Artificial platelets, and angry primates

An artificial pancreas now approved by the National Institute for Clinical Excellence - NICE - is set to help tens of thousands of children and adults living with type one diabetes. The ground-breaking device monitors a person’s blood glucose, and automatically adjusts the amount of insulin given to them through a pump. It’s all done wirelessly through bluetooth. It was developed by Cambridge University’s Professor Roman Hovorka...

Roman - So what we are talking about is a system which can control glucose levels in people who have type one diabetes. So we have a sensor which reads glucose. It goes into the computer program interpreter, which analyses the data, and then it has a device which delivers the right amount of insulin to give.

Chris - How does it all work and wire together and operate?

Roman - So we live now in the 21st century. So all the communication is Bluetooth. The computer program can sit on your smartphone.

Chris - In essence, then, a sensor is tasting the tissue fluid to work out what your blood glucose is. It's telling the computer program running on the smartphone that level all the time. And the smartphone is then computing what level of insulin to tell the pump to inject. Why is that better than the way we used to do it where you would just prick a finger, test a blood sample, inject insulin?

Roman - The reason is because we are all different and, not only we are all different, we are all different all the time. And glucose levels in your body need to be well controlled to avoid long term complications or health consequences. And it relates to the problem with your feet, with your kidney, with your eyes, with your heart. And it is very difficult for people to do so without this automatic management. They can do it, but it doesn't generally burn themselves out. Like parents would be waking up four or five times a night every night to manage and to check glucose levels. This device can do it on its own.

Chris - Presumably you've now got data showing that people who use this system achieve better long-term control of their glucose levels and therefore we think are likely to have better health outcomes compared to the more traditional way of managing diabetes.

Roman - So we should not forget that before 1920, when insulin was discovered, people uniformly died. Since then, there was a gap between the life expectancies, which until recently was about 20 years with type one diabetes. So that gap is closing and it's closing because we have better management of glucose levels and this technology will close the gap faster as well, but also will avoid the need of the mental burden of the challenge of managing it. Not just for the person with type one diabetes, but the whole family. So it's fantastic news before we have a cure, which would be fantastic as well. But right now this is the best thing we have before cure.

Chris - Is there a risk that people might get into bad habits if they know that there's something taking care of their blood glucose level all the time? Where previously they would've behaved themselves, is there a danger that they might adopt bad eating habits, just bad lifestyle habits, which are a plague of modern life anyway, but because they've got this looking after them, they can afford to take those risks, which previously they wouldn't have done.

Roman - Yeah, so our data does not show that. If anything, the amount of insulin needed is very similar. So we don't give more insulin without the system.

Chris - What's its fail safe mode? In other words, if it goes wrong, the phone goes flat or something. What does it do then to keep you safe?

Roman - That's a really important issue. Safety is the most important. On the insulin pump, which is part of the system, there is a pre-programmed rate. So if everything fails, people will still give the expected right amount of insulin.

Chris - With any kind of medical intervention, one always has to look at the bottom line - what is the cost? So what's the upfront cost of doing this? And is that more than mitigated by not just obviously the benefit to a person's quality of life, which you've outlined, but also if they avoid a lot of those healthcare related problems later because they've got better glucose control, does the saving later more than mitigate the increased upfront cost?

Roman - So the current calculation of the whole system is about 5,000 pounds per year, but some of the cost is already included because people are already allowed to use some of the technology itself. So the additional cost is much less than that. And indeed, if you look at the calculation, how much we save in terms of saving for long-term complications, how people live longer as well. So it's something which NICE considers when they do review the technology. And that was the reason why I accepted it because it is cost effective, the treatment as well. One thing which is becoming quite important. In the past it would be really just hard calculations. So you look at the cost. What is, for example, difficult to cost is the quality of sleep and quality of life. And I think it's very nice that the NICE right now acknowledges and they took the feedback from people with diabetes and their families and they didn't look just at their hard figures, but they also look at it, 'Hey, people can go back to work.' The moms, they can sleep through the night. All these things need to be added just to these hard calculations.

Large parts of North America witnessed a total solar eclipse this week - which occurs when the Moon passes between the Earth and the Sun, blocking our view of the Sun as it passes. The Moon’s ability to block our view of the Sun is an incredible cosmic coincidence. It happens because - by chance - the Sun and our Moon appear almost exactly the same size in Earth’s sky. So it’s perhaps a good time then to reappraise how the Moon was formed 4.5 billion years ago and what’s happened to it since, including the possibility that it’s turned itself inside out in the past! Here’s Jeff Andrews-Hanna at the University of Arizona…

Jeff - We're basically trying to understand the earliest evolution of the Moon and how the Moon got from what we think was initially an entirely molten body to the body we see today orbiting the Earth.

Chris - Because the Moon is thought to be the product of a giant cosmic collision. In fact, some people have written books called The Big Splat and things like that some four and a half billion years ago, isn't it?

Jeff - The Moon formed in some manner of giant impact. The details are uncertain. In one version of the story, a Mars sized body crashes into the proto Earth and the outcome of that is material gets ejected out into space and that material out in space very quickly, very rapidly and violently coalesces to form the Moon. But it's such a rapid and violent process, that initial Moon that formed orbiting the Earth would've been mostly or entirely molten. We say that the Moon had a magma ocean that may have extended to hundreds of kilometres depth, and may have gone all the way down to the core.

Chris - What is the evidence that that's either true or may be uncertain?

Jeff - The story of the magma ocean really begins with rocks that were brought back by the Apollo astronauts more than 50 years ago. That's where we really confirmed that the crust of the Moon, the bright surface of the Moon that we look up at at night, is made of a mineral called plagioclase. And plagioclase is really important because as that magma ocean crystallised, at some point plagioclase starts to form and it's light, it's buoyant, it floats. So it floats to the top of the magma ocean. That's why much of the surface of the Moon looks so bright when we look up at it at night. And that really, I think, clinches the story for the magma ocean. But the rest of that solidification process is really interesting because, as the Moon crystallises, different minerals form at different times and the final outcome basically leaves us with a Moon that is just wildly unstable and out of equilibrium. Because aside from that bright crust that we see sitting on the surface today, below that crust is the lunar mantle. And in the mantle after the magma, the ocean solidifies the most dense material, minerals rich in titanium and iron, are all sitting at the top of the mantle and the least dense material is all sitting at the bottom of the mantle. And that's just not where it wants to be. And so there's long been this idea of an overturn of the lunar mantle. That after the Moon solidifies, basically the mantle turns itself inside out. Everything on the bottom floats to the top. Everything at the top sinks to the bottom. And that's basically going to be the biggest thing that happens to the Moon between the Moon's formation and today. And it really sets the stage for everything else that happens.

Chris - This is a theory, of course. You've done some modelling to come up with this idea. Does this give us some testable hypothesis though? Some experiments we can now do in order to see if what you are proposing is what really happened?

Jeff - You know, there are a number of different models that have looked at this process of mantle overturn on the Moon. And this is a problem that goes back decades. For us though, this story really begins with a paper that some of my colleagues wrote a few years ago. So Nan Zhang had a model of this magma ocean overturned, and he published that work. And I looked at the figures, he showed the patterns of what he showed, and it reminded me of something that I'd seen in data from the Moon, in gravity data in particular. In his model as this really dense titanium rich material at the top of the mantle sinks into the interior, it sinks into this really unique pattern. It basically makes these sheetlike downwellings, it's basically like waterfalls or avalanches of titanium rich material. But it's not quite perfect. At the end of that process in his models, there are these long linear zones of dense, titanium rich material that are still sitting at the top of the mantle that didn't quite manage to sink down into the interior. And I saw those results and it immediately reminded me of something I'd seen in gravity data. About a decade ago I was analysing gravity data from NASA's GRAIL mission. And here we're looking at just subtle, subtle variations in the lunar gravity field. It's really important data because it tells us about what's going on below the surface. It basically tells you the distribution of mass and density beneath the surface. And when I analysed that gravity data 10 years ago, I saw this really interesting pattern on the near side of these long linear gravity anomalies intersecting in this polygonal pattern. And what I was seeing in the gravity data 10 years ago looked exactly like what Nan was showing in his Geodynamics models today,

Chris - You say on the near side of the Moon because the Moon is a bit weird, isn't it? Because although it's tightly locked, it always shows us the same face and that's where we get this whole concept of the dark side of the Moon. It's not dark at all, it's just that we don't see it. The other side looks very different to the side that we see now. Has that got anything to do with this?

Jeff - Yes, very much so. The Moon is fundamentally lopsided in almost every respect. When we look up at the Moon, we see these bright and dark patches and the dark patches are these mare volcanic lava flows. And there are very few of those lava flows on the far side. But, but this, this asymmetry, this lopsidedness goes beyond that. The near side is low in elevation and the far side is high in elevation. The near side has a thin crust. The far side has a thick crust. And very interestingly, the near side is really rich in titanium and the far side is generally poor in titanium. And so in these models, there's a giant impact on the far side that makes this giant impact base we call South Pole-Aitken. And that actually triggers, it catalyses, this migration of dense, titanium rich material towards the near side. And then it's on the near side that that all sinks into the interior.