase III. In fact, in TL07, TROPION-Lung07, this is Dato-DXd plus pembrolizumab with or without chemotherapy.
We have a doublet arm and a triplet arm in a patient population in TL07 that is less than or equal to 51% PD-L1. We will be able to see the performance of the doublet and the triplet. We also just, yes?
Sorry. Are. Yes, you're right. Then those trials, TL07 or TL08 for that matter, you guys are not including QCS. That's a huge difference.
Yeah, I see. Right. We have not announced any plan to include QCS in those studies. I think what you can see, though, in the data that I presented in TL02, is that even without QCS, you see that doublet regimen with evidence of durable responses in patients of different PD-L1 groups. You are correct. TL07 does not, and let's say TL08 for that matter, because that is also a doublet regimen, right? It is Dato-DXd and pembrolizumab in the PD-L1 50% and higher population. What I would say is that the data in TROPION-Lung02 does support TL07 and TL08, including the consideration that QCS is not currently included there. Yes.
Sorry. One more quick question. The reason that triplet patients did receive less, the less patients received the 6 mg versus 4 mg, is it because of the patient, the condition, such as brain mets?
Yeah. If you look at the design, there was a and that's on page 44. Patients in different cohorts received different doses. The idea in this phase I-B study is to explore across different cohorts different combinations. For instance, you can look at there was an attempt in cohort 3 and cohort 4, for instance, to look with carboplatin as the platinum agent with the different doses of Dato-DXd . You can look at cohorts 5 and 6, where the choice of platinum was cisplatin to look at either 4 mg or 5 mg per kilogram, 4 mg or 6 mg per kilogram. It is just inherent to the study design that there was an attempt to look at the dose and the potential influence on outcomes according to dose, and then to look at different platinum regimens.
It is not, let's say, intentional to give a higher dose or a lower dose based on, let's say, a baseline patient characteristic, more about the intent to systematically address the different dose with the different platinum agents in combination with dato and pembro.
Thank you very much.
Sure. Thank you.
The next question is from Matsubara-san from Nomura Securities. Matsubara-san, please go ahead.
Hi. I'm Matsubara from Nomura. Can you hear me?
Yes.
Okay. Thank you. Regarding page 18, the combination therapy of the HER2 and the Perjeta was effective, but why does the PFS reverse compared to the THAP regimen with long-term administrations?
Sorry. Can you repeat that question? I think I missed it. On page 18, you had a question again. Can you repeat it?
Is 18, the PFS, so the long-term administrations, so PFS reversed compared to THAP regimen. I mean the HER2 and the Perjeta.
Sorry, I'm not following the question. Again, my deepest apology. Could you repeat one more time, please?
Okay. Sorry. On page 18, sorry, I received the same figures. Okay. Wait.
Take your time. Take your time.
Yeah, yeah, yeah. Please figure, sorry, page 18. PFS is, I understand the ENHERTU and the Perjeta is effective compared to the THAP. In the long term, I mean, I can see the figures. Okay. Long term, like the 442 or 45, the lesser reversed.
I see the question. Yeah. So you're asking, I believe, that you were saying that it's a clear separation in the curves, the KM curves, and the medians, but the curves are kind of coming together. I think that's what you're suggesting. Yes. So indeed, and I think the discussion highlighted that and made an interesting comment. Actually, maybe it was the original presenter. What happens is this is actually an interim analysis of PFS. We don't even have this regimen actually achieves statistical significance with clinically meaningful results. That is T-DXd plus pertuzumab at the interim. What you see at the interim is you can see some of those hash marks, those vertical marks in the curves. This is evidence of patients being censored. We don't yet have the final number of events for the final analysis.
I mentioned the monotherapy arm is actually continuing to final analysis. In this case, this is interim analysis. What you see in those curves coming together is just that there is some censoring happening, which means that is reflecting immaturity. In fact, what is possible with greater follow-up and the addition of more progression-free survival events as patients experience progression-free survival events, we may see a change in those curves, and we may see a change even in the medians, even in the medians. This is just a reflection of the censoring and the evidence of not full maturity of the progression-free survival endpoint at the time of an interim analysis. That explains very likely the curves come together. This does not suggest that the PFS curves are, let's say, evidence of lack of benefit or something like that.
It's just that at the later time point, we reflect immaturity and censoring of PFS in the context of the interim analysis.
Okay. Understood. The second is breast cancers. I mean the combination therapy of TROPION and the KEYTRUDA showed efficacy to compare to the standard cares. How do you think about the result? I mean the potential of the DXd's.
Yes. I think you're referring to the ASCENT trial. So phase III study in frontline triple-negative breast cancer presented at this ASCO of sacituzumab or Trodelvy and pembro in PD-L1 expressing patient population. Yeah, there was, I think I believe it was a PFS hazard ratio of 0.65. I think the response rate was about 60%. There's evidence certainly of an active regimen there and what looks like a favorable benefit-risk profile. What I would say is that this year, we will also read out data in a phase III trial in the frontline setting of Dato-DXd in triple-negative breast cancer. We call that TROPION-Breast02. This study will be a little different. It'll be not in the PD-L1 expressing patients that are eligible for the immunotherapy agent.
It'll actually be dato monotherapy, Dato-DXd, TROPION-Breast02 monotherapy against chemotherapy in patients who are not candidates for immunotherapy. That trial, we don't have a specific month for that, but as we say, the first half of fiscal year 2025. Pretty soon, we'll have the opportunity to look at dato in a similar setting, albeit not the same population, but still triple-negative breast cancer. We'll be able to take a look. We think that dato has significant potential in this disease state. When we studied Dato-DXd in very late-line triple-negative breast cancer, we saw a promising response rate. We think that therefore advancing dato into an earlier line as monotherapy in TROPION-Breast02 has the potential to provide a new treatment option for patients if that result is positive.
What we know about these two drugs is that it's hard to compare them because we don't have a head-to-head trial. They're both TROP2 ADCs, but they differ actually in significant ways. They have different payloads. They actually have different linkers, and they have different pharmacokinetics. There is the potential for differentiation with Dato-DXd, but we have to await the results of the phase III study this year and the one that will come in the future that will combine Dato-DXd with durvalumab in the frontline of triple-negative breast cancer PD-L1 expressing in order to really understand the level of differentiation relative to Trodelvy. More to come, more to come.
Okay. Understood. Thank you.
The next question is from Wakao-san from JPMorgan. Wakao-san, please go ahead.
Hi. This is Wakao from JPMorgan. Thank you for taking my question. I have two questions. Firstly, about TL02. I want to ask about the specifics on QCS. So I view the QCS-based findings encoding. It appears that one of the key objectives of this analysis was to explore the clinical validity of the NMR cutoff used to distinguish QCS positive or negative. Could you confirm whether the NMR cutoff applied in TL02 is the same as the one being prospectively used in the Abundant trial? If so, based on the result observed in TL02, do you believe this threshold is sufficiently powerful to enable the Dato-DXd plus Infringe arm to demonstrate superiority over Keytruda in the Abundant trial?
Thank you. What we've said in this presentation today is that the cut point for the QCS TROP2 NMR in the TL02 retrospective exploratory analysis is the same cut point that was used for the TL01 monotherapy retrospective exploratory analysis. We didn't comment today on the specific cut point that's used in AVANZAR. Actually, it's not something that we can talk about today. However, that may be a good question for AstraZeneca itself that is running AVANZAR. What we can say is that when you look at retrospective analysis using the QCS assay in TL02, looking at the trend toward improving PFS and response rate in the NMR positive patients, we do think that that directionally supports AVANZAR, which is deploying the QCS assay in not the same, but a similar triplet regimen.
Indeed, the idea behind the use of that QCS assay is to increase the probability of technical success, as was the idea to focus the primary analysis in the non-squamous population after we learned that squamous patients did not benefit as much. We do believe, based on the cumulative data sets that we have that are early phase data sets with Dato-DXd immunotherapy combinations with and without platinum chemotherapy that include TROPION-Lung04 and TROPION-Lung02, these data are supportive of the AVANZAR study because you do have triplet regimens explored in TL04 with durvalumab that is actually being used in AVANZAR, and then the TL02 triplet that we showed today with pembrolizumab that is directionally supportive.
Knowing we have an endpoint structure in AVANZAR that will look at both the QCS positive, the NMR positive population, and the TROP2 biomarker positive population and the ITT, we believe this does indeed position us for the probability for the potential to have a positive study and does increase the probability of technical success of AVANZAR. Of course, it would be impossible for me to project the results of AVANZAR, and we will just have to see the results in the second half of the calendar year 2025. We believe the data sets that we have produced are directionally supportive.
Okay. Thank you. Second question about HASLI. Based on the result of HASLI, HASLI and Lung02 study, please provide implication regarding the approval potential of Dato-DXd TL05. I understand that HASLI and that are clearly different for the target, but the data from the HASLI DXD HASLI and Lung02 study is very similar with the result of the TL05, which is submitted. While TL05 appeared to be slightly more later line, TL05 application includes a second line setting similar to HASLI and Lung02. Given the current result and withdrawal of the HASLI application, I'm concerned about regarding the approval potential of TL05. Could you please comment on this point?
Yeah. These are definitely different drug programs. I mean, they did share in similarity, as you say, targeting late-line EGFR mutation positive non-small cell lung cancer. Unfortunately, with HER3-DXd, we did not have the confirmation of the overall survival in the HL02. That being said, I do not believe that the results of this other drug program in HER3-DXd should have a significant influence on our accelerated approval submission of Dato-DXd in late-line EGFR mutation positive non-small cell lung cancer. It is a different drug with a different safety profile. I do not really want to compare efficacy across the drugs because that would be very difficult to do in different studies.
What I can say is that the response rate of dato DXT in very late-line EGFR mutation positive non-small cell lung cancer, which to my recollection was in the 40% range, we believe, and with some durability, we believe provides the opportunity for a positive benefit-risk profile for that drug. We are not too far away from the PDUFA date, which is in July for that regimen. The bottom line is that I do not think the HER3-DXT results, even if they are in a similar patient population, they are with a different drug and a different program. I think we are encouraged by our submission of dato DXT in late-line EGFR mutation positive non-small cell lung cancer for accelerated approval.
Okay. Thank you. Understand.
The final question is from Mike Nedelcovych-san from TD Cowen. Mike, please go ahead.
Hi. Thank you for the questions. I have two. My first relates to HER3-DXd. Is it conceivable that the HER3 and Lung02 data set could be refiled using a retrospective biomarker like HER3-IHC? Or should we consider that biomarker effort truly exploratory? My second question is a broader one relating to the DXd platform. The discrepancy between ENHERTU's striking efficacy in a broad breast cancer population and HER3-DXd's modest efficacy in lung cancer so far is notable, especially as we try to gauge risk to the other DXd candidates. To what do you attribute this discrepancy, given that the linker and payload are shared? Is it differences in the drug-antibody ratio, dose, or therapeutic index, some other aspect of tumor biology? I know this is a complex question, and apples-to-apples comparisons are not possible.
What is your best explanation at this stage, and w hat does that mean for I-DXd and R-DXd?
Yeah. Sorry, Mike, can you repeat your first question again? I got your second one. Just repeat the first one again.
Sure. Sorry about that. Wondering if the biomarker efforts in HER3 and Lung02 could result in a refiling or if it's just exploratory.
Thanks, Mike. Yeah. Thanks for the repeat. Yeah. Basically, that analysis is the biomarker analysis being done is a retrospective exploratory analysis there. Usually in that case, from a regulatory standpoint, and I can't speak for all regulatory filings in all time, but in general, the pre-specification is important. Inclusion in the endpoint structure is important with the biomarker analysis to support registration. There are many other considerations to a companion diagnostic filing where we have analytical validation of an assay, we have clinical validation of an assay, and then we have pre-specification. I believe our efforts here for HER3-DXd are really more in exploratory fashion. If we were to discover retrospectively a biomarker that was very strong in its potential predictiveness, then we could potentially discuss with the regulatory agency.
In absence of that pre-specification, in absence of that biomarker being embedded into the endpoint structure, I think it would be less likely, less likely. Indeed, we are certainly very interested in patient selection for all our ADCs, including HER3-DXd. We will continue on. As soon as we can share data looking at patient selection using HER3 immunohistochemistry, we will do that. Your second question was related to the translatability of ENHERTU to other assets like HER3. It relates in a way, Mike, to the biomarker discussion we just had. ENHERTU is, of course, a different drug, and it has a predictive biomarker.
In HER2 being a HER2 ADC, I mean, we know this because if you look at the segments there, and Daiichi Sankyo with this drug has been really continuing to change the risk classification of HER2 of patient selection. If you look at the segments, right, HER2 expressing based on biomarker selection, HER2 low segment based on biomarker selection, HER2 ultra low based on biomarker selection. That is very different than HER3, where we have not shown data at this point in a broad set of patients with a biomarker selection approach. The first thing I would say is that one program is heavily entrenched, of course, with biomarker selection. The other program is not yet, and we're working on that.
The other thing is to say that even though we have the same deruxtecan linker payload system with, let's say, in HER2 and HER3, they have different targets. They have different targets. Those different targets relate to different aspects of tumor biology. Given the study of one DXd ADC in the same tumor as another DXd ADC, we may not see the same type of efficacy by virtue of the tumor biology associated with the specific target. I would note, Mike, though, that if you look at HER3 specifically, because you mentioned it, I did allude to the fact that at ESMO last year, we had the ICARUS-Breast01 trial. I think it was 99 patients in a phase II single-arm trial, second line plus metastatic hormone receptor positive breast cancer with a response rate that was just above 50%.
It was just above 50% without biomarker selection. I think that does tell you that with the specific example that you gave, that actually the DXT ADC platform does demonstrate that more than one drug can have activity in breast cancer. Indeed, one can imagine that if we were able, hypothetically, to find a patient selection strategy for HER3-DXT in breast cancer, then that might even have an additional impact on efficacy. More to come as we continue to explore patient selection across our DXT ADCs and beyond in HER2, where we have the patient selection strategy quite mature.
Great. Thank you.
Okay. We apologize that we aren't able to take questions from everybody, but we have run over the scheduled time significantly. We'll now conclude Daiichi Sankyo's ASCO 2025 highlights. Please contact our investor relations team for remaining questions. Thank you for joining today.
Thank you.