Hey, today's event is the Q2 presentation of Ascelia Pharma. With me today, I have the CEO of Ascelia Pharma, Magnus Corfitzen. It seems like the sound is much better, Magnus. Sorry about the inconvenience. We start all over again from slide one. Please carry on, Magnus.
Yeah. Thank you, and welcome everyone, and apologies for the issues with the sound. Hopefully, that has been fixed now. I'm happy here to present Ascelia Pharma and our Q2 update. I'll go into more detail in terms of the Orviglance situation and the way forward. We remain committed to developing Orviglance and bringing it to the patients in need. We'll be making a number of forward-looking statements here in the presentation today. Just the big picture, Ascelia is committed to identifying, developing, and commercializing novel drugs for people with rare cancer conditions. We have two drugs in development. Our lead program is Orviglance, which is a novel first-in-class diagnostic drug to enhance the liver MRI procedure signal for patients with severe kidney disease.
It has Orphan Drug Designation from the FDA. It will be the only product addressing an $800 million market opportunity. We've completed patient enrollment in phase III. We are working on getting the outcome from the study. The other program we have is Oncoral, which is a tablet-based chemotherapy. It has completed phase I development, will be moving into phase II . We have a clinical collaboration with Taiho Oncology and beyond gastric cancer, which is the lead indication, there are other opportunities that are very attractive. Ascelia is based in Malmö, in Sweden, and we are traded on Nasdaq Stockholm since 2019. In the quarter, in Q2, we had a focus on the scientific and medical conferences. We had the presentation of the Orviglance hepatic impairment study, which is testing how Orviglance.
what happens with Orviglance in patients where the liver function is compromised. The results from that study was presented both at the ESGAR Conference, the European Society of Gastrointestinal and Abdominal Radiology, and at the EASL, which is the Study of the Liver Conference in, in Europe. At the ESGAR Conference, we hosted a Q&A session with two experts in liver imaging, and that Q&A session is available on our website. After the completion of the quarter, we announced that we do not yet have the results of the SPARKLE study. This is disappointing, but we need to do a re-evaluation of the images, where we have the appropriate level of quality in the, in the image reading, in order to be able to make a conclusion on the study.
We also sent out a clarification just to because there seemed to be some misunderstanding based on the question we received, that this is not an issue with the image quality. It's not a question issue with the study. It's the reading phase that happens after the study, where we have independent radiologists looking at the images. There were some issues in that evaluation, which means that the results cannot be used to make any conclusions, unfortunately. I'll talk more about this today. First, just giving a sort of high-level recap on Orviglance. It's a targeting a well-defined unmet medical need for people who need a liver MRI procedure who have severe renal disease.
The global addressable market is $800 million, we have completed eight clinical studies in phase I and phase II, with a consistent profile in terms of efficacy and safety. The ninth clinical study, the SPARKLE phase III pivotal study, we have completed patient enrollment. All the images are in the database. All the safety data is in the database. The safety data, the adverse event profile, was consistent with what we've seen in previous studies, which we consider a very good result. Unfortunately, as mentioned, we don't have the conclusion on the efficacy because of the issues in the reading process. What went wrong? When we announced in early August that we need a re-evaluation to get the results, we reported that this is due to a high level of intra-reader variability.
This is something that was, you could say, the conclusion was reached after we went through the analysis of we looked at the data that was presented. This is part of the predefined protocol, quality control step, if you will, that is implemented in the SPARKLE study based on the FDA Guidance for Industry, and that we had also discussed in the design of the study with the FDA. This means that there is a, you know, an unacceptable level of variability for two of the three readers. Therefore, we cannot make a conclusion on efficacy. What we also announced is that all our focuses in Ascelia will be on the re-evaluation of the images, the new reading procedure. We're postponing other activities, and we are initiating cost-saving measures.
We will, in the middle of September, we will present a plan for when do we expect to have the result from the study. We need to enter into new contracts with specialist CROs, who will drive this process. We need to identify new readers, give them the proper training and instructions in how to do the reading of the, of the, of the images based on the protocol, and they need to have time, and it is a time-consuming process, to go through all the images.
Magnus?
Yeah.
If I can just interrupt you 2 seconds. What we see here on our left side on the slide, is that a reader, a radiologist looking at the different images, or is this just a showcase of how it looks?
This is a showcase of how it looks. Typically, a radiologist would have. I mean, I think this is probably a display option, but most software you would have, like, on one screen, you would have one image of the liver, and then you can basically scroll up and down to sort of simulate a 3D model. You move through and see the blood vessels and bile ducts and lesions, tumors, et cetera. This is more for illustrative purposes, but it illustrates also that when the radiologists are looking at this, they are sitting with a set of screens and looking at all the images.
Thanks a lot, Magnus. Please carry on.
In mid-September, we will present a timeline and also our financial projection. Our ambition is that the current capital that we have in the company should be sufficient. We don't know that yet, but that is the ambition, and we'll come back with that in mid-September. When we say high inter-reader variability, I mean, that's not something you talk about every day. Really, the way to measure that is to have a reader, a radiologist, the same radiologist, looking at the same set of images at two different time points. If they have a high level of consistency, you would expect them to have the same score. There may be, you know, slight variations, but when you then take the, you know, this is a group of patients whose images are presented twice.
You then look at the scores there, you would expect the average to be pretty similar. What we saw for two of the three radiologists in the evaluation we have just completed, we saw high variability, high difference between, you would say, first time they saw the image and second time they saw the image. For instance, first time they may see a high score and second time a low score, or the other way around, and that is not good for, you would say, getting consistency. And especially not if there's like some trends that is not having the same average. Nobody really knows what is the true, what is the accurate measure? Is it when they score high, or is it when they score low?
When there's a high level of inter-reader variability, it means that the scoring is unreliable. You don't know which one is the, you would say, the true one. When you have unreliable scoring, you cannot really evaluate the effect because you don't know what it is that you're measuring. Just to reemphasize, we have all the data, we have all the images, we have the clinical data that is needed. Everything is in the database, in the blinded database, it's the images in the database that have been evaluated by the radiologist, where we had, you would say, issues in that part. One of the radiologists evaluated the images and with an acceptable level of variability. You know, it's not an issue with the image quality.
That's also the feedback that we have received from, from the sites and the investigators that, that were involved in the study. This is extremely unfortunate, but it's not unknown. If you look at the scientific literature, this is a problem that happens from time to time. We are also aware that there are other drugs that have had similar issues in the past that have moved on to regulatory approval based on the same data. You know, we are obviously learning as much as we can. We don't have any names. We know this under confidentiality, because companies obviously would- if they are not obliged to share this, reading issues, then obviously they will not do so. As a publicly traded company, and we need to share this for transparency reasons.
Just to give you an example, I mean, this is, this is the evaluation phase. It's not the, the images. If the endpoint had been measured by, let's say, a blood sample, and, and we measured the blood samples, and that would be the primary endpoint. If we found out that the laboratory doing the blood sampling analysis, that their lab equipment had been malfunctioning, so it gave, you would say, inconsistent or to some extent random values from the reading of the blood sampling analysis, the obvious choice would be, okay, let's go to either another lab or the same lab, but with a well-calibrated lab equipment that could do the analysis with the required amount of precision. That would not be a surprise, and, and that would be a much faster process than doing the reading process.
The reading of the images is, unfortunately, a time-consuming procedure. They, they, you know, the approach is the same. When you have issues in the reading process that, you would say, disqualifies any outcome, you could say you cannot conclude anything, then you need to redo with the appropriate amount of quality to make sure that you have a result you can conclude on. That is what we are heading into. That's what we will provide in mid-September, the timeline and also the financial runway projection. We need to bring it to the patients. There is a high level of unmet medical need for liver imaging in patients with renal disease, and there is an addressable market opportunity of $800 million.
Our confidence, just to underline, our confidence in, in a positive outcome of efficacy in, in the SPARKLE study is unchanged. We've not seen anything that would lead us to believe otherwise, but we have had a quality in, in the, in the reading process. Now I'll switch gears and go to Oncoral, our tablet therapy for treatment of solid tumors. Oncoral is based on irinotecan, which is an approved and very potent molecule for treating certain solid tumors. Today, irinotecan is given as an infusion directly into the bloodstream every third week. The patient will go to a hospital, get a very large dose of irinotecan, which results in treatment effect of the tumor, but also in very substantial side effects.
What Oncoral is all about is using our patented, we say concept and approach, by taking irinotecan instead of an infusion, giving it as a tablet, and the patient will be able to take irinotecan every day. That means that they will get a smaller dose, but every day, and over time, they will have a comparable, we say, exposure of the active metabolite, and that will, we believe, drive improved efficacy and hopefully also a better safety profile. That is the purpose of Oncoral, and that's how it can help patients. Go briefly in here. What we're looking at, moving into further development is combining Lonsurf with Oncoral. Lonsurf is an approved tablet therapy for treatment of colorectal cancer and gastric cancer. Those indications are also where there's been seen effect for irinotecan.
That all-tablet therapy combination would be very interesting. As you can see here to the left, you see that Lonsurf, together with irinotecan, seem to have a synergistic effect in this animal model of gastric cancer, and that's what we want to explore in clinical development. We have made a clinical collaboration with Taiho Oncology, the developer and commercializing company of Lonsurf, where we will test the combination of Oncoral plus Lonsurf versus Lonsurf alone. If that has any improvements in terms of patient outcomes, we think that's a very attractive option for patients going forward with this all-tablet therapy. The potential in for Oncoral is even beyond gastric cancer.
There are a number of other cancer types where there's a high level of unmet medical need, where the treatment options are still not good enough, and where Oncoral, for some of these patients, will be a very attractive option. The potential of Oncoral is very significant, and we will initiate clinical development on Oncoral when we have the financial capacity to do so. Ending off the, you would say, the presentation with a financial update. We have been having using consistent cost per quarter, which is around SEK 30 million-SEK 40 million per quarter, depending a bit on the activity. Recently, they've been higher because of the high activity level in SPARKLE. As we are completing the SPARKLE study, we expect costs to be lower.
There will be some cost for the reevaluation, but we are initiating and have initiated a cost-saving measures. We focus almost exclusively on the reevaluation of the data. In mid-September, we will know more about the cost and the timelines, and then we will share it with you. Our ambition is that the current capital we have available should be sufficient, but we do not know until we are in mid-September. To finalize, you would say the outlook right now is that we got the last patients in SPARKLE in early March. Mid-September, we will share the plan, when it's finalized, for the timeline for the reevaluation of the images and the financial runway.
beyond Orviglance, which we remain very committed to, we believe in the value it has for the patients, and we are excited about the interest in the scientific and medical community about Orviglance. In addition, we have Oncoral, which is also a very attractive opportunity. With that, I'd like to end the presentation of, of the Q2, and be happy to take any questions.
Thanks a lot, Magnus, and really good and thoroughly run-through of the Q2 and your pipeline products. Interesting to get some news on Oncoral, even though you have, as I understand, is going to postpone it until you have more certainty about the Orviglance. There's a couple of questions out there. If we look into the actually problem, the readout, the radiologist's interpretation of the images, would you use... Because I know you put a lot of effort in educating the radiologist. You put a lot of effort in finding the best radiologist for this job. Would you use the same pool of radiologists, or will you change the radiologists?
We want to do sort of a new independent evaluation, which means that it will be new radiologists that are not biased. We want to have radiologists coming in, having no prior knowledge, and that's also consistent with, you could say, the, the, the guidance to the industry from the regulatory authorities. We will select new radiologists. We will train them and instruct them in the SPARKLE protocol, the reading manual. What does this score mean? What does this do? I would say, I think we, you know, we, we were working with, with, you would say, very high quality, very well reputed CROs in, in that process. I think definitely we are looking into, of course, every single step that could potentially have an impact in terms of...
I mean, there are so many parameters in terms of, you would say, having highly qualified radiologists. We did have that in the first read. It's also well known that, that, I mean, some radiologists are more, you would say, have higher variability. Are there way, ways to select them still, so in an independent way, so, so there's no bias, but just to make sure that we get the right ones. We will look into every, we are looking into, I can assure you, every tiny detail and step of the way. In the training sessions, anything we can do to improve, you know, do more thorough, extensive training, just to be on the safe side, retraining through the reading process, et cetera.
Everything we can do to minimize the risk, but it will be a completely independent review process compared to, to the first one. Obviously, the protocol is the same. What we're measuring is the same. We're not changing the, we say, the endpoints or, or anything like that, but we need to take every measure we can to minimize variability.
Thanks a lot, Magnus. So you're doing the analysis of what went wrong in the readout. You're probably doing that already now, as you mentioned. How fast can you actually educate these radiologists? Could you elaborate a little on that? Because, well, you have some impressions from what you did before summer, when you did this job. Is it two, three months, or, or what are we talking about?
Yeah, no, the educational process is the training process is typically much, much shorter. I mean, keep in mind that we are taking very experienced radiologists that, that have seen on liver images many, many times. What is an issue and what that, that we discovered, and also issue in, in other studies, but in, and, and widely accepted also by the regulatory authorities, is that there is a difference between looking at an image in a clinical trial and looking at an image in clinical practice. In a clinical trial, they need to evaluate the lesion border delineation and the lesion to liver contrast. They need to provide a score based on the criteria and the definitions of the score. When, as you know, the, the, the definitions, they, they should be clear.
They should provide a clear understanding of, what is the difference between score, you know, one and two. In clinical practice, a radiologist will be in a different situation. They are looking at, do we have a lesion? What kind of lesion is it? Then preparing a discussion, for instance, with the tumor board, where they have the oncologist and the surgeon, and they're discussing: Is this a patient who is a candidate for resection? What they're measuring is a little bit different from their routine clinical practice in terms of providing a score on the border delineation. That is well understood by everyone. That's why we have a training session where they need to understand the manual.
You could say, we need to make sure that they are adhering as much as, you know, as is absolutely possible to, to be stringent in terms of the criteria and so forth. We don't know, exactly... there's not sort of one, you could say, factor that when we look at things that, that say, "Oh, this is, this is kind of, this didn't work. We, we screwed up in the process." I think it's more about what we're looking at together with, you know, our Key Opinion Leaders, our imaging experts externally that we work with, is how can we improve, strengthen the process so that we minimize risk that something goes wrong, right? With the two readers that have high level of intra-reader variability, maybe the root cause is different from one between the other.
We don't know. But what we can see is all the time to sort of, you know, make sure that we reduce variability. That is the issue. It's, it's not, we want them to have high level of consistency. That is what we need to, to get the result. And we remain very confident and based on everything we've seen, that Orviglance will be efficacious in liver imaging.
With that said, with this high consistency, wouldn't it be possible to do machine learning or AI, as a popular word, and probably get a more consistent result, or is?
I think if there was, if it was in the FDA Guidance for Industry, and if there was, you could say, an approved and available software for this, that would be dream case scenario, because we would cut the reading process tremendously in time, probably also cost. It's not available.
Okay.
The guidance from the FDA, what they want to see is two out of three radiologists.
Okay.
When you're thinking about two out of three radiologists, that's also, you could say, could be interpreted as acknowledgment from the regulatory authorities that radiologists are like the, the rest of us. We're all human beings, people are subjective, and, and people have different, you would say, approaches. That is why you don't need, at least my take, you don't need three out of three, because there is variability.
Good. Thanks a lot, and interesting. If you could just elaborate a little on the cost side. I know you're coming back to this issue mid-September. For the audience here, I can mention as well, that Magnus will host our life science seminar next week, the 31st of August. Please, listening to Magnus here. Of course, you had a quite high burn rate, first half of 2023, due to the phase III study. You also mentioned that you're doing some cost initiatives. You're putting Oncoral on hold for a period. Are we looking into a run rate around SEK 20 million-SEK25 million, or am I too specific quarter wise?
Is it that in that ballpark, or can you shed some light on, on what we're looking into?
Yeah. No, I think. We are definitely looking into a significantly lower cost spent a quarter. There are some, as you mentioned, the first half of 2023 was high activity level in Sparkle. Also, you could say, completion of the, you would say, collecting all the data, doing the reading process. Going forward, we will focus more on the, only the, the reading process, because everything else has been done, so to speak. That will be. We're looking into every single cost that we have in the company, see how is this essential to make sure that we get to the reevaluation. I cannot really say anything about, you could say, the cost going forward. We will come back in mid-September.
I think that is the appropriate thing to do. What I can say is that it will definitely be a lower cost level that we will see going forward from what we have seen earlier this year.
Thanks a lot, Magnus. Is there any more questions from the audience? Please post them, and I will forward them to Magnus. Doesn't seem like there's any more questions. With that said, Magnus, thanks a lot for, for your presentation.
Thank you.
Thanks a lot for shedding some light on, on, what actually happened. There's a question coming in here. Well, there's a question here about the radiologist again. Are these radiologists the same that did the phase II study with the 20 patients?
No. No, they're not. Just to comment on the radiology selection process. We define, as a company, some criteria, and then the imaging CRO specialist company that is hired to do the reading evaluation, they select the readers, the radiologists. We have this arm's length for good reasons. We cannot use people that you would say, for instance, KOLs, that know the drug very well, and we're not using KOLs or investigators who have been, you know, have experience with Orviglance in the past. We want independent radiologists who have never seen the images to minimize any bias. You would say, I think that that's a standard way of doing.
We don't have any direct contact with, with the readers, and in the reading process, we don't even know who they are.
Okay. Yep. Thanks a lot.
Sure.
So well, that was actually the final question. So with that said, thanks a lot again, Magnus, for the thorough presentation. Thanks a lot for- thanks a lot to the audience. First of all, I apologize the technical issues in the beginning, and thanks a lot for all the good questions. I hope to see you all next week on our life science seminar, the 31st of August. I hope you all will have a really good day, and by that said, I will close the event. Thanks a lot.