June 26, 2023

Clinical Trials and Digital Twins

Clinical Trials and Digital Twins

This week’s episode dives into the concept of running a clinical trial before ever running a clinical trial. How is that possible? It’s an idea called digital twins & we have an expert interview to break it down.  Colin Hill, CEO and Co-Founder of...

This week’s episode dives into the concept of running a clinical trial before ever running a clinical trial. How is that possible? It’s an idea called digital twins & we have an expert interview to break it down.


Colin Hill, CEO and Co-Founder of Aitia shares how this AI-powered approach can make drug testing and discovery more efficient for diseases like Alzheimer’s, Parkinson’s, and more. 

Need help?

The successful patient is one who can get what they need when they need it. We all know insurance slows us down, so why not take matters into your own hands? Our Navigator is an online tool that allows you to search a massive network of health-related resources using your zip code so you get local results. Get proactive and become a more successful patient right now at the Patients Rising Helpline.

Have a question or comment about the show, or want to suggest a show topic or share your story as a patient correspondent?

Drop us a line: podcast@patientsrising.org

The views and opinions expressed herein are those of the guest(s)/ author(s) and do not reflect the official policy or position of Patients Rising, nor do the views and opinions stated on this show reflect the opinions of a guest’s current or previous employers. 



Colin Hill  0:00  
Can you run the clinical trial? Can you simulate the clinical trial ahead of time? Can you run the trial before you run the trial?

Colin Hill  0:13  

Bob Goldberg  0:14  
Welcome to the patients rising Podcast. I'm Bob Goldberg venturing out on my own, by my co host, Terry Wilcox, CEO of patients rising is out this week. But we still have a new episode for you today. Today's show is an exciting one, because we're talking about the future of healthcare. Now, for many centuries, clinical trials have operated pretty much the same way. There have been some variations to make the more effective and more personalized, but year after year hasn't been much change, really since the 19 century. However, my guest today, Colin Hill, CEO and co founder of ADEA explains how the intersection of artificial intelligence precision medicine could make clinical trials more efficient. And as a result changed the world. Here's my conversation with Colin Hill.

Bob Goldberg  1:11  
Colin, as we mentioned, at the top of the show, clinical trials have looked the same way for a really long time. And they're falling behind as diseases progress faster than the innovation for the trials. How is your company aiming to change that. 

Colin Hill  1:28  
So of course, clinical trials were created to establish the causal link between the treatment and a change in the clinical outcome for patients. Clinical trials have evolved, they've gotten more sophisticated over time. However, you know, as the diseases we're tackling have become more complex, as we're now dealing with rare diseases. There's a lot that's not accomplished, and a lot of challenges with the current way that clinical trials are conducted. And this has a big impact on patients and their caregivers, in terms of what new treatment options are coming forth, whether it's an Alzheimer's or Parkinson's, or pancreatic cancer, or glioblastoma, or what have you. And so, this is where AI, genomic, biotechnology, and the like are now starting to play a role to create a completely different way of discovering new drugs, and better designing clinical trials to now enrich and have those patients in who are going to be either developing the disease faster, or are more likely to respond to therapy. And so this is really at the heart of the mission of ATF, a company I co founded 23 years ago at a research I was doing at Cornell, coming out of the theoretical physics department. Really the core mission of the company is to reverse engineer digital twins, which are really computer models that capture the genetic and molecular interactions driving patient outcomes, driving the various clinical outcomes.

Bob Goldberg  3:28  
 Can you give us an example? I know specifically, Alzheimer's is a very complex disease and billions of dollars have been spent looking at just one disease mechanism. So how are you navigating that and making the process more efficient?

Colin Hill  3:43  
 One of the big challenges in Alzheimer's clinical trials is that we have a very high screen failure rate meaning from people who are potentially recruited being screened to come into a trial. And these are people who have a family history of Alzheimer's or they're having some early symptoms, or they're just worried well, the cognitive tests that are given to them initially before doing PET scans, which are the more definitive test for presence of beta amyloid plaque, only around 20 or 30% of potential patients are excluded at that stage from those tests before they go on to the very expensive PET imaging test. Now, with the PET imaging, roughly 80% of people are screened out from that test. So this adds up to essentially only one in 10. People who are initially screened for a clinical trial end up being really eligible for that trial. And that screening process takes a long time. It's very expensive, to the point that you're talking about sometimes years to recruit for a trial that's going to last for not that long. And so one of the key things he is doing with various partners including gap, the global Alzheimer's platform, is creating digital twins that are able to identify markers for pathology, does a given individual really have Alzheimer's? Or do they even have mild cognitive impairment? Or they have nothing? And instead of the very long process to try to determine if someone has disease? Can we determine that upfront? Can we determine that ahead of time? 

Bob Goldberg  5:30  
Now, when we talk about digital twins, how are you using those? Is it the screen compounds for the creation of new drugs, is find out more about what isn't working with some other drugs for one reason or another. 

Colin Hill  5:44  
So it's really both? First and foremost, we're using these digital twins to discover new drugs, full stop the therapeutic options, and a lot of diseases just don't exist, right? And so what do you do? What does a doctor do? What does a clinician What does a caregiver or patient do when there just isn't an effective treatment? And so first and foremost, ata exists to create new treatments to discover novel drugs going after novel biological pathways that have not been figured out yet. So that's first and foremost what we do. And we now have the ability to create early stage drug candidates with strategic partners. The second thing we do with these cohorts of digital twins and specific diseases, is we use them to simulate other companies drugs, and that then yields insights into potential response non response markers, that then are directly used in the design of future clinical trials, because one can have an effective treatment. But if you're putting it in the wrong set of patients, that drug is not going to work, it's not going to come to market. 

Bob Goldberg  7:02  
And then I think we're all curious, what's the future look like with something like this? How is AI going to change clinical trials, change medicine, and change patients lives? A decade or two decades from now? 

Colin Hill  7:18  
Look, what's going to be different is we've had this continual decline in productivity in terms of drugs coming through trials to market, and it's costing us more and more money, right. So with all the great advances in technology and medicine, you have this, what's really barely move this very kind of horrible statistic, in terms of our ability to translate new discoveries into into drugs, there's no other industry like it. And so this is not sustainable. And what's at the heart of why it's so bad, is that we're missing 95% of the circuitry of disease. And so was ends up being trial and error upon trial and error. Despite all of the great advances despite being able to develop COVID vaccines in record time. Despite now having new technologies like CRISPR, and having amazing advances in AI. These things have yet to come together to really change the paradigm of drug discovery and clinical development. What we expect to now come from our technology, which is I think the first true convergence of large scale multi omic human data, causal AI that's above and beyond things like large language models and chat GPT and really tailor made to answer this question. And of course, large scale supercomputing, that convergence is now producing the ability to discover drugs in a completely different way, with a much higher level of confidence to translate those discoveries faster and put them into well designed clinical trials with the right patient groups, and ultimately bring true breakthrough therapies to market for patients faster, and then even better match those drugs to patients in the real world.

Bob Goldberg  9:25  
Thank you to Colin Hill for joining us and giving us a sneak preview at the next couple of decades in medicine. And thank you for listening. We want to make sure you stay up to date on all of our new episodes, so make sure to click the Follow button. And don't forget the pass this episode along to your friends, family and fellow patient advocates. Next week, is the fourth of July. So no new episode. Instead, we'll be back on Monday, July 10. Until then, for the entire team had pasted is rising Bob Goldberg Happy Independence Day and stay healthy

Colin HillProfile Photo

Colin Hill

CEO, Aitia