An international team of researchers is developing an MRI-based method for the early detection of human brain tumours. The method is called CEST-MRI and it detects elevated glucose concentrations in humour tissues, quantitatively and at high spatial resolution. It does not expose patients to radiation, is non-invasive and relatively inexpensive.
Tumours need high levels of energy in the form of glucose, i.e. sugar, for growth. Locally elevated glucose levels in the tissue can therefore reveal where cancer cells are located. This finding is not new and is already being used in a method known as positron emission tomography, PET for short, which involves giving patients radioactively labelled glucose that accumulates in tumour tissue due to the tumour’s high metabolic activity. The radioactively labelled glucose molecules emit radiation that is detected and converted into imaging signals. Although this method is relatively sensitive, the drawback is that the patient is exposed to radiation, albeit in relatively low doses. In addition, the devices are expensive, so not all clinics can afford a PET scanner.
Scientists from the Max Planck Institute (MPI) for Biological Cybernetics in Tübingen are working with partners from across Europe on a project to develop a simple method for early diagnosis of tumours based on detecting elevated glucose levels in tumour tissue using magnetic resonance imaging (MRI). The project will be funded with around six million euros from the EU for the next four years and is headed up by Prof. Dr. Xavier Golay from University College London. Prof. Dr. Klaus Scheffler, who is head of a research group at the MPI for Biological Cybernetics involved in the project, explains the principle of CEST-MRI (CEST = chemical exchange saturation transfer): “Glucose has an influence on the water molecules in tissue, which results in the exchange of protons between glucose and water molecules. This changes the magnetic properties of the water molecules. Although this exchange process is not yet fully understood, it is possible to measure the magnetic field change reliably and quantitatively.” This is achieved by administering patients a certain amount of glucose in an infusion or drink. “The amount of glucose contained in about three to four hundred millilitres of commercially available soft drinks is enough to measure a time-delayed effect,” says Scheffler.
The researchers have shown in the animal model that the method works. However, MRI is not yet sensitive enough to detect the relatively small amounts of glucose present in small, early-stage tumours. The MPI is currently working on developing the sensitivity of the method. “There are 10 or so MRI parameters that we can change in order to influence the signals. For example, one of our goals is to improve the images by selectively manipulating the protons and changing the spin momentum. For this, we are also using numerical proton exchange simulations.”
The second challenge in the project is no less important: the method needs to be as robust as possible to enable widespread use in hospitals. “We want to develop a ‘push-button’ method, and this is not easy to achieve. The effect can be demonstrated quite well under very controlled laboratory conditions. However, this is not possible in hospitals, where it might lead to instabilities. This aspect still needs to be improved,” Scheffler admits.
The resolution needs to be good enough to accurately distinguish a tumour from surrounding healthy tissue, which would make it extremely valuable in surgical interventions. In addition, the project partners want to progress the method to the extent that it could be used to find out whether all or only specific tumour areas are metabolically active at the time of examination. Resulting data can be then used to draw conclusions on overall disease pathogenesis and better monitor tumour response to chemo- and radiotherapy. In principle, CEST-MRI can be used to detect molecules other than glucose, such as proteins for example. Other researchers from Tübingen are pursuing this approach in other projects, including a project with researchers from the Department of Neurology at Tübingen University Hospital that is investigating protein plaques in Alzheimer’s patients.
While the research group from Tübingen is focused on the diagnosis of highly malignant brain tumours, i.e. glioblastomas, their colleagues from London are working on making the method suitable for application in prostate and stomach tumours. “In Tübingen and London, we have identical 3-Tesla devices for making clinical measurements in healthy volunteers and patients. Here in Tübingen, we can also test the method with a 9.4-Tesla device, which gives us even better results than the 3-Tesla devices. However, this is currently being done out of pure scientific curiosity rather than for any direct relevance in clinical application as no hospital has such a device,” says Scheffler. In fact, the MPI for Biological Cybernetics, in cooperation with the Centre for Neurooncology at Tübingen University Hospital, was the first centre in the world to use a 9.4-Tesla device for examining human patients. At the time the two institutes purchased the 9.4-Tesla device, there were only three other institutions worldwide that owned a device with such a high magnetic field strength. The device in Tübingen is used for images with a particularly high spatial resolution of less than one millimetre. The 3-Tesla devices that are mostly used in the project are also common in private radiology practices.
While the clinical imaging method is mainly being developed by researchers from Tübingen and London, project partners at the University of Zurich are working on the improvement of the optical imaging process. Another team at the University of Tel Aviv reviews the new developments in the animal model, and partners at the University of Turin investigate the mechanism of glucose metabolism in tumours. EU-funded projects tend to involve both academic and industrial partners, and this is also the case here: an Italian company called Bracco Imaging carries out toxicological tests of the glucose variants and a French company called Olea Medical SA provides special image evaluation software to quantitatively process the images.