Quantitative functional MRI: developing non-invasive neuroimaging to map the human brain's consumption of oxygen

Diseases of the brain including neurological conditions, such as epilepsy, multiple sclerosis and dementia, and common psychiatric conditions such as depression and schizophrenia, have considerable personal, social and economic costs for the sufferers and their carers. Improving the tools at our disposal for quantifying brain function would help with diagnosis, choosing the right treatment for the patient and developing new, more effective, treatments. This proposal aims to develop a reliable non-invasive brain imaging method using magnetic resonance imaging (MRI) that maps, across the whole human brain with a spatial resolution of a few millimetres, the amount of oxygen that the brain is consuming. The rate of oxygen consumption, known as CMRO2, reflects neural activity and can change through disease processes. It provides a marker of disease and treatment related alterations in brain activity. Our proposed method would also map the functional characteristics of brain blood vessels whose health is crucial for the supply of oxygen and nutrients to the brain.

Until recently, it has only been possible to quantitatively map the human brain's metabolic energy use through positron emission tomography (PET), which relies on radioactive tracers. The application of such measurements is limited, as in order to minimise radiation doses, it cannot be applied many times in the same patients or healthy volunteers. This hampers the repeated study of disease or treatment progression and the study of normal brain development and aging. Our proposed method would avoid the use of ionizing radiation, would be cheaper than PET and more widely available, and would expand the applications of quantified CMRO2 mapping to more centres, leading to improved treatment targeting and potential healthcare cost savings.

We have performed some initial tests that show our proposed method to be feasible. It relies on mapping simultaneously the flow of blood to each part of the brain and the oxygenation of the blood leaving each part of the brain. Necessary for the measurement is the modulation of brain blood flow and oxygen levels, achieved by asking volunteers to breathe air enriched with carbon dioxide and oxygen. These procedures involve the volunteer wearing a face-mask but are safe and well tolerated. Our proposed method should yield additional information describing cerebrovascular properties compared to other recently-proposed methods. This means that it would require fewer assumptions which may be not be invalid in the diseased brain, giving our approach a wider scope of application and offering potentially richer clinical information.

This proposal optimises our method to ensure it is efficient and reliable for widespread research and eventually clinical use. We propose a close collaboration between physicists developing the neuroimaging methodology and clinical academic researchers who will help us to demonstrate its clinical feasibility in two common neurological diseases, epilepsy and multiple sclerosis (MS). About 70% of the project will be methodological development to optimise our image acquisition and data analysis strategy to yield accurate and repeatable measurements within about 10 minutes of scanning. The remaining 30% of the project will validate the method in groups of epilepsy and MS patients who volunteer to help us with our research. Validation will be performed by comparison with PET, the current 'gold standard.'

The project will develop and benefit from partnerships with academic and industrial researchers in the UK and internationally. In particular, the work has good potential for application in the drug development industry, a strong industrial sector in the UK, for the development of new and effective compounds to treat psychiatric and neurological disorders. This project would help maintain the UK at the forefront internationally of neuroimaging research, a position it has long held and from which it has benefitted.

We believe that the research outlined in this proposal would benefit several potential stakeholders outside our immediate academic research arena.

In the short term, basic and clinical neuroscientific researchers would benefit from a non-invasive neuroimaging tool that allows them to address their research questions in humans and animals in a way that radiotracer techniques have previously precluded, for example, through longitudinal studies in healthy volunteers and patients, and in animal studies where there is the opportunity to replace some autoradiographic procedures with imaging and thus reduce the number of animals used in research. The impact in academic research is a necessary step to the longer term and wider health and economic benefits that we would expect success in our project to offer.

The improved, cheaper and thus more widespread capacity in the healthcare sector to map human brain oxygen metabolism is likely to offer benefits to patients and the wider economy. In many cases an MRI-based method to measure cerebral oxygen consumption could replace radiotracer based clinical scans aimed at measuring brain metabolism with oxygen-15 or more commonly radio-labelled glucose (for glucose metabolism), by providing clinically comparable information. Replacement of a proportion of such scans would reduce radiation doses received by patients. However it would open the way to repeated scanning over time and thus the monitoring of disease progression and treatment effectiveness, allowing better and faster decisions to be made about the most effective treatment for the individual patient. Two specific examples, epilepsy and multiple sclerosis (MS), are addressed as 'proof-of-concept' in the present proposal.

As well as replacing some PET scans, metabolic imaging could become available for diagnosis, treatment planning and monitoring a wider range of brain diseases, including epilepsy, MS, head injury, cerebrovascular disease and neurodegenerative conditions such as dementia (e.g. Alzheimer's), as well as psychiatric conditions for which there is a current lack of diagnostic imaging tools. Many of these conditions affect young patients in their economically most active years, causing disability and a reducing their ability to work. Dementia, as well as being debilitating for the patient, imposes a high economic and healthcare burden on society and carers. An improved availability of the clinical tools to diagnose early, stratify patients and select the best treatment early on could reduce the impact of some of these conditions on the individual and on society, as well as reducing the overall health costs through the reduction of accumulated disability. New windows into brain function would also fuel the public's interest in neuroscience and create an educational impact across all age groups.

The improved understanding of cerebral metabolism offered by the provision of a tool to study it more widely may be expected to lead to hypotheses for testing new targeted rehabilitative interventions and drug therapies for the treatment of neurological and psychiatric conditions. Apart from the long-term health benefits this would be of great economic value to the drug discovery and development industry, an area of the world's knowledge-economy in which the UK is particularly strong. The success of our methods could make a considerable contribution to evaluating new compounds in the pharmaceutical industry. The development process for new treatments is an extremely expensive business particularly because of the cost of the high failure rate during late-stage clinical trials. Better imaging tools to provide biomarkers of drug action in small cohorts of subjects would help the industry predict better which compounds are likely to fail. Prof Wise has 12 years' experience in developing fMRI methods in concert with large pharmaceutical companies in the UK and internationally.