Many techniques are currently available to look at systems-level function in the brain, but all have their limitations. For example, PET provides a precise measure around a specific receptor of interest, however, there are many receptors for which we don’t have ligands that bind to them. Many drugs also have multiple targets and may be too diffuse to get good signals from PET scans. fMRI has good spatial resolution and reasonably good temporal resolution but is not selective to any drug targets and has uncertainties of interpretation in terms of hemodynamic confounds and neurovascular coupling. The following speakers demonstrated how novel multimodal imaging approaches can help to build bridges between receptor systems and neuroimaging techniques.
Professor Richard Wise, head of magnetic resonance imaging in Cardiff, began the session by describing quantitative fMRI methods which have been developed over the years to help overcome the limitations of the standard BOLD fMRI signal. These methods have wide applications in translational pharmacological research. fMRI is useful in two main ways: in basic neuroscience, looking at systems-level activity in the brain, and in developing markers for drug development, i.e. obtaining evidence for the action of drugs in the brain. Prof. Wise worked for a number of years on the mechanisms of pain and looked at the pharmacokinetic/pharmacodynamic response to pain stimuli. In those studies, responses in insula was found to be lowered with administration of opioid. -> Able to extract information related to the timescale of drug activity within the brain. Dr Wise told us he was a great advocate for multimodal imaging, as it combines the specificity of PET with the flexibility and functional information available from fMRI data. Blood oxygenation level-dependent (BOLD) activity from fMRI and cerebral blood flow (CBF) measures are likely not good enough as separate measures on their own to properly get at underlying neurophysiology, so a multimodal approach is important for progress in the field. Dr Wise continued by also highlighting new opportunities available with the use of 7 Tesla MRI scanners. These include: improved arterial spin labelling techniques and blood flow measures, the ability to obtain cortical layer specificity and whether certain drugs affect the input or output layers, the use of sodium MRI (nuclear magnetic resonance on the basis of sodium instead of the standard hydrogen), and also improved MR spectroscopy, e.g. better spectral separation of glutamate and glutamine.
Dr Gregor Gryglewski from Vienna described a study on the effects of SSRIs on the brain using hybrid PET and MR imaging. The first mechanism of action of SSRIs in the brain is binding to the serotonin transporter (SERT). SERT is the target of first-line antidepressant therapy, has been shown to lead to alterations in depression, and has potential as a biomarker. Previous study from Vienna lab showed reduced BOLD activation in the amygdala after three weeks of SSRI treatment. The objectives of their current study were to find predictors for treatment response in MDD, and on the methodological side, to find acute responses to SSRIs, and to see if there is a correlation between the PET signal (occupancy of SERT) and fMRI BOLD signal. Furthermore, they wanted to assess how this acute response might relate to the long-term clinical outcome. The study included 40 patients (unmedicated, first-episode) and 40 healthy controls, who each underwent two scans and received either citalopram or a placebo. An analysis of resting state data did not identify any regions with significant differences in regression coefficients between placebo and citalopram scans after correction for family-wise error. This result appears to be in line with: the sub-theraputic dosage used in the study (this dosage was chosen to not induce any nausea in participants), the lack of subjective effects reported by participants, as well as the latency of efficacy of the SSRI.
The final two speakers of the symposium were engineers by training and presented new methods demonstrating how multimodal imaging can be a powerful tool to quantify receptor mechanisms related to serotonergic and dopaminergic drugs.
Dr Ottavia Dipasquale from King’s College in London presented a study in which they applied a novel multimodal technique, called Receptor-Enriched Analysis of functional Connectivity by Targets (REACT), using resting state (RS-) fMRI data in conjunction with PET imaging. REACT is a two-step multivariate regression analysis in FSL that uses normative population-based PET maps to obtain subject-specific RS functional connectivity (FC) maps. The study was set up to assess changes in FC maps after administration of MDMA, a compound with a mixed profile of serotonergic function, compared to a placebo. PET-derived FC maps were created based on five serotonergic targets 5-HT1A, 5-HT1B, 5-HT2A, 5-HT4 and 5-HTT, per MDMA/placebo condition. Dr Dipasquale and colleagues found MDMA- related FC changes mainly for the 5-HT1A and 5-HTT targets. Furthermore, they found a positive relationship between MDMA-related FC changes for 5-HT1A and MDMA-related changes in pharmacokinetic (PK) plasma levels collected 45 minutes after drug administration. This new method, based on normative maps guided by molecular substrates, therefore provides us with a sophisticated way to examine drug-specific topography of functional connectivity in the brain.
Dr Christin Sander from Massachusetts General Hospital showed us results from a simultaneous PET/MRI study designed to identify, in vivo, the functional response at D2/D3 receptors of partial agonists, compared to full agonists and antagonists. Such partial agonists are currently being used as third generation antipsychotic drugs. Imaging was carried out on 3 non-human primates, which allowed the use of iron oxide as a contrast agent. Changes in cerebral blood volume (CBV) were determined from the MRI data. Receptor occupancy was obtained from the PET data using a simplified reference tissue model (with the cerebellum as reference) and allowed for dynamic binding changes. Full agonists and antagonists showed opposing CBV responses in the striatum; agonists were associated with a negative response and antagonists with a positive response. The partial agonist, aripiprazole, showed a positive but lower CBV response than the full agonist, but also higher D2/D3 receptor occupancy. Increasing dose amount of aripiprazole was found to be associated with higher receptor occupancy. Dr Sander went on to discuss receptor internalization, a homeostatic neuro-adaptation mechanism that occurs in response to agonist stimulation, and is known to affect both PET and MRI data in different ways. She proposed a neurovascular coupling model using combined PET/MRI data that can help distinguish between receptor internalization and receptor desensitization. All in all, this simultaneous PET/fMRI method seems like a useful tool to classify D2/D3 drugs according to their in vivo pharmacodynamic profile, and to tease apart drug-related receptor mechanisms such as internalization.