PET in neurology

Introduction

PET provides functional information about the brain, ranging from hemodynamic

information about blood flow and blood volume to metabolic data about glucose and

oxygen utilization, as well as metabolic processes such as protein synthesis. It also

permits the evaluation of pre- and post-synaptic neurotransmitter systems. Moreover, if

one simultaneously evaluates anatomy, physiology, and biochemistry of the brain in

patients exhibiting pathological behavioral symptoms, one should be able to obtain

information that may answer fundamental questions about its pathogenesis. Clinically,

PET has been used successfully in the grading of brain tumor, differentiation between

recurrent tumor and radiation necrosis, evaluation of dementia, evaluation of intractable

epilepsy, evaluation of Parkinson’s disease, schizophrenia, depression, functional

activation and many others.

Radiopharmaceuticals

1. Tracers for CBF Measurements

The oxygen-15 inhalation technique, a PET method first described in Europe by

Frackowiak et al, enables the evaluation in the same session and with absolute values,

blood flow (rCBF), oxygen consumption (rCMRO2), oxygen extraction ratio (OER), and

blood volume (rCBV). This method has been mainly applied to the evaluation of patients

with cerebrovascular disease to understand the pathological mechanisms underlying

stroke. However, its application in clinical practice is limited because of its complex and

somehow cumbersome technique. Furthermore, because of the high energy emitted by 15O, image resolution is approximately 10 mm, significantly lower than that obtained with 18F.

2. Tracers of Metabolic Substrates

The tracer most commonly used for PET studies of metabolic substrates has been 18F-

fluorodeoxyglucose (18F-FDG). The high-quality images obtained, the possibility of quan-

tifying glucose consumption in standardized units, and the relative simplicity of radiotracer

production are probably the main reasons for its frequent use in the evaluation of different

neurological and psychiatric disorders. The use of compartmental models enables measure-

ments of transport and metabolic rates in the brain for FDG. Unfortunately, these transport

and metabolic rates are different for glucose. However, if one knows the relationship

between FDG and glucose for transport and phosphorylation, the measurement of those

rates for glucose are within reach. Based on the principles of competitive substrate kinetics,

a correction constant (lumped constant) can be applied to FDG measurements for the

conversion into corresponding values for glucose.

It is also possible to measure protein synthesis in the brain with PET and amino acids

labeled with either 13N or 11C, e.g., 13N-Methionine and 11C-L-Leucine. The main

application of these tracers has been in the assessment of tumor metabolism.

3. Tracers for the Study of Neurotransmission

One of the most challenging chapters in functional brain imaging is the in vivo evalua-

tion of neurotransmission. This includes two main sites of interest: (1) evaluation of the

functional integrity of presynaptic neurons of different systems with radioligands for

several sites, such as 18F-DOPA, an intraneuronal marker of the dopa-decarboxylase; and

(2) study of postsynaptic neuronal response by using radioligands to study the

availability/distribution of postsynaptic receptors such as 11C-methylspiperone and 18F-

ethylspiperone as PET examples of ergocalciferol (D2) dopamine receptor ligands. Today,

there are several tracers available for the examination of various neurochemical transmitter

systems, including serotoninergic (ketanserin and nitroquipazine NQP), opiate (carfentanil),

and cholinergic (scopolamine and vesamicol) systems, as well as benzodiazepine receptors.

Clinical applications1.Detection of Brain Tumors

The incidence rate for primary brain tumors in the United States is 8 per 100,000 per

year. This translates into about 17,000 new cases annually. Gliomas are responsible for

9,000 of these cases and have an incidence rate of 4.4 per 100,000 per year. Astrocytoma is

by far the most common form of glioma. This neoplasm exists in pathologically distinct

forms that vary in terms of their growth rate. Hence patient survival is highly variable and

depends on the type of brain tumor and the relative proportion of various pathological cell

types within it. The presence of a suspected glioma can be confirmed with x-ray com-

puterized tomography, magnetic resonance imaging, or positron emission tomography.

X-ray computerized tomographic scanning with contrast media defines areas of the

brain where the structural integrity of the blood-brain barrier has been disrupted. This can

be seen when a contrast medium escapes from the blood vessels into the surrounding tissue.

This pathological change is a characteristic feature of blood vessels within the tumor or of

areas of the brain where normal tissue is displaced by tumor. The resolution attainable with

CT may not always clearly delineate the margins of a neoplasm if the blood-brain barrier at

the outer reaches remains sufficiently intact and prevents the escape of adequate amounts of

contrast agent.

Magnetic resonance imaging with or without a contrast agent (gadolinium-DPTA)

provides better resolution and definition than CT. Like CT, the use of contrast only

demonstrates deterioration of the blood-brain barrier; it does not demonstrate the actual

tumor or regions of necrosis. Tumors themselves, however, produce variable diagnostic

MR images, depending on the type of tumor and the extent of associated brain edema or

hemorrhage.

PET imaging with 18-fluorodeoxyglucose (18-FDG) is also a reliable method for

localizing tumors. Images of the tumor are produced with this positron emitter because

neoplastic glioma cells exhibit increased glucose metabolism relative to normal tissue.

Resolution of the histologic margins of the tumor, however, does not appear to be

substantially better than that obtained with MRI gadolinium-DPTA. Preliminary

experiments have shown that positron-emitting radiopharmaceuticals such as PK 11195-

labeled 11-carbon are selectively taken up by tumor cells and may allow for more complete

delineation of neoplasms in the near future. None of the currently available imaging

procedures can identify tumor cells outside the solid tumor mass.

The introduction of '8F-2-deoxyglucose (FDG) to measure the degree of cerebral glucose

use gave a new impetus to the metabolic imaging of brain tumors. Based on Otto Warburg's

suggestion that tumors have higher rates of aerobic glycolysis (lactate production) with

increasing degree of malignancy, several groups around the world have been using FDG

and PET to grade primary tumors in vivo and to distinguish between tumor recurrence and

postirradiation necrosis. Gliomas exhibit increased glucose utilization relative to

surrounding tissue. Studies of glucose metabolism with 18-FDG have shown that PET can

provide a non-invasive method of grading gliomas.

Deterioration of patients’ clinical condition because of either recurrence of a brain

tumor or postirradiation necrosis in the radiotherapy field is not rare. CT scan, MRI, and

cerebral angiography are unable to provide clear differentiation between these conditions.

Intracerebral lesions, either tumor recurrence or radiation necrosis, may show enhancement

with contrast-enhancing agents. The most likely tool to assist in the distinction between

metabolic active tumor (primary or recurrence) and postirradiation necrotic tissue is the

radionuclide method because of its physiological behavior. There are radionuclide tracers

able to study BBB properties, regional cerebral blood flow (perfusion), regional metabolic

rate of glucose use, and amino acid use by tumors. Assessments of glucose metabolism

with 18-FDG currently provide the best available noninvasive method for differentiating

radiation necrosis from recurrent tumor. When changes in postoperative neurological status

signal the spread of tumor beyond its initial site, PET may be a useful addition to

conventional methods in establishing a prognosis or determining if additional surgery is

indicated.

Other tracers labeled with positron emitters (e.g., 18F, '1C, or 13N) have been suggested for

the evaluation of brain tumors. They include sugar derivatives, amino acids, putrescence,

and receptor ligands. Among them, the amino acid 11C-methionine seems to be the most

promising. In 36 patients with cerebral tumors of different grades, "C-methionine uptake in

the tumors correlated with histopathological grades from multiple biopsies. In 23 cases,

delineation of the tumor was more accurate with PET and 11C-methionine than with CT

scan.

2. Evaluation of Dementia

Alzheimer’s disease (AD) is the leading cause of dementia in the United States. More

than 2,000,000 people are incapacitated by this disease. The age-specific prevalence rates

increase from about 5 percent to 7 percent at age 60 to somewhere between 20 percent to

40 percent at age 80. At present, no specific curative or palliative treatment exists for this

condition.

Only about 5 percent to 10 percent of cases of dementia that result from causes other

than AD are due to reversible conditions such as B-12 deficiency or hypothyroidism. CT or

MR imaging can identify most if not all of the space-occupying lesions, and vascular

lesions, that are responsible for the relatively small number of cases of dementia that result

from these causes.

PET has been extensively used in the evaluation of patients with dementia, particularly

those with Alzheimer's disease (AD), multi-infarct dementia (MID), dementia of the frontal

lobe type (FLD), and, more recently, acquired immunodeficiency syndrome (AIDS)

dementia complex (ADC). The usefulness of PET in these disorders is gaining greater

acceptance, especially in the early and differential diagnosis of the different types of

dementia. In AD, the accuracy of the clinical in vivo diagnosis is considered to be 80%,

with a consequent false positive ratio of 20%. The diagnosis of AD is based on several

criteria, including routine blood tests, neurological and neuropsychological evaluation, and

morphological imaging (CT or MRI), and is mainly made through the application of

exclusion criteria. However, in its early phase, patients usually present with mild memory

loss and/or personality changes, making the differential diagnosis with other disorders of

the elderly very difficult when based on clinical criteria alone. PET adds objective

information, enabling separation between cases of AD, MID, FLD, and sometimes ADC

PET Findings

The evaluation of AD with PET has particularly relied on the use of '8F-FDG, although also

on studies of cerebral blood flow and oxygen metabolism. The most common finding in

patients with AD is a reduction in glucose metabolism that is primarily seen in the tempo-

roparietal cortex. However, a wide variety of different degrees of glucose metabolism in

these patients has been reported and shown to correlate with the degree of dementia. Not

only does the degree of dementia correlate with PET/FDG patterns, but also, Alzheimer's

disease can be used as a model to evaluate in vivo the functional circuits underlying

cognitive functions, such as memory. Using this approach, a functional network of human

memory similar to the Papez circuit has been identified using PET. A European multicenter

study in AD patients using PET/FDG confirmed these findings and further showed this

technique to be able to consistently discriminate between patients and controls in all

participating institutions. Using a specially defined ratio for the quantitative analysis of

characteristic patterns of glucose hypometabolism, the overall accuracy was as high as

95.8%. Interestingly, this data analysis differed from the most common method, which uses

cortico/cerebellar ratios. The European study used a ratio between the average radioactivity

measured in the associative cortices (typically most affected in AD) and the radioactivity

averaged from samples of primary sensorimotor cortex, primary visual cortex, putamen,

and cerebellum (least affected regions). It appears that the use of such ratios may increase

the sensitivity of PET, i.e., the differentiation between AD and normals. This sensitivity is

probably caused by the fact that cerebellum is not completely spared in AD, i.e., it could be

functionally affected because of diffuse diaschisis, whereas averaging more areas usually

not affected by the disease could reflect the normal brain in AD more accurately.

An increasing number of studies with large series of patients using PET have shown a high

percentage of patients with the typical pattern of posterior temporoparietal glucose

hypometabolism. Improvement in the diagnostic accuracy of PET in dementia has been

achieved by using multiple regression and discriminant analysis of the data. Correct patient

classification was obtained in 87% of the population studied (including demented patients

and normal subjects). The specificity of the temporoparietal hypometabolism in AD

patients can be contested because (1) the same pattern may be found in diseases other than

AD and (2) in patients with probable AD other patterns may be encountered, such as

selective frontal hypometabolism and unilateral disease.

Temporoparietal hypometabolism has been described in demented patients with Parkin-

son's and Jacob-Creutzfeld disease. The differential diagnosis relies on the overall clinical

information, including presentation, neurophysiological and laboratory findings, and CT

and/or MR imaging. PET results must be interpreted with guidance from such information.

Contrary to AD, patients with multi-infarct dementia (MID) usually show multiple foci of

hypoperfusion and hypometabolism scattered throughout the cortex. These multiple areas

of ischemia are easily detected with MRI. MRI is the technique of first choice to detect

small cortical and subcortical infarcts, involving either grey or white matter. Unfortunately,

it is not rare, particularly in advanced cases, to find a combination of degenerative and

vascular dementia complicating the interpretation of radionuclide imaging data. Some

patients show a mixed pattern of focal deficits and bilateral temporoparietal

hypometabolism. In cases of dementia associated with Huntington's disease, hypoperfusion

and hypometabolism are characteristically present in the heads of caudate nuclei and

cortex, particularly in the frontal lobes. On the other hand, patients with progressive supra-

nuclear palsy (PSP) show a selective frontal hypometabolism. The latter is also found in

patients with Pick’s disease and in others with AD (even in the early phase). The

differential diagnosis may be established by neurological examination, which depicts

typical signs of extra-pyramidal disorders in PSP patients.

Reported unilateral abnormalities in patients with AD are not clearly understood. They

seem to be particularly frequent in very early phases of AD, with correlation between

hypometabolism and clinical symptoms. Frontal abnormalities have been observed in the

early phase of the disease with low specificity for the diagnosis of probable AD. It may also

be present in a wide variety of diseases and, more recently, it has been suggested that it

represents a special disease entity-dementia of frontal lobe type. This form of dementia is

reported to differ from AD in clinical presentation, neurological signs, profile of

psychological disability, electroencephalography, and PET findings. It may be more

frequent than previously recognized. Functional radionuclide imaging may play an

important role in the identification of this disorder.

Finally, the possibility of aiding in the differential diagnosis between dementia and

major depression has been one of the greatest expectations of radionuclide imaging. Some

reports seemed to raise these expectations by pointing out different patterns of glucose

metabolism and rCBF measured with PET in demented and depressed patients. However,

the manifestations of major depression differ from patient to patient, which is the most

probable explanation for the wide variety of glucose metabolism and rCBF patterns, thus

making data interpretation a rather difficult task. It appears that it is paramount to correctly

identify the clinical manifestations of depression on each patient to purify population

samples before any conclusions are drawn.

Comparison of Functional Imaging with Morphological Imaging in AD

X-ray CT and MRI are well-established tools to be used in the first screening of

dementia by differentiating some of its different causes (e.g., tumors, vascular lesions,

and others). In AD patients, MRI results may either be normal or may show cerebral

atrophy of variable degree, in addition to abnormal signal intensity in the periventricular

white matter. Functional abnormalities observed with PET are thought to be caused by

neuronal loss. However, because of the presence of marked atrophy in some cases,

correction of functional radionuclide images for the presence of atrophy has been

suggested and attempted. Alavi et a1 propose systematic correction for atrophy and

suggest that the new indices are better to differentiate between AD patients and normal

controls. These indices (atrophy-weighed, total-brain metabolism and absolute, whole-

brain metabolism) not only differentiate between patients and controls, but they also

correlate with severity of disease with adequate neuropsychiatric scales. The

combination of anatomical and functional information is shown to increase sensitivity

and specificity of diagnostic imaging, proposing its use to set up diagnostic criteria for

classification of dementia.

3. Evaluation of Intractable Epilepsy

Two-thirds of the 1.2 million to 2.4 million people in the United States who are

afflicted with epilepsy experience partial seizures. This means the abnormal activity that is

responsible for the seizure begins in a specific area of the brain (i.e. the locus). Partial

seizures are designated as simple or complex. Simple partial seizures present with sensory

and/or motor involvement of a single extremity, but there is no loss of consciousness.

Complex seizures involve varying states of altered consciousness. The patient may fall and

exhibit uncoordinated muscle contractions and spasms during this period. Sometimes the

patient remains upright and presents with brief staring episodes. Regardless of the

presentation, most patients lose awareness during this time.

In about half of the cases of partial epilepsy, the pathophysiological activity

responsible for the seizure begins in the temporal lobe and spreads to other parts of the

brain. According to recent estimates, approximately 20 percent to 40 percent of these

seizures are intractable; they cannot be controlled effectively with appropriate anti-

convulsant medications. Approximately 60,000 adult patients are classified as medically

intractable.

The most common surgical treatment for in tractable epilepsy is removal of the

affected portion of the temporal lobe. The majority (55 per-cent to 70 percent) of the

patients who have had temporal lobe resections are seizure-free for at least 5 years after

surgery. The combined mortality and morbidity attendant to these procedures is less than 5

percent.

Although no standard battery of preoperative tests has been established, most

presurgical evaluations require an initial neurophysiological assessment with telemetry and

video monitoring of spontaneous seizures. These methods require hospitalization and allow

for the simultaneous monitoring of electrical activity and semiology (i.e., behavior of the

seizure). These noninvasive approaches provide valuable information about the anatomical

origin of the seizure, but they do not accurately delineate the dysfunctional area. For

example, a seizure resulting from abnormal impulses that originate from subcortical

structures may spread over wide areas of the temporal and parietal lobes. Exclusive reliance

on this information could lead to the removal of more brain tissue than is necessary to

effect a cure. If the results of the monitoring paradigm and MR imaging lead to a diagnosis

of epileptic psychosis or a diffuse brain disease, patients are considered ineligible for

surgery.

Patients with initial findings that suggest the epileptogenic focus resides in the

temporal lobe are further evaluated with the Wada test to determine if the unaffected

temporal lobe is capable of memory function, and to identify the lobe responsible for

language function.

According to expert medical opinion, approximately 80 percent to 90 percent of the pa-

tients with temporal epilepsy that reach this stage of presurgical evaluation require

additional invasive electrophysiological monitoring. Recent estimates indicate that the use

of PET could reduce the need for invasive monitoring by approximately 50 percent. In the

remaining cases, where additional recording with depth and strip electrodes is necessary,

PET can aid in their direct placement.

At certain centers, more than 50 percent of the patients evaluated with PET undergo

surgical resection of the temporal lobe without invasive monitoring. With PET it is now

possible to determine correctly which temporal lobe is affected in about 85 percent of

patients with intractable epilepsy. PET is useful in defining the site of seizure origin within

the temporal lobe; it also can provide information that will allow for resections that are

smaller (i.e., tailored resections) than those of "standardized" protocols. However, the

diagnostic value of PET over other present day approaches has not been well evaluated.

Hemispherectomy can be used to treat certain forms of intractable epilepsy. In the

relatively small number of cases where this intervention is appropriate, it is of paramount

importance to distinguish the completely abnormal hemisphere from the one that is

completely or partially normal. Such distinctions are not always possible by standard

methods because both hemispheres usually appear normal on CT and MRI scans. Recent

studies with PET have demonstrated that it is now possible to make preoperative

assessments of relative differences in the functional integrity of the cerebral hemispheres so

that patients previously designated as inoperable can be offered a surgical option. Long-

term evaluations of this approach are underway.

A growing body of experimental evidence strongly suggests that PET has a role in the

clinical assessment of patients with so-called "infantile spasms." This disorder is classified

as a generalized form of epilepsy based on the observation that abnormal muscle

contractions that occur during the seizure involve the entire body. Recent clinical research

has shown, however, that a proportion of the pediatric patients who present with “infantile

spasms”, suggestive of a diffuse lesion, exhibit focal abnormalities amenable to surgical

resection.

PET Findings

The tracer most frequently used to study patients with epilepsy has been 18F-FDG, a

tracer particularly suited to PET imaging as it provides good quality images, with high

spatial resolution. In addition to that, blood flow and oxygen metabolism studies have also

been performed. More interesting seems to be those PET studies evaluating

neurotransmission (opiate and benzodiazepine receptors).

The reported sensitivity of PET studies to detect focal epileptic foci during the

interictal phase is approximately 80% to 90% in patients suffering from partial seizures. It

is higher (close to 100%) if patients are studied during seizures (ictal phase). Such findings

have been observed with metabolic (18F~FDG) and flow (13NH3) studies, whereas the use of

oxygen-15 inhalation technique showed bilateral deficits in the temporal cortex of patients

with unilateral foci. However, this must be considered cautiously because of the small

number of patients evaluated, only 10, as well as the low resolution of the images. FDG-

PET has been playing a very important role in the localization of epileptogenic foci since

the initial work reported by the UCLA group. FDG-PET compared favorably with

electroencephalogram. In 50 patients with partial epilepsy, PET and EEG detected

abnormalities in 35 and 36 patients, respectively. Among PET techniques, FDG is the most

frequently used in the evaluation of epileptic patients because of its higher resolution with

good quality images, which is essential for the exact localization of epileptogenic foci

required in the presurgical assessment.

Today, one of the most important diagnostic criteria of epilepsy refractory to treatment

is the presence of signal intensity abnormalities (gliosis) detected by MRI. This imaging

modality is already part of the routine work-up of these patients, because there are discrete

structural abnormalities in approximately 80% of cases of epilepsy refractory to treatment.

The comparison of x-ray CT, MRI, and PET findings within the same patient population

with refractory partial seizures showed that MRI detects more lesions than x-ray CT and

that PET could localize an epileptogenic focus in 13 of 14 patients with localizing EEG and

in 8 of 12 patients with nonlocalizing EEG.

Although focal hypometabolism in the interictal phase of epilepsy is often detected,

functional abnormalities in areas of the brain other than the epileptogenic focus may be

present. This finding has been reported in patients with mesial temporal lobe epilepsy

(TLE). In the first work by Sackellares et al, the investigators described the presence of

hypometabolism, which was more marked in the lateral than in the mesial temporal lobe,

and suggested the presence of functional pathways between the mesial and lateral temporal

cortex, which are altered in epilepsy of mesial temporal origin. However, the detection of

mesial temporal lobe activity is affected by the small size of these structures, and the spatial

resolution of the images in this work was relatively low (9.5 mm). More recently, in a study

of 27 patients with mesial TLE, Henry et a1 reported regional hypometabolism in 25 of 27

patients. The hypometabolic areas were ipsilateral to the seizure onset and included the

lateral temporal cortex (78% of patients), the mesial temporal cortex (70%), the thalamus

(63%), the basal ganglia (41%), and the frontal (30%), parietal (26%), and occipital (4%)

cortices. The investigators suggested a pathophysiological role for the thalamus in the

initiation or propagation of temporal seizures and the intricately cognitive dysfunction of

TLE.

Another region that has been reported to be hypometabolic in epilepsy (other than the

foci) is the cerebellum, which is usually bilaterally affected Although the investigators tried

to correlate the cerebellar hypometabolism with the effect of phenytoin (PHT), they found

only a weak inverse correlation between PHT levels and cerebellar glucose metabolism in

patients receiving this drug and concluded that cerebellar hypometabolism is only partly

caused by its effect. PET studies of neurotransmission in epilepsy have also been reported.

The study of benzodiazepine and mu-opiate receptors appears to be particularly interesting.

Although benzodiazepine receptors in the epileptogenic area are decreased, mu-opiate

receptors are increased. These studies may be useful not only for the identification of the

epileptogenic foci, but also for the understanding of the physiopathological mechanisms

underlying seizures. In the older studies with PET, few ictal scans were performed, which

all showed high sensitivity of the technique to localize epileptogenic foci. However, the

long period necessary for uptake and metabolism of FDG in the brain makes the technique

suboptimal for ictal studies and, in this regard, SPECT may present advantages.

4. Evaluation of Transient Ischemic Attack and Stroke

The three most common types of cerebrovascular disease are (1) cerebral infarction, (2)

transient cerebral ischemia without infarction, and (3) intracranial (subarachnoid)

hemorrhage. In all these types of cerebrovascular disease, the main cause of brain

(neuronal) damage is ischemia, which may be transient (TIA) or last for a significant length

of time. In any case, significant changes in brain perfusion occur. These changes, as well as

the consequent changes in glucose metabolism and oxygen extraction and use, can be

measured with the radionuclide method. PET studies made a valuable contribution to the

study of neurophysiological and neuropathological mechanisms in patients with

cerebrovascular disease. However, the development of new tracers to obtain rCBF maps

with SPECT opens new vistas to the investigation of patients with CVD at the District

General Hospital level.

X-ray CT scanning and MRI frequently show no abnormality in TIA. The former is often

negative in acute infarctions (<24 hours of onset), whereas rCBF studies immediately show

the perfusion deficits characteristic of the initial stages of transient or permanent ischemia.

In patients with CVD, the clinical history in conjunction with x-ray CT scanning and MRI

are capable of defining the type, location, and extent of the cortical and subcortical lesions

consequent to the vascular insult. It is usually possible to distinguish a lacunar from an em-

bolic or hemorrhagic infarction. Radionuclide functional imaging is also positive in the

very early phases of cerebral ischemic insult; however, it cannot differentiate the nature of

the lesion, which is paramount when therapeutic decisions have to be made. In addition,

both PET and SPECT will miss small size infarctions, particularly those in the white

matter. However, PET and SPECT may play unique roles in (1) the study of cortical

changes in areas distant from the site of ischemia, probably caused by deafferentation or

diaschisis, which may be either cortical or cerebellar; (2) the diagnosis and location of

acute cerebral infarctions a few hours after the vascular insult, when x-ray CT scan is

negative or equivocal and an immediate intervention is warranted; and finally, (3) the

prognostic assessment, studying reperfusion during the acute or subacute phases, as well as

other parameters.

PET Findings

Oxygen-15 inhalation technique to evaluate rCBF, rCMRO2, rOER, and rCBV in a single

session has been extensively used to study patients with CVD. When necessary and deemed

clinically important, FDG scans may be added at the end of the previous session. Although

highly sensitive in the detection of blood flow and glucose metabolism abnormalities in

patients with cerebral ischemia, PET has low specificity. Therefore, high-resolution struc-

tural MRI must always be used to add diagnostic clinical significance.

A significant contribution to the understanding of the physiopathological changes

occurring during the development of cerebral ischemia and stroke has been made by PET

oxygen-15 studies. When cerebral perfusion pressure starts to decrease, oxygen use

(rCMRO2) is maintained because of an increase of regional cerebral blood volume (rCBV)

first and because of oxygen extraction efficiency (rOER) afterwards. While this goes on,

patients may be totally asymptomatic. When ischemia is present (i.e., patients start to show

symptoms), rCMRO2 decreases to levels still sufficient for cell survival because of

compensatory mechanisms via increased rCBV and rOER. To promptly establish a

therapeutic intervention that could prevent cerebral infarction, the detection of this

particular change, reduction in rCMRO2, is fundamental and is called the therapeutic

window. If perfusion pressure does not improve in a short time or if collateral circulation is

not developed to supply the ischemic area, rCMRO2 decreases to a level at which it is no

longer sufficient for the survival of cells, thus, infarction is imminent. In this phase, rCBF

and rCBV are variable, changing from a hyperemic state to a final definitive absence of

perfusion in the late phase of infarction. In addition, this method is potentially useful for the

evaluation of patients during the subacute and chronic phases of stroke, in cases of

occlusion, and in stenoses of the carotid arteries. A well-known phenomenon in subacute

and chronic phases of CVD that was first described by PET is the diaschisis. This term,

which was first coined by Von Monakow, describes a transient reduction of neuronal

function occurring at a distance from the site of a circumscribed lesion in the brain. This

phenomenon was observed by electrophysiological experiments, but only the introduction

of PET gave renewed importance to remote effects or diaschisis by showing, for the first

time, the presence of crossed cerebellar diaschisis (CCD). Since then, CCD has been

reported by several investigators. Diaschisis may be recognizable in several areas of the

brain, e.g., the cortex ipsilateral to a subcortical lesion, the thalamus ipsilateral to a cortical

lesion, and the homotopic cortex contralateral to a cortical lesion. The neurological

functional counterpart of these metabolic findings, as well as the time line and frequency of

its occurrence, is not clear.

5. Evaluation of Dopamine system

The dopamine system is involved in the regulation of brain regions that subserve

motor, cognitive and motivational behaviors. Disruptions of dopamine function have been

implicated in neurological and psychiatric illnesses including substance abuse, as well as on

some of the deficits associated with aging of the human brain. This has made the dopamine

system an important topic in research in the neuroscience and neuroimaging as well as an

important molecular target for drug development.

Dopamine cells reside predominantly in the mesencephalon in three neuronal groups:

the retrobulbar, the substantia nigra (SN) and the ventral tegmental area (VTA). Dopamine

neurons from the SN project predominantly to the dorsal striatum and are mainly concerned

with initiation and execution of movements. Those from the VTA project predominantly to

limbic and limbic-connected regions including nucleus accumbens, orbital and cingulate

cortices, amygdala and hippocampus.tnd are involved with reinforcement, motivation,

mood and thought organization. Dopamine neurons from the retrohul bar area project to the

hypothalamus where they regulate hormone secretion from the pituitary. Dopamine cells in

the SN and in the VTA and their major projections are shown in Figure 1.

Dopamine is synthesized in the dopamine neurons where it is stored within vesicles

which protect it from oxidation by monoamine oxidase (MAO). Dopamine is released into

the synapse in response to an action potential and interacts with postsynaptic dopamine

receptors. The concentration of dopamine in the synapse is regulated primarily by its

reuptake by the dopamine transporters, to maintain low (nanomolar) steady-state

concentrations. Dopamine is also removed by oxidation by MAO A in neurons and by

MAO B in glia which surround the dopaminergic nerve terminals and by catecholamine O-

methyltransferase (COMT). Brain dopamine release is regulated by autoreceptor as well as

by other neuroanatomically distinct neurotransmitters through interactions with the

dopamine neuron.

Abnormalities in brain dopamine are associated with many neurological and psychiatric

disorders including Parkinson's disease, schizophrenia and substance abuse. This close

association between dopamine and neurological and psychiatric diseases and with

substance abuse make it an important topic in research in the neuroscience and an important

molecular target in drug development. PET enables the direct measurement of components

of the dopamine system in the living human brain. It relies on radiotracers which label

dopamine receptors, dopamine transporters, precursors of dopamine or compounds which

have specificity for the enzymes which degrade dopamine. Additionally, by using tracers

that provide information on regional brain metabolism or blood flow as well as

neurochemically specific pharmacological interventions, PET can be used to assess the

functional consequences of changes in brain dopamine activity. PET dopamine

measurements have been used to investigate the normal human brain and its involvement in

psychiatric and neurological diseases. It has also been used in psychopharmacological

research to investigate dopamine drugs used in the treatment of Parkinson's disease and of

schizophrenia as well as to investigate the effects of drugs of abuse on the dopamine

system. Since various functional and neurochemical parameters can be studied in the same

subject, PET enables investigation of the functional integrity of the dopamine system in the

human brain and investigation of the interactions of dopamine with other neurotransmitters.

Through the parallel development of new radiotracers, kinetic models and better

instruments, PET technology is enabling investigation of increasingly more complex

aspects of the human brain dopamine system.