Current State of OAI & Intro to Knowledgebase

Hi everyone. Thank you all so much for taking the time out of your busy schedules to attend our first live webinar hosted by the OAI CORE Knowledgebase. Today we're joined by Dr. Jeffrey Driban. Dr. Driban is the director of the OAI CORE Knowledgebase and professor in the department of population and quantitative health sciences at UMass Chan Medical School. His research explores novel risk factors and measures to facilitate more efficient clinical trials and a better understanding of osteoarthritis and its potential subtypes. Dr. Driban also aims to raise awareness about osteoarthritis and promote primary and secondary prevention strategies. Today, Dr. Driban will be talking about the current state of the OAI's data usage and an introduction to the OAI CORE Knowledgebase. Following his presentation, we will have a few minutes for Q&A. So, if you have any questions, feel free to put them in the Q&A box at the bottom right of your screen. And now, without further ado, I will let Dr. Driban take over. Thank you, Julieann. And thank you everybody for this opportunity and for attending this webinar.

Today, what I'd like to do is offer you an overview of the Osteoarthritis Initiative, especially for those who may be less familiar about the cohort, describe a little bit about the impact that the osteoarthritis initiative has had over the last 20 plus years, and then describe the OAI core knowledge base, who are we and what services are can we offer to you.

The Osteoarthritis Initiative was a public and private partnership funded by the NIH and private sector partners including Fizer, Merc, Glasosmith Klein and Novartis. The cohort had three overarching goals. First to enable a better understanding of the prevention and treatment of neostarthritis. Secondly, to support the investigation of the natural history of and risk factors for neostarthritis onset and progression and thirdly to determine the validity of traditional measures, novel biomarkers and surrogate endpoints.

To achieve these goals, the OAI investigators and staff recruited almost 4,800 participants between 2004 to 2006. The participants were 45 to 79 years of age at baseline and recruited from four locations in the United States. The investigators and staff purposefully recruited people from in into three subcohorts, the largest of which had about 3,300 participants without symptomatic radiographic tibal osteoarthritis in either knee, but had an increased risk for developing symptomatic neoi. The next cohort was about 1,400 participants with symptomatic radiographic tibmoral osteoarthritis in one or both knees. And then finally, there was a group of individuals who had neither symptoms nor radiographic OA in either knee nor o risk factors. And it's important to consider these subcohorts when thinking about the eligibility criteria that were used to select people and how that may influence your study design or analyses. But it's also important that when you're selecting people that you consider essential readings and other variables and that is something that the knowledge base can help you think through. As I mentioned, the Osteoarthritis Initiative started screening and recruiting participants between 2004 to 2006. The cohort was initially funded to have followup through the first 48 months with interim visits for a subsets of individuals at the 18 and 30-month visits. At the 48-month visits, almost 90% of the original cohort attended that visit. The OAI was then uh initially extended through the 96-month follow-up visit with telephone interviews at 60 and 84 months and in-person visits at the 72 and 96-month visit where about 76% of the original participants attended. After the 96-month visit, a series of grants supported the osteoarthritis initiative through the year 16 follow-up telephone interview when about 51% of the original cohort completed at least part of their interview. The OAI staff and investigators last engaged with participants in May of 2022.

This longitudinal cohort offers a unique resource with a lot of types of data including patient reported knee symptoms and function, other joint symptoms. They answered questions about general health as well as risk factors, health behaviors such as physical activity and psychosocial measures. They also completed physical performance assessments such as a 20 meter walk test. They had knee and hand examinations. They were asked at in-person visits to bring in their medications to complete a medication inventory form and answered questions about medications and supplements. They also provided biopecimens at visits up through the 72-month follow-up.

At each annual visit, um they had three Tesla MRI of their knee and thighs as well as knee x-rays. And the cohort also has adjudicated outcomes for knee and hip replacements and death has also been recorded. All this data is publicly available in a multimodal data ecosystem that is represented by three repositories that help create rich linkable data sets. The one that most people engage with is probably the NMIH data archive or NDA OAI website. This is where you can go to get the clinical data. You can download images from there. And in addition to the clinical data, there are also data from nested studies, many of which include outcome measures and bio other biomarkers. And you can see a link to these uh repositories along the bottom with the QR codes. And as a reminder, these uh slides will be shared through our video after today's webinar. The second repository is dbgap which is where you can find the genomewide association study data from the OI. And then thirdly there is a biopecimen repository where you can submit an application to request access to plasma serum and other uh biopecimens that are available up through the 72-month visit.

In addition to the imaging that was completed, there are also publicly available imaging assessments uh available on the NDA's OAI website. For example, for the knee radiographs, you can download semi-quantitative scoring such as Kellgren Lawrence grades. There's quantitative joint space width measurements at fixed locations and from moral tibial angle as a measure of static alignment. There's also publicly available uh semi-quantitative scoring for hips and the full limb films were measured for hip knee ankle angle which is a representation of static alignment where the 3T knee MRIs some subsets of the OEI have semi-quantitative scoring available typically it's the MOAKS scoring but some participants also have data using bloks or worms which are other common methods of scoring the MRIs. In addition to the semi-quantitative, there is also quantitative measurements of cartilage thickness and morphology for some participants, bone shape and trabecular morphology. Another two subsets of the OAI also had knee and hip dexa which allow for the reporting of femoral and perarticular tibial bone mineral density.

The biopecimens have also been used and some data is been released including the genetics data from the G-W was study. There's been a panel of OA related biomarkers completed mitochondrial hloypes telomer length from peripheral blood glucose sites vitamin D levels and intact parathyroid hormones have been released. It's also worth noting that the biopecimens have been used quite a bit and there may be other essays that have been completed but not included in the public release which is something that the knowledge base can help with by help helping you engage with the investigators who have used that data.

Now, one of the most productive and well characterized subsets of the OAI is a nested case control study of 600 people that was conducted by the FNIH OA biomarkers consortium. They completed extensive imaging and biopecimen assessments at baseline, 12 month, and 24 months. And I hope you'll join us next month when Dr. Dr. Jamie Collins will be talking about this unique um study and its data and how to use that data and how to use it along with the other OAI data. I hope you got a sense from those initial slides that the OAI really represents a valuable resource to the research community and it's had a significant impact in research. OAI data has contributed to more than 1300 publications. Since 2017, OAI related um papers have accounted for more than 100 publications per year. This is attributable to over 770 researchers who have published multiple times using OAI data and publications have come from more than 60 countries with and in this figure you can see yellow and green representing the countries that are uh been publishing the most with data from the osteoarthritis initiative.

It's had significant impact in some key areas including understanding risk factors in the natural history of osteoarthritis. Some of the other most cited papers have covered topics such as health economics, helping inform cost effectiveness analysis for OA treatments. In more recent years, we've seen a lot more papers describing the use of artificial intelligence with models to predict OA outcomes, assist with diagnosis, early detection, and image segmentation. The OAI also have been highly cited for its work in health and behavior, linking inactivity, sedentary behavior, and physical function in adults with knee osteoarthritis. And in alignment with the third goal of the Osteoarthritis Initiative, it's played a valuable role in biomarker identification and qualification. OAI data and images have been the basis of six of the seven recent FDA biomarker qualification submissions for osteoarthritis and two biomarkers have received approval for their qualification plan largely based largely on OAI data and resources.

Well, data collection has concluded for the OAI. Its potential for discovery is greater than ever because of the growing amount of information and data and resources that are available to the community. The next chapter is about empowering a wider research community to leverage these unique resources.

But we appreciate that there's a challenge to using the OAI. There is an immense amount of data and complexity to using the OAI resources that can sometimes be daunting to new users. On the one hand, it's fortunate that the OAI coordinating center staff and investigators created extensive work in documenting and harmonizing the data which sets us up um to use this great resource. But we also appreciate that it can be daunting to a new user to look at the multiple data sets that are available for download in the documentation. But on the other hand, it allows us to also help you even in this next chapter. So we acknowledge that there is a knowledge barrier to accessing the data and a nuance required to fully use the OAI resources which has led to underutilization in certain global regions and some academic disciplines.

A solution to this is a biomedical knowledge base which the NIH defines as a resource that extracts, accumulates, organizes, annotates, and links growing bodies of information related to and relying on core data sets. In our case, the osteoarthritis initiative. So in April 2025 the NIH funded the OAI collaborative osteoarthritis research enterprise or OAI CORE Knowledgebase. Our mission is to promote and facilitate the use of the OAI to accelerate discovery with a commitment to using fair data principles to achieve these goals. The OAI core faculty has over 15 years of NIAM's funding to use OAI resources with over 100 OAI-related publications representing almost 10% of OAI public related publications. Our faculty also includes expertise in community engagement, informatics and fair data principles. Now our faculty is also supported by two advisory committees. The first is the external advisory committee. These are faculty with extensive experience with the osteoarthritis initiative including investigators who have been involved with the OAI since its inception including members from the coordinating center and clinical sites. We also included people with in experience in community engagement and fa fair data principles. This committee helps us evaluate and advises us on providing quality services with transparency, accuracy, and efficiency. It also helps ensure that we meet our short and long-term goals and the research community's needs.

Our second committee is the community advisory committee which represents investigators around the world both early career and established uh investigators that have used the osteoarthritis initiative in the past or may be interested in using the osteoarthritis initiative in the future. This committee advises us on optimizing strategies to provide resources and services to meet the community's needs. They will also review user feedback and the results of each service uh to provide direction on how to incorporate findings into the services. So for example, after this survey, we will ask you to complete a survey about the webinar. We'll summarize those results and pre present them to the community advisory committee for feedback.

I hope now you have a little bit of a sense of who we are and I'd like to transition to what we can do to help you. On this slide, I just want to offer a big picture that we offer two types of services and on the next two slides I'll dive into these with a little bit more detail. The OAI Knowledgebase has two categories of work that we can offer. One is our online OAI resource library. These are the resources that you'll be able to access through the website without needing any interaction with the OAI core faculty. On the other hand, the OAI engagement services is where you would directly engage with an OAI CORE faculty member to achieve a goal. So, if we take a closer look at the resource library and the QR code on the bottom can help you access that page if you wanted to see it. includes things like our OAI archive, which will be a repository of foundational documents and other resources that future investigators may need from the OAI coordinating center and clinical sites. We also will have our educational ar aspect which is called oai.edu which includes presentations such as today's webinar and other educational opportunities to promote OAI resources and describe cutting edge research leveraging OAI data and that'll include the recording of today's webinar as well as shorter uh videos like two five 10 minute videos on how to access or use the other rep to use the repositories. The knowledge base will also issue an annual report that we call the state of the OAI which will describe the state of the OAI literature including novel uses, weaknesses we observe in reporting new investigators and to establish a baseline in the fall of 2025, we released an initial state of the OAI report which I hope uh you'll take time to check. In the next year or so, we're hoping to launch a few new features. one of which is the OAI explorer which is actually two different toolkits. The first is one that will allow you to calculate basic descriptive um characteristics of the OAI. So for example, if you wanted to look at education level by OAI clinical site, you can put those two variables into it uh our form and it'll offer you a cross tab table that shows you the distribution of education across sites. And if you wanted to, you could also stratify that by a another variable such as you might say just among people over 65 years of age. The second OAI explorer feature will be an OAI variable finder where you can type in something of interest such as walking and it'll report back all the different types of variables related to walking whether it's the walking test uh pace or pain with walking and it'll then show you which visits that data is available for at. And then you can go to the NDA OAI website, download the data and know where to look for those variables. We also hope to create an OAI biopecimen database which will allow investigators to see how much biopecimen is left at each visit for each participant. And we hope that that'll facilitate greater use of the OAI biopecimens. And then the last part of the resource library is the OAI learning lab. This is something that we've done internally but we're hoping to share which is small part uh data sets that are for teaching purposes that can be used in classrooms such as advanced epidemiology and machine learning and we'll offer a very parsimonious data set with some documentation to accompany it about what results you would expect. this we will expect people to complete an NDA data use certification for the faculty member and the students and we'll be uploading a video in the near future showing people how they can do that to help them

on the engagement services side of things we have two broad categories of services one is our OAI core office hours and the other is the OAI connector during OAI CORE office hours this is an opportunity where people can email us or schedule a time to meet with us to discuss available data, data limitations, and how to handle missing data. We'll also offer the assistance with creating parsimonious data sets and to a limited extent, we can also assist you with performing the analyses. Another service that we often do for ourselves that we're going to start offering to the uh research community is the ability to create parsimonious imaging sets based on specifications that are outlined. These services will require you to complete a data use certification through the NDA and we can assist you with that process. We're also offering the ability to assess OAI images with quantitative and semi-quantitative methods. This is one service that is a fee for service because it is a bit more of an extensive amount of hours of work on our team. We'll also assist people with developing imaging protocols for future studies based on the OAI protocol. We can also assist you with preparing your Brack applications so that you can um request biopecimens. Our team has experience with submitting multiple applications and I've reviewed these applications so we can help you prepare them as well as make the list of IDs that you need to submit with your application. We're also going to reach out to people who have published with NIAMS using OAI images or OAI biopecimens. to see if we can assist with organizing, annotating, and making data sets publicly available and harmonious to link with existing publicly available OAI data. We also hope to coordinate with journal editors in the near future to foster more reviewers with expertise in the OAI. Now for the OAI connector, this will be a searchable database where you'll be able to look up um who is doing what research with the OAI in the past. So for example, if you wanted to search for who has done stuff with hand OA in the using OAI resources, you could type in hand osteoarthritis and it will offer you a list of investigators who have published or been funded in that area in the past. At that point, you can reach out to them directly or if you'd like, the knowledge base can facilitate an initial meeting with that investigator and we'll offer to compensate them as a consultant for two hours to meet with you as an interested investigator. After that initial OAI connector session, we'll allow you and the other investigator to decide whether or not to pursue a new collaboration together or go your own way. We're also hoping to launch in the um the coming year meet the professor sessions where new users can meet experienced OAI investigators and we hope that this will be a catalyst to facilitate collaboration and new research.

Now I wanted to as we start to wrap up highlight the unique role that the knowledge base serves within the existing ecosystem for the OAI. The repositories are your go-to places to access the OAI resources. The knowledge base is available to make it easier to use these resources. We're here to help you. If you experience a or a problem or a challenge using a resource, let us know. We can help you troubleshoot it. If you have questions about OAI data or OAI resources or repositories, let us know. If we can't answer it, we'll reach out to the appropriate people to get the answers you need. The contributions of 48 almost 4,800 participants, the OAI investigators and staff have truly created a resource of immense potential. The OAI core knowledge base is committed to serving the research community to reduce barriers, foster the collaborations that will unlock the next wave of insights into OA aging and human health. Before I go to my last slide, I also just want to acknowledge that there's opportunities to use OAI data with other cohorts. And one of the cohorts that people often will work with with the OAI data is the multic-enter osteoarthritis study. And they're currently accepting um, grant applications right now that are due in April 6. And it's something to consider that you could potentially propose a project with the multic-enter osteoarthritis study and the OAI to increase sample sizes or to offer an opportunity to do some external validation. At this point, I'd like to thank all of you for your time. You can see on screen the link to our knowledge base, the our email, our LinkedIn page, and I hope you'll join us next month when Dr. Jamie Collins will present on working with data from the FNIH OA Biomarkers Consortium. And the QR code on screen will take you to the registration page. Thank you everybody.

Hi, thank you Dr. Driban for your presentation. We have a few minutes for Q&A. So we have a question here. Are there any biome specimen left from the baseline visit? There are some, but at this point what the um application is asking people to do is to list backup samples just in case they are running out. But that has been the visit that's probably been used the most and that is one of the things we're hoping with the biopecimen database is to kind of help people better understand which specimens are left. Another question does the OAI CORE charge for downloading databases? We don't charge for we are not the place to go for downloading a data set. If you want to access the OAI database down download the OAI data you go to the NDA website if you want to download all of your clinical data. What we can assist you with is if you need help making a more parsimonious data set for your project, we can meet with you and discuss the logistics of what variables and data you're looking for. And we can help merge the proper data sets together, make a smaller data set for that. And for that type of service, we do not charge a fee. It's part of what we're funded for by NOAMS. The one expectation on our part is that you would still go to the NDA website and complete a data use certification, which if you need help with that, we can do it while we're meeting. We can take you to the NDA website, show you where you log in, how you access the data use certification, and once that's been completed, we can ensure short delivery of that data set per your specifications.

Thank you. I think that's just about all the time we have today. So, thank you again Dr. Driban for your very insightful talk and thank you all for coming to our very first live webinar. We really appreciate you taking the time out of your day to come and we hope that you leave today with some useful information about the OAI and our knowledge base. And we'd also appreciate if you take the time to fill out a quick uh survey. I sent a link in the chat. Yeah, so that concludes today's webinar. Please look out for our upcoming webinar next month with Dr. Jamie Collins. And thank you all. Have a great rest of your day. Thanks, everybody.