'Bionic' pancreas could be used to treat diabetes

Medical practice

"An artificial pancreas could allow thousands of diabetes patients to live normal lives," the Mail Online website reports. A UK study involving a device that “top-ups” blood sugar levels with insulin has proved successful in people with type 1 diabetes…

"An artificial pancreas could allow thousands of diabetes patients to live normal lives," the Mail Online reports.

People with type 1 diabetes require lifelong insulin, as their body does not produce any. Insulin is a hormone that plays a key role in regulating the body’s blood sugar levels.

In a new study, the safety and effectiveness of a “closed-loop” insulin delivery system has been assessed.

Compared to a standard insulin pump, where the insulin delivery is programmed, the closed-loop system continuously measures sugar levels and automatically makes fine adjustments to insulin delivery in response. In effect, it acts like an artificial pancreas.

It can be challenging trying to keep insulin delivery at the right level in order to control blood sugar levels within the normal range, while avoiding blood sugar becoming too low (hypoglycaemia), particularly overnight.

The device improved blood sugar control overnight – importantly, it was not associated with hypoglycaemic episodes.

However, one of the trial’s limitations was its small size. In addition to this, it only examined the effects of the overnight closed-loop system compared with the standard pump over four periods of four weeks each. Longer-term studies examining the safety and effectiveness of this system in larger numbers of people with type 1 diabetes are now needed.

Where did the story come from?

The study was carried out by researchers from the Universities of Cambridge, Sheffield and Southampton, and King’s College London. It was funded by Diabetes UK.

The study was published in the peer-reviewed medical journal The Lancet.

While the Mail Online’s reporting of the study is broadly accurate, its headline: “Artificial pancreas could help stem the diabetes epidemic: Device could help patients lead normal lives by stopping need for constant insulin” is potentially misleading on multiple levels.

Firstly, “artificial pancreas” could be misinterpreted to mean that this is an artificial organ that is surgically transplanted into the person and can produce insulin to take the place of their own pancreas. In reality, the “closed-loop” insulin delivery system is designed to be worn outside of the body.

Secondly, the “diabetes epidemic” is usually taken to mean type 2 diabetes, which is associated with lifestyle factors such as being obese and lack of exercise. It is true that some people with type 2 diabetes can go on to need insulin; however, this particular study looked at people with type 1 diabetes.

The rise of people with type 2 diabetes can rightly be described as an “epidemic”. In contrast, the number of people who develop type 1 diabetes (which usually starts during childhood) in any given year has remained relatively static (around 24 in every 100,000 children).

Neither would this treatment “stem” the number of new cases of either type of diabetes.

Thirdly, the Mail says the treatment would “stop need for constant insulin”, which is not the case. In fact, this overnight closed-loop system delivers constant insulin. It has also only been used overnight, meaning the person continued delivering their insulin as normal during the day.

What kind of research was this?

This was a randomised crossover trial that aimed to see whether the use of a novel overnight insulin delivery system would help to improve blood glucose (sugar) control in people with type 1 diabetes.

Type 1 diabetes is an autoimmune condition where the body starts to produce antibodies that attack and destroy the insulin-producing cells in the pancreas. The body can therefore not make insulin, so the person relies on lifelong insulin injections to control their blood sugar. Type 1 diabetes most commonly develops in childhood.

It is different from type 2 diabetes, which is where the pancreas still produces insulin, but it either cannot produce enough, or the cells of the body are no longer sensitive enough to the actions of insulin to adequately control blood sugar. Type 2 diabetes is usually controlled by diet and medication, though some people with poor control also end up needing insulin injections, similar to people with type 1 diabetes.

As the researchers say, one of the main challenges with type 1 diabetes is keeping the right level of blood sugar control; people with this condition can face the challenge of complex daily insulin regimens and regular blood sugar monitoring.

One of the most common risks is when the blood sugar becomes very low (hypoglycaemia), which can cause varied symptoms, including agitation, confusion and altered behaviour, progressing to loss of consciousness. Hypoglycaemic episodes can often occur at night and after drinking alcohol, making it a particular risk for young people with diabetes.

This study was looking at an overnight “closed-loop” insulin delivery system – in other words, an artificial pancreas.

A small device is connected to the body through a standard insulin pump, and this delivers insulin under the skin without the need for continuous injections.

The wearer adjusts and programmes the amount of insulin to be delivered, according to their blood sugar levels.

The closed-loop system is different: a real-time sensor continuously monitors the person’s sugar level (by measuring the level in the interstitial fluid that surrounds body cells) overnight and then automatically increases or decreases insulin delivery in response to this, as would normally happen in the human body with a healthy pancreas.

Studies to date have suggested that the system is a safe and feasible option, and decreases the risk of hypoglycaemia.

This crossover randomised controlled trial aimed to see whether four weeks of unsupervised use of the overnight closed-loop system would improve blood sugar control in adults with type 1 diabetes.

The crossover design meant that the participants acted as their own controls, first receiving insulin with the closed-loop system or a standard insulin pump (control), then swapping over to the other group.

What did the research involve?

The study recruited 25 adults (18 years or over, with an average age of 43) with type 1 diabetes, who were used to using an insulin pump, monitoring their blood sugar and self-adjusting their insulin.

All participants first took part in a two to four week run-in period, where they were trained in the use of the insulin pumps and continuous sugar monitoring, and their treatment was optimised.

The trial was then divided into two subsequent four-week treatment periods, with a three to four week wash-out period in between, when they continued with their normal diabetes care regimen.

In the two treatment periods, the participants received continuous sugar monitoring and were randomly assigned to receive overnight insulin delivery with either the closed-loop system or a standard insulin pump (control).

The study was open-label meaning that participants and researchers knew what system was being used.

The participants received the treatment unsupervised and at home, though they stayed in the research clinic for the first night that they used the closed-loop system.

They were instructed to start the closed-loop system at home after their evening meal and discontinue it before breakfast the next morning.

The closed-loop system calculates a new insulin infusion rate every 12 minutes in response to the monitored glucose level.

The primary outcome examined was the time the person spent in the target optimum sugar range (3.9 to 8.0mmol/l) between midnight and seven in the morning.

Of the 25 people randomised, one person withdrew from the study, meaning that only 24 were available for analysis.

What were the basic results?

The time that participants spent in the target optimum sugar range during the seven-hour overnight period was higher when using the closed-loop system (52.6% of the time) compared to when they used the control pump (39.1%), with a significant difference of 13.5%.

The closed-loop system improved time spent in the target range in all but three participants. It also reduced the average overnight sugar level and the time spent above the target range, without increasing the time spent with a hypoglycaemic sugar level. The time spent with hypoglycaemia overnight (less than 3.9mml/l) was no different with the closed-loop and standard insulin pumps. The closed-loop system was found to deliver 30% more insulin during the night than the standard insulin pump.

There was no difference to total daily insulin delivery. However, when examining the full 24-hour period, when participants used the closed-loop system overnight their 24-hour blood sugar level was significantly reduced (by 0.5mmol/l), and their time spent within the target range was increased. People were also observed to have significantly lower levels of HbA1c (glycated haemoglobin – a longer-term indicator of blood sugar control over the past weeks to months).

There were no severe adverse effects associated with using the closed-loop system.

How did the researchers interpret the results?

The researchers conclude that “unsupervised overnight closed-loop insulin delivery at home is feasible and could improve [blood sugar] control in adults with type 1 diabetes”.


It can be a challenge for people with type 1 diabetes to keep insulin delivery at the right level, which is necessary to keep blood sugar levels within the normal range. Avoiding periods of hypoglycaemia can be a challenge, particularly overnight.

A further challenge is that the symptoms of type 1 diabetes usually develop during childhood. This means that children, especially teenagers, can often find the need to stick to a particular treatment “regime” and regularly monitor their blood sugar quite restrictive. However, without such treatment recommendations, they can be at risk of complications, such as hypoglycaemia.

Because of this difficulty, a device to help simplify the treatment of type 1 diabetes would be welcomed.

The device in question, the closed loop insulin delivery system, automatically makes fine adjustments to insulin delivery in response to the glucose level being continuously measured.

This crossover randomised controlled trial demonstrated that the closed-loop system improved blood sugar control overnight.

Though the closed-loop system was only used overnight, the effects also extended into the day, significantly reducing their 24-hour sugar level.

Importantly, it was not associated with hypoglycaemic episodes.

This study is also said to be the first to monitor safety and effectiveness of the closed-loop system when used unsupervised in the person’s own home over a four-week period. The participants continued all their daily activities and dietary patterns as normal during the study period, thereby assessing the system in a real-life situation without additional restrictions placed on the person.

However, there are some limitations, most notably the small sample size of only 25 participants. In addition to this, though the study period was fairly long, at four weeks, it was not long enough to monitor longer-term effects.

In particular, as the researchers acknowledge, though they monitored HbA1c, that shows the blood sugar control during the lifetime of the red blood cell, which is around four months, rather than four weeks.

This means that the short study design cannot reliably indicate whether closed-loop monitoring would have influenced longer-term blood sugar control as indicated by HbA1c.

A further limitation is that the technique was only used at night, between midnight and 7am, when each participant was resting/sleeping. It is unclear whether the technique would be responsive enough to cope with daytime activities that require greater adjustment of insulin control, such as eating and exercise.

Therefore, unfortunately, an insulin delivery system that would completely remove any need for the person to monitor their blood sugar or adjust their own insulin does not seem to be on the cards, at least for the immediate future. 

Despite these limitations, the results of this small study are encouraging. Studies involving a greater number of people and taking place over a longer duration are now required.

Article Metadata Date Published: Mon, 21 Aug 2017
Author: Zana Technologies GmbH
NHS Choices