Radiation Dosimetry of the 18 kDa Translocator Protein Ligand [18F]PBR111 in Humans

This study investigated the whole-body biodistribution and radiation dosimetry of the second-generation TSPO radioligand [18F]PBR111 in healthy volunteers. The translocator protein (TSPO, 18 kDa) is a mitochondrial marker overexpressed in activated microglia and astrocytes during neuroinflammatory processes such as Alzheimer's disease, multiple sclerosis, and other neurological or systemic conditions. [18F]PBR111, a fluorine-18 labeled imidazopyridine derivative, offers high affinity and specificity for TSPO but is sensitive to the rs6971 polymorphism, leading to different binding classes (high- and mixed-affinity binders). Establishing accurate human dosimetry was essential before extending its clinical use.

Six healthy volunteers (3 females, 3 males) free from chronic or acute inflammatory, infectious, or allergic conditions were recruited under local ethics approval (Geneva University Hospitals). All participants underwent genotyping for the rs6971 polymorphism to confirm the absence of low-affinity binders. Each subject received approximately 200 MBq of [18F]PBR111 intravenously, synthesized via an automated AllInOne module using [18F]fluoride produced on an IBA 18.5 MeV cyclotron. Radiochemical preparation included QMA cartridge trapping, elution with tetrabutylammonium hydroxide, azeotropic drying, nucleophilic substitution with the precursor, purification by semi-preparative HPLC, and formulation through dual C18 SepPak cartridges, ethanol/saline elution, and sterilizing filtration. The final product was obtained with high radiochemical purity and molar activity.

Dynamic whole-body PET/CT was performed on a Siemens Biograph™ Vision 600 Edge using continuous bed motion and 10 sequential passes of increasing duration over ~130 minutes. Low-dose CT provided attenuation correction. PET images were reconstructed with OP-OSEM (2 iterations, 21 subsets), incorporating TOF and resolution modeling, and expressed as SUV (body weight normalized). Organ segmentation was achieved using a deep-learning model applied to PET/CT data, covering major target organs (liver, lungs, spleen, kidneys, red bone marrow, GI tract, urinary bladder, heart wall, pancreas, adrenals, thyroid, etc.), with the remainder of the body included for dosimetric completeness.

Organ time-activity curves were derived from decay-corrected PET data, integrated, and analyzed using OLINDA/EXM v2.2 to calculate residence times and absorbed doses. Effective dose estimates followed ICRP-103 methodology. Statistical analyses included Mann-Whitney tests (sex effects) and Friedman tests with Dunn's multiple comparisons (organ contributions).