doi:10.

1093/brain/awu369 BRAIN 2015: 138; 653–663 | 653

Imaging acetylcholinesterase density in

peripheral organs in Parkinson’s disease with
11
C-donepezil PET

Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023

Trine Gjerløff,1 Tatyana Fedorova,1 Karoline Knudsen,1 Ole L. Munk,1 Adjmal Nahimi,1
Steen Jacobsen,1 Erik H. Danielsen,2 Astrid J. Terkelsen,2 John Hansen,3 Nicola Pavese,1,4
David J. Brooks1,4 and Per Borghammer1

See Stoessl (doi: 10:1093/awu392) for a scientific commentary on this article.

Parkinson’s disease is associated with early parasympathetic dysfunction leading to constipation and gastroparesis. It has been
suggested that pathological a-synuclein aggregations originate in the gut and ascend to the brainstem via the vagus. Our under-
standing of the pathogenesis and time course of parasympathetic denervation in Parkinson’s disease is limited and would benefit
from a validated imaging technique to visualize the integrity of parasympathetic function. The positron emission tomography tracer
5-[11C]-methoxy-donepezil was recently validated for imaging acetylcholinesterase density in the brain and peripheral organs.
Donepezil is a high-affinity ligand for acetylcholinesterase—the enzyme that catabolizes acetylcholine in cholinergic synapses.
Acetylcholinesterase histology has been used for many years for visualizing cholinergic neurons. Using 5-[11C]-methoxy-donepezil
positron emission tomography, we studied 12 patients with early-to-moderate Parkinson’s disease (three female; age 64  9 years)
and 12 age-matched control subjects (three female; age 62  8 years). We collected clinical information about motor severity,
constipation, gastroparesis, and other parameters. Heart rate variability measurements and gastric emptying scintigraphies were
performed in all subjects to obtain objective measures of parasympathetic function. We detected significantly decreased
11
C-donepezil binding in the small intestine ( 35%; P = 0.003) and pancreas ( 22%; P = 0.001) of the patients. No correlations
were found between the 11C-donepezil signal and disease duration, severity of constipation, gastric emptying time, and heart
rate variability. In Parkinson’s disease, the dorsal motor nucleus of the vagus undergoes severe degeneration and pathological
a-synuclein aggregations are also seen in nerve fibres innervating the gastro-intestinal tract. In contrast, the enteric nervous system
displays little or no loss of cholinergic neurons. Decreases in 11C-donepezil binding may, therefore, represent a marker of para-
sympathetic denervation of internal organs, but further validation studies are needed.

1 Department of Nuclear Medicine and PET Centre, Aarhus University Hospital, Denmark
2 Department of Neurology, Aarhus University Hospital, Denmark
3 Department of Health Science and Technology, Aalborg University, Denmark
4 Division of Brain Sciences, Imperial College, London, UK
Correspondence to: Per Borghammer, MD, PhD, DMSc,
Department of Nuclear Medicine and PET Centre, Noerrebrogade 44,
bygn. 10G, kaelderen, DK-8000 Aarhus C, Denmark
E-mail: perborgh@rm.dk

Keywords: autonomic nervous system; positron emission tomography; Parkinson’s disease; parasympathetic nervous system;
acetylcholinesterase
Abbreviations: AChE = acetylcholinesterase; DMV = dorsal motor nucleus of the vagus; RR = distance in ms between consecutive
normal R waves in QRS complexes; SUV = standardized uptake value

Received September 23, 2014. Revised October 29, 2014. Accepted November 02, 2014. Advance Access publication December 24, 2014
ß The Author (2014). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
654 | BRAIN 2015: 138; 653–663
Introduction
Parkinson’s disease is the second-most common neurode-
generative disorder and has a prevalence of 1% in those
over 65 years of age. It is clinically characterized by motor
symptoms of tremor, slowness and rigidity, but the early
involvement of the parasympathetic nervous system has
become increasingly apparent. It has been suggested that
the pathology of Parkinson’s disease, abnormal aggregation
of a-synuclein, initially starts in the autonomic nerve end-
ings of the gastro-intestinal mucosa, perhaps triggered by
an unknown toxin (Hawkes et al., 2007). The disease pro-
cess then spreads in a prion-like fashion via the vagal nerve
to the brainstem. This hypothesis could explain why 80%
of patients have an increased colonic transit time (Jost and
Schimrigk, 1991) and why 34% of patients with
Parkinson’s disease are constipated more than 20 years
before diagnosis (Savica et al., 2009). It also explains the
observation that the dorsal vagal motor nucleus (DMV) is a
primary target structure in the brainstem (Braak et al.,
2003). Furthermore, de novo, untreated patients with
Parkinson’s disease display impaired gastric emptying
(Tanaka et al., 2011), decreased saliva production
(Tumilasci et al., 2006), and decreased heart rate variabil-
ity (Kallio et al., 2000)—all symptoms of a dysfunctional
parasympathetic nervous system. The ability to image the
sympathetic nervous system, with 18F-dopamine PET
(Goldstein et al., 2002) and 123I-meta-iodo-benzyl-guan-
idine (MIBG) SPECT (single photon emission computed
tomography) has been instrumental in demonstrating car-
diac sympathetic dysfunction in Parkinson’s disease
(Haensch et al., 2009), even at a prodromal stage
(Miyamoto et al., 2006). An equivalent imaging technique
to visualize parasympathetic dysfunction is currently not
available, but could significantly improve our understand-
ing of the time course of the denervation in Parkinson’s
disease. Importantly, the ability to image a parasympathetic
nervous system deficit could become an important bio-
marker for diagnosing prodromal disease.
We have validated the novel PET tracer 5-[11C]-methoxy-
donepezil for the in vivo quantification of acetylcholinester-
ase (AChE) density in human peripheral organs (Gjerløff
et al., 2014), and similar validation studies were recently
carried out in rats (Watabe et al., 2014). Donepezil is a
non-competitive, high affinity, reversible antagonist of
AChE with low affinity for butyrylcholinesterase. 11C-done-
pezil undergoes very little peripheral metabolism during a
60-min PET scan (Hiraoka et al., 2009; Gjerløff et al.,
2014), and exhibits intense and highly specific binding to
various internal organs, including the gastro-intestinal tract
and pancreas (Gjerløff et al., 2014). Although AChE is not
exclusively produced by neurons, histological measure-
ments of AChE activity have been used for many years
to quantify integrity of the parasympathetic and enteric
nervous systems (Schmid et al., 1979), and the cardiac
conduction system (Pauza et al., 2013). An early study
T. Gjerløff et al.
demonstrated that sub-diaphragmatic vagotomy induced
a 50% decrease of AChE activity in the upper gastro-
intestinal tract of guinea pigs (Schmid et al., 1979), sug-
gesting that AChE PET imaging could provide a biomarker
of vagal functional integrity.
An imaging marker of vagal function is particularly rele-
vant to Parkinson’s disease, given that the DMV shows a
50% neuron loss and its cholinergic preganglionic para-
sympathetic neurons are known to be preferentially vulner-
able (Eadie, 1963; Gai et al., 1992). The DMV also has the
highest number of ubiquitin-positive neurons of any brain
nucleus in Parkinson’s disease, surpassing even the substan-
tia nigra. Pathological a-synuclein aggregates (Lewy bodies
and neurites) have been reported in the oesophagus,
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
stomach, pancreas and intestine (Beach et al., 2010), and
accumulating evidence suggests that these pathological
inclusions are present years before symptom onset
(Shannon et al., 2012; Hilton et al., 2014). The vast
majority (90%) of a-synuclein pathology in the gastro-
intestinal tract takes the form of Lewy neurites rather
than cytoplasmic Lewy bodies in enteric neurons
(Wakabayashi et al., 1988; Braak et al., 2006; Greene,
2014). The a-synuclein pathology exhibits a rostral-caudal
distribution, being most prevalent in the oesophagus and
stomach (Beach et al., 2010; Annerino et al., 2012). This
distribution closely mirrors the DMV innervation of the
gastro-intestinal tract (Hopkins et al., 1996). Taken to-
gether, these observations suggest that a-synuclein histo-
pathology within vagal efferents to the gut predominates
over histopathology in intrinsic enteric neurons. This view-
point is further supported by the fact that the myenteric
and submucosal plexus exhibits only a slight loss of enteric
neurons in Parkinson’s disease (Lebouvier et al., 2010;
Annerino et al., 2012).
In the present study, we used 11C-donepezil PET to quan-
tify AChE binding in the internal organs of patients with
Parkinson’s disease and healthy control subjects. AChE is
synthesized in the perikaryon of cholinergic neurons and
transported to dendrites and axons (Giacobini, 2000). As
a substantial fraction of efferent DMV neurons are already
lost in clinical Parkinson’s disease, and given that some or
even most of the remaining vagal axons are dysfunctional
because of the presence of Lewy neurites, we hypothesized
that AChE density would be measurably decreased in
organs receiving rich vagal innervation. We performed
60-min dynamic 11C-donepezil PET scans with the heart,
stomach, small intestine, liver, spleen, and kidneys in the
field of view. Arterial blood sampling was performed to
provide an input function and radio-metabolite levels
were measured to allow full PET kinetic modelling. The
first aim of the study was to explore differences in the
11
C-donepezil PET signal between patients with
Parkinson’s disease and healthy subjects. The second
study aim was to interrogate correlations between organ
11
C-donepezil uptake in Parkinson’s disease and severity
of symptoms and measures of parasympathetic function.
11C-Donepezil PET in Parkinson’s disease
The severity of motor and non-motor symptoms, includ-
ing hyposmia, REM sleep behaviour disorder, and consti-
pation, was rated with standard scales. Cardiovagal
dysautonomia was investigated by performing heart rate
variability measurements at rest and while subjects per-
formed deep breathing and a Valsalva manoeuvre.
Finally, solid meal 99mTc-labelled colloid gastric scintig-
raphy was performed to assess gastric emptying time,
which is known to be perturbed in patients with a para-
sympathetic deficit.
Materials and methods
The Central Denmark Region Committee on Health Research
and the Danish Health and Medicines Authority approved
the study protocol. The study was monitored by the GCP
unit at Aarhus University Hospital and registered in the
www.clinicaltrials.gov database (NCT02012595). All subjects
signed a written informed consent form.
Human study population
Twelve patients with idiopathic Parkinson’s disease (Hoehn
and Yahr stage 1–3) were recruited to the study. All were
diagnosed according to UK Brain Bank criteria (Hughes
et al., 2001). The patients were receiving the following
anti-parkinsonian medication: MAOB inhibitor (n = 1), dopa-
mine agonist (n = 1), levodopa + MAOB inhibitor (n = 2),
levodopa + COMT-inhibitor (n = 3), levodopa + dopamine
agonist (n = 5). The control group comprised 12 healthy age
and gender matched control subjects. Exclusion criteria
included systemic diseases, thoracic and abdominal cancers,
previous irradiation to the thoracic and abdominal region, pre-
vious major surgery to the gastro-intestinal tract, neurological
or psychiatric diseases, substance abuse disorders, or medica-
tions affecting acetylcholinesterase.
Clinical assessment and
questionnaires
Neurological evaluations were performed on all subjects prior
to inclusion. Motor symptom severity in the Parkinson’s dis-
ease group was rated with the updated MDS Unified
Parkinson’s Disease Rating Scale (MDS-UPDRS) (Goetz
et al., 2007). Dementia screening was performed using the
Mini-Mental State Examination. Symptoms of REM sleep be-
haviour disorder were assessed using the RBD behaviour ques-
tionnaire (RBDSQ) (Stiasny-Kolster et al., 2007). Constipation
severity was assessed using the Constipation Scoring System
(CSS) (Agachan et al., 1996) and by the sum of the seven
constipation items (Questions 9 to 15) on the Rome III ques-
tionnaire (ROME9–15; Rome, 2006). Chronic constipation was
diagnosed as defined by Rome III criteria (Rome, 2006).
Gastroparesis symptoms were assessed using the ROME III
nausea-vomiting module, which rates meal-related nausea,
vomiting, and food regurgitation. Olfaction was measured
using the Sniffin’ Sticks 16-item identification test (Hummel
et al., 1997).
BRAIN 2015: 138; 653–663 | 655
Gastric scintigraphy
All participants were examined after an overnight fast of 8 h
for solids and 4 h for liquids. Patients were withdrawn from
anti-parkinsonian medication for 4 12 h. Over 10 min the sub-
jects ingested a standardized solid meal of 120 ml egg white
omelette containing 19–37 MBq 99mTc-labelled colloid along
with two slices of bread, 30 g of jam and 120 ml of water.
Scintigraphy was performed with a Siemens Symbia T16
dual-head SPECT/CT gamma camera (Siemens), using a low-
energy high-resolution collimator. Summed 1 min anterior and
posterior images were obtained in the upright sitting position
at 0, 30, 60, 90, 120 and 180 min after finishing the meal.
Semi-automatic data analysis was done on a Hermes worksta-
tion (Hermes Medical Solutions). Total counts were defined on
each image from 2D regions of interest delineating the stomach
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
boundary. The mean gastric retention and the half-emptying
time (T1/2) were computed from time-activity curves.
Heart rate variability
Participants observed the same fasting restrictions as outlined
above. An electrocardiogram from lead II (sampling rate
1000 Hz) was recorded in the supine position during 10 min
rest, during two sessions of nine deep respiratory cycles (six
breath per min), and during two Valsalva manoeuvres (deep
breath followed by exhaling against 40 mmHg for 15 s with an
air leak to ensure an open glottis). Each test was separated by
a 5-min rest period. Precautions before autonomic testing are
described in detail elsewhere (Terkelsen et al., 2012). QRS
complexes were detected, noise or arrhythmic behaviour was
manually removed and heart rate variability measures esti-
mated as expressed elsewhere (Terkelsen et al., 2012).
Autonomic measures were expressed as mean RR (the distance
in ms between consecutive normal R waves in QRS com-
plexes), RMSSD (the square root of the mean squared differ-
ences of successive RR intervals), high frequency power
(0.15–0.4 Hz), total power, and the coefficient of component
variance in the high frequency power bands. The Valsalva
ratio was estimated as the longest RR-interval 30 s after com-
pleting the manoeuvre divided with the shortest RR-interval
induced by the Valsalva manoeuvre, selecting the largest re-
sponse in two tests. For each respiratory cycle the longest RR-
interval was subtracted from the shortest RR-interval and
mean of the five largest differences were estimated. The first
difference was deleted if it was more than twice the remaining.
PET/CT imaging
All participants fasted for a minimum of 8 h and abstained
from drinking for 4 h before PET. The synthesis of 11C-
donepezil and PET/CT imaging of the thorax and abdomen
has been previously described in detail (Gjerløff et al.,
2014). In brief, catheters were inserted in the radial artery
for blood sampling and in the contralateral cubital vein for
injection of 11C-donepezil. A 60-min dynamic PET scan was
performed for each subject using a standardized 60-s bolus.
The PET field of view included the upper abdomen/lower
chest region and covered the heart, liver, intestines, pancreas,
and stomach. Low-dose CT scan was performed for attenu-
ation correction and anatomical localization purposes. The
average injected dose of 11C-donepezil was 469  54 MBq
(range 372–531) for the patient group and 419  121 MBq
656 | BRAIN 2015: 138; 653–663
(range 205–561) for controls. To obtain the input function, 45
manual blood samples were drawn. Plasma and whole-blood
radioactivity was measured in a well counter (Cobra II,
Packard Instrument Co) cross-calibrated to the tomograph.
The parent fraction of 11C-donepezil was determined by
radio-HPLC in plasma extracts from samples taken at 2, 5,
10, 15, 20, 30 and 60 min. List-mode acquired PET data were
binned into 36 frames, 12  10 s, 6  30 s, 5  60 s,
5  120 s, and 8  300 s, and reconstructed using a 3D itera-
tive algorithm (three iterations, 21 subsets) including reso-
lution modelling (TrueX), attenuation correction, and
Gaussian filtering (3 mm). The final image resolution was
4.5 mm full-width at half-maximum. Blood and dynamic PET
data were decay-corrected to the scan start.
Volume of interest definition
Volumes of interest were manually defined (PMOD software)
based on anatomical CT scans using different approaches.
For the small intestine, two volumes of interest were defined,
as previously described (Gjerløff et al., 2014). One large PET
volume of interest was drawn to include all PET signal from
the small intestine at a sufficient distance from neighbouring
‘hot’ organs to avoid spill-over. Another, smaller volume of
interest was automatically derived by segmenting the CT
scan, using the large PET volume of interest as the constrain-
ing search space. The CT scans were smoothed using a 2 mm,
isotropic Gaussian filter, and a cut-off threshold of 20
Hounsfield units was used to define the small intestine.
Voxels, which did not derive from these organs (e.g. vessels),
were removed manually. The final time-activity curves in the
small intestine were obtained by scaling the curve from the
PET-volume of interest to this smaller CT volume. This pro-
cedure was chosen to compensate for frame-by-frame organ
movement due to peristalsis.
In the spleen, volumes of interest were defined in the paren-
chyma 10 mm from the outermost tissue border. The pancreas,
kidney cortex, and left myocardial ventricle were defined on
six consecutive slices using a 10 mm brush size in PMOD. In
the myocardium, we performed an additional subsegmentation
of the volume of interest into septal and lateral wall regions of
interest since a previous study, using 18F-dopamine PET,
demonstrated that sympathetic denervation in Parkinson’s dis-
ease is most profound in the lateral wall (Goldstein et al.,
2002). The anatomical pancreas volume was automatically
derived similar to the procedure described for the small intes-
tine. With the current PET/CT protocol, it was impossible to
derive meaningful image data from the stomach wall. In many
subjects, the stomach is situated adjacent to the pancreas and
liver, which display the highest level of 11C-donepezil uptake
(Fig. 1A). Thus, severe spill-over effects from the liver and
pancreas into the stomach volume of interest were seen in
some subjects. Furthermore, the stomach displays extreme
peristaltic movements leading to considerable PET-to-CT mis-
alignment, as the CT is obtained 460 min before the end of
the dynamic PET scan.
Standardized uptake value and kinetic analyses
Dynamic PET time–activity curves were derived from the vol-
umes of interest, and were kinetically analysed using arterial
metabolite-corrected blood and plasma input functions. We
used a reversible single-tissue compartment model, as we pre-
viously found this model to be optimal and to yield the best
T. Gjerløff et al.
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
Figure 1 Whole body PET. (A) Whole body 11C-donepezil PET
in a healthy volunteer (adapted from Gjerløff et al., 2014). (B)
Representative summed abdominal PET images (55–60 min post-
injection) from a healthy control (top) and a patient with Parkinson’s
disease (bottom).
fits (Gjerløff et al., 2014). Arterial plasma time–activity curves
were used as the input function for the compartmental models
and arterial blood time–activity curves to compute the
blood volume. The one-tissue compartment model derives K1
(ml ml 1 min 1), the influx rate constant of the tracer from
the plasma to the tissue compartment; k2 (min 1), the rate
constant of transfer of tracer out of the tissue compartment;
and Vb (ml ml 1), the fractional blood volume in the tissue
from the time–activity curves. We report estimates of total
distribution volume (Vd = K1 / k2 + Vb). Standard uptake
values (SUVs) were calculated based on body weight, i.e.
SUV = concentration (kBq/ml) / [injected dose (kBq) / body
weight (g)].
Image and statistical analyses
Statistical analyses were performed using STATA 13
(StataCorp). Group comparisons of demographic and clinical
parameters were performed using unpaired t-tests or the non-
parametric equivalent depending on type and normality of the
data. PET SUV data were analysed with multiple regression
analyses with correction for age and gender. Importantly, the
small intestine and pancreas SUVs were further corrected for
organ volume. The pancreas is a slim organ and quite variable
among subjects. Thus, in a very slim pancreas the standard
10 mm volume of interest in the centre of the pancreas
would potentially suffer more from activity spill-over effects
than in a corresponding larger pancreas. In the small intestine,
the inverse problem is true. As the small intestine PET time–
activity curve was scaled to the intestinal volume derived from
the CT, any group difference in intestinal volume could pro-
portionally bias the PET SUV values. Correlations between
SUVs and the kinetic parameters (K1, k2, Vd) were explored
using linear regression. Associations between PET SUV values
and gastric emptying time, ECG measurements, and clinical
scores were investigated with multiple regression analyses.
11C-Donepezil PET in Parkinson’s disease BRAIN 2015: 138; 653–663 | 657

Results PET standardized uptake value data

11
C-donepezil PET showed high signal in the liver, stomach
Demographic and clinical data are summarized in Table 1.
and pancreas, with more moderate tracer uptake in the
The healthy control subjects were age- and gender-matched
brain, salivary glands, heart and small intestine (Fig. 1).
to the patients, who had mild-to-moderate clinical disease.
Decreased 11C-donepezil uptake was visually apparent in
Table 2 summarizes data from questionnaires and clinical
the small intestine of most and the pancreas of several pa-
examinations. The Parkinson’s disease group was signifi-
tients with Parkinson’s disease (Fig. 2).
cantly more affected by nausea-vomiting and constipation,
Figure 3 displays small intestine and pancreas time–
as rated by the Constipation Scoring System and ROME III
activity curves and volume-corrected SUV values for the
questionnaires. Six patients and one control subject had
patients with Parkinson’s disease and healthy controls.
chronic constipation, defined by ROME III criteria
The small intestine SUV values in the Parkinson’s disease
(Rome, 2006). Most patients were hyposmic on testing
group showed a mean 35% decrease (P = 0.003), and a
and the group had significantly more symptoms of REM
mean 22% decrease was seen in the pancreas (P = 0.001;

Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023

sleep behaviour disorder.
Table 3). No significant difference was seen in pancreas
The Parkinson’s disease group had a significantly faster
volume (patients: 79  5 cm3, controls: 72  4 cm3;
mean gastric emptying time than controls. This group dif-
P = 0.29), but the patient group trended towards having a
ference was mainly caused by two patients, who had very
larger small intestine volume (patients: 146  10 cm3, con-
rapid emptying (T1/2 5 27 min). The mean Valsalva ratio
trols: 124  9 cm3; P = 0.065, Mann Whitney). As outlined
was significantly decreased in the Parkinson’s disease
in the ‘Materials and methods’ section, the SUV analyses of
group, signifying the presence of sympathetic cardiac dys-
the intestine and pancreas were corrected for organ volume
autonomia. No group differences were present in any of the
removing it as a confound when assessing group differences
cardiovagal measurements.
in SUV values (Supplementary Fig. 1).
The patient group showed a 9% decrease in myocardial
Table 1 Demographic and clinical data
SUV values (P = 0.044; Table 3), and a positive correlation
Parkinson Controls was seen with age in the myocardium. There was no dif-
Number of subjects 12 12 ference in the lateral wall-to-septum ratio between the
Gender (female/male) 3/9 3/9 groups (P = 0.99). No significant group differences in
Age (years) 68 (40–74) 65 (44–75) SUV values were seen in the spleen and kidney (Table 3).
MMSE 29 (24–30) 29.5 (27–30) No significant effects of age and gender were seen on SUVs
Disease duration (years) 4.5 (0.3–12) – in the intestine, pancreas, spleen, and kidney.
MDS-UPDRS (motor) 25.5 (9–56) –
Hoehn and Yahr (I/II/III) 1/8/3 –
Kinetic PET analyses
Values are median (range).
MMSE = Mini-Mental State Examination; MDS-UPDRS = Movement Disorder The PET kinetic parameters are summarized in Table 3.
Society Unified Parkinson’s Disease Rating Scale.
As previously reported (Gjerløff et al., 2014), a one-tissue

Table 2 Paraclinical parameters and questionnaires

Parkinson’s disease Controls P-value

RBDSQ 5 (2–12) 2 (0–7) 0.002
Olfaction score 7 (1–12) 12 (8–15) 0.000
CSS 7 (1–14) 1 (0–10) 0.005
ROME III
Nausea-vomiting 5 (4–23) 0 (0–3) 0.044
Constipation (9-15) 5.5 (0–20) 0 (0–14) 0.003
Gastric emptying T1/2 (min) 55 (20–102) 62 (52–113) 0.047
Cardiovagal function
Deep breathing (
-BPM) 7.7  5.3 9.4  3.2 0.37
Valsalva ratio 1.43  0.24 1.69  0.15 0.006
RR interval (ms) 1012  182 1037  216 0.78
RMSSD (ms) 27.2  15.0 29.6  26.6 0.82
HF-power (ms2) 360  539 319  345 0.78
CCV-HF (%) 1.37  0.77 1.54  0.88 0.67

Values are median (range) or mean  SD.

RBDSQ = REM Sleep Behaviour Disorder questionnaire; CSS = Constipation Scoring System; BPM = beats per minute; RMSSD = square root of mean squared differences of
successive RR intervals; HF = high frequency.
658 | BRAIN 2015: 138; 653–663 T. Gjerløff et al.

Figure 2 Intestinal PET. 11C-donepezil uptake in small intestine at 60 min post-injection in 12 healthy controls (A) and 12 patients with

Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023

Parkinson’s disease (B). All images are scaled to the same SUV threshold, and the hottest part of the small intestine is shown for each subject. Hot
organs (pancreas, liver) have been masked. The considerably decreased intestinal uptake is visually apparent in most patients.

Figure 3 Time–activity curves in small intestine and pancreas. (A) Time-activity curves in the small intestine from 12 patients with
Parkinson’s disease (PD) and 12 control subjects. (B) Small intestine SUV values at 60 min (grey area in A) were decreased in nearly all patients. (C
and D) Corresponding time–activity curves and SUV values from the pancreas.

compartment model yielded the most robust fits.
Significantly decreased mean 11C-donepezil Vd was seen
in the small intestine (P = 0.001), and a near-significant
mean Vd decrease was seen in the pancreas (P = 0.061)
(Supplementary Fig. 2). No group differences in Vd were
seen for the myocardium, spleen and kidney.
Across the two populations, a significant linear correl-
ation was seen between Vd and SUV in the pancreas, and
a near-significant linear correlation was found in the small
intestine (Fig. 4). After exclusion of a significant SUV out-
lier in the control group (P 5 0.05; Grubb’s test), the
intestinal Vd versus SUV correlation was significant
(r2 = 0.20, P = 0.03). Significant positive Vd versus SUV
correlations were also seen in the myocardium (r2 = 0.19,
P = 0.03), spleen (r2 = 0.26, P = 0.016), and kidney
(r2 = 0.49, P 5 0.001). These results suggest that simple
SUV values at 60 min post-injection are a reasonable
approximation of the distribution volume, as previously
reported (Gjerløff et al., 2014). Finally, no significant be-
tween group differences were seen in K1 values, whereas
the k2 values tended towards being higher in patients
(Table 3). Thus, the between group differences in SUV
11C-Donepezil PET in Parkinson’s disease BRAIN 2015: 138; 653–663 | 659

Table 3 PET SUV and kinetic parameters

Region SUV K1 k2 Vd
Parkinson Control P Parkinson Control P Parkinson Control P Parkinson Control P
Heart 6.8  1.3 7.5  1.0 0.044 0.77  0.14 0.71  0.22 0.43 0.017  0.005 0.017  0.007 0.947 49.4  9.9 45.1  8.7 0.275
Spleen 4.0  0.8 4.2  1.1 0.67 1.74  0.27 1.57  0.50 0.32 0.066  0.012 0.066  0.015 0.950 27.8  5.9 24.7  6.6 0.244
Kidney 2.6  0.6 2.8  0.7 0.44 1.43  0.15 1.54  0.28 0.25 0.085  0.018 0.091  0.030 0.597 18.1  4.4 18.1  2.7 0.978
Small intestine 7.1  0.86 11.0  3.2 0.003 0.58  0.16 0.65  0.18 0.32 0.009  0.004 0.007  0.003 0.059 66.4  15.4 111.9  40.0 0.001
Pancreas 16.5  2.8 21.1  4.2 0.001 1.49  0.24 1.45  0.65 0.65 0.012  0.002 0.010  0.004 0.119 126.2  31.7 167.0  64.2 0.061

Values are mean  SD.

Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023

Figure 4 SUV and Vd correlations in the small intestine and pancreas. (A) In the small intestine, a near-significant linear correlation
between Vd and SUV values was seen. (B) In the pancreas, a highly significant linear correlation between Vd and SUV values was seen.

Figure 5 SUV correlations with gastric emptying times and constipation scores. No significant correlations were seen between
intestinal SUV values and gastric emptying time T1/2 (A) or ROME9-15 constipation scores (B). PD = Parkinson’s disease.

values seems to be related more to faster tracer wash-out
(i.e. k2) in the patient group, than to differences in blood
flow (i.e. K1).
Correlations between PET and
clinical parameters
No significant correlations were detected between intestinal
or pancreatic SUV values and Constipation Scoring System
scores, ROME9–15 constipation score (Fig. 5A), ROME
nausea-vomiting score, or gastric emptying T1/2 values
(Fig. 5B) (all P-values 4 0.3). No significant correlations
were seen between myocardial SUV values and any of the
cardiac autonomic measurements, including deep breathing
(
-BPM), Valsalva ratio, RR-interval, RMSSD, HF-power,
or coefficient of component variance in the high frequency
power bands (all P-values 4 0.3). No correlations were
seen between the PET SUVs in any of the investigated
organs and olfaction or REM sleep behaviour scores.
Discussion
The present study provides the first in vivo PET imaging
evidence of decreased AChE density in the gastro-intestinal
660 | BRAIN 2015: 138; 653–663
tract of patients with Parkinson’s disease. The patients dis-
played a highly significant loss of 11C-donepezil signal in
the small intestine. The gut is richly innervated by the
vagus nerve, and the pattern of innervation displays a
rostral-caudal gradient being most dense in the lower
oesophagus, stomach, and upper small intestines
(Hopkins et al., 1996). This innervation pattern is similar
to the observed pattern of 11C-donepezil binding, which is
highest in the upper gastro-intestinal tract and lower in the
ileum and colon (Fig. 1A). AChE is produced in neuron cell
bodies and transported axonally to nerve terminals
(Giacobini, 2000). Thus, degeneration of vagal efferents
in Parkinson’s disease would be expected to decrease
vagus-derived AChE in the gastro-intestinal tract. In sup-
port of this viewpoint, it has been reported that subtotal
vagotomy resulted in a 40–50% decrease in AChE activity
in the myenteric plexus of the duodenum and upper colon
of guinea pigs (Schmid et al., 1979).
It should be emphasized, however, that AChE is not a
specific marker for vagal cholinergic parasympathetic nerve
terminals. An abundance of enteric neurons are cholinergic
and express AChE (Anlauf et al., 2003). Therefore,
decreased 11C-donepezil signal (i.e. AChE density) in the
gut could also be caused by loss of enteric neurons as
well as vagal efferents. However, previous studies have
failed to demonstrate any appreciable loss of enteric neu-
rons in Parkinson’s disease (Lebouvier et al., 2010;
Annerino et al., 2012). Moreover, the bulk of the gastro-
intestinal a-synuclein pathology seen in Parkinson’s disease
comprises Lewy neurites distinct from the enteric neurons
(Wakabayashi et al., 1988; Braak et al., 2006; Greene,
2014) probably originating from the vagus nerve (Beach
et al., 2010). Rodent studies have reported that only a mi-
nority (13–22%) of myenteric neurons were positive for
a-synuclein, whereas all preganglionic efferent vagal
axons and nerve terminals were a-synuclein positive
(Phillips et al., 2008). Furthermore, vagotomy dramatically
decreased the presence of a-synuclein positive axons and
varicosities in the myenteric plexus. Taken together, this
evidence suggests that a-synuclein positive nerve fibres in
the gut of patients with Parkinson’s disease are primarily of
preganglionic parasympathetic origin, in line with the
50% loss of efferent motor neurons in the DMV
(Eadie, 1963; Gai et al., 1992). The preferential loss of
preganglionic parasympathetic efferents from the DMV,
rather than a loss of enteric neurons, seems therefore to
be the most plausible explanation for our finding of
decreased 11C-donepezil signal in the small intestine of pa-
tients with Parkinson’s disease. However, even though the
number of cholinergic enteric neurons may remain intact in
Parkinson’s disease, it is possible that AChE levels are
downregulated in these enteric neurons or indeed in the
vagal nerve terminals themselves. One study reported upre-
gulation of acetylcholine production and downregulation of
AChE in inflamed rat intestine subsequent to nematode
parasite infection (Davis et al., 1998). Another study re-
ported downregulated AChE in circulating blood in
T. Gjerløff et al.
patients with inflammatory bowel disease (Maharshak
et al., 2013). To our knowledge, the relative magnitudes
of AChE expression in different cell structures of the intes-
tine in Parkinson’s disease has not been explored in detail,
and downregulation remains a competing explanation for
the 11C-donepezil signal decrease seen in the small intestine.
Of note, with the present PET/CT protocol, it was impos-
sible to extract valid SUV values or time–activity curves
from the stomach, due to peristalsis and it being adjacent
to the liver and pancreas. Thus, we cannot comment on
whether a similar between-group difference is present in
the stomach.
The pancreas has received little attention in the context
of Parkinson’s disease, but a-synuclein pathology has been
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
reported in this organ (Beach et al., 2010). We found a
significant decrease in the pancreas 11C-donepezil signal
of patients with Parkinson’s disease. The pancreas is a
mixed endocrine and exocrine organ, and receives rich
parasympathetic innervation from the DMV (Fox and
Powley, 1986). Immunohistochemistry studies have demon-
strated intense AChE staining throughout the endocrine
and exocrine pancreas (Delbro, 2012), and several studies
showed that AChE is concentrated in a dense network of
cholinergic nerve fibres in the pancreas, which was
most likely parasympathetic in origin (Hiramatsu et al.,
1993; Ushiki and Watanabe, 1997). The decrease in
11
C-donepezil signal could, therefore, point to parasympa-
thetic denervation of both the endocrine and exocrine
gland. However, other pancreatic cells have recently been
demonstrated to contain AChE. One study demonstrated
AChE expression in the lysosomal system of acinar cells
in the rat exocrine pancreas (Bendayan and Gisiger,
2001). To our knowledge, no histology studies have exam-
ined the distribution of cholinergic markers in the pancreas
of patients with Parkinson’s disease, so it is unknown
which cell structures exhibit decreased AChE expression.
Interestingly, Parkinson’s disease may be associated with
a higher prevalence of diabetes (Cereda et al., 2011), and
several early studies demonstrated that 450% of patients
with Parkinson’s disease show intolerance on the standard
oral glucose test (Lipman et al., 1974; Sandyk, 1993). A
pathological oral glucose response in Parkinson’s disease
could be caused by loss of the early post prandial insulin
response, which is mediated via the DMV and the vagal
parasympathetic efferents (Teff, 2008). We are currently
planning a second 11C-donepezil PET study of patients
with Parkinson’s disease and oral glucose tolerance tests
will also be performed.
The patients manifested only a 9% decrease in the myo-
cardial 11C-donepezil SUV values, far smaller than the de-
creases of sympathetic function reported in 123I-MIBG
SPECT and 18F-dopamine PET studies of the cardiac sym-
pathetic innervation (Goldstein et al., 2002; Haensch
et al., 2009). The exact interpretation of the cardiac 11C-
donepezil signal is unclear at present. The preganglionic
parasympathetic efferents primarily terminate on ganglia,
which are found in connective tissue adjacent to the atrial
11C-Donepezil PET in Parkinson’s disease
wall, pulmonary trunk and superior vena cava (Chow
et al., 2001). Using autoradiography of porcine heart
tissue, we previously demonstrated homogeneous 11C-done-
pezil binding throughout the left ventricular myocardium
(Gjerløff et al., 2014). This is in accordance with a previ-
ous report of AChE-positive nerves being the predominant
neural subtype observed throughout the endo-, epi-, and
myocardium in the porcine heart (Crick et al., 1999).
However, these nerves are mainly postganglionic nerves
and it is not clear whether this neuronal subpopulation
undergoes degeneration in Parkinson’s disease. Moreover,
the cardiovascular preganglionic parasympathetic neurons
arise predominantly from the nucleus ambiguus rather than
the DMV (Greene, 2014) and, in contrast to the DMV, no
appreciable loss of neurons is seen in the nucleus ambiguus
in Parkinson’s disease (Eadie, 1963). In addition, it has
recently been demonstrated that cardiomyocytes synthesize
and secrete acetylcholine in a paracrine fashion (Roy et al.,
2013). It seems likely that cardiomyocytes would also syn-
thesize AChE, although this was not explored by the inves-
tigators. In summary, the decrease in cardiac 11C-donepezil
SUV values was minor and not comparable to the dramatic
decreases of sympathetic innervation seen with PET and
SPECT imaging markers. We saw no group differences in
the lateral wall-to-septum ratio values in contrast to those
reported in 18F-dopamine PET studies (Goldstein et al.,
2002).
We detected no between-group differences in kidney and
spleen 11C-donepezil uptake. These highly perfused organs
display tracer kinetics characterized by a high initial uptake
and subsequent rapid wash-out (Gjerløff et al., 2014). In
humans, the major route of donepezil excretion is renal and
79% of the recovered dose is eventually found in the urine
(Wilkinson, 1999), with the remaining 21% found in
faeces. Therefore, estimation of 11C-donepezil uptake in
the renal parenchyma is difficult. In the spleen, the presence
of parasympathetic innervation is controversial, and may
target only the vessels (Bellinger et al., 1993). We previ-
ously demonstrated that lymphoid nodules are responsible
for most of the splenic 11C-donepezil uptake, consistent
with the presence of AChE in white blood cells (Gjerløff
et al., 2014). In the liver, 11C-donepezil time activity curves
and SUV values were similar between the groups (data not
shown). As 11C-donepezil is metabolized by the liver and
subsequently secreted into the bile, it is not possible to
conduct meaningful estimations of AChE density in the
liver. Importantly, we carefully checked that our 11C-done-
pezil analyses in the upper intestine was not confounded by
radioactive bile metabolites (for more details, see
Supplementary Fig. 3).
Small intestine SUV values did not correlate with disease
duration, motor symptom severity, constipation scores, or
gastric emptying time as measured by scintigraphy.
Myocardial 11C-donepezil uptake did not correlate with
cardiac autonomic measurements. Of note, our patient
group did not have delayed gastric emptying; in fact it
was rapid in some of the patients with Parkinson’s disease.
BRAIN 2015: 138; 653–663 | 661
It has been previously reported that patients with early
Parkinson’s disease display more prolonged gastric empty-
ing than patients at a more advanced disease stage in
whom normal or rapid emptying can be seen (Hardoff
et al., 2001). It has been speculated that this may be
related to levodopa use, and 10 of our 12 patients were
receiving levodopa. Our patients displayed significantly
decreased Valsalva ratios, but showed no significant differ-
ences in cardiovagal measures. It is possible that our
sample of patients with Parkinson’s disease had little myo-
cardial parasympathetic denervation despite the complaint
of constipation by six of the patients.
The reduced small intestine SUV values in the patient
group showed low variance (Fig. 3B) suggestive of a
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
‘floor effect’. This could have reduced our power to
demonstrate a correlation with clinical parameters. An
observation of tight data in the patient group but a wider
variance in the control values is also seen with both
123
I-MIBG (Miyamoto et al., 2006) and 18F-dopamine
(Goldstein et al., 2002) cardiac imaging in patients with
manifest Parkinson’s disease, and these imaging measures
also correlate poorly with measures of orthostatic hypo-
tension and sympathetic heart rate variability measure-
ments (Goldstein et al., 2002; Haensch et al., 2009).
Parkinson’s disease is also characterized by degeneration
of various central nuclei involved in autonomic regulation.
Thus, autonomic symptoms may arise from simultaneous
degeneration in both central and peripheral parts of the
nervous system. Moreover, the enteric nervous system
is complex and displays a great deal of plasticity.
Additionally, even though the brunt of a-synuclein path-
ology is in the form of Lewy neurites, populations of en-
teric neurons do contain a-synuclein inclusions and
this may influence the motility and function of the gastro-
intestinal tract. Chronic use of parkinsonian medication
and laxatives could also affect the presence and severity
of constipation. Thus, failure to find a simple relationship
between autonomic symptoms and PET imaging findings
is perhaps not unexpected. As pointed out above, we
cannot rule out competing causes for the 11C-donepezil
signal decrease, which would be another explanation for
the lack of correlations.
Our kinetic analyses of the PET time–activity curves
demonstrated positive correlations between SUV values
and Vd estimates from a one-tissue compartment model
(Fig. 4). The correlation was tight for the pancreas but
only modest for the small intestine. Extracting valid time–
activity curves from the small intestine was challenging
because of intestinal peristalsis (Gjerløff et al., 2014),
which added noise. Nevertheless, the between-group differ-
ence in small intestine Vd was highly significant.
Importantly, there was no group difference in K1, which
suggests that the group differences in Vd and SUV values
were not caused by differences in perfusion. The small in-
testine and pancreas k2 estimates were larger in the patient
group. Thus, tracer wash-out was in general faster in the
patients, which is consistent with decreased tissue density of
662 | BRAIN 2015: 138; 653–663
AChE. In summary, our findings demonstrate that SUV
values derived from static whole-body 11C-donepezil
uptake at 40–60 min post-injection yield a reasonable esti-
mate of tracer Vd.
Several limitations of the study need to be addressed.
Although AChE histology has been used extensively for
demonstrating cholinergic nerves, it is not a specific
marker of parasympathetic innervation. As discussed
above, other factors such as AChE down regulation in en-
teric neurons is an alternative explanation for our findings.
The validity of the present findings would also be rein-
forced if replicated using a tracer such as 18F-FEOBV
with affinity for the vesicular acetylcholine transporter.
This tracer was recently validated for human use, but to
date only brain kinetics have been explored (Petrou et al.,
2014). The binding affinity of 18F-FEOBV to peripheral
organs in humans has not yet been investigated in detail,
and rapid secretion of radioactive metabolites to the bile
could pose a challenge. Moreover, 18F-FEOBV binding is
also not specific for parasympathetic nerve terminals and
the tracer could accumulate in enteric cholinergic neurons
similar to 11C-donepezil.
In conclusion, using 11C-donepezil PET we detected
markedly decreased AChE density in the small intestine
and pancreas of patients with Parkinson’s disease at an
early-to-moderate disease stage. Given that the enteric ner-
vous system exhibits no appreciable neuron loss in
Parkinson’s disease, it seems likely that progressive
damage to parasympathetic nerve terminals and degener-
ation of the dorsal motor nucleus of the vagus is a major
contributor to the decrease in the 11C-donepezil PET signal.
However, other causes such as downregulation of AChE
expression in enteric neurons cannot be ruled out by this
study. Nevertheless, 11C-donepezil PET may be the first
successful imaging technique to visualize systemic parasym-
pathetic denervation, and has potential applications in
other disorders, including diabetic neuropathy. Future
11
C-donepezil PET studies of patients with de novo
Parkinson’s disease and prodromal subjects with REM
sleep behaviour disorder are warranted to determine
whether similar findings are present. If so, 11C-donepezil
PET could be a valuable imaging biomarker for the early
diagnosis of Parkinson’s disease.
Supplementary material
Supplementary material is available at Brain online.
Funding
The study was funded by the Danish Medical Research
Council, Lundbeck Foundation, Jascha Foundation, and
the Riisfort Foundation.
T. Gjerløff et al.
References
Agachan F, Chen T, Pfeifer J, Reissman P, Wexner SD. A constipation
scoring system to simplify evaluation and management of consti-
pated patients. Dis Colon Rectum 1996; 39: 681–5.
Anlauf M, Schafer MK, Eiden L, Weihe E. Chemical coding of the
human gastrointestinal nervous system: cholinergic, VIPergic, and
catecholaminergic phenotypes. J Comp Neurol 2003; 459: 90–111.
Annerino DM, Arshad S, Taylor GM, Adler CH, Beach TG,
Greene JG. Parkinson’s disease is not associated with gastrointes-
tinal myenteric ganglion neuron loss. Acta Neuropathol 2012; 124:
665–80.
Beach TG, Adler CH, Sue LI, Vedders L, Lue L, White Iii CL, et al.
Multi-organ distribution of phosphorylated alpha-synuclein histo-
pathology in subjects with Lewy body disorders. Acta
Neuropathol 2010; 119: 689–702.
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
Bellinger DL, Lorton D, Hamill RW, Felten SY, Felten DL.
Acetylcholinesterase staining and choline acetyltransferase activity
in the young adult rat spleen: lack of evidence for cholinergic in-
nervation. Brain Behav Immun 1993; 7: 191–204.
Bendayan M, Gisiger V. Demonstration of acetylcholinesterase mo-
lecular forms in a continuous tubular lysosomal system of rat pan-
creatic acinar cells. J Histochem Cytochem 2001; 49: 29–40.
Braak H, de Vos RA, Bohl J, Del Tredici K. Gastric alpha-synuclein
immunoreactive inclusions in Meissner’s and Auerbach’s plexuses in
cases staged for Parkinson’s disease-related brain pathology.
Neurosci Lett 2006; 396: 67–72.
Braak H, Del Tredici K, Rub U, de Vos RA, Jansen Steur EN,
Braak E. Staging of brain pathology related to sporadic
Parkinson’s disease. Neurobiol Aging 2003; 24: 197–211.
Cereda E, Barichella M, Pedrolli C, Klersy C, Cassani E,
Caccialanza R, et al. Diabetes and risk of Parkinson’s disease: a
systematic review and meta-analysis. Diabetes Care 2011; 34:
2614–23.
Chow LT, Chow SS, Anderson RH, Gosling JA. Autonomic innerv-
ation of the human cardiac conduction system: changes from infancy
to senility–an immunohistochemical and histochemical analysis.
Anat Rec 2001; 264: 169–82.
Crick SJ, Sheppard MN, Ho SY, Anderson RH. Localisation and
quantitation of autonomic innervation in the porcine heart I: con-
duction system. J Anat 1999; 195 (Pt 3): 341–57.
Davis KA, Masella J, Blennerhassett MG. Acetylcholine metabolism in
the inflamed rat intestine. Exp Neurol 1998; 152: 251–8.
Delbro DS. Expression of the non-neuronal cholinergic system in rat
beta-cells. Auton Neurosci 2012; 167: 75–7.
Eadie MJ. The pathology of certain medullary nuclei in Parkinsonism.
Brain 1963; 86: 781–92.
Fox EA, Powley TL. Tracer diffusion has exaggerated CNS maps of
direct preganglionic innervation of pancreas. J Auton Nerv Syst
1986; 15: 55–69.
Gai WP, Blumbergs PC, Geffen LB, Blessing WW. Age-related loss of
dorsal vagal neurons in Parkinson’s disease. Neurology 1992; 42:
2106–11.
Giacobini E. Cholinesterases and Cholinesterase Inhibitors. London:
Martin Dunitz Ltd; 2000.
Gjerløff T, Jacobsen S, Nahimi A, Munk OL, Bender D, Alstrup AK, et al.
In vivo imaging of human acetylcholinesterase density in peripheral
organs using 11C-donepezil: dosimetry, biodistribution, and kinetic
analyses. J Nucl Med 2014; 55: 1818–24.
Goetz CG, Fahn S, Martinez-Martin P, Poewe W, Sampaio C,
Stebbins GT, et al. Movement Disorder Society-sponsored revision
of the Unified Parkinson’s Disease Rating Scale (MDS-UPDRS):
Process, format, and clinimetric testing plan. Mov Disord 2007;
22: 41–7.
Goldstein DS, Holmes CS, Dendi R, Bruce SR, Li ST. Orthostatic
hypotension from sympathetic denervation in Parkinson’s disease.
Neurology 2002; 58: 1247–55.
11C-Donepezil PET in Parkinson’s disease
Greene JG. Causes and consequences of degeneration of the dorsal
motor nucleus of the vagus nerve in Parkinson’s disease. Antioxid
Redox Signaal 2014; 21: 649–67.
Haensch CA, Lerch H, Jorg J, Isenmann S. Cardiac denervation occurs
independent of orthostatic hypotension and impaired heart rate
variability in Parkinson’s disease. Parkinsonism Relat Disord 2009;
15: 134–7.
Hardoff R, Sula M, Tamir A, Soil A, Front A, Badarna S, et al. Gastric
emptying time and gastric motility in patients with Parkinson’s dis-
ease. Mov Disord 2001; 16: 1041–7.
Hawkes CH, Del Tredici K, Braak H. Parkinson’s disease: a dual-hit
hypothesis. Neuropathol Appl Neurobiol 2007; 33: 599–614.
Hilton D, Stephens M, Kirk L, Edwards P, Potter R, Zajicek J, et al.
Accumulation of alpha-synuclein in the bowel of patients in the pre-
clinical phase of Parkinson’s disease. Acta Neuropathol 2014; 127:
235–41.
Hiramatsu K, Watanabe T, Ohshima K. A histochemical study of the
distribution of acetylcholinesterase-positive nerves in the goat pan-
creas. Acta Anat (Basel) 1993; 147: 105–11.
Hiraoka K, Okamura N, Funaki Y, Watanuki S, Tashiro M, Kato M,
et al. Quantitative analysis of donepezil binding to acetylcholinester-
ase using positron emission tomography and [5-(11)C-methoxy]do-
nepezil. Neuroimage 2009; 46: 616–23.
Hopkins DA, Bieger D, deVente J, Steinbusch WM. Vagal efferent
projections: viscerotopy, neurochemistry and effects of vagotomy.
Prog Brain Res 1996; 107: 79–96.
Hughes AJ, Daniel SE, Lees AJ. Improved accuracy of clinical diagno-
sis of Lewy body Parkinson’s disease. Neurology 2001; 57: 1497–9.
Hummel T, Sekinger B, Wolf SR, Pauli E, Kobal G. ‘Sniffin’ sticks’:
olfactory performance assessed by the combined testing of odor
identification, odor discrimination and olfactory threshold. Chem
Senses 1997; 22: 39–52.
Jost WH, Schimrigk K. Constipation in Parkinson’s disease. Klin
Wochenschr 1991; 69: 906–9.
Kallio M, Haapaniemi T, Turkka J, Suominen K, Tolonen U,
Sotaniemi K, et al. Heart rate variability in patients with untreated
Parkinson’s disease. Eur J Neurol 2000; 7: 667–72.
Lebouvier T, Neunlist M, Bruley des Varannes S, Coron E,
Drouard A, N’Guyen JM, et al. Colonic biopsies to assess the neuro-
pathology of Parkinson’s disease and its relationship with symp-
toms. PLoS One 2010; 5: e12728.
Lipman IJ, Boykin ME, Flora RE. Glucose intolerance in Parkinson’s
disease. J Chronic Dis 1974; 27: 573–9.
Maharshak N, Shenhar-Tsarfaty S, Aroyo N, Orpaz N, Guberman I,
Canaani J, et al. MicroRNA-132 modulates cholinergic signaling
and inflammation in human inflammatory bowel disease. Inflamm
Bowel Dis 2013; 19: 1346–53.
Miyamoto T, Miyamoto M, Inoue Y, Usui Y, Suzuki K, Hirata K.
Reduced cardiac 123I-MIBG scintigraphy in idiopathic REM sleep
behavior disorder. Neurology 2006; 67: 2236–8.
Pauza DH, Saburkina I, Rysevaite K, Inokaitis H, Jokubauskas M,
Jalife J, et al. Neuroanatomy of the murine cardiac conduction
system: a combined stereomicroscopic and fluorescence immunohis-
tochemical study. Auton Neurosci 2013; 176: 32–47.
Petrou M, Frey KA, Kilbourn MR, Scott PJ, Raffel DM, Bohnen NI,
et al. In vivo imaging of human cholinergic nerve terminals with
BRAIN 2015: 138; 653–663 | 663
(-)-5-(18)F-fluoroethoxybenzovesamicol: biodistribution, dosimetry,
and tracer kinetic analyses. J Nucl Med 2014; 55: 396–404.
Phillips RJ, Walter GC, Wilder SL, Baronowsky EA, Powley TL.
Alpha-synuclein-immunopositive myenteric neurons and vagal preg-
anglionic terminals: autonomic pathway implicated in Parkinson’s
disease? Neuroscience 2008; 153: 733–50.
Rome F. Guidelines–Rome III Diagnostic Criteria for Functional
Gastrointestinal Disorders. Journal of gastrointestinal and liver dis-
eases: JGLD 2006; 15: 307–12.
Roy A, Fields WC, Rocha-Resende C, Resende RR, Guatimosim S,
Prado VF, et al. Cardiomyocyte-secreted acetylcholine is required
for maintenance of homeostasis in the heart. FASEB J 2013.
Sandyk R. The relationship between diabetes mellitus and Parkinson’s
disease. Int J Neurosci 1993; 69: 125–30.
Savica R, Carlin JM, Grossardt BR, Bower JH, Ahlskog JE,
Maraganore DM, et al. Medical records documentation of consti-
Downloaded from https://academic.oup.com/brain/article/138/3/653/332817 by guest on 24 July 2023
pation preceding Parkinson disease: A case-control study. Neurology
2009; 73: 1752–8.
Schmid W, van der Zypen E, Keller H. Die Wirkung einer subtotalen
Vagotomie auf den Plexus myentericus (Auerbach) verschiedener
Darmabschnitte. Acta Anat (Basel) 1979; 104: 36–51.
Shannon KM, Keshavarzian A, Dodiya HB, Jakate S, Kordower JH.
Is alpha-synuclein in the colon a biomarker for premotor
Parkinson’s disease? Evidence from 3 cases. Mov Disord 2012; 27:
716–9.
Stiasny-Kolster K, Mayer G, Schafer S, Moller JC, Heinzel-
Gutenbrunner M, Oertel WH. The REM sleep behavior disorder
screening questionnaire—a new diagnostic instrument. Mov Disord
2007; 22: 2386–93.
Tanaka Y, Kato T, Nishida H, Yamada M, Koumura A, Sakurai T,
et al. Is there a delayed gastric emptying of patients with early-stage,
untreated Parkinson’s disease? An analysis using the 13C-acetate
breath test. J Neurol 2011; 258: 421–6.
Teff KL. Visceral nerves: vagal and sympathetic innervation. JPEN J
Parenter Enteral Nutr 2008; 32: 569–71.
Terkelsen AJ, Molgaard H, Hansen J, Finnerup NB, Kroner K,
Jensen TS. Heart rate variability in complex regional pain syndrome
during rest and mental and orthostatic stress. Anesthesiology 2012;
116: 133–46.
Tumilasci OR, Cersosimo MG, Belforte JE, Micheli FE, Benarroch EE,
Pazo JH. Quantitative study of salivary secretion in Parkinson’s dis-
ease. Mov Disord 2006; 21: 660–7.
Ushiki T, Watanabe S. Distribution and ultrastructure of the auto-
nomic nerves in the mouse pancreas. Microsc Res Tech 1997; 37:
399–406.
Wakabayashi K, Takahashi H, Takeda S, Ohama E, Ikuta F.
Parkinson’s disease: the presence of Lewy bodies in Auerbach’s
and Meissner’s plexuses. Acta Neuropathol 1988; 76: 217–21.
Watabe T, Naka S, Ikeda H, Horitsugi G, Kanai Y, Isohashi K, et al.
Distribution of intravenously administered acetylcholinesterase
inhibitor and acetylcholinesterase activity in the adrenal gland:
11C-Donepezil PET study in the normal rat. PLoS One 2014; 9:
e107427.
Wilkinson DG. The pharmacology of donepezil: a new treatment of
Alzheimer’s disease. Expert Opin Pharmacother 1999; 1: 121–35.