David Klonoff: 0:14

Welcome to Diabetes Technology Report. This is the podcast that covers diabetes technology. I'm Dr David Klonoff. I'm an endocrinologist at Sutter Health and UCSF.

David Kerr: 0:26

We have a very special guest today and my co-interviewer, dr David Kerr, will introduce him pleasure in welcoming one of the giants of the artificial pancreas development, eyal Diso, who's currently Vice President of Innovation and Connected Care at Lilly Eyal. It's great to catch up with you once more. I think a lot of our listeners want to know how did you end up being interested in diabetes and diabetes technology specifically?

Eyal Dassau: 1:06

First of all, thank you, dr Kiran and Dr Khlona for having me here. Why I'm in diabetes or diabetes technology? I think that for me it's, like many others, motivated by a personal experience with diabetes not myself. My dad has type 1 diabetes for as far as I remember him. So as a young boy I learned how to identify hypoglycemia by just looking at him, how to dose insulin and to inject. Later on I had to apply even emergency treatments, including to him emergency treatments, including to him. So I was very motivated to kind of okay, what do we do with that? How do we improve the care for people with diabetes? And when the artificial pancreas program started, I was lucky enough to move to Santa Barbara to be part of that program and to be part of that initiative that all of you guys were part of it, to generate a new treatment modality for people with diabetes.

David Kerr: 2:17

I mean, the artificial pancreas has come a very, very long way and it's changed lives around the world. Where does it go from here? Do you think? What are the great barriers that we still need or the questions we need to solve when it comes to artificial pancreas systems?

Eyal Dassau: 2:36

So I think that, as you mentioned, there was great strides. I think that we have the first or second generations of systems out there. I think that we have the first or second generations of systems out there and to me it's just the starting point of what we can do with information, with data and with what we call Internet of Things. So we are living in a connected environment, we are living where information is flowing and the question is how we can nourish that information toward helping people with diabetes to reduce the burden of diabetes, improving the care and enjoying life. So spending less time on diabetes, worried about diabetes, and more time about other activities about other activities, eyal.

David Klonoff: 3:28

many people will say that the three areas where we could stand to see improvements in artificial pancreas are faster-acting insulin, better algorithms and better sensors. Which of those do you think could be improved the most? Where should we be putting efforts?

Eyal Dassau: 3:47

You just mentioned the three big ones. The question is what we're looking for from an improvement perspective, and it's maybe more personalized medicine. So when we speak about sensors and algorithms, it's how you tailor and algorithms is how you tailor modern algorithms to improve the life, but extend that even beyond what we call the artificial pancreas or AAD systems. Many are using injections, mdi therapy and other modalities.

David Klonoff: 4:28

How we can use information to better inform insulin management across the board. You know you've done work on an artificial pancreas that delivers glucagon.

Eyal Dassau: 4:39

Could you say something about that? It wasn't really an artificial pancreas. My work on the AAD space was more insulin-centric. We've done some work about using glucagon as a modality in hypoglycemia for patients that were under bariatric surgery, for post-bariatric surgery complications, and I think that's another example of how we can utilize technology to improve a certain group of individuals that needs that help and that, where automation can come to play and, reducing that fear of hypoglycemia, normalize life by utilizing a different drug, in this case glucagon and not insulin. Hopefully, what we envision is something like a patch that has a glucagon with a smart algorithm that will release it at the right time at the right moment.

David Kerr: 5:38

Going back to management of the needs, you know, in your title title it's got connected care. If I'm a person with diabetes, type 1, type 2, what does that mean to them? What can they look forward to?

Eyal Dassau: 5:58

So connected care is the ability basically to take the information that we have whether it's from an insulin modality, in this case, connected insulin pen glucose information and improve the way that we manage treatments, Improve them, Make them better and personalize that to the user's needs, Because each of us are very different. We have different objectives. We need to meet the users where they are by using technology, Making insightful insight from the data.

David Kerr: 6:42

So does that move into? It's not simply insulin. Look at the change in glucose you're getting into, looking at other lifestyle or patient-centered factor so you can really fine-tune the system.

Eyal Dassau: 7:00

I think that we need to consider not just glucose there's glucose, there's the insulin, there's the type of insulin that you're taking, the other contextual information that we are sharing or can be utilized via connected devices, again to improve management of diabetes, in this case, management of insulin delivery.

David Klonoff: 7:26

You know, there's a lot of interest now in classifying people with type 2 and even type 1 diabetes into subtypes. People in this area feel that there's not just type 1 and type 2, but there are many subtypes. Presumably, each of those subtypes would require or benefit from a different treatment. What do you think about that, and do you think that this movement of splitting the types of diabetes is compatible with artificial pancreas?

Eyal Dassau: 7:59

You know I can share probably my opinion here. We've seen throughout life the name, you know the definition of diabetes being modified multiple times with multiple names. So it might be a part of that cycle when I see technology and ability. I would tailor that to personalized medicine In that we, the first or second generation of AAD system, are designed for populations and now there's opportunity down the road to further personalize the systems toward the needs of the individuals. And that's maybe alluded to what you know, whether the type or subtype of you know that, where more advanced algorithms can come to play, learning the data and understanding how it can further modify the system to be tailored to the individual. So it's like a tailored suit, the system to be tailored to the individual. So it's like a tailored suit, it's not off the rack. It would fit to you Now easy to say lots of work needs to happen in that in order to get it into something that people can use, especially in a regulated environment.

David Klonoff: 9:19

Do you think that we're going to be seeing the use of GLP-1 receptor agonists increasing in type 2 and possibly even entering the approval group for type 1? What do you think is the future of those drugs?

Eyal Dassau: 9:40

It might be, I don't know. That's toward the clinicians in the room here and potentially other users would need to navigate that space and see how the utilization of GLPs will come in the future. I really can't comment on that. Okay, you're an engineer by background.

David Kerr: 10:01

I really can't comment on that. Okay, You're an engineer by background AI I happen to know that, and our listeners are really intrigued as to your vision of AI and diabetes. And when are we going to start noticing benefits from AI and what are your concerns about AI? Is it all hype at the moment or what's your thoughts on this?

Eyal Dassau: 10:30

I think that you know whether we're using right now AI, depending on what system you're using or where there's some elements of AI that are being used right now. When you look at artificial pancreas, you know predictions, learning elements there are. My assumption that we'll see more and more as time comes and we have more data going back to tailoring the algorithms to the individuals and how we can further predict lifestyle and ability to change with more information or make faster decisions. There is a it's not an easy task. There is challenges you may have you know mentioning, you know hallucinations or how you can trust the system, how you can verify and validate the system. So there'll be some work that will need to be done there.

Eyal Dassau: 11:24

I think that's part of the discovery process. There's always challenges and I think that, if I reflect back to the AID path, was it an easy task? No, it took multiple iterations and multiple steps by industry, academia collaborations to move it forward. So I would assume that we'll see some progress around the utilization of AI and different type of AI in the future of diabetes technology.

David Kerr: 12:00

I mean, I think that's a fair assumption at this stage. The other question, which is a practical question that people with diabetes constantly ask me, is the AP systems at the moment are not objects of beauty in the sense that they have a certain size and they're separate pieces. Do you see any change bringing everything to one device, much smaller, lasting longer? I'm just wondering what the vision for this is in the future, without giving us any industry secrets.

Eyal Dassau: 12:39

As a former designer in this space, I played or toyed with the idea of an all-in-one device. Going back many years ago I did some simulations of an all-in-one device. Can it be used or are there any interferences or not? This goes back to we need to serve the users, so we need the voice of the users, and I think that we see devices these days that are less medical devices, but they're still not so small, they're not so discrete. Can we make them smaller? Potentially, yes. Technology improved.

Eyal Dassau: 13:25

We are seeing some use of things that we didn't believe that would be reality using your phone as your screen, and that's minimized the footprint of your pump. So we see now more pumps that kind of found a, what I call a patch or semi patch pumps that are smaller, maybe even smaller in the future. Can they be integrated with a sensor? Might be. Hopefully. Somebody is working on it right now as we speak. There's the other opportunity and we've done some work on that in previous life at Harvard and UC Santa Barbara with implantable system, and I know that currently there's some work around that. There are definitely challenges there, but that's become kind of a really interesting proposition of an implanted system that you wouldn't see, provide the full benefit of an AAD system, but nothing is visible on your body, so body image is very important.

David Klonoff: 14:27

Hey Al, as we see more connected care, many people would like interoperability so that whatever device they use can send information to many platforms and their doctor can have access to many types of devices. And then we also see people concerned about cybersecurity, the more information that gets transmitted wirelessly. What do you think about those two areas?

Eyal Dassau: 14:53

You just mentioned. You know two big areas that are investigated how much you can enable interoperability of systems and across systems. Again, meeting the users where they are, as you mentioned, one of the users is the healthcare systems or the users themselves as patients, people with diabetes, and cybersecurity is around us across the board. So how do we make the system robust enough but also enable interoperability there? There are different approaches that are being utilized there and following guidance from the regulators to enable it, the regulators to enable it.

David Klonoff: 15:42

And one last question when do you think we'll be in 10 years or 20 years or 30 years Any of those based on how things are progressing in diabetes technology?

Eyal Dassau: 15:56

It's a good question. I think that we'll have an AAD systems that will be much smarter, much more predictive, as well as tailored toward personalization of needs. So what we probably initially dreamed and but had to propose on that more kind of ability, whether it's interface or the system will learn by itself via AI what is your true goals and what are your true needs and drive it toward that. I think that it's doable down the road, especially in 10 years from now.

David Klonoff: 16:41

Well, thank you for being interviewed and answering our questions. This completes today's interview for Diabetes Technology Report. You can find this report it's available on Spotify, the Apple Store and the Diabetes Technology Society website, so we look forward to you joining us for the next Diabetes Technology Report. I'm going to now say goodbye, Dr Kerr.

David Kerr: 17:07

Thank you very much, Eyal. That was really positive and thoughtful. Thank you very much. Thank you, Eyal.

Eyal Dassau: 17:13

Thank you guys. Thank you Dr Kerr and Dr Klonoff.