Computer generated transcript

Warning!
The following transcript was generated automatically from the content and has not been checked or corrected manually.

thank you. Hopefully everybody is suitably caffeinated now, Um, and most people are back in the room. That might just be a couple of people will filter in. So we're really privileged to have Professor McGary here from Denmark. So Linda is the director of the Danish Multiple Sclerosis Registry, which I'm sure we will hear more about in her talk, which is also collaborating wildly or white widely, I should say across the world, possibly wildly. Yeah, but it's it's really great to have you along and thank you for giving a talk today and thank you for inviting me. So I hope I can convince you that it's worse collecting data in registry because the Danish M s registry has existed for more than 70 years. And when we started collecting, it's not me. When my uh, leader started to collect data, we didn't actually know what to use it for. It was for prevalence incidents, mortality. Then we also started collecting data on disease modifying therapy because we had to, uh, document it for the health authorities that the drug works in order to get it reimbursed. So therefore, data collection was mandatory because we already had a practice, is it? So there was no question whether we want to or not collect data. And when we published the first observational study in I think it was 1999. Nobody ever cited it. They had no at all. So we thought, Okay, this area is that we just collect data and see what happens. But now it seems we have a renaissance of observational studies, so And how that influenced clinical practice. I think I see that it influenced clinical practice. At least in them are a lot. And as I can see the interest of the registries and also for international collaborations, uh, I think it's going. It came to stay, and then we see how we can use it. So But, uh, luckily, some of the things I wanted to say it's already said so I can focus on other things. So let's see better. I can find out, uh, use the keyboard. Okay. Thank you. And, uh, yes. So and this was already set, So I just mentioned that the gold standard was the randomized control trials. Uh, an observational studies are Hello. However, so then maybe an efficacy effectiveness Get that we heard from Orla that there is a selection. We want to know how the drug performs in the UN selected population, people with comorbidities older and younger people. And with the increasing number of therapies, there is absolutely not possible to do. Have to have study. It's expensive. It's, uh it's also not ethical to do placebo controlled trials. So we need to have a real word data, uh, to assess effectiveness and safety. So I see, uh, observational studies as a supplement to clinical trials. Yeah. So what is But what is big data? Big data. It's huge amount of data collection, which is collected by a non interventional day. And this is me during my PhD. While until 2009, 2000 and 14, we collected data manually into the Danish M s registry because my supervisor really wanted to make sure there is good quality. So he was checking every patient, whether everything was, uh, was correct, so that that was me working hard during my PhD. This is the first privilege study. So these these registries are very important and invest case their population based on nationwide and follow the same population for a long time, but also databases who just maybe just a clinical trial with an extension or data bases within the hospital There is in some countries are administrative claimed that I can be mind. That's the kind of big data medical records, like MRI images, Uh, and what we also heard I think they're I think in in not many years, we are going to collect data on variable devices. We have already been in a study with the Danish M s Center, the rather CNS study, where we have investigated whether we can find a proxy for the EDSS. Uh, I would say the Danish registries in 2015, there is an online data collection platform and it increased the data equality and data density and integrity a lot. So but how does it influence clinical practice? I used to say that registries has to be good for the patient, for the clinician's, for the authorities and maybe for the industry. So be in in Denmark. Uh, the Danish M s society partially funds the registry. Therefore, we make them annual reports. Those annual reports are presented to the politician in order to obtain funding also for the M s hospitals and also for the Council of Expensive, uh, hospital medication. Because in them are, uh it's a tax paid the healthcare so the patient got delivered or dispense the medication from the hospital, but in their back, they have national guidelines and are very they're very guided by the price. So I regularly present data to the authorities in order to adjust that guidelines, uh, and also has to be good for the clinician's. And, uh, there are more and more clinicians who ask questions whether, for example, we just got approved, uh, tomorrow, Bob and see. Okay, can we just see which, uh, characteristic of those patients, uh, that started of them Aubagio Can we see the patient who converted to S P. M s? But it's but let's get back to my topic. Oh, I have to Yes. Yes, yes. So I think I think in the increasing numbers of DMT, uh, we we need some good population based studies too, which can guide a guide. Our treatment decision. And I think if we look at what the big debt has contributed to clinical decision is when to start treatment when to switch treatment, whether we should stop treatment. And in this population, can we predict whether a DMT would work or not? And this was when to start treatment. This was one of my first PhD students. After that, we collected this 20 years data, and we had, uh, not very positive experience of publishing it. Then, uh, remember, I was talking with Gary Carter. He's a statistician at the narcotics, and his marina tried this early late, then see whether it can give you something. So we we, uh Com compare those patients treated early within two years after clinical onset Those who were treated late to to eight years after clinical onset and find that those treated early reached it, assessed my stones six, uh, later than those treated late. And we can we could also, uh, see an effect on mortality. So we convinced the Danish authorities that we do not have to wait for to relapsing to try to start the EMT, which was the rule in Denmark. You could not start patients, uh, after after one relapse, and, uh, but there was no treatment for C. I s patient and the same exercise. We, uh, repeated in the big data network that also, uh, is very active. And the M s base of the game is data network is a collaboration between five s registries the French, the Swedish, the Danish, the Italian and M s base. And since 2013, we have been working together to define the common data murder and two and two to define project where we can report reliable results on comparative effectiveness or other other scientific question, uh, touching clinical practices. So we did. A different approach not only are delayed but dividing the time from, uh, onset into 5 20 years. And again the figure talks for itself. It's, uh, starting early. Postponed disability. My stones. Yes. So it's easy to start early. But what should we start early? Should we start moderate or high efficacy? This was the next question. The first article was very well accepted. So I saw Then let's see whether whether there has any beneficial effect of starting high efficacy treatment early and that time we actually use high efficacy treatment, Uh, from the beginning, a bit of labor. So, uh, and this is not our study, I would say since we conducted the study there has been a huge interest to confirm better high efficacy from the beginning, uh is beneficial in, uh, preventing the relapsing and postponing disability? Actually, all these studies conducted in different populations all, uh, agree that the risk benefit ratio it's worth it that high efficacy treatment reduction reduce the inflammatory activity, the late disease progression. But we have some safety risk we have to take into account, because when we talk about the big data and, uh, observation of studies, we can see that sometimes the results are different because because they are conducted in different populations and we have we have done different exercise, especially with them as space, and we can see it. If you compare to populations. You may not get the same results and other population because the patient profile that was treated with a specific drug in one country may be different from the patient profile from another country. So but all these studies seems that high efficacy treatment is beneficial. But let's get back to it, because that because I'm not saying that we should use high efficacy treatment for everyone. So this was our study, the Danish M s registry and then we could compare 200 patients, uh, who started high efficacy be 200 who started moderate efficacy. And it was our first propensity score matching analysis, uh, and it it's a very clear, uh, there's all debt to the probability of relapse was lower. And, uh, also we postponed confirmed disability worsening. Uh, and we have another approach them as based, this time collaborating with the Swedish M s registry. So they looked at from two different time points. And see if we if we count the time from the start of DMT or from disease on set and, uh, see whether early, uh, whether we can see a difference of when we start the high efficacy treatment and, uh, as you can see or I have to get used to that. As you can see, the they are very nice diverging curve. So they also it's convincing that early timing of high efficacy treatment prevent patients, uh, functional deterioration. So but there are different, uh, different outcomes we can have in observational studies. And yes, I'm very happy that I heard about pros and other outcomes not only EDSS and relapses, because what we have an observational studies, they usually annual relapse rate time to first relapse confirmed. EDSS worsening confirmed SS progression or impairment. So but this study looked at conversion to S. P. M. S by timing of treatment. And also this was a study by SK Centers and also there's cohort. And even though we say transition to S P. M s is something up to discussion because it's very there is no one fixed time, There is no biomarker that can say now you transition to S P. M s. Uh but they looked at the predictors that can postpone conversion to S. P. M s. So treatment prevent early treatment within five years. Also, escalation of treatment within five years, even though you could sing. Yeah, but if you escalate that within five years, then you experience the disease breakthrough. But even though escalating to, uh, Hy effective drug, it was associated with a low probability of converting to STMS and also early intensive therapy. Even so again, very convincing results. And this was a comparative effectiveness study between Sweden and Denmark. And, as you know, Sweden and Denmark has been a, uh yeah, historical competition. So I am really sorry. We lost this competition. But I got suddenly journalist and the highest authorities calling. So it changed a lot of clinical practice because And that was Tim Spelman. The M s base registry said, Oh, but could be compared because you have, even though there are two countries who are genetically comparable and very similar structured and even use the same data collection platform. So the same variable the same data model, everything. But we have to totally different approach in Denmark. We are very cautious, and we really have to have evidence that everything worse. So we look more about safety, go slow and think about long term safety line is Swedish. Uh, they go for, uh, in that time, they have been treating 40% of the patients and rituximab, and this was, uh, patients between 2013 and 16. So they go for high efficacy of label. Uh, and, uh, we come and and now, uh, we have a different approach. As you can see, the selected people said we have to We have to mimic a clinical trial as good as we can. So only those below 5, 55 years, and, uh, you can see the think. Oh, sorry. You can see the figures. Uh, the hazard ratio for, uh, this was the relapse. The first is proportion of, uh, relapse. Every patient was lower than Sweden. Sort of a probability to have a relapse in Sweden. In the in, the in the population who started therapy between 13 and the 16 was lower. And you can also see the proportion of progression free. Patient was also, uh, 30% lower in Sweden. So, uh, and the difference was that in that time in Denmark, we started about 17% patients of high efficacy therapy from the beginning. And the Swedish was, uh I'm almost 40%. So in Denmark, the most used drug, that time was terrible for me there for a minute because that was guided by the national health authorities. Why in Sweden was natalizumab and Rituximab. So, um, but as you can see, this is this is, uh, actually last year, uh, data from the Danish M s registry, and you can see the highest of the columns is the proportion of newly diagnosed patients starting therapy within one year. And the green part of the Koran is the high efficacy treatment, so you can see an increased since the 1996 2000 so you can see more than 90%. We started very early and, uh, I looked at last week. 50% of the patients, uh, now start high efficacy treatment from the beginning because we also modified the criteria. What is highly occupation? And interestingly, I go. I got to have a conversation with Norwegian colleagues last week. And in Norway's they go in for everybody starting high efficacy. So they want to get approval. So let's see what's happening. I am not that radical, because I have. I have some. I think there are other things to take into consideration, but it changed. Clinical practice is a lot, uh, but when to switch there is no disease modifying therapy that promises 100% relapse freedom or 100% stopping of, uh, of disability worsening go EDSS worsening. And they don't even measure cognition. And, uh, hand function and life quality and, uh, fatigue. Uh, sleep. Other things. Uh, so why do we switch treatment? Failure is expressed as, uh, clinical activities or relapses and MRI activity. It's very rarely that progression indicates treatment, which in them markets. Uh, sometimes we do it some time. Or if it's intolerable intolerability of infusion or specific side effects. Adverse effects like infection or liver failure. Family planning, uh, compliance adherence issues and, uh, some psycho social reasons. So but I will only cover the treatments reaching for treatment failure. Because also, in the Danish M s registry, the median time to staying on a treatment is about 2.53 years also. And, um, of course, this is shorter now, because now we, uh, there is mandatory to monitor patients with MRI. So re baseline MRI after 33 to 6 months, then yearly MRI and a yearly basis. So when when it's a disease worsening, we have to act on that. So But we'll switch from and two which disease? Modifying therapy. And, uh, this study was shown, so I will not go into details. But this is a nice study from the Welsh cohort, uh, showing again that, uh, early intensive therapy was more beneficial than escalation therapy. So I will show another one. And this is the Italian, uh, M s registry who selected only treatment naive patients and, um, follow them up for 10 years, and they looked at that. Those who escalated or early intensive therapies, they had a better outcome. So again, switching was less beneficial than escalate. Uh, that that early intensive. Then how about letters? Which and that was one of our I think it's our studying. Yes, because then the oral disease modifying drugs became available in 2013. The authorities was discussing because but the flutamide and limited for Murad is not more effective than the injectables. So, uh, so we should just, uh we should just switch it literally if it if there are side effects. And then we also got another advice that it may be that some of the drugs anyway, they are effective, so we should not switch to high effective. But I switched to the auras, so we did two studies, and this is one of one of those where we looked at switching to lateral effectiveness or versus escalation and switching to hive and high effective, which were natalizumab and fingolimod, uh had better effect to preventing relapses and also time to three months confirm this E s S E s s worsening. So we needed some documentation to convince, uh, authorities to let us change clinical practice. And, uh, this was another study. Where then the patient started to ask us. Yes, but, uh, if I just switch because it's more convenient to do drugs would get switching gap and have have some give some risk of disease worsening. Then we looked at those who switch from injectables do to convenience two orders and those who continued injectables. And because we had a paper out the definitive humor. It was a bit more effective than the front of my. Then we had to separate the groups and see, Did we find a difference? And luckily, we could say to the patients know letters which, if if other reason than treatment failures has will not give you a risk of relapses and idiots is worsening. So I think I think there is a mess. Big body of evidence about starting early is important, starting high efficacy and escalating. So I think we are convinced, uh, with lots of observational study. But when to stop treatment, do we stop treatment? Do we consider stopping treatment? I think it's so difficult. I still have patients who have been in on injectables since 1996. There may be 60 on some of them 70 And, uh, I try to start this conversation every year, and then they ask, Yes, but what if What if I get worse? I now I got used to my injectables. And where? Why? Why? To stop it. Maybe as some of these patients as stable have no relapses, no MRI activity, Maybe they have, uh, s P m s with very slow progression. But if I cannot measure the progression with the media says it's only a subjective measure. We can stop by space safety reason, for example, some of the immuno suppressants in in older patients with stable disease We may, uh, affect the immune system and they get very infection. And we don't know the long term safety of malignancies, uh, and older patients with stable disease also for convenience. Uh, so and in some countries, they're also access related issues. So why to why to stop it? And do we have the resources for clinical and MRI monitoring after the cessation? Because some of the clinic clinic, at least in Denmark, they do not allow regular control of patients who are not in disease modifying therapy due to resources. So So that also keeps the patients on the drug. And will the patient be compliant to clinical and MRI monitoring? And so so there are lots of questions to ask. I think I think stopping treatment should be the next, uh, topic explored by registries. But I stick as a Should we stop or not to stop? So, uh, so the snow is Why should we stop? We should stop because the in there is a decrease in active MRI lesion with age. So we we should consider stopping patients in older patients with stable disease because there are evidence that the lesion law decrease with more than 50% compared to the younger. We know that relapses decrease very much after the age of 45 and we actually do not have sufficient evidence on effectiveness on the empty on aging M s because they were not included in the randomized controlled trials. However, now I'm not talking about not starting a DMT because a patient is a late onset, a very late onset m s. Now it's It's about stopping if the patient as stable, so and I can see I was confused with this You unless I just just listen. So why, uh, what can happen if we stop this rebound phenomenon? Can a cure, especially with some of the drugs, Uh, and Natalizumab Gilenya, I would say we just got accepted a paper in New England and, P uh, NMP about Gilenya rebound and found that actually the Gilenya rebound in the Danish population was only about 3.6%. But it's called Rebound the definition of rebound and those people who get disease reactivation in the cessation period where those who had treatment failure on Gilenya. So, um, we have we have to look at the exact numbers in highly active patient, even their orders. We should not treatment. And we also have to remember the stability on DMT is not a benign M s and we have some evidence and it's very interesting. Some of the different registries had has addressed this question very differently and nicely. So this is the M s base, and we know that the M s based are heterogeneous. And this study also was including all kinds of age groups. But they looked at, uh, more than 1000 patients with uh, 15 years follow up, and they did this Maginot structural model. So they added all the time of patients on treatment and separated for those periods without treatment and find that being on treatment, reduce the risk of disability a crew about 44% and also reduce the risk of impaired gait. So this would convince us to to stay on treatment. But all all patients were included, So not independent of age. There is. I think this is, uh, let me see. This is an Austrian study who consequently evaluated 220 patients who were on injectables, and they stopped due to convenience or side effects or no treatment failure. Also heterogeneous age group. And they followed up two years after discontinuation as they find out that about 44% of patients who had a relapse but then they looked at. But which patient did uh, presented a relapse as they find out that yeah, factors preventing relapses of those patients who did not experience the readers. But those who are older than 45 years did not have a relapse for more than five years and did not have a contrast enhancing lesions so they differentiated and see if younger patients with the previous relapse and MRI lesions stop because of side effects of convenience. The risk of disease reactivation is high. But if it's an older patient that we we may take the risk, uh, as it in the traditional outcomes, we could not see any clinical, uh, or MRI deactivation. And this is the Lorraine registry. And they only looked at patients over 50 about 500 those who are stable disease for three years and stop treatment and you can see they're paralyzed. So nothing happened, Uh, in in relapses again, we have to remember there are many outcomes we do not collect. But there is another approach. We talk about escalation, highly intensive therapy, induction. But we very rarely talked about the escalation. So I find this study very interesting, which emphasized the role of age. And this was the Rocky Mountain Center who had compared. And this is an interesting approach. They compared orals taking, uh, fingolimod and limited Humira with the infusible high efficacy rituximab and natalizumab and had some nice, uh uh figures with Luckily, it it is the same shape. So they looked at that the disease activity This is the This is the probability of disease activity on the X uh, on the Y X and the exact is the age. So we can see that the probability of disease activity decrease with age is and we can see at the relevant are the auras and the blue the, uh, injected birth. So again, starting the younger patients on highly injectables lowers the probability of disease activity and the same pattern we can see with escalation. Uh, and with high efficacy early So starting high efficacy early May on, you know, escalation may on the treat the young patients, while escalation actually works acceptable about, I think 54 years and higher efficacy early may over treat the older patients. So So we have to. We have to look at the age and also the induction therapy. It may be, it's it's good for the younger people, but then they may experience disease breaks too. However, when we look at the curve, the escalation therapy could work after the age of 50 so that could be an approach to try and to to see whether we have more data from clinical practice. So But now I talked a lot about registry data and all these watches, and and we we always we want some bio markers. We would like some predictors from the beginning, Uh, including MRI. And this is an approach, uh, big machine learning data approach where the authors really hoped that MRI could predict, So they included this less. Oh, uh, it's a kind of machine learning where you put a lot of variables, and then you exclude them, uh, one at the time and see what, uh, what can predict. So they included 35 30 37 baseline variables, including MRI. I'm sorry to say they find out that MRI the 1st 10 years could not predict disease. Uh, development. What? The measures. The variables that had, uh, of a prediction value, but actually those we collect into registries. So actually, I was said when I saw this paper that MRI did not, but I said okay, but all these all these so you can see we have a job at onset. We have the baseline. A d s s, uh, F s score. Uh, 25 ft walk. Sensory attack, Duration of first EMT. So how we look at this. It's, uh if it's, uh, always, uh, that's the worse prognosis, and it's if it's negative numbers, there's a better prognosis. So they have some outcomes, like lots of benign M s. And you can see the only the only, uh, thing that predict loss of, uh which prevent or it's for benign MSE sensory attacks. Or the other agent runs that high pyramid er they are all against benign m s and also so it's actually it's very expected. Also, developing aggressive mess is increasing with higher the s s high age it onset, um, and actually even with duration of first DMT because that means that it's a high disease activity. So and on the EDSS development, they find 12 different, uh, factors. And here they also included the baseline brain volume fraction. So, um, this I think this is a nice work, and, uh, it's also encouraged that what we collect is important. I hope that MRI can also MRI some MRI measures. And there are also other studies showing that can have a predictive value. But what we can back to my title. But we can use big data too, is to, uh to guide us the clinical a guide clinical practice and to give up, uh, on time answers. Therefore, sometime in one register, this emphasize is too small to answer scientific questions. Therefore, it's very important with the collaboration and see if I have a number of patients, then collaborate and generate some answers. So just to, uh, summarize it if we have the m s from disease onset, the disability progression, what we can see from, uh, observational studies debt from disease onset. In the younger age, we have a window of treatment opportunity. And with the increasing age, everything gets worse. We'll get comorbidities. We get, reduce the effectiveness of the EMT and, uh, more degeneration, less inflammation. So therefore, we should use that window of opportunity and remember not as not to generalize. So maybe, uh, an age and gender is two things which, actually very easy to, uh, sub group patient's, uh And, uh so my take home message that we should continue with collecting data into registries, uh, and include other, uh, other data and to the question whether these biomarkers included, uh, the M s registry is a part of the Danish M s centers of now the, uh we try to include at the genetic data and biomarkers. Uh, we have tried with the MRI, but that's, uh that's very difficult, because the MRI is not is a part of the radiological department, but we are We are having actually a study in Denmark where we try to look at the machine learning the images. Then let's see what, uh what's happening? So starting early and the time the high efficacy treatment where the inflammatory process is, uh, active. And, uh, but also remember that not everybody has a beneficial effect of high efficacy treatment as safety should be considered. And we I think we should dare to discontinue treatment, Uh uh, more often in older and stable patients and maybe put data together and see how it goes. I'm struggling to implement patient reported outcome in Denmark, and, uh, I did it in January. 750 patients has had an excess, and only 33 has completed. Uh, question is I? It's I know. And then the UK registered patient do clinical patients reported outcome. I don't know how to encourage the Ms Patient to do it. So again, the personalized treatment is very, very important. And this is my team at the Danish M s registry and, uh, thank you again for participating. It's really excellent, Melinda. Thank you. It's such a good overview of of all the work that has gone in, um, in Denmark and the collaborations with other places and how it's actually moved things on in clinical practice. So I think we're all really convinced. And what questions do we have with in the room? Um memory, Uh, the question I was thinking about what I was listening every center at different machines and different people. Of course. How do you make sure that the information that goes into the registries are harmonised and there are no differences, depending on what the machine to use? The MRI's memorized. That's a problem. And we couldn't but But now, for the the that was really a problem, because we don't have the images into the in the registry, we only have the number of the two regions and get enhancing. But after many years this year we actually succeeded to have a standardized MRI MRI protocol accepted. But remember the markets a small country, we have only 13 centers so and the M s community is so small. So if I see you, if I see some inconsistencies of data, I can just write and say, Could you please correct that? That doesn't make sense. So it's a very close community. So and that was very nice talking. And my question is perhaps a basic one with some explanation. And But I found it really interesting when you were talking about considering when to stop the d. M t. And you mentioned that one of the risks was having a rebound. And I was just wondering, What's the clinical definition of a rebound? Or what's the clinical part of the evidence that you would need to consider that one? That's a good question, because I have been answering referees for the last two months about water rebound. So I looked at, uh and I think sometimes we use a rebound in the wrong way because the rebound is an increase of clinical rebound. It's a crease of relapse activity, which is higher than before, starting the d m t. So we usually in clinical practice, we think about relapse, rebound. If I stopped Gilenya, then the patient get, uh relapse. It's also, even if the if the, uh, seriousness of a relapse has been much more serious, like a tomb effective lesions or or a very severe relapse that's also a rebound because of severity. So, uh, but but just. But if the patient had a relapse before starting Gilenya and stops Gilenya and have a relapse there before starting, maybe an antiseptic 20 that's not a rebound, because they have been in the same level. So and But I'm really as a radiological rebound one has been seen in a in a terrorism, um, and fingolimod also increasing of the lesion load after this continuation meant for that we are not that good in clinical practice to make on MRI. Maybe the first three months after this continuation. Therefore, we actually don't don't catch it very often, okay? And that's the first couple of minutes apologies they had given me, uh, get our patients able to opt out of the registry in Denmark. Yes. Yeah, they they are now, but they only had the access for Yeah, January. February. Yes. So now it's actually a wish from the patient and also for health authorities. Now we have to have patient informed. So we have a big study where we are as patient. What They want to, uh, add to the register we had as the clinician. Then we try to maybe at the two groups and it and the top, that the patients really want to see the data. But my hypothesis is when it's only that may be the younger and, uh, patients who are born in the digital era. And my I am curious how long they bother because it's very interesting. Then you have a new gadget to look at it. And then after a half year, I don't care how many steps I went. Uh, So So let's see that Yes, they have they so they can come back to me and say, Oh, I had a relapse. Why didn't you, uh, record? Uh, patients have to be, uh, no. Uh, no, no, they can't. Only if they if they don't want to be in the registry. We cannot give them the ent. So we'll so that that's the price for the Because I have to report to the health authorities about the quality of care. Otherwise, I we are being cut on the budget and other things. So I have to see that all we see a patient at least once a year. They perform an MRI, we do the antibody testing, and they referred to physiotherapist and different. And this quality indicators. It's not the authorities who defined it. But there is a There is a clinician from the center. So there are certain people and then we decide what's good quality. The authority says define good quality, define the standard of it, But then keep it. And I am the data police who has to yeah, has to go after them. And, uh so yes, so So And and that's the data is anonymous. So we don't We don't even have to ask the patients. Uh, and I'm also very lucky with the GDP are about the M s registry because in 1983 I have a document that we can collect all data without informed consent, and that's reality 2035. So I keep a practical how many people are involved in maintaining the register, assisting you with actually all these people. But But before I came, there was one person. For 30 years there was one person after 30 years. Then I I have It's me half half time. I now have a data manager and a statistician on epic hematologist and some PhD student. And they have a half secretary. She's the data coordinate. Or she, uh what? How many patients are on the, uh, we have 18,000 living patients, but we have 35,000 in the hole registry. So, uh, so actually, we we take data from the data collection platform. And every month we merged together with the old registry from the started in 1948. So So all the data, and then we also collaborate. So it's, uh yeah, we try to do it very efficient. Uh, but they have practice for 60 years. So, uh, and the young doctors are very, very much more keen on on on doing the data and asking for the, uh and, uh, really excellent we might have to get We might have to have a trip to Copenhagen to learn from your Yeah, Thanks for when you go back to the very start before when it was set up for it. Where did it come from? The compliment clinicians who working no This is actually a very funny story. So I mean, I know it was one young doctor called Ky who listed who was looking in 1948. And the professors, Uh, professor, Whenever he diagnosed, it's just open for, uh, diagnosed. Ms. Patient put. How do you say, Like a opinion, You know them. Them marked map. And he looked at this young daughter. Nobody's around the big cities. There must be a geographic distribution. Then he he has wrote a letter to all insurance companies, banks, even the Salvation Army that if anybody knows an M s, patients should report it as they get. I don't know, 10 pens, and he put his wife as a research assistant. I have all the all the, all the documentation in the cellar. It was really times when you brought nice letters to each other. So the his wife was sitting there, so it was one person and doing it his free time. And so if if your neighbor had a mess, you have to go, you could go to him and see I have a neighbor who has a mess. I got this money and then he collected all these journals that time was no GDP are. And then he find out that 25% of the MDs patients have not m s. Then he got the idea. We need a registry because otherwise 25% of the population gets a diagnosis that is devastating. And it's not true. They maybe have something else, uh, neuropathy or other thing. And then he did it, actually, Kona more. And so he did it himself. And, uh, and ask for the all the clinics, the records and, uh, put it into a paper registry, and we can see it. Then when it was 2000 mgs patients, it was 1956. It was prevalence day, and, uh, and he he made this doctor dissertation and in them but nobody, nobody saw the value. So the Corsica so the value of it and say, Oh, then I can ask the American society to fund the registry. So for 10 years, we got actually for funding from the U. S. M s society because then course came and traveled to the Pharaoh Islands to make the, uh so and then you'll just advised that. And I'm sure I'm sure some of you know your score, Kendricks. And he took it over. And then he said, Oh, but this is very old time. He made a bleed of programming. So and then the, uh, ms Society took it over in the seventies. And believe it or not, when I take it over in 2014, the health authority said that, but, uh, we don't need this registry because we have the administrative claim. So the registry is actually privately funded. So I am. I am doing all these applications, so I'm keeping it alive. And the m s society funds. Yeah, some of it. So that's the story. Thank you. We'll finish there, but thank you so much.