Artificial pancreas treats type one diabetes

Scientists have developed an implantable device that mimics the pancreas as a potential therapy for type one diabetes.
26 September 2017

Interview with 

Joan Taylor, De Montfort University


Heading inside the body now. About 40 million people worldwide have what’s called “type 1 diabetes”. These individuals cannot make the sugar-controlling hormone insulin, usually because their immune systems have attacked the tissue in the pancreas that normally produces this hormone. As a result, the body cannot regulate blood glucose levels itself and the person has to regularly measure their blood sugar and then inject artificial insulin to keep their glucose levels stable. This can be inconvenient, uncomfortable and embarrassing. But if they don’t control blood sugar they risk damaging their eyes, kidneys and blood vessels. So, there is a strong motivation to keep them in check. Chris Smith heard how, to solve this problem, Joan Taylor at De Montfort University has built a wrist-watch-sized implantable artificial pancreas, with no moving or electrical parts, that just uses a clever glucose-sensitive gel packed with insulin to do the job...

Joan - At the present time we’re looking at three implant sites. They’re all inside the body cavity near to the liver with blood vessel access and that’s really important because our real pancreases secrete insulin into the blood supply to the liver because it’s the liver that does the chemistry; it’s just that the pancreas is the supplier of the drug. We want to try and mimic that so we had an idea that we would produce a chemical substance that was responsive to glucose.

So we actually synthesised a polymer which is, in our case, a kind of semi solid plastic and that substance contains a glucose receptor and a glucose interactor so that the interactor makes contact with the receptor and can form a lock and key which unlocks when you add glucose to it. As glucose floods the polymer, then the polymer becomes much more permeable. What we did was to build that into a device that holds insulin, so that the insulin is forced to exit through this polymer to reach body fluids where the same polymer senses whether the glucose is high or low and then transmits the insulin either fast or slowly, depending on the need that it discovers on contact with body fluid. It’s completely different, it’s not electronic and it’s not biological so, because of that, it is actually very fast acting, it doesn’t have delays. And, because it’s not biological, it doesn’t need immunosuppression drugs to be given which is the case with transplanted pancreases.

Chris - How far have you got with this so far; is this just in a dish or are you actually doing this in animals now?

Joan - We did a lot of study over the last 20 years in dishes but it got to the point where my mentor of many years said to me “do you know, if you’re really saying this could be done, you’ve actually got to prove it.” It occurred to us that there was really no way of proving this with cells in a dish because what you need here is the device to respond to the biological system in its diabetic state, and that animal or volunteer to respond back. It’s an interactive system so you’ve got to do it in a live system. So what we’ve done is to use diabetic pigs and that work has really pleased us. We’re doing well with that.

Chris - So you have a pig with diabetes, you implant one of these devices - what happens to the pig?

Joan - The pig is quite happy; it doesn’t feel it. The device itself is about the size of a pocket watch and once it has recovered from the implant, we watch the blood sugar come down to normal levels. Then what we do is to challenge the pig’s diabetic system by giving it large amounts of treats and watching how well the hypoglycemia is controlled. The definition of being not diabetic is that a large burden of glucose should normalise within two hours and, in our experience, we have had the normalisation happening in less than an hour.

Chris - How much insulin can the device hold? In other words, between top ups, how long would a potential diabetic be controlled for?

Joan - In terms of actual volume it’s about 5 millilitres, but in terms of the time that would control symptoms for is about 50 days.

Chris - I suppose one risk is that one is walking around with, locked away inside this stuff, potentially a life threatening reservoir of insulin?

Joan - You are right there. That’s an inherent problem with the design is it presently exists but we have plans to modify our system so that instead of the insulin reservoir being liquid, it itself will be in gel form. So, should the device break, then there wouldn’t be an instant leakage of the insulin.

Chris - How long will the device potentially function for? In other words, how long have you taken it out trying this to know that the polymer doesn’t degrade, it doesn’t break down and  does reliably control blood sugar?

Joan - We have kept the gel in a simulated animal fluid for two years and it hasn’t degraded. What we did was to put it inside a membrane bag and that bag has pores in it that do not allow enzymes into the gel that would degrade it. It doesn’t allow bacteria in that would degrade it but it is large enough to let insulin out. So, based on that experiment, we think that the longevity of the gel is at least two years and because the gel is in good shape when we’ve taken that experiment apart again, we think it should be much longer too.

Chris - In terms of the pathway into the clinic, how far along that road are you?

Joan - You have to have a statistical number of animal experiments to show that the treatment is safe and effective and once we’ve finished our small number of experiments we will have to apply for grants to do that larger number. And then, once you have that evidence, you might be allowed to take it into phase one clinical trials, so I would think there’s five years work there. Then, once you’d got into phase one, there another similarly long period before it’s actually released onto the public in a bigger way so quite a long time I think. But what we hope is that we’ll be able to show that this thing actually works and, even if we don’t take it further, it will add to the knowledge out there, and that if we don’t then somebody will take this out into clinic. It’s advantages for diabetic sufferers are that not only does it keep the blood glucose steady, which is a major advantage, but that the whole thing is invisible from the outside and, really, what we just need now is the facility to develop what we’ve got.


Add a comment