Quantification of TSPO overexpression in a rat model of local neuroinflammation induced by intracerebral injection of LPS by the use of [(18)F]DPA-714 PET.

Abstract

[(18)F]DPA-714 is a radiotracer with high affinity for TSPO. We have characterized the kinetics of [(18)F]DPA-714 in rat brain and evaluated its ability to quantify TSPO expression with PET using a neuroinflammation model induced by unilateral intracerebral injection of lipopolysaccharide (LPS). Dynamic small-animal PET scans with [(18)F]DPA-714 were performed in Wistar rats on a FOCUS-220 system for up to 3h. Both plasma and perfused brain homogenates were analysed using HPLC to quantify radiometabolites. Full kinetic modelling of [(18)F]DPA-714 brain uptake was performed using a metabolite-corrected arterial plasma input function. Binding potential (BPND) calculated as the distribution volume ratio minus one (DVR-1) between affected and healthy brain tissue was used as the outcome measure and evaluated against reference tissue models. The percentage of intact [(18)F]DPA-714 in arterial plasma samples was 92 ± 4% at 10min, 75 ± 8% at 40min and 52 ± 6% at 180min. The radiometabolite fraction in brain was negligible (<3% at 30min). Among the models investigated, the reversible two-tissue (2T) compartment model best described [(18)F]DPA-714 brain kinetics. BPND values obtained with a simplified and a multilinear reference tissue model (SRTM, MRTM) using the contralateral striatum as the reference region correlated well (Spearman's r = 0.96, p ≤ 0.003) with 2T BPND values calculated as DVR-1, and showed comparable bias (bias range 17.94%, 20.32%). Analysis of stability over time suggested that the acquisition time should be at least 90min for SRTM and MRTM. Quantification of [(18)F]DPA-714 binding to TSPO with full kinetic modelling is feasible using a 2T model. SRTM and MRTM can be suggested as reasonable substitutes with the contralateral striatum as the reference region and a scan duration of at least 90min. However, selection of the reference region depends on the disease model used.