Katya Rubia: optimistic about using neuroimaging as a treatment
As if she were professor of science fiction, Katya Rubia’s neurofeedback team has developed a spacecraft that can be controlled, merely by activating the right frontal cortex. Unnecessary to grab a keyboard or game controller: the rocket reaches the planets by brain power. The game is operated by a mechanism for fMRI-neurofeedback applied to children, developed at the Institute of Psychiatry, King’s College, London. These revolutionary steps in using neuroscience for self-regulating treatment – not only for diagnostics – are part of the study that was just about ready at the ESCAP Congress in Madrid. Professor Rubia managed to include preliminary data in her keynote lecture.
Dr Rubia has been researching neuroimaging in ADHD over many years. In 2010 she decided it was time to start applying the then substantially advanced knowledge in a clinical environment. Her lecture ‘From bench to bedside’ at the first Eunethydis conference in Amsterdam marked her run up to neurofeedback therapy as an alternative to not so effective behavioural treatment and perhaps one day even to medication. Ahead of the real results, Rubia says preliminary data are “very promising”.
The brain adapts to medication
Katya Rubia: “We have reasons to search for an alternative to ADHD medication. We do not know enough yet about the effect of medication on the developing brain. Our meta-analysis of PET studies shows that studies that included long-term medicated patients with ADHD have found that the dopamine transporters are abnormally elevated in the basal ganglia relative to studies that included medication-naïve patients, who have reduced levels. This means that long-term medication may upregulate the dopamine transporters in the brain. We do not know exactly what this means functionally, but if they are abnormally high, you will end up with less dopamine because these dopamine transporters take up dopamine from the brain. In my view this means that the brain adapts to the stimulant medication. Which is not so surprising – we know that the brain adapts to other types of medication, taken for years and years. An adaptation, however, would mean that long term medication may loose effectiveness. I note that the long term effects of psychostimulants have not been demonstrated in the literature – American studies have shown, for example, that the effect of stimulant medication is superior to behavioural treatment after one year, but after three years the effect is no longer superior. If long-term stimulants lead to brain adaptation, this would support the typical practice of drug holidays. To stop medication during school holidays is a very good idea because this prevents the brain from adapting to the medication. On the more positive side, structural MRI studies comparing medicated and non-medicated patients have shown that medicated patients have more normal brains than unmedicated patients, which would suggest that medication has a positive effect on brain development. However, we need large-scale longitudinal studies – which currently do not exist – to establish the long-term effects of stimulant medication on the brain development of ADHD children. Currently we do not know enough about this. Given the concerns about long-term effects of stimulants on brain development, it would be wonderful if we would have a way to treat these children without having to give them medication for a long period of time.”
From bench to bedside
“The knowledge that has been built up during the last twenty years, is almost completely based on group studies and we were able to conclude that these groups had abnormalities in this or that region… But this does not really help us to use neuroimaging to diagnose an individual patient based on brain imaging data. That is why pattern recognition analyses have recently been applied to imaging, which is a new way of analysing imaging data. This method does not look for group differences, but it looks for patterns which characterize each group and in particular brain imaging patterns that differentiate the patients from the controls. And once the computer has learned which pattern characterizes each group, then the computer can actually make individual predictions – that is why this is also called machine learning. So the computer can take one individual patient and – based on his brain structure or function pattern – it will say whether he belongs to the controls or the patients. This, I think, is promising. We have now done a few studies that we will also present at the ESCAP conference, where we used this method on structural and functional imaging data, and we found that we can predict ADHD diagnosis with about eighty per cent accuracy. Also important is that we can also predict differential diagnoses. So, compared to autism, based on the brain pattern, we could predict which child has ADHD compared to autism and not just compared to controls. I must stress that this is very early days, because this was still based on relatively small numbers subjects and of course these methods need to be replicated and it has to be shown that they are robust across scanners and across countries. This work still needs be done.”
“If we look at the future and these methods turn out to in fact have good individual predictive power based on objective brain imaging data then I think the biggest potential is in predicting treatment rather than diagnosis. At the moment there is no good method to predict which child will react to which treatment: will it respond well to stimulants or will it respond to atomoxetine? I think this has great potential because it is very likely that the brain is closer to predicting treatment response than behaviour, for indeed the brain is the target of the treatment. So in the future, assuming that this will all work well, it may become possible to do improved diagnostics in difficult to diagnose cases – based on a brain scan we would be able to say ‘This child is more likely to have ADHD than autism’ or vice-versa. Or we could say: ‘This child has this pattern of abnormalities, so we know that he will be more responsive to stimulants’, or to atomoxetine for example. This would save a lot of cost and time, because if you know which treatment is best to use at first you will save weeks of trying out one treatment or the other. People argue that our brain scans are expensive, but it is as expensive to treat a child with any treatment that does not work. And we may be able to draw conclusions very quickly – for example after a simple, two hundred pounds scan. Scanners also become cheaper over time and there are even other methods we could possibly use, such as near-infrared spectroscopy.”
Will child psychiatry become more technology driven?
“Yes, I think it could be. And I think it would make sense. We are becoming more and more aware that child psychiatric disorders are brain developmental disorders. So there is always a brain basis to these disorders. And I think the future will move towards individualized treatments because every child is different and every child will have a different biological make-up and different biological abnormalities compared to healthy controls. And it would be more efficient to treat each child depending on his specific biological abnormalities. If we have better knowledge of which biological abnormalities respond to which treatment, then we could help children in a better way. Better than just trying it out in the blind, without knowing what is the exact biology corresponding to that specific child.”
“But I am even more optimistic about using neuroimaging as a treatment: as a neurotherapy. That’s the other way: we can use neuroimaging as a means for diagnosis or treatment prediction, but we can also use it as a real treatment itself. This I find the most exiting avenue for the future. We have now spent several decades to understand that children who have ADHD have deficits in very specific brain regions. For example, in our studies the right prefrontal cortex is the key area that is abnormal in ADHD. And this appears to be specific to ADHD compared to some other childhood disorder such as OCD or conduct disorder. What we are doing next is trying to manipulate this region. Not by drugs but by fMRI neurofeedback and one could also use brain stimulation methods. So we have found this area is underactivated in ADHD and we have also found that stimulant medication consistently helps to upregulate the activity of this region. So we try to mimic what stimulants do – but without stimulants – using neurofeedback with fMRI which to my knowledge has never been done in children.”
This is brain control to Major Tom
We are now training children with ADHD to self-regulate the activation in this right prefrontal cortex by a computer game. So whenever they manage to increase the activity in this brain region they get rewarded by something that is happening in this computer game. I cannot say anything about the results because we have not finished this study yet, but the interim findings show that children are well capable of increasing the activation of this region after fourteen sessions of neurofeedback – handsfree so to speak. The game shows a rocket sitting on the ground and neuroactivity will start moving the rocket to launch and move up into the sky, passing several planets, the moon etcetera. The game is that everytime the child increases the activation in the right frontal cortex, the rocket will move further up into the sky. And if they decrease activation it goes back towards the earth. The children really like this – in fact this is the first study in my twenty years of experience where we did not have a single drop out. The kids simply enjoy doing this. We let them find out for themselves what works best – sometimes we give suggestions like ‘Try concentrating’ and sometimes that will work, but with others relaxation works. It seems they all need to find their own personal strategy to get the rocket moving. I have tried it myself and I did not think I could do it, but I could. And it is an unbelievable experience to actually control this rocket by nothing else but your brain activation. That is the fun of it. It feels absolutely like magic.”
“International cooperation is absolutely crucial in this line of research. Without networks such as Eunethydis, an international cooperation of ADHD experts from all fields, we would not be able to reach where we are now. Traditionally the neuroimaging field in ADHD was based on everyone doing twenty children in their own labs. It is much more efficient if we work together and join databases. For example there is a project called Enigma – chaired by Barbara Franke at the Donders Institute – where she collects all the structural data on ADHD across many imaging centers of the world. This is a fantastic effort to actually join imaging data across many, many centres including thousands of children. Those findings will be far more conclusive than small scale studies done at one individual centre. We have our challenges of course, like different centres use different diagnostic methods and so forth, but on the whole this international cooperation is very positive and the best way forward.”
Guidelines for policy
“Apart from their conferences, Eunethydis also runs an important European guidelines group for ADHD. I am sure that ESCAP could also encourage the development of European guidelines, not only for ADHD. These guidelines will improve the exchange of knowledge between researchers, but ESCAP could also have an impact on European policies. At the moment we struggle with these anti-psychiatry motions, particularly in ADHD where people keep on saying it is a myth or accusing us of drugging children etcetera, partly based on lack of information of all the scientific evidence we have. ESCAP could support the view that ADHD is actually a brain disorder – I hope my research has contributed to this. In the mean time the public still seems to believe that ADHD does not exist and that we are over-diagnosing and over-drugging these children. This is partly based on misinformation; people seem to think that stimulant medication is like cocaine. What they do not know is that these drugs are prescribed in a way that they have not shown to be cause addiction in ADHD. Large scale, recent studies show that ADHD medication does not enhance any addiction. Also the side effects are less dramatic than the public thinks. I agree with Jan Buitelaar, who is in the Eunethydis guidelines group, that the ADHD medication we use is one of the best drugs we have in psychiatry, it works in seventy per cent of the patients. Still it has a very bad name because it is associated with abuse. This misconception among the public should be taken away by providing proper, evidence-based information. And, even worse, a number of national governments seem to have embraced this misconception and build their policy on it. There was even a White Paper presented in the UK government, trying to reduce the so-called ‘medicalization’ of ADHD in favour of behavioural treatments such as diets which have not shown much effects. This paper is not based on best scientific evidence and exaggerates social causes and the benefits of diet and food supplements when we know that there is a lot of evidence for the influence of genetics and brain abnormalities. In fact, the Eunethydis guidelines group published a treatment guidance that shows that non-pharmacological treatments are less effective than treatment with medication. Hence diets and other behavioural treatments do not work very well in ADHD. On top of that, people do not realize that herbal treatments and behavioural treatments also can have side-effects.”
Impact on governments
“ESCAP has the authority to develop evidence-based guidelines for treatment that hopefully could have an impact on national governments’ policies and take away these misconceptions. It could be very powerful to have ESCAP organize European guidelines just like the Americans do in the AACAP with their Practice Parameters. Sharing data and hosting expert groups could be something that ESCAP could also enhance, but that kind of cooperation happens usually in specialized associations, and that is why I see the role of ESCAP more on a political level.”
“The problem is, for instance in ADHD, that the same disorder is treated very differently in different countries while evidence has shown that the prevalence of ADHD is very similar across European countries and the cultural differences are minimal. There still are countries where child psychiatrists hardly prescribe stimulant medication for ADHD and in some places the only treatment is psychotherapy. It would be good, and better for the patients if here were generic global or European guidelines based on all the expertise knowledge we have. ESCAP could work together with Eunethydis and the other specialized groups to push ahead on the political aspects.”
Read the abstract, view the Madrid slides and watch the TV Interview: "Brain imaging in ADHD: disorder specificity, medication effects and clinical translation" (Madrid, 2015).