Assessing Synaptic Density in Alzheimer Disease With Synaptic Vesicle Glycoprotein 2A Positron Emission Tomographic Imaging

Abstract

Importance: Synaptic loss is well established as the major structural correlate of cognitive impairment in Alzheimer disease (AD). The ability to measure synaptic density in vivo could accelerate the development of disease-modifying treatments for AD. Synaptic vesicle glycoprotein 2A is an essential vesicle membrane protein expressed in virtually all synapses and could serve as a suitable target for synaptic density.

Objective: To compare hippocampal synaptic vesicle glycoprotein 2A (SV2A) binding in participants with AD and cognitively normal participants using positron emission tomographic (PET) imaging.

Design, setting, and participants: This cross-sectional study recruited 10 participants with AD and 11 participants who were cognitively normal between November 2015 and June 2017. We hypothesized a reduction in hippocampal SV2A binding in AD, based on the early degeneration of entorhinal cortical cell projections to the hippocampus (via the perforant path) and hippocampal SV2A reductions that had been observed in postmortem studies. Participants underwent high-resolution PET scanning with ((R)-1-((3-(11C-methyl-11C)pyridin-4-yl)methyl)-4-(3,4,5-trifluorophenyl)pyrrolidin-2-one), a compound more commonly known as 11C-UCB-J, for SV2A. They also underwent high-resolution PET scanning with carbon 11-labeled Pittsburgh Compound B (11C-PiB) for β-amyloid, magnetic resonance imaging, and cognitive and neurologic evaluation.

Main outcomes and measures: Outcomes were 11C-UCB-J-specific binding (binding potential [BPND]) via PET imaging in brain regions of interest in participants with AD and participants who were cognitively normal.

Results: Ten participants with AD (5 male and 5 female; mean [SD] age, 72.7 [6.3] years; 10 [100%] β-amyloid positive) were compared with 11 participants who were cognitively normal (5 male and 6 female; mean [SD] age, 72.9 [8.7] years; 11 [100%] β-amyloid negative). Participants with AD spanned the disease stages from amnestic mild cognitive impairment (n = 5) to mild dementia (n = 5). Participants with AD had significant reduction in hippocampal SV2A specific binding (41%) compared with cognitively normal participants, as assessed by 11C-UCB-J-PET BPND (cognitively normal participants: mean [SD] BPND, 1.47 [0.37]; participants with AD: 0.87 [0.50]; P = .005). These reductions remained significant after correction for atrophy (ie, partial volume correction; participants who were cognitively normal: mean [SD], 2.71 [0.46]; participants with AD: 2.15 [0.55]; P = .02). Hippocampal SV2A-specific binding BPND was correlated with a composite episodic memory score in the overall sample (R = 0.56; P = .01).

Conclusions and relevance: To our knowledge, this is the first study to investigate synaptic density in vivo in AD using 11C-UCB-J-PET imaging. This approach may provide a direct measure of synaptic density, and it therefore holds promise as an in vivo biomarker for AD and as an outcome measure for trials of disease-modifying therapies, particularly those targeted at the preservation and restoration of synapses.

Conflict of interest statement

Conflict of Interest Disclosures: Dr Chen reports research support from the Dana Foundation and research support from Eli Lilly and clinical trials from Merck, outside the submitted work. Dr Huang reports research grants from the UCB and Eli Lilly outside the submitted work. Drs Huang and Carson have a patent for a newer version of the tracer pending. Dr Carson reports grants from National Institutes of Health for the conduct of the study. Dr Carson also reports having received grants from AstraZeneca, Astellas, Eli Lilly, Pfizer, Taisho, and UCB, outside the submitted work. Dr van Dyck reports consulting fees from Kyowa Kirin, Roche, Merck, Eli Lilly, and Janssen and grants for clinical trials from Biogen, Novartis, Eli Lilly, Merck, Eisai, Janssen, Roche, Genentech, Toyama, and TauRx, outside the submitted work. No other disclosures are reported.

Figures

Figure 1.. Representative Total Uptake (VT) Parametric Synaptic Vesicle Glycoprotein 2A Positron Emission Tomographic (PET) Images

Representative coronal and axial images of parametric 11C-UCB-J–PET scans (VT) and magnetic resonance imaging scans in participants who were cognitively normal (A), participants with mild cognitive impairment (MCI) (B and C), and mild Alzheimer disease (AD) dementia (D and E). The cognitively normal participant had a negative carbon 11–labeled Pittsburgh Compound B (11C-PiB) scan, and all participants with AD had positive 11C-PiB–PET scans. The pseudocolor in PET images represents the intensity of 11C UCB-J binding (VT). Evident reduction of 11C-UCB-J binding in the hippocampus of participants with AD was noted compared with participants who were cognitively normal (the arrowhead denotes the right hippocampus). Various degrees of reduction can be visualized in the temporoparietal cortex of the participant with MCI (B and C), the right temporoparietal cortex of the participant with mild AD dementia (D), and the right temporal cortex of participant E. MCI indicates mild cognitive impairment.

Figure 2.. Voxelwise Analysis of Total Uptake (VT) and Perfusion (K1) With Statistical Parametric Mapping in Participants with Alzheimer Disease (AD) Compared With Cognitively Normal Participants

This Figure includes 9 participants with AD and 8 cognitively normal participants. A, Blue indicates the regions with statistically significant reduction of VT in participants with AD compared with those who were cognitively normal, rendered on brain surfaces at different projections. The statistically significant reduction of 11C-UCB-J binding in the hippocampus was consistent with the findings from region-of-interest analysis. B, Blue indicates a statistically significant reduction of K1 in the participant with AD compared with the cognitively normal participant. A broader regional reduction of K1 emerged in the temporoparietal cortices in the participant with AD.

Figure 3.. Comparison of Hippocampal Binding Potential (BPND) From Simplified Reference Tissue Model 2 (SRTM2) and Correlation With Clinical Findings

A, Comparison of BPND from SRTM2 in the hippocampus between participants with Alzheimer disease (AD) and cognitively normal participants. Error bars represent standard deviations. MCI indicates mild cognitive impairment. B, Association between hippocampal BPND from SRTM2 and episodic memory scores. A statistically significant correlation was observed between BPND in the hippocampus and an episodic memory score. C, Association between hippocampal BPND from SRTM2 and Clinical Dementia Rating–sum of boxes (CDR-SB) score.