New Treatments: Transcranial Magnetic Stimulation
In the second in our series on novel treatments, Anorexia Myths talks to consultant psychiatrist Dr Leo Chen about an exciting new trial looking at TMS for anorexia nervosa
AM: For a lay person, could you explain what TMS is?
LC: TMS stands for transcranial magnetic stimulation. The basic premise is that when you deliver an electrical current through an electrically conductive object, such as a magnetic coil, each time that current changes, there is a very brief flux of magnetic field around the magnetic coil. And if another conductive material is placed next to the magnetic coil, that magnetic field can induce a secondary current in the conductive tissue that it’s abutting and induce a current in that secondary circuit.
The brain is very electrically conductive. Firstly, human tissue contains a lot of water, and water is extremely conductive. More pertinently, the brain operates through a very complex and sophisticated array of connections, networks and circuits. If you applied TMS to say, my forearm— where there are networks and circuits of nerves that convey arm position, pain, temperature control, muscle movement — you will only affect the nerves that do those things. If you apply TMS, or any kind of electrical energy, to the brain, you can affect the circuits that are regulating much more complex human functions, such as thought and behaviour.
AM: So, because the brain is the centre of the nervous system, you’re targeting the source of whatever’s going on elsewhere in the body?
LC: Yes, that’s a good way of putting it. With TMS, you can deliver thousands of pulses within a short period of time. And when you fire lots of pulses, it affects the region that you’re stimulating and the networks that it’s connected to.
AM: How do we think TMS works on the brain?
LC: When you deliver lots of pulses, you strengthen connections and you strengthen the connectivity and the flow of electrical signals, nerve signals, along these networks. Moving away from scientific wording, I think of it as clearing a road or shovelling snow off a road.
AM: How established is this treatment?
LC: TMS has been around for about 40 years. The first machine was engineered in Sheffield, in the UK. Over the past 30 years, there have been lots of research efforts to see what would happen if we applied repetitive pulses to regions and circuits that are implicated in emotional regulation and depressed mood, to see if it might be an effective treatment for depression. In this time, it has been established as an evidence-based treatment for depression, particularly depression that has not been responsive to antidepressant therapies.
AM: So, we know that TMS can work for depression, but presumably it doesn’t work for everyone?
LC: When you apply TMS to depression networks in the brain, about 50 to 60 per cent of patients get considerably better.
AM: And do they have to continue using the TMS, or is it a limited treatment?
LC: It depends. People generally will have a course of 30 treatments, and the standard way is to do the 30 treatments five days a week over six weeks. But there are other protocols now where you deliver more sessions per day, and it seems to be just as safe and effective, if not more effective.
AM: How did you become interested in TMS as a potential treatment for anorexia nervosa?
LC: I can preface that by explaining how I got interested in TMS. I started out in my career as a clinical psychiatrist. I realised the limitations of many of our therapies, that they aren’t always as effective as we want them to be. I became very excited and interested in optimising our existing treatments or developing new treatments. I was exposed to TMS as part of my training, and then in my clinical practise. And I decided to home in on this space to work on improving its efficacy and efficiency in treating depression. That was a fair chunk of my PhD research.
More recently, our research lab in Melbourne was blessed to have received philanthropic funding from the Li family, who wanted to give back to mental health research, in particular the anorexia nervosa space, specifically looking at clinical trials. Their aim was to test novel therapies that can help people living with this condition rather than looking at pre-clinical animal models or genetic studies, which might have translational potential, but that translational potential usually comes many years down the track. They wanted something that people living with the condition can access now. With my research background, and expertise in brain networks and modulating these networks with brain stimulation, I was naturally inclined to conceptualise anorexia nervosa as a disorder of brain networks and problem solve how we can directly modulate this network for therapeutic effects.
AM: Is there a specific area of the brain that you are targeting for anorexia? If so, what is it?
LC: Yes, before anything else we need to identify the relevant network to target. This is what I think is probably the most exciting aspect of what we’re doing and why we’re cautiously confident we could see positive treatment outcomes.
As I said, with depression we’ve got a treatment that’s at least 50 to 60 per cent effective when we target depression brain targets — the left or right prefrontal cortex. Research efforts have ‘re-purposed’ those brain targets to see if applying TMS to them could treat other psychiatric conditions, such as post-traumatic stress disorder, anorexia nervosa, anxiety disorders etc and the results have generally been underwhelming. This makes sense to me because the brain is a complex array of networks. It is important to identify the right network for the cognitive and behavioural symptoms you are hoping to affect.
AM: So, how do you go about finding those networks?
LC: The technique called causal lesion network mapping. The premise is fairly simple. We find as many cases as possible of brain lesions (sites of brain injury) for a given behaviour, eg anxiety, depression, right arm paralysis, and map these sites to a normative brain atlas – think of it as a brain wiring diagram that is representative of healthy populations. This normative brain atlas is provided by the Human Connectome Project. To ensure we find as many cases as possible, published cases of these causal lesions are usually searched using comprehensive and systematic scientific literature searching methods. Mapping these lesions to a brain atlas allows us to see what networks they are connected to. We then overlap all these networks connected to each causal lesion, to statistically derive a causal brain network for that behaviour.
AM: So, you do a kind of Venn diagram and then you say, this is the common area to these patients.
LC: Yes, by overlapping the networks connected to each causal lesion, we can derive a causal brain network that is commonly connected to each causal lesion. When you do a systematic search for brain lesions for a particular behaviour, they’re always what we call heterogeneously distributed, which means they’re distributed across the brain, not localised to one brain region. But you can map these lesions to the networks that they’re connected to, and more often than not they are connected to the same networks. So over time you can overlap all the cases of brain lesions and the networks that they’re connected to and derive a causal network for that behaviour.
This was the first step in designing our clinical trial to treat AN. Dr Shan Siddiqi and Dr Joseph Taylor are psychiatrists working at the Brigham and Women’s Hospital in Boston, USA. Joe’s team started this work about 18 or so months ago. We did a systematic search in the published literature of all causal lesions for anorexia behaviours. We found 15 cases. As we expected, they were diversely distributed across the brain. Even though these published brain regions were not localised to one part of the brain, the published cases all described how, after a brain injury, people went on to develop behaviours of food aversion, weight loss and, in some, distortion of self-perceived body image.
As described earlier, we then mapped these lesions to find what circuits they were connected to. 15 out of 15 cases, very encouragingly, showed peak connections to a brain structure called the lateral hypothalamus. This is the part of the brain that drives appetite. Something else that was a little more surprising, but still really cool, was that 15 out of 15 cases showed a peak negative connection with the fusiform gyrus. This was unexpected. The fusiform gyrus does several things, but one of its main functions is in facial recognition. This led us to think that maybe we’ve uncovered a brain circuit that plays a role in driving appetite, but also in how people see themselves and appraise their facial appearance and maybe body image.
AM: So, do you think this could explain this idea that somebody with anorexia is not seeing themselves correctly?
LC: No, I think it’s a bit of a stretch to say that this dysfunction, and this part of the brain, is making people have distorted visual appreciation of what they’re seeing. I think it’s probably more likely that this part of the brain is very active in this population because they’re always thinking about or looking at themselves.
AM: So, presumably these are the areas you’re going to target with the TMS?
LC: No. The electromagnetic energy applied by TMS can only penetrate 2-3cm below the coil surface. In other words, it can only stimulate superficial brain cells lying around the outside of the brain. Both the hypothalamus and the fusiform gyrus are deep brain regions. We will therefore stimulate a superficial part of the brain that is connected to our causal brain network, of which the hypothalamus and the fusiform gyrus is a part.
AM: Are the patients in the trial going to be very underweight or are you looking at people who are more weight restored but still have AN?
LC: We’re specifically looking for people who have established anorexia nervosa and whose BMI is within a particular range in the underweight category.* We want to see if modulating the causal brain network we’ve uncovered could help people with their challenges with food, eating and restoring and/or maintaining a healthy weight.
AM: Will the people in the trial have a psychological treatment alongside the TMS?
LC: No.
AM: And is it a blind trial? Will there be a placebo arm?
LC: This is an open-label trial, so everyone will be getting the real TMS. This is a world-first proof-of-concept — meaning we need to first see if our approach is safe and effective. If so, the next step will likely be a larger, randomised trial with a sham stimulation control built into the study design.
AM: Can you describe what happens in treatment? Presumably the patients wear some kind of headset that delivers the treatment? Is it painful?
LC: The patient sits on a comfortable recliner chair and there’s a magnetic stimulator that’s connected to a coil. A coil is placed on the patient’s head, and it goes to work. The best way I can describe the sensation of receiving TMS is that it feels like someone tapping you very quickly on your head with a pencil. Each treatment session goes for about ten minutes. Overall, it is a very well tolerated treatment.
AM: What else can people in the trial expect?
LC: Consented participants will start by undergoing an MRI scan. We’ll take their height and weight, and we’ll do a variety of anorexia related cognition and body image surveys, psychological ratings, quality of life ratings. We analyse their MRI scan and plug our AN causal network into it. We then derive the individualised TMS target that is maximally connected to the causal network. The participant will receive 30 sessions of TMS applied to this target over three weeks. Each treatment session lasts for about ten minutes. The participants have a one-hour break and then have their second treatment for the day. After the second session, the participant is free to get on with the rest of their day. There are no post-treatment precautions, restrictions or other safety concerns.
We do that in month one, in what we call the acute phase. And then over months two, three, four, five, six, we bring them back in for ten additional sessions per month. At month 12, we follow-up with the participants to see how they have gone.
AM: When will the trial results be published?
LC: We should have some interim results, particularly of the weight and psychological treatment outcomes after the acute phase, say, 12 months from now.
AM: Most effective treatments so far have involved full nutritional rehabilitation. And I am wondering if it’s going to be possible to change the brain without the changing the body first?
LC: Me too. I too am open-mindedly sceptical whether our treatment approach will work. But treatments like family-based therapies, psychological therapies, medications and direct weight restoration measures are all trying to manage or reverse at best the behavioural manifestations. What we’re hoping to do here is directly target the brain network where anorexia psychology and behaviours are stemming from — getting at the root of the problem, so to speak.
AM: I think the magic bullet would be something that can help people to eat and gain weight without the trauma that this brings for people with anorexia. So, if TMS works for that, it would be amazing.
Thank you for taking an interest in anorexia nervosa and for this innovative research. And best of luck with the trial.
Dr Leo Chen is a consultant psychiatrist and senior research fellow at the Department of Psychiatry, School of Translational Medicine, Monash University in Melbourne, Australia. He is Director of Psychiatry Training at Alfred Health and Head of the Therapeutic Trials Team at the Multidisciplinary Alfred Psychiatry Research Centre, Melbourne, Australia
If you live in Australia and are interested in participating in the TMS trial for anorexia nervosa, please contact tms-trials@monash.edu. For more details, you can search for the “FUNCTIAN” trial at ANZCTR
*Please see the trial details if you wish to know the specific weight range