Purpose: Measurements of non-displaceable binding (V ND ) of positron emission tomography (PET) ligands are not often made in vivo in humans because they require ligands to displace binding to target receptors and there are few readily available, safe ones to use. A technique to measure V ND for ligands for the 18-kDa translocator protein (TSPO) has recently been developed which compares the total volume of distribution (V T ) before and after administration of the TSPO ligand XBD173. Here, we used XBD173 with an occupancy plot to quantify V ND for two TSPO radiotracers, [ 18 F]GE-180 and [ 11 C]PBR28, in cohorts of people with multiple sclerosis (MS). Additionally, we compared plots of subjects carrying high (HAB) or mixed binding (MAB) affinity polymorphisms of TSPO to estimate V ND without receptor blockade. Procedures: Twelve people with MS underwent baseline MRI and 90-min dynamic [ 18 F]GE-180 PET or [ 11 C]PBR28 PET (n = 6; three HAB, three MAB each). Arterial blood sampling was used to generate plasma input functions for the two-tissue compartment model. V ND was calculated using two independent methods: the occupancy plot (by modelling the differences in signal post XBD173) and the polymorphism plot (by modelling the differences in signal across presence and absence of rs6971 genotypes). Results: Whole brain V T (mean ± standard deviation) was 0.29 ± 0.17 ml/cm 3 for [ 18 F]GE-180 and 5.01 ± 1.88 ml/cm 3 for [ 11 C]PBR28. Using the occupancy and polymorphism plots respectively, V ND for [ 18 F]GE-180 was 0.11 ml/cm 3 (95 % CI = 0.02, 0.16) and 0.20 ml/cm 3 (0.16, 0.34), accounting for, on average, 55 % of V T in the whole brain. For [ 11 C]PBR28, these values were 3.81 ml/cm 3 (3.02, 4.21) and 3.49 ml/cm 3 (1.38, 4.27), accounting for 67 % of average whole brain V T .Sujata Sridharan and Joel Raffel are joint first authors.
Conclusions:Although V T for [ 18 F]GE-180 is low, indicating low brain penetration, half the signal shown by MS subjects reflected specific TSPO binding. V T for [ 11 C]PBR28 was higher and two thirds of the binding was non-specific. No brain ROIs were devoid of specific signal, further confirming that true reference tissue approaches are potentially problematic for estimating TSPO levels.