Brain imaging for psychiatrists

Part 2 – Structural (static) measures

Both MRI (= magnetic resonance imaging) and PET (=positron emission tomography) can be used to measure fixed elements of brain function.

Fig 1.
MRI scans reveal brain damage from alcohol

MRI can give detailed measurements of brain morphometry as used for instance in radiological diagnosis. A clear example of this is shown in fig 1 which demonstrates the massive loss of brain tissue in four patients with alcoholism as compared with an age and sex matched healthy control group. Similar brain shrinkage is shown in dementias such as Alzheimer's disease as seen in fig 2.

These morphometric measures can be made quantitative through statistical programmes such as VBM [volumetric brain mapping] that can estimate the density of brain tissue across the many thousands of voxels that make up an MRI image. However, DTI imaging can be used to delineate neuronal pathways across the brain [see part 1] and this is now becoming more quantitative with statistical procedures developing so that different patient populations can be compared.


Fig 2.
MRI scans reveal brain shrinkage in Alzheimer's disease patients

PET and SPECT (=single photon emission computerized tomography) can measure the distribution and density of proteins to which tracers bind.

Example of the evaluation of the receptors distribution: fig 3d shows a PET scan of dopamine D2 receptors in brain as identified by the binding of 11C-raclopride. The vast majority of these dopamine D2 receptors are detected in the striatum which makes sense as over 80% of all the dopamine in the brain is found in this region too. In contrast the GABA-A receptors have a very different distribution as seen in fig 4 which shows a PET scan of the benzodiazepine receptor tracer 11C-flumazenil. These are largely detected in cortex with relatively little binding seen in striatum and thalamus.


Fig 3.
Multiple PET tracers for the dopamine system

Fig 4.
PET scan measuring GABA-A receptors in normal brain and in people with panic disorder

Measuring receptor density with PET and SPECT has proved very useful in defining alterations of neurotransmitter receptors in disease states.

Example of the evaluation of receptor density: fig 3d shows the reduction in number of D2 receptors in patients with cocaine addiction, which is believed to contribute to the inability of cocaine addicts to restrain their drug use. In the right sided image in fig 4 we can see reduced benzodiazepine/GABA-A receptors in fronto/temporal brain regions in patients with panic disorder. Similar benzodiazepine/GABA-A receptor deficits are also found, though to a lesser extent in patients with General Anxiety Disorder and Post Traumatic Stress Disorder, as well as in epileptic foci. In all these states it is thought that the reduction in GABA-A receptors leads to reduced GABA-A inhibition which in turn leads to excessive glutamate excitation that ultimately provokes either anxiety or seizures depending on the brain regions affected.

PET tracers now exist to measure a number of neurotransmitter receptors and reuptake sites and a few enzymes. One of the most revealing studies has been those estimating the density of the serotonin 5-HT1A receptor in depression. These data are shown in fig 5 using the “glass brain” 3 axis projection system. The voxels that show a significant reduction in density in depressed patients as compared with age- and sex-matched healthy controls are shown in shades of grey with the darker the colour the more significant the reduction. This was the first demonstration of a serotonin deficit in living human brain in depression and was further extended by the same group revealing that the deficit did not recover on successful treatment. This suggests that the reduced serotonin receptor number may be a trait – rather than a state – marker of depression. We presume that the effect of SSRIs to lift depression is due to their increasing synaptic 5-HT concentrations so a level that offsets or compensates for the receptor deficiency. Also first degree relatives of the depressed subjects had a similar though lesser reductions on receptor number which suggests this possibly has a genetic basis.

Fig 5.
Density of the serotonin 5-HT1A receptor in depression

PET tracers can also be used to measure enzyme distribution and potential activity for example the evaluation of the enzyme distribution and activity. The best studied neurotransmitter system is that of dopamine because of its presumed role in addiction and schizophrenia. We can now measure most of the elements of this neurotransmitter system as shown in fig 3. 18F-dopa can be used to give an estimate of dopamine synthesis rate, 11C-deprenyl to measure MAO activity, 11C-raclopride to measure D2 receptors and 11C-bCIT to measure the density of dopamine reuptake sites.

Although this level of evaluation interrogation of the dopamine system has not revealed the cause of addiction, it has proved useful in aiding diagnosis of Parkinson's disease where 18F-dopa uptake into the striatum is reduced. Also the 11C-deprenyl tracer has been used to reveal that cigarette smoking blocks this enzyme which may explain the mood elevating effects of smoking Fig 3a.

Measures of the density of proteins can be useful in diagnosis as shown in Alzheimer's disease where the deposits of b-amyloid can be imaged using a tracer such as 11C-PIB as in the right side of fig 6. The left side of this figure shows the other imaging technique (18F-FDG PET) which is used in diagnosis by measuring cerebral metabolic rates of glucose.


Fig 6.
18F-FDG PET measures cerebral metabolic rates of glucose and 11C-PIB PET measures the density of b-amyloid deposits

Fig 7.
Using 11C-raclopride PET to estimate receptor occupation of D2 receptor antagonist

Left: Before treatment, Right: After treatment

Drug- receptor occupation studies

One of the most valuable uses of receptor imaging has been to estimate the fractional occupation that different drugs produce in the brain. For a range of drugs used in neurology and psychiatry we can now perform a form of dose-response measurement at the precise target site of drug action, a big improvement as previously we had to make estimates of brain target engagement from plasma concentration measurements. In fig 7 this is shown for D2 receptor antagonist. Both scans use 11C-raclopride to estimate the number of dopamine D2 receptors available for the tracer to bind to. In the untreated state a high level of 11C-raclopride binding is seen in the dopamine rich striatal regions. However after treatment with D2 receptor antagonist neuroleptic this is dramatically reduced because this drug competes with 11C-raclopride for the same binding site.

Studies such as these were instrumental in changing the dosing regimens of dopamine blocking drugs for psychosis because they revealed several important facts. First the receptor occupation needed to produce an antipsychotic effect was about 65%. But when occupation reached over 80% then adverse effects such as extra-pyramidal symptoms and elevated prolactin concentrations were seen. This means that the therapeutic window for these drugs is very narrow. Even a small increase in dose can switch a patient from being underdosed [i.e. occupation < 65% ] to overdosed [i.e. occupation > 80%].

The second advance was that it became apparent that certain doses of D2 receptor antagonist produced saturation of the dopamine receptors. Thus there was no benefit in using higher doses and as a result of this imaging knowledge the average dose of D2 receptor antagonist was markedly reduced.

Imaging of the serotonin reuptake site has shown rather different results because here it is necessary to get over 80% occupation by a reuptake blocker to produce significant lifting of depression. Different receptor occupation/effect relationships apply to different drugs so the full mu receptor opioid agonist produces it actions with very low [less than 5%] occupation of the receptor whereas the partial mu receptor opioid agonist requires occupation of > 50% of receptors.


PET and MRI imaging offer important new ways to explore brain structure and function in psychiatry and neurology. At present whilst MRI is in regular clinical use for diagnosis, PET is largely just a research tool, but some tracers e.g. F-DOPA and 11C-PIB are being used to aid diagnosis of Parkinson's disease and Alzheimer's disease respectively. The future development of tracers that use 18F instead of 11C will further spread the availability of these in clinical practice. This is because the much longer half-life of 18F [110 min v 20min] allows production of the radioactive tracer at a central cyclotron site with distribution to multiple hospitals that have PET scanners. In contrast 11C tracers have to be made on the same site as the scanner, so each site needs its own cyclotron.

Country selection
We are registering that you are located in Brazil - if that's correct then please continue to Progress in Mind Brazil
You are leaving Progress in Mind