Immunovia AB (publ) (IMMNOV) Earnings Call Transcript & Summary

Good afternoon, everyone. My name is Patrik Dahlen. I am the CEO of Immunovia, and I will be the moderator of this webinar. A very warm welcome to all of you to this Immunovia webinar. With me on this webinar call are Dr. Thomas King, who is our Medical Director; Dr. Linda Mellby, who is our Vice President of R&D; and Dr. Laura Chirica, who is our Chief Commercial Officer. This is the third webinar in a series of webinars that we have run since the beginning of this summer. Today's webinar is called IMMray PanCan-d Verification Study and Clinical Use. The agenda for the webinar is positioning of IMMray PanCan-d in clinical practice, this will be presented by Dr. King; the verification study versus clinical practice, this will be presented by Linda Mellby; and the voice of the pancreatic cancer experts about IMMray PanCan-d clinical utility, and this will be presented by Laura, who is our Chief Commercial Officer. Please remember that at the end of the webinar, you will have a chance to ask questions. Below the video screen is an envelope symbol. And clicking that, you can ask questions. With this short introduction, I would now like to hand it over to Dr. Thomas King. Tom?

Thank you, Patrik. I'd like to begin today by saying a few remarks about cancer detection tests in general and how those interact with negative and positive predictive value before I focus in on pancreatic cancer and PanCan-d's approach to it. Clearly, a clinically useful test must be able to separate healthy from diseased individuals. But in practice, that separation is never complete, as you can see at the bottom of the slide with the red and blue balls indicating diseased and healthy. For example, in cancer screening, pap smears traditionally have a sensitivity of only 55% but a specificity of 97%, so quite less than optimal in terms of sensitivity. Although the way they've been used in clinical practice, they've saved very many lives over the last 70 years that this test has been in place. Part of that reason is the sensitivity is made up for by serial testing. So every year or every other year, one minimizes the chance of missing a positive case but still finding lesions early enough to treat them, and in this case, usually to prevent cervical cancer. Prostate-specific antigen is also a test that's been used in screening for prostate cancer, probably much less successfully, only 21% sensitivity and 91% specificity. So as you can imagine, there are many false positives and false negatives. But certainly, some clinicians still view this as a useful test to do in older men to detect prostate cancer. How well a test separates healthy and diseased individuals can be expressed in a variety of different ways, probably the most useful way is in terms of receiver operating characteristics area under the curve analysis. With a qualitative test, the negative and positive predictive value is determined by the threshold that one chooses to call something a positive or a negative. Moving that either to the right or to the left will change the positive or the negative predictive value. But that's not all that determines the negative and positive predictive value, that those 2 are not completely an intrinsic property of the test itself but are quite dependent of the prevalence of disease or the number of percentage of diseased individuals in the population one is testing. Oftentimes, one doesn't know precisely what that percentage is, depending on how a patient presents. But the negative predictive value represents the proportion of individuals with a negative test result with a given test who are correctly diagnosed as negative. And the positive predictive value conversely represents the proportion of individuals with a positive test result who are correctly diagnosed. As you can see from the examples of these 2 cutoffs at the bottom with a disease prevalence of 25%, one can move the curve in 1 direction or the other, always trading sensitivity for specificity or vice versa specificity for sensitivity. Now I wanted to describe the 3 high risk groups that we've talked about many times that we're targeting for PanCan-d. The familial and hereditary group is defined either by 2 close family members with the disease or by specific genetic lesions. This is quite a heterogeneous group. As is shown at the bottom, the relative risk in these individuals ranges from 2 to greater than 100x the population average for pancreatic cancer. So while we talk about it as one group, it's still quite heterogeneous. Our data from our partners in pancreatic cancer surveillance programs indicate that the current prevalence of pancreatic cancer in those cohorts is about 1%. So that prevalence is important to keep in mind as we talk about how PanCan-d will work in different situations. The early symptoms cohort is defined by our colleagues really at University College London and more and more at other institutions throughout the world. This consists of a variety of different types of symptoms, most of which are rather nonspecific. So the prevalence of disease in this population is going to depend on which of these findings or these symptoms is present and whether more than 1 of them, 2 or 3 or 4 or some of them may be more suggestive than others. So again, a heterogeneous group. But based on our colleagues at University College London, we feel that this prevalence is about 3% as best estimate. The final risk group is nuanced diabetes later in life after age 50. And this is a well-defined prevalence of 0.85 from many studies. That prevalence last for the first 2 or 3 years after diagnosis, and it represents a very large population as type 2 diabetes is epidemic throughout the world and is increasing at a frightening rate. We've approached these 3 risk groups using 3 prospective studies. We've talked about before, PanFAM-1, PanSYM-1 and PanDIA-1, covering the main risk groups that we are targeting for the PanCan-d test. We anticipate beginning analysis of these cohorts from which we've collected a large number of samples in 2021. I next want to switch gears a little bit and talk about how prevalence in different disease groups is going to be important in terms of how PanCan-d is used. And I want to make the point that PanCan-d can be used in a variety of different ways, in different clinical situations to make decisions about escalating the diagnostic pathway or deescalating it in some situations, and in choosing the most appropriate types of additional testing to get to a definitive diagnosis. So I first wanted to begin with vague symptoms, as I've talked about, that's going to be a heterogeneous category because it depends on which symptoms are present and how many of them. So these individuals may have a prevalence of between 3 to perhaps more -- a higher percentage. And I've tried to indicate that by the prevalence triangle that's shown there. As we move from the left to the right of the slide, the disease prevalence is certainly going to increase. The exact numbers are always going to be hard to come by or hard to know precisely because patients may have a constellation of findings. So someone with vague symptoms may actually have had a CT scan previously, so they have some imaging abnormality as well. At least in the U.S., anybody who visits an emergency department with any kind of GI complaint is going to get a CAT scan. There's pretty much no way around it. So that's a relatively large cohort that would have a number of these findings. I wanted, for simplicity, to describe these individually so that we can get an idea of the diagnostic MLU that's operative in the U.S. today. So in someone with vague symptoms, a negative PanCan-d test may lead them to be discharged from clinical follow-up because they'll have enough confidence with a negative predictive value that this person doesn't have pancreatic cancer and has something else, so they can be worked up in that way. Conversely, if they had a positive test, that might lead to additional follow-up. Most likely, it would lead directly to imaging studies to further evaluate the patient. And if those are positive or suggestive, it might escalate from there to endoscopic ultrasound or EUS, where an endoscope is placed in the duodenum and an ultrasound probe is applied directly to the pancreas to image it in fine detail. The most invasive type of approach is a biopsy or fine needle aspiration, and that can be done either through the endoscope or directly through the skin through the back of the patient to directly sample the pancreas. As you can imagine, these different techniques become progressively more invasive, and ultimately, have dangers associated with them. So it's important to think about how one is going to apply these. Next, I wanted to consider high-risk individuals. So these would be individuals with familial histories and also with specific genetic lesions. Again, remembering that this is quite a heterogeneous group so some of these individuals are at very high risk, some just at high risk. So in this case, a PanCan-d test would either lead to continued clinical follow-up, perhaps additional imaging. Or if positive, probably to endoscopic ultrasound, which as I'll show you in a minute, is one of the most or is the most sensitive and specific imaging modality that one can do, but this is still an invasive test. With individuals with an imaging abnormality, a negative PanCan-d test typically would lead to additional follow-up and make sure there are no imaging changes over time. Positive, though, would escalate the clinical evaluation with endoscopic ultrasound and oftentimes, if there's a lesion present, an FNA or a biopsy. That escalation is really important because I really want to make the point that many individuals who ultimately die from pancreatic cancer spend more than a year trying to get a specific diagnosis of the disease. And that time is wasted. They could potentially receive curative surgery if they were diagnosed earlier. And clearly, their lives could be extended if they receive treatment earlier on in this process. Surveillance studies have shown that individuals diagnosed with pancreatic cancer in a surveillance program have a substantially longer survival than individuals who are diagnosed with pancreatic cancer not under surveillance. So that time interval is crucial, being able to find the disease early allows for their options in terms of potentially cure, but certainly in terms of life extension. In individuals with worrisome symptoms, those might have a constellation of many different findings, including jaundice and back pain, which may be suggestive of advanced disease. In those individuals, PanCan-d could serve a more confirmatory role and allow the choice of escalation of diagnostic entities and make an optimal choice in that way. What I really wanted to do with these series of slides is make the point that this is a complex diagnostic MLU. And the way the test can be used is different in different situations with different patients. Also, the sensitivity and specificity of the assay are going to play into the prevalence in these different situations. So how the test is going to be used in these situations is going to be different and is going to be dependent largely on the clinician's experience with the test and how they manage their patients. I wanted to show you, this is a diagram, I think you guys have seen before that we've shown where we've targeted the asymptomatic group, which would be the familial genetic disease; and the symptomatic group, which are individuals with symptoms, diabetes or imaging abnormalities. And we've identified a number of different points at which PanCan-d could interact with their diagnostic process. Here, we've labeled them a through epsilon. So many different points in the diagnostic process where they could interact. And that really goes back to the last slide, showing you the complexity of the diagnostic mLu. Now I wanted to also review for you the current diagnostic landscape without PanCan-d, without a specific test for pancreatic cancer. So imaging, I wanted to speak of first, we've already spoken about CT or computed tomography scanning, magnetic resonance imaging scanning and endoscopic ultrasound, where an endoscope is placed in the duodenum and the pancreas is imaged through the wall of the duodenum. These tests are really suboptimal. Certainly, CT and MRI, they have moderate sensitivity but very low specificity. So 43% specificity for CT scan and 63% specificity for MRI. So many false positive diagnoses are going to result from this type of imaging. Endoscopic ultrasound, which is an invasive process, has better sensitivity and better specificity, but still is not 100% or really even close to it in terms of specificity. So in terms of these imaging techniques, they're useful, they're what's available now, but they are far from perfect. And being able to augment them with additional types of testing that provides independent information would be very useful to clinicians and to patients. This is targeting an even more invasive process, either fine needle aspiration, where a skinny needle is placed into the pancreas either through an endoscope or through the skin of the back and targeting a lesion. This long needle is then put under negative pressure with the syringe, and it's moved back and forth in a cutting manner to obtain a cytologic sample, which can then be smeared on the slide and examined by an expert cytopathologist. As you can imagine, this is definitely an invasive procedure. Anytime one touches the pancreas, one risks the chance of causing acute pancreatitis, which is a quite severe illness requiring hospitalization and will be associated with some level of mortality. So this is serious business to do this. Even so, the sensitivity and specificity of this test is, for the most part, less than 100% for ultrasound guided FNA, 85% sensitivity and 90% specificity. And in the pie chart that I provided here, you can see that there are a substantial number of individuals who have either a false positive or a false negative test, highlighted in light blue here. Finally, I wanted to speak about the only blood-based marker that has some utility for pancreatic cancer, that is CA19-9. This shows an ROC area under the curve analysis for CA19-9 with a cut-off of 305. And what you can see is that much of the area under the curve is not cut off by this test. In this example, the diagonal line is 0 discrimination, complete discrimination, the line would follow the 2 axes. And I wanted to give this as an introduction to Dr. Mellby remarks about our verification study, showing how IMMray PanCan-d in individuals with PDAC performed as opposed to asymptomatic controls. And here, you can see that much more of the area under the curve is cut off by the line. With that, I'd like to turn things over to Linda.

Thank you. Tom. So I will now present for you the PanCan-D verification study and put the results into perspective of the diagnostic landscape just described by Tom. So the verification study was a multicenter case control study, covering 590 patients in total. The same samples were prospectively collected from 9 different EU and U.S.-based sites. In total, we had 195 PDAC stages , all stages and whereof 81 was PDAC, early stage, stage I and II. We had 212 healthy asymptomatic controls and 112 symptomatic controls. And the healthy controls were asymptomatic controls. And in that respect, they present the hereditary and the familial risk group in this study. The symptomatic controls were prospectively collected from gastrointestinal units and in secondary care centers. And these controls have symptoms of GISTIC -- of PDAC, but not being PDAC. Instead, they are diagnosed with various benign pancreatic and biliary diseases. This control group also include diabetes as being a symptom for pancreatic cancer. So I would like to stress that the diabetes we have within this control group is not from the non-risk group just described by Tom, but it is diabetes coming in having a symptom for pancreatic cancer. So the aim with the verification study was to clinically verify the PanCan-d for both the asymptomatic and the symptomatic control groups. And we did so by using a locked product configuration, we had a locked PanCan-d signature and the model cut-off that was locked and set in the CTMS study just finalized 1 year ago. As this is a verification study, you also have a verified software that automatically generates quality control measurements and decision values for your test. You have locked production processes, quality control methods, so you run this test according to a product setting and you really try to mimic as much as possible a real clinical situation. So I will now start to share the results for the symptomatic group, and you can interpret these results as a diagnostic aid for pancreatic cancer asymptomatic group. And results show that we could separate pancreatic cancer early stages, stage 1 and 2, versus symptomatic controls with a ROC AUC value of 0.84. This means with the locked cut-off we have for this test that we have a specificity of 81% and a sensitivity of 78%. And using a clinical relevance prevalence, as just described by Tom, this 3% prevalence is for the UCL Rapid Diagnostic Center, London, U.K. for this early symptomatic controls collected both in primary and secondary cares. And using this prevalence, together with the specificity and sensitivity values, we calculate a negative predictive value, NPV, and a positive predictive value, PPV. The negative predictive value was 99.2% and the positive predictive value was 11.3%. So with the specificity of 81%, you will have many fast positives. And with this current performance of the PanCan-d in this control group, it would put strength on the health organization. Although we believe that the negative prediction value that leads actually to missing out patients that have the disease is very important. But the positive news is that we have seen that there is potential for improvements for the test performance in this control group. And these improvements we are now working on and we will inform you and report up on these results once we have tested them. And I would also like to state that this is fully normal during a development phase. You have to remember that we don't have a validated test yet. So we are still in the development phase, and incidental improvements are absolutely normal and something you do with the required testing afterwards. I would also like to stress that this will not affect the sales start or our locked signature. I would now like to continue with the result of the asymptomatic group, and you can put these results in a perspective of -- as a diagnostic aid for surveillance of asymptomatic individuals. And first, I would like to say that the IMMray PanCan-d performance showed excellent classification performance in this group in separating early stage pancreatic cancer, stage 1 and 2, versus asymptomatic controls. We had a ROC AUC value of 0.94, and that gave us a very high specificity of 99%. And the high specificity is very important in a low prevalence disease. We also had a sensitivity of 78% for early-stage pancreatic cancer. So applying clinical relevant prevalences, as just described by Tom, we could calculate the NPV and PPV. And for the 1% prevalence, we had a 99.8% NPV and 44.1 PPV. And for the 3% prevalence, we had a 99.3% NPV and 70.7% PPV. And with this example, we can show that for the same test performance, we changed the PPV from 70% to 44%, only by changing the prevalence from 3% to 1%. So the NPVs and PPV values are very dependent on prevalence, and it's by putting the clinical performance of the test into a clinical setting that is relevant. And therefore, you need to select the prevalence from that are clinical relevant or the positioning of your test. So again, I would like to stress that this is really good results, and I will now like to present these in the context of the diagnostic landscape just presented by Tom. So taking the results from the verification study for the asymptomatic control group and comparing with the noninvasive imaging technologies Tom just talked about and also the more invasive approaches. We are assuming a prevalence of 1% -- PDAC prevalence of 1% as is in the familial hereditary risk group, apply that to 10,000 -- population of 10,000. This will mean that you have 100 that are -- that have the disease, and you have 9,900 that are well. And using the PanCan-d results, we -- that then means that we have 78 true positives and 22 false negatives. And these values are very similar to the other technologies, except for the endoscopic ultrasound that has really, really high sensitive -- sensitivity. And looking at the other end, we have excellent results when it comes to true negatives. We only have 99 false positives compared to the other technologies having 900 to several thousands of false positives. And these numbers really show the significant clinical utility of PanCan-d. And I would also like to stress that we are not aiming to replace imaging. This PanCan-d test will provide complementary information that could be used for the clinician in their decision-making process. So if we use these numbers to calculate negative predictive values and positive predictive values, we will again see that the NPVs are very high for all the technology since we have a very low prevalence disease. Whereas the PPV is more dependent on the test performance. And the PanCan-d show a much higher PPV than the other technologies, which again show the clinical utility of this test. So with this, I would like to send the word over to Laura to continue this presentation. Thank you.

Hello, and thank you for this opportunity. I am actually very grateful to be able to bring you the voice of the pancreatic cancer experts today about IMMray PanCan-d clinical utility. As soon as the verification study results were finalized, we turned to a number of our experts that we have been working with in pancreatic cancer and pancreatic diseases during many years, and they are representing the different risk groups and also the different clinical diagnostic needs that we have seen presented by Tom earlier. And today, I'm going to focus on just a number of highlights from these interviews. And I just want to point out that we are going to place on our homepage, exactly where you find the webinar link, you are going to find also a complete interview with Professor Steve Pereira from UCL from London, who is the founder and who is leading investigator for the early symptoms risk group. And we have chosen for today to just go through a number of the answers that are coming from representatives from the hereditary familial screening programs in the U.S. and in Europe. One U.S. expert and also Professor Alfredo Carrato from Ramón Y Cajal University Hospital in Madrid. Then we are going to also discuss and look through the answers from Professor Lise Lotte Gluud, who is working in Copenhagen at Hvidovre Hospital in Denmark, and she is one of the experts in pancreatitis, different types of differential diagnosis versus pancreatic cancer and the needs in this very important group. And we will also actually start by looking at the answers that we received from Dr. Stephen Pandol, who is one of the experts working in diabetes -- nuanced diabetes after the age of 50 or later older than that. And the connection into pancreatic cancer, which is so very important and now rising in importance. So if we start looking at the any of the diabetes risk group, as you have heard earlier, both Tom and Linda referred to the fact that we haven't actually presented any specific results to this particular risk group as it is. However, diabetics -- patients are included in the early symptoms group and quite many of them. And we also, as you heard Tom presenting, we have worked with a number of experts, one of which is here, in defining and designing our prospective study for nuance diabetes after 50, which we call PanDIA, which is collecting a large number over 6,000 patient samples. So looking at the answers that we received from Dr. Pandol, I think I will start by emphasizing the question, is it any clinical unmet need in this area, and is all the time referring to the diabetics -- nuanced diabetics group here and the answer is very clear. Yes, there is a need. And that is quite relevant because in this case, the diagnosis is late, the symptoms, or diabetes is being one of them, are not specific for the disease. And therefore, early diagnosis and, of course, surgery would lead to long-term survival, better outcome for the patient. I think another very interesting answer that he's giving us when seeing the data that we have that you have seen from Linda for this particular symptomatic group, we are answering to you foresee that this test would deliver clinical utility and the answer there is a good possibility that this test would contribute to improving outcome. And he refers to the prospective study that we are working on intensively. Then if we move on, I'm going to bring to your attention the answer to question 4, which relates to what more would you require in terms of data and proves to this test for the current clinical setting. And I will quote this test as it stands, could be used in the clinical setting to a diagnosis. If we then move on to the symptomatic risk group and Professor Lise Lotte Gluud, it's also extremely interesting her answers, because she is talking from working with this symptomatic patients coming to the clinic. And the need here is, of course, as high. It is very difficult to diagnose the patients in time, and they are diagnosed at a late-stage and very large number of them. So even here, the patients with symptoms and an obvious mass on diagnostic imaging are difficult to diagnose. So this is also extremely important in this group. I also want to raise the question too, in which we are asking for this particular results and this group and based on the data we have, do you foresee that this test would deliver clinical utility. So she is referring to the need for evidence, which we, of course, are working on and we need to show that. But on the other hand, it's very good to see that, I'm quoting her answer, "it is definitely promising". And what I think is important here is even state of the diagnostic imaging she says can sometimes give us results that are inconclusive when used as secondary screening or when making the primary diagnosis. And this relates very much back to Tom's description of the different combinations that are used for each patient. And in this case, of course, accurate blood-based tests could make an important difference. Question three, I think, is also relevant because it's a common answer that we received from the majority of the people we interviewed where we are asking so, how important would a blood-based test and she is answering also as what kind of clinical benefit, so it's a combination here where our blood-based test is important, not only to improve diagnostic accuracy, but also very important to reduce the number of necessary scans. And moving on, I will just raise one answer which use for question #5. And then the question relates to the test performance, is it appropriate for the presented risk groups? If not, what would you like to see improved? And what is extremely relevant here is the fact that it is important that the test should help exclude false positive patients, a high negative predicted value is very important. We have worked -- so all the experts we talk with, they all have worked with different types of tests with low PPVs. This is less important as we will be able to combine the test with patient characteristics, in this case, patient clinical information and diagnostic imaging. Now moving into the last group here where we are looking at the answers from people who are leading familial hereditary screening programs. One here is an U.S. expert and to the question about the clinical benefit. We also see here that for this particular high-risk cohort where we have asymptomatic individuals, a noninvasive option is the one that they are looking for, and it's extremely well received by them. Then if we are looking into how do you foresee this test to deliver clinical utility in the clinical setting of the familial screening group, here, the answer is relating to the use of the test in triage kind of utility where it would recommend the use of imaging, and then the other more invasive test that Tom was referring to, to be able to localize the tumor that we wouldn't know about in the beginning. Then also here, we find that in terms of the data and the proofs, there are not so much more than we have now. Of course, we are working together on the familial -- on PanFAM on the prospective clinical study there. But what is also interesting here is the fact that it would be important to use the argument that the high NPV is helping to forgo imaging if the blood test is negative. And we are going to end this extremely interesting voices of the pancreatic cancer experts with Professor Alfredo Carrato from Madrid, whom we have been also collaborating for quite a lot of time. He's one of the founders of the hereditary screening programs in Europe and also part of the European Pancratic Cancer guidelines. And he is also President of the Spanish Association for Research on Cancer. He is looking at this from the surgeon perspective, which is also extremely relevant in this group as the vast majority of the people leading screening programs are the surgeons. And here, the clinical unmet need, it's also confirmed and he sees it has a possibility to put the patients through neoadjuvant chemotherapy, the new type of treatment, that would then make them candidates for surgery. And in this way, enhance their cure rates and their overall survival time. Then I think it's also interesting to see the clinical benefits that he is looking for, where he is describing the discarding patient from being diagnosed of pancreatic cancer and avoiding what we were also discussing earlier the morbidity mortality associated with going through surgical procedures. And then I think I will just end with the answer to question 4, where we are looking at -- do you foresee that this has to deliver clinical utility in this group than the -- I think it's interesting the answer that says that the scientific community here, the people working with this, they're the ones to provide the characteristics of this population so that we will be able to offer the screening program with Immunovia's test. And lastly, for question #5, in terms of the proofs, Professor Carrato, who is, I think, if I remember correctly, is our second participant in PanFAM and part of designing PanFAM study here. He is, of course, referring to the fact that a long-term follow-up is needed for these healthy persons with a sequential blood test with an appropriate sensitivity specificity that would then help to look at positive results that would then take them into imaging to confirm or not the pancreatic cancer diagnosis. So we have reached the absolute end of our presentation. And I just want to say in conclusion that we have a unique test. Our IMMray PanCan-d that shows an outstanding performance for detecting early stages PDAC. And we are very excited that we are going to be able to bring this test to U.S. self-pay or the sales start in Q1 2021 and I believe 2021 will be a much better year for all of us. Thank you, and a session is open for questions.

Thank you very much, Tom, Linda and Laura. We are indeed open for questions. I have received a few questions already. The first question is from [indiscernible], and it's relating to the commercialization. Lately, Immunovia has communicated commercializations to start in Q4 2020 and lately Q1 2021. What does that actually mean related to the above, meaning commercialization to [indiscernible] is take orders, ship products and invoice customers. So briefly, if I just take this one, what it means for our case is that we will perform the validation study at Q1 2021. And at that time, we will be able to start the commercialization activities. We will then file for the final CLIA application and get the clear approval and be able to start testing patient samples in Q2. So that's what commercialization means to us, that in Q2, we will be doing the actual testing and the invoicing. The next question comes from Viktor Sundberg. There is a series of questions, and I will direct these to Linda. Can you clarify what improvements you will make in IMMray PanCan-d?

We will clarify these improvements once we have tested them. So we do improvements and we do the required testing. And after that, we will communicate, but as I said earlier, this will not affect sales start, and it will not affect the frozen signature.

And he further continues with this question. Does this include changes to the algorithm or the biomarker signature and would this imply -- and what would this imply for further studies and commercial launch? You responded to some of that already.

Yes, it will not affect the signature and not the bioinformatic algorithm or the cut-offs and not the sales start either. So we are on track.

And then his last -- or his third question in his mail, you have previously stated PPV of 78%, what assumptions did you assume in that calculation given today's results that do not reach a PPV of 78%?

We have never stated a PPV of 78%. We stated an NPV in the press release last time, not PPV. And this is the PPV we have in this webinar is correct and estimated on the clinical relevant prevalence.

Thank you, Linda. Your next question comes from [ John Balmer ]. Thank you for a good webinar. Will the same cut-level be used in the different risk groups PanDIA, familial and symptomatics? This is either for you, Tom or Linda. And Tom, you are on mute.

Linda, I think I should let you speak to that first from the development perspective.

You can start.

Okay. I think it's possible that we might use different cut-offs. Again, using the test in different clinical situations to make different types of decisions about escalation of diagnosis is going to be favored, in some cases, by a higher sensitivity or higher specificity. So I think that's an option for us. But we have not decided definitively on that as yet.

Yes. Currently, we have the same clinical --

Okay. Thank you. The next question comes from Felicia Rittemar. It's a series of questions. I'll read the first one. We note that the test performance in terms of sensitivity and specificity of Cologuard developed and commercialized by Exact Sciences has varied among studies conducted between 2010 and 2015, including case-control studies, pivotal study and studies performed after launch without any major changes to the test. Based on your knowledge and experience in the area, do you agree that this observation is standard in development and commercialization of diagnostics? And is the same logic applicable in Immunovia's case and going forward? Tom?

Yes. I think that, that is the case. I mean, they certainly may have looked at different prevalence populations. Oftentimes, it's going to be difficult to precisely specify what the prevalence in a given population is. So I suspect that some of the variance is due to that. It may also be due to how they actually performed the test although they may not have changed the test itself. The proficiency of the individuals performing it may have improved and that's something that one would often see after you launch a test that it becomes slightly better as people get more accustomed to doing it and more proficient.

Linda, you want -- anything you want to add?

No, I agree with Tom. And I think this is absolutely normal and something you do. You continuously improve and try to improve your tests also after launch. So this is nothing strange and you just -- you can do changes and then you do acquire testing according to your standards and quality requirements, and then you can implement these improvements.

And I can add, from my own perspective, having been in diagnostics in 35 years, that it is quite common also for screening methods in different areas to continuously see better and better results. And it's particularly common in -- when you are in -- approaching the launch phase that you continue to slightly improve the test going forward. So that is a very common observation in diagnostics. The next question from Felicia is the past few weeks -- in the past few weeks, there has been confusion/misinterpretation regarding the prevalence used to derive the predictive values. In my understanding, you have used the prevalence prevailing into clinical setting for the different high-risk groups where the test is intended to be used. Is this correctly understood? And can you elaborate on why this is the most relevant figure to use?

Well, I think clearly one wants to evaluate NPV and PPV in the population that the test is actually going to be used in. Now that said, these are heterogeneous populations. As we said in the familial group, the relative risk can vary from 2 or 3 up to over a 100. So it's not entirely the same group. And you'll never or very rarely know exactly what the prevalence is when you're testing individuals. Speaking as a clinician, certainly, these studies and the percentages are important in terms of clinicians deciding to use our test. But I think as a clinician and having worked with oncologists and surgeons and gastroenterologists over the last 30 years, a new test is something that they have to try out, they have to gain experience with and gain confidence in and decide how they're going to use it and what kind of role it's going to have in their own diagnostic process. That's a little less scientific, but I think it's the way almost all clinicians crack this medicine.

Thank you, Tom. And then her third question is, can you elaborate on how Immunovia's test is differentiated versus CA19-9? As I understand it, CA19-9 is mostly used for assessing treatment response and prognostication, not for early diagnosis.

That certainly is correct. CA19-9 is not specific enough to use for primary screening. And if one applies it to a general population, the number of false positive results is huge because it can be elevated in other situations of inflammation of the peritoneum and other types of GI conditions. What we've been able to do in the asymptomatic group, and Linda, you can speak to this more than me, is to really boost up the sensitivity and specificity of the assay substantially beyond where CA19-9 is. And that's the reason that we've been able to increase the positive predictive value of the test and really, I think, deliver a test that performs very well against anything else in the -- certainly anything noninvasive in the current diagnostic milieu.

Linda?

No, I agree with Tom again. I mean, it's not used for diagnosis more above following up on treatment. So I think our tests have a great value in comparison.

And before I move on to the next set of questions, I just would like to remind you all that there is an opportunity to ask questions. And as I did that, I got more questions. So I will just move on now. The next question comes from [ Patrick Dalimar Sterner.] How does Immunovia IMMray PanCan-d perform compared to competitive pancreatic markers? And do you foresee a threat of their product beating IMMray PanCan-d in clinical utility price or time to sell-start?

I guess I can speak to that initially. I mean, things that we consider competitors at this point, Grail and Thrive, are more screening tests for the general population to identify people who have cancer. They have some specificity in terms of saying what cancer that is, but that's more limited. So we actually view them as not entirely competitors but potentially situations where our tests might be used. If someone had a positive test there suggesting that they had a cancer, our test might be used to really nail down what the actual site is and direct the diagnostic workup without subjecting the individual to very invasive tests. So I think the answer is probably largely no at this point, at least for tests that we feel are close to market. For things that are far away, it's harder to know because we have less information about them.

Yes. Can I add just a couple of words?

Please.

Yes. So what's also very obvious about the very large Thrive and Grail, which we know are extremely large companies. And they are aiming, as Tom said, for screening several cancers in the same time. And their results to pancreatic cancer, they are, of course, much more limited and with a lower -- much lower number of patients and their studies are, of course, because they cover many other cancers in the same time. So there is a very good complementarity in this case. And we are focusing on PDAC and pancreatic cancer. And with this focus, we are unique and we continue to be so. And even though we have seen, of course, the launch predicted for 2021 for Grail, but nevertheless, again, we will look into complementarity more than competitive situation.

And as you point out, Laura, I mean, our experience in terms of the number of pancreatic cancer samples that we have evaluated is massive compared to them. I mean, since they're looking at many different cancers, they have a few pancreatic cancers. And at least in some of the initial publications, the sensitivity for detection of early disease did not look very good to me.

Very good. We're moving on to the next question. This comes from Michael Loughman. Can you give us an update on the pipeline products with RA and lung? We are continuing to collect samples for lung cancer. As you will appreciate, obviously, we are in the midst of a worsening COVID-19 situation around the globe. And this is, of course, particularly affecting lung cancer sample collections as such. But we continue our efforts and we will definitely update the market as soon as we have new information with regards to lung cancer. With regards to RA, the situation is very similar. We continue still to bring in samples, and we intend to update the market when we have conducted discovery studies there. So we very much look forward to presenting more data on our platform technology going forward for those 2 diseases or disease areas, I should say. Moving forward to the next question, which comes from Lars Hevreng. In asymptomatic population, how will doctors view the risk of false positives in general? Will it be seen as adding a lot of worry and work burden? Or will it matter less since the population is anyway subject to extensive clinical evaluation?

I think I could take that. I think it really depends on how suspicious the situation is. So we talked about in early symptoms for pancreatic cancer, most of them are rather nonspecific. So individuals that have only 1 or 2 of those symptoms, the likelihood that they have cancer or the prevalence in that population is going to be low. So I think that's going to be viewed rather differently than somebody that has a constellation of a number of different symptoms that seem to be more suggestive of cancer. And in that population, I think what our tests can do, at least a positive test, I think, would really accelerate their diagnosis. And that's something that's really important. Because I know -- I have known people who have really bounced around from doctor to doctor for over a year trying to get a diagnosis for their GI problems, and they ultimately die rather quickly of pancreatic cancer. So this is a very important inflection point where I think our tests can really have a dramatic impact on patients' lives.

Thank you, Tom. And we move on to the next question. It comes from Johan Blomquist. Can you comment on why the ROC AUC value has dropped in the PDAC versus symptomatic control group from 0.93 in the CTMS to 0.84 in the most recent study. Linda, you want to take that?

Yes. That's, of course, very difficult to explain. We run new cohorts every time we run a study. And for this particular cohort, we have this lower accuracy and ROC value. But as I said, we have discovered improvements that we are working on right now, and I have very great hopes to improve us significantly. So I'm not that worried about it right now.

I would say also that -- and Patrik, you can speak to this as well. As you move from a test that's in a research development situation to the actual application of a clinical test, there are a lot of things that you add to it in terms of quality control, and I think those things may help us as well. This is a very highly controlled test as we've developed it. The custom informatics that we built to analyze the samples gives us a lot of metadata that really allows us to say with a good deal of sensitivity and specificity how well an individual assay performs. So with a test like this, both pre analytical, analytical and post analytical, variation is important. And I think we're in a really strong position to have a very robust clinical test with all of that data.

And what's also important in this symptomatic group is that it's a very diverse group including a lot of diabetes, pancreatitis, different types of pancreatitis, liver diseases. So it's -- you need also a large number of samples since it is a very diverse group. So that's also one point that could be approved going forward.

Yes. Thank you, Tom and Linda. We move forward. We have one more set of questions from [ Patrick Dalimar Sterner.] What will the price per test be? Do you foresee a risk of healthcare not using the test because of high price and low impact on diagnostic pathway? Laura, you want to take that?

Yes. I mean, we have been running a large number of payer interviews and surveys that we have run in the U.S. and just to ensure that the price and the clinical benefit and the willingness to pay into this payer insurance coverage programs is well understood. And we have also validated the price in several other instances with the different customer target groups that we are going to aim for. And the price of $600 per test that we are going to sell for, it's an extremely well accepted price, and it's very much in line to the clinical value and the benefit it's going to offer.

And Patrick then continues with the question. Furthermore, have you calculated health economic impact i.e., will the test actually save healthcare systems money in excluding some people from unnecessary follow-up? That kind of question is most often highly relevant for clinicians. If I start on the health economics question, and then Laura, you can follow-up. We already, many, many years ago, 4, 5 years ago, started very deepgoing initiatives to examine and review the health economics impacts. We've worked with a lot of leading experts in the area. It is, however, clear that when you get the validated test and the test that you're actually going to use, and its performance is, of course, the sort of final basis on which you're going to do the final health economics calculations that is obvious. However, before I hand it over to Laura to complete the response, I would also say that in the case of the familial and hereditary risk group where with a very high sensitivity and with very high specificity can detect PDAC in Stage 1 and 2, we actually save lives. And as you are well aware, when you are able to demonstrate that, that is sort of the ultimate goal really for a company like ours and for the healthcare systems, and therefore, also for the health economy. Laura, you have anything to add?

Yes, I think it's -- yes, it's a very obvious case that is quite easy to calculate in the case of the hereditary familial, especially because it's also asymptomatic earlier, younger cohort. But I think we have actually worked with all the different groups. And we have done different kind of, as you can do, as Patrik was emphasizing, the need of having the final parameters and clinical performance so that you can actually finalize this. But we have done different kinds of assumptions and working together with health economy experts. And there is -- there are very good cases because if you also think about the way we were describing, you are as using the testing triage screening situation, using the testing, in that case, you are going to eliminate unnecessary scans which is one of the benefits that we are looking for. Then there is also the possibility of differential diagnosis and also confirmatory diagnosis, which is also accelerating the workup of the patients and also providing a lot of extremely valuable time for the clinicians. And how do you actually value that, that is also an enormous benefit to the way clinicians are going to judge using the test. On the other hand, I just want to work -- make one very important point because we have been in so many meetings and discussions with the payers and all kinds of -- both from national to regional, public and private, and every single one of them is having their own health economy experts. Every single one of them is going to perform one of these analyses as soon as we are having the validated test and the validated performance of the test. So it's going to be part of that evaluation. And of course, that's why we are preparing to provide the data to our filing process. But then it's going to be also confirmed by the experts they have.

Patrik, I just wanted to say about the U.S. environment that I think may be different than the European is certainly the insurance companies are going to look at health economics, and I think we can pass that test, but not the clinicians. Clinicians don't care. If they believe this test will help their patient, they will find a way to get it done. I mean, that I can say for certain. And perhaps we should be more penny wise. But as a group, I think we are not in the U.S. If they think it's valuable, they will get the test done.

Thank you, Tom. We have one more question from Felicia Rittemar. Since you haven't presented any spec or sens -- specificity or sensitivity for the NOD group can, one, based on biometric calculations from verification study draw any conclusions regarding utility for the NOD group?

No, I wouldn't do that. First of all, we don't have any patients from the energy group in the verification study. We have diabetes included among the symptomatic controls, but these are too few to take any statistical calculations and be sure of these. So we have the PanDIA, and we're really looking forward to run all these samples. We have over 5,000 collected. So we will have really good data. But in the current situation, I wouldn't do that.

Then we move to the next question, which comes from Peter [ Holmgren.] Will the test be improved continuously as you learn more about its performance?

I can speak to that, I think, in terms of the U.S. experience. Certainly, the core of the test, I think, will remain the same. But since we're launching the test as a laboratory developed tests in the United States, we have quite a lot of flexibility in terms of improving the test. Obviously, any kinds of improvements need to be validated. But as Patrik had said, once a test is launched, I think you learn more about it, you learn what works and what doesn't, both logistics and in terms of actually performing the assay. So I think those things we will take to heart, and we will work very hard to improve the test. But that's one big advantage, I think, of launching as a laboratory developed test rather than something that's CE Marked or FDA approved where you have very little latitude in making any changes to the way the test is performed.

Linda, you want to add some comments to that?

Yes. I see definitely that we will work on improvements in parallel. I mean, after we have launched, the R&D department will start to continue to work on improvements, and that's what you normally do. And since we have this opportunity to change in the LDT, that's a great opportunity for us.

Yes. And it's actually a very common practice in our industry, as we have mentioned before. So the next set of questions come from Viktor Sundberg. If you do not calibrate the algorithm or the biomarker signature, what other parts of the test could you calibrate to improve your results? Linda?

Yes. I can't comment on these improvements now. As I said earlier, we will comment and we will tell you when we have tested these performances and the improvement on this.

Thank you, Linda. What specificity level would you think would be acceptable in the NOD group for your test? Tom, you want to?

Well, I think that's hard to know at this point. It may vary depending on what the clinical situation, what kind of doctor there is. I can tell you, one thing we've seen very strongly as we've talked to primary care docs is those in concierge medicine are always on the lookout for an additional value test to offer to their clients to pay a bit extra to have special care. So in that situation, we think that we have a good opportunity. I think it's hard to know exactly what the sensitivity specificity needs to be to perform well in that group. What I think we need to do in terms at least of payers is we need to increase the percentage of individuals in those groups that have cancer. So we need to increase the prevalence to 3%, is the figure that's usually tossed around. Clearly, individuals with NOD later in life who also present with weight loss, are at very substantially increased risk for pancreatic cancer. That's a no-brainer group that we would certainly want to test. We're working with experts in the field to identify other ways to increase that prevalence to a point where things would make sense from an insurance payer point of view to fund this test. And that's going to be, I think, much more important for us than with the other indications for this test. In terms of familial testing and in symptomatic testing, this is a limited population of individuals. It still may be a large number. But if we think about all of the new diabetics diagnosed every year in the U.S. and in other countries, that number is huge. So I think in that situation, adhering to the payer recommendations is going to be very important for us. We also have to do a lot of education, and that's why we've been talking to primary care groups, making them aware of the risk of pancreatic cancer in NOD. And we plan to continue doing that. Many clinicians are really surprised by this because it hasn't -- while it certainly is in the medical literature, it hasn't been promulgated by the Diabetes Association as a strong association between late onset NOD and pancreatic cancer. So that's something that we're working on as well and I think we'll pay dividends. Certainly, when we had some of our key opinion leaders and myself speak at these groups, there's a tremendous amount of interest, particularly from the concierge stocks, but really from all of them. Most primary care docs in the U.S. or I think very many of them have had a patient die from pancreatic cancer. And that's an event that I think really affects them because this disease is so horrible.

And of course, in general terms, we can say that the NOD group is fairly large; from a U.S. and European perspective, roughly 3 million new patients every year and the prevalence is slightly below 1%. So it goes without saying that the specificity is, of course, a key performance indicator there. But obviously, we will learn a lot more when we perform the PanDIA prospective study on the fully validated test, and we look forward to reporting that. We continue with Viktor Sundberg. Any comment from your key opinion leaders on your side on the need for a mortality outcome based to convince USPSTF to switch its stance on screening for the NOD group? Do you want to take that briefly, Tom or?

Well, this is a deadly disease. So we pretty much know that the vast majority of people diagnosed with ductal pancreatic cancer are going to die from their disease within 5 years. It's certainly a very large percentage of them. So I think those numbers are out there. I think that diabetologists in ways, I think have been kind of sticking their head in the sand and ignoring this. I mean, this is a large population with a significant risk of pancreatic cancer. It's really surprising to me that there hasn't really been more movement in this direction. But I think getting a test out there with strong specificity is a way to really move that forward. And then maybe we are going to have to be the prime movers in doing that.

Yes. If we continue with Viktor Sundberg, last part of his question, considering that there are a few patients with familial and point mutation related risks in the U.S. currently undergoing surveillance for PDAC how do you propose to first identify and then convince the vast majority of remaining patients who are not yet known to the healthcare system to get tested? If I start and then Laura, you can follow-up. I think there's a high awareness of pancreatic cancer in general, not only the U.S. but also here in Europe. So in people with a family member who has had pancreatic cancer and died from it is very aware of the risk of also being affected. And there's very active patient groups, et cetera, who are looking at these types of family members. So I think the level of awareness is extremely high currently. With that, I'll ask Laura to follow-up.

Yes. And I think the very, very important point here is the work of the patient support organizations, and particularly in the U.S., they are doing a fantastic job in keeping in very tight contact with the family members and supporting them through this process with the relatives and then also afterwards. And there is a network that is supporting all these people in which we have been involved from literally day 1. And I believe very much that this collaboration is crucial for us in bringing this test to the right people who are going to be in need of this. We have here PanCAN, we have also pancreatic cancer action in the U.K., we have some souls in [indiscernible] Julie Fleshman and the ones who have been dedicating their lives literally to help these people that are left behind and that are in risk and that would need to know and understand the risk and what is possible for them to do out there. And they are doing really an amazing job. And this relation and the collaboration that we respect enormously has been helping us to reach out to a large number of these family members through the, what's so-called, the walks that we have been part with our colleagues dedicating so much time to this in the U.S. And then also now they are turning to these virtual events and we are continuing to be a very active supporter of them and being directly -- literally in direct contact with these people all the time. Tom, do you have anything that you'd like to add to this?

I think we're heading for a kind of grim inflection point because pancreatic cancer is thought to surpass colon cancer as the second leading cause of cancer death in the U.S. in the next decade, probably in the next few years. I suspect from the American Cancer Society and others, that will really grab people's attention. It has to. So I think that the horrible toll this disease takes in terms of rapidly killing people in their productive years is going to be more and more recognized as we go forward. But I certainly agree with Laura, and I've been to some of these walks as well. I mean, these organizations are really good, and they really keep it towards the forefront. And I think some of the things we've done as well, like lighting things purple for World Pancreatic Cancer Day. We lighted main bridges and other buildings throughout the world for that, and certainly, throughout the U.S. and Canada. So I think we do a lot of things to make people aware and to get them to recognize this terrible disease and what may be done about it.

Thank you, Tom and Laura. We move on. One more question from Felicia Rittemar. We note that you or Cancer Diagnostics are priced above USD 600 to USD 2,000 per test, even higher for genomic analysis. Do you think there is a price elasticity towards higher priced, high-performance test in the U.S. for PanCan-d?

I think you can always talk about price elasticity. But I think for all the work we have done in understanding the price and the value, I think where are, we are in a good place with our pricing strategy.

And from my side, I think, obviously, there should be price elasticity opportunities. It's something that we will continue to monitor and look at. And we do think that also other tests and their pricing structure also makes that possible for us to at least speculate in the fact that we might be able to move our price points up. But it's also -- we need to remember how it sort of balances with the health economic benefits that we bring, and those things do go together. We have a question from [ Victor Danielson.] Why did you not include new onset diabetes without early symptoms in the verification study? I would have thought that this is a relatively large proportion of your total targeted patient groups.

Well, I think that we've pursued an iterative strategy where we really have focused on the familial patients through the asymptomatics initially and then symptomatics, and then clearly for us, in terms of the education that we have to provide for clinicians in terms of NOD, even though it's a very large population, a lot of things need to be done, I think, before that. And we feel, I think, just getting the test out there for familial and genetic patients, we'll start to give clinicians experience with their test, and that experience is very important in terms of them deciding whether it's a useful test for them and their patients. So I think that our strategy is a good one in terms of introducing the test in a market and then expanding that market to include other indications where we need to do more work in terms of developing, as Laura mentioned, with PanDIA, developing the science and the numbers to support this. But also it's important for us to educate clinicians. And I think that we're making progress there, but we need to make more.

Thank you. Next question comes from [ Tim Sao.] Any update on the collection of samples for the validation study? Laura?

Yes. Well, I think what we could say right now is that there is a significant effect of COVID-19. It's -- probably you'll see it and hear about it all the time. In our case, we now see the closing of some of the collection sites that were really working fine. So in this case, even the Swedish. But I think we are still on a good track because we have been collecting during this time even at a lower rate. But I think we will definitely be able to have a drink, isn't it, Linda?

Yes, agree with you.

Yes. I also know that we will be where we need to be with regards to collecting all the samples for the validation study. The next question comes from [ Rudy Beutell.] I hope I pronounced that last time correctly. And it's for you, Tom. As soon as the test is available, what are the steps for self-payers to get tested in New York state, referral, question mark, test site, et cetera? So...

Well, our strategy really targets other states before New York. New York has a different licensing strategy. It really pretty much requires you to be licensed and accredited by the College of American Pathologists. The pathway for us to begin testing is to first get CLIA license that involves a state inspection that will allow us to start testing. Once we've done that, the College of American Pathologists should visit us. And I've been an inspector for the college for many years. So I don't expect any difficulties with that, but it takes time. So again, we're taking an incremental approach with most of the 50 states. The CLIA license is all we need. But in New York, you need additional. Florida, California, there's some specific licensing requirements that will take us a little bit longer. So in terms -- and Laura, you can speak, I think, in terms of our commercialization strategy, where we've targeted different geographic locations.

Yes. No, I think that's exactly how we have built our efforts. So New York is coming afterwards just because it has its own specific process.

Thank you, both. The next question comes from Lars Hevreng. Could you remind us about the time lines for initial data release from the 3 ongoing large prospective trials? So it's PanFAM, PanDIA.

Yes. I think what we have said all the time remains so far -- it's the same that we are going to analyze these cohorts as soon as we have a validated test. And that is what we are aiming to do and we are collecting very well and we have had the opportunity to collect also during the time when we didn't have COVID, thank God. So we could actually collect the samples that we needed.

Yes. And the plan is to come out with more details on the actual timing for reporting our results, and we'll do that very soon.

Yes.

And then the final question comes from [ Krista Nielsen.] It is, again, for you, Tom. It's -- this reminds me of a question we've already had, but I'm just going to bring it out, just to be sure we capture it. A question regarding CLIA approval process. Could you comment on this with a few words? How long do you expect this to consume in time? And is there a risk for Immunovia needing to deliver additional data or studies?

We have to complete a validation in the U.S. And I think we actually have almost enough samples in our U.S. banks to do that at this time, but we'll get additional samples from Lund as well. So that's the last thing we have to do. We've already done work on sample stability and transport on terms of interference, if there's hemolysis or other changes in the sample, we've evaluated that. We have that data and we have data on precision for our test. So most of the things we need to provide are there, plus the laboratory itself. We spent a number of years now really getting the laboratory in very good shape, very well documented in terms of what we do and having a strong quality management program in place. So the way that the licensure for CLIA works is we initially filed an application with the state of Massachusetts. They have status to inspect for CLIA. They have 30 days after that application is complete to visit us. And at that point, they will look around, see that we have everything in order and give us the go-ahead to start testing. At that point, they will return in somewhere around 90 days. And at that time, they're going to want to see that we have been doing testing, that we've been doing it appropriately. And then that's kind of the final certification. But as soon as they give us the go-ahead to start testing, we can start billing. And we will. So that really is the key for us to launch. We were a little more concerned about the ability of Massachusetts inspectors to come out with COVID. We actually have a fortunate location of our laboratory kind of just across the street from that building. We can see from intelligence that there are people in the building. So I think at this point, I'm a little less concerned about that than I was last fall. So I think we'll have a smooth ride through that. And as we said then, as soon as we have that, we'll apply to the College of American Pathologists, which is really -- I would consider to be the gold standard in terms of accreditation. That will take a little bit longer to get in place. That will be very important, I think, for discerning academic clinicians to accept our test because it's a much more detailed examination than the CLIA exam. And then after that, we will move on to New York, and Florida and California as makes sense for us.

Thank you, Tom. This actually concludes all the questions. I have no more questions in my inbox. I do want to thank Laura, Linda and Tom for your presentations and your responses to the questions. Thank you so much. And of course, a special thank you to all of you who have listened in through the webinar and participated and for asking such intelligent questions. And also, thank you very much for your continued interest in Immunovia. Thank you very much. This concludes this webinar. Thank you.

Thank you.

Thank you.

Thank you.