The Scope of Things
The Scope of Things
Episode: 39 - Blythe Adamson on Patient-Level Real-World Data for Multinational Oncology Research
In this episode of the Scope of Things, host Deborah Borfitz brings you the latest news on AI-recommended precision dosing, organoid drug testing aiding treatment selection for bowel cancer, an AI tool for stratifying lung cancer patients, using HIV drugs to treat Alzheimer’s disease, and the potential value of magic mushrooms to remedy the mood symptoms of Parkinson’s. Blythe Adamson, international head of outcomes research and evidence generation at Flatiron Health, also joins in to discuss groundbreaking work harmonizing patient-level real-world data across four countries to enable multinational oncology research.
News Roundup
CURATE.AI platform
FORECAST-2 clinical trial
- News on the Walter and Eliza Hall Institute website
AI tool for sorting cancer patients
- Study in Nature Communications
HIV drugs for Alzheimer’s protection
- Study in Alzheimer’s & Dementia
“Magic mushrooms” for Parkinson’s disease
- Study in Neuropsychopharmacology
Guest
Blythe Adamson, Ph.D., international head of outcomes research and evidence generation at Flatiron Health
- Flatiron Health enabling multinational oncology researh – article in Bio-IT World
Subscribe to the podcast on Apple Podcasts and leave a review to support the show. Join us at Scope Europe on October 14-15 in Barcelona - use code SOT10 for an additional 10% discount.
GUEST BIO
Blythe Adamson, PhD, MPH, Head of Outcomes Research and Evidence Generation, International at Flatiron Health & Founder of Infectious Economics
Dr. Blythe Adamson is the head of international outcomes research and evidence generation at Flatiron Health.
As a visionary senior leader at Flatiron Health, her team pioneered deep learning language models for extraction of clinical details from EHR documents, breaking the limits of what was possible for humans to do alone. Learning from the experience of millions of patients with cancer, they generate evidence of treatment effectiveness and value used by governments around the world. Dr. Adamson co-invented a patented clinical decision-support tool, enabled by machine learning, that is used by cancer clinics to benefit patients.
She holds degrees in microbiology, epidemiology, and pharmaceutical economics with a focus on infectious disease prevention. Dr. Adamson has held roles at the Bill and Melinda Gates Foundation Institute for Disease Modeling, the NIAID HIV Vaccine Trials Network, and Flatiron Health.
The Scope of Things podcast explores clinical research and its possibilities, promise, and pitfalls. Clinical Research News senior writer, Deborah Borfitz, welcomes guests who are visionaries closest to the topics, but who can still see past their piece of the puzzle. Focusing on game-changing trends and out-of-the-box operational approaches in the clinical research field, the Scope of Things podcast is your no-nonsense, insider’s look at clinical research today.
Hello and welcome to the Scope of Things podcast, a no-nonsense look at the promise and problems of clinical research, based on a sweep of the latest news and emerging trends in the field and what I think is worthy of your 30 or so minutes of time. I'm Deborah Borfitz, senior Science Writer for Clinical Research News, which means I spend a lot of time with my ear to the ground on your behalf and a lot of hours every week speaking to top experts from around the world. Please consider making this your trusted go-to channel for staying current on things that matter, whether they give us hope or cause for pause. In another five minutes or so, I'll be talking with Blythe Adamson, international Head of Outcomes Research and Evidence Generation at Flatiron Health, about its groundbreaking work harmonizing patient-level real-world data across four countries to enable multinational oncology research. But first the latest news, including AI-recommended precision dosing, organoid drug testing, aiding treatment selection for bowel cancer, organoid drug testing, aiding treatment selection for bowel cancer, an AI tool for stratifying lung cancer patients. Using HIV drugs to treat Alzheimer's disease and the potential value of magic mushrooms to remedy the mood symptoms of Parkinson's.
Deborah Borfitz:Researchers in Singapore have built a platform known as Curateai that creates clinically actionable patient profiles, or digital twins when given data on only two blood biomarkers together with drug type and dose. This could aid precision dosing for multiple conditions, including patients with advanced solid tumors. Per a small interventional clinical trial, finding a high rate of user adherence to the treatment recommendations. For some patients, the AI-proposed doses were lower on average than those suggested by standard of care guidelines. Beyond cancer, this dynamic dose adjustment approach has now been used in trials for hypertension, immunotherapy and cognitive training. The latest feasibility study will expand into larger randomized controlled trials to validate the effectiveness of the Curateai platform against traditional treatment regimens and factor in savings to the patient and healthcare system, in addition to the usual efficacy and patient outcome measures.
Deborah Borfitz:In Australia, meanwhile, researchers have launched a game-changing Forecast 2 clinical trial where tumor organoids many cancers grown in the lab from a patient's own tissue samples are being used to predict what drugs will work for newly diagnosed cases of bowel cancer before treatment begins. The findings could help replace current trial and error treatment practices with a more tailored and personalized approach to improve survival rates and quality of life for patients. Researchers previously validated organoid drug testing as an accurate tool in the treatment selection process for patients with advanced bowel cancer. The new trial will look at whether those results can be replicated in people who have recently been diagnosed. In collaboration with Regeneron Pharmaceuticals, investigators at Weill Cornell Medicine report on a new AI-based method for sorting lung cancer patients into groups that have similar characteristics before treatment and similar outcomes after treatment. The approach predicted outcomes from health record data on patients treated with immune checkpoint inhibitors and outperformed other methods published to date. The machine learning algorithm was trained on the de-identified health records of over 3,000 patients with lung cancer in a commercial database and sorted them into three groups based on survival time from the start of treatment. When applied to the new, smaller data set of lung cancer patients, it yielded almost identical groupings. The AI tool could find utility for patient stratification and clinical trials of new pharmaceuticals, as well as individual treatment selection.
Deborah Borfitz:New research indicates HIV drugs called nucleoside reverse transcriptase inhibitors, or NRTIs, could offer protection against Alzheimer's disease, based on the discovery that patients taking the medications are substantially less likely to develop the memory-robbing condition. An analysis of two large health insurance databases found the risk of developing Alzheimer's decreased every year by 6% and 13% respectively in patients taking NRTIs, even after adjusting for factors that might cloud the results by that yardstick. Taking the drugs could possibly prevent about 1 million new cases of Alzheimer's disease worldwide every year. The drug appears to prevent the activation of important immune system agents known as inflammasomes. The research team now plans to test its own inflammasome-blocking drug in upcoming clinical trials to assess its preventive power against Alzheimer's disease.
Deborah Borfitz:And finally, in a small pilot study, researchers in California have found that magic mushrooms, previously showing promise in treating depression and anxiety, could also help Parkinson's disease patients.
Deborah Borfitz:Psilocybin therapy meaningfully improved mood, cognition and motor symptoms and was well tolerated by study participants and, remarkably, the beneficial effects lasted for weeks after the drug was out of their system. It is the first time a psychedelic has been tested on patients with any neurodegenerative disease. Researchers point out that mood symptoms and Parkinson's are linked to faster physical decline and are a stronger predictor of patients' quality of life with the disease than the better-known motor symptoms. A larger randomized controlled trial, now underway at UC San Francisco as well as Yale, aims to enroll 100 participants and incorporate non-invasive brain stimulation, neuroimaging and other tools to understand how psilocybin impacts inflammation and neuroplasticity. As a reminder, links to the various articles and studies informing this news segment can be found in the show notes. It is now time for our chat with Blythe Adamson about the harmonized cancer-specific datasets created by Flatiron Health, which researchers anywhere in the world can now analyze in a secure, cloud-based, trusted research environment. Welcome to the show, Blythe.
Blythe Adamson:Thank you for having me, Deborah. It's great to be here.
Deborah Borfitz:The cross-border patient-level data sharing exercise we're talking about here was, by all accounts, a monumental feat because it involved four countries the US, of course, but also the UK, germany and Japan all with different regulations and privacy standards and all having different places for, and practices around data collection. I know it took a while to convince the scientific community that this was even possible. So take us back to the early days, if you would, when you were on the conference circuit sharing the news. What was the initial reception to the data sets and what did it take to flip the conversation from disbelief to the potentially enormous benefits to cancer researchers and patients?
Blythe Adamson:Well, when I first started presenting research studies about, for example, colorectal cancer in Japan having earlier age at onset than we'd ever seen historically, or looking at breast cancer in the UK and Germany, I thought that the people coming up to ask questions about the poster or our oral presentations would be wanting to know about the epidemiology of these diseases or the statistical methods that we used. But everyone first came up and said I can't even believe it. I didn't think data like this existed. I didn't think that it was that the regulations and laws within Europe and Japan would allow this to be possible, and so it was just so interesting to me because I had really prepared to talk about deep methods and epi. But really what people wanted to understand was how did you build a global healthcare data infrastructure in a way that was compliant with, we know, our stricter data privacy laws within these different markets and, honestly, it took Flatiron more than five years to navigate through this. Yes, exactly, to find the way to do this so that patient privacy is completely respected, that the processes of de-identification of the electronic medical record data or the anonymization of the records. It's now done locally within the country and then uploaded into a secure, compliant environment.
Blythe Adamson:So this has just been a key turning point, because previously, what people had been most familiar with was these federated data models, which is where Data is not allowed to leave the country, and so you have to, for example, do one analysis in Germany, one analysis in Spain, one analysis in France, and then you just print out the results and try to smush them all up together and they often just don't completely make sense, and this is now what technology has been able to overcome. So it's a really exciting moment and the way that I then had to step back and say okay, if you're not ready for me to show you research results yet, how can I figure out a way to more effectively communicate this? And it really came through the ISPOR EHR-derived data suitability checklist that's available and published by a task force from this professional society, and it really gives you a checklist to go through to ensure whether or not EHR data is fit for purpose for different research questions, regulatory decisions and it goes through how to characterize the data relevance, reliability, governance, provenance, relevance, reliability, governance, provenance. And so this is an extensive peer-reviewed paper that I wrote with my team and have now published to be able to go through every single item in the checklist to transparently communicate.
Blythe Adamson:Where does the CHR data come from? How is it governed? How is it kept secure? How is it curated so that people can see? Oh wow, we've never had anything like this before outside the United States.
Deborah Borfitz:And you just mentioned data curation. I want to deep dive on that just a little bit, because I think that's part of the distinctiveness of this approach here. Could you talk a little bit about that, how that was handled? I think our readers might, or our listeners might, appreciate hearing more about that.
Blythe Adamson:Right. So in contrast to other types of real world data that sometimes they send physicians surveys that say, okay, tell me about a patient you've had with multiple myeloma, when did they get diagnosed, what treatments did you give them we are integrating at the source of the EHR. So imagine, you know, installing servers and basements of hospitals and clinics across Germany where we really are getting real-time access and integrations to all of the clinical notes, the pathology reports, all this non-standardized documents. This is really where a lot of the treasures are in oncology. It's not in the structured billing data. The richness comes from these patients' stories that are often described in paragraphs of notes that these doctors are writing about them.
Blythe Adamson:So we do this longitudinal curation. So when we integrate, we curate and understand the whole retrospective history of these patients' experiences and prospectively follow them forward to start continuing to capture their experiences and following them over time. So, with regular refreshes, at least once a quarter, we're updating these data sets so that we can have real-time information, nuanced analyses that have the clinical details with provenance, where you can trace all the way back to the source of the document, the line number. You know, the place where we can verify this information is found.
Speaker 3:Are you enjoying the conversation? We'd love to hear from you. Please subscribe to the podcast and give us a rating. It helps other people find and join the conversation. If you've got speaker or topic ideas, we'd love to hear those too.
Deborah Borfitz:You can send them in a podcast review. So what are the clinical research opportunities you're most hoping to see pursued with the integrated data sets, and have you seen any movement in those directions? Absolutely.
Blythe Adamson:You know, interestingly it was the FDA who started asking for some of this information, rather than what you might expect is the EMA, and it came because, out of the increasing use of external control arms, so often for a single arm clinical trial that might be global, they may do a control arm that comes from real-world patient experiences, and previously many of those control arms were comparing real-world patients in the United States to a global clinical trial.
Blythe Adamson:But we know that there's a lot of variability in patient characteristics and the treatments used and healthcare systems between countries.
Blythe Adamson:And so the FDA had said wouldn't it be better if the real world external control arms had a more diverse and representative composition that's more comparable to the global single arm trial?
Blythe Adamson:So I'm already seeing the adoption and use by many life science companies in designing global external control arms to compare them to their global single arm trials, and we're also seeing this in comparative effectiveness studies within these different countries. So one of the new initiatives is the joint clinical assessment in Europe where regulatory and HTA decisions are happening together. They're wanting to enumerate what people call PICOs the population intervention comparator outcomes that are relevant for all the different countries in the European Union, and the dilemma they often have is the local standard of care in Latvia might not be something that's even used in the United States, and so we're using these curated data sets in UK, germany and Japan to surface more patients that have unique experiences and treatment pathways that we can learn from that may be very different from the United States, so it's really exciting to see the quick adoption and range of use cases that life science companies are now taking advantage of with global EHR-derived data.
Deborah Borfitz:How do researchers go about accessing these data sets and getting into this trusted research environment to do their analysis work, and are there any you know caveats we need to mention around that?
Blythe Adamson:Right, as I mentioned before, many are familiar with that federated model where you have to do separate analyses in every country, and it's such a relief to me now that we've technically overcome that limitation.
Blythe Adamson:So trusted research environments now are the way where, as you mentioned, it's cloud-based, compliant, secure, and this is where we upload the de-identified or anonymized EHR oncology data sets so that they can be pooled together at a patient level. So if you imagine a table that's describing the demographic characteristics with one row per patient, you can now pool together and combine all these different countries into one table, and that enables lots of different types of more rigorous methods for analysis. And it allows you, within the Flatiron Trusted Research environment for researchers, to use familiar tools like coding in R or Python within these environments, really being able to touch and visualize the patient level data within the trusted research environment. And then, once you have the results from your studies where you've pooled all the countries together, then you can export the results, so download those to your local computer to be able to use in publications. Regulatory submissions and this is the step that we use to ensure the data security and compliance with regulations like GDPR in Europe and AAPI in Japan.
Deborah Borfitz:Now are all cancers represented, or will they be at some point?
Blythe Adamson:Yes, we're increasing the number very rapidly. So I think right now we have eight different cancer types that have been curated and are research ready sitting in this tier trusted research environment where researchers are now asking questions about the effectiveness of different treatments. And the team at Flatiron is quickly adding more and more cancer types into this environment to do analyses and then also increasing the cohort sizes of them very quickly. So we've got a network of Flatiron sites within each of the countries. So, for example, in the UK it is a collection of NHS trusts, which is where cancer care is delivered in the UK. In Germany, for example, we have a huge network of lots of small community clinics, which is more of how cancer care is delivered in Germany. So the number of sites are increasing, patients are increasing and the cancer types that are research-ready to use is increasing as well.
Deborah Borfitz:a great segue to my next question, because we're talking about patient-level data from four countries. So let's get to the topic of transportability of real-world evidence across country borders. That's an area where Flatiron Health has been actively working, I know, with sponsor companies to determine when it is appropriate for one country to kind of borrow information from another. Is this the intention, then, with the data sets that you're developing here, that? Are they geographically diverse enough, or is the plan to repeat the harmonization exercise in yet more nations?
Blythe Adamson:Well, when I speak with regulators and HTA bodies, you know government decision makers all of them have a preference for local data. They would rather have real-world data from their own country to make a decision, but the dilemma is it's often not available, it's too sparse or insufficient for some reason. So they do have this dilemma of would we rather have nothing or do we make the assumption that it's okay to use data from another country to make a decision about cancer care in our country? And it's an assumption that has never been tested before. And so we really had a lot of benefit from doing a formal research collaboration with NICE in the UK, and all these other countries had so much interest in joining us, and many life science companies wanted to work together that we established a research consortium called the FORUM, which stands for Fostering Oncology, rwe Uses and Methods, where together we are doing a huge set of validation studies testing whether or not differences between two countries' healthcare systems is so big and impactful that it actually might change how long people survive with cancer the same patient. You know, if you live in Germany versus Denmark, versus Spain or the US, if you have identical patient characteristics, if you receive the exact same drug. Do you survive the same amount of time? Or are differences in the way nutrition, the environment, how people are cared for in the hospital do those things actually make a difference? And this is what the forum is studying.
Blythe Adamson:So we are going country by country for breast cancer, lung cancer, heme and looking at different combinations of countries to test whether or not.
Blythe Adamson:Once you control for the population characteristics being different, once you select for a specific type of cancer and drug that people are using and you control for everything, we are learning from the experience of patients in the US, predicting the outcomes of patients in a different country and then comparing that to their true overall survival in that country to see whether or not it actually was modified by differences in the healthcare system.
Blythe Adamson:So it's been pretty exciting. I mean, we've learned that in many cases that the real world evidence is transportable between countries. But there are some times where it's really not appropriate, like where we found where, when clinical guidelines between two countries are wildly different, like everyone in the US starts with drug A, then B, then C, but in Europe maybe they start with B, then D, then F. You know, you can't really adjust away when people receive completely different drugs, and so we're learning about what are the situations where you really do need local data to be able to inform decisions in that country referenced when life science companies are submitting to these different countries to support whether or not that assumption is appropriate that it's okay for them to borrow evidence from another country.
Deborah Borfitz:Wow, fascinating, interesting and what a work in progress. Right, we've covered a lot of ground here, blythe, so let me end with this when is this all going? Just how big and robust do you imagine the global data infrastructure will be in, say, another five years?
Blythe Adamson:I am really, really optimistic about this, and when I think about global healthcare data infrastructure, I feel really proud to be working at a private company that's solving this, because it really is bigger than what an individual country can do by themselves, and one of the keys is having harmonized, common data models across these populations and also having respect for true differences between these countries and how cancer care is delivered, how patients are different, how documentation styles are different.
Blythe Adamson:But I anticipate, you know, within five years there's going to be a much larger, more robust global health data ecosystem, and so much of that, I think, is getting super powered by large language models being one of the tools we use now for curation. In the past, so much of this was done manually by cancer nurses opening up charts, pulling out information. Now a lot of that is being used to validate LLMs doing some of those same tasks for us. So I think that it's a really exciting moment in history to be a researcher in this field, because the number of questions that we can answer that were never possible to understand before is just exponentially growing.
Deborah Borfitz:I hear your enthusiasm. This has been a truly enlightening and encouraging conversation, and you have succeeded in making it sound very cool to be a data geek. So I thank you for being on today's show and ask only that you leave us. As I told you in advance, I would do to leave us with your favorite data geek quote.
Blythe Adamson:Garbage in, garbage out. I work at Flatiron because it's the most beautiful data in the world that I've ever touched, and so I am so thrilled to be in a place with high quality data to be able to put out high quality findings.
Deborah Borfitz:You heard it here. A big thank you to everyone out there for listening, and if you are not subscribed to this podcast yet, please consider going to Apple Podcasts and doing so right now, so you don't miss your monthly dose of news and perspectives. You'll be hard-pressed to find anywhere else and, if you're up for it, I'd also be so very grateful if you leave a rating and review while you are there. One more thing before we go. If you liked today's conversation, it is only a glimpse of what you can expect from Scope Europe. Presenters and panelists, please plan to join us October 14th and 15th in Barcelona when clinical operations executives will be exploring the latest trends in clinical trial innovation, planning and operations. Save an additional 10% off any current rate by using the code SOT10. For more information, visit scopesummiteuropecom. Bye for now.