PET Quantification of [18F]VAT in Human Brain and Its Test-Retest Reproducibility and Age Dependence

Abstract

Molecular imaging of brain vesicular acetylcholine transporter provides a biomarker to explore cholinergic systems in humans. We aimed to characterize the distribution of, and optimize methods to quantify, the vesicular acetylcholine transporter-specific tracer (-)-(1-(8-(2-[¹⁸F]fluoroethoxy)-3-hydroxy-1,2,3,4-tetrahydronaphthalen-2-yl)-piperidin-4-yl)(4-fluorophenyl)methanone ([¹⁸F]VAT) in the brain using PET. Methods: Fifty-two healthy participants aged 21-97 y had brain PET with [¹⁸F]VAT. [³H]VAT autoradiography identified brain areas devoid of specific binding in cortical white matter. PET image-based white matter reference region size, model start time, and duration were optimized for calculations of Logan nondisplaceable binding potential (BP_ND). Ten participants had 2 scans to determine test-retest variability. Finally, we analyzed age-dependent differences in participants. Results: [¹⁸F]VAT was widely distributed in the brain, with high striatal, thalamic, amygdala, hippocampal, cerebellar vermis, and regionally specific uptake in the cerebral cortex. [³H]VAT autoradiography-specific binding and PET [¹⁸F]VAT uptake were low in white matter. [¹⁸F]VAT SUVs in the white matter reference region correlated with age, requiring stringent erosion parameters. Logan BP_ND estimates stabilized using at least 40 min of data starting 25 min after injection. Test-retest variability had excellent reproducibility and reliability in repeat BP_ND calculations for 10 participants (putamen, 6.8%; r > 0.93). We observed age-dependent decreases in the caudate and putamen (multiple comparisons corrected) and in numerous cortical regions. Finally, we provide power tables to indicate potential mean differences that can be detected between 2 groups of participants. Conclusion: These results validate a reference region for BP_ND calculations and demonstrate the viability, reproducibility, and utility of using the [¹⁸F]VAT tracer in humans to quantify cholinergic pathways.

Keywords: PET; acetylcholine; brain; cholinergic; human.

Graphical abstract

FIGURE 1.

[¹⁸F]VAT distribution in brain. (A–C) Sample images of composite average of SUV-parametrized images from 52 participants. Scales are varied to highlight cortical and subcortical distributions. Scale bars = 1 cm. (D) Average decay-corrected time–activity curves from select regions across 22 participants. Error bars indicate mean ± SEM.

FIGURE 2.

Representative control tissue [³H]VAT autoradiography from cortical white matter (A) and caudate (B). Images are presented adjacent to slices with 5 nM [³H]VAT + 10 μM vesamicol to demonstrate selective [³H]VAT binding. 1 nCi = 37 Bq.

FIGURE 3.

Seven erosion widths and 3 thresholds were used to generate potential WM_REF values. Mean region-spread function partial volume–corrected SUVs (A) and intraregional SEM (B) are shown to demonstrate region characteristics. (C) Reference region correlation with age.

FIGURE 4.

Effects of varying model start time and duration on Logan BP_ND. Logan binding potentials were averaged across all participants calculated for cortical gray matter (A) and putamen (B). Start times of 45 and 65 min are offset to demonstrate results ending at same time point. Error bars indicate mean ± SEM. Start times are color-coded as 25 min (gray), 45 min (blue), and 65 min (red). (C) Representative Logan plot, with regression line for 30- to 120-min model calculation shown in red. C_T(t) = activity in target tissue at time t; C_REF = activity in reference region at time t; k₂′ = population average reference region efflux constant.

FIGURE 5.

Representative Logan BP_ND demonstrating test–retest reproducibility of [¹⁸F]VAT. Data are putamen Logan BP_ND for each of 10 participants scanned 1 mo apart.

FIGURE 6.

Raw Logan BP_ND relative to age in putamen. Sample regression line is shown in red.