Diabetes developments – by Simon O’Neill

In a reguSimon-O'Neill-Director-of-Hlar blog series, Simon O’Neill, Diabetes UK’s Director of Health Intelligence and Professional Liaison, rounds up the latest diabetes technologies, medicines and treatments. Simon was diagnosed with Type 1 diabetes in 1992. Next week’s post will look at what’s happening in research.

The artificial pancreas – one step closer

Research published in the New England Journal of Medicine has demonstrated the benefits of a closed-loop insulin-delivery system, known as the ‘artificial beta cell.’ Studies showed that using this system improved glucose control and reduced hypoglycemia compared with a simpler sensor-augmented insulin-pump. Such benefits have been demonstrated before, but this time the closed loop system was worn in ‘real time’

The results come from two 12-week multi-center, randomized controlled studies, one in 33 adults who wore the systems continuously for 12 weeks, the other in 25 children and adolescents who used the systems only at night. These are the longest such studies conducted to date, and they are the first in which people wore the systems while completely unsupervised, with no remote monitoring of their blood glucose levels and no limits on their food or physical activity. The studies, normally restricted to a hospital or hotel setting, also allowed the individuals to go wherever they wanted during the study.

Both studies used algorithms that calculate the amount of insulin needed to achieve desirable glucose levels. Every 12 minutes, based on data from a continuous glucose monitor (CGM), the algorithm calculated an insulin infusion rate that was automatically sent wirelessly to the insulin pump.

In the adult study they used a “hybrid closed-loop” approach. The adults still estimated their individual carbohydrate consumption and gave their own pre-meal insulin using a standard bolus calculator, as they would normally. However, the closed loop is continually working in the background, so, if they under or over-estimated their insulin needs, the system was able to compensate.

The study looked at the amount of time glucose levels were between 4 and 10 mmol/l for the adults and 4 and 8mmol/l for the children and adolescents. (The upper target was lower for the children/adolescents as they weren’t expected to be eating overnight). There was an 11% increase in the number of adults using the system who achieved this target range, and a 25% increase in the children and adolescents. This led to a reduction in HbA1c in adults from 60 to 56 mmol/mol (7.6-7.3%) and slightly less in the children/adolescents, from 62 to 60 mmol/mol (7.8-7.6%). There was also a significant reduction in the amount of time spent in hypoglycaemia below 2.8 mmol/l in the adult group but not in the younger people.

Interestingly, most of the components used, such as CGM and pumps, are already commercially available, so the challenge is now creating a commercial system that can also incorporate the algorithm.

Type 1 Diabetes – Causes?

Type 1 diabetes (T1D) is a complex condition and it isn’t fully understood why it develops in some people and not in others. We do know that genes play a role (at least 50 have been implicated) and T1D occurs more commonly in people who have a particular HLA complex. HLA stands for human leukocyte antigen, and antigens’ function is to trigger an immune response in the body. There are several HLA complexes that are associated with T1D, and all of them are on chromosome 6.  However, we know that for identical twins, who share a full set of genes, when one twin has T1D, the other twin only has it 30%–50% of the time.

So having the genetic tendency isn’t enough on its own and we know that the immune system, which produces antibodies to fight infection, is also involved. T cells are in charge of making the antibodies, and also help in fighting the infecting agent by recognizing antigens. However, if the infecting agent has some of the same characteristics as the beta cells, which make insulin, then the T cells can actually turn against the beta cells and destroy them. As yet, no-one knows what the infecting agent or agents might be but we know that this process is involved as people with certain specific antibodies are more likely to develop T1D.

Diabetes related antibodies include islet cell autoantibodies, insulin autoantibodies, autoantibodies targeting a form of glutamic acid decarboxylase (GAD), autoantibodies targeting the phosphatase-related IA-2 molecule, and zinc transporter autoantibodies. The more antibody types you have, the more likely you are to develop T1D. If you have three to four antibody types, your risk of progressing to T1D is between 60%–100%.

This process often takes several years, depending on how quickly the autoimmune response can destroy beta cells. It was generally thought that signs and symptoms of T1D appeared when approximately 90% of beta cells had been destroyed – though newer evidence suggests that they may appear even when 50% of the cells are still active.

Many people report developing the symptoms of T1D after an infection or a stressful event and believe that this is the actual cause of their diabetes. Although the infection or other stressor may have been the final straw for an already compromised pancreas, it isn’t actually the cause of the T1D, which is likely to have been developing for some time. What is most likely is that the stress/infection causes the body to respond, releasing hormones which make the body more insulin resistant, raising blood glucose levels. This means that the remaining beta cells are suddenly called upon to produce more insulin and cannot cope with the increased demand, leading to insulin insufficiency. This may be why T1D is commonly diagnosed in puberty, when the increasing amounts of sex hormones may make the body very insulin resistant.

Often, once insulin treatment has been initiated and blood glucose levels return to normal, the remaining beta cells are able to function again, reducing the need for injected insulin. This is known as the honeymoon period and generally lasts until the remaining islet cells are destroyed, sometimes for up to two years. In fact new research looking at ongoing insulin production shows that many people with T1D do continue to produce very small amounts of insulin for many years. This is most likely if diagnosis occurs later in life and is least likely to happen in those who are diagnosed at a very young age. This may be a reflection of how aggressive the autoimmune response has been.

Understanding how T1D progresses is vital in both trying to understand the cause of the condition but also to see if it is possible to identify people who are at high risk of this autoimmune response and whether the response can be prevented or slowed, or whether the honeymoon period can be maintained for longer. One study, called TEDDY, is looking at young children who are known to have the genetic vulnerability for T1D. They are then monitoring them until they are 15 to see if those who go on to develop T1D and share any exposure to environmental factors (such as maternal infections or childhood illnesses) that might be the trigger for the development of T1D in genetically susceptible individuals. The study has already suggested that giving new born babies probiotics (live bacteria that may aid digestion) in their first 27 days of life may reduce the odds of developing T1D by 60% for children with the DR3/4 genotype. Children without that genetic makeup didn’t see any benefit and use after the first 27 days seemed to have no effect.

If such susceptible people could be identified, there might be three opportunities to intervene. Some groups are looking at screening all children for T1D in a defined population, with genetic risk, to see if they can find those who already have the auto antibodies but who haven’t yet started to destroy their pancreatic cells. The second group would be those who were already destroying cells but still had enough beta cells to produce insulin, so have no symptoms of undiagnosed T1D. And the third group would be those who have symptoms but may be in the honeymoon period, with functioning beta cells still in place.

Of course there are ethical questions about screening for T1D when there is, as yet, no treatment that can effectively prevent development of the condition. However, many researchers feel that such early identification will help in the search for preventative treatments – such as the ‘vaccine’ that Diabetes UK are funding with the help of Tesco. And the opposite is also true – if we develop a ‘vaccine’ that can switch off or slow down the immune process that kills beta cells, how will we know who will benefit from it unless we start to identify them now?

It may be that this won’t be a long term cure for T1D, but if it could slow down the progress of the condition, it could ensure many years of near normal control which many feel would lead to longer term benefit. People with T1D in the Diabetes Control and Complications Trial (DCCT), who managed with much tighter blood glucose control for several years, appear to have developed a protection against complications, which has been called ‘metabolic memory’. In long-term follow up, even though their day to day control is now no better than the original control group in that trial, their risk of complications still seems to be lower. What seems important is to get that near normal control as early as possible and for as long as possible.

One of the big hopes is ‘big data’. Care.data is a programme of work led by NHS England and the Health and Social Care Information Centre which will securely bring together health and social care information from different settings. The hope is that such data will help us to better understand diseases and develop drugs and treatments that can change lives, as well as to plan services and compare the quality of care across the country. The big challenge is to get enough people signed up to such initiatives and enough relevant data to be shared to be meaningful and able to answer the questions that need to be addressed.

Potential new devices

Google have filed a patent for a blood glucose-measuring system that circumvents the need for finger-pricking. The device, known as the “Needle-Free Blood Draw” would work by sending a surge of gas into a barrel containing a micro-particle that pierces the skin. Once blood is released from the skin, it’s sucked up into the negative pressure barrel. According to preliminary designs within the patent, the device could work as a small cylinder placed on the fingertip or be worn on the wrist.

The patent doesn’t yet define exactly what medical application it will be used for, but it does state that “such an application might be used to draw a small amount of blood, for example, for a glucose test.”

Whether this patent will ever turn into a functioning device is unclear. Google hold patents on many ideas, and not all will be developed but it suggests that the firm is considering the diabetes market as a serious opportunity for them, following on from previous announcements and partnerships with Dexcom and Sanofi.

A much smaller company, working with Lilly, are trying to develop a smart insulin pen that can talk with a bolus dose calculator. Realising that the majority of people worldwide are still using insulin pens rather than insulin pumps, Companion Medical are trying to bring some of the benefits of pumps to those taking injections and hope to have a product approved in the near future.

The idea is a bluetooth-enabled, reusable insulin pen which is able to pair with a bolus calculator app on a smart phone. The hope is that the pen will be able to store a year’s worth of information; will automatically transfer information on the dose taken to the app (even if the pen and phone are separated and then come back in range); and temperature alarms if the insulin gets too hot or cold. The app tells you how much insulin you currently have on board, shows recent insulin doses and helps you calculate your next bolus dose automatically. The app will also give dose reminders and can produce reports to share with healthcare professionals.

Do it yourself?

People with diabetes are taking diabetes technology into their own hands and having good results with their work.

Nightscout is a crowd-sourced, DIY movement to remotely monitor Continuous Glucose Monitoring (CGM) data. It was created by some parents of children with T1D who wanted to be able to view their child’s CGM data remotely and who felt that the device companies were not moving forward on this technology fast enough. Their hashtag is #wearenotwaiting. There are now 15,000 families worldwide linked in to the movement, with around 850 families in the UK. It has continued to be developed, maintained, and supported by volunteers.


Originally, Nightscout was a solution specifically for remote monitoring of Dexcom G4 CGM data. Today, there are Nightscout solutions available for a variety of devices, including Dexcom G4, Abbott Freestyle Libre and Medtronic in the UK and UK parents are also working on the Medtronic MiniMed640G. The goal of the project is to allow remote monitoring of a T1D’s glucose level using existing monitoring devices and using freely available shareware which has been developed by members of the group. Different solutions exist for different systems and users may have to build simple electronic devices, using commonly available components and a bit of soldering! But help is always on hand for those who have never done anything technical before.

A recent development is being able to transmit data from the CGM system to a Pebble or iWatch. This means that information on your blood glucose levels is always accessible just by glancing at your watch and many people think this has been the most significant advantage of the system.

A recent study of some of the members of the Nightscout movement have shown some impressive outcomes. As well as the vast majority of users finding the technology “extremely empowering”, enabling parents to return to work, have a social life and sleep, using Nightscout has also seen reductions in HbA1c of between 0.7-1.2%.


You might also like