This is the time for diabetes technology to improve the story for people with type 1 diabetes, says paediatric endocrinologist and Chief Medical Officer for Medtronic, Dr Francine Kaufman. She takes us on a journey from the days before we knew the importance of glycaemic control to the technology advances that are bringing us ever closer to an artificial pancreas.
Back to the beginning
When paediatric endocrinologist Dr Francine Kaufman first started practicing there was a lot of debate about the importance of glycaemic control.
But the DCCT trial and the follow-on EDIC trial put an end to that debate by showing that what we do for patients now has an effect on their outcomes later on, Dr Kaufman told delegates who were attending a dinner symposium sponsored by Medtronic as part of the ISPAD/APEG conference in Brisbane last week.
“Intensive control somehow — we don’t know exactly what it is so we call it metabolic memory or some change in glycation or epigenetics — means that the intensive controlled cohort now almost 30 years later has less complications,” Dr Kaufman, told delegates.
“The message is that we have to improve glycaemic control and get HbA1c as good as it can be, potentially as early as possible,” she said.
We’re not there yet with HbA1c targets
Data from the type 1 exchange registry involving 26,293 subjects with type 1 diabetes over 67 sites in the USA illustrates that children do a little bit better at achieving glycaemic control than adults, particularly when they are at the age where parents can control their glycaemia.
However, like a lot of countries, the registry data shows that these children tend to get lost when they hit early teenage years, Dr Kaufman explained.
“In the USA 60 is the new 40, and 25 is the new 12…[glycaemic control] is not much different between a young adult and a teenager anymore… they tend to come back under control around the age of 30,” she said.
“But if you look overtime at HbA1c and at those enrolled between 2010 and 2012 and then outcomes two years after you can see that we haven’t actually done any better”, she noted.
“HbA1c hasn’t actually improved so although we keep making leaps in advancing all kinds of new insulins, new meters, new pumps, we haven’t actually changed the outcome of our patient cohort,” she said.
The microcirculatory and cardiovascular complications associated with diabetes are well understood but many studies now also show that children with type 1 diabetes have some hit to their cognitive function.
This was most pronounced in the children who are diagnosed with early onset diabetes, Dr Kaufman told delegates.
For instance a meta-analysis by Gaudieri et al.  published in 2008 in Diabetes Care found children with type 1 diabetes performed moderately lower compared to controls. It also found that seizures didn’t really seem to explain all the reduction in cognitive performance.
Another meta-analysis by Augustina M. A. et al.  also published in Diabetes Care in 2005 showed that chronic hypoglycaemia, diabetes complications and recurrent severe hypoglycaemia were considered to be linked to cognitive dysfunction.
“If we analyse what’s going on it’s perhaps a combination of severe hypoglycaemia, episodes of diabetic ketoacidosis — maybe even the initial episode — and perhaps some impact of hypoglycaemia in some patients,” Dr Kaufman remarked.
“We’ve got a lot of issues that we have to face with our patients particularly those diagnosed young…We see some cognitive outcomes, we know about all the other damage to the eyes, to the kidneys, to the nerves, to the cardiovascular system and it’s a pretty daunting issue.”
Technology gives better outcomes
“There are several studies that show that type 1 diabetes patients who are on pumps seem to do better on average in every age group compared to those not on pumps”, Dr Kaufman said.
For example a meta-analysis by Pickup et al.  published in Diabetic Medicine showed multiple daily injections (MDI) had a 4.19 times higher rate of severe hypoglycaemia than pumps.
And a more recent observational study from Sweden published in the BMJ this year by Steineck et al.  found that among people with type 1 diabetes use of insulin pump therapy was associated with lower cardiovascular mortality than treatment with multiple insulin injections.
Evidence also showed that patients who used continuous glucose monitoring (CGM) also do better as far as reducing HbA1c, she told delegates.
“While these were not randomised controlled trials it did show that when we use technology the outcomes are better for patients”, she said.
The journey towards an artificial pancreas
Medtronic began by developing an integrated pump and sensor as a platform where they could really start to build automation — with the goal of eventually getting to the full artificial pancreas, Dr Kaufman explained.
“The MINIMED VEO system was intuitive in that once a patient hit a set threshold insulin delivery stopped, allowing the patient to recover potentially from a minimal amount of hypoglycaemia, before it turned into a longer duration or an episode with a more significant impact”, she told delegates.
“The next innovation with the MINIMED 640G System^ was to prevent patients from reaching the threshold by stopping insulin delivery with a predicted algorithm 30 minutes in advance. Once the patients’ glucose levels recovered insulin delivery automatically resumes.
Some early studies showed that 80 percent of the time patients will not hit the pre-set low threshold with this system”, Dr Kaufman said.
“But this is all about stopping insulin… the bigger chasm to leap over is can we actually deliver insulin,” she said.
Towards a closed loop system
Dr Kaufman told delegates that the ultimate aim is to have a device that can determine every five minutes what packet of insulin it wants to deliver, or not deliver, to patients.
“Medtronic had originally planned to first develop an overnight closed loop system where there aren’t the disturbances of food or physical activity, she explained.
However, when they went to the FDA for feedback the regulatory body determined it was too difficult to define night because essentially “night” was different between people.
Coming in and out of a system like this could be harder for a patient than one that was set for a full 24 hours, they thought.”
This led Medtronic to develop a hybrid closed loop system where the patient: informs the device that they are going to eat, estimates what they think they are going to eat, and gives the pre bolus. The device then uses this information in an algorithm to determine all of the basal or background insulin needed for the rest of the day.
“The algorithm will look at what is the glucose level now, what’s the rate of change, what’s it predicted to be and how much insulin do I have on board and how sensitive am I to that insulin that I just delivered,” Dr Kaufman explained.
Medtronic is now in a pivotal trial with their hybrid closed loop system. The company is working towards getting an approval in the US first and then the rest of the world.
Continuing to look for better answers
The bottom line is that we now have a lot of data to support the importance of controlling diabetes, whatever domain you look across, said Professor Kaufman.
“With that knowledge, with the tools we have today we’re not able to get there. The Type 1 exchange shows us that for teenage and young adults we’re not able to get the targets we want,” she said.
Because of this it is compelling to continue to look for better answers, and while we would all like to find a cellular cure it’s eons away, she said.
“We are sitting now on enough data to show us that pumps improve things, and sensors improve things even more.”
Automation is the next 10, 15, 20 years maybe longer before we get the breakthrough to where we need to be with patients.
Their outcomes will be better and it should make severe hypoglycaemia and diabetic ketoacidosis a thing of the past, she said.
But the path to getting there has been challenging, and there will be more challenges along the way.
For instance, it has been frustrating that instead of going from A straight to B, there needed to be incremental advances.
“It’s taken so long because the algorithm had to be right, the settings had to be improved, the systems had to be innovative… but this is the time of diabetes technology to improve the story for people with type 1 diabetes” she said.
^ Components sold separately.
- PA Gaudieri, R Chen, Tammy F. Greer, Clarissa S. Holmes, Cognitive function in children with type 1 diabetes: a meta analysis Diabetes Care 31: 1892-1897, 2008.
- Augustina M. A. Brands et al. The effects of diabetes on cognitive performance Diabetes Care 2005; 28: 726-735.
- Pickup, J.C. Sutton A. J et al. Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion Diabetic Medicine 2008 25: 765-774.
- Steineck l. et al. Insulin pump therapy , multiple daily injections and cardiovascular mortality in 18 168 people with type 1 diabetes: Observational study BMJ 2015: 350:h3234 doi: 10.1136/bmj.h.3234.