Evaluation of [¹⁸F]SynVesT-2 for imaging SV2A in the human brain: kinetics, test-retest reproducibility, and binding specificity

Objectives: PET imaging agents for synaptic vesicle glycoprotein 2A (SV2A), such as [¹¹C]UCB-J, allowed for the in vivo detection of synaptic density changes, which are correlated with a variety of neuropsychiatric diseases. We recently translated a novel SV2A radioligand, [¹⁸F]SynVesT-2 to human studies. The aim of this study was to evaluate the test-retest reproducibility of this SV2A tracer and the extent of specific binding in the brain through a blocking study.

Methods: Five healthy volunteers were scanned on a High Resolution Research Tomograph (HRRT) scanner twice on different days, after injections of [¹⁸F]SynVesT-2. Two of the volunteers were also scanned following injection of levetiracetam (20 mg/kg, i.v.). Dynamic scans were 2-hours long with collection of arterial blood samples for measurements of the arterial input function (AIF) and plasma free fraction (f_p). Individual MR images were co-registered to a brain atlas to define the regions of interest (ROIs) for generation of time-activity curves (TACs), which were fitted with the 1-tissue compartment (1TC) model to derive regional distribution volumes (V_T). Regional nondisplaceable binding potential (BP_ND) was calculated from 1TC V_T values, using centrum semiovale (CS) as the reference region ¹.

Results: [¹⁸F]SynVesT-2 was synthesized with high molar activity (189±83 MBq/nmol, n=12). Parent fraction of [¹⁸F]SynVesT-2 in plasma was 28±8% (n=12) at 30 min post injection (p.i.), which is similar to [¹¹C]UCB-J and [¹⁸F]SynVesT-1 ^2,3,4. Plasma free fraction was high (f_p: 0.31±0.04, n=12), and also in line with those of [¹¹C]UCB-J and [¹⁸F]SynVesT-1 ^2,3. [¹⁸F]SynVesT-2 crossed the blood brain barrier quickly, with a peak standardized uptake value (SUV) of 8 in the thalamus and putamen at 7 min p.i. Regional TACs were fitted well with the 1TC model. And V_T values ranged from 2.0 mL/cm³ in the CS to 7.9 mL/cm³ in the putamen. The mean absolute test-retest variability (aTRV) for V_T was 5.5% across the 11 ROIs analyzed. Regional BP_ND values ranged from 3.00±0.32 in the putamen to 1.71±0.22 in the hippocampus, with a regional mean aTRV of 10.6%. The mean BP_ND values of [¹⁸F]SynVesT-2 are 78±5% of those of [¹¹C]UCB-J, 66±4% of [¹⁸F]SynVesT-1, and more than 2-fold higher than those of [¹⁸F]UCB-H ⁵.Shortening of the dynamic scan time from 2 h to 30 min had negligible impact (-0.3±4.3%) on BP_ND estimation. This is significantly shorter than the 60 min required for [¹¹C]UCB-J and [¹⁸F]SynVesT-1³. To eliminate arterial blood collection and simplify the scan protocol for multicenter clinical investigations, we evaluated the option of using SUVR-1 as a surrogate for BP_ND, by comparing the averaged SUVR-1 values from different imaging windows with the BP_NDvalues from the full dataset and found that the optimal static imaging window is 20-50 min p.i., with 1.2±1.6% difference between SUVR-1_{(20-50 min)} and BP_ND.Levetiracetam blocked 85% of the available SV2A binding sites, based on the Lassen plots. The estimated V_ND/f_p is 4.86 mL/cm³, indicating that about 20% of the tracer uptake in CS at baseline is attributable to specific uptake, which is lower than the 35-40% for [¹¹C]UCB-J ¹.

Conclusions: The SV2A PET tracer [¹⁸F]SynVesT-2 demonstrated imaging characteristics in human brain suitable for large scale clinical trial investigations, with high brain uptake, fast kinetics, high specific binding, and low nonspecific uptake. Additionally, the pharmacokinetic properties of [¹⁸F]SynVesT-2 requires only a short 30 min dynamic scan to derive physiological parameters, or the use of a short static scan (20-50 min p.i.), to reliably estimate the specific binding without arterial blood collection.