Welcome to Farmer Talk Radio. This podcast is focused on lessons learned in the development and introduction of closed loop systems from the twenty twenty four pod Partnership Opportunities and Drug Delivery Conference. For more information about the pod conference, editorials, podcasts, or webcasts, please visit Drug desh Delivery dot org. Thank you and enjoy the podcast.
Good afternoon, everybody. Thank you for sticking around towards the end here. It's a privilege to be able to moderate this panel today on a topic very near to many of our careers and hearts on looking at closed loop systems across multiple therapeutic areas. Lessons learned. I'll introduce myself first and then ask my fellow panelists to do the same.
Chris Kovalcik I lead this systems engineering organization within Abbott Diabetes Care, responsible for the development and commercialization of the entire portfolio of diabetes care products, including the Freestyle Every Glucose monitoring system and their associated cloud and mobile application services.
Hi. Stephen Russell.
I'm an associated professor at Harvard Medical School and over the last few years, the chief medical officer of Beta Bionics, which is a company that makes an automated insulin delivery system.
Hi.
Am Hindley Pinkson from the bio Innovation Hub nov Ive Externally Innovation.
Hey Ben Lane from Key Technologies. I'm VP of Product Development. We've been around for a little over twenty five years, and much of that time we've been developing diagnostics and drug delivery devices, so we've been in this market for quite some time.
Hi, my name is Louiterritor. I'm a pH Kennedy at Harvard Medical School and MIT working with Bob Linger and Geo Traversal on this closed loop drug delivery device, which we'll be talking about shortly.
Okay, So to start us off to some common background. When we talk about a closed loop system, you generally need three or four ingredients to do so. Need to be able to measure a metric or variable, diagnose a dose or a therapy, deliver that dose, and you also need a drug to be able to do that in some cases. So if we look at what therapeutic ecosystem has been at the forefront of closed loop space for
the last twenty years or so, it's been diabetes. So first question I'd like to ask doctor Russell, why has diabetes been so successful in terms of not only development of closed loop systems, but commercial uptake and getting it out to patients compared to other therapy areas.
Sure, I think the reason that diabetes has been at the forefront is, first of all, you mentioned we have to measure something. The physiologically important variable in diabetes is glucose, blood glucose, and fortunately blood glucose is in equilibrium with interstitial fluid glucose, which then can be measured in a minimally invasive way with continuous glucose monitoring devices, and they've gotten progressively more accurate and more reliable, to the point
that now the accuracy is excellent. The next piece is that for delivery, we have insulin, which can also be delivered subcutaneously.
It's absorbed very well.
The pharmacokinetics are pretty good for the ultra rapid insulin analogs, and so that means that the control loop isn't too long. I mean, if you measure something and then deliver something, but then the drug that you deliver doesn't have an effect for a day, that's not going to work. Fortunately, insulin works within reaches peak and blood within an hour or so, and you can get a relatively short control loop.
And then the reason it's had uptake is that it's really hard to control glucose by yourself as a patient.
Diabetes is completely unique, and.
That we've got this really powerful, really dangerous drug, insulin, and then we tell patients to figure out how much of it to give themselves multiple times a day, which is absolutely crazy and doesn't happen anywhere else in medicine. And so there was a real unmet need for changing that us the the development of closed loop systems and their uptake once they were developed, I think, yeah, so I.
Would just echo that that the burden of treating yourself every day and adjusting many times throughout the day is really big. And and then maybe also that the way that it came about that we have closed loop today, it didn't came in one day. We tried back in the seventies where there's something you carried on your back. And and then you say, insulin came out and was a business in itself. Then the pins were there, pumps
came out, was a business in the themself. And first the ten years after the CDMs came out, we were able to get it approved as a closed loop system.
So going along those lines with the success of diabetes and closed loop on it. I'll start with you, what are some of the areas we've seen limitations or areas for improvement before we segue into other therapy areas.
Okay, yeah, So then I think if first I will just trying to skitch out where we are today, and I think Steve did a bit of that with the With the you would say that it starts with that we are giving a bolus in from from the pump, and then when do you see that effect on the CTM since you need to wait around one and a half hour before it has been throughout the system, so there is a big time delay in the system. It works, and people are adjusting the meal habits and shows so
they don't get too much fluctuation there. But then the in the we say then what could come that could change this? I think the first thing would be to get the time to lay down and and the biggest element is the insulin. And right now we have a dedicated insulin in in clinical trials and that that will work so fast that they will not be available in pins because they might give a hypoglycemia before you're finished with your meal. So so we can cut down a
lot of time delay on the insulin. Then I would say on the pump side, we have an infusion set. It sits on for three to seven days. CTMs can sit on for fourteen days. Maybe you could picture that the infusion set could reach the same level as CDM. On the CTM sensors, you see a trend on moving from subcutis interstitial liquid measurement and some of the new ones going into and when you go into dermal you will gain five to ten minutes more in the reduced
time delay. And then you can say, if you look not on the system but the alternatives, then there's cell therapy making Bezza cells and putting them in the pocket in the in the body, and there's a glucose sensitive incident on its way.
So if we consider everyone sitting in this room, the majority of us work for organizations that work across a variety of therapy ecosystems. So I'd like to segue into to lewis what are if you had to garner a guess, what's the next therapy that's right for closed loop infiltration outside of diabetes.
Definitely, I'd say in chology is a space that really is just prime to have close loop therapies start to be implemented. So I'm sure many of us have experienced someone in our life receiving chemotherapy, but they don't think many of us know is how they are dose. So chemotherapies have a very narrow therapeutic window, as many of us likely know. They work by killing cells through just kind of like means that would also kill healthy cells as well. But amazingly, the way that they're dose right
now is on the body service area basis. And what that means is if you have Lebron James six nine two hundred and five two hundred and fifty pounds and happy the.
Dwarf, if Happy the Dwarf was four six.
Five hundred pounds, that'd get the same amount of chemotherapy. And the reason for that is the basis that they're doc on is using ideal body service area that comes from this equation from nineteen sixteen.
So these two cousins to.
Boy and to Boy, they empirically fit data to nine people. They took height and weight and body surface area and empirically fit this equation to it. And that's the standard of care today for figuring out how to dose patients with chemotherapy across the world that the data fire were
in many other centers. So it really seems like oncology, just given the nature of the cytotoxicity of these drugs and just the center of characters being in the state where it's using this equation from one hundred years ago, or an abbreviated version of the pharma that was driving in nineteen seventy one when calculators weren't around, so it was a much easier way to get to BSA, but
it was based on the Debois and Dubois parmula. So I'd say for those reasons, and the fact that the patient's generally in an a fusion care center, so you have the infusion pump puck directly up to them, so you have a way that you could modulate DELS to keep the concentration of drug within the therapeutic window. Really makes oncology a very interesting space.
Yeah, it's interesting to think about these different time constants. You know, diabetes is a very very long time constant, sensors in there for a couple of weeks, whereas what you're talking about is more more acute. Acute maybe the wrong word but very short time caused. And there's plenty of other applications where you know that's time constant is on the treatment windows on the order of hours and not you know, days, weeks, years.
Right, intermittent in that case, whereas somebody with type one diabetes has to be under insulin treatment for the entire rest of their lives. So whatever system you come up with has to be wearable and easily usable, and something that somebody can doesn't create more burden than it reduces.
So going off that thread, the last panel talked a lot about considering the user and the development of these systems. So if we think across different therapy areas, the user plays a different role, right, So we talk about diabetes, the user is very intensely aware and often more educated on the technology than some of the people that develop
the systems. If we think about other therapy areas like neurodegenerative or on cology, like you mentioned, open question to the panel, how do you see the role of these user playing in the different close loop applications across these therapy areas well.
One thing I think is that we're trying to reduce the role of the user and Type one so.
The first.
Automated insulin delivery technologies are so called hybrid systems, so they modulate basil insulin. You need both basil and bolus insulin. Basil dealing with the basil insulin needs like overnight, and then bowl is dealing with meals and so forth. The systems typically manage the basil part and then the user is responsible for the bowls part. Our system, the islet is designed to take over one hundred percent of the insulin dosing and reduce the need for the user to
get involved. And what we're trying to solve there is that only twenty percent of people with type one are able to maintain the kind of glucose control they're supposed to do because people with really good quantitative skills, really good attention to detail, so forth, people with great executive function they do fine. Everyone else doesn't. So taking the user out of the loop more and more is the direction we're trying to move in for diabetes.
What about a therapy area such as Alzheimer's where the user could be the person with cognitive impairment or a caregiver. Anybody want to comment on, how do you see closeup playing in the Alzheimer's space.
Yeah, I mean that's a super challenging patient population, right, I mean, diabetes has this massive patient population. I think, you know, we've done a very good job with the vast majority of that group. But there's certain are sub certainly are subgroups pediatric, you know, cognitive issues, dexterity issues where managing the treatment but then also just applying the pump, applying the sensor becomes very challenging for that patient group and there's definitely room for improvement there.
The question with things like you know, neurodegenerative diseases is what to measure, right. I'm interested in Parkinson's because my mother suffers from that and she has to take this complicated regimen of multiple doses. V just I guess recently got approval for doodopa, which is an sebt Q infusion, so that seems very familiar to me.
In the insulin space.
You can infuse this drug supultaneously maintain steady levels. But if you were going to try and turn that into a closed loop system, what would you measure? You know, would you measure interstitial dopamine That's not really where it functions, right, So would you measure tremor maybe you could do something like that. I think that becomes the challenge for these other fields.
You could say that if it's not a fully closed loop, but just some of the components, like using the CDM for for other for other things, then I think that the that the we have tried out to use a CDM for tight trading once weekly insulin to figure out quite fast where to go and then go up in a few steps. Yeah, so really long acting components getting them to the right level fast. Uh. And one other element could be to use the CTM sensus for anti obesity treatment.
So segueing to lessons learned, one of the areas we've talked about so far as the readiness of the technology.
So back to the diabetes space.
A lot of the technologies used to do any of these aspects of closed loop development have been readily available for the last twenty years. But even still, if you look at the companies playing in this game, very few if any, develop all of the technology all internally and
requires a lot of partnership. So if Ben, I'll go to you, if we look at focusing on improving technology in these other areas, what are some of the advices or lessons learned that you've seen that helps not only develop this technology, but be able to partner effectively to get these solutions out to patients for other areas.
Yeah, that's a great question.
I think, you know, the collaboration of all the partners that are involved in this ecosystem, it's a complicated beast raight and the ability to rapidly develop and deploy a system depends on early engagement between all those parties. So you're developing a holistic system from the beginning as opposed to disparate development that then gets tied together.
At the end.
And it's a very difficult challenge. We see it all the time in diagnostics where the you know, the assay is developed and then the consumables developed and the instruments developed, and there's not that holistic from the beginning viewpoint, and I, you know, I don't have a lot of great answers there. It's very challenging because very often the pharma company is not in the in the hardware space.
And the drug delivery.
Company is not in the diagnostic space, and so tying all that together early to develop that holistic, efficient system is particularly challenging.
The regulators can help a lot though that's another place we haven't discussed that.
We had an advantage in.
The diabetes space because the FDA created an it's this five ten K pathway for closed loop systems, and they created something called interoperable CGM, an ICGM with so it's a five to ten K with special controls and essentially if it could communicate with one of the devices in the x ecosystem and if it met certain accuracy criteria, then you could get this ICGM status. And as a company making a closed loop device, we are able to now do the study with one ICGM and then substitute
another ICGM into our system without a regulatory filing. We just did a letter to file, so that really accelerated
the whole process tremendously. You know, we launched with one CGM, we got the spec for another CGM, and launched four months later with a second CGM, and then we got the spec for a third and then launched a number of months later for that, and now we're we've got three CGM sensors that are all available, so it's user choice that wouldn't have been possible without the FDA stepping in and creating that regulatory framework.
You could also say that by going through a partner's you open up a free, free competition and the BIST, the pump company can can pick the BIST CTM company and I and you get it even bitter system than if they try to do it internal.
Yeah, this is an interesting space because you have companies who are partners and competitors at the same time, sometimes even on the same product. Switching to another technology question with this for you, So you mentioned oncology, and in terms of companion diagnostics, to measure the metric you're proposing requires a sophisticated technology in the clinic to be able to measure. Do you see start with oncology, Do you
see that becoming a low cost product eventually? And what do you think would it take to get away from a diagnostic machine to measure the variable versus a wearable What.
Are your thoughts there?
Yeah, absolutely, So for context, we've developed this closed loop drug delivery system to control of the concentration of chemotherapy and we've shown it in vivo we published it earlier
this year. So in the system, we draw a blood from the animal or eventually a patient and then we do this raptive the blood preprocessing step to then put it onto a high performance to liquid chromatography MasSpec or HBOCMS system, which has been as many of the people in this audience likely know, a gold standard by the FDA to determine the concentration of drug for many many
different clinical trials over the years. So we really wanted to design the sensor or the overall system around the sensor because of the development and the trust and to the points that we're mentioned earlier, the fact that you really need this sensor, and that's one of the things that really helped in the diabete space to enable closer the control. So we designed this overall system around LCMS, and then the loop gets closed by taking that concentration
reading and put it into the control algorithm. So with that, that's one of the major questions we always get is LCMS is very expensive. The high end ones that you'd likely want in the clinic are just shy of about a million dollars, So the upfront cost is large, But when you start thinking about the lifetime of these types of devices, they'd likely last with good maintenance on the
order of about a decade. You can imagine you can do about two maybe three patients a day with the system comes down to about one thousand uses over a year, ten thousand over a decade, So the cost per patient is approximately one hundred dollars if you work it out according to that way. Of course, there's many other variables
that go into account with it. You'd have to have personal cost, but the cost of the system isn't as large as you'd initially think, and there's likely if this system were ever to reach it to the clinic, they'd likely be work continuing to bring the cost of the
LCMS device down. So definitely opportunity i'd say is the ability to make low cost LCMS devices is definitely something that's interesting, and that's just one of many different types of approaches you get take for the censor, But we really decided to focus on that just because it is very well trusted, able to be applied to multiple different drugs, and there's allowed of us to develop this system and apply it to a chemotherapy space.
You also have to think about the value.
So in the diabetes space, you know, these devices are not nearly as expensive as that, but there are a lot of people because each person needs one and multiple over their lifetime. Usually, but we're reducing the risk of long term complications, and you, I believe would be reducing the risk of chemotherapy related complications because of overdosing or
inadequate treatment of a cancer from underdosing. So I think you can easily make the argument that a one hundred dollars is worth the reduction to an insurance company is worth the reduction of those those complications.
Yeah.
Absolutely, There's been this additional approach that's been developed therapeutic drug monitoring, where you take the concentration of the end of one around of chemotherapy and then don't suggest for the next round of chemotherapy. And even with that, which if you think about it, that's the closed system. It's a sampling rate in the case of five few of once per two weeks. It's a very basic control algorithm
that's normally for five few. The clinical trials I'm aware of that have been done, it's a rule spase algorithm, so if you're a certain percent above, you decrease by x percent of the dose. So even within those trials,
they've shown benefit in terms of tox and efficacy. So I completely agree with you, and in the paper we put out, we did some preliminary analysis on that with certain assumptions of how much benefit we could apply based on those TDM studies, and it does when you take everything into account, does appear that it could be cost effective, even given the high costs involved.
An area like to go into a little bit is around access. So when the first hybrid closed loop technology is launched in the diabetes space, if you purely look at the physics and the technology, it's a no brainer that every single person with diabetes would therapeutically benefit from this system. However, if you look at the percent of the population of persons with diabetes that are on such
a system is incredibly low, a single digit percent. How do we increase access to these technologies by building patient trust not only a diabetes space, but in other therapeutic areas where the concept of a closed loop system would be completely foreign to an individual receiving the therapy.
Well, I'm going to mention one thing, which is that percentage being that low. That may be true for diabetes as a whole, but if you look at type one we're we're at about thirty percent, So it's really pretty substantial adoption in the type one space. The other thing is that it's kind of a unique situation because, as
was mentioned, CGM was always already available. Insulin pumps were already of aailable Obviously, the insulins were already available and the people were using them, they just weren't using them together. And when the closed loop devices came out, they used the same billing codes as the dumb pumps, right, So the the insurer was getting increased value for no increase in cost. So the value proposition was pretty compelling in
that case. I don't know if that's that's probably not going to be the case for some of these other ones.
You can say that I tried to look up how much longer would a person live if if they go from a pin system to a closed loop system, and I might not have got all the data right, but what it's all around the too too too quality adjusted life years in And then I think in in US we were most insurance are willing to pay fifty two hundred thousand per year quality adjusted lifespan improvement. And so you will say this should be enough recent enough science to YE to get it out there.
We have about three and a half minutes left, so I like to open it up for questions from the floor.
Either.
My name is Sheldon Oberg and with bets to biomedical but in a former life I used to head the Electronic Diabetes Insulin Delivery System Electronic Mini Men. The last panel was on digital health, and as we know, the insulin industry has been using connected systems for well over two decades. I'm curious on your insights on how digital health has helped and improved that patient barka, So I'll start.
So, I think one area is providing access to information. So by having connected applications or just having the data stored in a cloud, it's made life easier for not only the persons with diabetes using the systems, but also they're endocrinologists that they see. We could spend a whole panel talking about the difficulty with the minimum amount of time that a patient gets with an endocrinologist, and I'm hoping you can.
Talk about that.
But having access to this information, getting feedback on how to create these reports and graphs and how to make
it very available is very helpful for people. Also simplicity, I think one of the dangers, as was mentioned in the last panel, is sometimes we connect devices for connection sake rather than actually thinking about what's needed, and the reality is a lot of people that are on closed loop diabetes systems don't want to have it in true life, so providing them with every single bell and whistle and alarm and alert actually is a high level of dissatisfaction
for the user. So I think it's important that we're mindful of, yes, we can have these digital companion applications, but being cognizant of how to use them effectively to provide a wealth of information while having limited intrusion to the user base.
One thing that I think is pretty interesting about the digital health applications for this use case is that so called follow apps are really impactful, which means that somebody can have their CGM glucose information followed by another person, often a parent, if it's a child, or a spouse, or even just another friend with diabetes, and so they watch out for each other and if some you know, if an hypo alarm for your friend goes off, and you know.
You can call them up and say, hey, are you okay?
You seeing this? And I think that that's tremendously valuable. It kind of expands the network of caregivers beyond the people who are immediately present to any anyone you want.
Frankly, yeah, I would second that.
My five year old niece is Type one and the connected ability to have her parents be able to monitor is just tremendous.
It really is.
Is there an there one more question?
I work for a design agency and we'll be looking, in fact, for the last sort of a year or so into biosensing and.
The closed loops.
I mean, in fact, I came across a company it's only down the road. Then they created this sensor where you put in just a plastic cube coming out from a wound after an operation. You would tell you about the chemistry of the liquid coming out. That will tell you how well is the actually you know the hill that is taking place, which is fantastic and it's also a throwaway center if you like. So I was I want to ask the panel, how where are we going
with a biosensing I think we are there. There are sort of many examples out there, but I haven't seen a lot of the sort of like closed loop systems. I mean, we kept talking about which is great about type one diabetes and so on, how can biosensing help us move into other bigger areas and bigger applications.
Thank you, Eric, Do you want to take this?
Yeah, So, I think there's a that they'll becoming a number of new symstotypes and some day they call it electronic nose, a sensor that can that can sense a number of different things and then by looking a little bit on on the fingerprick from then then they can then they can miss a lot of metabolics and key tones and and other stuff. And yeah, so so I think there's a lot of UH since US on its way. Also the CTM since US are trying to build in more sensors within the existing system.
So yeah, there's a lot of interest in a key tone sensor for type one diabetes space so you can detect failure of insulin delivery, but also for lactate for the UH, the lifestyle UH or or athletic improvement. So having lactate sensors measurements of lactate is really something that's only done by pro athletes or or high level athletes now, but you could conceivably have c g M like measurements of lactate which could potentially improve athletic endeavors.
And it's yet to be seen.
I mean, both of the major CGM companies have launched direct to consumer devices for people without diabetes, with the idea that being able to see what particular types of food due to your blood glucose would change your habits.
And I can tell you that it's done that for me.
After seeing what some of my favorite breakfast foods did to my glucose, I changed my habits considerably, So I think there definitely is a role for this.
We hope you enjoyed the podcast. For more information about the pod, conference, editorials, podcasts, or webcasts, please visit drug desh Delivery dot org. Thanks for listening.