DUKPC highlights part 1: Untangling the complexity of diabetes
Earlier this month – along with over 3,000 healthcare professionals and researchers – we packed our bags and headed to Liverpool for our Professional Conference (DUKPC). Over three action packed days we heard from leading experts on what’s new in diabetes research. Here’s what we learnt about progress to untangle the complexity of diabetes.
Getting personal with Type 2 treatments
Precision medicine was a hot topic this year. It’s all about moving away from a one-size-fits-all approach and tailoring treatments to the individual.
Prof Mark McCarthy brought home just how complex Type 2 diabetes is. In most people there’s no single factor which explains why they develop the condition. Instead the cause of an individual’s Type 2 is driven by a muddle of different risks, and everyone with the condition has a unique mix of these factors. Scientific advances are beginning to help us unpick someone’s exact mix, which would allow doctors to think much more cleverly about how to best treat them.
Genetics is an important part of this. Prof McCarthy told us there are over 400 genes that bump up your risk of developing Type 2 diabetes. Researchers are developing genetic risk calculators, where they can look at how many of these genes someone has and add them up to determine their risk of the condition. And potentially we could parcel this risk up further, and pinpoint who has genes that put them at highest risk of quicker progression or complications.
We also heard that genetics can determine how well people with Type 2 responds to treatments. Prof Ewan Person has shown that based on your genes, you need smaller or larger doses of metformin to do the exact same job.
Prof Andrew Hattersley rounded off the session by thinking about precision medicine on a simpler level. He’s shown that characteristics like our age, gender or BMI, can also help us predict treatment responses. For example, Prof Hattersley found that men of a healthy weight seem do better with Sulphonylureas, while women who are overweight do better with Thiazolidinediones.
So we can start to think about personalising treatments now. But to make things more precise, and make sure people with Type 2 diabetes benefit as much as possible, we need to combine what we know about an individual’s characteristics with information about their genes.
At the moment we don’t have a way to easily and cheaply read everyone’s genetic information. Once this becomes part of routine health care, scientists think precision medicine in diabetes will really take off.
Breaking down into subtypes
Just before DUKPC 2018, research hit the headlines that suggested Type 2 diabetes could be made up of four subtypes. At this year’s conference, PhD researcher John Dennis ran through what he’s found out since then.
He wanted to double check that the same subtypes, or clusters, could be found again in a different pool of people with Type 2 diabetes. To do this, he looked at existing data from over 4,000 people with Type 2 who had taken part in a previous research study. He did find similar clusters and a similar proportion of people in each group, backing up the previous study.
He then asked if the subgroups were actually useful – could they help improve and personalise Type 2 diabetes care? John found the clusters could help predict someone’s progression of Type 2 diabetes (measured by increases in their HbA1c). But in fact the age at which someone is diagnosed predicts this just as well. And it’s way easier to use age than trying to figure out which subgroup someone’s in.
He then found that the clusters weren’t good at predicting who would go on to develop kidney disease. Or at predicting how people would respond to different Type 2 medications. Once again using simple characteristics – like age, gender and BMI – were much more helpful.
When it comes to breaking down Type 2 diabetes John concluded, “all that clusters is not gold”.
Getting the diagnosis right
C-peptide is a molecule our body makes while it produces insulin. You can use C-peptide measurements to find out how much insulin you produce: people with Type 1 diabetes have very low c-peptide levels, while people with Type 2 have higher levels.
This also means a C-peptide test can differentiate between Type 1 and Type 2 diabetes. In the past, these tests were expensive and inaccurate – but scientists in Exeter have changed this.
Prof Andrew Hattersley explained that the first test they developed used urine samples; stable at room temperature for a few days, and painless and easy to collect. Then, they discovered a way to make blood samples more stable too. You could collect the samples during a diabetes clinic, and patients didn’t need to fast beforehand. Now, they’re developing a finger prick blood test that’s not quite as sensitive but it stable for 10 whole days at room temperature – great for collecting and sending samples across the UK.
So we have a quick, easy way to distinguish between Type 1 and Type 2 diabetes – something lots of healthcare professionals need support with. When should they use it?
Dr Angus Jones highlighted that people with Type 1 diabetes still produce insulin during the ‘honeymoon phase’, so it’s not a great test at the point of diagnosis – that’s why we need antibody tests. But he suggests that if someone is diagnosed with Type 2 diabetes and needs insulin therapy within three years, that’s a warning sign for a potential misdiagnosis and a C-peptide test would confirm the right diagnosis.
One of Prof Mark Strachan’s patients was diagnosed with Type 1 diabetes, but they finally realised she had monogenic diabetes – Prof Strachan wishes they’d spotted it sooner.
So he set out to run C-peptide tests on all of their 1000 patients with Type 1 diabetes, to check if anyone else had been misdiagnosed. So far, they’ve tested over 750 people and have ‘reclassified’ several people with Type 2 diabetes or rare forms of genetic diabetes. He was surprised by just how many of his patients didn’t have Type 1.
Bottom line, healthcare professionals are struggling to correctly diagnose every case of diabetes – and we have a reliable test that can help. But it’s really important that we think about what it means to reclassify someone’s diabetes, so that they’re given all of the support they need in the most sensitive way.
Pregnancy and diabetes
For all mothers, the weight of their baby depends on the length of the pregnancy, and other factors such as their own height and weight. But women with diabetes are more likely to have a baby that is larger than usual for the number of weeks of pregnancy. This carries risk for both the mother and baby during pregnancy and birth. That said, most babies who are large for their age are delivered without any problems at all.
One way of reducing the risk of having a large baby is to try and keep blood sugar within safe ranges, and researchers are working to make that simpler with diabetes tech. The CONCEPTT trial compared how well 248 pregnant women with Type 1 diabetes were able to control their blood sugar using either insulin injections, or an insulin pump. They also measured other important pregnancy factors like blood pressure, blood sugar levels of the babies and admissions into neonatal intensive care units.
Mothers using insulin injections had better blood sugar control and were less likely to have high blood pressure during pregnancy. Their babies were also less likely to have low blood sugar or be admitted into intensive care for more than a day after being born.
Many would assume that insulin pump therapy would be more effective for managing blood sugar, and in turn, improve the health of the mother and her baby. In CONCEPTT, that wasn’t the case. Nevertheless, 93% of the women who took part had successful, healthy births irrespective of which insulin delivery method they used. But there’s still more we need to learn about using the latest tech in the best way to help make pregnancy in diabetes simpler.
Keep an eye out next week for part 2 in our DUKPC blog series, covering new diabetes treatments.