Good afternoon and good morning. Thank you for joining our analyst briefing to cover data presented at the European Society for Medical Oncology in recent days. Today's discussion will be focused only on recently presented clinical data. We will not be providing any commentary on performance as we report our Q3 results tomorrow. We are joined today by Nina Mojas, our President of Global Product Strategy, Hesham Abdullah, Senior Vice President and Global Head of Oncology R&D, and Ahmed Rabie, Senior Vice President of Global Product Strategy for Oncology. We are also joined by Eric Richards, Richard Davidson, and Claus Becker from the Oncology R&D team for the Q&A portion of this session. For the agenda, Hesham will take you through the data, and then Nina will provide some closing remarks before opening up for your questions. I will now hand over to Hesham.
Jeff, thank you very much, and it's a pleasure to be with everyone. Good afternoon to those in Europe, good morning to those in the U.S. And thank you very much for joining us. Of course, we are just back from Madrid. Very exciting ESMO. More than 21 abstracts, and a number of different oral presentations as well too from the GSK oncology pipeline. I am happy to take maybe the next 25-30 minutes to walk you through some of the key data from our oncology portfolio that was recently presented at ESMO in Madrid. With that, I think we could probably start off first with ris-rez, which is our B7-H3 ADC, of course. I think first and foremost, it's important to highlight that we actually presented some data from across our phase I-II development program.
This is the GSK development program, global development program, more specifically as well too. We actually had data from our phase I-II study, and specifically in small cell lung cancer. I think what's probably important to highlight is this data certainly showed consistency between the global data set that we have generated in terms of objective response rate and second line plus small cell lung cancer with a 61% response rate in the 8 mg per kg dose cohort. This is actually the dose that will be used and is being used currently in the second line phase III small cell lung cancer study across our own global development program. Then, of course, data that's been generated recently, at World Lung from our partners, Hansoh, and the ARTEMIS-008 study, where we saw an objective response rate of 58%.
I think probably the first key headline to take is certainly the data from our emerging global development program is very reassuring and confirms certainly the clinical activity that's being observed in the China Hansoh development program. Next slide, please. This is, of course, the data that was recently presented by Hansoh from the ARTEMIS-008 trial, looking at ris-rez versus topotecan in second-line small cell lung cancer. I think probably what's important to highlight and remind everyone of course, is the study was positive, demonstrating a statistically significant and clinically meaningful overall survival benefit for ris-rez relative to topotecan, with a median overall survival of 18.5 months versus 10.3 months on the topotecan arm. As you can probably see from the Kaplan-Meier curves, the curves separate early. That separation is maintained over time.
At a landmark of about 12 months, which is certainly the median duration of follow-up on the study, there were about maybe 64.5% of patients still alive on the ris-rez arm versus 43.3% on the topotecan arm. Next slide, please. This is a slide showing certainly a consistency of the overall survival effect across key subgroups from the ARTEMIS-008 trial. What I'd probably call your attention to is the fact that about maybe 17% of these patients that were recruited into the study did have brain metastases at study entry. What we did observe was that there was certainly a consistent treatment effect across this subgroup as well too, with a hazard ratio of 0.36 for overall survival, but generally, certainly consistent survival benefit observed across all predefined subgroups, whether it be based on age, smoking history, prior PDL1 therapy, and of course, performance status as well.
Next slide, please. Key secondary endpoints, of course, PFS by independent and investigator review, quite consistent, with a hazard ratio of 0.33 for the BICR-assessed PFS, 0.35 for investigator-assessed PFS, with about maybe a four month delta on median PFS, certainly, whether it be based on BICR or investigator assessment as well too. Next slide, please. Of course, objective response rate, disease control, and duration of response. What I'll probably call your attention to on this slide, of course, is the objective response rate that's been observed and I've noted previously, 58.3% for ris-rez versus 12.8 for the control arm with single-agent topotecan. I would say maybe close to a fivefold difference in the magnitude of effect on response, supported of course by the median duration of response, close to about maybe seven months and the disease control rate as well too.
One point to highlight here, which I think is really important in terms of how we think about the extrapolation of data from China. What you see in this study specifically is there's a really strong internal consistency of results. If you look at the primary endpoint and the magnitude of effect that's observed on overall survival with a hazard ratio of 0.46, key secondary endpoints like PFS with a hazard ratio of 0.33, and then this fivefold difference in response rate as well too. That strong internal validity and consistency within the phase III study itself gives us certainly strong confidence in its external validity as well too, and broader application across other countries and regions. Next slide, please. Maybe just a quick summary then.
Again, of course, just the caveat as always of looking at cross-study comparisons and the differences that may exist between baseline characteristics across each of these studies, just taking that into account. But at least when we look at a high level, the median overall survival from the ARTEMIS-008 compared to some of the emerging data across the B7-H3 class, but also some of the existing data as well too, in second-line small cell lung cancer with immune cell engagers. No doubt the median overall survival with ris-rez at 18.5 months, certainly quite well-positioned in that regard. We have two phase III studies in small cell lung cancer, an ongoing phase III and second-line small cell lung cancer, which is the EMBOLD-SCLC-301 trial, which is currently recruiting.
Then a second study in first-line small cell lung cancer that will be initiated very shortly before the end of the year in the first-line setting as well too. Next slide. So one of the additional data sets that were actually presented as well too at ESMO was an analysis that we conducted from our own global development program looking at the incidence of ILD, given that this is certainly an event of interest across the TOPO1-based class of ADCs as well too. When we have looked at this across our certainly data set, whether it be across all solid tumors or specifically in non-small cell and small cell lung cancer, we do see a relatively low incidence of grade three or greater ILD events at 0.9%, so less than 1%.
Then, of course, in the overall at least aggregate population from our own development program, across all solid tumors, a 4% incidence of all grade events. Again, the majority of these events were, of course, mild or moderate grade one and two in nature. I think when you look at also the data that has been published from the Hansoh experience in China, whether it be with the ARTEMIS-003 or ARTEMIS-002 studies, these are phase I and II trials. We do see a relatively low incidence of grade three or greater ILD events as well too. Of course, ARTEMIS-008 being the exception to that with an incidence of 4% of grade three or greater ILD. Next slide. This was, of course, all presented on a poster at ESMO that included, of course, the ILD data for reference. Next slide, please.
Then, of course, data that was presented just yesterday in an afternoon presidential symposia at ESMO. I would say quite a landmark result, just given certainly how challenging this disease is, osteosarcoma, quite poorly prognostic. We have not seen any advances in standard of care over the past 50 years, as one leading sarcoma expert had told us. I think that is really important to highlight and emphasize as well, too. Next slide, please. So this basically was a study looking at osteosarcoma patients in an advanced metastatic second-line disease setting. So patients had received certainly greater than one to two prior lines, specifically at least two prior lines. Patients were randomized two to one, either to receive ris-rez or certainly standard of care gemcitabine docetaxel chemotherapy. Then the primary endpoint of the study was PFS by blinded independent review.
Key secondary endpoints included overall survival and PFS and objective response rate. Of course, given the sample size of the study, which is 122 patients, the study is not powered for overall survival. Next slide, please. Key baseline demographics. When we look at the median age of these patients, it is ranging between 21- 22 months, and I think that is really important to highlight. Just given the patients that are typically impacted by osteosarcoma, we have about maybe a quarter of patients that were actually adolescents between 12- 18 years of age. So certainly, at least the incidence tends to also occur in younger patients, as evidenced by the data. I think one thing to probably also highlight is this is a patient population that no doubt had a relatively high disease burden, as evidenced by the fact that the majority of these patients had pulmonary metastases.
And of course, is also heavily pretreated. The majority of these patients had received prior chemotherapy and targeted therapy as well, too. Next slide, please. The primary endpoint, BICR-assessed PFS, was met, demonstrating a statistically significant and clinically meaningful improvement in PFS with close to a fourfold improvement, five point five months versus one point months in terms of the medians.
Again, you see the curves separate relatively early and certainly stay separated until that six month landmark. I think it's important to highlight, of course, the median duration to follow-up at this point in time is about maybe seven point one months. Next slide, please. Tumor response assessment by BICR as well, too. The objective response rate is 15% on the ris-rez arm and 2% on the gemcitabine docetaxel arm. Now, one thing to bear in mind, and especially with this type of tumor, just given the location of the tumor itself in the bone.
Oftentimes, one of the challenges is really getting enough exposure to be able to penetrate these lesions, especially given the thickness of the extracellular matrix that exists around the tumor and the capsule within the bone itself as well, too. So I would say probably objective response rate is not necessarily the most meaningful way to assess the activity of the drug, but rather, of course, a disease control rate may be a better proxy along with PFS. And here, as you can probably see, the disease control rate is at 76% on the ris-rez arm versus 33% on the control arm. I think one other point to highlight, the dose that was used of ris-rez in this study was 12 mgs per kg.
I mentioned and alluded to the fact that higher exposure may be needed here, given the need to be able to better penetrate, of course, the tumor, especially given its location and its site as well, too. Next slide, please. Overall survival data, currently immature. It's about, I would say, 16%-17% mature at this point in time, with about 10 events having occurred in each arm. The study did allow for crossover, so we did note that 17 patients in the control arm have already crossed over to ris-rez treatment following progression on gemcitabine. And despite that, we're still seeing certainly a very encouraging trend at this point in time on overall survival with a hazard ratio of 0.51. And again, awaiting further follow-up, just given the relative immaturity of the overall survival at this point in time. Next, please.
Safety profile, I would say certainly as expected, given the dose that was used here, but also the typical Topo I-based payload assessed toxicities or related toxicities all predominantly relating to myelosuppression, including, of course, neutropenia and, of course, thrombocytopenia, being the key ones. I think the one point to highlight here that's really important as well, too, despite certainly the higher grade three event rates across the myelosuppression, discontinuation rates were quite low at 4%. Next slide, please. Moving on to our B7-H4 ADC, mo-rez. We did provide an update on the data that was actually presented at SGO in April. So specifically, the platinum-resistant ovarian cancer cohort with longer follow-up now, about maybe a little over 13 months of follow-up from that cohort. Just maybe a quick, I would say, brief on the design of the study, of course. So this is our phase I-A/I-B study.
There was an abbreviated dose escalation, then a movement into dose optimization and expansion across three different dose levels: 2.8 mg, 4.8 mg, and 5.8 mg q3w. That will be the focus at least of the data that will be presented on the following slides, and specifically the platinum-resistant ovarian cancer cohort. Baseline characteristics, maybe just to call out the fact that a few key elements to this. One, the majority of these patients had received certainly prior bevacizumab, and more than half had received a prior PARP inhibitor as well, too. The other, of course, just to note here is that the majority of patients, close to about maybe 80%, had received two or more prior lines of therapy, including 36% of patients that had actually received three or more prior lines of therapy as well, too.
A large portion of patients had tumor tissue available for B7H4 expression evaluation. Next slide, please. What you'll see on here, of course, is the waterfall plots for objective response rate. I think the one point to call out more specifically here is that these are confirmed objective response rates across all dose levels. What do we see? Well, what we see is certainly an objective response rate of 36% across the 2.8 mg dose, 57% across the 4.8 mg dose, and 62% across the 5.8 mg per kilogram dose, which is the registrational dose that we're taking forward in this indication. There's also a trend towards shorter times to response relative to dose with the shortest time to response on the 5.8 mg per kg dose level. Also a trend towards depth of response as well, too, relative to dose in that respect as well, too.
What's probably also important to highlight here is the fact that responses, as you'll see on the following slides, were observed irrespective of B7H4 expression levels as well. We did have actually four patients that had received prior mirvetuximab treatment. Two out of the four patients developed a partial response. Next slide, please. This is certainly just data specifically on duration of response. Again, what we do see is a dose response, certainly trend here as well, too, with the median duration of response, of course, being longest and strongest, certainly on the 5.8 mg per kg dose level at 12.2 months. Along with, of course, the probability of duration of response at the six month landmark as well, too, with certainly that separation being evident from the Kaplan-Meier plot on the right-hand side of the slide. Next slide, please. The same holds true for PFS.
We saw the longest PFS on the 5.8 mg per kg dose level with a median PFS a little over 11 months and a probability of PFS or being progression-free at six months of 80%. I think all of these data in aggregate reinforce our certainly confidence in selecting the 5.8 mg per kg dose level for the platinum-resistant ovarian cancer indication as well, too, along with a few of the other phase III studies that we're initiating, including the second-line endometrial cancer study and the platinum-sensitive ovarian cancer study as well, too. Next slide, please. This is data that was also presented at a separate oral presentation that was more focused on translational biomarkers from this phase I-A/I-B study as well, too.
And again, as you can probably see, we had a range of different, I would say, on the left-hand side of the slide, B7-H4 expression levels represented in the sample size of patients that were treated. Then on the right-hand side of the slide, and again, based on the color coding, as you could probably tell, responses were observed across a range of expression levels, including depth of response as well. Next slide. The safety profile, as expected, and based on the topoisomerase warhead, dominated by GI and also myelosuppression events. I think just a few key points to call out in this respect. First, across all dose levels, the 2.8 mg, the 4.8 mg, and the 5.8 mg dose levels, the incidence of all grade ILD was 7%, and specifically, at the 5.8 mg per kg dose level, it was 2%.
And the majority of these events, eight out of the nine events that had actually taken place, were grade one/two in nature. Only one event was grade three or greater. One thing to also call out as well is really just in terms of the phase I study design itself. It did not mandate the prophylactic use of growth factors to mitigate neutropenia, as is evidenced by the data that we have in hand, and that is G-CSF was only used in 8% of patients at the 5.8 mg per kg dose level. And I think for the phase III studies, we are being much more prescriptive and strongly recommending the use of G-CSF to mitigate the neutropenia from the start of treatment as well. And this is being implemented across all five phase III studies. Next slide, please. And so just a quick recap.
I've mentioned, of course, that we've initiated five phase III trials across platinum-resistant ovarian cancer, platinum-sensitive ovarian cancer, first-line HR proficient maintenance ovarian cancer, and then also second-line endometrial cancer and first-line MMR proficient maintenance endometrial cancer as well. Next slide. Moving forward to dostarlimab, and of course, this is data from the AZUR-1 study. I think everyone's well aware, of course, that we had announced last week the regulatory approval in the U.S., and then the data followed over the weekend with an oral presentation as well, from the AZUR-1 trial in the locally advanced dMMR MSI-high rectal cancer patient population. Next slide, please. And maybe just to provide a little bit of context, I think this is certainly a patient population that is typically treated with what is multimodal therapy, a combination of chemoradiotherapy and surgery.
Unfortunately, of course, the tolerability challenges and comorbidities that has on patients is quite difficult, resulting in certainly gastrointestinal dysfunction, certainly bowel, and also urinary dysfunction, and certainly in certain cases also impacting fertility as well. Historical, I would say, clinical complete response rates for this patient population range between 20%-25%. And I think that's important when we think about how to contextualize the results from the AZUR-1 study as well. Next slide, please. Baseline characteristics for this patient population. I think probably first and foremost, the median age was about maybe 61 and a half years. But I think what's probably important is actually to look at also the confidence intervals around that. The range is about maybe 23-92, indicating, of course, that we're seeing these cases even in younger patients as well.
It also highlights, of course, this emerging trend of younger onset disease across multiple types of cancers, including colorectal cancer as well. The majority of patients had bulky stage three disease coming into the study at the time of entry. One thing to point out as well, too, in terms of the types of tumors. About 15% of patients had a mucinous type of tumor, which tends to probably be a bit more challenging or obscure in terms of how it shows up in the MRI scans as well, making it a bit more difficult for physicians to be able to assess clinical complete response. Next slide, please. The primary endpoint of the study was clinical complete response at 12 months. Patients actually received nine doses of dostarlimab or about maybe six months of treatment.
The cCR12 rate by investigator assessment, at this time was 68%, as evidenced by the light blue bars. These are the patients that actually were able to achieve cCR12 and had enough follow-up for them to be assessable for that endpoint. I think the one point to highlight on the left-hand side in the black bars as well, too, is that there is a number of different patients that have achieved a clinical complete response but do not yet actually have enough follow-up to be assessed for cCR12 at this time. So we will continue to be following patients moving forward as well, too. Next slide, please. These are the swim lane plots certainly demonstrating the durability of responses that have been observed across the AZUR-1 patient population.
I think what is probably important to highlight here is, of course, is that 98% of these cCRs were durable, and to date, no recurrences have actually occurred so far. Next slide, please. I think probably just to provide a little bit more color around the type of benefit that patients were deriving. So cCR12 is an endpoint that is typically assessed through two different methodologies. One, I mentioned the MRI. The other, of course, is through an endoscopic assessment. In certain instances, when there may be a bit of ambiguity around being able to declare a cCR12, or there is maybe uncertainty from the physicians around it, oftentimes what they will do is they will actually biopsy the patient as well, too, and get access to tissue.
So there is a segment of patients in the dark blue bars that actually have what is called a pathologic complete response, at the time of the biopsy and surgery as well, too. Limited number of patients that had actually undergone that of just about 12. But when you look at the combined cCR 12 plus pCR rate, 76%, which is certainly, again, if you are comparing it to the 20%-25% clinical complete response rate that is historically observed, quite significant and quite clinically meaningful, especially when you think about the fact that the majority of these patients were able to not only avoid chemotherapy, but also avoid surgery as well, too. Next slide, please. Then finally, I think probably just important to be able to see what this looks like. I mentioned, of course, the two different approaches to assessing, certainly clinically, the clinical complete response.
One is through the MRI, the other is through the endoscopic exam. This is exactly what you are looking at. Basically, this is an endoscopic assessment of the tumor in the rectum. Starting first maybe on the left-hand side, what you are seeing is, this is actually one of these mucinous tumors, which are typically difficult to assess on an MRI scan. As you can see, certainly both from the anterior and also posterior walls, significantly encroaching, and causing an obstruction within the rectal canal as well. Then basically the baseline evaluation after getting treatment, the patient getting treatment with dostarlimab for six months, nine doses. What you can see is there is a start of maybe normalization of the rectal mucosa and epithelium. Specifically, you see that manifested through the flattening, the smoothening as well, with probably some very residual scarring.
I would probably say even slight residual evidence of the tumor. Then finally, of course, after the second follow-up assessment, complete normalization, and no evidence of the tumor exists as well. Certainly that gives you a sense of what this looks like in terms of the assessment of cCR as well. The safety profile, certainly very consistent with what is known about dostarlimab and immune-mediated adverse events. Important to highlight, low incidence of grade three events, with these nine doses of dostarlimab as well. So a low tolerability burden for patients. Next slide, please.
Just to conclude, certainly, we have seen an evolution, in what will be the standard of care for locally advanced dMMR MSI-high rectal cancer patients from where it was previously, as I had mentioned, in the context of the prior treatment paradigm with chemotherapy and surgery to where we now see more than 2/3 of patients achieving a cCR 12, 76% of patients achieving a cCR 12 plus pCR, and certainly quite strong durability of these clinical complete response rates with a low adverse event burden for patients as well. Next slide. I think maybe just one reminder, from another data set that was presented at the World Lung Conference as well. Very exciting data from our Nuvalent acquisition, as everyone will probably recall.
Zidesamtinib or JIDEYTRO was actually approved in late July in the U.S., two months ahead of the PDUFA date, which I think is important to highlight as well. Here is actually data being presented from the TKI-naïve cohort in advanced metastatic ROS1-positive non-small cell lung cancer. Very quickly, I think probably the baseline characteristics as expected, with about maybe a little over half of patients being females, and certainly the same for patients being non-smokers as well. Certainly, a good portion of these patients being fit in terms of status, with an ECOG performance status of zero or one. About 17% of patients having baseline CNS metastases as well, and the majority of patients having advanced disease, and a quarter having previously received chemotherapy or chemo-immunotherapy. Next slide, please.
Just to highlight, of course, the waterfall plot from 94 patients that are TKI-naïve, an objective response rate of 94% based on independent review, with a CR rate of 15%, which is actually close to doubling the CR rates that we have observed with third-generation TKIs as well too, which I think is important to highlight. Next slide, please. Duration of response, also extremely encouraging at this point in time, and specifically looking at the landmark duration of response at 12 months, being 86% with the median duration of response not yet reached at this point in time. Next slide, please. The same holds true for progression-free survival, looking at the landmark PFS at 12 months of 90%, and the median PFS not reached. Now, one thing that probably for us to think about and consider as well too is how this drug was actually designed.
In terms of the medicinal design of the drug, it was really designed to have this brain-penetrant activity while also be sparing as well too, to minimize any CNS effects, or side effects that is, while also making sure that it addresses intracranial or brain metastases as well. That was certainly evident in the ARROS-1 study in this TKI-naïve cohort. In the subset of patients that did have brain metastases at baseline, intracranial objective response rate was 100%, and certainly the intracranial CR rate, complete response rate was 70%, which is quite profound. Then, of course, the median duration of response in these intracranial metastases is not, or at least duration of response in the brain metastases has not been reached. Tolerability-wise, I would say certainly a very good tolerability profile, low incidence of grade three events.
I think probably I will call your attention to the right-hand side of the slide where we did see a really low discontinuation rate, only at 4%, and specifically that related to treatment-related adverse events at 1%. Dose reductions also were infrequent at 11% and 12% as well, too. With that, I think what I could probably do is hand it over to Nina for closing remarks.
Thank you, Hesham. I am reflecting on the data and the amazing data that were presented, but also on Hesham’s ability to present all of this in such a short period of time. So thank you, Hesham. I think it is quite impressive. Okay, before we go into Q&A, this slide is trying to capture the importance and the impact of the data. So to focus first on the first two boxes, so ris-rez in second-line small cell lung cancer, why are these two data points important? The first one, you have seen that in the Western cohort, so Western population of second-line small cell lung cancer, we are seeing 61% response rate. That correlates very well with 58% response rates that has been observed in the Chinese phase III study, ARTEMIS-008.
And that study has shown unprecedented, the longest median overall survival that has been seen in the second-line small cell lung cancer. I think this data give us confidence in the activity and the promise of ris-rez in second-line small cell as we are going to witness from the phase III study, which is ongoing at the moment. So very high confidence in second-line small cell for ris-rez. Another set of data, ris-rez in relapsed refractory osteosarcoma. For anyone of you who were there yesterday at the session and listening to the discussant, I think you get the full picture of what that disease is and how devastating it is, and the lack of progress over the last decades. For ris-rez specific, it is an illustration of this asset being a pipeline in an asset with yet another indication, very difficult to treat indication, successfully tackled.
JEMPERLI, over the last few years, we have heard repeatedly the magic of JEMPERLI in the dMMR locally advanced rectal cancer. Now pivotal data in a study that on its own is quite unique, have confirmed the efficacy, and we are actually just in the process of launching following the approval in the U.S. JIDEYTRO, ROS1 inhibitor in TKI-naive, was presented at WCLC a couple of weeks ago, again, demonstrating best-in-class efficacy. Based on this data, FDA is going to review the first line label, resulting hopefully very quickly in the approval for indication for TKI-naive patients. Last but not least, mo-rez. Actually, mo-rez had very good ESMO with data presented in PROC or platinum-resistant ovarian cancer.
Looking at the data, response rates, plus duration of response and PFS that has been observed indicate that this asset is at the top end, if not the best in efficacy in this setting. For mo-rez specifically and ovarian cancer, obviously PROC indication is the beginning. ADCs have intense development programs in ovarian cancer. The question is going to be, where is which asset positioned, either in platinum-sensitive ovarian cancer or eventually in newly diagnosed, so first-line ovarian cancer. With this, again, I am just going to thank Hesham for an amazing presentation and the read-through. I think we are going to open the questions, and Jeff is going to manage that.
Yes. Thank you for those of you who have not raised your hand and have a question, please feel free to raise your hand and we will try to get through as many of you as we possibly can. I see James Gordon is up first from Barclays. James, you will be unmuted and you can ask your question.
Great. Thanks a lot for taking the question. One on ris-rez and one on mo-rez, please. In ris-rez, the efficacy looks good, but I think there is a 42% grade three AE rate in the phase I/II. Does that look challenging? It does not look like it is lots of ILD. That looks all right, but it looks like these are the things that 42% of patients had to have dose interruptions or delays. Might it be a bit too high a dose, and is that something you think you can improve in terms of tolerability in phase III, and would you want to try and improve it before taking something that challenging into phase III? That is the first question, please. The second question was on mo-rez. The B7H4, where the data looks strong in ovarian and endo.
But then on Friday, Genmab had also very strong phase III efficacy data for rina-s, its folate ADC, and there are a few other ones of those in development from other people as well. How do you think about folate ADCs versus B7H4s for gynecological cancer? Because it looks like at best, mo-rez is probably about the same, but they will already be on the market and quite well-established. Is it going to come down to some sort of subgroups or something like that, or might it be you do a folate, then a B7H4?
James, thank you very much for the question. I will start off first maybe just addressing the question around ris-rez. I think probably 42% that you might be referring to is coming from the osteosarcoma study, and I think maybe as I have just highlighted as well, the dose that was used in that trial was actually the 12 mg per kg q3w dose. That is not the dose that is currently being used across a number of key tumor types that we are exploring. Specifically, we are actually using the 8 mg per kg dose level q3w across both small cell and prostate cancer as well. I think from our perspective, again, most of the, I would say, safety, tolerability, at least, profile is very much being driven by the myelosuppression and the neutropenia.
From our perspective, certainly the use of growth factors in the phase III studies should certainly help mitigate that. At the same point in time, we are not using the high dose, the 12 milligram per kilogram dose that Hansoh used in their osteosarcoma phase III trial. We are actually using the 8 mg per kg dose level. In terms of the second question that you actually had as well, and specifically regarding mo-rez. First and foremost, of course, I think we are all well aware that we saw that the announcement from Genmab around the RAINFOL study as well, and announcing positive results. From my perspective, first and foremost, I think it is important for patients to be able to have treatment options, and especially with this emerging class of antibody drug conjugates.
With that in mind, I think we have to look at the data that we have at hand. There are things that we know and things that we do not know. The things that we know are the hazard ratios from that trial for PFS and overall survival. The things that we do not know are the medians. However, I think, again, acknowledging the caveats that exist with cross-study comparisons, what we do have is actually phase I data from across both prog programs. When we look at the phase I data, specifically with mo-rez, which we just presented over the weekend in PROC, we saw a confirmed objective response rate of 62%. From the phase I rina-s lead study, it was 46%. That is one.
Second, when we looked at also PFS as well, too, I think as you probably saw, at the 5.8 mg per kg dose level, the median PFS was 11.1 months for mo-rez, and then for rina-s, it is about maybe nine and a half months. I think we just have to keep in mind that at this stage, we have to wait and see what the data looks like from their phase III trial. Certainly, I think what has probably become evident is that there is some level of relationship between clinical activity in the context of response rates and depth of response, but also certainly at least with these ADCs, the duration of response and the PFS as well, too.
At this stage, we certainly feel pretty comfortable with the level of clinical activity that we are observing, and then how that is manifesting in duration of response at a little over 12 months, and then PFS a little over 11 months as well, too. In the context of subgroups, again, I think for us, when we looked at expression levels, as you probably saw from the data itself, we see no correlation between clinical activity and expression levels. Broad activity across, whether it be B7-H4 expression low or high. I think that is data to be awaited from their phase III trial as well, too, to see what, if any, relationships may or may not exist with FR alpha expression levels as well, too.
I would just add for PROC is the first step for all of these ADCs in ovarian cancer. That eventually, hopefully, the efficacy is going to confirm for many of these to be moved into earlier line of therapy, and this is probably where the more important competition is going to happen. As you know, studies are only starting in these settings as we speak.
Excellent. Let's move to next question from Shyam Kotadia from Goldman Sachs. Shyam, we will unmute you.
Yeah. Thank you so much for the question. Can you hear me?
Yes.
Perfect. There is a lot of focus on novel combinations in oncology with ADCs and IO backbones. How do you think about this potential in gyn onc? We have seen activity with VEGF in ovarian cancer, so could B7-H4 be a potential partner for VEGF PD-1 bispecifics? Thank you.
Yeah. Shyam, thank you very much for the question. I think we will start off first by saying, no doubt that the standard of care has evolved in endometrial cancer, and we have seen that, of course, with the advent of checkpoints. As you may probably be aware, our first-line endometrial study in the MMRp segment of patients, it is a maintenance trial, but it is also looking to combine mo-rez or B7-H3 with the existing standard of care, PD-1 agents. That is probably one component of it. The second probably is getting directly to your question, which relates to could combining with bispecifics in this space be an opportunity? Certainly we have seen some preliminary data from other competitors, including Pfizer as well.
There seems to, at least again, given the single-arm nature, uncontrolled nature of the study, there seems to be a signal of additive effects between PD-1, VEGF and chemotherapy. It is something that we certainly take into account. Maybe the one thing that I would just highlight, of course, is that we do have an existing collaboration with Summit Therapeutics. Specifically that is, at this time, directed more towards the ris-rez program, and exploring combinations of ris-rez with ivonescimab in lung cancer across both small cell and non-small cell lung cancer as well, too. I think it is something that we would take into account, take into consideration, and maybe think about.
Excellent. We will go next up for a question from Peter Verdult from BNP Paribas. Peter? You are unmuted, Peter. You can ask. You there, Peter?
Can you hear me?
Yes.
Sorry, guys. I am actually at ESMO and, yeah, schoolboy error on the mute function. Sorry. Peter Verdult, BNP. Thanks for the presentation and time. Just a couple for Hesham, please. Lots of doc excitement around the B7-H3 ADCs, but docs saying it is too early to differentiate, and perhaps pointing to some of your competitors having stronger data in patients with brain and liver mets. I would love to hear where you agree and disagree with the idea of there being a lack of differentiation right now amongst the B7-H3 players. Secondly, on development plans for mo-rez, just to follow on from James's question earlier on, not on folate itself, but look, anyone who was at Madrid, and I am sure you guys were, it was a tidal wave of ADC data, lots of different targets, but a lot of them are Topo I- payloads.
So the idea of are you really going to do sequential therapy if you failed on a Topo I- payload and the idea of acquired resistance. Would love to get a little bit of thoughts there in terms of how that shapes your development plans. Thank you.
Yeah.
Peter, thank you very much. Let me start off first with the question around ris-rez and B7-H3 and the class of agents and differentiation as well too. I think first and foremost, we probably need to re-baseline in where we are today. Where we are today is, we have dosed more than 1,100 patients across our development program, the GSK Global Development Program. Hansoh has two positive phase III studies out of two phase III trials that have read out data across second line small cell lung cancer and across osteosarcoma. Those are facts that we can start with. I think certainly, the data that we have seen, at least from our own small cell lung cancer experience, seems to also give us confidence in the translatability of those results moving forward as well too.
With regards to differentiation, I think probably maybe just it is helpful because I called it out during the data presentation as well too. I think it is helpful that maybe if we can go back to the slide specifically with the subgroups, because I know you touched on, for example, patients with either brain mets or liver mets. I think it is a good one to maybe just look at in the context of the data itself, from across the ARTEMIS-008 trial. Specifically, in one of the pre-defined subgroups. Maybe 17% of patients that were actually recruited into the study had brain metastases at baseline. The hazard ratio for overall survival in that segment of patients was 0.36.
I think from that standpoint, certainly, we feel pretty comfortable in terms of the drug's effects, not only systemically, but also in patients with brain mets at baseline. I think that is probably important to touch on. With regards to differentiation relative to other agents in the class, I think really when you think about, and we will touch on this with mo-rez as well too, when you think about physicians always think about two things. They think about efficacy and the type of benefit that patients are experiencing, but also tolerability and safety as well too. We have touched on the ILD, which I think certainly, across the class has been variable, including with other types of platforms. I think maybe that is one thing that we have to take into account is maybe not all linker payloads, Topo 1 payload platforms are created equal.
I think that's one thing that we need to consider. When you look at, for example, the incidence of ILD across our own global development program that was presented in ESMO, 4% all grades, 0.9% grade three or greater. That's the safety side of it. In terms of efficacy, I think we've also touched on at least the data from the second line small cell, ARTEMIS-008 study. Median OS, 18.5 months, when it's benchmarked either to other competitor trials or existing standard of care in that setting with immune cell engagers as well too. I think probably, certainly the data, the clinical data will be a key driver of decision-making for physicians. I think that, at least as you look at the class as well too, safety will become another key differentiator.
I think over time, no doubt, as we continue to evaluate and assess biomarkers, translational endpoints in other tumor types, this is something that will no doubt become a key part of how physicians make choices along with combinations, combination partners. A good example of that, for example, is as we're moving into frontline small cell lung cancer, we're considering three different opportunities. One, which is the now combinations with PD-L1. Two, looking at combinations with bispecifics. Then three, of course, looking at combinations with immune cell engagers. I think that will be equally important for physicians as they make choices moving forward. In regards to mo-rez, I think you touched on a fact, which is, there's been an explosion of antibody-drug conjugates in gyn onc malignancies. I think we have to all acknowledge that and be grounded in it.
Again, like I said, I actually think that that's a good thing for patients from that standpoint. I think at the end of the day, I think physicians are going to use the same criteria. They're going to use safety and efficacy. I think when we look at across a number of different classes of agents, whether it be TROP2, whether it be- Looking at other B7-H4s, whether it be the CDH-6 class or others, certainly safety is going to be a question. It's going to be a question whether or not there's, for example, some that come with ocular toxicity, some that come with ILD, some that come with stomatitis. Those are all key variables that they will take into account, and that's what they tell me when I speak to them. The second really will be the efficacy.
I think we've seen a variation in the efficacy as well, too, in the performance. Now, we don't know if that's related to internalization of the target or the antigen itself. We don't know if there may be some slight differences in the Topo I-based payloads or not. But I think one of the things that we saw over the weekend was, no doubt, these ADCs are certainly evolving the environment within the tumor. You probably saw the data that was presented from the mo-rez phase I study, where you saw that certainly it was inducing some immunological changes within the tumor microenvironment as well, too. We'll continue to learn about that, but I guess probably what I'm saying is, typically across the entire ADC class, we've seen response rates range between 40%-60% or 65%. Mo-rez is delivering a 62% objective response rate.
It is in the upper tier of that class. I think that's something that's important to anchor to as well, too, at least in platinum-resistant ovarian cancer. I think for us, again, continued work on biomarkers, translational endpoints, over time to continue to better assess whether or not there's any additional ways to optimize benefit risk is something that we're going to continue to evaluate and look at.
Thank you.
Excellent. Let's go to the next question. Thanks, Peter. From Seamus Fernandez at Guggenheim. Seamus, we'll unmute you. Please ask your question.
Sounds good. Can you guys hear me okay?
Yes.
Great. My question is actually on the evolution of treatment in ovarian cancer. We have sort of the PARP1/2s in the maintenance setting. These ADC data are obviously establishing a unique new bar, potentially for treatment. Just interested to know how you're thinking about the sort of utility of a PARP1/2, like Zejula, in the context of these evolving data. Also, there are more highly selective PARP1s that were revealed at ESMO. How are you thinking about the potential for selective PARPs emerging as potential paradigm-changing or shifting assets? Or is your focus really going to be exclusively on advancing your ADC programs in combination with other therapies like immunotherapy, ahead of pursuing potential PARP combinations? Thanks.
Yeah. Seamus, thank you very much for the question. I'll start off first by saying, no doubt, the advent of PARP inhibitors has had a really meaningful impact in advanced ovarian cancer patients. I would say probably much more, of course, in the BRCA mutant and HRD-positive segment of patients, and it'll continue to be certainly a key part of the decision-making and treatment for that segment as well, too. Now, one thing I maybe just want to reiterate and highlight as well, too. When we think about Zejula is actually a PARP inhibitor that is more selective for PARP1 than it is for PARP2. I just want to make sure that I point that out first and foremost.
In terms of how we're thinking about the landscape, I think for us, really, no doubt Zejula has continued to do well in those segments where it's approved currently, whether it be in the U.S. or Europe. At the same point in time, our focus is definitely on mo-rez moving forward. That includes not only thinking about later lines of therapy, but also moving into earlier lines as well, too. I think maybe one thing that's maybe a bit underappreciated is the fact that we actually have access not only to data that we're generating across our own development program, but also the Hansoh development program as well, too. I can tell you that I see this data with the team, including, of course, some of my colleagues in R&D that are with me today, Richard Davidson, Eric Richards and Claus Becker.
We see this data in real-time, and I can tell you there's really interesting data that we see, of course, in platinum-sensitive ovarian cancer. The data hasn't been published yet. It gives us additional confidence in thinking about ways that we can continue to evolve the treatment paradigm. For example, displacing platinum-based chemotherapy in this disease segment of patients. Thinking about different combinations as well, too, moving forward, across both ovarian and endometrial cancer as well, too. That's really where our focus is moving forward.
Excellent. Thanks. Thanks, Seamus, for the question. We only have a few minutes left. Let's try to go through three final questions here, first with Graham Parry from Citi. Graham?
Great, thanks. Just wondering what your thoughts are on the IMDELLTRA Imfinzi combination in maintenance in small cell lung cancer, and whether that changes the recruitment of patients that you think you need for your first-line small cell lung cancer or even second-line studies for ris-rez. Secondly, just your thoughts on what drove the lower grade three ILD rates in your trials versus the ARTEMIS-008 China studies. Just differences in how patients are managed with ILD. Thank you.
Graham, thank you very much for the question. I'll start off first with the data from Amgen and IMDELLTRA in the first-line maintenance setting. Again, great to see such an improvement in median overall survival, and PFS as well. I think, if you ask me, is this, of course, going to change how physicians practice and especially in this maintenance setting? Absolutely, it will in the short term, no doubt about it. But I think probably at the same point in time, as we're thinking about first-line development, there is an element that's missing here, that may not necessarily have been accounted for, which is actually the start of treatment, the initiation of treatment, and the challenges that currently exist with chemotherapy or, which can typically consists of a platinum-based doublet of cisplatin as well.
Not only is it the tolerability burden, and the challenges that it actually provides for patients, but also the fact that, oftentimes, in this disease setting, as you're well aware, patients respond quite a bit to it. But the responses are, of course, very short-lived. So, there's a lot of limitations that exist in that regard. What are we trying to do? Well, at least for us with ris-rez, one of the things that certainly we're going to be aiming to do is trying to help evolve the treatment paradigm at induction. Induction and then maintenance, not just maintenance. I think probably, for us, I mentioned earlier that we're going to be looking at combinations, whether it be with existing standard of care, PD-L1, bispecifics, given the emerging data, and access to those through Summit or certainly based on the IMDELLTRA data.
Thinking also about immune cell combinations as well. With regards to recruitment into the studies, I think this is something that we've certainly already taken into account, including on our second-line phase III study, which is currently recruiting. We are certainly recruiting or looking to recruit a proportion of patients that would be tarlatamab exposed. It's something that we've already addressed, had engagements with health authorities with, and factored into the program. With regards to ILD, I think there's a number of different variables that we have to take into account. Parts of it, of course, is as you look at specifically the ARTEMIS-008 study, you might wonder, why was the incidence higher than expected relative to the experience that we had on our global development program?
Well, bear in mind that about maybe 37% of patients recruited into the ARTEMIS-008 study on the ris-rez arm had actually received prior radiotherapy, which is quite, I would say, prognostic in terms of the incidence of ILD. The second, I would say, is also the fact that about maybe 13% of patients on the ris-rez arm actually were patients that had limited stage disease and then progressed. Typically, these patients are treated with definitive or chemoradiotherapy that's administered with curative intent. Again, the radiotherapy theme there is important to highlight. The other element in this, of course, is the follow-up. So there are about maybe 12 months of follow-up from the study, so certainly relative to other agents that have published phase III results that had shorter follow-up on the trial as well. That's certainly another variable to take into account.
Finally, of course, it also really depends on data collection and which types of preferred terms for the incidence of ILD were used. Are you using broader, as we call it, search criteria in the database, to look for events? The fact that these events actually that Hansoh reported were never adjudicated, and I think that's something that's important to highlight as well. So, for us, in terms of our phase III development program, we've certainly done a lot of education with the investigators, awareness about ILD, monitoring patients as well. I think you probably see that borne out in the analysis that we presented at ESMO as well.
Even in the small cohort of patients that were treated either with a 6 mg per kg or 8 mg per kg dose in second-line small cell lung cancer, the incidence of all grade ILD was actually 2%. Again, this is a smaller subset of patients, about 50 patients, but it gives you a bit more of the direct small cell lung cancer read-through as well.
Excellent. We're just there on time, but let's try to squeeze in if we can, two real quick questions. First, we'll go with Steve Scala from TD Cowen. Steve?
Thank you so much. I actually have two bigger picture oncology questions, which are hopefully appropriate for this call and may be best directed for Nina or Ahmed. There's a general investor perception that China-only data is not a good predictor of global results. But GSK's and other companies' data at ESMO seem to be pushing back on that perception. So I'm just wondering, how does GSK view this? Should China-only data be viewed as always reliable or only in certain circumstances, and if so, what circumstances? Secondly, given GSK's extensive experience in vaccines, does the company have any preclinical efforts directed at therapeutic neoantigen therapies? Thank you.
Yeah. Hi, Steve. On China data, I think everybody will go through the journey on a data-specific level. The reason why we actually did the global cohort is exactly to understand to which level we can be confident that the China data do represent this specific China data.
To represent something that we can count on. I think Hesham is probably in a better place to respond to this. But everything that we see in second-line small cell lung cancer, if that's the specific ask, gives us high confidence that genuinely phase III study from China are actually solid data that are reproducible. Regulatory question is something completely different, and I think I'd probably take it. I'm sure Hesham can make a comment, but in terms of trust or confidence in the data, I think this is exactly why we did the Western cohort to confirm that. But Hesham, I don't know if you want to add.
Yeah. No, thank you, Nina. Just to build on what you were highlighting, there's typically just a few variables that we take into account when we're trying to think about or consider extrapolation of data. I think first, you have to, of course, look at the baseline characteristics of the patient population and how it compares typically what you observe in a Western patient population. I think that's one. I think at least when we looked at the ARTEMIS-008 data, relatively at a high level, quite consistent with what you'd observe in a global population. Second, I think you have to also look at, for example, the standard of care, the control arm that's being used on the trial as well too, and how it behaves as well too. I think probably the data point here with ARTEMIS-008, the control arm is a topotecan control arm.
The most recent data set that's actually been published from a global phase III study, is from a study called the LAGOON trial, with lurbinectedin. The study actually was a negative study in second-line small cell lung cancer, but the control arm was actually a physician's choice of topotecan or irinotecan. It actually delivered a median overall survival of about maybe 10.7 months, relative to the topotecan control arm in ARTEMIS-008, which delivered a median of 10.3 months. So you see that consistency again. Then the reference I made during the presentation really around what are we seeing in the internal validity of the results from the trial, this consistency in treatment effect, the magnitude of effect as well too. So it's on the primary endpoint, it's on the key secondary endpoints, and it carries through, and the magnitude of effect is strong.
I think all of those are good indicators and variables that help you think about the strength of translatability. On the question of how regulators view the data from China. Well, I think probably the one thing that we've seen is, there is this regulatory, I would say, dissonance that exists in the context of how different regions or countries may view them differently. I think we've seen certainly, data from China serve registrational purposes, for example, in regions like Europe, in regions like Japan as well too. Not necessarily, of course, in the U.S., but I think it's certainly important to have that perspective.
The question about
There was a question, yeah, the second part of the question. Yes. Vaccines, yeah. I think it is an important point to highlight, of course, just given the heritage that we have in this space and specifically, also the mRNA platform that currently exists. I would probably call to your attention to the fact that we actually have an existing, I would say, academic partnership with Oxford. Specifically, it is a translationally oriented immuno prevention program, evaluating and assessing neoantigens that are most likely to elicit an immune response relative to cancer vaccines as well, too. So, that is currently ongoing. It is a multi-year collaboration as well, too, and it is an area we continue to evaluate and assess on an ongoing basis relative to what is happening externally in the competitive landscape. Of course, over the weekend, we saw the data being presented from Moderna.
I just want to make sure that I acknowledge that as well, too. And more recently, we have also seen data that has been also not positive in other tumor types, including in colorectal cancer as well, too. I think there is still a question that remains around how much can be extrapolated from a highly immunogenic and immune-sensitive tumor like melanoma relative to other tumor types that may either be less immune sensitive, colder, or having lower tumor mutational burden as well, too. But we are assessing and evaluating different considerations in that regard.
Great. Just one quick final question then if Kerry Holford can come off mute, from Berenberg. Kerry, we will give you the last question here as we try to wrap up.
Thank you very much for squeezing me in. Clearly, a lot of focus more broadly on ADCs this year at ESMO, not least within your pipeline. So maybe to end on a sort of big picture question, if we ask you to take a 10 year view, how much of traditional chemo do you think could ultimately be replaced by ADCs? And GSK within that, what sort of share do you ultimately target within that ADC field? Thank you.
Hi, Kerry. Everybody's looking at me here in the room
Sorry
to try to come up with an answer. Look, I don't think we are approaching this from the angle of what is the total amount of chemotherapy that is going to be replaced. Yes, ADCs will replace a significant portion, but ADCs will not fully replace chemotherapy. It's going to be very indication-specific, sometimes even line-of-therapy-specific. Definitely very tumor-specific. I think, with the 10 years view that you quoted, we are probably looking at beyond Topo I, and then what are other drug conjugates that are going to end up on the ADCs with even multi-mode of actions in the same target. What we do see is obviously great benefit, but we also need to be aware ADCs are not completely benign therapies. There are toxicities that are coming with ADCs.
I think we need to look at what is going to be the duration of therapy, how much can be tolerated. I'm not sure I would look at it fully from how much of chemotherapy can it replace. It's very likely it will go into earlier lines of therapies because that's where the drug, where the shot on goal for potential cure is. I would say, at the moment, out of the totality of oncology, it's still a minority, unfortunately.
It's just going to be another platform that could complement or could add to different therapies or different tumor types, and we're still starting with this.
Excellent. Well, thank you. Thank you for sticking with us for an extra 10 minutes, and we'll conclude the call at this point.