Industry Leaders Share Top Trends in Clinical Research for 2025

Das: Effective and well-designed clinical trials is essential for the commercialization stage, and particularly when it comes to making sure that the drugs and therapies, as designed and commercialized, are getting to the right patients, which means the links between the clinical stage and the commercial stage, that's becoming stronger, and the gap between those two things are becoming much less. The first area where we are going to see a lot of investment rightly, is in data strategy and in integration of data itself. First and foremost, the industry has to be prepared for increasing volumes of real-world data and real-world evidence getting integrated in a real-time process into measuring the effectiveness of trials, as well as adapting or modifying them. We're definitely going to see a lot more on data standardization. That's actually a problem. Before you can do AI (artificial intelligence), you need to get your data into order, check the quality of it, the governance of it, the updating of it, and so forth, so making sure that that organizations have enough data infrastructure to handle that sheer diversity of data sources, which, by the way, we see every day, the diversity of data sources, the unstructured nature of it, the non-standardization of it, can be a really, really huge problem. Then in 2025 I predict we're going to see a lot of investment, of course, also in patient diversity and inclusion, where you're strengthening strategies for diverse patient recruitment, looking at more decentralization so you can reach people where they're at, you want to reduce the barriers that a typical patient will face in getting to a trial and adhering to that regimen through the length of the trial. You're also going to see a lot on building relationships with underserved communities, and I think this is important in working with local health systems, with nonprofits to access the communities which might have a disease prevalence that isn't reflected. Last, but certainly not the least is regulatory preparedness. Regulations are getting more complex and more prescriptive and more demanding, but monitoring FDA guidance on novel trial designs and DCT approaches, staying current with international regulations, so that you have a very smooth commercialization process, as well as data privacy. As we use more AI to analyze data that has PHI and does not, making sure that you're able to actually align with those data privacy requirements, which is essential for trust is going to be very important. I could probably go on, but these are some of the key items that I expect to see a lot of progress being made on.

Jadhav: I would say, as we move into 2025 there's two generic themes that I see in terms of how real-world evidence and real-world data is going to be utilized in clinical research. I think one is around the data and expanding the type of data that we're using for clinical research. The second one is what I like to bucket it under, helping productionize, using external/synthetic control arms in the overall process. When you're looking at the data side of the house, historically, a lot of the data that has been used for clinical research, a little bit more mainstream, has been more structured data, seeing a good movement into leveraging unstructured data. There's different types of unstructured data, things such as physician notes, as an example, is an unstructured data source that we are leveraging in terms of helping facilitate improved clinical research. There's also other unstructured data sources, different data sources like imaging data as well, and we're starting to use, and see usage of imaging data quite a bit in terms of helping identify patients with certain diseases, understanding their disease progression as well, and then rolling that into clinical research. The good thing about imaging data is it's like an unbiased data source. From that perspective, the credibility, the quality of including that is very, very good.

The second bucket is around external control arms, synthetic control arms. It's been around for quite a while, overarching use of this for the particular components of the clinical trial where they're actually not getting the treatment, so the non-treatment arms are leveraging real-world data to understand and compare and contrast to the treatment arms that you put forward there. Historically, recruiting patients where you're not providing them the treatment, as you can imagine, is very, very expensive and very time consuming overall. I think that's the area where you look at synthetic control arms, and leveraging real-world data can help be a good proxy, or can help augment and allow you to compare and contrast to the treatment arm. The good thing with real-world data and leveraging it for these type of control arms is that the volume is very, very large, and the quality and the longitudinality of the data is pretty good as well. If anything, the compare and contrast, the treatment arm versus a non-treatment arm can actually be a high-quality fashion as well. These are some of the key trends that we're seeing as we move into 2025 in terms of helping improve clinical research by leveraging real-world data.

Tableman: I think one of the things we're really seeing right now is related to data interoperability, so we have historically had a set of data standards, particularly getting data submission ready from a CDISC perspective, but we have also now seen with the FDA putting forth the guidance on the ICH M11 standardization around the data for protocol collection as well, so that data, and then there's an opportunity, really, from an interoperability perspective, to map all the way through, so you have the protocol now in the ICH M11 format, you can map that to CDISC. We've done that as well as USDM, and we have for the schedule of activities, mapped all the way through to this, to the actual SOA itself, so the ability to then, from a FHIR perspective, be able to share that data, which this has all been things that the industry has been talking about for a number of years, and it is now finally coming to fruition. The benefit of that data interoperability is supporting generative AI. It's really allowing you to use data in downstream systems to really automate that digital data flow. Again, stuff that we've been talking about but haven't really been able to fully recognize, I think we're going to see that happen in 2025.