It often takes several weeks to discover whether a medicine for a psychological or neurological condition actually works. By the time that the decision is taken to try a different anti-depressant, a patient has often had to wait for a long time without his or her complaints being alleviated, but while still suffering disturbing side effects of the medication. It is frustrating for the patient, frustrating for the doctor and also frustrating for those who develop new medicines. Can it not be done quicker?
Dr Serge Rombouts: ‘For a resting state fMRI scan a test person does not have to do anything at all. The only instruction is: 'Relax, but don't fall asleep.' Images are then made of the brain as a whole.'
Brain at rest
MRI specialist and neuroscientist, Dr Serge Rombouts thinks it can. He has received a Vidi subsidy from NWO to research whether, using a new MRI application, he can immediately see what is happening in the brain after substances are administered which affect the central nervous system. This new offshoot of the MRI branch is the Resting State fMRI (RS fMRI); using this technique you can make images of the 'brain at rest'.
Within one day
Rombouts: ‘Hopefully, this will allow us to see within a day what a particular drug is doing in the brain. But we can also see whether two types of medications which at first sight seem to have the same effect - for example they both improve memory - actually do two very different things in the brain.
‘Relax, but don't fall asleep’
Resting state fMRI is a type of functional MRI. Functional MRI makes images of brain activity; it is possible to see fron the oxygen concentration in the blood which areas of the brain are active when a test candidate carries out a specific cognitive task, such as naming an object, or recognising a face. For a resting state fMRI scan, a test candidate does not need to do anything at all. The only instruction is: 'Relax, but don't fall asleep.' Images are then made of the complete brain.
Development of medicines
What is so important about this distinction? Rombouts: ‘Normal functional MRI has a number of significant limitations, which mean that it does not lend itself well to the development of medicines for the central nervous system. You work with tasks. This means that the effects of the medication which you see taking place are always dependent on the task being carried out. You don't always want to have this task in the way; you want to see the direct effects. Also, you restrict yourself to the area of the brain which is activated by the specific task. With resting state fMRI, you do not have this interaction between medication and task. And you see the effects in all areas of the brain, not only in very specific areas which are determined in advance.'
‘Resting state pharmaco-MRI’
‘And this would have enormous advantages for drug research,' adds Professor Joop van Gerven. Van Gerven is conducting drug research at the Centre for Human Drug Research, and is one of the partners in the project. 'In the early stages of drug development, very little can be said about their therapeutic effectiveness. So, a new medicine is first tested on healthy volunteers. We have to examine whether the medication reaches the part of the brain where it needs to have an effect. Currently, all kinds of indirect techniques are used, such as EEGs (recording brain activity), or neuropsychological tests.
These show that a new medication has an effect on the brain, at what dosage the effects take place, how long the effects last and what the side-effects are But such an approach has the same disadvantages as the 'task-dependent' fNRI: the answer you get is dependent on the task. At the present time there hundreds of different techniques in use in drug research and there is absolutely no standard. With resting state fMRI you can see directly what part of the brain is affected by the drug. If it really is shown to work, resting state pharmaco-MRI may well become the standard method in the development of new medicines for disorders of the nervous system.'
Dr Serge Rombouts works at the radiology department of the LUMC, and the psychology department of the Faculty of Social and Behavioural Sciences. He is also director of the Leiden Institute for Brain en Cognition (LIBC), in which four faculties are represented. Of the eight Visi subsidies awarded to Leiden this year, four were for LIBC-related research.
New vision of how the brain works
Rombout's research is pioneering. RS-fMRI has never previously been applied to drug development, but this variation on fMRI has itself only been in existence for a couple of years. It is based on a new perspective on how the brain works.
Background noise is not noise
Rombouts: ‘The dominant view in research into brain activity is that the brain is above all reflexive; it responds to stimuli from the environment. What happens below this is often seen as background noise, or not important. But a different vision, which is becoming increasingly popular, is that the operating of the brain is largely intrinsic. The surface noise is actually not noise at all, but the activity of a well-ordered network of brain areas. The activity which we call background noise may well tell us much more about how the brain works than the increase in activity in a particular area of the brain when a task is carried out. But as yet very little research has been done into this. We want to know what this noise means.'
Rombouts already knew about the resting state fMRI from his previous research at the VU (Vrije Universiteit Amsterdam). He received a Veni subsidy for this research at the time. Rombouts: ’I made some digressions and with my co-workers I studied the effects of the stress hormone cortisol on resting state activity. When I came to Leiden, Mark van Buchem, our Neurology Professor, told me about the Centre for Human Drug Research, where Joop van Gerven is working as Professor of Neuro-Psychopharmacology. I subsequently contacted Joop and told him about the resting state fMRI. He recognized the possibilities immediately and explained what we should do to apply this to medicinal research.’
3 ‘resting state’ networks in the brain, seen from the front of the head and from the back. A network in blue, one in red/yellow, and one in green. The areas of the same colour exhibit a strong mutual ’connectivity’ when at rest.
Together with a team including the two research assistants, Rombouts will examine the effects of six substances on the brain, the actions of which have been studied thoroughly already: alcohol, morphine, THC ( the active chemical compound in Marijuana), the anti-psychotic risperidon, the anti-depressant citalopram, and scopolamine, a substance that slows a person down and has an adverse effect on the memory.
Gold standard: PK/PD
The results will be checked against the gold standard of Pharmacology: pharmacokinetic/pharmacodynamic modelling. This is a mathematical description of the change in the concentration of administered medicine in the blood over the course of time, related to the change of the pharmacological effect over the course of time. ’We will now examine whether there are certain parts of the brain which correlate in the same way,’ Rombouts remarks.
Patients and test subjects
Rombouts and his co-workers will test 15 patients and 6 groups of 12 healthy test-subjects. This will enable them to calculate group averages. ‘But what we ultimately hope to achieve is a personalised medicine: a medicine individually tailored to each patient.’
In and out of the scanner
The research will not be easy, neither for the test-subjects and patients nor for the researchers. An MRI scan ise not pleasant for people who cannot stand noise or who suffer from claustrophobia. Rombouts: ’But our patients, psychiatric patients, will have to spend half a day going in and out of the scanner, because we need to see the whole curve showing the increase and the decrease in concentrations. But the process of scanning is challenging in itself as the medicine is often applied intravenously. The pump must remain outside the MRI scanner, but the tube must go inside. It is definitely not a standard scan.’
Investigating without a model
But the data analysis is the most complicated aspect, he continues. ’It is standard practice to work with a model. You expect something to happen. You have someone do calculations for 30 seconds and then pause. You get an on-off signal. Now we have to investigate without any model whatsoever. But we do have specialists in Oxford with whom I collaborate. They are developing the necessary techniques. But this is only one of the difficulties. We still have the connection with the PK/PD modelling to solve.’
Cite This Page: