¹⁸F-PBR111 PET Imaging in Healthy Controls and Schizophrenia: Test–Retest Reproducibility and Quantification of Neuroinflammation

Abstract

Activated microglia express the translocator protein (TSPO) on the outer mitochondrial membrane. ¹⁸F-PBR111 is a second-generation PET ligand that specifically binds the TSPO, allowing in vivo visualization and quantification of neuroinflammation. The aim of this study was to evaluate whether the test-retest variability of ¹⁸F-PBR111 in healthy controls is acceptable to detect a psychosis-associated neuroinflammatory signal in schizophrenia. <b>Methods:</b> Dynamic 90-min ¹⁸F-PBR111 scans were obtained in 17 healthy male controls (HCs) and 11 male schizophrenia patients (SPs) during a psychotic episode. Prior genotyping for the rs6917 polymorphism distinguished high-affinity binders (HABs) and mixed-affinity binders (MABs). Total volume of distribution (V_T) was determined from 2-tissue-compartment modeling with vascular trapping and a metabolite-corrected plasma input function. A subgroup of HCs (<i>n</i> = 12; 4 HABs and 8 MABs) was scanned twice to assess absolute test-retest variability and intraclass correlation coefficients of the regional V_T values. Differences in TSPO binding between HC and SP were assessed using mixed model analysis adjusting for age, genotype, and age*cohort. The effect of using different scan durations (V_{T-60 min} versus V_{T-90 min}) was determined based on Pearson r. Data were mean ± SD. <b>Results:</b> Mean absolute variability in V_T ranged from 16% ± 14% (19% ± 20% HAB; 15% ± 11% MAB) in the cortical gray matter to 22% ± 15% (23% ± 15% HAB; 22% ± 16% MAB) in the hippocampus. Intraclass correlation coefficients were consistently between 0.64 and 0.82 for all tested regions. TSPO binding in SP compared with HC depended on age (cohort*age: <i>P</i> < 0.05) and was increased by +14% ± 4% over the regions. There was a significant effect of genotype on TSPO binding, and V_T of HABs was 31% ± 8% (HC: 17% ± 5%, SP: 61% ± 14%) higher than MABs. Across all clinical groups, V_{T-60 min} and V_{T-90 min} were strongly correlated (<i>r</i> > 0.7, <i>P</i> < 0.0001). <b>Conclusion:</b>¹⁸F-PBR111 can be used for monitoring of TSPO binding, as shown by medium test-retest variability and reliability of V_T in HCs. Microglial activation is present in SPs depending on age and needs to be adjusted for genotype.