Abstract
P001. Increased parietal amyloid burden and metabolic dysfunction in Alzheimer's disease with early onset
VU University Medical Center, Amsterdam, The Netherlands
Background
Alzheimer's disease (AD) with early onset often presents with a distinct cognitive profile, potentially reflecting a different distribution of underlying neuropathology. The purpose of this study was to examine the relationships between age and both in vivo fibrillary amyloid deposition and glucose metabolism in AD.
Methods
Dynamic [11C]Pittsburgh compound-B (PIB) (90 minutes) and static [18F]fluorodeoxyglucose (FDG) (15 minutes) scans were obtained in 100 AD patients and 20 healthy controls. Parametric non-displaceable binding potential images of [11C]PIB and standardized uptake value ratio images of [18F]FDG were generated using cerebellar grey matter as reference tissue. Nine [11C]PIB negative patients were excluded. The remaining patients were categorized into younger (n=45, age: 56±4) and older (n=46, age: 69±5) groups, based on the median age (62) at time-of-diagnosis.
Results
Younger patients showed more severe impairment on visuo-spatial function, attention and executive function composite scores (p<0.05), while we found a trend towards poorer memory performance for older patients (p=0.11). Repeated measures ANOVA showed no main effect of age for [11C]PIB or [18F]FDG, suggesting that overall, the extent of amyloid deposition or glucose hypometabolism did not differ between groups. Regional distributions of [11C]PIB and [18F]FDG (both p for interaction <0.05) differed between groups, however, largely due to increased [11C]PIB binding and decreased [18F]FDG uptake in the parietal cortex of younger patients (both p<0.05). Linear regression analyses showed negative associations between visuo-spatial functioning and parietal [11C]Pittsburgh compound-B binding for younger patients (standardized β: −0.37) and between visuo-spatial functioning and occipital binding for older patients (standardized β: −0.39). For [18F]fluorodeoxyglucose, associations were found between parietal uptake with visuo-spatial (standardized β: 0.55), attention (standardized β: 0.39) and executive functioning (standardized β: 0.37) in younger patients, and between posterior cingulate uptake and memory in older patients (standardized β: 0.41, all p<0.05).
Conclusions
These in vivo findings suggest that clinical differences between younger and older AD patients are not restricted to topographical differentiation in downstream processes, but may originate from distinctive distributions of early upstream events. As such, increased amyloid burden, together with metabolic dysfunction, in the parietal lobe of younger AD patients may contribute to the distinct cognitive profile in these patients.
Regional (A) [11C]PIB BPND and (B) [18F]FDG SUVr in younger and older Alzheimer's disease patients.
P002. Differential impact of apolipoprotein E genotype on distributions of amyloid load and glucose metabolism in Alzheimer's disease
VU University Medical Center, Amsterdam, The Netherlands
Background
Apolipoprotein E (APOE) ε4 is the major genetic risk factor for sporadic Alzheimer's disease (AD), but its impact on pathophysiological processes and neuronal function is not completely understood. The purpose of this study was to examine the relationships between APOE ε4-dose and in vivo distributions of both fibrillary amyloid burden and glucose metabolism in the same AD patients.
Methods
84 AD patients underwent dynamic (90 minutes) 11C-Pittsburgh compound-B (PIB) and static (15 minutes) [18F]fluorodeoxyglucose (FDG) PET scans, and APOE genotyping. Only [11C]PIB-positive patients were included. Parametric non-displaceable binding potential (BPND) images of [11C]PIB and standardized uptake value ratio (SUVr) images of [18F]FDG were generated using cerebellar grey matter as reference tissue. Regions-of-interest are the frontal, parietal, temporal, posterior cingulate, and occipital cortices. AD patients were categorized into APOE ε4 negative (n=22), heterozygous (n=40), and homozygous (n=22) groups.
Results
Multivariate ANOVAs with adjustment for age, gender, and MMSE, showed main effects for APOE ε4-dose for distributions of both [11C]PIB (p<0.05) and [18F]FDG (p<0.01). More specifically, ANOVAs of individual regions showed increased [11C]PIB BPND in the frontal cortex of APOE ε4 non-carriers compared with APOE ε4 carriers (p<0.05). By contrast, APOE ε4 carriers had reduced [18F]FDG uptake in the occipital cortex (p<0.05) and a borderline significant effect in the posterior cingulate (p=0.07) in a dose dependent fashion.
Regional [11C]PIB BPND in APOE ε4 negative, homozygous, and heterozygous AD patients.
Conclusions
We found a reversed APOE ε4-dose effect for amyloid deposition in the frontal cortex, whilst increased APOE ε4-dose was associated with more profound metabolic impairment in the occipital cortex and posterior cingulate. These findings suggest that APOE genotype has a differential impact on the distribution of amyloid plaques and regional glucose metabolism in AD.
P003. Effects of the partial agonist antipsychotic on dopamine synthesis capacity in human brain measured by PET with [C-11]DOPA
National Institute of Radiological Sciences, Chiba, Japan
Background
The partial agonist antipsychotic drugs of dopamine D2 receptors can modulate dopaminergic neurotransmission as functional agonists or functional antagonists. Effects of antipsychotics on the presynaptic functions of dopaminergic neurotransmission might also be related to their therapeutic effects. Recently, we found that the antagonist antipsychotic drug risperidone can be considered to stabilize dopamine synthesis capacity in healthy human subjects [1]. In the present study, changes in dopamine synthesis capacity by the partial agonist antipsychotic drug aripiprazole were measured by PET.
Methods
PET studies were performed on 12 healthy men under resting condition (baseline study) and oral administration of single dose of aripiprazole of 3–9 mg, (drug challenge study) on separate days. In each study, both PET scans with [C-11]raclopride and [C-11]DOPA were performed sequentially. The occupancy of dopamine D2 receptors by aripiprazole was calculated from binding potential values in the striatum for baseline and drug challenge studies with [C-11]raclopride determined by the SRTM method using the cerebellum as a reference region. The uptake rate constant, Ki, for [C-11]DOPA in the striatum indicating the dopamine synthesis capacity was estimated by graphical analysis using the occipital cortex as a reference region.
Results
The occupancies of dopamine D2 receptors were 53%–77%. The dopamine synthesis capacity Ki were 0.0128±0.0016 and 0.0128±0.0014 (1/min) for the baseline and drug challenge studies, respectively, and no significant change in Ki by aripiprazole was observed. No significant correlation between occupancies of dopamine D2 receptors and changes in Ki by aripiprazole was observed. A significant negative correlation was observed between the baseline Ki and the change in Ki by aripiprazole (r=−0.65). The plasma concentrations of aripiprazole during [C-11]raclopride and [C-11]DOPA PET studies ranged from 9.3 to 40.4 ng/mL (23.7±11.3 ng/mL, mean±SD) and from 9.1 to 39.7 ng/mL (21.5±11.0 ng/mL), respectively.
Conclusion
The negative correlation between the baseline Ki and the change in Ki by aripiprazole, and smaller coefficient of variation of Ki in drug challenge studies than in baseline studies indicate that aripiprazole can be assumed to stabilize the level of dopamine synthesis capacity same as antagonist antipsychotic drugs. An abnormal responsivity in both phasic and tonic dopamine release, which might be related to the modulation of dopaminergic neurotransmission, has been considered in the pathophysiology of schizophrenia [2]. Therapeutic effects of aripiprazole on schizophrenia might be related to stabilizing effects on dopaminergic neurotransmission responsivity in dopamine release.
P004. Regional distribution of dopamine D2 receptor occupancy by aripiprazole: a PET study using [11C]raclopride and [11C]FLB457
National Institute of Radiological Sciences, Chiba, Japan
Background
Second-generation antipsychotics demonstrate clinical efficacy with fewer extrapyramidal side effects compared with first-generation antipsychotics. One of the explanations is the hypothesis of preferential extrastriatal dopamine D2 receptor binding by second-generation antipsychotics. In the present study, we focused on aripiprazole, which has a unique pharmacological profile with partial agonism at dopamine D2 receptors and the minimal risk of extrapyramidal side effects. Past studies using [18F]fallypride have reported inconsistent results regarding regional differences of dopamine D2 receptor occupancy by aripiprazole [1,2]. To test the hypothesis of preferential binding to extrastriatal dopamine D2 receptors by aripiprazole, we examined its regional dopamine D2 receptor occupancies in healthy young subjects.
Methods
We performed PET studies on 12 healthy male subjects under resting condition (baseline study) and oral administration of single dose of aripiprazole of 6 mg (drug challenge study) on separate days. Dopamine D2 receptor bindings were measured using [11C]raclopride for the striatum (caudate head and putamen) and [11C]FLB457 for the extrastriatal regions (midbrain, thalamus, parahippocampal gyrus including amygdala, anterior part of the cingulate cortex, frontal cortex, temporal cortex and parietal cortex). For each PET scan, the binding potential (BPND) was calculated using SRTM method using the cerebellum as a reference region. In drug challenge study, the plasma concentrations of aripiprazole and its main metabolite, dehydroaripiprazole, were determined.
Results
Our data showed that dopamine D2 receptor occupancies in the striatum measured with [11C]raclopride were 70.1 and 74.1%, with the corresponding values for the extrastriatal regions measured with [11C]FLB457 ranging from 46.6 to 58.4%. The plasma concentrations of aripiprazole and dehydroaripiprazole in [11C]raclopride PET studies were 29.4±4.8 ng/ml and 1.4±0.6 ng/ml, and the corresponding values in [11C]FLB457 PET studies were 25.6±2.2 ng/ml and 1.7±0.7 ng/ml, respectively.
Conclusion
Preferential extrastriatal dopamine D2 receptor occupancy by aripiprazole was not observed. Considering previous simulation study of non-negligible specific binding in the reference region [3] and difference in plasma concentrations of aripiprazole, dopamine D2 receptor occupancies in both the striatal and extrastriatal regions did not differ. Our data suggest partial agonism at dopamine D2 receptors is the most likely explanation for the minimal risk of extrapyramidal side effects in the treatment by aripiprazole.
P005. Cerebrovascular dysfunction during somatosensory stimulation associated with β-amyloid accumulation in APP transgenic mice
1National Institute of Radiological Sciences, Chiba, Japan; 2University of Electro-Communications, Tokyo, Japan
Background
Cerebrovascular dysfunctions have been implicated in Alzheimer's disease. 1 In order to explore mechanistic links between cerebrovascular dysfunctions and pathogenesis of Alzheimer's disease, we performed a repeated longitudinal evaluation of CBF response to whisker stimulation and accumulation of β-amyloid in the somatosensory cortex of amyloid precursor protein (APP) transgenic and wild-type (WT) mice.
Methods
Long-term measurement of CBF response and behavior activity was performed 2 every 2 to 4 weeks with laser-Doppler flowmetry (LDF) in awake APP transgenic mice aged from 3 to 27 months and WT mice aged from 3 to 37 months. Whisker stimulation was performed to provoke CBF response. On a separate date, amyloid and microvessels were fluorescently labeled with newly developed probe and sulforhodamine 101, respectively, and two-photon imaging (1024 by 1024 pixels) was performed with a z-step size of 4 μm (Figure 1). Vascular and parenchymal amyloid deposits were evaluated separately. The thickness of vascular amyloid was measured at several points along a particular vessel as the difference between the outer diameter of the vessel wall and the outer diameter of the amyloid. Fraction of the vessels covered by amyloid was calculated as the ratio of the length of amyloid-positive sections along the vessel wall against the total vessel length. The parenchymal amyloid was evaluated by calculating the area covered by labeled amyloid in the parenchymal tissue.
Cerebral vessels covered by amyloid.
Results
We observed that percentage increase of CBF during stimulation in APP transgenic mice declined with age from 3 months (21%) to 27 months (2%), while animal locomotion was preserved. In contrast, percentage increase of CBF in WT mice was relatively stable from 3 months (17%) to 37 months (16%). There was a significant difference in the mean percentage increase of CBF between APP transgenic mice and WT mice aged 15 months or older (P<0.01). In concurrence with this alteration, accumulation of amyloid became detectable in the parenchyma and vessel wall of small arteries at 14 months. Parenchymal amyloid gradually accumulated and the vascular amyloid expanded from the arteries to arterioles from 14 to 19 months. The thickness of the vascular amyloid at 19 months was greater by 60% than that at 14 months. The fraction of the vessels covered by amyloid at 14 months and 19 months were 62% and 99%, respectively. The size of parenchymal amyloid at 19 months increased by 70% relative to that at 14 months.
Conclusions
The present results indicate that an age-dependent decline of cerebrovascular function was associated with accumulation of β-amyloid in the brain. We hypothesized that cerebrovascular dysfunction in APP transgenic mice is caused by: i) neural dysfunctions due to parenchymal amyloid deposition; or ii) a mechanical restriction of cerebral vasodilatation by vascular amyloid. To elucidate this issue, it will be required to investigate CBF response to CO2 inhalation in APP transgenic mice showing attenuation of evoked CBF response to whisker stimulation.
P006. Studying the interaction of neurotransmitter networks with PET: an example on serotonin and opioid systems
1Turku PET Centre, Finland; 2University of Helsinki, Finland; 3University of Turku, Finland
Background
Brain operates as a decentralized network and brain functions are simultaneously modulated by several different neurotransmitter systems. Yet, we lack direct in vivo information on neurotransmitter interaction especially at the molecular level. We introduce novel voxel-based positron emission tomography methods for studying internal neurotransmitter network structure and intercorrelations of different neurotransmitter systems in the human brain. We chose serotonin transporter and μ-opioid receptor for this analysis because of their interaction on the cellular level and similar regional distribution in the brain.
Methods
21 healthy subjects underwent two consecutive positron emission tomography scans using [11C]MADAM, a serotonin transporter tracer and [11C]carfentanil, a μ-opioid receptor tracer. First, voxel-by-voxel intratransmitter correlations (hub and seed analyses) were used to study the internal structure of opiate and serotonin systems. Neurotransmitter system interactions were analyzed by computing voxel-level opiate-serotonin intercorrelations.
Results
Regional μ-opioid receptor binding potentials were uniformly correlated throughout the brain. However, our analyses revealed non-uniformity in the serotonin transporter intracorrelations and identified a highly connected local network (midbrain-striatum-thalamus-amygdala). Intercorrelations between the opiate and serotonin tracers were also regionally different, and significant neurotransmitter interaction was found in anteromedial thalamus, amygdala, dorsal anterior cingulate cortex and in dorsolateral prefrontal cortex i.e. in regions relevant for several neuropsychiatric disorders, especially affective disorders.
Conclusions
This methodology enables in vivo mapping of connectivity patterns within and between neurotransmitter systems. Quantification of functional neurotransmitter balance may be a useful approach in etiological studies of neuropsychiatric disorders and also in drug development as a biomarker-based rationale for targeted modulation of neurotransmitter networks.
P007. History of cigarette smoking associated with increased limbic GABAA receptor availability
1Neuropsychopharmacology Unit, Division of Experimental Medicine, Department of Medicine, Imperial College London, UK; 2MRC Clinical Sciences Centre and Division of Medicine, Imperial College London, Hammersmith Hospital, UK; 3MRC Clinical Sciences Centre and Division of Medicine, Imperial College London, Hammersmith Hospital, UK & The Neurodis Foundation, CERMEP Imagerie du Vivant, Lyon, France
Background
Cigarette smoking presents a significant worldwide healthcare challenge. Preclinical, genetic association and clinical trials studies provide considerable evidence for the involvement of the GABA system in the pathophysiology of nicotine addiction (Agrawal et al, 2009; Herman et al, 2011; Markou et al, 2004). However there have been few in vivo neurochemical imaging studies of the GABAergic system in volunteers with a history of cigarette smoking (Esterlis et al, 2009; Staley et al, 2005). We have previously shown that the availability of limbic GABAA receptors in the living human brain can be measured using [11C]-Ro-15 4513 positron emission tomography (PET) (Lingford-Hughes et al, 2002) and that that the availability of both GABAA α1 and α5 receptor subtypes can be estimated from the [11C]-Ro-15 4513 PET signal (Myers et al, 2012). In this study we used a database of [11C]-Ro-15 4513 PET scans to investigate the availability of limbic GABAA receptors in volunteers with a history of cigarette smoking.
Methods
Eight [11C]-Ro-15 4513 PET scans from volunteers with a history of cigarette smoking were compared with twelve scans from volunteers who were non-smokers. All scans were analysed using a region of interest analysis restricted to limbic areas previously implicated in nicotine addiction.
Results
Volunteers with a history of cigarette smoking had significantly higher [11C]-Ro-15 4513 binding in the presubgenual cingulate and parahippocampal gyrus, and at a trend level in the insula, nucleus accumbens and subgenual cingulate (Figure 1). In six abstinent previous smokers (‘ex-smokers’), [11C]-Ro-15 4513 binding was significantly higher in all limbic regions studied with lower GABAA α1 subtype availability in the nucleus accumbens and presubgenual cingulate and higher GABAA α5 subtype availability in the amygdala, anterior cingulate, nucleus accumbens and presubgenual cingulate. The average number of lifetime cigarettes smoked or the mean number of cigarettes smoked per lifetime year did not correlate with total, α1 or α5 subtype [11C]-Ro15-4513 availability in any region studied.
[11C]-Ro15-4513 VT values in the presubgenual cingulate, parahippocampal gyrus, insula and nucleus accumbens for volunteers with a history of cigarette smoking and non-smokers.
Conclusions
These results suggest that limbic GABAA receptor availability is increased in volunteers with a history of cigarette smoking which may reflect either increased expression of GABAA receptors in these areas or decreased endogenous GABA levels. They also suggest downregulation of synaptic limbic GABAA α1 subtype receptors and upregulation of extrasynaptic limbic GABAA α5 subtype occurs in ex-smokers. The findings in ex-smokers imply that alterations in limbic GABAA receptor availability continues with abstinence indicating that this may be a trait marker for nicotine addiction or that alterations in GABA function resulting from cigarette smoking persist.
P008. Altered serotonin release and serotonin 1A receptor density in CaMKIIα-deficient mice assessed by a comparison of in vivo PET and in vitro autoradiographic data
1Molecular Imaging Center, National Institute of Radiological Sciences, Chiba, Japan; 2Institute for Comprehensive Medical Science, Fujita Health University, Toyoake, Aichi, Japan
Background
The alpha isoform of calcium/calmodulin-dependent protein kinase II (CaMKIIα) is the most important serine/threonine-specific protein kinase for the regulation of neural plasticity signaling downstream of excitatory amino acid receptors. Mice deficient in CaMKIIα manifest several serious emotional dysfunctions, such as lack of fear response and hyperaggression. These abnormal behaviors are presumably caused by dysregulated serotonergic neurotransmission, but detailed molecular mechanisms are unclear.
Methods
In the present study, binding of radioligands to serotonin 1A (5-HT1A) receptors in the brain of mice heterozygous for the null mutation (hKO) of CaMKIIα was analyzed by in vivo positron emission tomography (PET) and in vitro autoradiography in a comparative manner in order to clarify functional changes of the serotonergic system by the deficiency of CaMKIIα.
Conclusions
These results implicate decline in serotonergic transmissions via 5-HT1A receptors as a mechanism by which in vivo binding of [18F]MPPF in the CaMKIIα hKO brain was augmented relative to its in vitro binding.
P009. No evidence for additional blood-brain barrier P-glycoprotein dysfunction in Alzheimer's disease patients with microbleeds
VU University Medical Center, PO Box 7057, 1007 MB Amsterdam, The Netherlands: 1Department of Nuclear Medicine & PET Research; 2Department of Neurology & Alzheimer Center; 3Department of Neurology & Alzheimer Center, Epidemiology & Biostatistics; 4PET Centre, Uppsala University Hospital, 751 85 Uppsala, Sweden
Background
Alzheimer's disease (AD) is characterized by amyloid-beta (Aβ) accumulation in both brain parenchyma and cerebral blood vessel walls, the latter known as cerebral amyloid angiopathy (CAA). Microbleeds (MBs) are thought to reflect underlying CAA [1]. Impaired clearance of Aβ from the brain is thought to play an important role in pathogenesis of AD [2, 3]. It has been shown that P-glycoprotein (Pgp), a blood-brain barrier (BBB) efflux transporter, is involved in Aβ transport [4] and decreased BBB Pgp function was found in AD patients using (R)-[11C]verapamil and PET [5]. Decreased Pgp function has also been hypothesized to promote CAA development. The purpose of the present study was to investigate differences in Pgp function between AD patients with and without MBs.
Methods
Eighteen [11C]PIB positive AD patients underwent MRI scans (1.5 T Sonata scanner, Siemens Medical Solutions) including a coronal T1-weighted 3-D MPRAGE for co-registration and region of interest definition and a susceptibility weighted imaging sequence to assess presence of MBs. Dynamic 60 min PET scans were acquired using an ECAT EXACT HR+ scanner (Siemens/CTI, Knoxville, USA) after intravenous injection of 340±37 MBq (R)-[11C]verapamil. Arterial blood was withdrawn continuously and manual samples were taken to obtain a metabolite corrected plasma input curve. Kinetic analysis was performed using nonlinear regression of the standard two-tissue compartment model, including a blood volume component, and fixing the non-specific volume of distribution (= K1/k2) to the mean whole brain grey matter value. Possible differences in K1/k2 ratio were investigated. The non-displaceable binding potential (BPND) was used as outcome measure. Group differences were calculated using Mann-Whitney U tests.
Results
Of the 18 AD patients, 12 had no MBs (AD MB- group) while 6 did have MBs (AD MB+ group) of which 3 patients had a single lobar MB and 3 had multiple lobar MBs (3, 4 and 9, respectively). There were no differences between AD MB- and AD MB+ groups with respect to age, Mini-Mental State Examination (MMSE) scores, injected dose and specific activity of (R)-[11C]verapamil (Table 1). No significant differences were found in BPND between AD MB- and AD MB+ groups (Table 2) and results were essentially the same after PVE correction. K1/k2 ratios did not differ (global cortical brain region: AD MB- 0.27±0.08; AD MB+ 0.29±0.07, p 0.71) between both groups.
Conclusions
No differences were found in BPND of (R)-[11C]verapamil between AD patients with MBs and those without. These results provide no evidence for additional Pgp dysfunction in AD patients with CAA characteristics.
P010. Comparison between neuromelanin-related MRI signal and dopamine transporter binding measured by PET in humans
Molecular Imaging Center, National Institute of Radiological Sciences, Chiba, Japan
Background
The presynaptic function of dopaminergic neurons can be estimated from radiotracer binding to the transporter in PET. 1 In substantia nigra pars compacta (SNc), dopaminergic neurons contain the neuromelanin pigment made from oxyradical metabolites of monoamine neurotransmitters such as dopamine. 2 It has been shown that neuromelanin-related signal is measurable in vivo using the 3T MRI. 3 Both the dopamine transporter and neuromelanin exist in the dopaminergic neurons in SNc. This preliminary study reports the relationship between the binding potential (BPND) of dopamine transporters measured with [18F]FE-PE2I PET and a neuromelanin-weighted MRI ratio (NM-ratio) in human SNc.
Methods
Both the PET and MRI experiments were performed on four normal controls (NCs) and seven Parkinson's disease patients (PDs) (NCs: 63.5±2.1, PDs: 70.1±6.1 years old). A dynamic PET scan was performed for 90 min after intravenous injection of [18F]FE-PE2I. Neuromelanin-weighted (NMW) images were acquired with a 3T MRI scanner using a 2D fast spin echo sequence (TR/TE: 550/11 ms, resolution: 0.45 × 0.64, slice thickness: 2.5 mm). T1-weighted images were also acquired to aid spatial registration between PET and NMW images. Registration and inter-subject anatomical normalization were performed with the statistical parametric mapping software (SPM8). 4 The BPND of dopamine transporter was calculated with a simplified reference tissue model using the cerebellum as a reference region. The anatomically normalized BPND and NMW images were averaged across subjects before regions-of-interest (ROIs) were defined on the averaged images. The NM-ratio was defined as the ratio of the signal intensities in the SNc and decussation of the superior cerebellar peduncles. The NM-ratio and BPND were averaged over the bilateral SNc ROIs. Statistical tests were performed with custom Matlab scripts. A p-value of less than 0.05 was considered significant.
Results
The NM-ratios of the NCs and PDs were 1.11±0.03 and 1.08±0.02, respectively (mean±SD). The BPND of the NCs and PDs were 0.40±0.15 and 0.25±0.05, respectively. Both parameters showed a statistically significant difference between NCs and PDs (Student's t-test, NM-ratio: p=0.036, BPND: p=0.027). The scatter plot of NM-ratios and BPND shown in the attached figure suggests that multimodal imaging with NMW MRI and [18F]FE-PE2I PET of the SNc can distinguish PDs from NCs. Pearson's correlation coefficient between NM-ratios and BPND were 0.22, −0.53 and 0.03 for all participants, NCs and PDs, respectively. There was no statistically significant correlation between the NM-ratio and BPND for any of the groups.
The relation between neuromeranin-related MRI signal (NM-ratio) and binding potential of [18F]FE-PE2I of normal controls (NC) and Parkinson's disease patients (PD).
Conclusions
The NM-ratio may give additional information apart from the presynaptic function measured by dopamine transporter imaging. However, the role of neuromelanin in the dopaminergic system is still controversial, and further studies are necessary.
P011. GABAergic correlates of speech production
Arash Fazl1,
1Mount Sinai School of Medicine, New York, New York, USA; 2Otolaryngology Department, Mount Sinai School of Medicine, New York, New York, USA; 3PET Department and Clinical Center, National Institutes of Health, Bethesda, Maryland, USA
Background
Although the role of gamma-aminobutyric acid (GABA) in regulation of neuronal excitability is well recognized during a wide range of behaviors, its influences on neural control of speech remain unknown. We investigated the GABAergic function associated with speech production by mapping the correlation between GABA-A receptors and functional brain activity in healthy humans.
Methods
Twenty healthy native English speaking right-handed volunteers (mean age 53.2 years, 12 females/8 males) participated in three scanning sessions: PET with [11C]flumazenil (FMZ); fMRI during the sentence repetition; and resting-state fMRI. The whole-brain PET images were acquired following the bolus injection of 20 mCi FMZ while subjects rested with their eyes closed in the quiet environment for 60 min (GE Advance tomograph). PET data analysis included registration of averaged PET images to each subject's structural T1-weighted MRI, calculation of FMZ binding potential (BP) using simplified reference tissue model with pons as a reference region of non-specific activity, and normalization of the resultant BP maps to a standard space.
Whole-brain resting-state images (rs-fMRI) were acquired using a gradient-weighted echo planar imaging (EPI) pulse sequence on a 3.0 Tesla GE scanner (150 contiguous volumes, TR 2 s, TE 30 ms, slice thickness 4 mm, FOV 240 mm, in-plane resolution 3.75 mm, duration 5 min). Resting-state time series were processed accounting for head motion, hardware-related noise and physiological noise.
FMRI data during production of meaningful English sentences were acquired using a gradient-weighted EPI pulse sequence using BOLD contrast and sparse-sampling event-related design (TE 30 ms, TR 10.6 s with 8.6 s for task production and 2 s for image acquisition, FOV 240 × 240 mm, slice thickness 4.0 mm). Task-related responses were analyzed using multiple linear regression; the group activation maps were generated using a mixed-effect design ANOVA (p≤0.05, corrected).
Whole-brain voxelwise Spearman correlation coefficients were computed between speech-induced BOLD signal, resting-state low frequency oscillations and FMZ binding, thresholded at a corrected p < 0.025.
Results
Significant positive relationships between BOLD signal (both speech-related and resting-state) and FMZ binding were found in the parietal operculum, supplementary motor area, precuneus, and superior temporal gyrus, while negative correlations were observed in the inferior frontal gyrus, posterior cingulate cortex and cerebellum. The extent and significance of these correlations were larger during sentence repetition compared to the resting state, which may indicate the enhancement of GABAergic influences on neural activity during actual task production. In addition, significant positive relationships between speech-induced BOLD signal changes and FMZ binding were observed in the laryngeal sensorimotor cortex, supramarginal and angular gyri, inferior parietal lobule, putamen, caudate nucleus and cerebellum, whereas negative correlations were found in the superior parietal lobule, middle temporal gyrus and anterior cingulate cortex.
Conclusions
This is the first study to identify the GABAergic function associated with speech control. Our results suggest that GABAergic transmission may influence the neural activity at the different stages of speech and language control, from auditory perception to motor production.
P012. Nature or nurture? Determining the heritability of human dopamine function: an [18F]-DOPA PET study
1Psychiatric Imaging Group, MRC Clinical Science Centre, Imperial College London, Hammersmith Hospital, London, UK; 2Centre for Neuroimaging Sciences, Institute of Psychiatry at King's College London, London, UK; 3Department of Psychology, University of Virginia, Charlottesville, VA, USA; 4Neuropsychopharmacology Unit, Centre for Pharmacology and Therapeutics, Division of Experimental Medicine, Department of Medicine, Imperial College London, London, UK; 5Division of Experimental Medicine, Department of Medicine, Imperial College London, London UK
Background
Striatal dopamine function is important for personality (Reeves et al., 2007), cognitive function (Cools and D’Esposito, 2011) and behaviour (Gjedde et al., 2010); and abnormalities are linked to a number of neuropsychiatric disorders (Howes and Kapur, 2009; Martinez et al., 2007). However, no studies have examined the relative influence of genetic inheritance and environmental factors in determining striatal dopamine function. Using [18F]-DOPA positron emission tomography (PET), we sought to determine the heritability of presynaptic striatal dopamine function by comparing variability in uptake values in monozygotic (MZ) twins to dizygotic (DZ) twins.
Methods
Nine same sex MZ and ten same sex DZ twin pairs underwent high resolution [18F]-DOPA PET to index presynaptic striatal dopamine synthesis capacity. Uptake values for the overall striatum and functional striatal subdivisions were determined by a Patlak analysis using a cerebellar reference region (Egerton et al, 2010). Heritability (h2), shared environmental effects (c2) and non-shared individual-specific effects (e2) were estimated using a region of interest (ROI) analysis and a confirmatory parametric image analysis.
Results
Overall striatal heritability estimates from the ROI and parametric analyses were 0.4 and 0.33 respectively. We found a distinction between striatal heritability in the functional subdivisions, with the greatest heritability estimates occurring in the sensorimotor striatum (right: 0.51, left: 0.63), and the greatest non-shared individual-specific effect in the limbic striatum (right: 0.69, left: 0.82) where heritability was estimated at zero. Variation in overall presynaptic striatal dopamine function was more strongly influenced by non-shared individual-specific effects (e2: 0.6) than by genetic inheritance, with shared familial environmental effects having no effect (c2: zero).
Conclusions
These findings underline the importance of individual-specific environmental factors for striatal dopaminergic function, particularly in the limbic striatum, with implications for understanding neuropsychiatric disorders such as schizophrenia and addictions.
P013. Flumazenil is a weak substrate for P-glycoprotein in humans: a PET study in pharmacoresistant patients with unilateral MTS
F.E. Froklage1,2, A. Postnov3, R. Boellaard3, R.C. Schuit3, N.H. Hendrikse3, J.C. Reijneveld2, J.J. Heimans2,
1SEIN – Stichting Epilepsie Instellingen Nederland, Heemstede, The Netherlands; 2Department of Neurology, VU University Medical Center, Amsterdam, The Netherlands; 3Department of Nuclear Medicine & PET Research, VU University Medical Center, Amsterdam, The Netherlands; 4Department of Clinical Pharmacology Pharmacy; 5Division of Pharmacology, LACDR, Leiden University, Leiden, The Netherlands
Background
It has been hypothesized that pharmacoresistance in patients with epilepsy is due to limited transport of anti-epileptic drugs across the blood-brain barrier (BBB) as a result of increased activity of multidrug efflux transporters, such as P-glycoprotein (P-gp). Recent preclinical data suggest that flumazenil might be a P-gp substrate. At present, [11C]flumazenil is used clinically as a PET tracer for assessment of GABAA receptor mediated inhibition in epilepsy and for localization of an epileptic focus prior to resective surgery. If flumazenil is indeed a P-gp substrate, increased P-gp function at the BBB could confound interpretation of clinical [11C]flumazenil PET scans.
Methods
Ten pharmacoresistant patients with unilateral mesial temporal sclerosis were included. Two dynamic [11C]flumazenil scans (16 frames, 60 minutes) were acquired on the same day using an HR+ scanner (CTI/Siemens). The morning scan was performed under baseline conditions, the afternoon scan 110 minutes following intravenous administration of 2 mg/kg tariquidar, a P-gp inhibitor. During each scan arterial blood was monitored continuously using an online samplerl. Additional manual samples were used to generate a metabolite corrected arterial plasma input function. In five patients both [11C]flumazenil scans were preceded by [15O]water scans to assess possible changes in blood flow. T1 MRI scans were acquired on a SONATA 1.5T scanner (Siemens) and subsequently used for co-registration (Vinci) and volume of interest delineation (PVElab).
Results
No significant differences in blood flow between morning and afternoon scans were observed. Analysis of [11C]flumazenil studies using the one-tissue plasma input model demonstrated a slight increase in distribution volume (VT) from 5.3±0.8 to 5.7±0.3 with p=0.1 (paired t-test). Using the simplified reference tissue model with pons as reference tissue showed a slight decrease in binding potential (BPND) from 4.7±1.0 at baseline to 4.2±0.5 after tariquidar (p=0.2). The volume of distribution of non-specifically bound [11C]flumazenil (K1/k2) was estimated by taking the ratio of VT and 1+BPND. This increased significantly (p=0.003) from 0.92±0.14 at baseline to 1.12±0.10 after tariquidar.
Conclusions
The significant increase in K1/k2 after tariquidar suggests that flumazenil may be a substrate for P-gp. Whether this small increase in pharmacoresistant patients is clinically relevant remains to be determined.
Acknowledgements
European Project (FP7) EURIPIDES.
P014. Mechanisms linking cannabis use with psychosis and anxiety: studies of the dopaminergic and cannabinoid systems
1MRC Clinical Sciences Centre, London, UK; 2Institute of Psychiatry Kings College London, UK
Background
Cannabis use is associated with a dose dependent increase in the risk of schizophrenia that exceeds that seen with any single gene variant so far identified and the induction of psychotic and anxiety symptoms. However, it is also clear that only some individuals are vulnerable to the effects of cannabis. Rodent studies indicate that cannabis, and its psychotogenic constituent THC, act on the dopaminergic and cannabinoid systems in the brain.
Method
In the first study we used [18F]-DOPA PET to measure dopamine synthesis capacity in regular cannabis users, and measured the severity of psychotic symptoms induced by smoking cannabis. In a second study we used [11C]-MePPEP PET to measure cannabinoid receptor-1 (CB-1) availability, and fMR imaging to measure the effects of THC on fear processing in non-cannabis using individuals.
Results
In study 1 dopamine synthesis capacity was significantly reduced in cannabis users compared with matched healthy controls (p=0.016), and showed no relationship with the severity of psychotic symptoms induced by cannabis. In study 2 CB-1 receptor availability in the amygdala was directly related to the effects of THC on fear processing in the amygdala (r=+0.5, p<0.05), and to the severity of anxiety symptoms induced by THC (r=+0.5, p<0.05).
Conclusions
These findings indicate i) that the mechanism linking cannabis use to psychotic symptoms is not mediated by the same dopaminergic alterations seen in schizophrenia; and ii) that the link between cannabis and anxiety is mediated by the regulation of emotional processing by cannabinoid receptors in the amygdala.
P015. Affinity of commonly-prescribed benzodiazepines for the 18 kDa translocator protein (TSPO): implications for imaging studies
1Neuropsychopharmacology Unit, Imperial College London, UK; 2Imperial College London, UK; 3Imanova Ltd, London, UK
Background
Levels of TSPO binding are being investigated in patients with neuropsychiatric disorders. It is therefore important to quantify the potential occupancy of the TSPO by benzodiazepines, commonly prescribed in such conditions, to determine whether participants on these medications should be excluded from or their impact accounted for in PET studies. At a non-physiological temperature (4°C), the benzodiazepine, diazepam has nanomolar affinity (K i =27.43 nM±8.92) for the TSPO in human brain (1). However, the affinity of benzodiazepines for the TSPO under physiological conditions, ie. 37°C, in human brain is unknown. An additional point of interest is whether benzodiazepines exhibit lower affinities for Low Affinity Binders (LABs) who have an Ala147Thr polymorphism in the TSPO gene which decreases the affinity of TSPO tracers, than for High Affinity Binders (HABs) (2). The aim was to establish the affinity of commonly-prescribed benzodiazepines for the TSPO in human brain under physiological conditions.
Methods
Binding affinity of diazepam, chlordiazepoxide, desmethyldiazepam (active metabolite), lorazepam, midazolam, and oxazepam for the TSPO were measured in post-mortem human brain tissue from donors with no neurological disease, previously characterised as HABs (n=4) or LABs (n=4). Brain tissue was homogenised as described in (3). Tissue membranes were resuspended at a concentration of ∼500 ug/ml and incubated with [3H]PK11195 (5 nM) and unlabelled benzodiazepines at a range of concentrations (1 mM–3 nM) for 60 minutes (37°C). Specific binding was defined using unlabelled PK11195 (10uM). Assays were terminated via rapid filtration and bound radioactivity determined using a Beta scintillation counter. K i was calculated using the Cheng-Prussof equation (4) and a PK11195 K d (29 nM) previously established (3).
Results
Affinities (K i ) of benzodiazepines for the TSPO are shown in Table 1. The highest affinity was demonstrated by diazepam (K i =719 nM). Midazolam, chlordiazepoxide, lorazepam, desmethyldiazepam and oxazepam exhibited much lower affinities (K i >1μM). All compounds with reliable affinity data, demonstrated higher affinity for the HAB than the LAB TSPO.
Affinity of commonly-prescribed benzodiazepines for the TSPO.
Conclusions
Benzodiazepines have a low affinity for the TSPO in likely in vivo conditions. However our study suggests occupancy of TPSO around 10% may occur with high doses of diazepam (eg 80 mg/d) and chlordiazepoxide (30 mg). Within-subject comparative studies are unlikely to be affected, but group comparisons where changes of the same magnitude are expected may be if diazepam/chlordiazepoxide doses are high.
P016. The effects of different cellular environments on opioid receptor binding
Neuropsychopharmacology Unit, Imperial College London, UK
Background
Imaging endogenous opioid peptide (EOP) release using positron emission tomography (PET) would increase our understanding of the opioid systems' role in health and disease. [11C]Carfentanil and [11C]diprenorphine binding is suggested to be sensitive to fluctuations in EOP levels.1,2 Receptor internalisation may contribute to signal changes observed during endogenous release studies. 3 We assessed the binding affinities of five EOPs to rodent opioid receptors (ORs) using [11C]carfentanil and [3H]diprenorphine. Additionally we examined the in vitro binding parameters of both these radioligands in cellular environments representative of those experienced by a receptor following agonist-induced internalisation and assessed the extent each cellular compartment may contribute to overall basal signal observed with [3H]diprenorphine.
Methods
Rat whole brain binding assays were performed using [3H]diprenorphine and [11C]carfentanil. Saturation studies: a range of concentrations (0.003–10 nM) of both ligands were performed in the presence of three buffers representative of different cellular compartments: Extracellular-50 mMTris-HCl, 140 mMNaCl, 5 mMKCl, 1.5 mMMgCl2, 1.5 mMCaCl2, pH 7.4, 37°C; Intracellular-50 mMTris-HCl, 10 mMNaCl, 140 mMKCl, 0.5 mMMgCl2, pH 7.0, 37°C; Endosomal-20 mMMES, 10 mMNaCl, 140 mMKCl, 0.5 mMMgCl2, 0.003 mMCaCl2, pH 6.0, 37°C. To determine EOP affinity: unlabelled peptides were used in the presence of [3H]diprenorphine and [11C]carfentanil (both 0.3 nM) in extracellular buffer at a range of concentrations: β-endorophin (10 pM–10 μM), endomorphin-1 (3 pM–100 μM), endomorphin-2 (3pM–100 μM), met-enkephalin (3 pM–100 μM), leu-enkephalin (3 pM–100 μM). To achieve plasma-membrane, microsomal and cytosolic cell compartments, subcellular fractionation assays were performed according to Laduron. 4 For each fraction, radioligand binding assays were performed using [3H]diprenorphine (5 nM) and Western blot analysis (30 μg/well) using polyclonal antibodies for μ/δ/κ.
Results
A significant reduction in OR density (B max ) was observed in the endosomal versus the extracellular condition for both radioligands (Table 1; p<0.001/p<0.05 for [3H]diprenorphine/[11C]carfentanil). A trend for a reduced affinity (KD) for [11C]carfentanil but not [3H]diprenorphine was observed in the endosomal environment versus extracellular and intracellular (Table 1; p<0.3). The reduced in vitro binding potential exhibited by these radioligands in the endosomal versus the extracellular environment suggests following agonist-induced translocation of ORs to sub-cellular compartments, both radioligands would exhibit a reduced binding ability. Competition studies revealed the presence of two binding sites for all peptides tested with the μ/κ/δ-ligand, [3H]diprenorphine (K i : endomorphin-1: 45±6 nM/7±1 μM; endomorphin-2: 65±10 nM/19±89μM; met-enkephalin: 125±35 nM/18±6 μM; leu-enkephalin: 200±56 nM/44±18μM) with the exception of β-endorphin (K i : 265±0.06 μM); whilst one binding site was observed for all peptides with the μ-ligand, [11C]carfentanil (β-endorphin: 22±15 nM; endomorphin-1: 34±15 nM; endomorphin-2: 58±20 nM; met-enkephalin: 177±110 nM; leu-enkephalin: 125±103 nM). Under basal conditions fractionation studies suggested the majority of total cell protein were located in the cytosol, intense μ/δ/κ immunoreactivity was also observed here. Interestingly, [3H]diprenorphine binding suggested the cytosolic fraction would not contribute to overall tissue binding (<1%), but the majority of total tissue binding would be from plasma-membrane bound (∼70%) and microsomal (∼29%) ORs, a finding consistent to that observed previously with the D2-dopamine system. 5
Comparison of different cellular environments on OR binding.
Mean±SEM, [3H]diprenorphine (n = 4), [11C]carfentanil (n = 3). Data analysis performed using GraphPad Prism 5.0, statistics using SigmaStat one-way ANOVA and Tukey post-test. Binding Potential = Bmax(nM)/Kd(nM).
P017. Association between 5HT1B receptors and cocaine dependence in human subjects
1Yale University School of Medicine, New Haven, Connecticut, USA; 2Bristol-Myers Squibb, Wallingford, Connecticut, USA; 3New York University School of Medicine, New York, New York, USA
Background
Preclinical evidence implicates the 5-HT1B receptor in cocaine dependence (CD). This project has now completed a dataset gathered in humans to explore whether differences exist in 5HT1B receptor distribution in CD as compared to healthy control (HC) subjects.
Methods
14 medically healthy, non-treatment seeking CD subjects (mean age=41±6) were compared to age matched 14 historic HC subjects (mean age 41±8) with no past or current history of cocaine/illicit substance abuse. CD subjects were admitted to a research-dedicated psychiatric inpatient unit where they resided for the duration of the study. After an acclimatization period, participants received a MRI and then a PET scan with [11C]-P943 for purposes of quantifying brain 5-HT1B binding potential (BPND), a measure of available receptors. In order to account for possible partial volume effects, a binary grey matter mask (GMM) was also employed.
Results
Initial regional ROI analyses of [11C]-P943 PET imaging data were performed for amygdala, caudate, anterior cingulate cortex, frontal cortex, hippocampus, hypothalamus, pallidum, putamen, and thalamus and yielded statistically significant reductions in BPND in both frontal cortex (−11%, p=0.01) and anterior cingulate cortices (−17%; p=0.005) in CD subjects. Additional analyses using a grey matter masking (GMM) technique retained statistically significant reductions in [11C]-P943 BPND in the anterior cingulate (−11%, p=0.03) and frontal cortex (−12%, p=0.01), with temporal cortex (−8%; p=0.05) findings emerging.
Conclusions
This study is the first, to our knowledge, to focus on the role of the 5HT1B receptor availability in cocaine dependent humans. These results are consistent with prior findings of frontal cortical dysfunction in cocaine dependence and point to the possible importance of 5HT1B receptors in its pathophysiology.
P018. PET imaging of cerebral nicotinic acetylcholine receptors (nAChRs) in early Alzheimer's disease (AD) assessed with the new radioligand (−)-[18F]-norchloro-fluoro-homoepibatidine ([18F]Flubatine)
1Department of Nuclear Medicine, University of Leipzig, Leipzig, Germany; 2Department of Psychiatry, University of Leipzig, Leipzig, Germany; 3Central Institute for Electronics (ZEL), Research Center Juelich, Juelich, Germany; 4Advanced Biochemical Compounds (ABX), Radeberg, Germany; 5Helmholtz-Zentrum Dresden-Rossendorf, Dresden and Leipzig, Germany
Background
Post mortem studies have shown a degeneration of cholinergic neurons in the brain of AD-patients. Further evidence suggests that the loss of nAChRs is a major contributor to the cognitive deterioration in AD, whereby the alpha4beta2-nAChR subtype is thought to be the most severely reduced in the onset of AD. Using 2-[18F]F-A85380 PET we showed a significant decline in alpha4beta2-nAChRs in early AD-patients which correlated significantly with the loss of cognitive function (Sabri et al. 2008; Kendziorra et al. 2010). However, this tracer was not well suited as a biomarker in a routine clinical set-up for early AD-diagnosis because of unfavourable properties (slow kinetics, long acquisition times up to 7 hours, limited alpha4beta2-receptor-selectivity). We, therefore developed the new radiotracer (−)-[18F]NCFHEB (denominated as [18F]Flubatine) which is an epibatidine derivative with low toxicity in humans), with significantly improved brain uptake, nAChR affinity and selectivity (Brust et al. 2008). Here, we present the results of the worldwide first ongoing [18F]Flubatine-PET study in humans.
Methods
16 mild AD-patients (NINCDS-ADRDA, age 74.4±6.6, MMSE 23.7±2.8) and 11 age-matched healthy controls (HC, age 69.6±5.1, MMSE 28.5±0.9) underwent [18F]Flubatine-PET (370 MBq, 3D-acquisition, ECAT Exact HR+). All were nonsmokers and naïve for central acting medication. In each subject, 4 scans (41 frames) were acquired from 0–270 min post injection and motion correction was performed with SPM2. Kinetic modeling was applied to the VOI-based tissue-activity curves generated for 29 brain regions (irregularly anatomically defined via MRI-coregistration) using a one tissue compartment model with measured arterial input-function. Total distribution volume (DV) and binding potential (BP, reference region: corpus callosum) were used to characterize specific binding. Additionally, parametric images of DV were computed (Logan plot).
Results
Image quality of [18F]Flubatine scans was clearly superior to 2-[18F]F-A85380, and a 20 minutes scan already adequate for visual analysis. All 29 regions were well described with one tissue compartment. PET data acquired over only 90 minutes were sufficient to estimate all kinetic parameters of all VOIs (30 minutes sufficient for modelling of all cortical VOIs respectively) precisely indicating a fast binding kinetic (much faster than for 2-[18F]F-A85380). DVs in HCs increase as expected with receptor density: Corpus callosum (DV: 5.32±0,66), temporal (8.92±0.45), posterior cingulate (9.03±0.55), pons (10.97±0.95), thalamus (24.92±3.56). The AD-patients showed significant BP reductions in distinct cortical regions (p<0.05) compared to HCs.
Conclusions
Due to significant shorter acquisition time, higher brain uptake, faster kinetics and superior image quality [18F]Flubatine appears to be a much more valuable tracer than 2-[18F]F-A85380 to image alpha4beta2-nAChRs in humans. Furthermore, full kinetic modelling (1-tissue compartment model) is accurately possible within 90 minutes in all VOIs, and within 30 minutes in all cortical VOIs. In keeping with its diagnostic properties, early AD-patients show declines of alpha4beta2-nAChRs in distinct cortical regions typically affected by AD-pathology. These results indicate that Flubatine-PET could have a great potential to be tested as a biomarker for early AD-diagnosis.
P019. Decreased μ-opioid receptor binding in brain reward circuitry in morbid obesity
1Turku PET Centre, Finland; 2University of Turku, Department of Radiology, Finland; 3University of Turku, Department of Medicine, Finland; 4Brain Research Unit, Low Temperature Laboratory, Aalto University School of Science and Technology, Finland; 5Department of Biomedical Engineering and Computational Science, Aalto University School of Science and Technology, Finland
Background
Obesity has reached epidemic proportions worldwide, yet little is known about the brain mechanisms that predispose to overeating. Functional imaging studies suggest that overeating might be caused by an imbalance between the reward circuit (including the ventral striatum) and inhibitory networks (including the prefrontal and orbitofrontal cortices). The endogenous opioid system, particularly acting via the μ-opioid receptor, mediates reward, and animal studies have suggested that μ-opioid receptors are involved in the pathophysiology of obesity. Whether μ-opioid receptors are altered in human subjects with obesity is unknown.
Methods
To examine whether morbid obesity is associated with changes in brain μ-opioid receptor binding we recruited 17 morbidly obese women (mean body mass index [BMI] 41.9, mean age 43) and 9 normal-weight women (mean BMI 23.6, mean age 41). We measured brain μ-opioid receptor binding using positron emission tomography (PET) with [11C]carfentanil, a selective radioligand for μ-opioid receptors. Radioactivity was measured with the GE Heatlhcare Discovery 690 PET/CT scanner for 51 min, and receptor binding was expressed in terms of BPND, which is the ratio of specific to non-displaceable binding in brain. BPND was calculated for each voxel using the simplified reference tissue model (SRTM) with occipital cortex as the reference region, and parametric BPND maps were compared between the groups using ROI analysis and full-volume statistical parametric mapping with SPM8.
Results
Morbidly obese patients had on average 37% lower [11C]carfentanil BPND than did control subjects in several brain regions relevant for reward processing in brain, including ventral and dorsal caudate nucleus. Full-volume analysis revealed that BMI correlated negatively with BPND in multiple components of the reward circuit including thalamus, caudate nucleus, amygdala and anterior cingulate cortex, further suggesting a direct link between low μ-opioid binding and obesity.
Conclusions
We found decreased μ-opioid binding in morbidly obese patients in several brain regions implicated in reward processing. Low μ-opioid tone may represent a neural substrate for decreased hedonic responses to eating in obesity, and could thus make these individual susceptible to overeating in order to gain the desired hedonic response. Identifying the brain mechanisms involved in overeating is critical for developing new psychological and pharmacological treatments to curb the obesity epidemic. In the next phase of the study, we will examine whether these changes are reversible after bariatric surgery for obesity. Whether decreased μ-opioid is a cause or consequence of obesity remains to be determined.
(A) Brain regions showing negative association between BMI and BPND. (B) Scatterplot illustrating the relationship between BMI and BPND in thalamus.
P020. Short-term SSRI treatment reduces 5-HT1A autoreceptor binding in the treatment of major depressive disorder
1Department of Psychiatry, Columbia University College of Physicians and Surgeons, New York, New York, USA; 2Department of Molecular Imaging and Neuropathology, New York State Psychiatric Institute, New York, New York, USA; 3Center for Neurobiology and Behavior, Columbia University College of Physicians and Surgeons, New York, New York, USA; 4Department of Biostatistics, Mailman School of Public Health, New York, New York, USA; 5Department of Radiology, Columbia University College of Physicians and Surgeons, New York, New York, USA
Background
Chronic administration of serotonin selective reuptake inhibitor (SSRI) to rodents causes desensitization or downregulation of 5-HT1A autoreceptors in the raphe nuclei (RN). We have previously reported elevated 5-HT1A binding in medication-naïve, but not recently-medicated, subjects with major depressive disorder.
Methods
5-HT1A binding (BPF) was quantified in medication-free subjects using PET with [11C]-WAY-100635 before and after treatment of major depressive disorder (MDD) with an SSRI. 23 subjects completed both [11C]WAY-100635 PET scans with a metabolite-corrected arterial input function and depression severity was rated before and after the treatment course.
Results
13/23 (56%) subjects had a 50% reduction in score on the 24-item HDRS (from 24.3±7.2 to 11.7±7.5; p<0.001 paired-samples t-test). No treatment effect on autoreceptor binding was found in the entire group (df=1,22; F=2.58; p=0.123). After removing subjects (n=4) with known recent history of SSRI exposure, 5-HT1A autoreceptor BPF in the raphe was reduced on SSRI treatment (df=1,18; F=5.12; p=0.036). Degree of reduction in 5-HT1A autoreceptor BPF was unrelated to improvement in depression (df=1,16; F=1.27; p=0.276).
Conclusions
Downregulation of 5-HT1A autoreceptor binding by SSRI treatment of major depression was detected only in not recently treated patients using the antagonist ligand [11C]WAY-100635 and is consistent with animal studies and previous findings. Future studies should determine whether there is receptor desensitization by using an agonist ligand that binds selectively to the high affinity conformation of this receptor.
P021. Effects of NMDA receptor antagonists on the binding of agonistic and antagonistic radiotracers to striatal dopamine D2 receptors
1Molecular Imaging Center, National Institute Radiological Sciences, Chiba, Japan; 2Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden
Background
Ketamine and phencyclidine (PCP) act as noncompetitive antagonists to the N-methyl-D-aspartate (NMDA) receptor. Ketamine and PCP both bind to the PCP recognition site in the Ca2+ channel domain, and exert psychotomimetic and anesthetic effects at relatively low and high doses, respectively. In our previous monkey PET study, we noticed that ketamine and xylazine anesthesia increased the binding potential of the dopamine D2 receptor agonist [11C]MNPA (Tokunaga et al., J Neurosci 2009). This could be in line with the view that NMDA receptor antagonists produce psychotomimetic effects through alterations of the dopaminergic neurotransmission. However, mechanisms by which NMDA receptor antagonists modulate D2 receptor binding remain unclear. In the present study, we have investigated the effect of ketamine (at an anesthetic dose) and PCP (at a psychotomimetic dose) on the binding of agonistic and antagonistic radioligands to striatal D2 receptors. In addition, we have pursued the neurochemical basis of this crosstalk between NMDA and D2 receptors.
Methods
Male Sprague-Dawley rats weighing 300–400 g were operated in order to fixate their heads during the PET measurement under a conscious state. PET scans of rats were performed with a microPET FOCUS220 system for 90 min after administration of agonistic ([11C]MNPA) and antagonistic ([11C]raclopride) radiotracers. Anatomical regions of interests (ROIs) were manually defined on the striatum and cerebellum in PET images coregistered with MR images using PMOD® software. Binding potential (BPND) values were obtained using the multilinear reference tissue model (MRTM), with cerebellum used as reference region. Rats were also treated with ketamine (90 mg/kg/h) and PCP (3 mg/kg/h) starting 30 min before the scan start using repeated intravenous boluses. Microdialysis studies were performed on conscious rats which were pretreated with ketamine and PCP, similar as in the PET measurements. Dialysate samples were collected from the striatum every 10 minutes, and DA contents were measured by HPLC with an ECD system. Autoradiographic analysis of coronal brain sections was performed using brains obtained from untreated and treated rats. Striatal [11C]raclopride binding was densitometrically quantified.
Results
During ketamine anesthesia, striatal BPND values for both agonistic and antagonistic radioligands were increased by approximately 20%, while ketamine-induced changes of DA contents were barely detectable by microdialysis in the rat striatum. In vitro autoradiography studies demonstrate a ketamine-induced increase in the striatal Bmax for [11C]raclopride, but no change in Kd. The PCP-induced increase ofBPND for [11C]MNPA was however significantly larger than that for [11C]raclopride.
Conclusions
Ketamine did not modify dopamine levels as measured by microdialysis. The similar increase in agonistic and antagonistic radioligand binding indicated that ketamine, at an anesthetic dose, does not alter the functional states of D2 receptors (i.e. high-affinity and low-affinity states). The autoradiography results confirm that ketamine enhances radioligand binding in vivo due to an increase in Bmax. In contrast, PCP-enhanced radioligand binding may result from a change in the functional states of D2 receptors, which could be an effect characteristic of PCP receptor ligands at a psychotomimetic dose.
P022. Anti-CD20 therapy reduces lesion volume and activated microglia in rodent models of multiple sclerosis
1F. Hoffmann-La Roche Ltd, Basel, CH, Switzerland; 2University of Oxford (Oxford, UK); 3Genentech, San Francisco, California, USA
Background
To establish a rational basis for the suppression of B-cell function by anti-CD20 therapy in multiple sclerosis (MS), it is essential to determine whether this therapy treats both pattern I and pattern II type MS-lesions and whether such therapy reduces extra-lesional activated microglia, which cause damage to axons. Pattern II lesions are associated with deposition of antibody and complement, whereas pattern I lesions are thought to be mediated principally by CD4+ myelin-reactive Th1/Th17 cells and macrophages in a B-cell independent fashion. The purposes of this study were to determine whether anti-CD20 therapy would inhibit lesion development and microglial activation in a pattern I MS-like lesion (ie, Th1 pattern I-type delayed type hypersensitivity (BCG-DTH)) and in a pattern II MS-like lesion (ie, an antibody-mediated focal fMOG-EAE) rodent model of MS.
Methods
To induce pattern I MS-like lesion, rats were injected stereotaxically into the left striatum with heat-killed BCG. Four weeks after the injection of BCG, the focal DTH lesion was initiated by an intradermal injection of BCG in complete Freund's adjuvant into both hind legs. To induce pattern II MS-like lesion, rats were injected subcutaneously at the base of the tail with MOG peptide. Twenty-one days after injection of MOG, rats were stereotaxically injected with a mixture of recombinant rat tumor necrosis factor-α and IFN-ã into the corpus callosum. Animals received either anti-CD20 therapy (ie, anti mCD20; 5 mg/kg/iv on days 1 and 7)) and were sacrificed on day 12. Tissue was either perfusion fixed with paraformaldehyde or used as fresh tissue, which was post-fixed as cryosections. To detect alterations in activated microglia, immunohistochemistry was carried out using the OX-6 antibody and in situ hybridization using [125I]DPA-713 (ie, compound selective for translocator protein (18 kDa) (TSPO) which co-localizes with activated microglia). Specific binding was confirmed by incubating adjacent sections with PK 11195.
Results
Anti-CD20 treated rats displayed significantly reduced lesion volume compared to vehicle treated rats in both the pattern I (66%; p<0.0001) and pattern II (57%; p<0.01) models. Areas of microglia activation were significantly reduced in anti-CD20 treated compared to vehicle treated rats, as shown by both immunohistochemistry and in situ hybridization methods. Anti-CD20 treated rats displayed reduced [125I]DPA-713 binding compared to vehicle treated rats in both the pattern I (40%; p<0.0001) and pattern II (25%; p<0.006) models.
Conclusions
These results demonstrate that anti-CD20 therapy blocked the development of pattern I and pattern II MS-like lesions in rodent models of MS. The suppression of B cell function by anti-CD20 therapy in pattern I MS-like lesions, which hitherto would not have been considered B-cell dependent, suggests that B cells play an important role in lesion development that is independent of local CNS antibody production. In addition, B-cell depletion by anti-CD20 therapy reduced extra-lesional microglia activation in both rodent models of MS. Further, these results suggest that an in vivo assessment of microglial activity in patients with MS may provide a quantifiable and longitudinal measure of patients grey matter microglial activation and suppression during drug therapy.
P023. An open-label PET study to evaluate serotonin transporter (SERT) occupancy following escalating doses of Desvenlafaxine (Pristiq®)
Gordon Frankle1, Brigitte Robertson5, Gary Maier4, Jennifer Paris3, Deanna Asmonga3, Chi-Min Chen3, Maureen May3, N. Scott Mason2, Chester A. Mathis 2 and
1Departments of Psychiatry and Radiology; 2Department of Radiology; 3Department of Psychiatry, University of Pittsburgh, Pittsburgh, Pennsylvania, USA; 4Sunovion Pharmaceuticals Inc, Marlborough, Massachusetts, USA; 5Shire Pharmaceuticals Inc, Wayne, Pennsylvania, USA
Background
Desvenlafaxine (Pristiq®) is a venlafaxine metabolite, approved for the treatment of major depressive disorder, which acts as a serotonin (5-HT) and norepinepherine (NE) reuptake inhibitor. This study examined the SERT occupancy achieved by desvenlafaxine at daily doses of 25 mg, 50 mg, 100 mg and 150 mg using [11C]DASB positron emission tomography (PET).
Methods
Eight healthy male volunteers participated in this study (age 27±9 years). Each subject underwent three PET scans with [11C]DASB. First, a baseline, off-medication, scan was performed. Within 21 days of the baseline scan, subjects were admitted to the research unit at the University of Pittsburgh for 7 days. Each subject participated in one of two dose groups (N=4 per group) during which they were administered, in an open label, unblinded manner, two doses of desvenlafaxine. On Days 1–3 subjects received a lower dose, then on Days 4–6 the dose of the study drug was increased such that total daily doses of 25 mg, 50 mg, 100 mg and150 mg were studied.
Two [11C]DASB PET scans were perform during the inpatient stay, after 72 hours at each dose level, to coincide with trough steady state level of the study drug; on Day 4 and on Day 7. Subjects were scanned for 100 min on the EXACT HR+ in 3D mode. Arterial input function was measured and corrected for metabolites. Regions of interest (ROIs) were traced on coregistered MRI. ROIs included the midbrain, thalamus, amygdala and striatum. [11C]DASB binding potential (BPND, unitless) was derived, calculated using the simplified reference tissue method with the cerebellum as the reference region. Occupancy of the SERT was calculated as the change in BPND (BPND) from each individual's baseline scan to the on-medication scan relative to the baseline BPND value as BPND=(BPND BASE−BPND MED)/BPND BASE.
Results
Desmethylvenlafaxine significantly reduced regional distribution volumes and region BPND values in a dose-dependent manner. The relationship between the average occupancy, total daily dose and plasma levels of desvenlafaxine are given in Figures 1 and 2, respectively. The total daily dose required to provide 50% SERT occupancy was 14.4 mg. Deriving the Ki values using the free (unbound) plasma concentration of desvenlafaxine resulted in a Ki value of 17.8 nM.
Conclusions
Previous studies indicate ∼80% SERT occupancy by antidepressants at minimum therapeutic doses (Meyer et al, AJP 2004); at the recommended therapeutic dose of 50 mg, desvenlafaxine demonstrates a SERT occupancy of 80.5%±10.2%. At higher doses SERT occupancy consistently exceeds 80%, raising the possibility that dose increases may benefit individuals who do not respond to the initial 50 mg starting dose of desvenlafaxine.
P024. Dose dependent occupancy of central dopamine D2 receptors by the dopamine stabilizer (−)-OSU6162
1VU University Medical Center, Department of Nuclear Medicine & PET Research, Amsterdam, The Netherlands; 2VU University Medical Center, Department of Neurology, Amsterdam, The Netherlands, 3University of Gothenburg, Sweden, and VU University Medical Center, Amsterdam, The Netherlands
Background
Dopamine stabilizers inhibit dopaminergic effects in case of elevated dopaminergic activity, whilst enhancing dopaminergic signaling under conditions of low dopaminergic tone. Consequently, it has been hypothesized that dopamine stabilizers may reduce dyskinesias in Parkinson's disease and at the same time prevent an increase in symptoms. The aim of this study was to determine the optimal dose of the dopamine stabilizer (−)-OSU6162 required to achieve significant D2 receptor occupancy.
Methods
Ten male healthy volunteers (mean age 55.4 years. range 51–59 years) underwent dynamic 60 minute [11C]raclopride scans on an ECAT EXACT HR+ scanner (Siemens/CTI) before and one hour after an oral dose of (−)-OSU6162. Both PET scans were performed on the same day. The dose of (−)-OSU6162 varied between subjects, ranging from 15 to 90 mg. Parametric images of non-displaceable binding potential (BPND) were generated using a basis function implementation of the simplified reference tissue model (RPM). Cerebellum grey matter was used as reference tissue. Volumes of interest (VOI) were defined using an automated MRI based template (PVE lab) and then projected onto the parametric [11C]raclopride BPND images. Striatal BPND was used as outcome measure. D2 receptor occupancy (%) was calculated as 100 × (1−BPOSU/BPbaseline), where BPbaseline and BPOSU represent BPND values before and after (−)-OSU6162 administration, respectively.
Results
As shown in the Table, a dose dependent occupancy of (−)-OSU6162 except for one of the subjects was found with a maximum occupancy of around 30%.
Conclusions
In vivo, only a relatively small proportion of human striatal dopamine D2 receptors is available for binding with (−)-OSU6162. Although the mechanism is unknown, data could be explained against the background of the different forms and locations of the D2 receptor. More pharmacological dissection is needed to better understand this intriguing phenomenon. The maximal occupancy by (−)-OSU6162 of about 30% of dopamine D2 receptors may have important pharmacotherapeutic implications.
P025. The novel dopamine D2/D3 receptor agonist radioligand [18F]MCL-524 is sensitive to dopamine levels
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2McLean Hospital, Harvard Medical School, Belmont, Massachusetts, USA; 3University of Toronto, Toronto, Ontario, Canada
Background
Dopamine D2 and D3 (D2/D3) receptor agonist PET radioligands have so far predominantly been radiolabeled with the relatively short-lived radioisotope carbon-11. An agonist radioligand labeled with a longer lived radioisotope, e.g. fluorine-18, is of interest to allow for prolonged investigation of the interaction between radioligand binding, endogenous dopamine and receptor trafficking. We recently reported the development of the novel D2/D3 receptors agonist radioligand [18F]MCL-524 [1]. In this study we evaluated the potential for quantitative analysis of [18F]MCL-524 binding to D2/D3 receptors in the non-human primate brain.
Methods
On two experimental days four PET measurements were performed in two female cynomolgus monkeys. On each day two baseline PET measurements were performed after intravenous injection of respectively [11C]MNPA [2] and [18F]MCL-524. The PET measurements were performed for two hours using the HRRT. Arterial blood was obtained for measurement of the metabolite-corrected arterial input function after injection of [18F]MCL-524. Distribution volumes (VT) were obtained for [18F]MCL-524 by applying compartment modeling (1TCM and 2TCM). The here-reported binding potential values (BPND) were obtained by the 2TCM and the simplified reference tissue model (SRTM), using the cerebellar cortex as reference region. On two additional experimental days we evaluated [18F]MCL-524 binding for dopamine D2/D3 receptor specificity and dopamine level sensitivity. [18F]MCL-524 was therefore i.v. injected to the same two animals after an i.v. administration of raclopride (1.0 mg/kg, at −20 min) or D-amphetamine (1.0 mg/kg, at −25 min).
Results
Administration of [18F]MCL-524 resulted in a radioactivity accumulation which was high in the striatum but lower in the cerebellum. Striatal BPND values for [18F]MCL-524 obtained with the 2TCM were 2.36/2.22. BPND values obtained with SRTM were higher after injection of [18F]MCL-524 than for [11C]MNPA, respectively 2.22/2.10 and 1.19/1.56. Administration of D-amphetamine and raclopride reduced [18F]MCL-524 binding with 43% (BPND, SRTM and 2TCM) and 87% (BPND, SRTM), respectively.
Conclusions
The receptor binding of [18F]MCL-524 can be quantified using kinetic modeling and reference region approaches. Pretreatment challenges demonstrated that [18F]MCL-524 binding is D2/D3 receptor specific and sensitive to amphetamine-induced changes in dopamine level. We conclude that [18F]MCL-524 has unique characteristics which makes it suitable for studies evaluating the interaction of radioligand binding, endogenous dopamine and receptor trafficking.
P026. Test-retest variability of [11C]raclopride in rat using PET/MRI
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2University Medical Center Groningen, Department of Nuclear Medicine and Molecular Imaging, The Netherlands; 3CROmed Ltd, Budapest, Hungary; 4AstraZeneca R&D, Södertalje, Sweden
Background
Small animal PET imaging is a rapidly growing research field within neuroscience. One great advantage is that e.g. long-term treatment effects or the progression of disease can be investigated in the same animal longitudinally. However, in order to evaluate these changes over time, the within-subject test-retest variability is an important factor to be considered. The purpose of this study was therefore to perform a test-retest study in rat using the dopamine D2/D3 receptor antagonist radioligand [11C]raclopride.
Methods
A 63-minute PET scan was initiated immediately upon intravenous injection of 11C-raclopride and after completion of the PET, a MRI was performed using a nanoScan® PET/MRI scanner (Mediso Ltd, Budapest Hungary) [1]. One week later, the same procedure was conducted in the same animals. The PET scans were co-registered with the MRI using PMOD 3.2 (Zurich, Switzerland). The individual MRI images were normalised to a template MRI (T2) in PMOD and a region of interest (ROI) template for rats was used to delineate areas with high binding of 11C-raclopride (i.e. caudate, putamen, nucleus accumbens core and septum). Binding potential (BPND) of 11C-raclopride was calculated with the simplified reference tissue model (SRTM) using cerebellum as a reference region. A test-retest variability (TRV) was calculated for the three ROIs and the intersubject variability (COV) was calculated by dividing the standard deviation (SD) by the mean. In addition, the intraclass correlation coefficient (ICC) was calculated using SPSS, which compares the relative variation within subjects to between subjects.
Results
Four male Wistar rats (384.6±28.8 g) were intravenously injected with 11C-raclopride twice, one week apart (injected radioactivity dose: 21.9±3.5MBq, injected mass: 0.005–0.09 ìg, n=8). The BPND, TRV of BPND (%), COV and ICC were calculated for each of the three delineated regions: Caudate/Putamen (BPND: 1.77±0.24, TRV: 11.1±5.2%, COV: 13.8%, ICC: 0.65), Accumbens (BPND: 1.17±0.24, TRV: 12.3±11.8%, COV: 20.5%, ICC: 0.77) and Septum (BPND: 0.82±0.09, TRV: 6.39±3.94%, COV: 11.1%, ICC: 0.83). The test-retest within-subject variability for BPND varied between 6.39–12.3% in regions with high specific binding of 11C-raclopride as compared to the intra-subject variability between 11.1–20.5% in the same regions. The average intraclass correlation coefficient was 0.75, showing a greater variability between subjects as compared to differences within-subject.
Conclusions
The results from the present study are in line with previous published studies for test-retest of [11C]-raclopride in mice [2] and shows that 11C-raclopride could be used to investigate changes in the dopamine system in the same animal over time.
P027. Further validation of the serotonin sensitivity of [11C]AZ10419369 receptor binding in non-human primates
Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden
Background
We recently demonstrated that the 5-HT1B receptor antagonist PET radioligand [11C]AZ10419369 is sensitive to (±)-fenfluramine-induced changes in serotonin concentration [1]. In a second study we have reported the development of an improved methodology, using a bolus infusion approach, to allow for estimation of changes in serotonin concentration during equilibrium conditions [2]. Using the improved methodology, we demonstrated that the efficient serotonin releaser (±)-fenfluramine (5.0 mg/kg) decreased BPND values of e.g. the occipital cortex with an average of 39% in three monkeys [2]. In the current study, we evaluated the changes in serotonin concentrations after a single dose of the stimulant D-amphetamine or the selective serotonin reuptake inhibitor (SSRI) escitalopram using [11C]AZ10419369 in non-human primates.
Methods
A total of 28 PET measurements were performed using a bolus infusion protocol of [11C]AZ10419369 in three cynomolgus monkeys. On each of fourteen experimental days a baseline PET measurement was followed by a pretreatment PET measurement in which D-amphetamine (1.0 mg/kg, −25 min over 10 min, n=7) or escitalopram (2.0 mg/kg, −45 min over 30 min, n=7) was intravenously administered before the bolus radioligand injection. Emission data were acquired for 125 min using the HRRT PET system. The binding potential (BPND) was calculated for the time frame of 63–123 min with the cerebellum as reference region.
Results
Administration of D-amphetamine and escitalopram caused a significant reduction in regional BPND values (RM ANOVA, P < 0.0001; paired t-test p<0.05). BPND values in the occipital cortex decreased significantly by 32±10% and 12±8% after administration of D-amphetamine and escitalopram, respectively.
Conclusions
We conclude that 5-HT1B receptor binding of [11C]AZ10419369 can be altered by serotonin concentration modifying drugs other than (±)-fenfluramine. This study thereby confirms that [11C]AZ10419369 can be used to assess serotonin concentrations in the living brain. The effect of the SSRI escitalopram was significantly smaller than for the stimulants (±)-fenfluramine and D-amphetamine, which should be considered when translating the methodology to human subjects. The selection of a stimulant may be a more suitable serotonin challenge when compared to the use of a SSRI.
Acknowledgements
The research leading to these results has received support from the Innovative Medicine Initiative Joint Undertaking under grant agreement n° 115008 of which resources are composed of EFPIA in-kind contribution and financial contribution from the European Union's Seventh Framework Programme (FP7/2007–2013).
P028. Visualizing the central 5-HT1B receptor with PET in rats: effects of cyclosporin A on [11C]AZ10419369 uptake
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2University Medical Center Groningen, Department of Nuclear Medicine and Molecular Imaging, Groningen, The Netherlands; 3CROmed Ltd., Budapest, Hungary; 4AstraZeneca R&D, Södertalje, Sweden
Background
The radioligand [11C]AZ10419369 [1] has been demonstrated to be a good marker of drug-induced occupancy at central 5-HT1B receptors in the primate brain [2]. It has similar affinites for human and monkey receptors, but a preliminary rat experiment showed that [11C]AZ10419369 did not penetrate the Blood Brain Barrier (BBB). We have therefore explored the possibility that [11C]AZ10419369 is a substrate for P-glycoprotein (Pgp) by pre-treating the animals with the Pgp blocker cyclosporin A.
Methods
Cyclosporin A was used to investigate the possibility of imaging 5-HT1B receptors with [11C]AZ10419369 ligand in rats. Pre-treatment of male Wistar rats with Cyclosporin A (50 mg/kg) was made 15 minutes before the start of the 93 min PET scan using a nanoScan® PET/MRI scanner (Mediso Ltd, Budapest Hungary) [3]. Using PMOD 3.3 Fusion and Kinetic modules (PMOD Zurich, Switzerland) PET scans were co-registered to its own MRI, and the individual MRI's were normalised to PMOD's rat brain MRI template which was used to acquire time activity curves (TAC) in volumes of interest. BPND values for different brain regions were estimated with the simplified reference tissue model (SRTM) using the cerebellum as reference region.
Results
Without pretreatment of cyclosporin A, very low binding of [11C]AZ10419369 was found in all brain areas (Whole brain SUV=0.45; SD=3.6; n=2). With cyclosporin A pre-treatment, binding of [11C]AZ10419369 was higher in the Whole brain (SUV=1.39; SD=0.29; n=2). The highest was in the Thalamus (SUV=1.84; SD=0.45; BPND=0.68; SD=0.16; n=2), Periaqueductal Gray (SUV=1.79; SD=0.38; BPND=0.56; SD=0.06; n=2), Midbrain (SUV=1.78; SD=0.44; BPND=0.58; SD=0.10; n=2), Superior Colliculus (SUV=1.76; SD=0.38; BPND=0.51; SD=0.08; n=2), Septum (SUV=1.71; SD=0.51; BPND=0.64; SD=0.23; n=2), Caudate/Putamen (SUV=1.71; SD=0.43; BPND=0.69; SD=0.14; n=2), Hypothalamus (SUV=1.67; SD=0.40; BPND=0.60; SD=0.10; n=2), Nucleus Accumbens Core (SUV=1.67; SD=0.44; BPND=0.67; SD=0.18; n=2), Ventral Tegmental Area (SUV=1.66; SD=0.46; BPND=0.54; SD=0.11; n=2) and lowest in Cerebellum (SUV=1.19; SD=0.20; n=2).
Conclusions
[11C]AZ10419369 does not penetrate the rat BBB under normal circumstances, although previous monkey and human studies have shown brain uptake in vivo and the radioligand has been used to succesfully image central 5-HT1B receptors and to measure drug-induced occupancy with PET. With the use of a P-glycoprotein blocker, cyclosporin A, it is possible to increase the brain uptake of [11C]AZ10419369 and evaluate its binding to the 5-HT1B receptor in the rodent brain.
P029. Characterization of [11C]CIMBI-36 as a 5-HT2A receptor agonist PET radioligand in the non-human primate brain
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2Cimbi and Neurobiology Research Unit, Rigshospitalet and University of Copenhagen, Denmark
Background
Recently, the first 5-HT2A receptor agonist PET radioligands have been developed, with [11C]CIMBI-36 reported as the most promising in pig brain so far [1]. In this study we evaluated different approaches for quantification of [11C]CIMBI-36 binding to central 5-HT2A receptors in non-human primates. Additional experiments were performed to evaluate the sensitivity of [11C]CIMBI-36 binding to (±)-fenfluramine-induced changes in 5-HT concentration.
Methods
For evaluation of the quantification of [11C]CIMBI-36 binding, a total of 6 PET measurements were performed in 3 female rhesus monkeys. On 3 experimental days, a baseline measurement was followed by a pretreatment measurement with the 5-HT2A receptor selective antagonist ketanserin (0.75–1.5 mg/kg). After i.v. bolus injection of [11C]CIMBI-36, PET measurements were performed for 2 hours in the HRRT PET system. Arterial blood was obtained for measurement of the metabolite-corrected arterial input function. In the second part of the study 4 PET measurements were performed in 2 rhesus monkeys. On 2 days a baseline measurement was followed by a pretreatment study with the 5-HT releaser (±)-fenfluramine (5.0 mg/kg). In these experiments no arterial blood samples were obtained. Quantification of [11C]CIMBI-36 binding was performed with kinetic modeling using tissue-compartment models, Logan graphical analysis and different reference tissue models. The outcome measures were the total distribution volume (VT), nondisplaceable distribution volume (VND) and non-displaceable binding potential (BPND) values using the cerebellar cortex as reference region.
Results
Baseline PET measurements demonstrated good brain uptake and a regional distribution in accordance with the known 5-HT2A receptor distribution. The 2-tissue compartment model was the preferred model for the description of the regional time activity curves. BPND values estimated with reference tissue models correlated with the corresponding values obtained with kinetic modeling, although the reference methods slightly underestimated BPND values. Administration of ketanserin decreased the binding in 5-HT2A receptor rich brain region, e.g. neocortex (BPND, 77±7%), but did not decrease VT of cerebellum (VND). Fenfluramine decreased BPND values (SRTM) by ∼60% in the neocortex of two monkeys.
Conclusions
This work indicates that [11C]CIMBI-36 receptor binding is specific to 5-HT2A receptors. [11C]CIMBI-36 receptor binding can be quantified using kinetic modelling, and reference tissue models can be applied with the cerebellar cortex as reference region. Fenfluramine-induced changes in 5-HT concentration decreased the binding of [11C]CIMBI-36 suggesting that [11C]CIMBI-36 receptor binding is sensitive to increased endogenous 5-HT concentration.
P030. Tiagabine does not modify [11C]Ro 15-4513 and [11C]flumazenil binding to a major extent in anesthetized non-human primates
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2Pharmaceutical Division, Hoffman-La Roche, Basel, Switzerland; 3Pfizer Neuroscience Research Unit, Groton, Connecticut, USA
Background
A non-invasive method suitable for evaluation of GABA levels in the living brain is of interest to enhance understanding of the etiology and treatment of neuropsychiatric disorders. Frankle et al. (2009), previously demonstrated that the GABA transporter I (GAT I) inhibitor tiagabine enhances the binding of the benzodiazepine antagonist [11C]flumazenil in human subjects [1]. This enhanced binding has been proposed to be related to tiagabine-enhanced GABA levels increasing the affinity of GABAA receptors for benzodiazepine ligands, the so called ‘GABA shift’. [11C]Ro 15-4513 [2] is a partial inverse agonist at the GABAA receptor, with relative selectivity for the α5-subunit (GABRA5). In the current study we evaluated the effect of tiagabine on [11C]Ro 15-4513 and [11C]flumazenil binding in anesthetized non-human primates.
Methods
Four female rhesus monkeys were examined during intravenous ketamine/xylazine anesthesia on six experimental days. On each day a baseline positron emission tomography (PET) measurement was followed by a pretreatment PET measurement during which tiagabine (1.0 mg/kg) was intravenously administered. All four monkeys were examined with [11C]Ro 15-4513 and to date two monkeys with [11C]flumazenil. After an intravenous bolus injection of the radioligand, PET measurements were performed for 95 minutes in the HRRT PET system. Arterial blood was obtained for measurement of the metabolite-corrected arterial input function. Quantification of radioligand binding was performed by applying the simplified reference tissue model (SRTM) with an outcome measure of the non-displaceable binding potential (BPND) values. The pons was used a reference region, similarly to previous reports for [11C]Ro 15-4513 [3] and [11C]flumazenil [4].
Results
Baseline PET measurements demonstrated good brain uptake and a regional distribution in accordance with the literature. Administration of tiagabine did not modify the radioligand binding to a significant extent. Tiagabine caused, however, a small decrease, from 2.4 to 2.2 (−9%), in the mean BPND values (SRTM) of six different brain regions (ventral striatum, occipital cortex, dorsal lateral prefrontal cortex, hippocampus, anterior cingulate cortex and insular cortex) obtained after injection of [11C]Ro 15-4513 (n=4). In comparison, tiagabine caused a minor increase in the mean BPND values (SRTM) of six different brain regions, from 2.7 to 2.8 (+5%), obtained after injection of [11C]flumazenil (n=2). The preliminary results of the kinetic modeling support the use of a reference region approach.
Conclusions
This study indicates that tiagabine does not modify or has only minor effect on [11C]Ro 15-4513 and [11C]flumazenil binding in anesthetized non-human primates. Further work is ongoing and will include validation of the here-applied reference region approach by performing compartmental modeling. The development of a radioligand with higher sensitivity to modified GABA levels is desirable.
Acknowledgements
The research leading to these results has received support from the Innovative Medicine Initiative Joint Undertaking under grant agreement n° 115008 of which resources are composed of EFPIA in-kind contribution and financial contribution from the European Union's Seventh Framework Programme (FP7/2007–2013).
P031. Sensitivity of the benzodiazepine inverse agonist PET ligand [11C]Ro15-4513 to changes in endogenous GABA
1University of Bristol, UK; 2Imperial College London, UK
Background
Measuring endogenous GABA in the brain is possible using MRS, though the main contributor to the signal is metabolic and lacks sensitivity in characterising changes in the synaptic pool. We have investigated the potential of the inverse agonist [11C]Ro15-4513 benzodiazepine PET ligand to measure changes in synaptic GABA, using tiagabine to block the uptake transporter GAT1 and increase GABA as reported for the non-selective benzodiazepine ligand [11C]flumazenil (Frankle et al., 2009). We have shown in a preclinical study that tiagabine significantly reduced [3H]Ro15-4513 uptake in the hippocampus (Tyacke et al., 2009). Since the actions of GAT1 are mostly synaptic (Conti et al., 2004), we hypothesised a decrease in the synaptic α1 signal with elevations in endogenous GABA, consistent with GABA.
Methods
12 healthy male volunteers (age mean±S.D.: 50±8 years) were recruited for a double-blind, randomised, placebo-controlled PET study. Subjects underwent two 90-minute PET scans at least one week apart, with a 15 mg tiagabine challenge or identical placebo. Dynamic images of 24 frames were acquired, with discrete blood samples collected throughout for radioactive counts and metabolite analysis, as well as 15 minutes of continuous arterial blood sampling from the start of the scan. These data were used to calculate individual plasma input functions for estimation of ligand binding parameters. Subjects also underwent a T1-weighted MRI scan for structural reference to an atlas of 84 regions. We applied previously described methods of estimating specific binding of [11C]Ro15-4513, using compartmental modelling and spectral analysis to isolate both the GABARα1 and GABARα5 subtypes (Myers et al., 2012). This division is possible due to the different kinetic properties of [11C]Ro15-4513 at each site.
Results
Tiagabine resulted in a significant reduction in α1 distribution volume globally (RMANOVA F=34.91, p<0.0001). Post-hoc paired t-tests found significant decreases in α1 binding in the hippocampus (t=2.358, p<0.05), amygdala (t=2.488, p<0.05) and anterior cingulate gyrus (t=2.285, p<0.05). Global increases in the α5 signal were also seen (F=42.46, p<0.0001).
Conclusions
This suggests that tiagabine-induced increases in GABA result in displacement of [11C]Ro15-4513 binding from the synaptic GABARα1 to the high-affinity extra-synaptic GABARα5. These conclusions are supported by evidence that the shift is more apparent in regions with high concentrations of GABARα5 and GAT1, such as frontal cortical and limbic areas. This is supported by further analysis using a 2-tissue compartmental model, to demonstrate a strong relationship between the changes in k2, due to the reduction in binding at GABARα1, and changes in k3/k4, a measure of specific binding at GABARα5. In this study we have demonstrated that [11C]Ro15-4513 PET can detect increased synaptic GABA through reduced binding at the synaptic GABARα1 and increased binding at the GABARα5 subtype. Regional changes in endogenous GABA in neurological and psychiatric disorders can be investigated using this technique.
P032. PET imaging of brain inflammation during epileptogenesis in the rat
Marie-Claude Grégoire1, Andrew Katsifis1, Christian Loc'h1, Tien Pham1, Paul Callaghan1 and
1LifeSciences, ANSTO, Menai, NSW, Australia; 2Translational Neuroscience and Molecular Imaging Centre, University of Antwerp, Belgium
Background
Evidence from animal and clinical studies has been accumulating which supports the hypothesis that inflammatory processes within the brain might constitute a common and crucial mechanism in the pathophysiology of seizures and epilepsy. Recently, inflammatory processes in the brain have been suggested as a target for epilepsy therapy. PET imaging offers the unique possibility to evaluate brain inflammation longitudinally in a non-invasive translational manner. [18F]-PBR111, with high specificity for the translocator (TSPO) or peripheral benzodiazepine (PBR) receptor, expressed on activated microglia, has suitable in vivo characteristics to be used as a brain inflammation imaging biomarker. This study aimed to investigate brain inflammation early on during epileptogenesis in the post Kainic Acid-induced Status Epilepticus (KASE) model with postmortem histology/autoradiography and in vivo [18F]-PBR111 PET.
Methods
SE was induced (N=13) by repeated low-dose injections of KA, while controls (N=9) received saline. The extent of TSPO expression and microglia activation was assessed with in vitro [125I]-CLINDE as well as in vitro and ex vivo [18F]-PBR111 autoradiography on the one hand and OX-42 immunohistochemistry on the other hand, 7 days post SE. In a subgroup of rats, [18F]-PBR111 PET imaging with metabolite corrected input function was performed before postmortem evaluation. Volume of distribution of several regions of interest were calculated and compared with the partition coefficient at a late time point (50 min).
Results
Animals with severe SE showed a huge and significant overexpression of TSPO in vitro in relevant brain regions such as hippocampus and amygdala (P<0.001), while animals with mild symptoms displayed a smaller increase in TSPO in amygdala only (P<0.001). TSPO expression was associated with OX-42 signal but without obvious cell loss. Similar increases in volume of distribution in key regions such as hippocampus (P<0.05) and amygdala (P<0.01) were observed with in vivo PET imaging compared to in vitro and ex vivo measurements (Figure 1). Good correlation was obtained between kinetic modelled volume of distribution values and the partition coefficient at a single time point.
Evaluation of brain inflammation by use of in vivo [18F]-PBR111 PET imaging (coronal plane) with corresponding postmortem autoradiography and immunohistochemistry in a representative control and KASE rat, seven days post SE induction. Arrows indicate the different brain regions named on the ipsilateral hemisphere. Abbreviations: THAL, thalamus; AMYG, amygdala and PIR, piriform cortex.
Conclusions
Both postmortem and in vivo methods substantiate that brain regions important in seizure generation display significant microglial activation early during epileptogenesis in the KASE model. Moreover, this work enables future longitudinal investigation of the role of brain inflammation during epileptogenesis and the chronic epileptic state. Finally, it will also allow for in vivo follow-up of anti-inflammatory treatment strategies.
P033. Withdrawn
P034. In vivo quantification of striatal dopamine D2 receptor expression and [11C]raclopride affinity using a single injection protocol
1Department of Preclinical Imaging and Radiopharmacy, Eberhard Karls University, Tuebingen, Germany; 2Australian Nuclear Science and Technology Organisation, ANSTO LifeSciences, Sydney, Australia; 3Department of Medical Genetics, Eberhard Karls University Tuebingen, Germany
Background
High resolution small animal PET has emerged as a valuable tool to study receptor expression (appB max ) and [11C]raclopride affinity (1/appK d ) in rats and mice using multiple injection protocols. Since multiple injection experiments are expensive, time consuming and require at least three scans per animal, we aimed to identify an analysis approach that meets the requirements of technical simplicity and greatest possible accuracy for routine laboratory protocols. Here, we investigated the partial saturation approach (PSA), a single injection method, over a range of occupancy levels to validate the stability of the method for [11C]raclopride PET imaging in mice. In addition, we tested the feasibility of the PSA after pharmacologically induced changes of synaptic DA concentrations and in two transgenic mouse models with altered D2 receptor expression.
Methods
Eight mice underwent 4 [11C]raclopride PET scans with increasing raclopride doses, ranging from 2.6 to 550 nmol/kg using an injected activity of 427±42 MBq/kg. Individual time activity curves were generated from the striatum and the cerebellum, which were used to estimate the free (F) and bound (B) tracer concentration. The general equilibrium equation: B/F=(Bmax−B)/KdVr–dB/dT·(1/koff·F) includes a residual term which was used as an indicator of the dynamic equilibrium state to choose a constraining time window and occupancy level for Scatchard analysis. Eight other animals participated in test-retest studies on two separate days with the mass dose kept constant at 44±3 nmol/kg. To identify the impact of increased DA availability on appB max and appK d values, 12 mice underwent two scans, at baseline and under challenge conditions: (i) 30 min (n=4) and 4 h (n=4) after 3 mg/kg d-amphetamine (AMPH) i.p. and (ii) 10 min after 3 mg/kg methylphenidate (MP) i.v. (n=4). To calculate changes of D2 receptor expression using the PSA two mouse models of human disease were used: (i) the R6/2 mouse model of Huntington's disease: six transgenic mice and three healthy controls were scanned at 5 and 7 weeks of age and the injected dose was adjusted to obtain about 50% receptor occupancy, (ii) the 6-OHDA mouse model of Parkinson's disease: six mice were lesioned with 3 μg 6-OHDA in the right MFB and PET scans were performed after three weeks of recovery.
Results
Except for the early data points, the bound ligand concentration showed a linear decrease over the duration of the PET scan over a wide range of occupancy levels (30–60%). Most stable results were obtained at an injected dose of 44 nmol/kg, which corresponds to about 50–60% receptor occupancy. Therefore, the injected dose had to be adjusted in mice with low receptor expression. Test-retest experiments revealed a reproducibility of 78% and 81% for appB max and appK d , respectively (2 mice were not analyzable). The results from challenge scans and mouse models are provided in Figure 1.
Conclusions
Our results demonstrate that the PSA is a very attractive method for [11C]raclopride PET imaging in mice to assess changes in appB max and appK d after pharmacological interventions and in neurodegenerative or psychiatric disorders using a single bolus injection protocol.
P035. The impact of dopamine on behaviour during an aggression provocation paradigm: an [18F]FDOPA PET study
RWTH Aachen University, Germany
Background
Although aggressive behaviour is both a prominent and challenging syndrome in mental disorders (e.g. antisocial personality disorder, schizophrenia, ADHD etc.) and a frequent behavioural pattern in healthy populations there is a lack of knowledge and treatment options in this regard. Aggressive behaviour can be clustered into an instrumental and a more reactive (impulsive) prototypic mechanisms. Whereas impulsivity could be well related to serotonergic deficiency, in the long run also dopaminergic mechanisms were postulated in the modulation of aggression. However, so far no studies have been conducted that used in-vivo molecular imaging techniques in combination with behavioural aggression tasks in humans. The present investigation combined an [18F]FDOPA-PET with a well validated paradigm which standardized provoced intersubject-aggression.
Methods
A group of healthy males was included (n=22). One proband saw through the principle of the behavioural task and consequently had to be excluded. The final sample (N=21) had an age of 25.5±5.5 years (mean±SD). All subjects underwent a single [18F]FDOPA-PET scan including arterial plasma sampling and metabolite detection (acquisition time: 124 min.). All subjects were without any pharmacologic challenge except carbidopa pre-treatment. Directly before the scan, the subjects underwent the Point Subtraction Aggression Paradigm (PSAP). In short, this is an evaluated monetary reward game against a putative adversary which habitually tries to cheat. The proband can react by punishment (money subtraction), by pressing a defensive button, or by continuing his money-making behaviour (resilience). The PET-dynamics were analyzed according to the “inlet/outlet-model” of Kumakura et al. (2005) to obtain the net uptake of [18F]FDOPA (K), the total distribution volume (VD) and kloss in the striatum and midbrain.
Results
The subjects showed mean K values of 0.018±0.004, VD-values of 6.2±1.8, and kloss-values of 0.0034±0.0011 in the NC (midbrain: K=0.008±0.003, VD: 1.6±0.6, and kloss-values of 0.006±0.006). Significant correlations between dopamine synthesis capacity (K-values) and aggressive responses on the PSAP were found in several regions (most prominent in the midbrain: r=−0.640; p=0.002). The sum of aggressive and defensive actions correlated also negatively with the K-values. Importantly, we found additionally a significant positive correlation between the investment into monetary, incentive responses on the PSAP and dopamine synthesis capacity (K-values) especially in the midbrain (r=+0.618, p=0.003). Personality factors (measured by TCI) did not substantially contribute to the relationships.
Conclusions
Apparently, lower presynaptic dopamine synthesis capacities distract healthy subjects from reward-oriented behaviour during aggressive provocation and let them shift to interactional behaviour. This action may be an aggressive attack which in this paradigm does not provide any direct monetary reward or a defensive strategy. Subjects with higher dopamine synthesis capacity are less likely to react on the counterpart's aggression and proceed with their own more profitable bar-pressing (resilience) action. This investigation suggests that the influence of dopamine on aggression is not directly linked with a reward-expectancy to harmful-behaviour but is inversely linked to the distractibility from goal/reward-directed behaviour.
P036. Preliminary evaluation of the quantification and metabolism in human subjects of the dopamine transporter radioligand [18F]FE-PE2I
1Karolinska Institutet, Department of Clinical Neuroscience, Stockholm, Sweden; 2AstraZeneca R&D, Södertälje, Sweden
Background
[18F]FE-PE2I is a high-affinity (Ki=12 nM) radioligand for the dopamine transporter (DAT) 1 already validated in non-human primates and human subjects.2,3 DAT quantification with [18F]FE-PE2I can be performed with only 60 min of imaging data, due to its fast kinetic properties.3,4 [18F]FE-PE2I also shows favorable metabolism as compared with [11C]PE2I. 5 In rats, the hydroxyl-methyl (CH2OH-) radiometabolite of [11C]PE2I has been found to enter the brain and likely bind to the DAT. 6 In non-human primates, at HPLC analysis, [18F]FE-PE2I displays lower abundance, compared with [11C]PE2I, of a lipophilic radiometabolite peak. 2 Thus, the quantification of [18F]FE-PE2I is likely less affected by the production of radiometabolites. 2 We have recently optimized the HPLC conditions and identified by LCMS/MS the two metabolites corresponding to the CH2OH- and the carboxy- (COOH-) analogue of FE-PE2I. The aims of this study were: 1) to investigate the quantification of the binding of [18F]FE-PE2I to the DAT in the human brain using the high-resolution research tomograph (HRRT), and 2) to confirm that the metabolism of [18F]FE-PE2I is associated with negligible production of the CH2OH-radiometabolite.
Methods
Two healthy controls (S01/S02, female, age: 57/64 y, weight: 74/59 kg) were studied with [18F]FE-PE2I (221/185 MBq). PET measurements (93-min duration) were conducted using the HRRT system. Venous blood samples were drawn 5 min before and 1, 3, 5, 10, 20, 30, 45, 60, 75, and 90 min after tracer injection, for measurement of free fraction (fP) and metabolite analysis. Regions-of-interest were drawn on caudate, putamen, ventral striatum, midbrain, and cerebellum. Quantification was performed with the simplified reference tissue model, using the cerebellum as reference region. The outcome measure was the binding potential (BPND).
Results
[18F]FE-PE2I showed high uptake and fast wash-out from the brain (Figure A, B). BPND values in S01-S02 were the following: caudate: 3.7–3.1; putamen: 4.4–3.7; ventral striatum: 4.1–3.1; midbrain: 0.8–0.7. The fP was 3.3% (S01) and 5.3% (S02). The HPLC analysis (Figure C, D) showed that <10% of the plasma radioactivity at 60 min was due to the parent. The major radiometabolite peak (60% at 60 min) was polar with retention time (Rt) of 0.9 min. The CH2OH-radiometabolite (Rt=4.4 min) represented <1.5% of total plasma radioactivity, whereas the COOH-radiometabolite (Rt=3.7 min) was ∼8-to-10% at 60 min.
Time-activity curves (A,B) and metabolism (C,D) of [18F]FE-PE2I.
Conclusions
Preliminary human data using the HRRT system confirm that [18F]FE-PE2I is a suitable radioligand for DAT quantification. The fast kinetic properties and the BPND ∼4 in the striatum and ∼0.8 in the midbrain in aged controls suggest that [18F]FE-PE2I is suitable for DAT quantification in Parkinsońs disease. The negligible production of the CH2OH-radiometabolite seems to confirm that the quantification of [18F]FE-PE2I is not affected by lipophilic radiometabolites and can be performed with reference tissue models.
P037. Periventricular Flumazenil binding for predicting postoperative outcome in individual patients with temporal lobe epilepsy and hippocampal sclerosis
Josiane Yankam Njiwa1, Nicolas Costes2, Sandrine Bouvard2,3, Jérôme Redoute2, Hélène Catenoix4, Philippe Ryvlin3,5 and
1Neurodis Fundation, Lyon, France; 2CERMEP-Imagerie du vivant, Lyon, France; 3Université Lyon 1, INSERM, CNRS, Centre de Recherche en Neuroscience de Lyon, France; 4Service de Neurologie Fonctionnelle et d’Epileptologie, Hôpital Neurologique Pierre Wertheimer, Hospices Civils de Lyon, France; 5Hospices Civils de Lyon, France
Background
[11C]Flumazenil (FMZ) PET images the benzodiazepine site of GABAA receptors, present on most neurons. Cortical decreases of FMZ binding are associated with epileptogenic foci. Previous studies have shown that white matter (WM) FMZ binding detected by voxel-by-voxel-analysis (SPM) 1) correlates strongly with the number of WM heterotopic neurons determined histologically, 2) was present in the temporal lobe WM in 11/18 individual patients with temporal lobe epilepsy and normal MRI, 3) was present in periventricular areas in 7/44 patients with extratemporal epilepsy, and 4) periventricular increases detected by PET in MRI-negative patients had identical appearance to MRI-visible periventricular heterotopia. In a previous study not seizure free (NSF) outcome after temporal lobe resection for hippocampal sclerosis (HS) was associated – at the group level – with increased preoperative periventricular WM [11C]FMZ-VT (Hammers et al. Epilepsia 2005). Here, we first replicate those results in a new cohort on methodologically less demanding summed radioactivity images (SRI). We then extend the work to enable refinement of the post-operative prognosis for individual patients, based on preoperative FMZ-PET.
Methods
[11C]FMZ SRI (40–55 minutes) were available for 16 patients with HS (nine left-sided; diagnosed on MRI) and 41 healthy subjects. All patients had undergone epilepsy surgery and had at least 23 (median 68) months' follow-up. SRI were analyzed using SPM8 with explicit masking to include the WM, and correction for global radioactivity via a group-specific ANCOVA. For group analyses, the seven right HS patients' images were flipped prior to normalization to lateralize the epileptogenic side to the left for all patients. All patients were also individually compared against controls. Periventricular increases were assessed with a specific mask and different cutoffs for distinguishing NSF and seizure free (SF) patients.
Results
Nine patients were SF after surgery (Engel IA) and seven were NSF (Engel not IA)). Comparison of NSF versus SF showed areas of increased FMZ binding around the posterior horn of the ventricles ipsilaterally (z=2.53) and contralaterally (z=4.44) to the seizure focus, replicating the earlier findings. In addition, some clusters were located around the anterior horns. Compared with controls, SF had two clusters at threshold p=0.001 (0.60/0.27 cm3, zmax=3.81) in periventricular areas. In contrast, NSF had two ipsilateral clusters (1.38/1.98 cm3, zmax=4.77), and three contralateral clusters (1.79/0.74/0.26 cm3, zmax=4.72). Periventricular increases (PI) occurred in similar locations in individual patients. At the optimal threshold of p=0.01 with a cluster pmax=0.5 and zmin=3.5), five (71%) of seven individual NSF patients and one (11%) of the nine SF patients showed PI compared with controls (accuracy 81%). One (2%) of the 41 controls showed PI at this threshold.
Periventricular increases (overlaid on the SPM8-CSF template: (A) SF patients versus controls (zmax=3.97 at p<0.001). (B) NSF patients versus controls (zmax=4.77 at p<0.001). (C) Individual NSF patient (zmax=5.13 a at p<0.01).
Conclusions
The association between PI of [11C]FMZ radioactivity concentration with NSF outcome after temporal lobe resection for HS has been confirmed in an independent cohort on simple activity images. These results suggest usefulness of FMZ-PET for individual preoperative counseling with significant clinically relevant accuracy.
P038. Quantification of cholinergic degeneration in Alzheimer's disease using [18F]Fluoroethoxybenzovesamicol autoradiography
1Douglas Mental Health Institute, McGill University, Montreal, Quebec, Canada; 2Université du Québec à Montréal, Canada; 3Montreal Neurological Institute, Quebec, Canada
Background
Septohippocampal and basalocortical cholinergic degenerations have been largely described in Alzheimer disease (AD). These neurochemical degenerations are known to occur early in the course of the disease and to correlate well with its clinical symptoms (Dournaud, Delaere et al. 1995). In an effort to detect and quantify these cholinergic alterations, the PET radiotracer [18F]Fluoroethoxybenzovesamicol ([18F]FEOBV) has been developed, and shown to be fitting for the brain cholinergic terminals in rodents and primates (Kilbourn, Hockley et al. 2009). Here, we aimed to further evaluate [18F]FEOBV in differentiating AD patients from controls by using autoradiography on post-mortem brain tissues.
Methods
Tissues from the striatum, prefrontal cortex (PFC) and hippocampus were sampled from the brains of healthy controls (n=23) and AD individuals (n=13) from the Douglas-Bell Canada Brain Bank (Douglas Mental Health University Institute). All tissue slides were incubated for 60 minutes in [18F]FEOBV (1147 Ci/mmol), then washed in a cold buffer and dried. The sections were thereafter exposed for five days to phosphor imaging plates (See Figure 1 for hippocampus example). Seven regions of interest were drawn on the digitized sections: white matter (as a non-specific binding baseline), caudate, putamen, CA1, CA3, dentate gyrus, and neocortex. Average values for each region were compared cross-sectionally using a one-way Analysis of Variance. We hypothesized a decrease of binding for AD tissues in all three hippocampal regions as well as in the cortex, with no significant changes in the white matter or the striatal nuclei.
Typical hippocampus sample with [18F]FEOBV autoradiography.
Results
In AD tissues, a binding reduction of 20.13% was observed in CA1 (F=13.39, p=0.001), 25.16% in CA3 (F=7.15, p=0.014), and 24.5% in the PFC (F=22.07, p<0.0005). There was no significant change in the white matter, putamen, caudate nucleus or dentate gyrus (see Figure 2).
[18F]FEOBV binding for AD and control tissues, for all seven regions of interest. Significant differences are observed in CA1 and CA3 of the hippocampus, as well as in the prefrontal cortex.
Conclusions
The decreased [18F]FEOBV binding in CA1, CA3 and PFC is concordant with the cholinergic degeneration of the septohippocampal and basalocortical systems in AD. The lack of effect in the dentate gyrus, while surprising, may be explained by a reactive synaptogenesis of the remaining cholinergic fibers (Scheff 2003). [18F]FEOBV is a promising marker for the detection of brain cholinergic systems alterations.
P039. Interaction of inhibitory and excitatory receptors of the serotonergic system: a multi-tracer PET study
1Department of Psychiatry and Psychotherapy, Medical University of Vienna, Austria; 2Department of Nuclear Medicine, Medical University of Vienna, Austria; 3Department of Radiology and Psychiatry, New York University School of Medicine, New York, USA; 4Institute of Neuroscience and Medicine (INM-2), Research Centre Jülich, Germany
Background
The function of serotonergic receptors in the brain depends on several factors, including their functional interdependence with other homologous and heterologous receptors. Because of ethical and methodological limitations our understanding of neurotransmitter interconnections in the living human brain largely benefits from studies of nonhuman primates and other species. Enough evidence from animal and genetic studies is available to support the hypothesis of a local and systemic interplay in protein expression and internalization of the different subsystems within the serotonergic system. In this study we hypothesize that there are also positive and negative feedback mechanisms modulating the expressions of different receptor subtypes and their interplay on a systems level of the whole brain. Here, we quantified important receptor subtypes, including 5-HT1A, 5-HT1B and 5-HT2A (major inhibitory and excitatory serotonergic receptors) as well as 5-HTT (target of many antidepressants) to investigate associations within the serotonergic system.
Methods
PET scans were performed on 83 healthy subjects (age=27.7±6.8 years; 55% males; for 5-HT1A: 35 subjects; for 5-HT2A: 19 subjects; for 5-HT1B: 11 subjects; for 5-HTT: 18 subjects) using the selective radioligands [carbonyl-11C]WAY100635, [18F]Altanserin, [11C]P943 and [11C]DASB, respectively. After motion correction and spatial normalization in SPM8, dynamic PET scans were quantified from radioactivity concentrations by a multi-linear reference tissue model (MRTM2; 5HT1A, 5-HT1B, 5-HTT; BPND) and a bolus/infusion approach (5HT2A; BPP) in PMOD 3.4. A standard template in MNI stereotactic space served for ROI delineation. The ROI template follows the anatomical parcellation scheme according to Brodmann comprising 41 regions. To evaluate the association between serotonergic binding proteins, partial correlations of mean Brodmann area BP values were calculated between two sets of receptor BPs controlling for the third and 5-HTT.
Results
We found significant partial correlations between 5-HT1A and 5-HT2A controlled for 5-HT1B and 5-HTT (r=0.70, p<0.001), 5-HT1A and 5HT1B controlled for 5-HT2A and 5-HTT (r=−0.69, p<0.001), and 5-HT1B and 5-HT2A controlled for 5-HT1A and SERT (r=0.84, p<0.001), respectively. For details, see Figure.
Partial regression plots between 5-HT1A and 5-HT2A binding potential controlled for 5-HT1B and 5-HTT (left), 5-HT1A and 5-HT1B binding potential controlled for 5-HT2A and 5-HTT (middle), and 5-HT1B and 5-HT2A binding potential controlled for 5-HT1A and 5-HTT (right). All p-values<0.001. Partial regression plots adjust simple scatter plots for variables of noninterest. Notably, this adjustment implies that plotted values may yield positive as well as negative values (mean centered). Data points are Brodmann area BPs, averaged for both hemispheres.
Conclusions
Our results from in vivo data underline consistent interactions among serotonergic receptors of healthy subjects. We found strong positive correlations of both inhibitory subtypes (5 HT1A and 5 HT1B) with the excitatory 5-HT2A receptor subtype suggesting a balanced distribution of these opposing proteins not only locally but throughout the entire cortex represented by BP of 41 Brodmann areas. Interestingly, such relations were observed only when controlling for the 5-HTT level and the other receptor, while we did not find as strong results by regular correlations. This emphasizes a systemically organized interaction instead of independent receptor mechanisms. Remarkable is the negative correlation between the excitatory receptors 5 HT1A and 5 HT1B as well. They serve to regulate the serotonin release either presynaptically (5 HT1B, autoreceptors) on serotonergic neurons or postsynaptically (5 HT1A, heteroreceptors) on target neurons. It seems their predominance vary locally, however, on a globally scale they complement each other in the inhibitory performance. Our findings highlight the importance to evaluate interactions between multiple receptor subtypes to improve our understanding of entire neurotransmitter systems.
P040. Brodmann area database of the serotonergic system in healthy subjects using multi-tracer PET
1Department of Psychiatry and Psychotherapy, Medical University of Vienna, Austria; 2Department of Nuclear Medicine, Medical University of Vienna, Austria; 3Institute of Neuroscience and Medicine (INM-2), Research Centre Jülich, Germany; 4Department of Radiology and Psychiatry, New York University School of Medicine, New York, USA
Background
The highly diverse serotonergic system with 15 different receptor subtypes is implicated in the pathophysiology of most neuropsychiatric disorders including affective and anxiety disorders, obsessive compulsive disorder, post-traumatic stress disorder, eating disorders, sleep disturbance, attention deficit/hyperactivity disorder, drug addiction, suicidal behavior, schizophrenia, Alzheimer, etc. Alterations of the interplay between various pre- and postsynaptic receptor subtypes might be involved in the pathogenesis of these disorders. However, there is a lack of comprehensive standard values using standardized procedures. In the current PET study we quantify 3 receptor subtypes, including the major inhibitory (5-HT1A and 5-HT1B), the major excitatory (5-HT2A) receptors and the transporter (5-HTT) in the human brain of healthy subjects to provide a standard database.
Methods
PET scans were performed on 83 healthy subjects (age=27.7±6.8 years; 55% males; for 5-HT1A: 35 subjects; for 5-HT2A: 19 subjects; for 5-HT1B: 11 subjects; for 5-HTT: 18 subjects) using the selective radioligands [carbonyl-11C]WAY100635, [18F]Altanserin, [11C]P943 and [11C]DASB, respectively. After motion correction and spatial normalization in SPM8, dynamic PET scans were quantified from radioactivity concentrations by a multi-linear reference tissue model (MRTM2; 5HT1A, 5-HT1B, 5-HTT; BPND) and a bolus/infusion approach (5HT2A; BPP) in PMOD 3.4. A standard template in MNI stereotactic space served for ROI delineation. The ROI template follows the anatomical parcellation scheme according to Brodmann comprising 41 regions [1].
Results
Standard values (mean, SD, range) for each receptor and region are given in the table. Briefly, the ROI analysis yielded high 5 HT1A binding in Brodmann area (BA) 10, 11, 20–22, 25, 28, 35–38, 42, 48 (mainly frontal and temporal regions, insula), for the 5-HT2A receptor in BA 17, 21, 22, 31, 37, 39, 40, 42, 44–47 (mainly frontal, occipital and temporal regions), for the 5 HT1B receptor in BA 9, 10, 17, 18, 31, 32, 39, 40, 42–47 (mainly frontal, occipital and temporal regions), and for the 5-HTT in BA 25, 28, 34 (olfactory, amygdala), respectively.
Binding potential values for the major inhibitory (5-HT1A, 5-HT1B) and excitatory (5-HT2A) receptor subtypes and the transporter (5-HTT) measured by the selective radioligands [carbonyl-11C]WAY, [18F]Altanserin, [11C]P943 and [11C]DASB.
Regions of interest follow the scheme according to Brodmann. Left and right hemisphere are averaged.
Conclusions
This database grants quantitative information considering the major inhibitory (5-HT1A, 5-HT1B) and excitatory (5-HT2A) serotonin receptor subtypes and the 5-HT transporter in healthy subjects. This knowledge can improve our interpretation of the alterations taking place in the serotonergic system during neuropsychiatric disorders. It emphasizes on the existence of widely variant topologies, functions and interactions of the binding proteins in studying the disorders characterized by major serotonergic involvement. Moreover, the database contributes to standard definitions of reference ranges and provides a template. Besides, the standard ROI atlas permits a comparison between the tracers, which is unbiased in terms of ROI delineation, size and location.
P041. Whole body biodistribution and radiation dosimetry in humans of a new PET ligand, [18F]-FEPPA, to image translocator protein (18 kDa), and influence of genotype
Romina Mizrahi1,2, Pablo Rusjan1,
1Centre for Addiction and Mental Health, Toronto, Ontario, Canada; 2Department of Psychiatry, University of Toronto, Ontario, Canada
Background
[18F]-FEPPA is a novel PET radiotracer for investigating the translocator protein (18 kDa, TSPO). The aim of this paper was to estimate the human radiation dosimetry for [18F]-FEPPA from PET whole body biodistribution measurements.
Methods
A Biograph HiRez XVI tomograph was employed to acquire PET whole body data following the injection of [18F]-FEPPA (ca. 190 MBq) into six healthy subjects (4 males, 2 females). The PET acquisition duration was 120 minutes over 10 discrete time frames. Following the reconstruction of the emission image data, eight organs were identified in the images as exhibiting uptake above background. Regions of interest were delineated on these organs and time–activity curves (TAC) generated. The TAC data were normalized for the injected activity, specific organ density and volume, from which cumulated activities (previously known as ‘residence times’) were calculated. The cumulated activities were then used in OLINDA/EXM 1.1 to calculate the internal dose assessment. In addition, genotyping of rs6971 was performed using peripheral leukocytes to determine binding affinity class for each subject.
Results
Of the 6 subjects, one had an atypical [18F]-FEPPA biodistribution, similar to that of “low affinity binders” previously reported with the TSPO tracer [11C]-PBR-28. The mean effective radiation dose was estimated to be 20.9±2.7 μSv/MBq when excluding the “low affinity binder”, and 20.3±2.9 μSv/MBq upon including the “low affinity binder”. For the effective dose equivalent the corresponding values are 29.5±5.8 μSv/MBq, excluding the “low affinity-binder”, and 27.8±6.7 μSv/MBq upon including the “low affinity binder”. The organ receiving the highest dose was the spleen at 81.98±23.52 μSv/MBq averaged over all the subjects. All the results are presented as mean±std.dev. We observed agreement between the TSPO Ala147Thr genotypes and [18F]-FEPPA whole body biodistribution, with the single “low affinity binder” being the sole carrier of the minor allele (T/T) and the subjects exhibiting high affinity binding carrying the major allele (A/A).
Conclusions
The estimated radiation dose for [18F]-FEPPA is consistent with those for other neuroreceptor ligands labelled with F-18. Based on conventionally accepted dose limits, [18F]-FEPPA may be used for multiple PET scans in the same subject. TSPO protein Ala147Thr polymorphism determines whole body biodistribution for [18F]FEPPA.
P042. In vivo imaging of dopamine D3 receptors in alcoholism using [11C]PHNO-PET, and a selective D3 receptor antagonist
D.
1Neuropsychopharmacology Unit, Centre for Pharmacology and Therapeutics, Imperial College London, UK; 2Clinical Imaging Centre, GlaxoSmithKline, London, UK; 3FMRIB centre, Department of Clinical Neurology, University of Oxford, UK; 4Department of Imaging, Charing Cross Hospital, London, UK; 5CEDD for Neurosciences, GlaxoSmithKline, Verona, Italy
Background
Evidence from animal studies supports a role for the dopamine D3 receptor (DRD3) in alcohol reinforcement or liking. 1 Sustained voluntary alcohol drinking in rats has been associated with an upregulation of striatal DRD3 gene expression, 2 and selective blockade of DRD3 in rat reduces ethanol preference, consumption, 3 and cue-induced reinstatement of alcohol self-administration. 4 We compared the DRD3 status in the brains of alcohol dependent patients (ADP) to that of healthy controls (CTR) using the DRD3-preferring agonist positron emission tomography (PET) radioligand [11C]PHNO and a selective DRD3 antagonist, to enable the separation of DRD3 and DRD2 contribution to the [11C]PHNO PET signal.5–7
Methods
Sixteen male ADP subjects (abstinent for 415, range 39 to 893, days) and 13 healthy non-dependent males matched for age and smoking status were scanned with PET following the administration of [11C]PHNO pre- and 3hr-post the administration of a selective DRD3 antagonist (GSK598809 60 mg p.o.). A two-tissue compartmental model was used to derive regional volumes of distribution (VT). The regions examined were: Substantia nigra, hypothalamus, pallidum, thalamus, ventral striatum, putamen, caudate, and cerebellum.
Results
The cerebellar VT demonstrated a significant decrease following the administration of GSK598809 (15.9±9.2% in ADP and 11.4±6.1% in CTR), hence further analyses were conducted on regional VT values. The CTR and ADP groups were well matched for age (41.5±10.3 vs 42.4±9.4 year). The baseline scans for each group were well matched for injected mass of PHNO (20±4 vs 19±3 ng/kg), and injected [11C]PHNO activity (210±58 vs 201±60 MBq). Baseline [11C]PHNO VT was significantly higher in ADP when compared with CTR in hypothalamus (16.5±4 vs 13.7±2.9, p=0.015), a region in which the [11C]PHNO signal in healthy controls almost entirely reflects DRD3 availability. There were no other significant regional differences in VT between the two groups. No significant relationships between duration of alcohol abstinence and baseline regional VT values in hypothalamus or any other brain region were detected.
Conclusions
The reductions in regional VT following a single 60 mg dose of GSK598809 were consistent with those observed in previous studies across all regions. 9 The regional changes in VT following DRD3-blockade were consistent between the two groups, indicating the regional fractions of DRD3 are similar in the two groups and the increased [11C]PHNO binding in the hypothalamus in APD is explained by elevations of the DRD3 in this group. To our knowledge, this is the first study where brain DRD3 levels are assessed in alcohol dependent individuals. In the light of the existing pre-clinical support for a role of selective DRD3 antagonism in controlling addiction behavior, this preliminary finding of regionally increased DRD3 binding is relevant to future therapeutic strategies targeting the DRD3.
P043. Norepinephrine binds to 11C-PIB labeled β-amyloid plaques in post-mortem Alzheimer's disease brain
University of California Irvine, USA
Background
Alzheimer's disease (AD) brain is characterized by the accumulation of Aβ-amyloid senile plaques (SP) and neurofibrillary tangles (NFT). Accumulation of SP and NFT affects neurotransmitter-receptor functions in the AD brain. We examined interaction of the neurotransmitter, norepinephrine (NE) with 11C-PIB labeled SP in post-mortem human AD brain.
Methods
Human post-mortem brain samples (AD, n=4, age 77–89, SP Stage C and controls, n=4; age 81–90 SP Stage 0-A) were obtained from our ADRC. Brain slices (7 μm thick) of hippocampus and frontal cortex of the frozen tissue were obtained on a Leica 1850 cryotome. Slides were washed in TBS and stained with anti-Aβ antibody 4G8 (Covance, Princeton, NJ). 11C-PIB was produced in specific activities of approx. 1 Ci/μmole and autoradiographic studies were carried out by exposing tissue samples on storage phosphor screens. Adjacent slices were incubated with 11C-PIB (5–25 μCi/cc) in 40% ethanol at 37°C for 0.5 hr. For nonspecific binding 100 ìM PIB was used. Competition with NE (1 nM to 500 μM) was carried out. For comparison dopamine, serotonin, nisoxetine, citalopram and propranolol were also used. Slices were washed with cold water, 70%−90%−70% ethanol, water for 2,1,1,1,1 min, respectively. Air dried slices were exposed overnight and regions of interest were drawn on the slices and extent of binding of 11C-PIB was computed as DLU/mm2 using OptiQuant program (Packard Instruments Co). Percentage change in binding of 11C-PIB under different challenge conditions (11C-PIBDrug) was compared to control experiment (11C-PIBCont; no drug treatment). Thus, effect of various drugs on “PIB binding site” was compared to baseline as %PIBocc=(11C-PIBCont−11C-PIBDrug/11C-PIBCont) × 100.
Results
Presence of Aβ-amyloid plaques in the hippocampus and frontal cortex of AD subjects was confirmed using 4G8 immunostains while control subjects had few (Figure 1). Significant 11C-PIB binding was found in AD hippocampus and frontal cortex slices and was consistent with immunostaining. The control subjects sections had little 11C-PIB binding and mostly resembled nonspecific binding in the AD brain in the presence of 100 μM PIB. Ratio of 11C-PIB binding in AD brains versus control brains was found typically to be >5 and in the presence of PIB this ratio was reduced to approx. 1 (Figure 1). Norepinephrine competed with 11C-PIB most strongly in all the subjects studied, in both hippocampus and frontal cortex with an IC50 of approx 10 to 50 μM. Effect on 11C-PIB by dopamine, serotonin, nisoxetine (norepinephrine uptake inhibitor), citalopram (serotonin uptake inhibitor) and propranolol (β-adrenergic blocker) were much smaller or absent at a concentration of 100 μM compared to NE (−70 to −80%). Norepinephrine displayed a dose related effect on 11C-PIB suggesting competitive binding.
11C-PIB in AD brain: 4G8 antibody immunostain of AD hippocampal sections showing SP (A); 11C-PIB binding to postmortem human AD brain hippocampal sections, in vitro autoradiographs (B); Displacement of 11C-PIB by norepinephrine (100 μM) in postmortem human AD brain hippocampal sections, in vitro autoradiographs (C); Nonspecific binding with 100 μM PIB (D); Graph showing 11C-PIB binding in AD and Control brain hippocampal sections in the absence and presence of norepinephrine and PIB.
Conclusion
11C-PIB binding site in the human post-mortem AD brain has a particular affinity for norepinephrine. Further experiments are needed to assess the physiological impact of norepinephrine binding to the SP in AD. Our preliminary results indicate that the Aβ-amyloid plaque presents itself as a competing binding site for NE and may interrupt NE transmission or have a role in the pathogenesis of AD.
Acknowledgements
NIH/NIA R01AG029479, R33AG030524, P50 AG16573.
P044. Three weeks of daily exposure to bright light does not alter striatal serotonin transporter binding in healthy Scandinavians
1Neurobiology Research Unit and Center for Integrated Molecular Brain Imaging (Cimbi), Rigshospitalet and University of Copenhagen, Denmark; 2PET and Cyclotron Unit, Copenhagen University Hospital, Rigshospitalet, Denmark; 3Cimbi and Danish Research Centre for MR, Copenhagen University Hospital, Hvidovre, Denmark; 4Department of Cellular and Molecular Medicine, Copenhagen University, Denmark
Background
The serotonin (5-HT) transmitter system modulates a number of psychophysiological functions including mood, sleep, arousal, appetite and circadian rhythms. Lack of daylight is prominent at high latitudes and this may be perceived as an environmental stressor associated with a high frequency of seasonal affective disorder that can be treated with bright light (BL) therapy. By means of 11C-DASB PET, we and others have in cross-sectional studies observed season dependent fluctuations in the serotonin transporter (SERT), with high striatal binding around winter solstice and low binding around summer solstice. In addition, we have identified a genotype-dependent interaction with these environmental stressors with SERT binding being dependent on the carrier status of the 5-HTTLPR promoter polymorphism (Kalbitzer et al 2010). This gene∗environment paradigm predicts the SERT fluctuations with a negative correlation between SERT binding and daylight minutes in carriers of the short 5-HTTLPR allele (S-allele), but less so in homozygote carriers of the long allele (L-allele). The aim of the present study was to examine if three weeks of daily exposure to BL versus placebo light was associated with a reduction in SERT binding and if any such change was depending on 5-HTTLPR carrier status.
Methods
Seventy healthy males were pre-screened for inclusion and based on their 5-HTTLPR triallelic carrier status they were invited for participation. Of these, 24 healthy male volunteers without any history of psychiatric or neurological disorders participated in the study that took place in Copenhagen between November and January. Exclusion criteria were significant medical history, known retinal pathology and use of photosensitizing medications. All subjects had a normal neurological exam and unremarkable brain MRI. The healthy males were divided into three groups: Eight LA/LA-carriers received BL (10,000 lux light) and 16 S-allele carriers were randomized to exposure to either placebo (1,000 lux light) or BL for half an hour every morning between 7AM and 9AM for three weeks. Compliance was assured through daily text reminders and an online logbook system. Participants were investigated with MRI and 11C-DASB on an HRRT PET scanner (mean inj. dose: 589±29, range 421–609 MBq) before and after three weeks of BL or placebo light. Quantification was done with MRTM2 that generates the outcome parameter BPND. BPND before and after BL/placebo was compared with a two-tailed paired t-test.
Results
Data are given in Table 1. The three groups were comparable in age and daylight minutes on the date of their first PET-scan. The numerical 3 weeks BPND test-retest variability in the placebo treated S-allele carriers was modest: 6.3±3.9% (2.4–12.9) in caudate and 8.6±5.3% (3.6–17.0) in putamen. There was no significant intervention-associated change in SERT BPND in any of the groups.
Age, daylight minutes and change in BPND in the three different groups. Values are given as mean±standard deviation and range (in parentheses)
Conclusions
We did not observe a significant effect of intervention with three weeks of BL on striatal SERT binding, irrespective of 5-HTTLPR genotype. Either the effect of BL is insufficient to mimic the daylight exposure in the summer, or season-related factors other than light determine striatal SERT fluctuations.
P045. Abnormalities in GABAergic neurotransmission and glucose metabolism in patients with essential tremor in PET study using [18F]Flumazenil and [18F]FDG
Yu Kyeong Kim, Hyun Woo Kwon,
Department of Nuclear Medicine and Neurology, Seoul National University College of Medicine, South Korea
Background
Essential tremor (ET) is one of the common neurological disorders characterized by postural and kinetic tremor and motor incoordinations, causing mild, often severe, disturbances of daily activity. However, its pathological mechanism is still unknown. Recently, an alteration of GABAergic neurotransmission system has been proposed as a key role in the generation of tremor. Our study aims to investigate the regions with significant changes in GABAergic neurotransmission and glucose metabolism using [18F]Flumazenil (FMZ) PET and [18F]FDG PET, respectively, and to explore how these regional alterations are related as an underlying pathophysiology of the disease.
Methods
Twelve patients (M;F=4:8, age=55.8±12.1 yr) diagnosed as essential tremor and age-matched 18 healthy controls (M:F =5:13, age=56.2±6.1 yr) were enrolled. All subjects performed both FDG PET and FMZ PET. Regional cerebral metabolism and receptor availability of GABA/BZ were compared in voxel-wised manner using SPM2 and region of interest (ROI) analysis.
Results
In patients with ET, significant hypometabolism was identified in the bilateral primary and supplementary motor cortex and and left posterior parietal lobe (p<0.005, uncorrected, extend threshold k>100). On the other hand, no significant regional hypermetabolism compared with controls was noted in patients. However, a trend of cerebellar hypermetabolism showed at lesser stringent criteria, whereas cortical and cerebellar GABA receptor availability showed a trend of increase in patients with ET compared with healthy controls even though the difference did not reach the significance level in voxel-based analysis.
Conclusions
Our study demonstrated a metabolic abnormality with GABAergic alteration in the brain regions of motor related cortices and cerebellum. Our findings with [18F]FMZ-PET support the idea that the role of cerebellum in modulation of motor activity initiated by other brain regions is dysfunctional due to the abnormality in GABAergic neurotransmission, and it causes the kinetic tremor characteristics in ET patients. Also, a close relationship in the pattern of the distribution of [18F]FMZ and [18F]FDG tracer in ET patients remains to be further investigated.
P046. The effects of a serotonin 2C receptor agonist on endogenous and amphetamine-induced dopamine levels in rhesus monkeys
1PET Center, Yale University, New Haven, Connecticut, USA; 2 Neuroscience, Pfizer Worldwide Research and Development, Groton, Connecticut, USA
Background
A body of research has demonstrated the effects of serotonin (5HT) systems on striatal dopamine (DA) release. Specifically, 5HT2A and 5HT2C serotonin receptors have been shown to exert opposite effects on striatal dopamine (for review see Boureau & Dayan, 2011). PET imaging in rats showed that 5HT2C antagonists increase striatal DA, while 5HT2A antagonists attenuate amphetamine-induced DA elevation in the striatum. The effects of 5HT2C agonists would thus be expected to mirror those previously shown for 5HT2A antagonists, namely, to decrease striatal DA. Tyrosine depletion studies in humans and rats have demonstrated that decreases in amphetamine-induced DA release in the striatum are detectable using PET imaging (Leyton, 2004; Le Masurier, 2004). The current study aims to investigate the effects of Vabicaserin, a selective 5HT2C agonist, on striatal DA and DA release. We hypothesize that Vabicaserin will attenuate striatal increases in DA following amphetamine administration, and may also decrease levels of endogenous DA in the absence of amphetamine.
Methods
The current study in rhesus monkeys utilizes comparable experimental designs in two separate PET investigations, one using the D2 receptor radioligand 11C-PHNO and one using the D2/D3 receptor radioligand 18F-Fallypride (both bolus injections). In each investigation, monkeys underwent four scans: 1) baseline (radioligand only), 2) bolus plus constant infusion of Vabicaserin (1 mg/kg) starting ∼80 minutes before the scan, 3) bolus injection of amphetamine (0.4 mg/kg) ∼5 minutes before scan, 4) Vabicaserin infusion and amphetamine injection administered ∼80 minutes and ∼5 minutes before scan, respectively. In the 11C-PHNO investigation, a 5th scan was performed on each monkey with a double dose of Vabicaserin (2 mg/kg). Thus far, two monkeys have completed all 5 scans in the 11C-PHNO investigation. Scan lengths for 11C-PHNO and 18F-Fallypride scans were 120 and 180 minutes, respectively. Data from both investigations were analyzed by extracting time activity curves from striatal and extrastriatal regions of interest, and calculating binding potentials (BPND) using the simplified reference tissue model (SRTM) with cerebellum as the reference region.
Results
Preliminary results from the 11C-PHNO investigation in two rhesus monkeys revealed Vabicaserin-induced decreases in endogenous DA in caudate (5.2%) and putamen (8.0%). For Vabicaserin doses resulting in plasma levels between 0 and 200 ng/mL, we observed a dose-dependency in caudate and putamen. Higher plasma levels of Vabicaserin resulted in greater decreases in endogenous DA. Vabicaserin co-administration with amphetamine was reliably performed in one monkey. Following amphetamine treatment, caudate and putamen showed a 39% and 41% increase in DA, respectively, compared to baseline. Following a co-administration of amphetamine and Vabicaserin, a comparatively lower 36% increase in DA from baseline was seen in both regions. 18F-Fallypride studies of extrastriatal regions are ongoing.
Conclusions
Preliminary data suggest that decreases in striatal DA as well as attenuations in amphetamine-induced DA release after administration of a 5HT2C agonist (Vabicaserin) are detectable with PET imaging using 11C-PHNO. We continue to study the possibility of using PET to assay the effects of 5HT2C agonists in humans, as Vabicaserin is emerging as a potentially useful drug treatment of psychiatric disorders such as schizophrenia (Shen, 2011).
P047. Evaluation of D3 versus D2 receptor binding in rats and D2R knockout mice using in vivo PET imaging with [18F]desmetoxyfallypride and [18F]fallypride
University of California Irvine, USA: 1Radiological Sciences; 2Microbiology & Molecular Genetics
Background
[18F]Desmetoxyfallypride (DMFP) and [18F]fallypride are moderate and high affinity receptor radioligands that binds to both D2 and D3 receptors (Mukherjee et al. 2001). Evaluation of relative binding of D2 to D3 receptors in vivo with established tracers presents scientific interest. We investigated dopamine D3 receptor distribution and the potential for quantification of [18F]fallypride and [18F]DMFP binding to the D3 receptors in wild type (WT) and homozygous D2 deficient (KO) mice via dynamic PET. Separate baseline and competition studies were conducted with [18F]DMFP in normal rats.
Methods
One WT mouse and two KOs (33±1 g) were imaged with [18F]fallypride (2.7±1.1 MBq) for 120 min and with [18F]DMFP (3.0±0.7 MBq) for 90 min in the Inveon preclinical scanner. Sprague-Dawley rats (395±16 g) received 2 [18F]DMFP (29.0±1.0 MBq) scans on 2 separate days (3 weeks apart). The first day scan was a control. On second day the animals were rescanned after pre-injection of BP897 (68-fold partial D3R agonist) at low (0.25 mg/Kg) and high dose (1 mg/Kg), received as IV bolus 15 min before [18F]DMFP. All PET images were co-registered to a mouse MR template (Ma et al. 2008). All animals also received CT scans for attenuation and scatter correction. Regions of interest were drawn on dorsal striatum (DStr), ventral striatum (VStr), and hypothalamus (Hyp). In mice binding potentials (BPND) were computed using interval ratio method (10–90 min) (Ito et al. 1998). Graphical analysis was used for rat data (Logan et al. 1996). Cerebellum was used as a reference region.
Results
The BPNDs of [18F]fallypride were 4.05 (DStr), 2.98 (VStr), 0.97 (Hyp) in the WT and 0.04±0.02 (DStr), 0.44±0.05 (VStr), 0.76±0.27 (Hyp) in the KO. The BPNDs of [18F]DMFP were 1.44 (DStr), 1.11 (VStr), 0.62 (Hyp), in the WT and < 0 (DStr) 0.18±0.20 (VStr), 0.44±0.04 (Hyp) in the KO. The average D3R/(D2R+D3R) fraction was (16±18) % in VStr and (71±7) % in Hyp. In rats, the decrease in BPND at low dose of BP897 measured with [18F]DMFP was 41% (DStr) and 61% (VStr). The decreases at high BP897 dose were 70% (DStr), and 49% (VStr).
Conclusions
Mice studies showed low but detectable binding in VStr and moderate binding in Hyp of KO mice, attributed to D3R. The low or negative BPND values in DStr for both [18F]fallypride and [18F]DMFP indicate the absence of D2R in low affinity states in KO mice. The KO mouse appears to be a good model for evaluating the D3R only binding in vivo. The rat imaging showed that BP897 has a higher in vivo affinity for D3R at low dose than at higher dose as indicated by differential changes in BPND in VStr versus DStr. Further baseline and blocking imaging studies are underway aimed to more accurately asses the specificity of both tracers for D3R in vivo in WT animals.
Acknowledgements
Research supported by NIH EB006110.
P048. Differential roles of dopamine D1- and D2-like receptors in impulsivity: a preliminary PET study
1Department of Psychiatry and Biobehavioral Sciences and Semel Institute at University of California at Los Angeles, USA; 2Veterans Administration Greater Los Angeles Healthcare System, California, USA; 3Department of Molecular and Medical Pharmacology at University of California at Los Angeles, USA; 4Department of Physics at University of California at Irvine, USA; 5Department of Psychology at University of California at Los Angeles, USA
Background
Impulsivity, a personality trait that is characterized by the tendency to act without forethought, is exhibited by patients with neuropsychiatric disorders that feature dysfunction of dopaminergic neurotransmission. These include schizophrenia, mood disorders, attention deficit-hyperactivity disorder and addictions. Recent studies have provided evidence for roles of cortical dopamine D1-like receptors (D1Rs) (1) and striatal D2-like receptors (D2Rs) (2, 3) in influencing impulsive behaviors, but a complete picture of the contributions of these dopaminergic systems to self-control (and impulsivity) in healthy human subjects is lacking. We therefore addressed this question using positron emission tomography (PET) with [11C]NNC112 and [18F]fallypride to quantify availability of D1Rs and D2Rs, respectively, in participants who gave self-reports of impulsivity on a well-established assessment tool.
Methods
Sixteen healthy volunteers (9 men), 29.6±8.8 (mean±SD) years of age, completed the Dickman Impulsivity Inventory, which provides subscales for both functional and dysfunctional impulsivity. MRI scans were acquired for coregistration and quantitation of PET data. Ten subjects underwent [11C]NNC112 PET scans, and thirteen underwent [18F]fallypride PET scans. Volumes of interest (VOIs) for the caudate nucleus, putamen, and nucleus accumbens, were defined on individual MRI scans and transferred to the coregistered PET images. Binding potential (BPND) was calculated for each region using the simplified reference tissue model, with the cerebellum as the reference region. Voxel-wise BPND maps for each radiotracer were also generated and analyzed using SPM 8.
Results
VOI analysis of striatal binding of both tracers showed no correlations with functional impulsivity, but notable associations with dysfunctional impulsivity were observed. A negative correlation with [18F]fallypride BPND in all striatal regions [caudate (r=0.64, p=0.02), putamen (r=0.65, p=0.02), accumbens (r=0.61, p=0.03)] was observed. Voxel-wise regression analysis also showed a negative correlation between [18F]fallyprideBPND in the striatum and dysfunctional impulsivity (p<0.01, uncorrected, Figure 1). For [11C]NNC112 BPND, VOI analysis revealed no significant correlations with dysfunctional impulsivity. However, voxel-wise regression analysis showed a positive correlation between [11C]NNC112 BPND in the medial orbitofrontal cortex and dysfunctional impulsivity (p<0.01 uncorrected, Figure 1).
Statistical parametric maps resulted from voxel-wise regression analysis between BPND and Dickman Impulsivity Inventory (subscale: dysfunctional impulsivity) score. Voxels with positive and negative correlation are shown on a standard brain in the left ([11C]NNC112) and right ([18F]fallypride), respectively. Data are shown at a liberal threshold (p < 0.05, uncorrected) for display. The scale bars represent values of T.
Conclusions
These preliminary findings support the view that the Dickman functional and dysfunctional impulsivity scales measure different subconstructs of impulsivity. They also extend findings that D1Rs in the medial orbitofrontal cortex and D2Rs in the striatum play important roles in impulsivity (1–3), and suggest that D1Rs and D2Rs may act differently on impulsivity.
Acknowledgements
This work was supported by the Consortium for Neuropsychiatric Phenomics (NIH Roadmap for Medical Research grants UL1-DE019580, RL1DA024853).
P049. Mannitol infusion and blood brain barrier permeability
1Aarhus University Hospital, PET Center, Denmark; 2Aarhus University Hospital, Psychiatric Hospital and PET Center, Denmark
Background
Endothelial cells build the blood brain barrier (BBB) and effectively inhibit compounds entering the brain. The “barrier” results from the selectivity of the tight junctions between endothelia cells in CNS vessels that restrict the passage of solutes. Several approaches have been developed to overcome this barrier and to improve the effectiveness of chemotherapy for brain tumors. Osmotherapy applying intracarotid mannitol (20–25%) infusion results in an opening of the tight junctions and gives an approximate 10 min window with increased BBB permeability. The mannitol infusion causes the contents of the artery to be hyperosmotic to the cell. Water leaves the cell and enters the artery in order to recreate an osmotic equilibrium. This loss of water causes the cells to shrivel and shrink, stretching the tight junctions between the cells. It was the aim to investigate if a mannitol infusion might be used in brain PET studies to improve the brain uptake of PET tracers with slow or no brain uptake.
Methods
All procedures involving animals were performed under a license granted from the Danish Ministry of Justice. The effect was investigated both in a pig model and a rat model. In pig model, the effect on brain uptake of F-18 labeled FLT was chosen and radioactivity uptake and distribution was followed in a PET scanner. Mannitol was administered as intravenous infusion. In rat model, a Ga-68 labeled glycosylated peptide was applied briefly after a bolus injection of mannitol and brain radioactivity uptake and distribution was followed in μPET scanner. In another set-up, rats were injected with the Ga-68 labeled peptide, sacrificed after 30 minutes, brain dissected, and radioactivity measured in a well-counter.
Results
Generally, mannitol affected the physiological functions as expected: in both species we observed an increased urine production. In the pig model, the loss of fluid was corrected by increased saline infusion. Regarding brain uptake, we found in all experiments that the brain uptake was not affected considerably. For example, in the rat studies, brain uptake increased only by roughly 0.3%. However, even though the increase was very small, there is an indication of modified brain permeability by mannitol infusion even in PET experiments. It might be argued that to observe a bigger effect, an intracarotid infusion of mannitol is needed. Under the given conditions, the BBB is open for influx and efflux. Therefore, this approach might need a target in the brain that efficiently traps the applied tracers. Furthermore, with the mannitol infusion, the BBB is open for all compounds including the anesthesia. Especially the anesthesia might alter the brain uptake in ways that make PET scans difficult to interpret.
Conclusions
We found that mannitol affects physiological functions of anesthetized animals used for PET brain imaging, including a minor increase in opening of the BBB with slightly enhanced brain uptake of selected PET radiotracers. However, this procedure induces several new complications, resulting in experimental conditions that may be difficult to control and interpret.
P050. Initial assessment of the dopamine system in fibromyalgia
1Indiana University School of Medicine, Indianapolis, Indiana, USA; 2University of Wisconsin, Madison, USA
Background
Fibromyalgia (FM) is a chronic pain disorder that can be debilitating. It is estimated that FM affects 2% of the population. Many of the medications prescribed for FM are highly addictive, have limited clinical efficacy, and do not treat the cognitive symptoms of FM. The neurobiological substrates of FM are currently unknown. However, recent evidence indicates that altered dopaminergic transmission may play a role in pain disorders. Given that dopamine (DA) is essential for cognitive function, it is possible that the cognitive complaints of FM patients are mediated, in part, by abnormal dopaminergic functioning. To date, few studies have assessed DA function in vivo in FM patients. The objective of this study was to ascertain how the DA system in FM subjects differs from that of healthy controls. We hypothesized that (1) FM subjects would have a higher basal dopaminergic tone (reflected by lower D2 receptor availability), and (2) that they would exhibit relatively lower cortical DA release in response to a working memory (WM) task. We predicted that, in FM subjects, D2 availability and WM task-induced DA release would be related to subjective pain ratings and performance on the WM task, respectively.
Methods
Seven patients with a diagnosis of FM (28.9±7.4 y.o.) and six healthy CON (29.4±6.2 y.o.) completed study all procedures. Subjects underwent a MP-RAGE MRI and two [18F]-Fallypride (FAL) PET scans: one during a baseline attentional task (“0-back”), and one during a working memory task (“2-back”). Task order was counter-balanced across subjects. Parametric binding potential (BPND) images were generated from the dynamic FAL data using Logan graphical analysis, with the cerebellum as the reference tissue. SPM8 was used to compare baseline data between groups and to determine relative changes in DA levels during the WM task.
Conclusions
These pilot data suggest that symptoms of FM may be associated with abnormalities in both basal DA tone and altered patterns of DA release during a working memory challenge. We are currently studying how the results in FM subjects may relate to pain ratings and WM performance.
P051. Molecular imaging genetics of the serotonin-1A receptor investigating the common rs6295 single nucleotide polymorphism
1Medical University of Vienna, Department of Psychiatry and Psychotherapy, Austria; 2Medical University of Vienna, Department of Nuclear Medicine, Austria; 3Ludwig-Maximilian-University Munich, Genetics Research Center, Germany
Background
The inhibitory 5-HT1A receptor plays a central role in serotonergic neurotransmission [1]. The transcriptional regulation of the HTR1A gene by allele-specific factors has been shown for Deaf-1 that represses at the HTR1A C(−1019)G SNP rs6295 [2,3]. There is a region-specific regulation by Deaf-1, repressing 5-HT1A in dorsal raphe nuclei (DRN) and enhancing postsynaptic 5-HT1A heteroreceptor expression [2,3]. Deaf-1/NUDR cannot bind with the G(−1019) allele [2], therefore we would expect an increase of 5-HT1A receptor binding potential (BPND) in DRN and a decrease in cortial areas [3]. The aim of this PET study was to demonstrate that the functional genetic variation rs6295 affects 5-HT1A BPND.
Results
22% of the subjects were CC, 48% CG and 30% GG carriers. No significant SNP effects were observed for voxel-wise 5-HT1A maps (F<7.97, p>0.001 uncorrected) nor when grouping alleles (p>0.001 uncorrected). Similarly, no significant effects were found when adjusting whole-brain maps to individual 5-HT1A BPND in DRN (p>0.001 uncorrected). Also, no significant differences were observed for the ROI-based analysis of DRN (p>0.39), see Figure.
Scale indicates 5-HT1A receptor binding potential (BDND). The white cross (left) indicates the dorsal raphe region corresponding to the graph (right).
Conclusions
We found no effect of the SNP rs6295 on 5-HT1A BPND. Our data indicate that in healthy subjects (in contrast to patients with depression [4]) this antagonistic effect of genotype between pre- and postsynaptic receptors may be compensated by negative feedback mechanisms via 5-HT1A autoreceptor-mediated inhibition of serotonergic firing and serotonin level-dependent 5-HT1A heteroreceptor expression [1].
P052. Preliminary analysis of the difference of amyloid beta detectability by 11C-pittsburgh compound-B PET (PIB PET) in different image display conditions
1Division of Clinical Radiology, Gunma University Hospital, Maebashi, Gunma, Japan; 2Department of Diagnostic Imaging and Nuclear Medicine, Gunma University Graduate School of Medicine, Japan
Background
For image display of 11C-Pittsburgh Compound-B PET(PIB PET), distribution volume ratio (DVR), standardized uptake value ratio from 0 to 70 minutes (SUVR 0–70) and standardized uptake value ratio from 50 to 70 minutes(SUVR 50–70) are usually used. In this study, the difference of the detectability of amyloid beta deposition was investigated among these three types of images.
Methods
Image sets from 25 examinations (25 patients) were included in this study. All examinations were performed using a PET/CT scanner (Discovery STE, GE Healthcare, USA). PIB was produced in our own cyclotron (CYPRIS HM-18, Sumitomo heavy industry, Japan) and intravenously administered to patient, followed by dynamic image acquisition just after injection. DVR image was constructed by PMOD software (PMOD technologies, Switzerland) and compared with SUVR images. Image conditions for 3 types of images are fixed to show similar color gradients. Top image matrix values for DVR, SUVR 0–70 and SUVR 50–70 were adopted as a representative value for each image and correlations of these values from three different images were evaluated. In visual qualitative analysis, concordance and disconcordance of each three images in the detection of amyloid beta were compared.
Results
Significant correlation was noted between SUVR 0–70 and SUVR 50–70 images. Significant difference of image matrix values of 3 different images were noted each other. In these 3 image conditions, SUVR 50–70 showed worst detectability rate of amyloid beta and also indicated less concordance rate with other images.
Conclusions
Although mild correlation was noted between SUVR 0–70 and SUVR 50–70 images, image matrix value showed big difference between different image display conditions. In conclusion, to improve the diagnostic performance, careful consideration and correlation of 3 different images are required in routine clinical condition.
P053. Effect of a nicotine vaccine on nicotine binding to the beta2-nAChRs in vivo in human tobacco smokers
1Yale University, New Haven, Connecticut, USA; 2University of Toronto, Canada; 3Inst Neurodegen D/O, New Haven, Connecticut, USA; 4University of Minnesota, Minneapolis, Minnesota, USA
Background
Nicotine acts in the brain to promote smoking primarily by binding to beta2-containing nicotinic acetylcholine receptors (β2∗-nAChRs). The effects of nicotine from smoking one tobacco cigarette are significant (80% of β2∗-nAChRs occupied for >6 h). This likely contributes to the maintenance of the smoking dependence and low cessation outcomes. Development of nicotine vaccines has provided potential for alternative treatments. We used [123I]5IA-85380 and SPECT to evaluate the effect of 3′-AmNic-rEPA (Nabi Biopharmaceuticals) on the amount of nicotine that binds to the β2∗-nAChRs in smokers.
Methods
Eleven smokers (36.1±12.9 yo; 19±11 cig/d for 10±7 y) who were dependent on nicotine (Fagerström Test of Nicotine Dependence score=5.5±3; plasma nicotine 9.1±5 ng/mL) participated in 2 SPECT scan days: before and after immunization with 4–400 μg doses of 3′-AmNic-rEPA over 4 months. On the day of each SPECT scan, 3 30-min baseline emission scans were obtained, followed by administration of IV nicotine (1.5 mg/70 kg) and up to 9 30-min emission scans. β2∗-nAChR availability was quantified as VT/fP and receptor occupancy was derived using the Lassen plot approach.
Results
Immunization led to asignificant 12.5% reduction in β2∗-nAChR occupancy by nicotine (F=5.19, p=0.05) with an estimated 23.6% reduction in available nicotine to enter the brain. Significant positive correlations were observed between β2∗-nAChR occupancy by nicotine and nicotine injected before but not after vaccination (p=0.05 vs. p=0.98). There was a significant reduction in the daily number of cigarettes and desire for a cigarette (p=.01 and p=.04, respectively).
β2∗-nAChR occupancy by nicotine after IV nicotine administration on two SPECT scan days (average across subjects): 1) before immunization (left – closed circles) and 2) after immunization (open circles). Lines represent mean±standard deviation. There was a significant 12.5% decrease in nicotine's binding to the receptors after the immunization. Difference in binding was calculated as [VT/fp rafter / VT/fp before] ∗ 100=% difference.
Conclusions
This proof-of-concept study demonstrates that immunization with nicotine vaccine can reduce the amount of nicotine binding to β2∗-nAChRs and disrupt the relationship between nicotine administered vs. nicotine available to occupy β2∗-nAChRs.
P054. Disrupted adenovirus-based anti-cocaine vaccine: assessment of efficacy based on PET dopamine transporter (DAT) imaging and cocaine challenge
Weill Cornell Medical College, New York, New York, USA: 1Departments of Radiology; 2Departments of Genetic Medicine
Background
We have recently developed a disrupted “adenovirus cocaine vaccine” (dAd5GNE) and preclinical studies have demonstrated that a standard prime-boost regimen of dAd5GNE evokes high titer anti-cocaine antibodies (anti-coc). In order to test the efficacy of dAd5GNE, we have performed DAT-PET in non-human primates, before and after vaccination using [11C]PE2I, a cocaine analog with high specificity to DAT.
Methods
Prior to vaccination, 4 rhesus monkeys had DAT-PET for 90 min with [11C]PE2I (5 mCi; SA: 10–15 Ci/μmole). At each session, DAT-PET was performed before and after cocaine challenge (1 mg/kg). Following vaccination with dAd5GNE, identical DAT-PET studies were repeated 2–3 times over a period of 4 months. Based on PET-MRI fused images and using the cerebellum as the reference region, PE2I binding potential (BP) for caudate and putamen were estimated for each PET study with and without cocaine challenge.
Results
Prior to vaccination, cocaine challenge significantly reduced PE2I BP (9.1±1.5 vs 4.0±1.0) and an estimated DAT occupancy was 55±12%. Following vaccination, cocaine challenge effects a significantly smaller reduction in BP (DAT occupancy <10%). There was variation in the plasma anti-coc among the 4 monkeys with 2 having titers ∼106. In general, the DAT occupancy following vaccination showed good correlation with the anti-coc titers for both caudate (r=0.72) and putamen (r=0.91). Additional vaccinations increased anti-coc levels and decreased DAT occupancy following cocaine challenge.
Conclusions
This primate study demonstrates that vaccination with dAd5GNE evokes high anti-coc titers sufficient to sequester intravenous administered cocaine in blood and preventing access to its receptors in the brain. DAT-PET with [11C]PE2I provides a reliable in vivo assay to assess the efficacy of dAd5GNE.
P055. Regional dopamine release and dopamine D2 receptor levels in normal weight, overweight and mildly obese subjects
1Vanderbilt University School of Medicine; 2Department of Psychology, Vanderbilt University; 3Department of Radiology, Vanderbilt University School of Medicine; 4Department of Electrical Engineering and Computer Science, Vanderbilt University; 5Department of Psychology, Vanderbilt University, Nashville, Tennessee, USA
Background
It has been suggested that obesity is similar to drug abuse in having impaired dopamine (DA) neurotransmission, particularly decreased DA D2 receptor levels, consistent with a reward deficient state in obesity (Wang, 2001; Volkow, 2008). However, a recent fMRI study reported increased food cue activation in adolescents at high risk for obesity suggesting that increased DA release and increased behavioral salience of food may lead to obesity (Stice, 2011). In addition an inverse parabolic curve of reward sensitivity versus BMI with a peak at BMI's in the low to mid 30's has been reported (Davis, 2008). These studies suggest that increased salience of food and increased food reward may lead to obesity. No study to date has reported the relationship of DA release to BMI in the development of obesity.
Correlation of right putamenal DA release versus BMI, r=0.581, P=0.023.
Methods
To examine this issue, we examined both regional DA D2 receptors in 34 healthy subjects (mean age=26.4, range 18–38; mean BMI=24.8, range 19–35; M=18, F=16), and amphetamine induced DA release using oral amphetamine (0.43 mg/kg) in 16 healthy subjects (8M, 8F; mean age=24.3, age range =21–32; mean BMI=25.2, BMI range of 19–35) using [18F]fallypride PET studies. Regional BPND's were calculated using the reference region method. Regions of interest were delineated for the caudate, putamen, ventral striatum, amygdala, substantia nigra, medial thalamus, and temporal cortex.
Results
For the group as a whole no significant correlations of BMI with available DA D2/3 levels were found for any region after covariation for age. The right amygdala had an r=−0.317 with a trend level P=0.068, uncorrected for multiple comparisons. No other region achieved a trend level comparison. Except for the right temporal cortex (r=0.112), all regional correlations of DA D2 levels with BMI were negative, −0.015 to −0.317. Age covaried correlation coefficients between regional d-amphetamine induced DA release and BMI were positive for all regions, except the right temporal cortex (r=−0.17), ranging up to r=0.581 (P=0.023) in the right putamen and r=0.568 (P=0.027) in the left substantia nigra. Correlation coefficients of 0.411 and 0.391 were seen in the right ventral striatum and caudate which approached a trend level.
Conclusions
These results and those of previous studies suggest that DA neurotransmission may undergo an evolution with the development of obesity; increased DA neurotransmission may be seen during the development of obesity but with the onset of extreme obesity, i.e. BMI's of greater than 45–50, impaired DA neurotransmission may occur.
P056. Serotonin transporter promotor variants are not associated with transporter binding in human brain: a PET and postmortem brain study
Columbia University/NYSPI, New York, New York, USA
Background
Serotonin transporter promotor low expressing variants are associated with depressive disorders. PET studies report low transporter binding in major depression. Childhood adversity is also associated with low transporter binding in major depressive disorder (MDD). We sought to determine the relationship between serotonin transporter gene promotor variants and serotonin transporter binding in the brain, in both in vivo PET studies using [11C]DASB, and postmortem brain samples from patients with MDD and controls. To our knowledge this is the first study to combine PET and postmortem studies to measure a biological phenotype associated with reported functional gene variants.
Methods
MDD suicides (N=18) and controls (N=40) without a psychiatric diagnosis were genotyped for three functional variants in both the 5′ and 3′ promotor regions. The sum of high expressing alleles at any of the three polymorphisms was used as the gene measure potentially related to expression level and therefore binding. Autoradiography with [3H]cyanoimipramine quantified binding in gyrus and sulcus of Brodmann areas (BA) 9 and 47. Live subjects: patients with a lifetime diagnosis of MDD and healthy volunteers were genotyped as above and underwent PET with [11C]DASB to quantify serotonin transporter binding. Mixed effect regression analyses were performed with log binding levels as response variable and number of high-expression alleles (0–4) as predictor, testing for possible effects on binding by brain region, and adjusted for gender and diagnosis.
Results
Serotonin transporter binding was not associated with allele load variable in postmortem groups. Although the interaction effect between allele load and brain region was significant (p=.0421), the correlation was not significant in either brain region (BA9: beta estimate=−0.05, t=−0.93, df=66, p=0.3560; BA47: beta estimate=−0.002, t=−0.04, df=66, p=0.9682). Acquisition and analysis of PET data is ongoing and results will be presented at the meeting.
Conclusions
We found no association between high expressing promotor allele load and serotonin transporter binding in postmortem human brain. Although the sample size is small, the direction of the small effect is the opposite of that hypothesized. We previously reported that in the brain of rhesus macaques, the carriers of the low-expressing rh5-HTTLPR alleles exhibited higher mean 5-HTT CpG methylation (Kinnally, 2010). And that high methylation correlated with low transporter binding and when the model accounted for DNA methylation then genotype did not correlate with binding. Low expressing alleles were associated with higher DNA methylation. Our present findings raise the possibility that the same may be true for human brain and this is a subject for future study.
P057. Dopamine D1, D2 and D3 receptor densities in the striatal and extra-striatal regions of the aged healthy human brain
Washington University School of Medicine, St. Louis, Missouri, USA
Background
Imaging of D2 versus D3 and of D3 versus D2 receptors in living brain has been hampered by the lack of selective D2 or D3 ligands. The dopamine receptor antagonists [11C]raclopride and [18F]fallypride and the full agonist, [11C](+)-PHNO, bind with similar affinity to D2 and D3 receptors and typically measure D2/D3 receptor binding potentials. The benzamide analog, 4-(dimethylamino)-N-(4-(4-(2-methoxyphenyl)piperazin-1-yl)butyl)benzamide (
Methods
In this study, we measured the D2 and D3 receptor densities in the striatal regions (caudate and putamen) and in extra-striatal regions (thalamus and globus pallidus) of 11 aged healthy human brains (ranged from 77 to 107 years old) using quantitative autoradiography with the D2/D3 mathematical model developed in our group.
Results
We found the regional density (fmol/mg tissue) of dopamine D2 and D3 receptors respectively were: precommissural caudate: 79 and 38; postcommissural caudate: 68 and 28; precommissural putamen: 82 and 39; postcommissural putamen: 76 and 38; nucleus accumbens: 93 and 55; globus pallidus external part: 18 and 18; globus pallidus internal part: 9 and 21; thalamus: 3 and 33. We also determined the dopamine D1 receptor densities in those striatal and extra-striatal regions using [3H]SCH23390 which is selective for the dopamine D1 receptor; serotonin 5-HT2 receptors were blocked with the ketanserin. We found specific dopamine D1 receptor binding (fmol/mg tissue) was highest in nucleus accumbens: 126 followed in descending order by precommissural caudate and putamen: 115 and 116; and postcommissural caudate and putamen: 86 and 90. Extra-striatal D1 receptor binding was significantly lower than the binding in the D1 receptor-rich striatal regions, the extra-striatal thalamus: 14; globus pallidus external part: 16; globus pallidus internal part: 29. There was a strong linear correlation (R2 > 0.78) between the average dopamine D1 and D3 receptor densities in the thalamus, while the dopamine D2 receptor density (3 fmol/mg tissue) appeared to be negligible.
Conclusions
Our systematic measurement of dopamine receptor subtypes in the striatal and extra-striatal regions will assist ongoing validation of PET radiotracers selective for D2 (versus D3) and D3 (versus D2) imaging of dopamine receptor subtypes in the central nervous system.
Acknowledgements
Research funded by NIH grants MH081281, DA29840 and P50 AG05681.
P058. Evaluating NA release with C-11 yohimbine: combined in vivo PET-microdialysis studies
1University of British Columbia, Vancouver, Canada; 2Aarhus University, Denmark; 3Aarhus PET Center, Denmark; 4Lundbeck A/S, Denmark, Copenhagen, Denmark; 5Aarhus Univesity Hospital, Denmark
Background
Noradrenaline plays a crucial role in multiple brain functions and in numerous neurological disorders. Yet, little remains known of its regulation and responses to behavioral and pharmacological challenges. We recently reported that the binding of [11C]yohimbine, a selective antagonist tracer of the alpha 2 receptors in tracer dose, is decreased following amphetamine challenge. We hypothesized that this change reflects competition with amphetamine-induced release of NA. In the current study, we performed combined microdialysis/PET studies to assess concurrently yohimbine binding and the extracellular concentrations of NA and DA in regions of high (cortex and thalamus) NA innervation and low NA/high DA innervation (striatum).
Methods
Isofluorane-anesthetized Gottingen minipigs were positioned in a PET-compatible stereotaxic headholder and a high resolution CT performed in a Siemens clinical PET/CT. Microdialysis probes (CMA70) were placed stereotaxically in thalamus, striatum and cortex and perfused with artificial CSF (2 μl/min). Samples were collected every 10 min throughout the course of the experiment and immediately frozen until assay. After a 2–3 hrs equilibrium period, three 90 min [11C]Yohimbine (200–300 MBq in 10 mL, injected mass: <1 μg per scan) scans were acquired, at baseline and following a pharmacological intervention (amphetamine (1–2 mg/kg), a non specific NA/DA releaser or nisoxetine (1 mg/kg), a specific NET inhibitor) at 30 and 180 min after challenge. Vital signs were monitored throughout the course of the study. The animal was humanely euthanized at the end of the experiment to verify probe integrity and location. Samples were analyzed with HPLC for NA and DA and their metabolites. Yohimbine total distribution volume (DVT) were obtained from thalamus, striatum and several cortical regions as previously described using a standard arterial plasma input function.
Results
Both pharmacological challenges induced a significant decrease in yohimbine binding, presumably from competition by the endogenous ligand: cortical and thalamic regions showed the greatest decrease (>20%) while the striatum had a more moderate decrease (8–15%) in amphetamine-treated pigs. Nisoxetine induced a similar decrease (about 20%) in all brain regions. Dialysis samples revealed significant increases in NA extracellular concentrations after both challenges. DA was also significantly increased after amphetamine challenge, more specifically in striatum. Comparably to raclopride, yohimbine binding did not return to baseline in the 2nd post amphetamine challenge scan despite return of the extracellular DA and NA concentrations towards baseline.
Conculsions
This data suggest that [11C]yohimbine may be a potential tracer to evaluate acute variations in NA concentrations after acute pharmacological challenges. This study, showing a similar behavior of binding of yohimbine and raclopride in response to amphetamine, also suggests that, like raclopride, yohimbine may reflect intrasynaptic concentrations of the ligand.
P059. Occupancy of naltrexone at kappa opioid receptors may predict efficacy in reducing drinking in alcoholics
1Yale University, New Haven, Connecticut, USA; 2University of Glasgow, UK
Background
Naltrexone (NTX), a non-specific opioid antagonist that binds dose-dependently to Mu, Delta, and Kappa opioid receptors, has been shown to be efficacious in the treatment of alcohol dependence. The fraction of drinkers who respond favorably to NTX treatment is small to moderate but for those who respond, the drug helps them to reduce their drinking. We recently introduced a new PET tracer, [11C]LY2795050, which is selective for the kappa opioid receptor (KOR). (See Kim et al., abstract for selectivity data.) The goal of this study is to further our understanding of how NTX reduces drinking and specifically to examine the kappa-mediated component of NTX efficacy in non-treatment seeking, alcohol-dependent, heavy drinkers (age range: 24–47).
Methods
Two [11C]LY2795050 scans, baseline and post-NTX, were performed on 5 male drinkers. Each scan was 2 hours long on the Siemens HRRT following a bolus injection of tracer (9 mCi). Occupancy of KOR by NTX was calculated by comparing the volumes of distribution (VT) between the two scans via modified Lassen plots (Cunningham, 2010). The baseline and post-NTX scans were separated by 6–11 days of NTX treatment which ramped up over the first 3 days to a maintenance dose of 100 mg/day thereafter. Subjects also participated in two alcohol drinking paradigms (ADP; following Krishnan-Sarin, 2007) conducted on the days immediately prior to the baseline scan and following the post-NTX scan. Each ADP began with a priming drink followed by three one-hour self-administration periods; during each period subjects were given the choice between a total of 4 alcoholic drinks or $12 (each drink being worth $3). Occupancy of NTX was plotted vs fractional change in number of drinks consumed during the baseline to the post-NTX ADP.
Results
The mean absolute reduction in drinks from before to after NTX was 3.2 and the mean percentage reduction was 79%. In our first 5 subjects, there was a non-significant correlation (R=0.74) between occupancy of KOR by NTX and reduction in drinks from baseline to post-NTX session. There was no correlation between baseline VT and fractional reduction in drinking.
Percent reduction in drinks from baseline to post-NTX vs occupancy of NTX at KOR.
Conclusions
NTX helps some alcoholics to reduce their drinking. The occupancies of NTX at the Delta and Mu Opioid Receptors have been investigated previously with 11C-Me-naltrindol and 11C-carfentanil, respectively. In the case of the Mu site, an acute dose of 100 mg of NTX occupied a mean of 94.9% of Mu receptors across a cohort of recently detoxified alcoholics (Weerts, 2008). Given the very high occupancy of MOR by NTX and the minimal variability across subjects, it is unlikely that occupancy at MOR is the primary mediator of differential efficacy of NTX – which varies widely. Thanks to the recent development of a highly selective KOR antagonist tracer (Zheng, submitted) and a well-established laboratory test for treatment efficacy, we have begun investigating the contribution of kappa to the efficacy of NTX in the hope that this investigation will help guide development of future pharmacotherapies.
P060. Effects of amphetamine-induced dopamine release and raclopride-induced dopamine blockade on psychiatric symptoms and tics of subjects with and without Gilles de la Tourette syndrome
Johns Hopkins University School of Medicine, Baltimore, Maryland, USA
Background
In prior reports subjects with Gilles de la Tourette syndrome (TS) demonstrated elevated striatal dopamine release (DAR) and more psychiatric symptoms [Brief Psychiatric Rating Scale (BPRS)] after the administration of intravenous amphetamine. To improve the accuracy of the measurements throughout each scan, this study, using the more comprehensive “area under the curve (AUC)” representation of behavior and movement ratings, compares the effects of amphetamine (increases DAR) and raclopride-induced dopamine blockade (DAB) on tics and behaviors in subjects with and without TS. We developed DAR and DAB change variables for each rating score represented as ratios similar to DAR computed as {[nondisplaceable binding potential (BPND) after saline] – [BPND after amphetamine]}/{BPND after saline}. It is hypothesized, based on proposed abnormalities of dopaminergic transmission, that subjects with TS will have greater behavioral and movement effects following DAR and DAB.
Methods
24 subjects (12 TS, 12 controls) underwent 3 PET scans: 1) low-specific activity (LSA) [11C]raclopride to obtain a quantitative estimate of partial occupancy of dopamine D2/3 receptors in the brain; 2) high-specific activity (HSA) [11C]raclopride preceded by 5 min with an injection of 10 mL 0.9% NaCl (marker for resting tonic dopamine D2/3 receptor state); and 3) HSA [11C]raclopride preceded by 5 min with an injection of 0.3 mg/kg amphetamine (simulates the excited phasic state). Ratings of stimulant effects, motor and phonic tics on the Yale Global Tic Severity Scale (YGTSS), and psychiatric symptoms (BPRS) were performed at baseline and during each ten-minute interval of each ninety-minute scan. In contrast to prior studies, which utilized the difference between the highest scores of the phasic and tonic states to compute DAR change variable, this report estimated the area under the curve (AUC) to compute the DAR and DAB change variables. The
Results
Both subjects with and without TS exhibit a negative or zero change variable for all ratings under all conditions. Subjects without TS exhibit a greater negative magnitude of change variables on the Stimulant Effects Scale and Motor and Phonic Tic Scores than the subjects with TS in response to both DAR and DAB. These differences are significant under DAB.
Conclusions
Subjects with TS may have (A) higher baseline tonic levels of stimulant effects and tics and (B) consequently lesser relative phasic levels of stimulant effects and tics than subjects without TS. The relationships of behaviors and tics to intrasynaptic and extrasynaptic dopamine in DAR and DAB need clarification. Striatal dopaminergic dysfunction contributes to the symptoms of subjects with TS and OCD in response to dopamine release and dopamine blockade.
Acknowledgements
Research Support provided by PHS Grants MH078175, NS38927, DA00412, AA12839, GCRC (NIH/NCRR M01RR00052), The Brain and Behavior Research Foundation (NARSAD), Essel Foundation, Tourette Syndrome Association.
P061. Modulation of brain activity by DOPAergic volume transmission
1University of Copenhagen, Denmark; 2Aarhus University, Denmark; 3University of Oslo, Norway
Background
We present the novel perspective of DOPA as a volume transmitter (Bergersen & Gjedde 2012) that regulates brain activity. Metabolites are volume transmitters of regulatory information when specific criteria are fulfilled in terms of stereoselective interactions with proteins, including unlimited passage across cell membranes by means of facilitated diffusion, unlimited distribution in brain tissue, and interaction with enzymes or receptors. DOPA fulfills these criteria as volume transmitter of information that redistributes monoaminergic activity in brain tissue (Goshima et al. 1986; Ng et al. 1970).
Results
This means that all the noradrenergic and serotonergic neurons of the cortex, and the serotonergic and dopaminergic neurons of the striatum, produce dopamine in proportion to the local DOPA concentration, as shown in the Figure.
DOPA diffuses among at least 5 different cell types with TH or DDC activity, or both, in which it is now clear that DOPA is synthesized, metabolized, or both. By this action, DOPA distributes and adjusts the generation of the DA, NA, and 5HT modulators across large volumes of brain tissue.
Conclusions
Treatment of Parkinson's disease with DOPA's provides the most emphatic evidence of volume transmission with additional effects that are independent of DOPA's conventional function as a precursor of dopamine and noradrenaline in the respective neurons. Cells that contain only TH appear in the striatum and elsewhere in parkinsonism where they provide additional DOPA (Darmopil et al. 2008). Endo- and exogenous DOPA supports the synthesis in, and release from, serotonergic and noradrenergic neurons of dopamine, in relation to the appearance of dyskinesia in an animal model of parkinsonism (Nahimi et al. 2012).
P062. PET imaging of D2 receptor in Rett syndrome and in Mecp2-deficient mice
1Kennedy Krieger Institute, Baltimore, Maryland, USA; 2Johns Hopkins University, Baltimore, Maryland, USA; 3University of Copenhagen, Denmark
Background
Rett Syndrome (RTT) is a developmental disorder caused by mutations in the MECP 2 gene. Motor findings include stereotyped limb movements, dystonia, dyskinesias, and progressive rigidity. Since dopamine (DA) has been implicated in motor abnormalities, we focus on the role of DA in RTT.
Methods
To determine what role alterations in DA plays in these motor alterations, we characterized the expression of DA D2 receptors and the (DA) transporter (DAT) in patients with RTT and in the Bird mouse model of Mecp2 deficiency. For PET imaging of D2 receptors, PET scans with [11C]-3N-methylspiperone, ±haloperidol (for Bmax) was used for women with RTT and age-matched controls, and [11C]Raclopride was used for wildtype (WT), Mecp2-null, and Mecp2-heterozygous (HET) mice. For DAT, [11C] WIN35,428 was used for human studies and [11C] Methylphenidate) for mouse studies.
Results
The results show that the density of D2 receptors is decreased in women with RTT, and in young (7–10 weeks) HET and Mecp2-null mice compared to age-matched controls and WT mice. Moreover, D2 receptor Bmax declined with age in human controls as expected (Wong et. al 1984, 1986) and WT mice, but not in women with RTT, or HET mice. PET studies of DAT did not reveal significant differences in binding densities in women with RTT or HET mice. However, Mecp2-null mice did have significantly lower densities of DAT, indicating that the absence of Mecp2 affects the expression of the DAT. Unlike D2 receptors DAT did not show age–related changes in expression.
Conclusions
The results show that Mecp2-deficiency, or loss, or both, affect DA neurotransmission. Most studies revealed that levels of DA are decreased in RTT. Thus, the decreases in D2 receptor expression and lack of a change in DAT suggest dysfunction of this neurotransmitter system. DA D2 receptors normally facilitate movement by inhibiting the indirect pathway in the basal ganglia. Decreased expression of DA D2 receptors would lead to inhibition of movement and may account for the rigidity seen in RTT. Taken together, our findings suggest that dopaminergic mechanisms play a role in the motor abnormalities observed in patients with RTT. These results also suggest that DA D2 receptors may in part be a target for pharmacologic therapy in RTT.
P063. Chronic L-DOPA treatment prevents increased binding of [11C]MDL100907 to 5-HT2A receptors in a rat model of Parkinson's disease
1Department of Nuclear Medicine and PET Centre, Aarhus University Hospitals, Aarhus University, Denmark; 2Department of Neuroscience and Pharmacology, University of Copenhagen, Denmark; 3Center of Functionally Integrative Neuroscience, Aarhus University Hospitals, Aarhus University, Denmark
Background
6-hydroxydopamine (6-OHDA) lesions increase 5-HT2A receptor mRNA which is preventable by treatment with high doses of L-3,4 dihydroxyphenylalanine (L-DOPA) (Zhang et al, 2007). However, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced depletion of dopamine (DA) in macaques does not increase 5-HT2A receptors, and chronic treatment with L-DOPA actually increases 5-HT2A expression in this animal model (Huot et al, 2012). Further studies are needed to address these conflicting findings of the role of 5-HT2A−receptors in the DA depleted brain. [11C]MDL100907, a highly selective 5-HT2A−receptor antagonist was used in post-mortem autoradiography experiments to test if the increase in 5-HT2A receptor mRNA results in increased expression of the receptor itself and if the predicted elevation of the receptors can be prevented by chronic treatment with low doses of L-DOPA.
Methods
Sprague Dawley rats were unilaterally lesioned with 6-OHDA and were treated with either L-DOPA (8 mg/kg) and benzeraside (15 mg/kg) or saline for 5 weeks. Autoradiography on fresh frozen brains was performed as previously described in detail by (Strome et al, 2005 and Lopez-Gimenez, et al, 1997). Total binding was assessed by incubating 20 mm coronal slides in 0.64 nM [11C]MDL100907 in a 50 mM Tris HCL buffer (pH=7.4) for 30 minutes at 25–30 degrees Celsius.
Results
One-way ANOVA showed a significant difference between means (P=0.0097) and Neuman-Keuls post-test showed significant difference between the non-lesioned hemisphere of L-DOPA or saline treated rats compared to the lesioned hemisphere of saline treated rats (see Figures 1 and 2). [11C]MDL 100907 binding in the lesioned hemisphere was reduced from 18.34 PSL/area in saline treated rats to 15.31 PSL/area in L-DOPA treated rats but this did not reach statistical significance.
Coronal sections of unilateral 6-OHDA lesioned rat brain labelled with [11C]MDL 100907. The right side with increased binding of [11C]MDL 100907 indicates the hemisphere of dopamine depletion.
The effect of DA depletion on the expression of 5-HT2A receptors after saline of L-DOPA treatment. Image intensity is expressed in photon-stimulated luminescence per area (PSL/AREA).
Conclusions
The finding that 5-HT2A receptor expression is increased in the DA depleted striatum is consistent with the finding of increased 5-HT2A receptor mRNA in 6-OHDA-lesioned animals. The prevention of part of this increase by L-DOPA treatment suggests an inhibitory role of DA in the expression of 5-HT2A receptors, presumably by down regulation of 5-HT2A receptor mRNA synthesis.
P064. The ventral striatum shows initial preservation of dopamine transporters but indistinguishable progression rates of degeneration in Parkinson's disease as measured with [11C]Methylphenidate PET
1Radiology, Johns Hopkins University, Baltimore, Maryland, USA; 2Parkinson's Disease Research Unit, Thomas Jefferson University, Philadelphia, Pennsylvania, USA; 3Neurology, Johns Hopkins University, Baltimore, Maryland, USA
Background
Previously we reported sparing of the ventral striatum (vS) in idiopathic Parkinson's disease (PD) at the time of diagnosis (JNM 46:127P) and suggested that this subdivision may have a different susceptibility for the dopamine denervation in PD. This study aims to confirm the findings with larger sample sizes and age-matched control subjects, and compare the rates of disease progression in vS against other striatum subdivisions during relatively early stages of the disease using [11C]Methylphenidate (MP), a marker of the dopamine transporter (DAT).
Methods
Patients with idiopathic PD (n=39; age: 59.5±9.3 years, time since diagnosis of approximately 3 years) and healthy control subjects (HC; n=23; age: 60.5±8.6 years) were studied with PET for 90 min following a bolus injection of [11C]MP. The unified Parkinson's disease rating scale (UPDRS) motor score was 18.4±4.4 and Hoehn and Yahr stage was 2.2±0.3 for PD subjects. A subset of PD patients with standard of care treatment (i.e., levodopa±dopamine agonist) completed [11C]MP scans at both 2 and 3 years from the initial scan (n=6). Volumes of interest (VOIs) were defined on MRI for the caudate nucleus (CN), putamen (Pu), and cerebellum (Cb) manually and divided into limbic (vS), associative (anterior putamen, aPu, and anterior and posterior caudate nucleus, aCN and pCN), and motor (posterior putamen, pPu). Regional binding potential (BPND) was obtained by the multilinear simplified reference tissue method (JCBFM 22:1271) using Cb as the reference region.
Results
Regional BPND values of PD and HC subjects are listed in the table for the five subdivisions (mean±SD). As reported repeatedly, posterior putamen (pPu) showed the most severe loss of [11C]MP binding compared to HC. [11C]MP BPND was most preserved in vS in PD (i.e., 75% of the control value). BPND decreased linearly over the three year observation period (R2>0.997) at the rates listed in the Table, except for pPu which showed almost no change during the period (R2=0.44). Despite relative sparing of vS early in the course of the disease, vS showed indistinguishable rates of progression compared to aPu, aCN, and pCN.
Conclusions
This study demonstrated a sparing of ventral striatal DAT binding in PD early in the disease. However, vS displayed a rate of progression that was indistinguishable from other striatal subdivisions (except for pPu) over a three year observation period. Therefore, even though there appears to be early sparing of vS dopamine terminals using a DAT biomarker, the rate of loss of DAT binding appears to not be different from other striatal subregions as the disease progresses.
P065. Dopamine release and serotonin interactions in Tourette's syndrome: new relations to obsessive compulsive disorder
1Johns Hopkins University, Baltimore, Maryland, USA; 2University of Pittsburgh, Pennsylvania, USA; 3University of Copenhagen, Denmark
Background
Previous work from the lab (Wong, et al, Neuropsychopharmacology: 33(6), 2008) revealed changes of dopamine release (DArel) and densities of monoamine receptors/transporters in Tourette's syndrome with or without Obsessive Compulsive Disorder (TS+/−OCD). We expanded the number of healthy controls (HC) and patients studied, including an OCD-only comparison group. Specifically, we determined striatal DArel, dopamine D2/D3, 5HT2A, dopamine transporter (from previous work) and cortical serotonin transporter (SERT) across 4 groups, healthy controls (HC), OCD and TS +/− OCD.
Methods
For dopamine, we studied 11 HC (5 females, 6 males, age (mean +/− SD) 27.4 +/−7.8), 15 TS (3 females, 12 males, age 25.8 +/− 3.1), and 9 TS + OCD (2 females, 7 males, age 37.0 +/− 14.3) and 12 OCD only (5 females, 7 males, age 43.31 +/− 15.1) subjects. The serotonin studies included 10 HC (4 females, 6 males, age 32.9 +/− 8.04), 5 TS (1 female, 4 males, age 28.8+/− 5.8), 12 TS + OCD (4 females, 8 males, age 33.71 +/− 10) subjects. A subset of 7 TS subjects completed both dopamine and serotonin PETs within 1 year. To calculate DArel, 2 PETs were performed with 11C raclopride (RAC), preceded by IV saline or 0.3 mg/kg amphetamine. We used McN5652 (MCN) for SERT and 11C MDL 100,907 (MDL) PET for 5HT2A BPND. DArel was calculated with SRTM and SERT and 5HT2A BPND were estimated using parametric methods (Zhou, et al., 2003). MANOVA was followed by post-hoc comparisons (Bonferroni corrected).
Main panel: DArel across all regions, by group. ∗Indicates a sig (p < 0.05) increase vs. controls, ∗∗indicates a sig. increase over controls and TS. Insert: DArel vs. tic severity in TS +/− OCD subjects. Inset: DArel increased with tic severity.
Results
MANOVA analysis of DArel showed that diagnostic group was a significant source of variance (p=0.01), while region was not. An inhibition plot analysis confirmed that the change in RAC BPND was not different across regions but that TS differed from other groups (p=.019). Post-hoc comparisons shows: for the DArel, we found TS > HC in the posterior caudate (PC) (P<0.05), and ventral striatum (VS) (p<0.05), both bilaterally. For TS+OCD, DArel was > HC in right VS (P<0.05). For DArel in the right VS, we found TS+OCD > TS-OCD (p<0.05). OCD only was >HC and TS+OCD in right PC (p<0.05). In temporal, frontal, and parietal cortices, TS+OCD SERT BPND >HC (p<0.05) and also > TS (p<0.05). In the subset of 7 TS+/−OCD who underwent both DA and 5HT PETs, we observed two (0.1>p<0.5) trends: DArel was approximately inversely proportional to SERT BPND (r =−0.5), and directly proportional to 5HT2A BPND (r=0.4). We also found significant correlation of overall tic severity with DArel in the right PC (r=0.76, p=0.007) and left VS (r=0.59, p=0.04), consistent with increased phasic DArel in proportion to tics.
Conclusions
These results expand on previous work and are not inconsistent with the hypothesis that TS+/−OCD have lower SERT density in the cortex and higher DArel than HC, specifically in the posterior caudate and ventral striatum. In addition, preliminary results suggest a relationship between DA and 5HT in TS and OCD.
P066. Voxel level concordance-discordance analysis of [11C]PiB and Florbetapir PET
University of Pittsburgh, Pennsylvania, USA, Departments of: 1Radiology; 2Biostatistics; 3Psychiatry
Background
The objective was to apply a non-parametric multimodal correlation test to explore voxel cross-modality relationships (concordance and discordance) between [11C]PiB (or PiB) and Florbetapir PET.
Methods
The ADNI database supplied PiB and Florbetapir PET images acquired for: A) 22 subjects (6 NC, 16 MCI) with a baseline PiB scan and a subsequent baseline Florbetapir scan (3.0±0.3 years apart) and B) 22 subjects (6 NC, 16 MCI) with a PiB scan acquired before Florbetapir scan (1.4±0.6 years apart). Sixteen subjects in dataset B were members of dataset A, but none of the 44 images were identical. Standardized uptake ratio (SUVR) images were generated over 50–70 min post-injection using the cerebellum as reference. PET SUVR images were spatially normalized to a MR template using the individual's structural MRI. Subject amyloid retention status (amyloid(+) vs. amyloid(−)) for both PiB and Florbetapir data was based on regional iterative outlier analyses of the PiB data. Voxel-wise 2-sample t-tests were performed to generate group difference t-maps for both PiB and Florbetapir (Contrast: amyloid(+) > amyloid(−)). The degree of concordance-discordance was summarized using a combining t-value function within a permutation test framework [1]. Clusters of concordance-discordance were projected into normalized space and mapped to anatomical locations (p<0.05 FWE corrected). Concordance reflects significant t-map differences for both PiB and Florbetapir. Discordance reflects significant PiB group differences without Florbetapir differences, and vice versa.
Results
Concordance was significant in high amyloid retention areas of cortex and was consistent for the short-term and long-term scan interval, with a greater extent of concordance for the shortest interval of about 1 year. No concordance was found in white matter regions. Discordance was observed at gray/white matter boundaries for the short-term scan interval and primarily in white matter areas for the long-term scan interval.
Conclusions
Group differences in retention of the amyloid imaging agent for the amyloid(+) vs. amyloid(−) subjects were highly concordant in amyloid-bearing areas of cortex for short and long-term intervals between PiB and Florbetapir scanning, despite brain changes that may arise during the interval (e.g., atrophy, amyloid accumulation), that likely contribute to discordance.
P067. Amphetamine-induced inhibition of [11C]yohimbine binding in rat brain
1Department of Nuclear Medicine and PET Centre, Aarhus University Hospitals, Aarhus University, Denmark; 2Department of Neuroscience and Pharmacology, University of Copenhagen, Denmark; 3Department of Radiology and Radiological Science, Division of Nuclear Medicine, Johns Hopkins Medical Institutions, Baltimore, Maryland, USA
Background
Yohimbine is a selective alpha-2 adrenoceptor antagonist, which can be labeled with carbon-11 and used as a positron emission tomography tracer (Jakobsen et al. 2006). Amphetamine is an inhibitor of dopamine and norepinephrine reuptake sites in mammalian brain. The administration of amphetamine can raise extracellular concentrations of dopamine and norepinephrine at cortical and subcortical sites. Yohimbine does not cross the blood-brain barrier of the rat appreciably, however the penetration of the blood-brain barrier can be enhanced with cyclosporine A, an inhibitor of p-glycoprotein. To test the hypothesis that amphetamine administration can displace binding of [11C]yohimbine to rat alpha-2 adrenoceptors, we administered cyclosporine to increase blood-brain barrier permeability and used rodent PET to determine yohimbine binding before and after amphetamine administration.
Methods
This pilot study included three female Sprague Dawley rats (225–250 g), anaesthetized by isoflurane inhalation and pretreated with cyclosporine A (1 mg/kg i.v.). Dual PET sessions with [11C]yohimbine were performed at baseline and after amphetamine challenge (2 mg/kg i.v). Each PET session included i.v. bolus injection of [11C]yohimbine (20–40 MBq) followed by a 90 minute emission recording in 25 frames. Nine arterial blood samples were drawn from the femoral artery to obtain the plasma activity input function. Cerebral cortex, striatum and cerebellum were chosen as volumes of interest (VOI) after co-registration of PET to an MRI atlas of the rat bran. Volumes of distribution were acquired by Logan graphical analysis (Logan et al., 1990) of data from the interval 40–90 min. The post-amphetamine volumes VT(inhibition), were plotted as function of baseline volumes VT(baseline) for all VOI. We then applied the inhibition plot (Gjedde and Wong, 2000),
in which s is the inhibitor's occupancy and VND is the volume of non-displaceable tracer, both of which are estimated by the linear regression of VT(inhibition) vs. VT(baseline). The estimate of VND was then used to calculate the binding potentials at baseline, BPND, and inhibition, BPND.
Results
The upper panel depicts the volume of distribution, which declined approximately 45% in all VOIs after amphetamine administration, with a reference VND estimated at 0.234 mL/g. The lower panel illustrates the binding potentials at baseline (black bars), which decreased similarly in all 3 regions after amphetamine (grey bars).
Conclusions
After permeabilization of the blood-brain barrier of the rat, the results show that [11C]yohimbine binding to alpha-2 adrenoceptors can be measured with microPET. Furthermore, acute amphetamine treatment reveals both the sensitivity of yohimbine binding to competition from endogenous monoamines and the sensitivity of monoamine accumulation to amphetamine's inhibition of the norepinephrine and perhaps dopamine reuptake sites.