Improving function in Huntington's disease through neurofeedback: using real-time fMRI to enhance cortical plasticity in early stages of the disease

Huntington's disease (HD) is a rare, inherited, neurological condition where progressive brain damage leads to severe difficulties with movement, emotion and thought. The most characteristic sign of HD is uncontrolled movement of the limbs but subtle movement, emotion and thinking difficulties can start many years before this. There is no known cure for HD; a few drug treatments can help with some of the symptoms but have many deleterious side effects.

HD is characterized by slow, progressive atrophy of the brain, which theoretically allows the brain the opportunity to re-organize and compensate for the disease-related structural and functional changes. In this proposal, we will develop state-of-the-art technologies using functional magnetic resonance imaging (fMRI) and biofeedback to enhance such compensation and potentially alleviate symptoms. fMRI is a well-known technique that allows activity in the brain to be recorded non-invasively while people perform a task. Normally the analyses of these data are carried out off-line over a period of several weeks. However, recent developments allow real-time analysis of brain images and enable us to feed back the level of activity in selected brain areas to an individual while they lie in the brain scanner. Such real-time feedback can be used to train people to control their own brain activity through repeated sessions where they learn to enhance the activity that is fed back. This has been successfully used to allow people to control activity in many different regions of the brain and has also been applied to clinical conditions such as Parkinson's disease, depression and pain management with encouraging initial results for symptom management.

In this proposal we will use such advanced neurofeedback fMRI techniques to train individuals with symptomatic HD to control the level of activity in motor cortex, or the coupling of motor cortex with areas deep in the brain that are affected early in HD. Our underlying hypothesis is that if individuals with HD can learn to regulate the activity of brain regions that underpin their disease and contribute to the manifestation of certain symptoms, then the symptoms should improve.

Apart from the urgent need for new therapies, one advantage of studying HD is that genetic testing can establish with 100% accuracy whether an individual carries the mutant gene and will therefore develop HD at some point in their life. This enables us to not only monitor disease progression many years before an individual manifests the disease (pre-symptomatic stage), but also to develop interventions that can be applied at very early stages of the disease, when the brain damage is still limited and function is still well preserved. We know that structural and functional brain changes precede the manifestation of overt clinical signs by many years. This means that there is a mechanism of compensation at work, enabling gene-carriers to maintain a normal level of function despite degeneration. As the disease progresses and brain atrophy increases, this mechanism starts to break down and clinical signs start to become more evident. This is another gradual process and pre-symptomatic HD gene-carriers gradually move to a peri-symptomatic phase (with soft motor signs and mild cognitive impairment) and then eventually express unequivocal signs of the disease.

Our aim is to determine through a proof of concept study that the training protocol can be tolerated and followed by HD patients, and show that the intervention can improve symptoms in early stage HD. If the results are positive we will then proceed to develop this intervention more generally for both early and pre-symptomatic stages of the disease. Because neurofeedback training is a non-invasive intervention with no side effects, it is an ideal candidate for both pre-symptomatic treatment and in combination with other treatments at later stages of the disease.

Huntington's disease (HD) is a genetic, neurodegenerative condition that leads to extensive brain atrophy, starting from the striatum and gradually spreading throughout the brain. Clinically it is characterized by progressive motor impairment, e.g. chorea, cognitive decline and neuropsychiatric symptoms. There is currently no known cure for HD and treatments prescribed for symptom management have significant side-effects. This project will deliver "proof of concept" testing of a novel, non-invasive intervention: neurofeedback training using real-time functional MRI (rt-fMRI), which will induce neuroplasticity and could help patients better manage the disease symptoms. Patients will be scanned using fMRI and will be trained to regulate the activity of specific brain regions through receiving in-scanner real-time feedback about the activity of these regions. We hypothesise that training patients to regulate the activation of brain regions whose activity has been disrupted by the disease (e.g. premotor cortex and striatum), will lead to improvements in behaviour and slowed disease progression. Such rt-fMRI has already been used for the treatment of other clinical conditions (e.g. Parkinson's disease, chronic pain, depression) with positive results on behaviour and symptom management without any side-effects. The proposed study will be the first "proof of concept" for HD and will provide preliminary evidence on the feasibility and efficacy of the intervention in early stage HD. If the results are positive, then these data will support further development with a large scale, randomized controlled trial in both early stage and premanifest HD. HD is an excellent model of neurodegenerative diseases as there is a precise genetic test, which allows prospective study of disease progression particularly in the premanifest phase. Our results may therefore have broad applicability and be extrapolated to other more common neurodegenerative diseases such as Alzheimer's disease.

The proposed project is the collaborative effort between different research groups working on MRI physics, HD and cognitive neuroscience. As such the results from the project are relevant to not only the HD research and patient community, but also to research in neuroplasticity, brain functional reserve and neuroimaging methods for the study of neuroplasticity. At present neurofeedback training is being tested as treatment for and has already partly demonstrated efficacy in Parkinson's disease (Subramanian et al., 2011), depression (Linden et al., 2012) and chronic pain management (deCharms et al., 2005). Therefore our systematic approach to training and technical improvements will be highly beneficial for these applications and for the further development of neurofeedback training using rt-fMRI in the clinical setting more generally.

If our results are positive, we expect to actively involve supportive MRI manufacturers (see letter of support by Siemens), who will be able to deploy this technology in the clinical setting through integration into MRI scanner software. Dr Weiskopf has a long-standing collaboration with Siemens and is PI on a joint academia-industry project with Siemens on implementation of novel scanning methods for clinical use.

Developments in the technology of rt-fMRI, such as increase in signal to noise and contrast to noise ratio and prospective head motion correction, will also be relevant for other applications of rt-fMRI, other than neurofeedback training. One such application is the communication with locked-in patients to establish whether a patient is in a vegetative or locked-in state. Improvements in rt-fMRI technology will provide a better tool for communication with these patients.

In HD, we have predictive genetic testing that can establish with absolute certainty whether a person will develop the disease later in life, therefore HD is a good model for research in the development of a preventative treatment. Neurofeedback training is a safe preventative treatment that could be also used for other neurodegenerative conditions, where there are identifiable risk factors, but no clear diagnosis prior to symptom onset, such as Alzheimer's disease. Therefore the outcomes of the research will be of interest in the development of low-risk preventative treatments for other neurodegenerative diseases as well.