Diabetes developments (new technologies, medicines and treatments) – by Simon O’Neill
In a regular blog series, Simon O’Neill, Diabetes UK’s Director of Health Intelligence and Professional Liaison, rounds up the latest diabetes news.
This week Simon updates on new technologies, medicines and treatments.
Verily (Google) Smart Lenses
Back in 2014, I wrote about Google’s contact lens blood glucose monitoring solution and explained that this wasn’t a new concept, with other companies looking at the technology since 2011 but ‘de-prioritising’ development at a later date. This was hailed by several papers as the next great breakthrough in diabetes care.
Verily, the new name of Google’s Life Science Section, began working with Novartis on this problem in 2014 but little has been heard about the project since. However, it is running behind schedule, which perhaps isn’t that surprising, as there are many obstacles to overcome in producing such a novel technology.
One issue is that the microchip that would crunch the data needs to be incredibly small to fit into the lens, as well as having to house memory and have its own power management system. The company will also have to create a minute battery that can be fitted into the lens and somehow recharged wirelessly. There are also questions about the relationship between the level of glucose in tears and in the blood, so that a valid comparator can be made in order to give useful information to the user. It isn’t much help knowing there is a lot of glucose in your tears if that has no bearing on the amount in your blood.
There is also the issue of wearing these lenses overnight, as many current lenses have to be removed daily. And, of course, with all that technical stuff built in to the lens, you will still have to be able to see through it, which raises the interesting question of how will this work with those who already wear contact lenses for sight correction?
Novartis have now officially announced that the project is behind schedule and that no trials in patients are yet in the pipeline. At the same time Verily are working with several other diabetes partners on technology which may be nearer the market – so perhaps we will see the Google lenses finally bite the dust.
A recent workshop at the European Medicines Agency (EMA) has produced a report on the future role of personalised medicine and the actions that will be needed to move towards developing, authorising and using this type of technology.
So what is personalised medicine? We know that certain people react to drugs in different ways but currently all advice on dosage and side effects is generic, based on a large research population who have trialed the drug. Over 6% of acute hospital admissions are as a result of severe adverse reactions to medicines.
One example is codeine. The speed with which this drug is converted into morphine in the body appears to depend on genetic factors, so for some people a standard dose will have no pain reducing effect while in others it can lead to adverse effects. So is there a way that we can tailor the medicine to meet the particular needs of an individual?
Although there isn’t an agreed definition of personalised medicine, it is often described as a model which considers an individual’s phenotypes (such as our physical appearance) and genotypes (our genetic makeup) to tailor treatment which is right for that person at that time. This type of approach is being used in the development of some rare cancer treatments or ‘orphan drugs’, which are used in very rare conditions.
The idea is an interesting one, and it is certainly seen as the future direction of medicines, but the workshop highlighted many of the challenges that this model raises, not just in developing new medicines but also in how we test them, approve them and make sure that they reach the people who will most benefit from them. Bodies such as the EMA and FDA are already working to help the development of these approaches, developing reference standards and libraries of big data around genetics and biological samples.
But just having large amounts of data isn’t enough, and we will also have to develop new ways of analyzing data to discover relationships and patterns that can be used in drug development. There is also the question of data confidentiality and how personal information can be protected.
We will also need to do much more to understand how genomic data can help us to identify predictive biomarkers that will allow us to accurately and reliably make diagnoses, rather than just relying on symptoms as we largely do now. Can we also develop tools that enable us to identify patients at risk (much as we are starting to do in T1D) so that treatments can be developed for prevention as well as treatment or cure?
The workshop also discussed the potential issues with developing appropriate clinical trials, as looking for very specific groups within a larger population will need smaller trials on those groups, making it harder to gain sufficient evidence of safety and effectiveness. Trials will also have to become much more patient focused, answering real life questions that matter to individuals rather than looking for an overall clinical effect, such as lowering HbA1c.
Medicines targeted at very specific groups of patients will also be more expensive, as the pharmaceutical companies will not have the same large markets to recoup their costs. So can hard pressed health systems actually afford this type of technological advance, even if it may mean better outcomes for some individuals?
Finally the group considered the implications for patients. They will need to be able to make informed decisions about their treatment options, but will they have the knowledge to do this? How do we educate different groups of people, so that they can make the best choices for their own health – and how do we educate HCPs so that they understand who will be best suited to a certain treatment option? Diseases are currently classified by signs and symptoms rather than the underlying molecular or genetic causes, so we will need to think about how we train HCPs going forward.
There is certainly much to think about and many issues to overcome – but the potential to have much more effective medicines in the future that are really targeted to the needs of the individual.
Apple Watch for diabetes?
It has been reported that the CEO of Apple, Tim Cook, has been seen wearing a prototype blood glucose device that pairs with the company’s existing Apple Watch. Speaking with students at the University of Glasgow, he said that he had been wearing a CGM device to help understand how his blood sugar responded to particular foods, and how he could adjust his diet to control it. Interestingly his comments also included the assumption that if you could monitor your blood glucose in reaction to food, you might be able to take steps to adjust your lifestyle before, presumably, T2D developed.
Earlier reports had stated that Apple had hired a team of experts to create a non-invasive glucose monitor, so the assumption has been that this technology might be on the horizon and reports of feasibility trials have already circulated.
The health benefits of the Apple Watch are certainly of interest to the company and many developers are producing apps that will use the capabilities of the watch to give added health benefits. A recent study showed that the Cardiogram app, used via an Apple Watch, correctly identified atrial fibrillation with a 97% accuracy, 98% sensitivity and 90.2% specificity. This is a problem that affects over 2.7M people in the US, so a huge potential market for Apple.
Metformin is the first line medical treatment in the management of Type 2 diabetes and is believed to work by reducing the amount of glucose made in the liver, which would in turn lower blood sugar levels. However, scientists know that slow-release metformin is equally effective even though only small amounts of the drug reach the liver and metformin also works in people who have genetic variants that stop it reaching the liver at all.
Recent research into T2D remission through weight loss have noted that, as well as losing weight, the gut microbiota also appear to change and could be part of the reason why the remission has occurred. Several researchers have therefore investigated whether there is another method of action in metformin and have studied its effect on gut bacteria.
40 volunteers who were newly diagnosed with T2D were randomly assigned to take metformin or a placebo for four months as well as following a low-calorie diet. During the study gut bacteria were collected and assessed. Those taking metformin saw a dramatic change in the bacteria present, with particular growth of strains of Akkermansia and Bifidobacterium. Lab experiments also demonstrated that both types also grew faster in the presence of metformin.
Interestingly the research team then took stool samples from three people both before and after a course of metformin. These were then fed to mice that were on a high-fat diet, to mimic T2D, using a technique known as faecal transplant. This is thought to pass beneficial bacteria from the gut of a healthy individual to a sickly one. Mice that received the transplants from people on metformin seemed to improve their glucose tolerance, while there was no effect on the mice with the non-metformin samples.
The researchers suggest that this shows that metformin at least partly works by encouraging the growth of gut bacteria such as Akkermansia. However, they still don’t know how the bacteria might be improving glucose tolerance or whether people with T2D might benefit from modifying their diets to encourage the growth of these particular bacteria.
In separate work, the role of metformin in people with T1D has also been considered. A large meta-analysis, looking at the use of metformin in children and adolescents with T1D found a slight reduction in daily insulin dose and a slight fall in BMI but no real impact on HbA1c. However, there also seemed to be a slight increase in both hypos and DKA. A small study looking at adults found similar short term benefits but that these had disappeared within 10 years. So there may be a potential benefit of metformin in certain people with T1D, perhaps especially in those who are overweight, but more and larger research studies are needed to prove this.