Relationship of Age and Synaptic Density:A ¹¹C‑UCB‑J PET Study in Healthy Controls with Partial Volume Correction

Introduction: ¹¹C-UCB-J is a PET ligand for synaptic vesicle glycoprotein 2A (SV2A) which may be a quantitative in vivo synaptic density marker (1-3). In an initial evaluation of age effects (4), without performing partial volume correction (PVC), subcortical synaptic density (BP_ND) was found decreasing due to age effects. Here, we applied PVC to ¹¹C-UCB-J PET in healthy controls (HC), which aimed to separate the age effects of age on synaptic density and volume loss. Materials and Methods: A cohort of 72 HCs underwent ¹¹C‑UCB‑J PET with the high resolution research tomograph (HRRT; dose: 14.5±4.8 mCi). The age range was 21 to 83 years (y) with 16 elderly subjects (> 60 y). Individual T1 weighted MR images were analyzed by FreeSurfer (FS) to segment the gray matter (GM) into 85 regions of interests (ROIs). Iterative Yang (IY) (4) PVC was performed on each dynamic frame. This algorithm corrects the original image frame-by-frame in a voxelwise manner based on the spill-in and spill-out for each FS ROI, assuming a Gaussian point spread function (PSF) of 3.0 mm in full-width-half-max. Arterial samples were used to measure the metabolite-corrected input function and plasma free fraction (f_P). V_T images were produced with the one-tissue compartment model. The FS ROIs were merged into 12 GM ROIs: frontal, occipital, parietal, and temporal cortices, anterior and posterior cingulate cortices, hippocampus, precuneus, caudate, putamen, thalamus, cerebellum. The optimized centrum semiovale (CS) ROI in AAL (5) was applied to individual MR spaces and BP_ND was calculated with the CS as reference region. Age effects in V_T, K₁, BP_ND, and ROI volume were evaluated with regional correlations and t tests without multiple comparison correction.

Results: Tissue volume was significantly decreased with age in all GM ROIs except occipital cortex and cerebellum. Frontal cortex showed the most significant and largest decrease (mean: 158 mL, 4.4% reduction per decade (RPD), p < 0.0001). The mean RPD of GM ROI volumes was 2.8% (range: 0.9 - 4.4%, 37% SD). Average GM V_T was 20.2 mL/cm³ (13% SD) without PVC and 24.0 mL/cm³ (13% SD) with PVC. Average GM K₁ was 0.37 mL/min/cm³ (14% SD) without PVC and 0.43 mL/min/cm³ (14% SD) with PVC. V_T reduction with age in CS was 2.2 % RPD without PVC (mean: 4.2 mL/cm³,16% SD, p = 0.04) and 2.0 % RPD with PVC (mean: 4.2 mL/min/cm³ ,16% SD, p = 0.06); there was no significant K₁ reduction in CS or f_P with age. There was a significant correlation between V_T with PVC in CS and f_P (r = 0.41, p < 0.001), and the age effects of V_T/f_P in CS were eliminated (without PVC: 1.4% RPD, p>0.10, with PVC: 1.2% RPD, p > 0.2). Thus, we attribute the age effect on V_T to changes in plasma free fraction and subsequently used V_T/f_P as the primary outcome measure. Average GM V_T/f_P was 72.7 mL/cm³ (13% SD) without PVC and 86.6 mL/cm³ (13% SD) with PVC. V_T/f_P in all GM regions showed significant reduction with age(without PVC: 2.2% PRD, p < 0.05, with PVC: 2.3% PRD, p <0.05) except anterior cingulate, hippocampus, and putamen, Caudate showed the most significant reductions with age (without PVC: 3.8% RPD, p < 0.001, with PVC: 3.7% RPD, p < 0.001). The volume reduction in caudate was 3.8% RPD (p < 0.0005). V_T/f_P with or without PVCresults are summarized in the table as mean±SD (% RPD).

Conclusions: Age-related synaptic density decreases were found in frontal, occipital, temporal, and posterior cingulate cortices, caudate, thalamus, and persisted after PVC, which corrected for age-related tissue volume reduction. The SV2A/synaptic density reduction in the caudate appears to be larger than in other brain regions. References 1. Nabulsi et al., JNM, 2016. 2. Finnema et al., Sci Transl Med, 2016. 3. Finnema et al., JCBFM, epub, 2017. 4. Carson et al., SNMMI 2018. 5. Toyonaga et al., SNMMI 2018.