 Good afternoon, everyone. My name is Ana Vallejo, I'm the Communications Manager for Myeloma Patient Europe. Welcome to this webinar in the ADN 2020 webinar series. As you know, our Amodinal Meeting was cancelled due to the coronavirus outbreak and we are delivering our educational program in a different way. Today's webinar on big data, the harmony example will be given by Ana Rodriguez, Institute of Cancer, Molecular and Biology, which is in Salamanca, Spain. For the information, this webinar will be fully recorded and will be uploaded to the Myeloma Patient Europe website, which, as you know, is www.mp.eu.org. It will be also available in our YouTube channel, so if you would like to share the webinar with someone or to work it again, you can go to our website or our channel to find that file. Before we start, I would like to explain briefly the webinar agenda, especially for those who during this webinar for this time. As you know, the webinar is scheduled from 6 to 7, so the presentation will last about 20 to 40, 45 minutes, and then I will open the decision for questions. There are basically two ways to ask questions to the doctor, and one of them is using the microphone as I'm doing right now, the microphone in your computer, so just press the right hand button that you will see on your screen. I will unmute you and you will ask that you can ask the question to your doctor, to the doctor. If you don't want to use the microphone, the other possibility is to be inviting in the question master window that you will see also in your screen. I will receive all those questions and I will read them so the doctor can answer them. On behalf of Myeloma Patient Europe, I would like to thank you for your collaboration and for your time to give this webinar. Thanks again and thank you very much. Okay, thank you so much Ana. I hope everyone can hear me well, okay, and good evening everybody. I'm Ana Rodriguez, as Ana said, I'm a postdoctoral researcher in Salamanca. I'm also involved in harmony and that's why I'm here today in this webinar. First, I would like to say that it's really a pleasure having been invited to present Harmony and I also want to thank NPE for making this possible, for organizing this webinar, which I think is a great idea, especially in this difficult time for everybody, and I hope that you all are okay and thank you for joining us today. So let's start. Oh, I don't know what is happening. Okay, so the main ideas of my talk are these. First, I will talk about blood cancers, then I will try to explain what's harmony and the genetic data. Of course, I'll talk about big data in general and in healthcare. I will explain what's harmony, big data platform, and how it works. Also, I will give you some details about the ongoing myaloma project in harmony. I will talk about what's next in the project and the expanding collaborations, and I will talk a little bit about our next project or the second chapter of harmony. And finally, although I forgot to write it down, sorry, I will be happy to take all the questions you have. So blood cancers or hematological cancers account for about 40% of cancer cases in children and about one third of cancer death. Blood cancers include diseases such as leukemia, lymphoma, and myaloma, all of which have an impact on the production and function of blood cells. Developing and optimizing treatments for these life-threatening diseases, many of them rare, can be a complicated process. That's why collecting and harmonizing high-quality data on outcomes and assisting treatments is crucial, but it's often hampered by the lack of data, the radity of hematological malignancies, as well as the variation in healthcare practice throughout Europe. So these three represents a huge challenge to clinicians, to researchers, and regulators. And here it is where harmony takes place. Harmony Alliance is funding through the Innovative Medicines Initiative, IMI, which is Europe's largest public-private initiative, which aims to speed up the development of better and safer medicines for patients. IMI supports collaborative research, project and business works of industrial and academic experts in order to boost pharmaceutical innovations in Europe. Within IMI, there are several programs, BD4VO, which stands for Big Data for Better Outcomes. It's a comprehensive European research program aiming to support healthcare system transformation through the use of big data. BD4VO will develop platforms for integrating and analyzing diverse data sets focusing on outcomes that matters to patients. BD4VO has four disease-specific projects so far. Roadmap for Alzheimer's disease, big data at heart for cardiovascular diseases, harmony for hematological malignancies, and pioneer for prostate cancer. These four disease-specific projects are supported by DO-IT, which is Accordination and Support Action, and EDEN, which will be a federated network of relevant data sources. So let's talk about harmony. The Harmony Alliance is a public-private partnership for big data hematology established in January 2017. It's a European network of excellence consisting of 53 public organizations. This large number of stakeholders across different European countries makes a fragmenting playing field from basic hematology research all the way to approval processes for new medicines. And involving every stakeholder group is necessary to meet patients' needs. That's why people involved in this project in harmony come from very different backgrounds and the patient is always at the center of the project. So we divided the work into eight separate but interconnected teams or work packages, responsible, for example, for the project management, data analytics, dissemination, legal issues, and so on. And each work package is composed of specialists and organizations that represent high-level expertise in their particular field. As I just said, harmony approach is patient-centric, and that's why patients are present in the project. They have boys and they have many things to say, but how are patients represented in harmony? They are represented by means of these seven patients organizations, which is a unique group of seven European patient umbrella organizations working in different areas of hematological diseases within the Harmony Alliance. They combine years of expertise in patient-led scientific research and policy advocacy, and this is called the patient cluster. The patient cluster and their patient communities are involved in, for example, in the definition of outcomes, in the design of the research project, the information to the patients. And in addition, the patient cluster also makes sure that there is a reliable bi-directional flow of information within and within the Alliance. So the most important thing is that for the first time in a major research project, the patient community sits at the table as truly equal partner to researcher, industry, and regulators. So at Harmony, we are working together to collate data from all over Europe from as many as as many patients with blood cancers as possible. For blood cancers, big data means gathering into one single database, clinical, genetic, and molecular information on patients and diseases, which is currently contained in a number of individual databases from clinical trials and registry in different countries. Another goal of Harmony is to define the outcomes that should be collected and reported, representing the priorities of clinicians, industry, health authorities, and patients alike. What Harmony wants is to harmonize this definition at European level. Harmony also works to increase the application of the genetic data in the clinical practice. I will see later on what's the meaning of the mixed data. And another urgent need of hematological malignancy is the draft development. And Harmony is working towards the speed-up of this process. So, omics data seems to be a crucial thing for Harmony, but what's the meaning of omics data? To answer this, first, I have to talk a little bit about the human genome and human gene sequencing. A genome is the genetic material that makes up living organisms. It is contained in chromosomes, and chromosomes are made from a chemical substance called DNA, and DNA is a sequence of smaller units. It's dot here, called bases. The DNA bases combine to form genes, and these contain these genes contain the extraction that are passed from one generation to the next. So, all our genes together are known as our genome. So, you can imagine how important it is to know how we are built. And why the human genome project was created. The human genome project was an international effort to discover the exact makeup of the genetic material. This project involves scientists from all around the world who work together to achieve their aims. And the project began in 1919, and the first draft of the human genome was published in 2000. The step after sequencing is decoding to figure out what the sequence of letter means. After that, scientists must determine which part of the sequence fits on which human gene and what his gene does. But although knowing how our genetic material is made is a huge, huge breakthrough, 20 years later, we are still sequencing. We are still applying this technique to better understand the diseases. In this case, cancer. Last February, we saw the results of a massive analysis of the entire genomes of over 2,600 people with 38 different types of cancer. Why are we still sequencing? Why is omics data still so important? Because health outcomes are multifactorial. There are several factors involved. The literature highlights five major categories of health determinants, social circumstance, environmental and physical influences, behavior, genetics, and medical care. So it is a challenge to estimate the contribution of each factor. Besides this, we need to take into account that there is also relevant data outside the medical system. And this data also contributes to health outcomes. So 60% of data are from outside, are exogenous. But which is really important, and I want to stress now, is that 30% are genomic factors, which is a lot. In fact, in our daily life, we can see how these data are recorded, for example, through smartphones, by personal tracking devices, etc. And healthcare is one area where big data has the potential to make dramatic improvements in the quality of life. But big question, what does big data mean? Okay, big data is a collection of large and complex data sets, which are difficult to process using common databases or traditional tools. So big data refers to sets of data that are too massive to be handled with traditional hardware. Every day, we can see how big data influences on our lives. And some of the most useful innovations of the past 20 years have been made possible by the use of massive data tools, combined with the computer technology. For example, we have become used to finding almost any information we need through the internet. You can locate nearly everything immediately by using a search engine, such as Google. But Google couldn't assist without the ability to process massive quantities of information and extremely rapid speed. Another area that has changed our lives forever is online shopping. We have, for example, Amazon, where we can buy almost every product we use in our daily lives online. We have it delivered too. And even better, we can see the reviews provided by other customers to decide whether to buy a product or not. A website such as amazon.com must process quantities of information that would have been unthinkable just a few years ago, in a quick and efficient way. So both are examples of big data in our daily life. The factors that distinguish big data from other types of data are first volume. Clearly, with big data, the volume is massive. Big data systems are designed to work with very big volumes of information. They are built in such a way as to make it easier to add new data as they are produced. Second one, velocity. Velocity refers to the speed at which data is gathered and we get the data. The architecture of big data systems allows near real-time processes. They are designed to store data in such a way that the speed of analysis is several times greater than that of traditional systems. The third place, we have variety. Almost any time of database can be used in a big data system. Variety refers to the fact that the contents of a big data set may consist of a number of different formats, including spreadsheets, videos, music leaves, and so on. Storing a huge quantity of these incompatible types and different types is one of the major challenges of big data. And last one, velocity, because big data systems add quality controls to data before the data are processed in order to assure that the results can be trusted. In recent years, big data was defined by these four v's, but now there is a fifth one. This is value because there is no sense in having a big data system if the information generated has no value because that itself is of no use. It needs to be converted into something valuable to extract information. So you can say that value this v, this number five v is the most important one. So to sum up, the main message is that big data refers to data sets that are too massive for traditional data set management system. And big data requires more sophisticated approaches than those used in the past to handle the information. So the next question is why do we need data and why big data and why harmony? To fight blood cancers in an effective way, as I said, we have to understand the disease inside out. And as I said in the beginning of this talk, the critical questions in the field of blood cancers can only be answered by studying large number of patients. Spread all across Europe, there are databases from clinical studies and public registry containing clinical and biological data from thousands of blood cancer patients. So the Harmony Big Data Big Data Platform launched in June 2018 is the key to enabling access to a massive amount of high quality data by collecting and engaging multiple individual data sources from hospitals, from public registry, from pharmaceutical companies, and from University Medical Centers. Because just using high quality data, we will allow the performance of meaningful analysis. So in Harmony, we anonymize these data and assemble them in one in one harmonized Big Data Platform. The Big Data Platform observes some of the highest standards and principle for data safety to ensure maximum protection for data donors. So the main characteristics are the data platform is secure, is reliable, is safe, is private, and is anonymous. How is this possible? This is possible thanks to some characteristics of the database. This is a centralized platform, is isolated from internet, uses big data technology, segregated management things are in charge of the database. We use tool for advanced analysis and we use a common data model because we want to harmonize all the datasets we record. So the Harmony Big Data Big Data Platform is a central repository where the anonymous data donated by our partners and associated members is collected securely following all legal and ethical requirements, harmonized, and then analyzed. And as a summary of the talk till now, I will say what makes the Harmony Alliance unique. There are several things that make Harmony a unique project but if I had to say something, I will say that is the large number of stakeholders across different European countries from different backgrounds, medical doctors, IT people, regulators, patients, everyone has something to say in their lives. So let's move to more specific things. I'm not sure if I have already mentioned this but seven hematological malignancies are integrated into the Harmony project. So Harmony Alliance uses big data technologies to improve the treatment of seven blood cancers. Acute myeloplukemia, chronic lymphocytic leukemia, myelodysplastic syndromes, multiple myeloma, acute from myelocytic leukemia, acute lymphoblastic leukemia, and not whole thin lymphoma. And Harmony has research projects addressing specific questions for these hematological malignancies. So Harmony works by research question. So there are dedicated research project teams who have started to analyze the data and conduct clinical research. For example, in chronic lymphocytic leukemia, to understand the pathophysiology of the patients where we evaluate a target panel containing the most frequently mutated genes in CLF. Or for example, in multiple myeloma where we validate and all further improve the revised international staging system. I'll talk more in depth about this in a minute. Now I'd like to explain briefly how this works. I know this is a business life but I will try my best. First, we need a research question. Okay, so as I just said, Harmony works by question. So we need a proposal with questions coming from any member of alliance. The proposal is reviewed and if it's good, it's approved. And then we ask for data coming from the partners. And as a complaint, the data is gathered, the partners send the partners send us the data and is uploaded onto the big data platform. Depending on the research question, we can be talking about different variables. For example, demographics, genetics, quality of light, etc. So inside the platform, the data is processed, which means that is uploaded, is anonymized. We also have to check the quality of the data because as I said, we need high quality data is harmonized because each data set comes from a different provider. So we need to give them all the same look or the same structure. And finally, the data is ready for the analysis. At this moment, we have over 20,000 cases in the Harmony platform and we are waiting for more. As 45,000 data sets from patients with blood cancers have been identified and are being transferred to the Harmony Big Data Platform for analysis. Okay, at this point, you must be thinking, okay, this is great, great project, there are a lot of partners as they call their blah, blah, but what about myeloma? Okay, so let's talk about multiple myeloma projects in Harmony. As I said, Harmony works by project. In multiple myeloma, the challenge is that the outcome of multiple myeloma patients is heterogeneous. In 2015, a research certification algorithm named Revised International Staging System was developed by Palumbotol. In this work, newly diagnosed multiple myeloma patients were identified. And we saw that this patient population represents an admit clinical need. So the main goal of the project was to provide an updated report on the RISS prognostic role and highlight potential improvements. More specifically, the scientific questions were, are the variables defining RISS all the same? Is this staging system still valid after an extended follow up? Are there other variables predicting overall survival? What is the impact of treatment on these variables? So data from different European cooperative groups were collected through the European myeloma network and registered in the big data platform developed by Harmony. The primary endpoint of this analysis was overall survival. And you can see here the variables recorded. These patients, these are the patients analyzed approximately 1,200 more than in Palumbo study. Coming from 14 clinical trials and these patients had a median age of 65 years and a median follow up of 74 months. Okay, this slide, let's talk about the results, but this slide just illustrates the characteristics of the patient included in the study. Regarding treatment, here you can see the treatment distribution. Half of the patients in red receive an immunomodulatory treatment. A quarter in blue receive a protein inhibitor. And approximately another quarter in green receive both. And approximately half of the patients 45% were transplant ineligible. And regarding risk factors, the baseline risk factors used are ISS, where ISS, ISS, ISS, LD, LDH levels and cytogenetic alterations that they did by piece. First, we wanted to know if all the variables used to calculate the ISS were the same, if they had the same importance. To do this, we perform a multivariate analysis and the results are represented here. Okay, I don't know if you are familiar with this analysis and this representation, but basically the line here represents zero and everything on the right is in favor of high risk and everything on the left is low risk. The farther away from the line, the highest importance. So if we focus on the chromosomal abnormalities, the chromosomal abnormalities that in our ISS were defined to have impact the so-called high risk chromosomal abnormalities, we can see that evaluating. Translocation 1416, 414 translocation and 17B deletion positivity, just 414 translocation and 17B deletion confirm their role as independent risk factor. Why 1416 translocation did not? Therefore, we define high risk chromosomal abnormalities as 414 translocation and 17B deletion. Based on this, we split the patients into three groups stratified by our ISS. So stage one, two and three and we analyze their overall survival. As you can see here, the prognostic role of our ISS was also confirmed in this larger cohort. It means that adding the 1200 patients that were not included in the original report validated the stage insist. And finally, we tested whether additional factors can impact overall survival. So let's take a look at the blue line. So the newly diagnosed multiple myeloma patients with an immunoglobulin A monoclonal component, so a worse overall survival than the rest of the patient. The same happened when a baseline of creatinine clearance lower than 45 milliliters per minute independently predicted overall survival. And the amplification of chromosome 1Q effect on overall survival was also important. And finally, patients with a poor prognostic performance stages were at higher risk of their as well. That we not only confirm the prognostic role of our ISS within the largest cohort of newly diagnosed multiple myeloma patients analyzed so far, but also we detected other independent overall survival predictors that can help us to further refine the current prognostic method. Here you can see the roadmap of myeloma multiple myeloma in the harmonic project. So first we have the approval of the proposal. Second, we have the then we need to transfer the data to the platform. And we are here. The aim is to have big data, not only a big data set and less not stop here because we want more projects because this pilot project demonstrated that data collection and harmonization is feasible across Europe, European multiple myeloma cooperative groups. So we are really happy to say that work work is in progress. And last year in 2019, we celebrated the midterm success of the Harmony Alliance. And what will happen for the remainder of the project? So apart from finishing the current projects, we have new sets of research question identified by our partners and patient organization. We are also expanding the collaborations with cooperative groups with other IMI projects and real world data. We are also receiving expressions of interest and proposals for collaboration all over the world as that I have from Korea, from Australia, where they want to build a similar network. And also great news is the approval of a new project or the second chapter of Harmony, which is called Harmony Plus, whose main goal is to expand to new diseases such as Hawking Informa, Chronic Myeloid leukemia and so on. It's a three year project and is organized in a similar way. And it will start later this year. Harmony also has a strong social media presence. So I invite you to visit our website to join our community, to watch our videos, to join on Twitter or on LinkedIn. Harmony has achieved many things so far, but I don't want to be boring and repeat it again. So I'd like to recap the main points of this presentation. And here we have what I consider the five most important achievements of Harmony during these two years, which are engagement, data, building the big data platform, creation of value. And what I think is the most important one, realizing the potential of big data in hematology. So by the end of its life cycle in December 2021, Harmony aims to capture data records of 100,000 patients with blood cancer. And that's all. Thank you all for your attention. And I'll be happy to answer your question. Thank you very much. Sana, for your presentation, now I will open the short question. So if you have any questions, just please remember to remind you that we have two ways to ask questions. One of them is using the right hand button. I will unmute you and you can ask the question. And the other one is, as you know, doing writing. So write your question and answer with them and I will read them to the doctor. We have here are some questions that are writing you in your presentation. The first one, you mentioned clinical trial and release trees to gather data. But is data coming also from real life? Is real world evidence part of big data? At this moment, can you hear me well? Yes. At this moment, we have clinical trial data and data from registry. But we are planning to incorporate real world data. It means data from hospital or the data from other registries. Because now, maybe this is a retrospective way of looking at the data. So in the future, we are planning to incorporate real world data in a more prospective way. I don't know if I explain myself very well. But yes, real world data is the next step. Thank you very much. Next question. What type of quality of life and patient-related outcomes data can be captured? Why is this important? Okay. As I said, the most important thing in harmony is the patient. So that's why in the project, there is one world package devoted to patients. Patients are also involved in harmony. So they can have their research question. They can have proposals. Quality of life, I think, is a very important thing for patients. So it's not only important in clinical trials or in the data we are used to have. It's important. The follow-up is important. I don't know how many patients you have. But in real life, for the patient, the quality of life, if they live better or worse, it's also important. So we are paying attention also to quality of life indicators. Because the important thing is not only that patients live longer, but the important thing is that patients live well. So that's why patients are also involved in the project. Because maybe they are the most implicated in deciding the quality of life indicators, which is what or which indicators are important for them. So we are open to research projects. And at this moment, the patient organizations are working on the outcome definitions. And they need to decide what quality of life indicators are important for them. Thank you very much for your answer and the next question. Gathering and analyzing big data sounds very expensive. Is this something doable for all countries? Sorry, I can't hear you well. Yes, can you hear me now? Yes, yes. Gathering and analyzing big data sounds very expensive. So the question said, is something doable for all countries? Is something doable for? For all countries that don't have money to do that or not the same resources as others? Yes, yes. So the thing is that we gather the data. I mean, we gather the clinical data and we gather also the genetic data. But we think that it's sometimes or always is expensive because it's not so easy to get the clinical data and it's not so easy to get the genetic data. So the thing is that we compensate for the effort that the hospitals or the registry or the working groups are making to get the data. So depending on the quality of the data that you offer me, you offer harmony, we are going to give you money. So we know that there is a lot of effort behind the gathering of data, behind the collection of the data. So we want to compensate for it. So I think it's something that it can be done in in more or less in all countries or in all groups because clinical data is not so easy to get the clinical data. It's time consuming and it's difficult. But for the genetic data and for other indicators, we can also compensate or give some money for each record. Thank you very much. The next question. After a harmony project has finished, will data continue to be added to the dataset? Yes. The thing is that harmony project is a five-year project. So it's from 2017 till the end of 2021. So after that, the data is going to be kept in the database for 10 years. And also we are working on a sustainability plan to get this, to have this. And also that's why we have the second chapter of harmony, the Harmony Plus. But the Harmony Plus will have benefits, will take some advantage from Harmony, from the database, big database platform, which Harmony has already built. And this Harmony Plus is going to be for three years more. So yes, the data is going to be kept longer than the life cycle of the project. Well, thank you very much. The next question. How do you participate in data in this project? How do you ensure the privacy of data in the line? Okay. Yes. This is a very good question. And yes, this is something that we were really worried about. And as you all probably know, the GDPR started in May, two years ago. And this law was designed to protect privacy. So one thing that we were really, really worried was the security of the big data platform, because you know that this data is private. So it's not nice. It's not good to share with anybody. So the first step that we asked for is that if you as a data provider, you are sending us the datasets, we don't want any record that can be identified. So it has to be anonymous. For example, we don't want demographic data. So you have to take out all the data, all the records that can identify a patient. So this is the first step. So the data is anonymized from the origin. So it means that the data provider has to give us, has to give to Harmony the anonymized data. But as we were not satisfied with this first anonymization process, so Harmony, we are not receiving the datasets, but you as a data provider are sending the datasets to an honest broker, to another another figure. And this intermediate person or enterprise has to anonymize it again. So the data is anonymized at the origin. The second step is again anonymized. So it gives you so you can, you know as a data provider, you know your data. But once you give the data to the honest broker, the honest broker anonymized it again. And then is that in this third step, that the honest broker gives us to Harmony the data. So Harmony is completely blind and we can't identify the patient. We can't identify the data we receive. And this completely anonymized data is uploaded into the big data platform. Thank you very much for your answer and the last question, how in our pharmaceutical companies provide their data to the Harmony project? Can you hear me? Yes, now I can, I can, yes. Okay, well the last question, how in our pharmaceutical companies provide the data to the Harmony project? Is it to get data from them? Okay, so the pharmaceutical companies have the data from clinical trials or for the trials they have made to test their drugs. Okay, so in fact, it was not so easy at the beginning, but I think that they have understood that it's important to share the data. So, yes, they are, they are more in favor now in saving the data. And for example, we have data from Novartis, from AML. We can, we have also data from Seljín, Taqueda. So it's, I think that it's a slow process or it has been a slow process, but as Harmony has been going on, it's been more easy. Thank you very much. That was the last question. Thank you Rana for this interesting webinar. Just remind you all that this webinar has been recorded and will be available in the MBE website, which is www.MBE.gov course, and also now with you to try it. So thank you all for joining us today. Thank you, Rana, for your talk and have a nice evening. Thank you very much. Thank you very much, Anna. And thank you all for attending. Thank you very much. Have a nice evening.