The diabetes tech revolution: How closed-loop systems are transforming care

medical innovations and inspiring stories in the field of healthcare. I’m your host Gina Furnival, and in today’s episode, we’re turning our attention to diabetes and in particular a device often known as the ‘artificial pancreas’ or the ‘closed-loop’ system.

Now, diabetes is an area that I’ve got some familiarity with: I work on our platform, touchENDOCRINOLOGY, and also for our endocrinology journal. But I’m delighted to be joined today by my colleague, Sola, who probably knows a great deal more than me about the subject. Hi, Sola. How are you?

[Sola Neunie] I’m well, Gina, thank you. Thank you for inviting me to join you both.

[Gina Furnival] Oh, my pleasure. So, Sola, you’ve been doing educational activities in the field of diabetes for some time. So when did you first start hearing about these systems?

[Sola Neunie] I first started hearing about these in 2016 when the first system was approved for use. But since then, there’s been so much excitement in this space, and it’s still a very much evolving field.

[Gina Furnival] Indeed. From what I’ve been reading, it seems like it’s been very rapid, from research right into clinical practice. So today, we are honoured to welcome a pioneering figure in diabetes management, a leader whose commitment to leveraging cutting-edge technologies has and still does significantly improve the lives of those living with diabetes.

So please join me in giving a warm welcome to our guest, Dr Diana Isaacs. Hi, Diana. So pleased you could join us today.

[Diana Isaacs] Thank you. Thanks so much for having me.

[Gina Furnival] Our absolute pleasure. So for those that might not know, Diana, you’re a specialist endocrinology clinical pharmacist, based in Ohio at the Cleveland Clinic, where you coordinate the CGM program and co-direct the Endocrine Disorders in Pregnancy clinic and also play a pivotal role in shaping diabetes management through your position as a Director of Education and Training in diabetes technology.

So could you tell us a bit about how you found yourself in this field?

[Diana Isaacs] Yeah. Well, you know, I was inspired back when I was in pharmacy school seeing what I learned about diabetes and just seeing how many people have it. In the US, it’s about 11% of the population and then another 38% percent have pre-diabetes, so it’s affecting so many people.

And I think I realized early on just through education and training, we can make such an impact. And when I came to Cleveland Clinic about 8 years ago, it was clear there was such a need for technology for someone to kind of take ownership of that space.

And, yeah, I found myself just getting deep into it. And since then, it has totally exploded. And so it’s been very, very exciting to be involved in.

[Gina Furnival] Indeed. So you’re well known for advocating, aren’t you, and campaigning for access for use of these latest technologies and therapeutics. Could you tell us a bit about your work there?

[Diana Isaacs] Yeah. So I’m involved in a lot of professional organizations like the American Diabetes Association (ADA). I had the great privilege of being on their professional practice committee for three years, which is the committee that updates the ADA standards of care, the guidelines.

And then I’m really involved with the Association of Diabetes Care and Education Specialists. I currently serve on their Board of Directors.

But I think the big thing is there are these great technologies available, but, unfortunately, only about 40% (at least in the US), only about 40% of people with type one diabetes, are using these automated insulin delivery systems, and worldwide, it’s even less.

And so there’s a lot of disparities too that we see in terms of underserved populations having less access to them. And so, my whole thing is that we should let people know what the options are and make sure we’re offering them and also supporting them because different people have different needs in terms of education and training.

[Gina Furnival] And so before we get into those technologies we just touched on there, for those that might not be familiar that are listening today, could you tell us a bit about diabetes and how treatments are designed to work to manage the condition?

Yeah. So, I mean, generally, most there’s different types, but mostly type 1 and type 2 are the most common. And for anyone with type 1 diabetes, it’s an absolute insulin deficiency, which is usually related to an autoimmune process that’s attacking the body cells that make insulin. So in that case, they need insulin. That is the treatment, whether it comes as injections or it comes as an insulin pump, but they need insulin. Now, type 2 diabetes is usually more associated with insulin resistance. And so in type 2, there’s a lot of other therapies that we can use, a lot of orals and non-insulin injectables.

Although, through the progressive nature of type 2 diabetes, if someone’s been living with it for a long time, 20 years, 30 years, 40 years, often those cells that make insulin stop working, and they also may need insulin replacement therapy.

And so in terms of how you treat it then, what you’re trying to achieve when you’re managing the condition? Well, what we’re trying to achieve is glucose levels of a person with a healthy pancreas, but what we realize is with our current therapies, it’s very difficult to do that. So, what we aim to do is achieve what we call the target glucose range, which is between 70 and 180 milligrams per deciliter.

And because insulin, it’s a tricky drug, it can cause blood sugars to go too low, and so we have to be very careful with it. And so the goal for most people is actually 70% in this range, between 70 and 180, and when people have access to a continuous glucose monitor to monitor their glucose, then we can vary. We have all this information and we can accurately determine if they’re meeting that percentage and range. But in general, by achieving that, it reduces the potential complications that can come along with diabetes. And so, if glucose levels are too high for too long, that can cause complications with the kidneys, with the eyes, it can cause a condition called gastroparesis, neuropathy. So we’re really trying to reduce the risk of those complications and help people live long, healthy lives.

Well it’s clear from all those things that can happen that it’s really important to manage these levels in your patients. So as you mentioned before, it’s type 1 diabetes where you are insulin dependent, and it’s really where these automated technologies really fit at the moment, where the shift’s really been. So how was type 1 typically managed by patients before these closed-loop automated systems were available?

Yeah. So, I mean, we were relying on insulin, and many years ago, we didn’t even have as good of types of insulin. We had insulins that peaked at the wrong times and just had a greater risk of hypoglycaemia or low blood sugars.

But, yeah, most people were doing the standard, generally at least 4 injections a day, taking a long-acting insulin, and then taking a mealtime insulin.

And often, people are asked to take even more insulin than that because, let’s say, they had a high blood sugar at another time or they were gonna eat an additional snack. Pretty much any time a person’s eating anything that has carbohydrates in it, that could raise their blood sugars, they need to take insulin. So generally, at a minimum, 4 times a day, but someone could take 6, 7, 8 injections or more a day. So it’s a lot. And typically, these injections are with insulin pens or with vials and syringes.

[Gina Furnival] In terms of conventional insulin pumps, there was something before these automated closed-loop versions, wasn’t there?

[Diana Isaacs] So we have had insulin pumps for several years, and traditional insulin pumps did not connect with a continuous glucose monitor to automate insulin delivery. So it involved the healthcare professional putting in settings, so putting in different background rates of insulin, putting in different carbohydrate ratios, correction factors and and all these different things in hopes of achieving these target glucose levels. I think the challenge is that you can put in rates, and you there’s calculations to make great estimates on these rates, but the problem is you take the same person and 2 days in a row, 3 days in a row, they may require very different amounts of insulin based on other factors, like their activity levels, the amount of time they slept, their stress levels. So you try your best with those pumps, but often, if there’s a lot of variability, it’s very tricky to achieve those glucose targets.

[Gina Furnival] As you say it sounds super tricky, not only for the clinician, but also the patient that’s having to manage this in their daily life. So that’s quite a burden on them. So when we’re thinking of the impact of living with and managing diabetes, what effects did you commonly observe in patients’ lives?

I mean, people with diabetes are people just like everyone else. So anything, any stressors, job changes, things happening with their kids, with their parents, getting sick, that can be really tricky. And so all of the kind of usual stressors that can happen in life happen to people and can cause their glucose levels to fluctuate, making it really challenging. And then on top of that, just managing daily life, we’re asking people to inject, to give themselves insulin, to keep track of how often carbohydrates they’re eating, which helps to calculate the right amount of insulin.

And then we ask people, you know, we have to be so careful with physical activity. In general, we encourage physical activity, but that can lower glucose levels too much. And so kind of trying to work around eating enough to not go low, but not eating too much, just spike up too high. So it really is, it can be a big, it can feel like a big burden. It’s a lot that we ask people with diabetes to do.

[Gina Furnival] In terms of the psychosocial impact then, obviously, you’ve got, you know, to monitor all that. In terms of sleep quality, I’ve heard that could be impacted, and cognition?

[Diana Isaacs] Absolutely, and if someone’s glucose levels are fluctuating a lot, it’s going to be really hard to concentrate, and so you could imagine how it affects kids in school, people on their jobs. So yeah, it can have a huge impact.

[Sola Neunie] As we touched on earlier, the first automated insulin system was launched just under 8 years ago. How did the first automated insulin delivery system operate to manage diabetes, and what impact did that have on diabetes care?

Great question. So I remember that vividly. So we had the Medtronic 670G and it really truly was revolutionary at the time. Our first fully automated insulin delivery. Prior to that, there were systems that started to connect with a continuous glucose monitor, but all they could do was suspend insulin. This was our first system that not only could suspend insulin, but also could increase background insulin based on someone’s glucose levels. So it really was revolutionary, and it increased the time people spent in that target range. 

It wasn’t perfect, as you would expect with the first of anything. And so some of the challenges were that it was very easy to get out of the ‘closed-loop’. So in order for automation to work, you have to stay in automation. Right? Well, what happened is if someone’s glucose went too high or too low, that could trigger them to go out of the automation. And usually, the way to get back in, they had to do finger sticks. And so the first system did require quite a few finger sticks, which I would say was a limitation and caused people to not always be in that automated system. And the ones that spent more time in automation did, generally did, really, really well. But I recall it was a challenge working with some of the people to try to really stay in automation. Otherwise, then their glucose levels would fluctuate more.

[Sola Neunie] And those first generation hybrid ‘closed-loop’ systems were obviously a landmark step in the journey towards the more sophisticated systems that we have today. How do the newer ‘closed-loop’ systems and those technologies differ from the first generation insulin pump that you mentioned just now, which were a significant step, obviously, towards today’s advancements?

Yeah. So the second one that we got in the US was the Tandem, t:slim X2 with Control-IQ technology, and that really was a big improvement because first of all, the continuous glucose monitor that it worked with, did not require finger stick calibration. So immediately, we removed one barrier of needing to do finger sticks.

Secondly, it was markedly improved in terms of not getting kicked out of the automation. In general, people would stay in automation as long as they have – the key is you have to be wearing that continuous glucose monitor.

So those were things that really, really helped. And then I think even the algorithm was much, you know, much improved. It was the first algorithm that not just adjusted the background insulin, but started to incorporate automatic correction. So if someone was running higher, instead of kind of waiting for that background insulin to kick in, that system every hour will give a correction dose to try to bring someone into range faster. So that was, I mean, that was a great improvement. People are still using that system and enjoying it. 

And then the next system that we obtained after that is called the Omnipod® 5. And what’s really unique about that system is it’s a tubeless insulin delivery. And so many people, their barrier to using an insulin pump is that tubing. They, you know, people say, well, their kids are going to pull at it, their pets are going to pull at it, or they just don’t want to feel like it’s too visible, people are going to see that. So that was a great improvement.

That algorithm also works with the Dexcom [CGM system], so it didn’t require the finger sticks. And in general, people sometimes get kicked out, but in general, it’s still better than the original. You know, the original [Medtronic] 670G, people stay in automation, most of the time. So that’s fantastic.

And since then, we’ve had new innovations. So for example, with the t:slim X2, gaining mobile phone access to being able to give boluses through the phone and not have to pull out your pump all the time. So it helps to make it more discreet, make it easier on the user. 

And then, also another recent innovation was the Beta Bionics iLet. And this pump is really unique because, this pump actually doesn’t require settings. So even though we have all these automated insulin delivery systems, we still have to program them as a backup in case someone’s not in automation, so we still put in basal rates, correction factors, carbohydrate ratios, all those things. The iLet doesn’t actually have any of that, you connect it to a continuous glucose monitor, you put in the person’s weight, and then you slide a button that says go bionic, and it learns, essentially learns, how much insulin the person needs. And then for boluses, instead of having to put in actual carbohydrates, a person will put in that they’re eating a usual meal, they’re eating less, or they’re eating more. So that has just, I mean, that is a really innovative system.

And so, yeah, we’ve made tremendous progress over the years.

[Sola Neunie] And are there any real-world data or even clinical trial data that show how those innovations compare to the kind of older, more traditional methods?

So in terms of data, there’s not much data that compares system-to-system, but we do have a tonne of real-world data.

And what also is unique about these systems is that they’re Bluetooth connected. They stream to the cloud so that we have this data. We can see data from hundreds of thousands of users. And so the data is very impressive.

And actually, going back to the Medtronic; so with the Medtronic, the original was the 670G. Since then, now they are, well, they have the 770G and now the 780G. So the 780G is so much improved from their original version.

The algorithm is, like, it’s completely different, and it does such a good job of keeping a person in range. And it has what’s called meal-detection technology, where if someone, you know, you put you estimate your carbohydrates, but there’s so much variability, so it’s really hard to be perfect. Even when reading food labels, they can be off by at least 10%, and so it’s common someone might put in too little or too much, but this needle-detection technology will make up for that, and also it has automatic corrections that actually occur every 5 minutes.

So that’s been a huge improvement, and the fact they have an updated CGM, so people don’t get kicked out of automation, if at all, or very frequently anymore at all. So anyway, this has been around in Europe and the UK for a few years, and so there’s some great real world-data showing people are spending on average, like, at least 75% in range, especially when they have the optimal settings put in. I mean, some populations are spending 80% or more in range. And those are averages, which means some people are spending way more time than that. So it’s really very impressive. There is also real-world data with Omnipod® 5, with Tandem [t:slim] Control-IQ, and the data, I mean, it’s impressive. And this is the reason that guidelines now are recommending that people with type 1 diabetes should absolutely be offered these systems, because they are far superior to what we can do with injections.

So I mean, we must, and I’m very impressed in the UK, there’s also been so much advocacy that now these systems are going to be offered to people with type 1 diabetes. So, it’s just it, it’s just been incredible.

[Sola Neunie] Fantastic. And are there any particular groups of patients who benefit the most, like, amongst children or older adults, pregnant women, for example?

So definitely children. Children, and also what we’re learning is early on in the diagnosis, like, not waiting until someone fails other therapies, like, getting people on these as early as possible, can definitely lead to the best outcomes.

So definitely children have high benefit, although adults certainly do as well. Now pregnancy is really interesting. So I do a lot of management in pregnancy, and there is a system in the UK called CamAPS [FX hybrid closed-loop app] and that system was studied in a randomized controlled trial with Helen Murphy and her team called the AiDAPT trial, showing that it is clearly superior to the standard of care, to other insulin pumps or to multiple daily injections.

We, unfortunately, don’t have that system in the United States, and one of the challenges is in pregnancy, you need much lower targets. So, the target range of pregnancy is actually 63-140, versus outside of pregnancy is 70-180. And so with all of the systems, other than this CamAPS, it’s been really challenging because the lowest target our systems go to is really with the 780G – it goes down to a 100. With other systems, like Omnipod 5 goes to a 110, iLet also is a 110. So those really aren’t stringent enough for pregnancy.

That being said, we are using them anyway, because it’s often still better than the control we get just through traditional injections or through a traditional insulin pump. There is one new pump that was recently FDA approved called, Sequel [Med Tec], Twiist, in the US, and this is going to work with what’s called the Tidepool loop algorithm, which is actually an algorithm that was created by the DIY looping movement. 

So this is a group of people with diabetes who were very frustrated with the slow approvals of these automated insulin delivery systems. They had the technology for years before they were approved, and they were kind of quote unquote ‘hacking’ insulin pumps, creating their own mobile apps for automation and connecting them with continuous glucose monitors. But you know what? They developed a very good algorithm and they actually did a study, an observational study, collecting the outcomes and submitted that to the FDA and gained approval. And so this algorithm is going to work with the Twiist insulin pump, and their glucose target goes down to 87. So I think that’s going to be, that’s great for pregnancy, and so I look forward to the potential use in pregnancy with that lower glucose target.

[Sola Neunie] Amazing! Is there anything else in terms of the challenges and patient burden? I’ve heard about the kind of intrusiveness of the alarm and things like that. How do your patients respond to, you know, equipment burden?

It’s a very good question because it’s still, the thing is you sometimes hear the term artificial pancreas thrown out there with about these systems, and unfortunately that’s really far from the truth. When I think of an artificial pancreas, I think, okay, you throw something on and now you don’t have to think about diabetes anymore, and that’s really very far from how these systems work. With all of them, we are asking people to count their carbohydrates and enter them or at least announce, like, with the iLet, at least announce that they’re eating. That’s really because the action time of insulin is not fast enough, and so if someone never announces that they’re eating, their glucose will rise and then it’s going to take extra time for it to be able to come back down.

And then there’s also even just the filling of these devices, like, you have to fill them up with insulin. So you’re taking insulin from a vial, you’re filling them up. If the device has tubing, you have to run it through the tubing. It can get kinked. There’s a cannula that is running the insulin under the skin. Well, if that gets kinked or bent, that means the person’s not going to get insulin, so they have to kind of problem solve and troubleshoot, like, do I smell insulin on my arm? Is it leaking? Why is my glucose going up? 

So there still is a tremendous burden, even though these make it easier to manage glucose levels, and I don’t think we can ignore that. There is a lot of discussion in this space about, kind of the balance of tighter glucose numbers versus reducing the burden, and many companies, what they’re working on is, is there a way we could not have people bolus? Like, they could just wear the device, somehow it can sense a little bit better when glucose is rising, and then a person wouldn’t have to put in their carbs.

Now the problem or the challenge with this is time in range will usually go down. And so it’s kind of like, well, what’s acceptable? Would 60% time in range be acceptable if a person really didn’t have to do much, as opposed to, like, if you can get 80% with bolusing? So there’s these discussions kind of taking place.

I think a lot of people would also be interested in a device that had a continuous glucose monitor and insulin pump all in one instead of having to wear two different sites because that can itch skin, like, that could be uncomfortable wearing things all the time. So, one thing though that I am excited about, so very recently, so the Eversense® [E3] CGM, continuous glucose monitor, is an implantable CGM with a transmitter that’s on top that you can just take off anytime you want because it’s in the sensor is actually implanted under the skin in an office procedure.

So we’ve had this for a little while in the US, and my understanding is you used to have it in the UK, but then after COVID, I think it went away. Hopefully, it’ll come back. But recently it got integrated CGM status, which means it’ll be able to pair with these insulin pumps. So I could see that being kinda nice because you could just disconnect anytime you want versus traditional CGM. You wear it for 7-14 days. And even though you could say, well, why not just pull it off? Every time you pull it off, you have to put in a new sensor, which is a cost. It can be a high cost, and you have to wait a warm up time for it. And I don’t know, I’m excited about this potential with the Eversense®.

[Gina Furnival] I think that point leaves me quite nicely on to, what’s the future going to be like? So you’ve touched on this already, but there’s obviously other things going on. I’ve read about these dual-hormone closed-loop systems. Is this something that excites you for the future?

[Diana Isaacs] Yeah. So the iLet is currently a single hormone system, but it’s actually the intention is to make it a dual hormone, and they’ve done preliminary studies with glucagon and with insulin. And so, right, when you think about the true artificial pancreas, like a pancreas also produces glucagon to kind of help regulate those glucose levels. And so the thought would be that you could really get much tighter numbers if you had glucagon because then you could get more aggressive with the insulin delivery, and then if it starts to come down too much, you could put out glucagon.

So there is going to be a study with the iLet. The iLet is actually made with two chambers, so it can hold glucagon. So there’s a study planned, a randomized controlled trial to test this, and that will be really interesting. I think we’re all eagerly waiting to see how that pans out. 

One of the challenges is that with glucagon, you are looking at then another site. Like, you would need a site to deliver the insulin, a site to deliver the glucagon, and then you’re wearing the CG app. So and then also we don’t know, like, glucagon is used as an emergency rescue medication, but it’s an infrequent use. We don’t know what the effects are of someone getting, like, some glucagon every day and also the costs associated with it. So I think we’re all eagerly seeing how it plays out because it has the potential to really be amazing, but there’s kind of all these questions.

So we will see. Beyond that, though, I think we’re gonna see more sophisticated algorithms where the continuous glucose monitors continue to get more accurate. The algorithms are getting smarter. I think we’re gonna start seeing more artificial intelligence embedded with predictions. And so, all of that hopefully is going to lead to better management and hopefully reduced burden for the person with diabetes.

[Gina Furnival] And how about type 2 diabetes? Are we thinking it is going to go into that area too, are we hoping?

[Diana Isaacs] So, yeah. Actually, the major insulin pump companies are all doing studies in type 2 diabetes as well. So I think the first thing is we should utilize non insulin therapies, and there’s been so much growth with the incretin therapies, the GLP-1 receptor agonists. So we should be utilizing those, first, but through the progressive nature of type 2 diabetes, it is very common that people do require insulin.

And so these are being studied in type 2 diabetes showing very promising outcomes, and I personally have a lot of patients with type 2 diabetes that are utilizing these systems and doing very, very well. So I think that’s an area we’re going to see more growth in as well.

[Gina Furnival] Undoubtedly it is a very exciting area to be in and to watch, and it’s crucial that I suppose we keep discussing it and provide educational support, as you mentioned, for patients to ensure they’re adopted. So big thank you, Diana, for taking the time today to speak to us, and we wish you continued success in all your work in advocating for broader access to these technologies.

[Diana Isaacs] Thank you so much for having me.

If you enjoyed today’s episode, please do subscribe.

[Sola Neunie] Yes. And we’d also genuinely love to hear from you. So please do go to LinkedIn and search for Touch Medical Media and follow our podcast series by going to Podbean and searching for Visionary Voices, insights from health care professionals, and leave us a review.

[Gina Furnival] Well, that’s it for today. As always, thank you for listening, and see you all again soon.