Follow-up of FDG PET imaging in idiopathic REM Sleep Behavior Disorder (RBD)

Movement disorders or neurodegenerative brain diseases in general are difficult
to diagnose at early disease stages. One reason for this is the slow
development (years) of such conditions with only few complaints or unclear
signs at the beginning. Well-known examples of these diseases are Parkinson*s
disease (PD) and Alzheimer*s disease (AD). But there are many less-frequent
conditions as well. An accurate diagnosis as early as possible is, however,
necessary to avoid superfluous further procedures or unjustified treatments,
and, most importantly, to begin with potential protective therapeutic
strategies as they may emerge in the course of time. A correct diagnosis is
equally important to reliably inform the patient about his/her condition and to
anticipate prognostic actions.
Sophisticated neuroimaging methods of the brain provide an inroad into this
problem, in that they can detect pathological changes in the brain pertinent to
the group of diseases investigated here. Structural scan methods like X-ray,
CT, and MRI scans are not helpful in early disease stages, since at that stage,
no clear anatomical changes can be noticed apart from slight-to-moderate global
atrophy. At later stages, some diseases may show structural abnormalities in
some of these conditions. But by that time, the clinical picture is usually
already clear. MRI and CT brain scans, however, can be very useful to exclude a
range of other diseases, mainly of vascular origin which would confound the
results if they were included in the study. In addition: at the research side,
advanced MRI sequences may allow for atrophy corrections or investigation into
white matter components (e.g. diffusion features) which could contribute to
defining the pathological processes.
Radiotracer PET scans, on the other hand, can reflect regional brain
biochemical activity depending on the type of tracer applied. The PET tracer
[18F]-fluorodeoxyglucose (FDG) allows the measurement of regional cerebral
metabolic rate of glucose. FDG is a glucose analogue with physiological aspects
almost identical to glucose. It is transported from the blood to the brain by a
carrier-mediated diffusion mechanism. Glucose is then phosphorylated to
glucose-6-PO4 , and FDG to FDG-6-PO4 , catalysed by hexokinase. While glucose
phosphate is metabolized further to carbon dioxide and water, FDG phosphate is
not a substrate for any enzyme known to be present in brain tissue, and is
trapped for some longer time and therefore a useful imaging marker for the
first step of glycolysis. It is the rate-limiting step for the transportation
of the energy source of the brain. No glycogen is available. Therefore, if in a
location of the brain something goes wrong, it will immediately be reflected in
altered local glucose needs, and thus glucose transport to that brain region.
Large regional brain tissue problems can be seen visually. Small changes,
however, will be invisible, but can still be measured because of the
sensitivity of the method.
In neurodegenerative brain diseases, determined brain regions degenerate with
the characteristic consequence of specific patterns of altered metabolic brain
activity. This happens before clear structural changes can be detected with
other imaging techniques. Measurement of glucose consumption with FDG PET
imaging thus allows us to identify disease-specific cerebral metabolic brain
patterns in several neurodegenerative brain diseases at an early disease stage.
Several methods have been used to identify these metabolic brain patterns.
First, univariate methods like voxel-based statistical parametric mapping (SPM)
were used to identify group differences between patients with neurodegenerative
brain diseases and controls (1,2). However, Scaled Subprofile
modelling/principal component analysis (SSM/PCA), a multivariate method, not
only identifies group differences, but is also able to identify relationships
between relatively increased and decreased metabolic activity within different
brain regions in combined samples of patients and control scans (see Chapter 7
of this protocol *Statistical analysis*) (3,4). By using the SSM/PCA method,
metabolic disease-specific patterns have been developed for several
neurodegenerative diseases (5-7). In close collaboration with the New York
group of Eidelberg, we have installed and operationalized the SSM/PCA method at
our department. This has resulted in the build-up of the GLIMPS project in
collaboration with various RUG and UMCG departments (Target, NeuroImaging
Center (NIC), Department of Nuclear Medicine and Molecular Imaging, Department
of Neurology, and the department of mathematics). In clinical practice, we can
now check each subject for the metabolic patterns of multiple system atrophy
(MSA), progressive supranuclear palsy (PSP), PD, and AD.
REM-sleep-behaviour disorder (RBD) is a parasomnia characterized by apparent
dream-enacting behaviours and loss of normal REM sleep muscle atonia (8). For
establishing the diagnosis, polysomnography (PSG) is required and represents
the clinical gold standard. RBD can be idiopathic, but is commonly associated
with neurodegenerative disorders characterized by *-synuclein deposition,
including PD, MSA, and Lewy body dementia (DLB) (9,10). In a significant
proportion of cases, RBD occurs prior to the development of clinically evident
parkinsonism or dementia, and therefore idiopathic RBD may represent an early
warning of PD, MSA, or DLB. The estimated 5-year risk of developing
neurodegenerative disease after a definite diagnosis of RBD varies between 18 *
45 percent (11,12). This is also depending on the interval between RBD symptom
onset and diagnosis in the patients. A recent study of Schenck et al. (13)
shows that 80.8 percent of patients initially diagnosed with idiopathic RBD
eventually developed parkinsonism/dementia after a mean interval of 14 years,
with a range of 29 years. Thus, idiopathic RBD in middle-aged and older adults
commonly heralds future parkinsonism. However, the interval between the
diagnosis of RBD and the development of a neurodegenerative disease can be very
long. With idiopathic RBD being possibly an early feature of parkinsonism, and
the importance of an early diagnosis in parkinsonism, it is now possible to
investigate whether metabolic patterns are present in patients with idiopathic
RBD, and if they could contribute to an earlier diagnosis.
The final goal is to predict which RBD patient develops which condition (PD,
DLB, or MSA) at which time interval. This prediction may be reached with
FDG-PET PDRP expression alone, but could perhaps be improved by taking into
account other known markers of phenoconversion as well: loss of olfactory
function (predictive of PD and DLB, but not MSA), and loss of presynaptic
dopaminergic markers ( DAT SPECT; predictive of PD, DLB and MSA).

1) Network patterns will be extracted from the FDG PET data images of 100 RBD
participants. The key measure will be the expression of the PDRP pattern at the
time of the clinical diagnosis RBD (polysomnographically confirmed). The basis
for this objective is given by the results of REMPET1.
2) To observe how many subjects will phenoconvert to PD, DLB or MSA during 3
years of clinical follow-up. Here it will be very helpful to consider the 30
patients who already participated in REMPET1 in previous years . The PDRP
results will be compared to the DAT-SPECT and olfactory results.
3) A DAT-SPECT scan of the brain will be performed to establish the degree of
nigrostriatal dopaminergic activity.
4) An MRI scan of the brain at baseline will be necessary to exclude other
diagnoses and to be able to correct for brain atrophy.
5) Depending on the results, REMPET3 might be considered.

Importantly, in the end we may also be able to identify the RBD patients who
will *never* convert into one of the mentioned diseases and thus will have a
favourable prognosis.
Since the *substance of predictability* will only be available after this
study, comments to the participants concerning the possible future developments
of their individual conditions cannot (and will not) be made. As was done in
REMPET 1, all data will be anonymized prior to the imaging analyses, and the
outcome of the pattern analysis will be disclosed neither to the patient, nor
to the treating neurologist.

Participants with RBD (or expected to have RBD) will be recruited from
referrals to the University Medical Center Groningen (UMCG) or the other Dutch
participating centers. As part of their regular clinical work-up, all subjects
will get an RBD Screening Questionnaire (RBDSQ) (see Appendix 1) (20). If it
concludes a suspicion for RBD (five or more questions answered with *yes*),
they will get a PSG to validate the diagnosis of RBD. If they already have a
PSG-validated diagnosis of RBD, the RBDSQ and PSG will not be done again in the
context of the study. Occasionally, there will be an exception to this, if the
original PSG used for diagnosis was incomplete or incorrectly performed, even
if the available registration may have been sufficient for diagnosis of RBD. In
these cases, the subject may be asked to undergo a properly-done PSG. They will
be asked for permission to do so.
The treating clinician will inform suitable RBD patients about the study and
ask if they are willing to participate. If yes, the treating physician will
inform the research group at the UMCG. They will send the subject the written
information. The subjects will have the opportunity to ask questions by
telephone. The subjects will have two weeks to decide about participation in
the study. Voluntary written informed consent will be obtained from each
subject before performing any study-related procedures. There will be an
independent physician available for further questions (Dr. B.M. de Jong,
neurologist at the UMCG). The subjects will not incur direct costs by
participating in the study.
If the RBDSQ concludes no suspicion for RBD, then subjects cannot participate.
Subjects who are diagnosed with RBD after the video-PSG will undergo a
neurological (UPDRS I-III) and cognitive examination (MoCA). Furthermore, a
smell test (Sniffin* Sticks Test) will be done with all subjects.
Subjects who can be included, will undergo FDG-PET, DAT-SPECT, and MRI scans
according to protocol.

Video-Polysomnography (PSG)
The PSG will be registered during the afternoon, evening, and overnight.
Subjects will be connected to a computerized system with several sensors and
electrodes. The heart rate, respiratory rate, oxygen saturation, presence or
absence of snoring, body position, movements, and the EEG will be recorded. A
video registration will be made as well.
After the PSG, the data will be scored by a technician of the respective sleep
labs or Clinical Neurophysiological Departments of the participating centers.
In addition, the PSGs will anonymously be collected at the University of
Marburg for central reading and quantification.

FDG-PET imaging
Positron emission tomography (PET) is a nuclear medicine imaging technique
which produces a three-dimensional image or map of functional processes in the
body. The system detects pairs of gamma rays emitted indirectly by a
positron-emitting radionuclide, which is introduced into the body on a
biologically active molecule. Images of tracer concentration in 3-dimensional
space within the body are then reconstructed by computer analysis. In this
study, the biologically active molecule is FDG, an analogue of glucose, and the
concentrations of tracer imaged give brain metabolic activity.
FDG-PET imaging will be performed with subjects fasting overnight. At the UMCG,
subjects are scanned in 3D mode using the Siemens Biograph mCT (64 slice)
PET/CT scanner. 200 MBq FDG in 4 mL saline is injected intravenously in a vein
of the arm, and emission data is collected according to the GLIMPS protocol.
Image acquisition will be performed in a resting state, and plasma glucose
levels will be measured.
Subjects will be exposed to radiation. For one FDG-PET scan, subjects will be
exposed to 3,8 mSV. According to recommendations from the International
Commission of Radiological Protection (ICRP) (21), a dose of 3,8 mSV for
volunteers in biomedical research falls into a moderate risk category (1-10
mSV).
The patients who are included in other centers will be scanned locally
according to a similar scan protocol as described above for the UMCG.


DAT-SPECT imaging:
Single-Photon Emission Computed Tomography (SPECT), like PET, is a nuclear
medicine imaging technique which produces a three-dimensional image or map of
functional processes in the body. The system detects the gamma rays directly
emitted by radionuclides (as opposed to positron-emitting radionuclides in PET)
which is introduced into the body on a biologically active molecule. Images of
tracer concentration in 3-dimensional space within the body are then
reconstructed by computer analysis. In this study, the biologically active
molecule is Ioflupane (FP-CIT), which contains a radioactive isotope of iodine
(123I). The concentrations of tracer imaged give the affinity to presynaptic
dopamine receptors, specifically to the dopamine transporter (DAT). DATs,
located on dopaminergic nerve endings, participate in the reuptake mechanism of
dopamine into presynaptic terminals and are modulated by concentrations of
endogenous dopamine. A decrease in DAT density in the striatum has been
associated with PD. DAT imaging can therefore be used as a marker for the
degree of malfunction or loss of dopaminergic nerve endings (28).
DAT-SPECT imaging of Dutch patients will be performed at the Academisch Medisch
Centrum (AMC) in Amsterdam by Prof. dr. Jan Booij. German patients will be
scanned in Germany locally. The radiotracer (123I-ioflupane; 123I-FP-CIT;
marketed as DaTSCAN) will be injected as a bolus in a vein of the arm (at the
AMC a very sensitive brain-dedicated SPECT is used, and consequently a
relatively low dose of 111 MBq is applied). The participants do not have to
fast. Then the images will be acquired 3 hours later (29). In this time
interval, there are no restrictions for the subjects. To acquire the images of
the brain (which will take typically 30-40 min), the subjects will be lying on
their back on the bed of the camera, with their head positioned in the bore of
the SPECT system. Subjects will refrain from taking any medications which could
interfere with DAT-binding (30, see Table 1; chance is low that they have to
refrain). Each subject will first have their thyroid gland blocked with 200 mg
of sodium perchlorate at least 30 minutes prior to radionuclide injection, in
order to prevent free radioactive iodine from accumulating in the thyroid.
Subjects will be exposed to radiation. For one DAT-SPECT scan, subjects will be
exposed to approximately 4,44 mSV. According to recommendations from the
International Commission of Radiological Protection (ICRP) (21), a dose of 4,44
mSV for volunteers in biomedical research falls into a moderate risk category
(1-10 mSV) (31).

MRI imaging
During the study, each subject will undergo an MRI of the brain within two
weeks of the PET scan. The intent of the MRI is to facilitate anatomical
localization and regional analysis by co-registering onto the PET image. The
outcome of the reference MRI scan is not a dependent variable in this study.
Data acquisition will be performed using a 3Tesla Philips Intera MR scanner.
For PET analysis, only a structural T1 weighted MRI is needed.
*Note: Participants should not have any magnetic material (iron after trauma
surgery, abdominal clips, certain tattoos, etc.) in their bodies.


After metabolic brain imaging, Statistical Parametric mapping (SPM12, Wellcome
Department of Imaging Neuroscience, Institute of Neurology, London, UK)
implemented in MATLAB (version 2012b; MathWorks, Natick, MA, USA) will be used
for image processing. In-house written code that was previously validated will
be used to calculate PDRP expression scores (also implemented in MATLAB).

Benefits and Risks Assessment, Group Relatedness
The patients will not incur direct costs by participating in the study. The
burden associated with participation includes the associated travel time over
the course of several visits, and lying in the scanners for several hours, as
described in the protocol. The risks associated with participation include
venepuncture-related pain and hematoma. There is great experience with the
proposed scan investigations in normal diagnostic work-up of patients with
parkinsonisms and significant side effects are not known; The FDG-PET scans in
the context of the REMPET1 study were, as expected, performed without any
problems. It needs to be emphasised that the participating subjects are not
handicapped the way that clinical parkinsonian patients are, but rather suffer
from RBD signs and symptoms at night. The condition can be very disturbing
during sleep, but it poses no problem during daytime. These persons are thus,
in practical terms, in *fit condition.* Benefits associated with participation
in the study includes clarification of disease course * in particular, for
those RBD subjects who are unlikely to convert to PD/DLB/MSA.
This study cannot be conducted without the participation of a sufficient number
of subjects who are suffering from RBD. This study can contribute to a better
understanding of RBD.